d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
2b519b1707
freebsd-dev/sys/dev/ahci/ahci.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

2914 lines
82 KiB
C
Raw Normal View History

Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/*-
sys/dev: further adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 2-Clause license, however the tool I was using misidentified many licenses so this was mostly a manual - error prone - task. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts.
2017-11-27 14:52:40 +00:00
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
*
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
* Copyright (c) 2009-2012 Alexander Motin <mav@FreeBSD.org>
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer,
* without modification, immediately at the beginning of the file.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/module.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/bus.h>
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
#include <sys/conf.h>
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
#include <sys/endian.h>
#include <sys/malloc.h>
#include <sys/lock.h>
#include <sys/mutex.h>
newbus: Move from bus_child_{pnpinfo,location}_src to bus_child_{pnpinfo,location} with sbuf Now that the upper layers all go through a layer to tie into these information functions that translates an sbuf into char * and len. The current interface suffers issues of what to do in cases of truncation, etc. Instead, migrate all these functions to using struct sbuf and these issues go away. The caller is also in charge of any memory allocation and/or expansion that's needed during this process. Create a bus_generic_child_{pnpinfo,location} and make it default. It just returns success. This is for those busses that have no information for these items. Migrate the now-empty routines to using this as appropriate. Document these new interfaces with man pages, and oversight from before. Reviewed by: jhb, bcr Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D29937
2021-06-23 02:27:57 +00:00
#include <sys/sbuf.h>
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
#include <sys/sysctl.h>
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
#include <machine/stdarg.h>
#include <machine/resource.h>
#include <machine/bus.h>
#include <sys/rman.h>
#include "ahci.h"
#include <cam/cam.h>
#include <cam/cam_ccb.h>
#include <cam/cam_sim.h>
#include <cam/cam_xpt_sim.h>
#include <cam/cam_debug.h>
/* local prototypes */
static void ahci_intr(void *data);
static void ahci_intr_one(void *data);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
static void ahci_intr_one_edge(void *data);
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
static int ahci_ch_init(device_t dev);
static int ahci_ch_deinit(device_t dev);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static int ahci_ch_suspend(device_t dev);
static int ahci_ch_resume(device_t dev);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
static void ahci_ch_pm(void *arg);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
static void ahci_ch_intr(void *arg);
static void ahci_ch_intr_direct(void *arg);
static void ahci_ch_intr_main(struct ahci_channel *ch, uint32_t istatus);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
static void ahci_begin_transaction(struct ahci_channel *ch, union ccb *ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void ahci_dmasetprd(void *arg, bus_dma_segment_t *segs, int nsegs, int error);
static void ahci_execute_transaction(struct ahci_slot *slot);
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
static void ahci_timeout(void *arg);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void ahci_end_transaction(struct ahci_slot *slot, enum ahci_err_type et);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
static int ahci_setup_fis(struct ahci_channel *ch, struct ahci_cmd_tab *ctp, union ccb *ccb, int tag);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void ahci_dmainit(device_t dev);
static void ahci_dmasetupc_cb(void *xsc, bus_dma_segment_t *segs, int nsegs, int error);
static void ahci_dmafini(device_t dev);
static void ahci_slotsalloc(device_t dev);
static void ahci_slotsfree(device_t dev);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
static void ahci_reset(struct ahci_channel *ch);
static void ahci_start(struct ahci_channel *ch, int fbs);
static void ahci_stop(struct ahci_channel *ch);
static void ahci_clo(struct ahci_channel *ch);
static void ahci_start_fr(struct ahci_channel *ch);
static void ahci_stop_fr(struct ahci_channel *ch);
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
static int ahci_phy_check_events(struct ahci_channel *ch, u_int32_t serr);
static uint32_t ahci_ch_detval(struct ahci_channel *ch, uint32_t val);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static int ahci_sata_connect(struct ahci_channel *ch);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
static int ahci_sata_phy_reset(struct ahci_channel *ch);
static int ahci_wait_ready(struct ahci_channel *ch, int t, int t0);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
static void ahci_issue_recovery(struct ahci_channel *ch);
static void ahci_process_read_log(struct ahci_channel *ch, union ccb *ccb);
static void ahci_process_request_sense(struct ahci_channel *ch, union ccb *ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void ahciaction(struct cam_sim *sim, union ccb *ccb);
static void ahcipoll(struct cam_sim *sim);
Mark MALLOC_DEFINEs static that have no corresponding MALLOC_DECLAREs. This means that their use is restricted to a single C file.
2011-11-07 06:44:47 +00:00
static MALLOC_DEFINE(M_AHCI, "AHCI driver", "AHCI driver data buffers");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
#define recovery_type spriv_field0
#define RECOVERY_NONE 0
#define RECOVERY_READ_LOG 1
#define RECOVERY_REQUEST_SENSE 2
#define recovery_slot spriv_field1
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
static uint32_t
ahci_ch_detval(struct ahci_channel *ch, uint32_t val)
{
return ch->disablephy ? ATA_SC_DET_DISABLE : val;
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
ahci_ctlr_setup(device_t dev)
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
{
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
struct ahci_controller *ctlr = device_get_softc(dev);
/* Clear interrupts */
ATA_OUTL(ctlr->r_mem, AHCI_IS, ATA_INL(ctlr->r_mem, AHCI_IS));
/* Configure CCC */
if (ctlr->ccc) {
ATA_OUTL(ctlr->r_mem, AHCI_CCCP, ATA_INL(ctlr->r_mem, AHCI_PI));
ATA_OUTL(ctlr->r_mem, AHCI_CCCC,
(ctlr->ccc << AHCI_CCCC_TV_SHIFT) |
(4 << AHCI_CCCC_CC_SHIFT) |
AHCI_CCCC_EN);
ctlr->cccv = (ATA_INL(ctlr->r_mem, AHCI_CCCC) &
AHCI_CCCC_INT_MASK) >> AHCI_CCCC_INT_SHIFT;
if (bootverbose) {
device_printf(dev,
"CCC with %dms/4cmd enabled on vector %d\n",
ctlr->ccc, ctlr->cccv);
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
}
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
/* Enable AHCI interrupts */
ATA_OUTL(ctlr->r_mem, AHCI_GHC,
ATA_INL(ctlr->r_mem, AHCI_GHC) | AHCI_GHC_IE);
return (0);
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
ahci_ctlr_reset(device_t dev)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
struct ahci_controller *ctlr = device_get_softc(dev);
Add support for AHCI BIOS/OS Handoff. This allows clean handoff from BIOS implementing some asynchronous I/O to the OS AHCI driver. During attach driver declares OS ownership request and waits from 25ms to 2s for BIOS to complete operation and release the hardware. MFC after: 2 weeks
2020-04-01 03:27:47 +00:00
uint32_t v;
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int timeout;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Add support for AHCI BIOS/OS Handoff. This allows clean handoff from BIOS implementing some asynchronous I/O to the OS AHCI driver. During attach driver declares OS ownership request and waits from 25ms to 2s for BIOS to complete operation and release the hardware. MFC after: 2 weeks
2020-04-01 03:27:47 +00:00
/* BIOS/OS Handoff */
if ((ATA_INL(ctlr->r_mem, AHCI_VS) >= 0x00010200) &&
(ATA_INL(ctlr->r_mem, AHCI_CAP2) & AHCI_CAP2_BOH) &&
((v = ATA_INL(ctlr->r_mem, AHCI_BOHC)) & AHCI_BOHC_OOS) == 0) {
/* Request OS ownership. */
ATA_OUTL(ctlr->r_mem, AHCI_BOHC, v | AHCI_BOHC_OOS);
/* Wait up to 2s for BIOS ownership release. */
for (timeout = 0; timeout < 80; timeout++) {
DELAY(25000);
v = ATA_INL(ctlr->r_mem, AHCI_BOHC);
if ((v & AHCI_BOHC_BOS) == 0)
break;
if ((v & AHCI_BOHC_BB) == 0)
break;
}
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
/* Enable AHCI mode */
ATA_OUTL(ctlr->r_mem, AHCI_GHC, AHCI_GHC_AE);
/* Reset AHCI controller */
ATA_OUTL(ctlr->r_mem, AHCI_GHC, AHCI_GHC_AE|AHCI_GHC_HR);
for (timeout = 1000; timeout > 0; timeout--) {
DELAY(1000);
if ((ATA_INL(ctlr->r_mem, AHCI_GHC) & AHCI_GHC_HR) == 0)
break;
MFp4: Add set of chip IDs, known to support AHCI.
2009-11-11 10:44:09 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
if (timeout == 0) {
device_printf(dev, "AHCI controller reset failure\n");
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (ENXIO);
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
}
/* Reenable AHCI mode */
ATA_OUTL(ctlr->r_mem, AHCI_GHC, AHCI_GHC_AE);
Introduce Annapurna Labs AHCI support Overview: * implemented quirk for forcing SATA interface enable * restore value to status register - this enables link autonegotiation Modifications: * devid:vendorid field * quirk for forcing PI setting (BIOS is doing that on PC-like systems) * write to capabilites field to enable phy link initialization Submitted by: Wojciech Macek <wma@semihalf.com> Reviewed by: imp, mav Obtained from: Semihalf
2015-03-17 18:59:47 +00:00
if (ctlr->quirks & AHCI_Q_RESTORE_CAP) {
/*
* Restore capability field.
* This is write to a read-only register to restore its state.
* On fully standard-compliant hardware this is not needed and
* this operation shall not take place. See ahci_pci.c for
* platforms using this quirk.
*/
ATA_OUTL(ctlr->r_mem, AHCI_CAP, ctlr->caps);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
return (0);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_attach(device_t dev)
{
struct ahci_controller *ctlr = device_get_softc(dev);
Fix ahci(4) driver attach to controller with 32 ports. Incorrect sign expansion in variables that supposed to be a bit fields caused infinite loop. Fixing this allows system properly detect maximal possible 32 devices configured on AHCI HBA of BHyVe. That case did not happen in a wild before due to lack of hardware AHCI HBAs with 32 ports. Approved by: re (gjb@) MFC after: 1 week
2016-07-07 22:10:10 +00:00
int error, i, speed, unit;
uint32_t u, version;
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
device_t child;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->dev = dev;
Initialize variables before resource_int_value(). Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com>
2014-09-12 08:37:21 +00:00
ctlr->ccc = 0;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
resource_int_value(device_get_name(dev),
device_get_unit(dev), "ccc", &ctlr->ccc);
Improve AHCI Enclosure Management and SES interoperation. Since SES specs do not define mechanism to map enclosure slots to SATA disks, AHCI EM code I written many years ago appeared quite useless, that always bugged me. I was thinking whether it was a good idea, but if LSI HBAs do that, why I shouldn't? This change introduces simple non-standard mechanism for the mapping into both AHCI EM and SES code, that makes AHCI EM on capable controllers (most of Intel's) a first-class SES citizen, allowing it to report disk physical path to GEOM, show devices inserted into each enclosure slot in `sesutil map` and `getencstat`, control locate and fault LEDs for specific devices with `sesutil locate adaX on` and `sesutil fault adaX on`, etc. I've successfully tested this on Supermicro X10DRH-i motherboard connected with sideband cable of its S-SATA Mini-SAS connector to SAS815TQ backplane. It can indicate with LEDs Locate, Fault and Rebuild/Remap SES statuses for each disk identical to real SES of Supermicro SAS2 backplanes. MFC after: 2 weeks
2019-06-23 19:05:01 +00:00
mtx_init(&ctlr->ch_mtx, "AHCI channels lock", NULL, MTX_DEF);
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Setup our own memory management for channels. */
Fill rman range start/end values. It makes devinfo output more readable.
2010-05-22 08:30:47 +00:00
ctlr->sc_iomem.rm_start = rman_get_start(ctlr->r_mem);
ctlr->sc_iomem.rm_end = rman_get_end(ctlr->r_mem);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->sc_iomem.rm_type = RMAN_ARRAY;
ctlr->sc_iomem.rm_descr = "I/O memory addresses";
if ((error = rman_init(&ctlr->sc_iomem)) != 0) {
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
ahci_free_mem(dev);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (error);
}
if ((error = rman_manage_region(&ctlr->sc_iomem,
rman_get_start(ctlr->r_mem), rman_get_end(ctlr->r_mem))) != 0) {
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
ahci_free_mem(dev);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
rman_fini(&ctlr->sc_iomem);
return (error);
}
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
/* Get the HW capabilities */
version = ATA_INL(ctlr->r_mem, AHCI_VS);
ctlr->caps = ATA_INL(ctlr->r_mem, AHCI_CAP);
Fix AHCI 1.2 version checks. This should be mostly cosmetic. Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com> MFC after: 1 week
2012-09-12 09:20:37 +00:00
if (version >= 0x00010200)
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
ctlr->caps2 = ATA_INL(ctlr->r_mem, AHCI_CAP2);
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
if (ctlr->caps & AHCI_CAP_EMS)
ctlr->capsem = ATA_INL(ctlr->r_mem, AHCI_EM_CTL);
Introduce Annapurna Labs AHCI support Overview: * implemented quirk for forcing SATA interface enable * restore value to status register - this enables link autonegotiation Modifications: * devid:vendorid field * quirk for forcing PI setting (BIOS is doing that on PC-like systems) * write to capabilites field to enable phy link initialization Submitted by: Wojciech Macek <wma@semihalf.com> Reviewed by: imp, mav Obtained from: Semihalf
2015-03-17 18:59:47 +00:00
if (ctlr->quirks & AHCI_Q_FORCE_PI) {
/*
* Enable ports.
* The spec says that BIOS sets up bits corresponding to
* available ports. On platforms where this information
* is missing, the driver can define available ports on its own.
*/
int nports = (ctlr->caps & AHCI_CAP_NPMASK) + 1;
int nmask = (1 << nports) - 1;
ATA_OUTL(ctlr->r_mem, AHCI_PI, nmask);
device_printf(dev, "Forcing PI to %d ports (mask = %x)\n",
nports, nmask);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->ichannels = ATA_INL(ctlr->r_mem, AHCI_PI);
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
/* Identify and set separate quirks for HBA and RAID f/w Marvells. */
Invert AHCI_Q_NOBSYRES quirk meaning, waiting for readiness by default. I gave up to update list of Marvell chips that require this quirk. The final nail was growing number of PCIe/M.2 SSDs where Marvell chips have PCI IDs of different vendors. MFC after: 1 week H/W donated by: I/O Switch
2014-09-05 13:45:14 +00:00
if ((ctlr->quirks & AHCI_Q_ALTSIG) &&
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
(ctlr->caps & AHCI_CAP_SPM) == 0)
Invert AHCI_Q_NOBSYRES quirk meaning, waiting for readiness by default. I gave up to update list of Marvell chips that require this quirk. The final nail was growing number of PCIe/M.2 SSDs where Marvell chips have PCI IDs of different vendors. MFC after: 1 week H/W donated by: I/O Switch
2014-09-05 13:45:14 +00:00
ctlr->quirks |= AHCI_Q_NOBSYRES;
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
if (ctlr->quirks & AHCI_Q_1CH) {
ctlr->caps &= ~AHCI_CAP_NPMASK;
ctlr->ichannels &= 0x01;
}
if (ctlr->quirks & AHCI_Q_2CH) {
ctlr->caps &= ~AHCI_CAP_NPMASK;
ctlr->caps |= 1;
ctlr->ichannels &= 0x03;
}
if (ctlr->quirks & AHCI_Q_4CH) {
ctlr->caps &= ~AHCI_CAP_NPMASK;
ctlr->caps |= 3;
ctlr->ichannels &= 0x0f;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->channels = MAX(flsl(ctlr->ichannels),
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
(ctlr->caps & AHCI_CAP_NPMASK) + 1);
if (ctlr->quirks & AHCI_Q_NOPMP)
ctlr->caps &= ~AHCI_CAP_SPM;
if (ctlr->quirks & AHCI_Q_NONCQ)
ctlr->caps &= ~AHCI_CAP_SNCQ;
- Spec tells that CCC interrupt is edge triggered. Acknowledge it as such. - Do not try to enable CCC if it is not supported.
2010-03-21 18:18:58 +00:00
if ((ctlr->caps & AHCI_CAP_CCCS) == 0)
ctlr->ccc = 0;
Add support for "LED" enclosure management messages, defined by the AHCI. When supported by hardware, this allows to control per-port activity, locate and fault LEDs via the led(4) API for localization and status reporting purposes. Supporting AHCI controllers may transmit that information to the backplane controllers via SGPIO interface. Backplane controllers interpret received statuses in some way (IBPI standard) to report them using present indicators.
2011-05-17 22:07:45 +00:00
ctlr->emloc = ATA_INL(ctlr->r_mem, AHCI_EM_LOC);
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
/* Create controller-wide DMA tag. */
A dma tag alignment should be 1 (not 0) when a device can dma at any address.
2014-11-06 18:15:16 +00:00
if (bus_dma_tag_create(bus_get_dma_tag(dev), 1, 0,
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
(ctlr->caps & AHCI_CAP_64BIT) ? BUS_SPACE_MAXADDR :
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL, NULL,
BUS_SPACE_MAXSIZE, BUS_SPACE_UNRESTRICTED, BUS_SPACE_MAXSIZE,
Check if the device is marked as dma-coherent in the FDT, and if so, let busdma know, so that on architectures where dma isn't always coherent, we know we don't have to write-back/invalidates cachelines on DMA operations. Reviewed by: andrew, mav
2017-04-26 16:13:22 +00:00
ctlr->dma_coherent ? BUS_DMA_COHERENT : 0, NULL, NULL,
&ctlr->dma_tag)) {
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
ahci_free_mem(dev);
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
rman_fini(&ctlr->sc_iomem);
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (ENXIO);
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
}
- Spec tells that CCC interrupt is edge triggered. Acknowledge it as such. - Do not try to enable CCC if it is not supported.
2010-03-21 18:18:58 +00:00
ahci_ctlr_setup(dev);
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
Restore order of interrupt setup. Minor problems can result by setting up the interrupts too early: Reviewed by: mav@ Sponsored by: Netflix
2014-09-06 18:20:50 +00:00
/* Setup interrupts. */
Return the error from ahci_setup_interrupt in ahci_attach Previously ahci_attach returned a hard coded ENXIO instead of the value from ahci_setup_interrupt. This is effectively a NOOP change as currently ahci_setup_interrupt only ever returns 0 or ENXIO, so just there to protect against any future changes to that. Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:32:57 +00:00
if ((error = ahci_setup_interrupt(dev)) != 0) {
Restore order of interrupt setup. Minor problems can result by setting up the interrupts too early: Reviewed by: mav@ Sponsored by: Netflix
2014-09-06 18:20:50 +00:00
bus_dma_tag_destroy(ctlr->dma_tag);
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
ahci_free_mem(dev);
Restore order of interrupt setup. Minor problems can result by setting up the interrupts too early: Reviewed by: mav@ Sponsored by: Netflix
2014-09-06 18:20:50 +00:00
rman_fini(&ctlr->sc_iomem);
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (error);
Restore order of interrupt setup. Minor problems can result by setting up the interrupts too early: Reviewed by: mav@ Sponsored by: Netflix
2014-09-06 18:20:50 +00:00
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
i = 0;
for (u = ctlr->ichannels; u != 0; u >>= 1)
i += (u & 1);
ctlr->direct = (ctlr->msi && (ctlr->numirqs > 1 || i <= 3));
resource_int_value(device_get_name(dev), device_get_unit(dev),
"direct", &ctlr->direct);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Announce HW capabilities. */
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
speed = (ctlr->caps & AHCI_CAP_ISS) >> AHCI_CAP_ISS_SHIFT;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(dev,
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
"AHCI v%x.%02x with %d %sGbps ports, Port Multiplier %s%s\n",
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
((version >> 20) & 0xf0) + ((version >> 16) & 0x0f),
((version >> 4) & 0xf0) + (version & 0x0f),
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->caps & AHCI_CAP_NPMASK) + 1,
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
((speed == 1) ? "1.5":((speed == 2) ? "3":
((speed == 3) ? "6":"?"))),
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->caps & AHCI_CAP_SPM) ?
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
"supported" : "not supported",
(ctlr->caps & AHCI_CAP_FBSS) ?
" with FBS" : "");
Added output of device QUIRKS for CAM and AHCI devices during boot. Reviewed by: mav Approved by: pjd (mentor) MFC after: 2 weeks
2013-05-18 23:36:21 +00:00
if (ctlr->quirks != 0) {
device_printf(dev, "quirks=0x%b\n", ctlr->quirks,
AHCI_Q_BIT_STRING);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (bootverbose) {
device_printf(dev, "Caps:%s%s%s%s%s%s%s%s %sGbps",
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->caps & AHCI_CAP_64BIT) ? " 64bit":"",
(ctlr->caps & AHCI_CAP_SNCQ) ? " NCQ":"",
(ctlr->caps & AHCI_CAP_SSNTF) ? " SNTF":"",
(ctlr->caps & AHCI_CAP_SMPS) ? " MPS":"",
(ctlr->caps & AHCI_CAP_SSS) ? " SS":"",
(ctlr->caps & AHCI_CAP_SALP) ? " ALP":"",
(ctlr->caps & AHCI_CAP_SAL) ? " AL":"",
(ctlr->caps & AHCI_CAP_SCLO) ? " CLO":"",
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
((speed == 1) ? "1.5":((speed == 2) ? "3":
((speed == 3) ? "6":"?"))));
printf("%s%s%s%s%s%s %dcmd%s%s%s %dports\n",
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->caps & AHCI_CAP_SAM) ? " AM":"",
(ctlr->caps & AHCI_CAP_SPM) ? " PM":"",
(ctlr->caps & AHCI_CAP_FBSS) ? " FBS":"",
(ctlr->caps & AHCI_CAP_PMD) ? " PMD":"",
(ctlr->caps & AHCI_CAP_SSC) ? " SSC":"",
(ctlr->caps & AHCI_CAP_PSC) ? " PSC":"",
((ctlr->caps & AHCI_CAP_NCS) >> AHCI_CAP_NCS_SHIFT) + 1,
(ctlr->caps & AHCI_CAP_CCCS) ? " CCC":"",
(ctlr->caps & AHCI_CAP_EMS) ? " EM":"",
(ctlr->caps & AHCI_CAP_SXS) ? " eSATA":"",
(ctlr->caps & AHCI_CAP_NPMASK) + 1);
}
Fix AHCI 1.2 version checks. This should be mostly cosmetic. Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com> MFC after: 1 week
2012-09-12 09:20:37 +00:00
if (bootverbose && version >= 0x00010200) {
Decode some bits defined in AHCI 1.3.1 Device Sleep Technical Proposal. For now this is only conmetics to report HBA capabilities (Haswell?). Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com>
2013-07-25 10:29:40 +00:00
device_printf(dev, "Caps2:%s%s%s%s%s%s\n",
(ctlr->caps2 & AHCI_CAP2_DESO) ? " DESO":"",
(ctlr->caps2 & AHCI_CAP2_SADM) ? " SADM":"",
(ctlr->caps2 & AHCI_CAP2_SDS) ? " SDS":"",
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->caps2 & AHCI_CAP2_APST) ? " APST":"",
(ctlr->caps2 & AHCI_CAP2_NVMP) ? " NVMP":"",
(ctlr->caps2 & AHCI_CAP2_BOH) ? " BOH":"");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/* Attach all channels on this controller */
for (unit = 0; unit < ctlr->channels; unit++) {
child = device_add_child(dev, "ahcich", -1);
Change the way how "not implemented" AHCI channels handled. Instead of completely skipping them, create ahcich devices for them to allocate unit numbers, but mark them as disabled to prevent driver probe and attach. Last time some BIOSes tend to report unused channels as "not implemented". This change makes ahcichX devices numbering consistent, independently of connected disks. It makes per-channel driver hints usable and CAM devices wiring possible on such systems.
2011-11-17 20:46:51 +00:00
if (child == NULL) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(dev, "failed to add channel device\n");
Change the way how "not implemented" AHCI channels handled. Instead of completely skipping them, create ahcich devices for them to allocate unit numbers, but mark them as disabled to prevent driver probe and attach. Last time some BIOSes tend to report unused channels as "not implemented". This change makes ahcichX devices numbering consistent, independently of connected disks. It makes per-channel driver hints usable and CAM devices wiring possible on such systems.
2011-11-17 20:46:51 +00:00
continue;
}
device_set_ivars(child, (void *)(intptr_t)unit);
if ((ctlr->ichannels & (1 << unit)) == 0)
device_disable(child);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Create a AHCI attachment for nvme. Intel has created RST and many laptops from vendors like Lenovo and Asus. It's a mechanism for creating multiple boot devices under windows. It effectively hides the nvme drive inside of the ahci controller. The details are supposed to be a trade secret. However, there's a reverse engineered Linux driver, and this implements similar operations to allow nvme drives to attach. The ahci driver attaches nvme children that proxy the remapped resources to the child. nvme_ahci is just like nvme_pci, except it doesn't do the PCI specific things. That's moved into ahci where appropriate. When the nvme drive is remapped, MSI-x interrupts aren't forwarded (the linux driver doesn't know how to use this either). INTx interrupts are used instead. This is suboptimal, but usually sufficient for the laptops these parts are in. This is based loosely on https://www.spinics.net/lists/linux-ide/msg53364.html submitted, but not accepted by, Linux. It was written by Dan Williams. These changes were written from scratch by Olivier Houchard. Submitted by: cognet@ (Olivier Houchard)
2019-08-21 22:18:01 +00:00
/* Attach any remapped NVME device */
for (; unit < ctlr->channels + ctlr->remapped_devices; unit++) {
child = device_add_child(dev, "nvme", -1);
if (child == NULL) {
device_printf(dev, "failed to add remapped NVMe device");
continue;
}
device_set_ivars(child, (void *)(intptr_t)(unit | AHCI_REMAPPED_UNIT));
}
ahci(4): Allow enclosure emulation without hardware. After 53f5ac1310e allowed SATA device mapping to enclosure slots, it may have sense to provide enclosure device emulation even without real hardware interface like SGPIO just for purposes of physical device location tracking (still assuming straight cabling). MFC after: 1 week Sponsored by: iXsystems, Inc.
2021-12-17 20:24:56 +00:00
int em = (ctlr->caps & AHCI_CAP_EMS) != 0;
resource_int_value(device_get_name(dev), device_get_unit(dev),
"em", &em);
if (em) {
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
child = device_add_child(dev, "ahciem", -1);
if (child == NULL)
device_printf(dev, "failed to add enclosure device\n");
else
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
device_set_ivars(child, (void *)(intptr_t)AHCI_EM_UNIT);
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_generic_attach(dev);
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (0);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_detach(device_t dev)
{
struct ahci_controller *ctlr = device_get_softc(dev);
Move the device_delete_all_children() function from usb_util.c to kern/subr_bus.c. Simplify this function so that it no longer depends on malloc() to execute. Identify a few other places where it makes sense to use device_delete_all_children(). MFC after: 1 week
2011-11-19 10:11:50 +00:00
int i;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Detach & delete all children */
Rename device_delete_all_children() into device_delete_children(). Suggested by: jhb @ and marius @ MFC after: 1 week
2011-11-22 21:56:55 +00:00
device_delete_children(dev);
Move the device_delete_all_children() function from usb_util.c to kern/subr_bus.c. Simplify this function so that it no longer depends on malloc() to execute. Identify a few other places where it makes sense to use device_delete_all_children(). MFC after: 1 week
2011-11-19 10:11:50 +00:00
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Free interrupts. */
for (i = 0; i < ctlr->numirqs; i++) {
if (ctlr->irqs[i].r_irq) {
bus_teardown_intr(dev, ctlr->irqs[i].r_irq,
ctlr->irqs[i].handle);
bus_release_resource(dev, SYS_RES_IRQ,
ctlr->irqs[i].r_irq_rid, ctlr->irqs[i].r_irq);
}
}
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
bus_dma_tag_destroy(ctlr->dma_tag);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Free memory. */
rman_fini(&ctlr->sc_iomem);
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
ahci_free_mem(dev);
Improve AHCI Enclosure Management and SES interoperation. Since SES specs do not define mechanism to map enclosure slots to SATA disks, AHCI EM code I written many years ago appeared quite useless, that always bugged me. I was thinking whether it was a good idea, but if LSI HBAs do that, why I shouldn't? This change introduces simple non-standard mechanism for the mapping into both AHCI EM and SES code, that makes AHCI EM on capable controllers (most of Intel's) a first-class SES citizen, allowing it to report disk physical path to GEOM, show devices inserted into each enclosure slot in `sesutil map` and `getencstat`, control locate and fault LEDs for specific devices with `sesutil locate adaX on` and `sesutil fault adaX on`, etc. I've successfully tested this on Supermicro X10DRH-i motherboard connected with sideband cable of its S-SATA Mini-SAS connector to SAS815TQ backplane. It can indicate with LEDs Locate, Fault and Rebuild/Remap SES statuses for each disk identical to real SES of Supermicro SAS2 backplanes. MFC after: 2 weeks
2019-06-23 19:05:01 +00:00
mtx_destroy(&ctlr->ch_mtx);
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
return (0);
}
void
ahci_free_mem(device_t dev)
{
struct ahci_controller *ctlr = device_get_softc(dev);
/* Release memory resources */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ctlr->r_mem)
bus_release_resource(dev, SYS_RES_MEMORY, ctlr->r_rid, ctlr->r_mem);
Introduce support for MSI-X interrupts in AHCI - Allocate resources for MSI-X table and PBA if necessary - Add function ahci_free_mem() to free all resources Reviewed by: jhb, mav Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3009
2015-07-22 09:46:22 +00:00
if (ctlr->r_msix_table)
bus_release_resource(dev, SYS_RES_MEMORY,
ctlr->r_msix_tab_rid, ctlr->r_msix_table);
if (ctlr->r_msix_pba)
bus_release_resource(dev, SYS_RES_MEMORY,
ctlr->r_msix_pba_rid, ctlr->r_msix_pba);
ctlr->r_msix_pba = ctlr->r_mem = ctlr->r_msix_table = NULL;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_setup_interrupt(device_t dev)
{
struct ahci_controller *ctlr = device_get_softc(dev);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
int i;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Check for single MSI vector fallback. */
if (ctlr->numirqs > 1 &&
(ATA_INL(ctlr->r_mem, AHCI_GHC) & AHCI_GHC_MRSM) != 0) {
device_printf(dev, "Falling back to one MSI\n");
ctlr->numirqs = 1;
}
Clamp ahci max irq's to AHCI_MAX_IRQS This prevents the possiblity of any overruns on the statically allocated struct irqs field. Differential Revision: D838 MFC after: 2 weeks X-MFC-With: r276012 Sponsored by: Multiplay
2014-12-21 16:15:29 +00:00
/* Ensure we don't overrun irqs. */
if (ctlr->numirqs > AHCI_MAX_IRQS) {
device_printf(dev, "Too many irqs %d > %d (clamping)\n",
ctlr->numirqs, AHCI_MAX_IRQS);
ctlr->numirqs = AHCI_MAX_IRQS;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Allocate all IRQs. */
for (i = 0; i < ctlr->numirqs; i++) {
ctlr->irqs[i].ctlr = ctlr;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ctlr->irqs[i].r_irq_rid = i + (ctlr->msi ? 1 : 0);
Use r271207 optimization only for MSI-enabled HBAs. It was found that VirtualBox' AHCI does not allow nterrupt to be cleared before the interrupt status register is read, causing interrupt storm. AHCI specification allows to skip this register use when multi-vector MSI is enabled and so interrupting port is known. For single-vector MSI that is not stated explicitly, but if the port is only one, it is obviously known too.
2014-10-06 10:58:54 +00:00
if (ctlr->channels == 1 && !ctlr->ccc && ctlr->msi)
Save one register read (AHCI_IS) for AHCI controllers with only one port. For controllers with only one port (like PCIe or M.2 SSDs) interrupt can come from only one source, and skipping read saves few percents of CPU time. MFC after: 1 month H/W donated by: I/O Switch
2014-09-06 19:43:48 +00:00
ctlr->irqs[i].mode = AHCI_IRQ_MODE_ONE;
else if (ctlr->numirqs == 1 || i >= ctlr->channels ||
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
(ctlr->ccc && i == ctlr->cccv))
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->irqs[i].mode = AHCI_IRQ_MODE_ALL;
Do not consider the last interrupt shared if there are enough interrupts for all channels.
2016-07-17 12:31:31 +00:00
else if (ctlr->channels > ctlr->numirqs &&
i == ctlr->numirqs - 1)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->irqs[i].mode = AHCI_IRQ_MODE_AFTER;
else
ctlr->irqs[i].mode = AHCI_IRQ_MODE_ONE;
if (!(ctlr->irqs[i].r_irq = bus_alloc_resource_any(dev, SYS_RES_IRQ,
&ctlr->irqs[i].r_irq_rid, RF_SHAREABLE | RF_ACTIVE))) {
device_printf(dev, "unable to map interrupt\n");
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (ENXIO);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
if ((bus_setup_intr(dev, ctlr->irqs[i].r_irq, ATA_INTR_FLAGS, NULL,
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
(ctlr->irqs[i].mode != AHCI_IRQ_MODE_ONE) ? ahci_intr :
((ctlr->quirks & AHCI_Q_EDGEIS) ? ahci_intr_one_edge :
ahci_intr_one),
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
&ctlr->irqs[i], &ctlr->irqs[i].handle))) {
/* SOS XXX release r_irq */
device_printf(dev, "unable to setup interrupt\n");
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (ENXIO);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
While AHCI specification tells that multi-vector MSI doesn't use global IS register, nVidia chipsets have different oppinion, requiring every interrupt to be acknowledged there. While there, add interrupt descriptions in multi-vector MSI mode.
2010-01-10 16:05:05 +00:00
if (ctlr->numirqs > 1) {
bus_describe_intr(dev, ctlr->irqs[i].r_irq,
ctlr->irqs[i].handle,
ctlr->irqs[i].mode == AHCI_IRQ_MODE_ONE ?
"ch%d" : "%d", i);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
return (0);
}
/*
* Common case interrupt handler.
*/
static void
ahci_intr(void *data)
{
struct ahci_controller_irq *irq = data;
struct ahci_controller *ctlr = irq->ctlr;
- Spec tells that CCC interrupt is edge triggered. Acknowledge it as such. - Do not try to enable CCC if it is not supported.
2010-03-21 18:18:58 +00:00
u_int32_t is, ise = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
void *arg;
int unit;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (irq->mode == AHCI_IRQ_MODE_ALL) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
unit = 0;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (ctlr->ccc)
is = ctlr->ichannels;
else
is = ATA_INL(ctlr->r_mem, AHCI_IS);
} else { /* AHCI_IRQ_MODE_AFTER */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
unit = irq->r_irq_rid - 1;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
is = ATA_INL(ctlr->r_mem, AHCI_IS);
In AHCI_IRQ_MODE_AFTER mode do not clear interrupts below. This is probably a NOP change since IS register is not activery used for interrupts below the shared, but it looked odd to clear interrupts we did not handle.
2016-07-17 12:45:58 +00:00
is &= (0xffffffff << unit);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
}
- Spec tells that CCC interrupt is edge triggered. Acknowledge it as such. - Do not try to enable CCC if it is not supported.
2010-03-21 18:18:58 +00:00
/* CCC interrupt is edge triggered. */
if (ctlr->ccc)
ise = 1 << ctlr->cccv;
Clear all ports interrupt status bits in single write. Clearing one by one causes additional MSIs messages sent if several ports asked for attention same time. Time window before clearing is not important, as these interrupts are level triggered by interrupt source.
2009-12-21 21:27:56 +00:00
/* Some controllers have edge triggered IS. */
if (ctlr->quirks & AHCI_Q_EDGEIS)
- Spec tells that CCC interrupt is edge triggered. Acknowledge it as such. - Do not try to enable CCC if it is not supported.
2010-03-21 18:18:58 +00:00
ise |= is;
if (ise != 0)
ATA_OUTL(ctlr->r_mem, AHCI_IS, ise);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
for (; unit < ctlr->channels; unit++) {
if ((is & (1 << unit)) != 0 &&
(arg = ctlr->interrupt[unit].argument)) {
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
ctlr->interrupt[unit].function(arg);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
}
Create a AHCI attachment for nvme. Intel has created RST and many laptops from vendors like Lenovo and Asus. It's a mechanism for creating multiple boot devices under windows. It effectively hides the nvme drive inside of the ahci controller. The details are supposed to be a trade secret. However, there's a reverse engineered Linux driver, and this implements similar operations to allow nvme drives to attach. The ahci driver attaches nvme children that proxy the remapped resources to the child. nvme_ahci is just like nvme_pci, except it doesn't do the PCI specific things. That's moved into ahci where appropriate. When the nvme drive is remapped, MSI-x interrupts aren't forwarded (the linux driver doesn't know how to use this either). INTx interrupts are used instead. This is suboptimal, but usually sufficient for the laptops these parts are in. This is based loosely on https://www.spinics.net/lists/linux-ide/msg53364.html submitted, but not accepted by, Linux. It was written by Dan Williams. These changes were written from scratch by Olivier Houchard. Submitted by: cognet@ (Olivier Houchard)
2019-08-21 22:18:01 +00:00
for (; unit < ctlr->channels + ctlr->remapped_devices; unit++) {
if ((arg = ctlr->interrupt[unit].argument)) {
ctlr->interrupt[unit].function(arg);
}
}
Clear all ports interrupt status bits in single write. Clearing one by one causes additional MSIs messages sent if several ports asked for attention same time. Time window before clearing is not important, as these interrupts are level triggered by interrupt source.
2009-12-21 21:27:56 +00:00
/* AHCI declares level triggered IS. */
if (!(ctlr->quirks & AHCI_Q_EDGEIS))
ATA_OUTL(ctlr->r_mem, AHCI_IS, is);
AHCI: Fix AHCI driver for ARM. On ARM, we must ensure proper interdevice write ordering. The AHCI interrupt status register must be updated in HW before registers in interrupt controller. Unfortunately, only way how we can do it is readback. Discussed with: mav Approved by: kib (mentor) Differential Revision: https://reviews.freebsd.org/D4240
2015-11-29 11:28:04 +00:00
ATA_RBL(ctlr->r_mem, AHCI_IS);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/*
* Simplified interrupt handler for multivector MSI mode.
*/
static void
ahci_intr_one(void *data)
{
struct ahci_controller_irq *irq = data;
struct ahci_controller *ctlr = irq->ctlr;
void *arg;
int unit;
unit = irq->r_irq_rid - 1;
if ((arg = ctlr->interrupt[unit].argument))
ctlr->interrupt[unit].function(arg);
While AHCI specification tells that multi-vector MSI doesn't use global IS register, nVidia chipsets have different oppinion, requiring every interrupt to be acknowledged there. While there, add interrupt descriptions in multi-vector MSI mode.
2010-01-10 16:05:05 +00:00
/* AHCI declares level triggered IS. */
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ATA_OUTL(ctlr->r_mem, AHCI_IS, 1 << unit);
AHCI: Fix AHCI driver for ARM. On ARM, we must ensure proper interdevice write ordering. The AHCI interrupt status register must be updated in HW before registers in interrupt controller. Unfortunately, only way how we can do it is readback. Discussed with: mav Approved by: kib (mentor) Differential Revision: https://reviews.freebsd.org/D4240
2015-11-29 11:28:04 +00:00
ATA_RBL(ctlr->r_mem, AHCI_IS);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
}
static void
ahci_intr_one_edge(void *data)
{
struct ahci_controller_irq *irq = data;
struct ahci_controller *ctlr = irq->ctlr;
void *arg;
int unit;
unit = irq->r_irq_rid - 1;
/* Some controllers have edge triggered IS. */
ATA_OUTL(ctlr->r_mem, AHCI_IS, 1 << unit);
if ((arg = ctlr->interrupt[unit].argument))
ctlr->interrupt[unit].function(arg);
AHCI: Fix AHCI driver for ARM. On ARM, we must ensure proper interdevice write ordering. The AHCI interrupt status register must be updated in HW before registers in interrupt controller. Unfortunately, only way how we can do it is readback. Discussed with: mav Approved by: kib (mentor) Differential Revision: https://reviews.freebsd.org/D4240
2015-11-29 11:28:04 +00:00
ATA_RBL(ctlr->r_mem, AHCI_IS);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
struct resource *
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_alloc_resource(device_t dev, device_t child, int type, int *rid,
Convert rman to use rman_res_t instead of u_long Summary: Migrate to using the semi-opaque type rman_res_t to specify rman resources. For now, this is still compatible with u_long. This is step one in migrating rman to use uintmax_t for resources instead of u_long. Going forward, this could feasibly be used to specify architecture-specific definitions of resource ranges, rather than baking a specific integer type into the API. This change has been broken out to facilitate MFC'ing drivers back to 10 without breaking ABI. Reviewed By: jhb Sponsored by: Alex Perez/Inertial Computing Differential Revision: https://reviews.freebsd.org/D5075
2016-01-27 02:23:54 +00:00
rman_res_t start, rman_res_t end, rman_res_t count, u_int flags)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
struct ahci_controller *ctlr = device_get_softc(dev);
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
struct resource *res;
Convert a long to rman_res_t, fixing a sign extension bug. ahci.c had one signed long, which was passed into rman, rather than u_long. After the switch of rman_res_t from size u_long to size uintmax_t, the sign extension caused ranges to get messed up, and ahcich* to not attach. There may be more signed longs used in this way, which will be fixed as they're reported. Reported by: pho
2016-03-20 14:21:07 +00:00
rman_res_t st;
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
int offset, size, unit;
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
bool is_em, is_remapped;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
unit = (intptr_t)device_get_ivars(child);
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
is_em = is_remapped = false;
Create a AHCI attachment for nvme. Intel has created RST and many laptops from vendors like Lenovo and Asus. It's a mechanism for creating multiple boot devices under windows. It effectively hides the nvme drive inside of the ahci controller. The details are supposed to be a trade secret. However, there's a reverse engineered Linux driver, and this implements similar operations to allow nvme drives to attach. The ahci driver attaches nvme children that proxy the remapped resources to the child. nvme_ahci is just like nvme_pci, except it doesn't do the PCI specific things. That's moved into ahci where appropriate. When the nvme drive is remapped, MSI-x interrupts aren't forwarded (the linux driver doesn't know how to use this either). INTx interrupts are used instead. This is suboptimal, but usually sufficient for the laptops these parts are in. This is based loosely on https://www.spinics.net/lists/linux-ide/msg53364.html submitted, but not accepted by, Linux. It was written by Dan Williams. These changes were written from scratch by Olivier Houchard. Submitted by: cognet@ (Olivier Houchard)
2019-08-21 22:18:01 +00:00
if (unit & AHCI_REMAPPED_UNIT) {
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
unit &= AHCI_UNIT;
Create a AHCI attachment for nvme. Intel has created RST and many laptops from vendors like Lenovo and Asus. It's a mechanism for creating multiple boot devices under windows. It effectively hides the nvme drive inside of the ahci controller. The details are supposed to be a trade secret. However, there's a reverse engineered Linux driver, and this implements similar operations to allow nvme drives to attach. The ahci driver attaches nvme children that proxy the remapped resources to the child. nvme_ahci is just like nvme_pci, except it doesn't do the PCI specific things. That's moved into ahci where appropriate. When the nvme drive is remapped, MSI-x interrupts aren't forwarded (the linux driver doesn't know how to use this either). INTx interrupts are used instead. This is suboptimal, but usually sufficient for the laptops these parts are in. This is based loosely on https://www.spinics.net/lists/linux-ide/msg53364.html submitted, but not accepted by, Linux. It was written by Dan Williams. These changes were written from scratch by Olivier Houchard. Submitted by: cognet@ (Olivier Houchard)
2019-08-21 22:18:01 +00:00
unit -= ctlr->channels;
is_remapped = true;
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
} else if (unit & AHCI_EM_UNIT) {
unit &= AHCI_UNIT;
is_em = true;
}
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
res = NULL;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
switch (type) {
case SYS_RES_MEMORY:
Create a AHCI attachment for nvme. Intel has created RST and many laptops from vendors like Lenovo and Asus. It's a mechanism for creating multiple boot devices under windows. It effectively hides the nvme drive inside of the ahci controller. The details are supposed to be a trade secret. However, there's a reverse engineered Linux driver, and this implements similar operations to allow nvme drives to attach. The ahci driver attaches nvme children that proxy the remapped resources to the child. nvme_ahci is just like nvme_pci, except it doesn't do the PCI specific things. That's moved into ahci where appropriate. When the nvme drive is remapped, MSI-x interrupts aren't forwarded (the linux driver doesn't know how to use this either). INTx interrupts are used instead. This is suboptimal, but usually sufficient for the laptops these parts are in. This is based loosely on https://www.spinics.net/lists/linux-ide/msg53364.html submitted, but not accepted by, Linux. It was written by Dan Williams. These changes were written from scratch by Olivier Houchard. Submitted by: cognet@ (Olivier Houchard)
2019-08-21 22:18:01 +00:00
if (is_remapped) {
offset = ctlr->remap_offset + unit * ctlr->remap_size;
size = ctlr->remap_size;
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
} else if (!is_em) {
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
offset = AHCI_OFFSET + (unit << 7);
size = 128;
ahci(4): Allow enclosure emulation without hardware. After 53f5ac1310e allowed SATA device mapping to enclosure slots, it may have sense to provide enclosure device emulation even without real hardware interface like SGPIO just for purposes of physical device location tracking (still assuming straight cabling). MFC after: 1 week Sponsored by: iXsystems, Inc.
2021-12-17 20:24:56 +00:00
} else if ((ctlr->caps & AHCI_CAP_EMS) == 0) {
break;
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
} else if (*rid == 0) {
offset = AHCI_EM_CTL;
size = 4;
} else {
offset = (ctlr->emloc & 0xffff0000) >> 14;
size = (ctlr->emloc & 0x0000ffff) << 2;
if (*rid != 1) {
if (*rid == 2 && (ctlr->capsem &
(AHCI_EM_XMT | AHCI_EM_SMB)) == 0)
offset += size;
else
break;
}
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
st = rman_get_start(ctlr->r_mem);
res = rman_reserve_resource(&ctlr->sc_iomem, st + offset,
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
st + offset + size - 1, size, RF_ACTIVE, child);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (res) {
bus_space_handle_t bsh;
bus_space_tag_t bst;
bsh = rman_get_bushandle(ctlr->r_mem);
bst = rman_get_bustag(ctlr->r_mem);
bus_space_subregion(bst, bsh, offset, 128, &bsh);
rman_set_bushandle(res, bsh);
rman_set_bustag(res, bst);
}
break;
case SYS_RES_IRQ:
if (*rid == ATA_IRQ_RID)
res = ctlr->irqs[0].r_irq;
break;
}
return (res);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_release_resource(device_t dev, device_t child, int type, int rid,
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
struct resource *r)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
switch (type) {
case SYS_RES_MEMORY:
rman_release_resource(r);
return (0);
case SYS_RES_IRQ:
if (rid != ATA_IRQ_RID)
style (9) nits Use return (val); instead of return val; Differential Revision: D838 MFC after: 2 weeks Sponsored by: Multiplay
2014-12-21 16:38:29 +00:00
return (ENOENT);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (0);
}
return (EINVAL);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_setup_intr(device_t dev, device_t child, struct resource *irq,
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int flags, driver_filter_t *filter, driver_intr_t *function,
void *argument, void **cookiep)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
struct ahci_controller *ctlr = device_get_softc(dev);
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
int unit = (intptr_t)device_get_ivars(child) & AHCI_UNIT;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (filter != NULL) {
printf("ahci.c: we cannot use a filter here\n");
return (EINVAL);
}
ctlr->interrupt[unit].function = function;
ctlr->interrupt[unit].argument = argument;
return (0);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_teardown_intr(device_t dev, device_t child, struct resource *irq,
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
void *cookie)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
struct ahci_controller *ctlr = device_get_softc(dev);
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
int unit = (intptr_t)device_get_ivars(child) & AHCI_UNIT;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ctlr->interrupt[unit].function = NULL;
ctlr->interrupt[unit].argument = NULL;
return (0);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_print_child(device_t dev, device_t child)
{
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
intptr_t ivars;
int retval;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
retval = bus_print_child_header(dev, child);
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
ivars = (intptr_t)device_get_ivars(child);
if ((ivars & AHCI_EM_UNIT) == 0)
retval += printf(" at channel %d", (int)ivars & AHCI_UNIT);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
retval += bus_print_child_footer(dev, child);
return (retval);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
int
newbus: Move from bus_child_{pnpinfo,location}_src to bus_child_{pnpinfo,location} with sbuf Now that the upper layers all go through a layer to tie into these information functions that translates an sbuf into char * and len. The current interface suffers issues of what to do in cases of truncation, etc. Instead, migrate all these functions to using struct sbuf and these issues go away. The caller is also in charge of any memory allocation and/or expansion that's needed during this process. Create a bus_generic_child_{pnpinfo,location} and make it default. It just returns success. This is for those busses that have no information for these items. Migrate the now-empty routines to using this as appropriate. Document these new interfaces with man pages, and oversight from before. Reviewed by: jhb, bcr Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D29937
2021-06-23 02:27:57 +00:00
ahci_child_location(device_t dev, device_t child, struct sbuf *sb)
Report ATA/SATA channel number to NewBus at location string.
2010-05-22 07:32:47 +00:00
{
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
intptr_t ivars;
Report ATA/SATA channel number to NewBus at location string.
2010-05-22 07:32:47 +00:00
Fix AHCI Enclosure Management, broken by r351356. ivars value of -1 was used to distinguish EM device, and r351356 left some wrong checks for it. Give EM device separate flag there instead.
2019-08-28 22:04:04 +00:00
ivars = (intptr_t)device_get_ivars(child);
if ((ivars & AHCI_EM_UNIT) == 0)
newbus: Move from bus_child_{pnpinfo,location}_src to bus_child_{pnpinfo,location} with sbuf Now that the upper layers all go through a layer to tie into these information functions that translates an sbuf into char * and len. The current interface suffers issues of what to do in cases of truncation, etc. Instead, migrate all these functions to using struct sbuf and these issues go away. The caller is also in charge of any memory allocation and/or expansion that's needed during this process. Create a bus_generic_child_{pnpinfo,location} and make it default. It just returns success. This is for those busses that have no information for these items. Migrate the now-empty routines to using this as appropriate. Document these new interfaces with man pages, and oversight from before. Reviewed by: jhb, bcr Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D29937
2021-06-23 02:27:57 +00:00
sbuf_printf(sb, "channel=%d", (int)ivars & AHCI_UNIT);
Report ATA/SATA channel number to NewBus at location string.
2010-05-22 07:32:47 +00:00
return (0);
}
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
bus_dma_tag_t
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
ahci_get_dma_tag(device_t dev, device_t child)
{
struct ahci_controller *ctlr = device_get_softc(dev);
return (ctlr->dma_tag);
}
Improve AHCI Enclosure Management and SES interoperation. Since SES specs do not define mechanism to map enclosure slots to SATA disks, AHCI EM code I written many years ago appeared quite useless, that always bugged me. I was thinking whether it was a good idea, but if LSI HBAs do that, why I shouldn't? This change introduces simple non-standard mechanism for the mapping into both AHCI EM and SES code, that makes AHCI EM on capable controllers (most of Intel's) a first-class SES citizen, allowing it to report disk physical path to GEOM, show devices inserted into each enclosure slot in `sesutil map` and `getencstat`, control locate and fault LEDs for specific devices with `sesutil locate adaX on` and `sesutil fault adaX on`, etc. I've successfully tested this on Supermicro X10DRH-i motherboard connected with sideband cable of its S-SATA Mini-SAS connector to SAS815TQ backplane. It can indicate with LEDs Locate, Fault and Rebuild/Remap SES statuses for each disk identical to real SES of Supermicro SAS2 backplanes. MFC after: 2 weeks
2019-06-23 19:05:01 +00:00
void
ahci_attached(device_t dev, struct ahci_channel *ch)
{
struct ahci_controller *ctlr = device_get_softc(dev);
mtx_lock(&ctlr->ch_mtx);
ctlr->ch[ch->unit] = ch;
mtx_unlock(&ctlr->ch_mtx);
}
void
ahci_detached(device_t dev, struct ahci_channel *ch)
{
struct ahci_controller *ctlr = device_get_softc(dev);
mtx_lock(&ctlr->ch_mtx);
mtx_lock(&ch->mtx);
ctlr->ch[ch->unit] = NULL;
mtx_unlock(&ch->mtx);
mtx_unlock(&ctlr->ch_mtx);
}
struct ahci_channel *
ahci_getch(device_t dev, int n)
{
struct ahci_controller *ctlr = device_get_softc(dev);
struct ahci_channel *ch;
KASSERT(n >= 0 && n < AHCI_MAX_PORTS, ("Bad channel number %d", n));
mtx_lock(&ctlr->ch_mtx);
ch = ctlr->ch[n];
if (ch != NULL)
mtx_lock(&ch->mtx);
mtx_unlock(&ctlr->ch_mtx);
return (ch);
}
void
ahci_putch(struct ahci_channel *ch)
{
mtx_unlock(&ch->mtx);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static int
ahci_ch_probe(device_t dev)
{
device_set_desc_copy(dev, "AHCI channel");
Reduce priority of ATA/SATA drivers. Legacy ata(4) -> BUS_PROBE_LOW_PRIORITY; more functional ahci(4), siis(4), mvs(4) -> BUS_PROBE_DEFAULT; BUS_PROBE_VENDOR leave for vendor drivers. MFC after: 2 weeks
2015-03-23 19:47:52 +00:00
return (BUS_PROBE_DEFAULT);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
static int
ahci_ch_disablephy_proc(SYSCTL_HANDLER_ARGS)
{
struct ahci_channel *ch;
int error, value;
ch = arg1;
value = ch->disablephy;
error = sysctl_handle_int(oidp, &value, 0, req);
if (error != 0 || req->newptr == NULL || (value != 0 && value != 1))
return (error);
mtx_lock(&ch->mtx);
ch->disablephy = value;
if (value) {
ahci_ch_deinit(ch->dev);
} else {
ahci_ch_init(ch->dev);
ahci_phy_check_events(ch, ATA_SE_PHY_CHANGED | ATA_SE_EXCHANGED);
}
mtx_unlock(&ch->mtx);
return (0);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static int
ahci_ch_attach(device_t dev)
{
struct ahci_controller *ctlr = device_get_softc(device_get_parent(dev));
struct ahci_channel *ch = device_get_softc(dev);
struct cam_devq *devq;
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
struct sysctl_ctx_list *ctx;
struct sysctl_oid *tree;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
int rid, error, i, sata_rev = 0;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
u_int32_t version;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->dev = dev;
ch->unit = (intptr_t)device_get_ivars(dev);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
ch->caps = ctlr->caps;
ch->caps2 = ctlr->caps2;
Add AHCI attachment code for Allwinner A10/A20 SoCs. The Allwinner SoC has an AHCI device on its internal main bus rather than the PCI bus. This SoC is somewhat underdocumented, and its SATA controller is no exception. The methods to support this chip were harvested from the Linux Allwinner SDK, and then constants invented to describe what's going on based on low-level constants contained in the SATA standard and guess work. This SoC requires a specific AHCI channel setup in order to start the operations on the channel properly. Clock setup and AHCI channel setup idea came from NetBSD. Tested on Cubieboard 2 and Banana pi (and attachment on Cubieboard by Pratik Singhal). Differential Revision: https://reviews.freebsd.org/D737 Submitted by: imp Reviewed by: imp, ganbold, mav, andrew
2015-07-03 14:11:01 +00:00
ch->start = ctlr->ch_start;
Change the way in which AHCI+PATA combined controllers, such as JMicron and Marvell handled. Instead of trying to attach two different drivers to single device, wrapping each call, make one of them (atajmicron, atamarvell) attach do device solely, but create child device for AHCI driver, passing it all required resources. It is quite easy, as none of resources are shared, except IRQ. As result, it: - makes drivers operation more independent and straitforward, - allows to use new ahci(4) driver with such devices, adding support for new features, such as PMP and NCQ, same time keeping legacy PATA support, - will allow to just drop old ataahci driver, when it's time come.
2009-11-16 15:38:27 +00:00
ch->quirks = ctlr->quirks;
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
ch->vendorid = ctlr->vendorid;
ch->deviceid = ctlr->deviceid;
ch->subvendorid = ctlr->subvendorid;
ch->subdeviceid = ctlr->subdeviceid;
Fix small typo. Submitted by: Artem Belevich
2010-11-22 23:04:25 +00:00
ch->numslots = ((ch->caps & AHCI_CAP_NCS) >> AHCI_CAP_NCS_SHIFT) + 1;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
mtx_init(&ch->mtx, "AHCI channel lock", NULL, MTX_DEF);
Initialize variables before resource_int_value(). Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com>
2014-09-12 08:37:21 +00:00
ch->pm_level = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
resource_int_value(device_get_name(dev),
device_get_unit(dev), "pm_level", &ch->pm_level);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
STAILQ_INIT(&ch->doneq);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (ch->pm_level > 3)
callout_init_mtx(&ch->pm_timer, &ch->mtx, 0);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
callout_init_mtx(&ch->reset_timer, &ch->mtx, 0);
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
/* JMicron external ports (0) sometimes limited */
if ((ctlr->quirks & AHCI_Q_SATA1_UNIT0) && ch->unit == 0)
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
sata_rev = 1;
Add support for SATA part of Marvell 88SE912x controllers to ahci(4). Limit early revisions from 6Gb/s to 3Gb/s by default, or they negotiate only 1.5Gbps, when 3Gb/s devices connected. Add dummy driver for PATA part of these controllers, preventing generic driver attach them. It causes system freeze when SATA controller used after PATA was touched.
2010-01-26 15:25:24 +00:00
if (ch->quirks & AHCI_Q_SATA2)
sata_rev = 2;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
resource_int_value(device_get_name(dev),
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
device_get_unit(dev), "sata_rev", &sata_rev);
for (i = 0; i < 16; i++) {
ch->user[i].revision = sata_rev;
ch->user[i].mode = 0;
ch->user[i].bytecount = 8192;
ch->user[i].tags = ch->numslots;
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
ch->user[i].caps = 0;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
ch->curr[i] = ch->user[i];
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
if (ch->pm_level) {
ch->user[i].caps = CTS_SATA_CAPS_H_PMREQ |
CTS_SATA_CAPS_H_APST |
CTS_SATA_CAPS_D_PMREQ | CTS_SATA_CAPS_D_APST;
}
Improve SATA Asynchronous Notification feature support in CAM: - make SATA SIMs announce capabilities to handle SDB with Notification bit; - make PMP driver honor this SIMs capability; - make SATA XPT to negotiate and enable this feature for ATAPI devices. This feature allows supporting SATA ATAPI devices to inform system about some events happened, that may require attention. In my case this allows LG GH22LS50 SATA DVR-RW drive to report tray open/close events. Events reported to CAM in form of AC_SCSI_AEN async. Further they could be used as a hints for checking device status and reporting media change to upper layers, for example, via spoiling mechanism of GEOM.
2011-04-13 16:20:54 +00:00
ch->user[i].caps |= CTS_SATA_CAPS_H_DMAAA |
CTS_SATA_CAPS_H_AN;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
}
Refactor enclosure manegement support in ahci(4). Move it out into separate subdevice ahciem. Emulate SEMB SES device from AHCI LED interface to expose it to users in form of ses(4) CAM device. If we ever see AHCI controllers supporting SES of SAF-TE over I2C as described by specification, they should fit well into this new picture. Sponsored by: iXsystems, Inc.
2012-07-26 13:44:48 +00:00
rid = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (!(ch->r_mem = bus_alloc_resource_any(dev, SYS_RES_MEMORY,
&rid, RF_ACTIVE)))
return (ENXIO);
Fix channel initialization in FBS mode. Due to reading initialized variable, FIS receive area was always allocated as 256 bytes, suitable for command-based switching, instead of 4096 bytes, required for FIS-based switching. This caused memory corruption in case of port multipliers used on FBS-capable HBAs (Marvell). MFC after: 1 week
2016-09-07 13:51:34 +00:00
ch->chcaps = ATA_INL(ch->r_mem, AHCI_P_CMD);
version = ATA_INL(ctlr->r_mem, AHCI_VS);
if (version < 0x00010200 && (ctlr->caps & AHCI_CAP_FBSS))
ch->chcaps |= AHCI_P_CMD_FBSCP;
if (ch->caps2 & AHCI_CAP2_SDS)
ch->chscaps = ATA_INL(ch->r_mem, AHCI_P_DEVSLP);
if (bootverbose) {
device_printf(dev, "Caps:%s%s%s%s%s%s\n",
(ch->chcaps & AHCI_P_CMD_HPCP) ? " HPCP":"",
(ch->chcaps & AHCI_P_CMD_MPSP) ? " MPSP":"",
(ch->chcaps & AHCI_P_CMD_CPD) ? " CPD":"",
(ch->chcaps & AHCI_P_CMD_ESP) ? " ESP":"",
(ch->chcaps & AHCI_P_CMD_FBSCP) ? " FBSCP":"",
(ch->chscaps & AHCI_P_DEVSLP_DSP) ? " DSP":"");
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_dmainit(dev);
ahci_slotsalloc(dev);
mtx_lock(&ch->mtx);
For some uniformity move ahci_ch_init() call under the lock. Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com> MFC after: 2 weeks
2015-02-26 12:51:05 +00:00
ahci_ch_init(dev);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
rid = ATA_IRQ_RID;
if (!(ch->r_irq = bus_alloc_resource_any(dev, SYS_RES_IRQ,
&rid, RF_SHAREABLE | RF_ACTIVE))) {
device_printf(dev, "Unable to map interrupt\n");
Fix attach errors handling. Found with: Coverity Prevent(tm) CID: 3424
2010-06-05 08:02:19 +00:00
error = ENXIO;
goto err0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
if ((bus_setup_intr(dev, ch->r_irq, ATA_INTR_FLAGS, NULL,
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ctlr->direct ? ahci_ch_intr_direct : ahci_ch_intr,
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ch, &ch->ih))) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(dev, "Unable to setup interrupt\n");
error = ENXIO;
goto err1;
}
/* Create the device queue for our SIM. */
devq = cam_simq_alloc(ch->numslots);
if (devq == NULL) {
device_printf(dev, "Unable to allocate simq\n");
error = ENOMEM;
goto err1;
}
/* Construct SIM entry */
ch->sim = cam_sim_alloc(ahciaction, ahcipoll, "ahcich", ch,
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_get_unit(dev), (struct mtx *)&ch->mtx,
Partially workaround ASMedia HBA error recovery. Taking closer look on my ASM1062 I found that it has bunch of issues around error recovery: reported wrong CCS, failed commands reported as completed, READ LOG EXT times out after NCQ error. This patch workarounds first two problems, that were making ATAPI devices close to unusable on these HBAs. MFC after: 2 weeks
2017-01-25 18:16:17 +00:00
(ch->quirks & AHCI_Q_NOCCS) ? 1 : min(2, ch->numslots),
MFp4: Check SNCQ HBA capability bit when reporting NCQ support to CAM.
2009-11-14 20:06:04 +00:00
(ch->caps & AHCI_CAP_SNCQ) ? ch->numslots : 0,
devq);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ch->sim == NULL) {
Fix attach errors handling. Found with: Coverity Prevent(tm) CID: 3424
2010-06-05 08:02:19 +00:00
cam_simq_free(devq);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(dev, "unable to allocate sim\n");
error = ENOMEM;
Fix attach errors handling. Found with: Coverity Prevent(tm) CID: 3424
2010-06-05 08:02:19 +00:00
goto err1;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
if (xpt_bus_register(ch->sim, dev, 0) != CAM_SUCCESS) {
device_printf(dev, "unable to register xpt bus\n");
error = ENXIO;
goto err2;
}
if (xpt_create_path(&ch->path, /*periph*/NULL, cam_sim_path(ch->sim),
CAM_TARGET_WILDCARD, CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
device_printf(dev, "unable to create path\n");
error = ENXIO;
goto err3;
}
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (ch->pm_level > 3) {
callout_reset(&ch->pm_timer,
(ch->pm_level == 4) ? hz / 1000 : hz / 8,
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_ch_pm, ch);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
mtx_unlock(&ch->mtx);
Improve AHCI Enclosure Management and SES interoperation. Since SES specs do not define mechanism to map enclosure slots to SATA disks, AHCI EM code I written many years ago appeared quite useless, that always bugged me. I was thinking whether it was a good idea, but if LSI HBAs do that, why I shouldn't? This change introduces simple non-standard mechanism for the mapping into both AHCI EM and SES code, that makes AHCI EM on capable controllers (most of Intel's) a first-class SES citizen, allowing it to report disk physical path to GEOM, show devices inserted into each enclosure slot in `sesutil map` and `getencstat`, control locate and fault LEDs for specific devices with `sesutil locate adaX on` and `sesutil fault adaX on`, etc. I've successfully tested this on Supermicro X10DRH-i motherboard connected with sideband cable of its S-SATA Mini-SAS connector to SAS815TQ backplane. It can indicate with LEDs Locate, Fault and Rebuild/Remap SES statuses for each disk identical to real SES of Supermicro SAS2 backplanes. MFC after: 2 weeks
2019-06-23 19:05:01 +00:00
ahci_attached(device_get_parent(dev), ch);
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
ctx = device_get_sysctl_ctx(dev);
tree = device_get_sysctl_tree(dev);
SYSCTL_ADD_PROC(ctx, SYSCTL_CHILDREN(tree), OID_AUTO, "disable_phy",
Mark some sysctls as CTLFLAG_MPSAFE. MFC after: 2 weeks
2021-08-11 00:44:27 +00:00
CTLFLAG_RW | CTLTYPE_UINT | CTLFLAG_MPSAFE, ch,
Mark more nodes as CTLFLAG_MPSAFE or CTLFLAG_NEEDGIANT (17 of many) r357614 added CTLFLAG_NEEDGIANT to make it easier to find nodes that are still not MPSAFE (or already are but aren’t properly marked). Use it in preparation for a general review of all nodes. This is non-functional change that adds annotations to SYSCTL_NODE and SYSCTL_PROC nodes using one of the soon-to-be-required flags. Mark all obvious cases as MPSAFE. All entries that haven't been marked as MPSAFE before are by default marked as NEEDGIANT Approved by: kib (mentor, blanket) Commented by: kib, gallatin, melifaro Differential Revision: https://reviews.freebsd.org/D23718
2020-02-26 14:26:36 +00:00
0, ahci_ch_disablephy_proc, "IU", "Disable PHY");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (0);
err3:
xpt_bus_deregister(cam_sim_path(ch->sim));
err2:
cam_sim_free(ch->sim, /*free_devq*/TRUE);
err1:
bus_release_resource(dev, SYS_RES_IRQ, ATA_IRQ_RID, ch->r_irq);
Fix attach errors handling. Found with: Coverity Prevent(tm) CID: 3424
2010-06-05 08:02:19 +00:00
err0:
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_release_resource(dev, SYS_RES_MEMORY, ch->unit, ch->r_mem);
mtx_unlock(&ch->mtx);
Add missing mtx_destroy() on channel attach failure.
2010-10-25 07:41:21 +00:00
mtx_destroy(&ch->mtx);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (error);
}
static int
ahci_ch_detach(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
Improve AHCI Enclosure Management and SES interoperation. Since SES specs do not define mechanism to map enclosure slots to SATA disks, AHCI EM code I written many years ago appeared quite useless, that always bugged me. I was thinking whether it was a good idea, but if LSI HBAs do that, why I shouldn't? This change introduces simple non-standard mechanism for the mapping into both AHCI EM and SES code, that makes AHCI EM on capable controllers (most of Intel's) a first-class SES citizen, allowing it to report disk physical path to GEOM, show devices inserted into each enclosure slot in `sesutil map` and `getencstat`, control locate and fault LEDs for specific devices with `sesutil locate adaX on` and `sesutil fault adaX on`, etc. I've successfully tested this on Supermicro X10DRH-i motherboard connected with sideband cable of its S-SATA Mini-SAS connector to SAS815TQ backplane. It can indicate with LEDs Locate, Fault and Rebuild/Remap SES statuses for each disk identical to real SES of Supermicro SAS2 backplanes. MFC after: 2 weeks
2019-06-23 19:05:01 +00:00
ahci_detached(device_get_parent(dev), ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
mtx_lock(&ch->mtx);
xpt_async(AC_LOST_DEVICE, ch->path, NULL);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
/* Forget about reset. */
if (ch->resetting) {
ch->resetting = 0;
xpt_release_simq(ch->sim, TRUE);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
xpt_free_path(ch->path);
xpt_bus_deregister(cam_sim_path(ch->sim));
cam_sim_free(ch->sim, /*free_devq*/TRUE);
mtx_unlock(&ch->mtx);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (ch->pm_level > 3)
callout_drain(&ch->pm_timer);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
callout_drain(&ch->reset_timer);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_teardown_intr(dev, ch->r_irq, ch->ih);
bus_release_resource(dev, SYS_RES_IRQ, ATA_IRQ_RID, ch->r_irq);
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
ahci_ch_deinit(dev);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_slotsfree(dev);
ahci_dmafini(dev);
bus_release_resource(dev, SYS_RES_MEMORY, ch->unit, ch->r_mem);
mtx_destroy(&ch->mtx);
return (0);
}
static int
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
ahci_ch_init(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
uint64_t work;
/* Disable port interrupts */
ATA_OUTL(ch->r_mem, AHCI_P_IE, 0);
/* Setup work areas */
work = ch->dma.work_bus + AHCI_CL_OFFSET;
ATA_OUTL(ch->r_mem, AHCI_P_CLB, work & 0xffffffff);
ATA_OUTL(ch->r_mem, AHCI_P_CLBU, work >> 32);
work = ch->dma.rfis_bus;
ATA_OUTL(ch->r_mem, AHCI_P_FB, work & 0xffffffff);
ATA_OUTL(ch->r_mem, AHCI_P_FBU, work >> 32);
/* Activate the channel and power/spin up device */
ATA_OUTL(ch->r_mem, AHCI_P_CMD,
(AHCI_P_CMD_ACTIVE | AHCI_P_CMD_POD | AHCI_P_CMD_SUD |
((ch->pm_level == 2 || ch->pm_level == 3) ? AHCI_P_CMD_ALPE : 0) |
((ch->pm_level > 2) ? AHCI_P_CMD_ASP : 0 )));
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_start_fr(ch);
ahci_start(ch, 1);
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
return (0);
}
static int
ahci_ch_deinit(device_t dev)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
struct ahci_channel *ch = device_get_softc(dev);
/* Disable port interrupts. */
ATA_OUTL(ch->r_mem, AHCI_P_IE, 0);
/* Reset command register. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop(ch);
ahci_stop_fr(ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_CMD, 0);
/* Allow everything, including partial and slumber modes. */
ATA_OUTL(ch->r_mem, AHCI_P_SCTL, 0);
/* Request slumber mode transition and give some time to get there. */
ATA_OUTL(ch->r_mem, AHCI_P_CMD, AHCI_P_CMD_SLUMBER);
DELAY(100);
/* Disable PHY. */
ATA_OUTL(ch->r_mem, AHCI_P_SCTL, ATA_SC_DET_DISABLE);
return (0);
}
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
static int
ahci_ch_suspend(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
mtx_lock(&ch->mtx);
xpt_freeze_simq(ch->sim, 1);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
/* Forget about reset. */
if (ch->resetting) {
ch->resetting = 0;
callout_stop(&ch->reset_timer);
xpt_release_simq(ch->sim, TRUE);
}
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
while (ch->oslots)
msleep(ch, &ch->mtx, PRIBIO, "ahcisusp", hz/100);
ahci_ch_deinit(dev);
mtx_unlock(&ch->mtx);
return (0);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static int
ahci_ch_resume(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
mtx_lock(&ch->mtx);
ahci_ch_init(dev);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
Improve suspend/resume support. Make sure controller is idle on suspend and reset it on resume.
2010-05-21 13:29:28 +00:00
xpt_release_simq(ch->sim, TRUE);
mtx_unlock(&ch->mtx);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (0);
}
static device_method_t ahcich_methods[] = {
DEVMETHOD(device_probe, ahci_ch_probe),
DEVMETHOD(device_attach, ahci_ch_attach),
DEVMETHOD(device_detach, ahci_ch_detach),
DEVMETHOD(device_suspend, ahci_ch_suspend),
DEVMETHOD(device_resume, ahci_ch_resume),
- Const'ify the ahci_ids table. - Use DEVMETHOD_END. - Use NULL instead of 0 for pointers. MFC after: 3 days
2014-12-28 20:42:28 +00:00
DEVMETHOD_END
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
};
static driver_t ahcich_driver = {
"ahcich",
ahcich_methods,
sizeof(struct ahci_channel)
};
ahci: Remove unused devclass arguments to DRIVER_MODULE.
2022-04-19 20:39:53 +00:00
DRIVER_MODULE(ahcich, ahci, ahcich_driver, NULL, NULL);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
struct ahci_dc_cb_args {
bus_addr_t maddr;
int error;
};
static void
ahci_dmainit(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
struct ahci_dc_cb_args dcba;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
size_t rfsize;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
int error;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Command area. */
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dma_tag_create(bus_get_dma_tag(dev), 1024, 0,
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR,
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
NULL, NULL, AHCI_WORK_SIZE, 1, AHCI_WORK_SIZE,
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
0, NULL, NULL, &ch->dma.work_tag);
if (error != 0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
goto error;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dmamem_alloc(ch->dma.work_tag, (void **)&ch->dma.work,
BUS_DMA_ZERO, &ch->dma.work_map);
if (error != 0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
goto error;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dmamap_load(ch->dma.work_tag, ch->dma.work_map, ch->dma.work,
AHCI_WORK_SIZE, ahci_dmasetupc_cb, &dcba, BUS_DMA_NOWAIT);
if (error != 0 || (error = dcba.error) != 0) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_dmamem_free(ch->dma.work_tag, ch->dma.work, ch->dma.work_map);
goto error;
}
ch->dma.work_bus = dcba.maddr;
/* FIS receive area. */
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (ch->chcaps & AHCI_P_CMD_FBSCP)
rfsize = 4096;
else
rfsize = 256;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dma_tag_create(bus_get_dma_tag(dev), rfsize, 0,
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR,
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
NULL, NULL, rfsize, 1, rfsize,
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
0, NULL, NULL, &ch->dma.rfis_tag);
if (error != 0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
goto error;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dmamem_alloc(ch->dma.rfis_tag, (void **)&ch->dma.rfis, 0,
&ch->dma.rfis_map);
if (error != 0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
goto error;
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dmamap_load(ch->dma.rfis_tag, ch->dma.rfis_map, ch->dma.rfis,
rfsize, ahci_dmasetupc_cb, &dcba, BUS_DMA_NOWAIT);
if (error != 0 || (error = dcba.error) != 0) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_dmamem_free(ch->dma.rfis_tag, ch->dma.rfis, ch->dma.rfis_map);
goto error;
}
ch->dma.rfis_bus = dcba.maddr;
/* Data area. */
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
error = bus_dma_tag_create(bus_get_dma_tag(dev), 2, 0,
Create controller-level DMA tag, handling range of supported addresses. That simplifies logic for channels and gives the bus information about what device actually allocated the tag. Submitted by: jhb@
2013-04-10 20:38:15 +00:00
BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR,
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
NULL, NULL,
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
AHCI_SG_ENTRIES * PAGE_SIZE, AHCI_SG_ENTRIES, AHCI_PRD_MAX,
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
0, busdma_lock_mutex, &ch->mtx, &ch->dma.data_tag);
if (error != 0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
goto error;
return;
error:
Fix use of busdma(9) KPI in ahci(4). Use BUS_DMA_NOWAIT for loads at initialization time. Report actual numeric error code if any problem occurs at the initialization. Reported and tested by: pho Reviewed by: mav Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18741
2019-01-07 02:39:40 +00:00
device_printf(dev, "WARNING - DMA initialization failed, error %d\n",
error);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_dmafini(dev);
}
static void
ahci_dmasetupc_cb(void *xsc, bus_dma_segment_t *segs, int nsegs, int error)
{
struct ahci_dc_cb_args *dcba = (struct ahci_dc_cb_args *)xsc;
if (!(dcba->error = error))
dcba->maddr = segs[0].ds_addr;
}
static void
ahci_dmafini(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
if (ch->dma.data_tag) {
bus_dma_tag_destroy(ch->dma.data_tag);
ch->dma.data_tag = NULL;
}
if (ch->dma.rfis_bus) {
bus_dmamap_unload(ch->dma.rfis_tag, ch->dma.rfis_map);
bus_dmamem_free(ch->dma.rfis_tag, ch->dma.rfis, ch->dma.rfis_map);
ch->dma.rfis_bus = 0;
ch->dma.rfis = NULL;
}
if (ch->dma.work_bus) {
bus_dmamap_unload(ch->dma.work_tag, ch->dma.work_map);
bus_dmamem_free(ch->dma.work_tag, ch->dma.work, ch->dma.work_map);
ch->dma.work_bus = 0;
ch->dma.work = NULL;
}
if (ch->dma.work_tag) {
bus_dma_tag_destroy(ch->dma.work_tag);
ch->dma.work_tag = NULL;
}
}
static void
ahci_slotsalloc(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
int i;
/* Alloc and setup command/dma slots */
bzero(ch->slot, sizeof(ch->slot));
for (i = 0; i < ch->numslots; i++) {
struct ahci_slot *slot = &ch->slot[i];
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
slot->ch = ch;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
slot->slot = i;
slot->state = AHCI_SLOT_EMPTY;
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
slot->ct_offset = AHCI_CT_OFFSET + AHCI_CT_SIZE * i;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
slot->ccb = NULL;
callout_init_mtx(&slot->timeout, &ch->mtx, 0);
if (bus_dmamap_create(ch->dma.data_tag, 0, &slot->dma.data_map))
device_printf(ch->dev, "FAILURE - create data_map\n");
}
}
static void
ahci_slotsfree(device_t dev)
{
struct ahci_channel *ch = device_get_softc(dev);
int i;
/* Free all dma slots */
for (i = 0; i < ch->numslots; i++) {
struct ahci_slot *slot = &ch->slot[i];
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
callout_drain(&slot->timeout);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (slot->dma.data_map) {
bus_dmamap_destroy(ch->dma.data_tag, slot->dma.data_map);
slot->dma.data_map = NULL;
}
}
}
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_phy_check_events(struct ahci_channel *ch, u_int32_t serr)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if (((ch->pm_level == 0) && (serr & ATA_SE_PHY_CHANGED)) ||
((ch->pm_level != 0 || ch->listening) && (serr & ATA_SE_EXCHANGED))) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
u_int32_t status = ATA_INL(ch->r_mem, AHCI_P_SSTS);
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
union ccb *ccb;
Reset port on disconnect event, to abort any running requests.
2010-01-29 12:47:04 +00:00
if (bootverbose) {
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if ((status & ATA_SS_DET_MASK) != ATA_SS_DET_NO_DEVICE)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "CONNECT requested\n");
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
else
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "DISCONNECT requested\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
if ((ccb = xpt_alloc_ccb_nowait()) == NULL)
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
return (0);
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
if (xpt_create_path(&ccb->ccb_h.path, NULL,
cam_sim_path(ch->sim),
CAM_TARGET_WILDCARD, CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
xpt_free_ccb(ccb);
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
return (0);
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
}
xpt_rescan(ccb);
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
return (1);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
return (0);
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_cpd_check_events(struct ahci_channel *ch)
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
{
u_int32_t status;
union ccb *ccb;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_t dev;
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if (ch->pm_level == 0)
return;
status = ATA_INL(ch->r_mem, AHCI_P_CMD);
if ((status & AHCI_P_CMD_CPD) == 0)
return;
if (bootverbose) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
dev = ch->dev;
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if (status & AHCI_P_CMD_CPS) {
device_printf(dev, "COLD CONNECT requested\n");
} else
device_printf(dev, "COLD DISCONNECT requested\n");
}
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if ((ccb = xpt_alloc_ccb_nowait()) == NULL)
return;
if (xpt_create_path(&ccb->ccb_h.path, NULL, cam_sim_path(ch->sim),
CAM_TARGET_WILDCARD, CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
xpt_free_ccb(ccb);
return;
}
xpt_rescan(ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_notify_events(struct ahci_channel *ch, u_int32_t status)
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
{
struct cam_path *dpath;
int i;
Add Asynchronous Notification support for controllers without SNTF capability by snooping SDB FIS receive area. It should be even faster then regular way, but less reliable.
2009-12-06 23:56:54 +00:00
if (ch->caps & AHCI_CAP_SSNTF)
ATA_OUTL(ch->r_mem, AHCI_P_SNTF, status);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if (bootverbose)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "SNTF 0x%04x\n", status);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
for (i = 0; i < 16; i++) {
if ((status & (1 << i)) == 0)
continue;
if (xpt_create_path(&dpath, NULL,
xpt_path_path_id(ch->path), i, 0) == CAM_REQ_CMP) {
xpt_async(AC_SCSI_AEN, dpath, NULL);
xpt_free_path(dpath);
}
}
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(struct ahci_channel *ch, union ccb *ccb)
{
mtx_assert(&ch->mtx, MA_OWNED);
if ((ccb->ccb_h.func_code & XPT_FC_QUEUED) == 0 ||
ch->batch == 0) {
xpt_done(ccb);
return;
}
STAILQ_INSERT_TAIL(&ch->doneq, &ccb->ccb_h, sim_links.stqe);
}
static void
ahci_ch_intr(void *arg)
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = (struct ahci_channel *)arg;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
uint32_t istatus;
/* Read interrupt statuses. */
istatus = ATA_INL(ch->r_mem, AHCI_P_IS);
mtx_lock(&ch->mtx);
ahci_ch_intr_main(ch, istatus);
mtx_unlock(&ch->mtx);
}
static void
ahci_ch_intr_direct(void *arg)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = (struct ahci_channel *)arg;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
struct ccb_hdr *ccb_h;
uint32_t istatus;
Use a local STAILQ for unlocked done CCB processing in ahci direct mode Previously it was possible for issues e.g. use after free, to result from processing the done queue while not holding the channel lock. While this should never happen in practice, unexpected code flows which result in two threads processing from the same queue may be possible. We now use a local STAILQ to prevent this ever being an issue. Sponsored by: Multiplay
2014-09-27 19:14:22 +00:00
STAILQ_HEAD(, ccb_hdr) tmp_doneq = STAILQ_HEAD_INITIALIZER(tmp_doneq);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
/* Read interrupt statuses. */
istatus = ATA_INL(ch->r_mem, AHCI_P_IS);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
mtx_lock(&ch->mtx);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ch->batch = 1;
ahci_ch_intr_main(ch, istatus);
ch->batch = 0;
Use a local STAILQ for unlocked done CCB processing in ahci direct mode Previously it was possible for issues e.g. use after free, to result from processing the done queue while not holding the channel lock. While this should never happen in practice, unexpected code flows which result in two threads processing from the same queue may be possible. We now use a local STAILQ to prevent this ever being an issue. Sponsored by: Multiplay
2014-09-27 19:14:22 +00:00
/*
* Prevent the possibility of issues caused by processing the queue
* while unlocked below by moving the contents to a local queue.
*/
STAILQ_CONCAT(&tmp_doneq, &ch->doneq);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
mtx_unlock(&ch->mtx);
Use a local STAILQ for unlocked done CCB processing in ahci direct mode Previously it was possible for issues e.g. use after free, to result from processing the done queue while not holding the channel lock. While this should never happen in practice, unexpected code flows which result in two threads processing from the same queue may be possible. We now use a local STAILQ to prevent this ever being an issue. Sponsored by: Multiplay
2014-09-27 19:14:22 +00:00
while ((ccb_h = STAILQ_FIRST(&tmp_doneq)) != NULL) {
STAILQ_REMOVE_HEAD(&tmp_doneq, sim_links.stqe);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
xpt_done_direct((union ccb *)ccb_h);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
static void
ahci_ch_pm(void *arg)
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = (struct ahci_channel *)arg;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
uint32_t work;
if (ch->numrslots != 0)
return;
work = ATA_INL(ch->r_mem, AHCI_P_CMD);
if (ch->pm_level == 4)
work |= AHCI_P_CMD_PARTIAL;
else
work |= AHCI_P_CMD_SLUMBER;
ATA_OUTL(ch->r_mem, AHCI_P_CMD, work);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_ch_intr_main(struct ahci_channel *ch, uint32_t istatus)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
uint32_t cstatus, serr = 0, sntf = 0, ok, err;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
enum ahci_err_type et;
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
int i, ccs, port, reset = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
/* Clear interrupt statuses. */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_IS, istatus);
/* Read command statuses. */
Depending on combination of running commands (NCQ/non-NCQ) try to avoid extra read from PxCI/PxSACT registers. If only NCQ commands are running, we don't really need PxCI. If only non-NCQ commands are running we don't need PxSACT. Mixed set may happen only on controllers with FIS-based switching when port multiplier is attached, and then we have to read both registers. MFC after: 1 month
2013-03-25 08:50:51 +00:00
if (ch->numtslots != 0)
cstatus = ATA_INL(ch->r_mem, AHCI_P_SACT);
else
cstatus = 0;
if (ch->numrslots != ch->numtslots)
cstatus |= ATA_INL(ch->r_mem, AHCI_P_CI);
/* Read SNTF in one of possible ways. */
Read Asynchronous Notification statuses only if Port Multiplier or ATAPI device are connected. ATA disks are not using ANs, while the extra register read operation is quite expensive.
2013-03-25 13:58:17 +00:00
if ((istatus & AHCI_P_IX_SDB) &&
(ch->pm_present || ch->curr[0].atapi != 0)) {
Add Asynchronous Notification support for controllers without SNTF capability by snooping SDB FIS receive area. It should be even faster then regular way, but less reliable.
2009-12-06 23:56:54 +00:00
if (ch->caps & AHCI_CAP_SSNTF)
sntf = ATA_INL(ch->r_mem, AHCI_P_SNTF);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
else if (ch->fbs_enabled) {
u_int8_t *fis = ch->dma.rfis + 0x58;
for (i = 0; i < 16; i++) {
if (fis[1] & 0x80) {
fis[1] &= 0x7f;
sntf |= 1 << i;
}
fis += 256;
}
} else {
Add Asynchronous Notification support for controllers without SNTF capability by snooping SDB FIS receive area. It should be even faster then regular way, but less reliable.
2009-12-06 23:56:54 +00:00
u_int8_t *fis = ch->dma.rfis + 0x58;
if (fis[1] & 0x80)
sntf = (1 << (fis[1] & 0x0f));
}
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Process PHY events */
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (istatus & (AHCI_P_IX_PC | AHCI_P_IX_PRC | AHCI_P_IX_OF |
AHCI_P_IX_IF | AHCI_P_IX_HBD | AHCI_P_IX_HBF | AHCI_P_IX_TFE)) {
serr = ATA_INL(ch->r_mem, AHCI_P_SERR);
if (serr) {
ATA_OUTL(ch->r_mem, AHCI_P_SERR, serr);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
reset = ahci_phy_check_events(ch, serr);
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
}
}
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
/* Process cold presence detection events */
if ((istatus & AHCI_P_IX_CPD) && !reset)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_cpd_check_events(ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Process command errors */
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (istatus & (AHCI_P_IX_OF | AHCI_P_IX_IF |
AHCI_P_IX_HBD | AHCI_P_IX_HBF | AHCI_P_IX_TFE)) {
Partially workaround ASMedia HBA error recovery. Taking closer look on my ASM1062 I found that it has bunch of issues around error recovery: reported wrong CCS, failed commands reported as completed, READ LOG EXT times out after NCQ error. This patch workarounds first two problems, that were making ATAPI devices close to unusable on these HBAs. MFC after: 2 weeks
2017-01-25 18:16:17 +00:00
if (ch->quirks & AHCI_Q_NOCCS) {
/*
* ASMedia chips sometimes report failed commands as
* completed. Count all running commands as failed.
*/
cstatus |= ch->rslots;
/* They also report wrong CCS, so try to guess one. */
ccs = powerof2(cstatus) ? ffs(cstatus) - 1 : -1;
} else {
ccs = (ATA_INL(ch->r_mem, AHCI_P_CMD) &
AHCI_P_CMD_CCS_MASK) >> AHCI_P_CMD_CCS_SHIFT;
}
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
//device_printf(dev, "%s ERROR is %08x cs %08x ss %08x rs %08x tfd %02x serr %08x fbs %08x ccs %d\n",
// __func__, istatus, cstatus, sstatus, ch->rslots, ATA_INL(ch->r_mem, AHCI_P_TFD),
// serr, ATA_INL(ch->r_mem, AHCI_P_FBS), ccs);
port = -1;
if (ch->fbs_enabled) {
uint32_t fbs = ATA_INL(ch->r_mem, AHCI_P_FBS);
if (fbs & AHCI_P_FBS_SDE) {
port = (fbs & AHCI_P_FBS_DWE)
>> AHCI_P_FBS_DWE_SHIFT;
} else {
for (i = 0; i < 16; i++) {
if (ch->numrslotspd[i] == 0)
continue;
if (port == -1)
port = i;
else if (port != i) {
port = -2;
break;
}
}
}
}
Depending on combination of running commands (NCQ/non-NCQ) try to avoid extra read from PxCI/PxSACT registers. If only NCQ commands are running, we don't really need PxCI. If only non-NCQ commands are running we don't need PxSACT. Mixed set may happen only on controllers with FIS-based switching when port multiplier is attached, and then we have to read both registers. MFC after: 1 month
2013-03-25 08:50:51 +00:00
err = ch->rslots & cstatus;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else {
ccs = 0;
err = 0;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
port = -1;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
sys/dev: minor spelling fixes. Most affect comments, very few have user-visible effects.
2016-05-03 03:41:25 +00:00
/* Complete all successful commands. */
Depending on combination of running commands (NCQ/non-NCQ) try to avoid extra read from PxCI/PxSACT registers. If only NCQ commands are running, we don't really need PxCI. If only non-NCQ commands are running we don't need PxSACT. Mixed set may happen only on controllers with FIS-based switching when port multiplier is attached, and then we have to read both registers. MFC after: 1 month
2013-03-25 08:50:51 +00:00
ok = ch->rslots & ~cstatus;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
for (i = 0; i < ch->numslots; i++) {
if ((ok >> i) & 1)
ahci_end_transaction(&ch->slot[i], AHCI_ERR_NONE);
}
/* On error, complete the rest of commands with error statuses. */
if (err) {
if (ch->frozen) {
union ccb *fccb = ch->frozen;
ch->frozen = NULL;
fccb->ccb_h.status = CAM_REQUEUE_REQ | CAM_RELEASE_SIMQ;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (!(fccb->ccb_h.status & CAM_DEV_QFRZN)) {
xpt_freeze_devq(fccb->ccb_h.path, 1);
fccb->ccb_h.status |= CAM_DEV_QFRZN;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, fccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
for (i = 0; i < ch->numslots; i++) {
ahci(4): Fix a typo in a source code comment - s/reqests/requests/ MFC after: 3 days
2023-01-11 09:46:58 +00:00
/* XXX: requests in loading state. */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (((err >> i) & 1) == 0)
continue;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (port >= 0 &&
ch->slot[i].ccb->ccb_h.target_id != port)
continue;
Revert interrupt reason check order back. ATAPI errors may set IF bit together with TFE.
2009-10-23 13:07:22 +00:00
if (istatus & AHCI_P_IX_TFE) {
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (port != -2) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Task File Error */
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (ch->numtslotspd[
ch->slot[i].ccb->ccb_h.target_id] == 0) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Untagged operation. */
if (i == ccs)
et = AHCI_ERR_TFE;
else
et = AHCI_ERR_INNOCENT;
} else {
/* Tagged operation. */
et = AHCI_ERR_NCQ;
}
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
} else {
et = AHCI_ERR_TFE;
ch->fatalerr = 1;
}
Revert interrupt reason check order back. ATAPI errors may set IF bit together with TFE.
2009-10-23 13:07:22 +00:00
} else if (istatus & AHCI_P_IX_IF) {
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (ch->numtslots == 0 && i != ccs && port != -2)
Revert interrupt reason check order back. ATAPI errors may set IF bit together with TFE.
2009-10-23 13:07:22 +00:00
et = AHCI_ERR_INNOCENT;
else
et = AHCI_ERR_SATA;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else
et = AHCI_ERR_INVALID;
ahci_end_transaction(&ch->slot[i], et);
}
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/*
* We can't reinit port if there are some other
* commands active, use resume to complete them.
*/
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (ch->rslots != 0 && !ch->recoverycmd)
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_FBS, AHCI_P_FBS_EN | AHCI_P_FBS_DEC);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
/* Process NOTIFY events */
To save small bit of CPU time, hide part of SNTF register read latency behind other reads.
2009-09-06 21:22:24 +00:00
if (sntf)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_notify_events(ch, sntf);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/* Must be called with channel locked. */
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_check_collision(struct ahci_channel *ch, union ccb *ccb)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
int t = ccb->ccb_h.target_id;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA)) {
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
/* Tagged command while we have no supported tag free. */
if (((~ch->oslots) & (0xffffffff >> (32 -
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->curr[t].tags))) == 0)
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
return (1);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* If we have FBS */
if (ch->fbs_enabled) {
/* Tagged command while untagged are active. */
if (ch->numrslotspd[t] != 0 && ch->numtslotspd[t] == 0)
return (1);
} else {
/* Tagged command while untagged are active. */
if (ch->numrslots != 0 && ch->numtslots == 0)
return (1);
/* Tagged command while tagged to other target is active. */
if (ch->numtslots != 0 &&
ch->taggedtarget != ccb->ccb_h.target_id)
return (1);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else {
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* If we have FBS */
if (ch->fbs_enabled) {
/* Untagged command while tagged are active. */
if (ch->numrslotspd[t] != 0 && ch->numtslotspd[t] != 0)
return (1);
} else {
/* Untagged command while tagged are active. */
if (ch->numrslots != 0 && ch->numtslots != 0)
return (1);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & (CAM_ATAIO_CONTROL | CAM_ATAIO_NEEDRESULT))) {
/* Atomic command while anything active. */
if (ch->numrslots != 0)
return (1);
}
/* We have some atomic command running. */
if (ch->aslots != 0)
return (1);
return (0);
}
/* Must be called with channel locked. */
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_begin_transaction(struct ahci_channel *ch, union ccb *ccb)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
struct ahci_slot *slot;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
int tag, tags;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Choose empty slot. */
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
tags = ch->numslots;
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA))
tags = ch->curr[ccb->ccb_h.target_id].tags;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ch->lastslot + 1 < tags)
tag = ffs(~(ch->oslots >> (ch->lastslot + 1)));
else
tag = 0;
if (tag == 0 || tag + ch->lastslot >= tags)
tag = ffs(~ch->oslots) - 1;
else
tag += ch->lastslot;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->lastslot = tag;
/* Occupy chosen slot. */
slot = &ch->slot[tag];
slot->ccb = ccb;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
/* Stop PM timer. */
if (ch->numrslots == 0 && ch->pm_level > 3)
callout_stop(&ch->pm_timer);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Update channel stats. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ch->oslots |= (1 << tag);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->numrslots++;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numrslotspd[ccb->ccb_h.target_id]++;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA)) {
ch->numtslots++;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numtslotspd[ccb->ccb_h.target_id]++;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->taggedtarget = ccb->ccb_h.target_id;
}
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & (CAM_ATAIO_CONTROL | CAM_ATAIO_NEEDRESULT)))
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ch->aslots |= (1 << tag);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
slot->state = AHCI_SLOT_LOADING;
Reform the busdma API so that new types may be added without modifying every architecture's busdma_machdep.c. It is done by unifying the bus_dmamap_load_buffer() routines so that they may be called from MI code. The MD busdma is then given a chance to do any final processing in the complete() callback. The cam changes unify the bus_dmamap_load* handling in cam drivers. The arm and mips implementations are updated to track virtual addresses for sync(). Previously this was done in a type specific way. Now it is done in a generic way by recording the list of virtuals in the map. Submitted by: jeff (sponsored by EMC/Isilon) Reviewed by: kan (previous version), scottl, mjacob (isp(4), no objections for target mode changes) Discussed with: ian (arm changes) Tested by: marius (sparc64), mips (jmallet), isci(4) on x86 (jharris), amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00
bus_dmamap_load_ccb(ch->dma.data_tag, slot->dma.data_map, ccb,
ahci_dmasetprd, slot, 0);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
} else {
slot->dma.nsegs = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_execute_transaction(slot);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/* Locked by busdma engine. */
static void
ahci_dmasetprd(void *arg, bus_dma_segment_t *segs, int nsegs, int error)
{
struct ahci_slot *slot = arg;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = slot->ch;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
struct ahci_cmd_tab *ctp;
struct ahci_dma_prd *prd;
int i;
if (error) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "DMA load error\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci_end_transaction(slot, AHCI_ERR_INVALID);
return;
}
KASSERT(nsegs <= AHCI_SG_ENTRIES, ("too many DMA segment entries\n"));
/* Get a piece of the workspace for this request */
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
ctp = (struct ahci_cmd_tab *)(ch->dma.work + slot->ct_offset);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Fill S/G table */
prd = &ctp->prd_tab[0];
for (i = 0; i < nsegs; i++) {
prd[i].dba = htole64(segs[i].ds_addr);
prd[i].dbc = htole32((segs[i].ds_len - 1) & AHCI_PRD_MASK);
}
slot->dma.nsegs = nsegs;
bus_dmamap_sync(ch->dma.data_tag, slot->dma.data_map,
((slot->ccb->ccb_h.flags & CAM_DIR_IN) ?
BUS_DMASYNC_PREREAD : BUS_DMASYNC_PREWRITE));
ahci_execute_transaction(slot);
}
/* Must be called with channel locked. */
static void
ahci_execute_transaction(struct ahci_slot *slot)
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = slot->ch;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
struct ahci_cmd_tab *ctp;
struct ahci_cmd_list *clp;
union ccb *ccb = slot->ccb;
int port = ccb->ccb_h.target_id & 0x0f;
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
int fis_size, i, softreset;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
uint8_t *fis = ch->dma.rfis + 0x40;
uint8_t val;
Fixed endianess issue in AHCI driver There were some bits that were being set in cmd_flags (a field of AHCI's command list structure) after cmd_flags was converted to little endian. On a big endian host, such as PowerPC, this would set the wrong bits. This was preventing AHCI driver from working on these hosts. Reviewed by: jhibbits Approved by: jhibbits (mentor)
2018-07-27 13:11:05 +00:00
uint16_t cmd_flags;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Get a piece of the workspace for this request */
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
ctp = (struct ahci_cmd_tab *)(ch->dma.work + slot->ct_offset);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Setup the FIS for this request */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (!(fis_size = ahci_setup_fis(ch, ctp, ccb, slot->slot))) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(ch->dev, "Setting up SATA FIS failed\n");
ahci_end_transaction(slot, AHCI_ERR_INVALID);
return;
}
/* Setup the command list entry */
clp = (struct ahci_cmd_list *)
(ch->dma.work + AHCI_CL_OFFSET + (AHCI_CL_SIZE * slot->slot));
Fixed endianess issue in AHCI driver There were some bits that were being set in cmd_flags (a field of AHCI's command list structure) after cmd_flags was converted to little endian. On a big endian host, such as PowerPC, this would set the wrong bits. This was preventing AHCI driver from working on these hosts. Reviewed by: jhibbits Approved by: jhibbits (mentor)
2018-07-27 13:11:05 +00:00
cmd_flags =
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
(ccb->ccb_h.flags & CAM_DIR_OUT ? AHCI_CMD_WRITE : 0) |
(ccb->ccb_h.func_code == XPT_SCSI_IO ?
(AHCI_CMD_ATAPI | AHCI_CMD_PREFETCH) : 0) |
(fis_size / sizeof(u_int32_t)) |
Fixed endianess issue in AHCI driver There were some bits that were being set in cmd_flags (a field of AHCI's command list structure) after cmd_flags was converted to little endian. On a big endian host, such as PowerPC, this would set the wrong bits. This was preventing AHCI driver from working on these hosts. Reviewed by: jhibbits Approved by: jhibbits (mentor)
2018-07-27 13:11:05 +00:00
(port << 12);
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
clp->prd_length = htole16(slot->dma.nsegs);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Special handling for Soft Reset command. */
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
(ccb->ataio.cmd.flags & CAM_ATAIO_CONTROL)) {
if (ccb->ataio.cmd.control & ATA_A_RESET) {
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
softreset = 1;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* Kick controller into sane state */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop(ch);
ahci_clo(ch);
ahci_start(ch, 0);
Fixed endianess issue in AHCI driver There were some bits that were being set in cmd_flags (a field of AHCI's command list structure) after cmd_flags was converted to little endian. On a big endian host, such as PowerPC, this would set the wrong bits. This was preventing AHCI driver from working on these hosts. Reviewed by: jhibbits Approved by: jhibbits (mentor)
2018-07-27 13:11:05 +00:00
cmd_flags |= AHCI_CMD_RESET | AHCI_CMD_CLR_BUSY;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
} else {
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
softreset = 2;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* Prepare FIS receive area for check. */
for (i = 0; i < 20; i++)
fis[i] = 0xff;
}
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
} else
softreset = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
clp->bytecount = 0;
Fixed endianess issue in AHCI driver There were some bits that were being set in cmd_flags (a field of AHCI's command list structure) after cmd_flags was converted to little endian. On a big endian host, such as PowerPC, this would set the wrong bits. This was preventing AHCI driver from working on these hosts. Reviewed by: jhibbits Approved by: jhibbits (mentor)
2018-07-27 13:11:05 +00:00
clp->cmd_flags = htole16(cmd_flags);
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
clp->cmd_table_phys = htole64(ch->dma.work_bus + slot->ct_offset);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_dmamap_sync(ch->dma.work_tag, ch->dma.work_map,
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
BUS_DMASYNC_PREREAD | BUS_DMASYNC_PREWRITE);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_dmamap_sync(ch->dma.rfis_tag, ch->dma.rfis_map,
BUS_DMASYNC_PREREAD);
/* Set ACTIVE bit for NCQ commands. */
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA)) {
ATA_OUTL(ch->r_mem, AHCI_P_SACT, 1 << slot->slot);
}
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* If FBS is enabled, set PMP port. */
if (ch->fbs_enabled) {
ATA_OUTL(ch->r_mem, AHCI_P_FBS, AHCI_P_FBS_EN |
(port << AHCI_P_FBS_DEV_SHIFT));
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Issue command to the controller. */
slot->state = AHCI_SLOT_RUNNING;
ch->rslots |= (1 << slot->slot);
ATA_OUTL(ch->r_mem, AHCI_P_CI, (1 << slot->slot));
/* Device reset commands doesn't interrupt. Poll them. */
if (ccb->ccb_h.func_code == XPT_ATA_IO &&
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
(ccb->ataio.cmd.command == ATA_DEVICE_RESET || softreset)) {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
int count, timeout = ccb->ccb_h.timeout * 100;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
enum ahci_err_type et = AHCI_ERR_NONE;
for (count = 0; count < timeout; count++) {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(10);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (!(ATA_INL(ch->r_mem, AHCI_P_CI) & (1 << slot->slot)))
break;
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
if ((ATA_INL(ch->r_mem, AHCI_P_TFD) & ATA_S_ERROR) &&
softreset != 1) {
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
#if 0
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
device_printf(ch->dev,
"Poll error on slot %d, TFD: %04x\n",
slot->slot, ATA_INL(ch->r_mem, AHCI_P_TFD));
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
#endif
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
et = AHCI_ERR_TFE;
break;
}
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
/* Workaround for ATI SB600/SB700 chipsets. */
if (ccb->ccb_h.target_id == 15 &&
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
(ch->quirks & AHCI_Q_ATI_PMP_BUG) &&
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
(ATA_INL(ch->r_mem, AHCI_P_IS) & AHCI_P_IX_IPM)) {
et = AHCI_ERR_TIMEOUT;
break;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
Enable optional soft reset in AHCI It occurred that some Marvell integrated controllers require additional time after soft reset to work properly. Introduce new quirk (AHCI_Q_MRVL_SR_DEL), that enable such operation. Submitted by: Konrad Adamczyk <ka@semihalf.com> Obtained from: Semihalf Sponsored by: Stormshield Reviewed by: mav Differential revision: https://reviews.freebsd.org/D9221
2017-01-25 10:32:57 +00:00
/*
* Some Marvell controllers require additional time
* after soft reset to work properly. Setup delay
* to 50ms after soft reset.
*/
if (ch->quirks & AHCI_Q_MRVL_SR_DEL)
DELAY(50000);
Invert AHCI_Q_NOBSYRES quirk meaning, waiting for readiness by default. I gave up to update list of Marvell chips that require this quirk. The final nail was growing number of PCIe/M.2 SSDs where Marvell chips have PCI IDs of different vendors. MFC after: 1 week H/W donated by: I/O Switch
2014-09-05 13:45:14 +00:00
/*
* Marvell HBAs with non-RAID firmware do not wait for
* readiness after soft reset, so we have to wait here.
Fix typo in comments. Submitted by: Benedict Reuschling <bcr@FreeBSD.org> MFC after: 6 days
2014-09-06 15:37:55 +00:00
* Marvell RAIDs do not have this problem, but instead
* sometimes forget to update FIS receive area, breaking
* this wait.
Invert AHCI_Q_NOBSYRES quirk meaning, waiting for readiness by default. I gave up to update list of Marvell chips that require this quirk. The final nail was growing number of PCIe/M.2 SSDs where Marvell chips have PCI IDs of different vendors. MFC after: 1 week H/W donated by: I/O Switch
2014-09-05 13:45:14 +00:00
*/
if ((ch->quirks & AHCI_Q_NOBSYRES) == 0 &&
(ch->quirks & AHCI_Q_ATI_PMP_BUG) == 0 &&
softreset == 2 && et == AHCI_ERR_NONE) {
AHCI: Use bus_dmamap_sync(9) when accessing DMA buffers. Reviewed by: mav Approved by: kib (mentor) Differential Revision: https://reviews.freebsd.org/D4240
2015-11-28 14:30:42 +00:00
for ( ; count < timeout; count++) {
bus_dmamap_sync(ch->dma.rfis_tag,
ch->dma.rfis_map, BUS_DMASYNC_POSTREAD);
val = fis[2];
bus_dmamap_sync(ch->dma.rfis_tag,
ch->dma.rfis_map, BUS_DMASYNC_PREREAD);
if ((val & ATA_S_BUSY) == 0)
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
break;
AHCI: Use bus_dmamap_sync(9) when accessing DMA buffers. Reviewed by: mav Approved by: kib (mentor) Differential Revision: https://reviews.freebsd.org/D4240
2015-11-28 14:30:42 +00:00
DELAY(10);
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
}
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (timeout && (count >= timeout)) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "Poll timeout on slot %d port %d\n",
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
slot->slot, port);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "is %08x cs %08x ss %08x "
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
"rs %08x tfd %02x serr %08x cmd %08x\n",
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
ATA_INL(ch->r_mem, AHCI_P_IS),
ATA_INL(ch->r_mem, AHCI_P_CI),
ATA_INL(ch->r_mem, AHCI_P_SACT), ch->rslots,
ATA_INL(ch->r_mem, AHCI_P_TFD),
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
ATA_INL(ch->r_mem, AHCI_P_SERR),
ATA_INL(ch->r_mem, AHCI_P_CMD));
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
et = AHCI_ERR_TIMEOUT;
}
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* Kick controller into sane state and enable FBS. */
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
if (softreset == 2)
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
ch->eslots |= (1 << slot->slot);
ahci_end_transaction(slot, et);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return;
}
/* Start command execution timeout */
Prevent overflow issues in timeout processing Previously, any timeout value for which (timeout * hz) will overflow the signed integer, will give weird results, since callout(9) routines will convert negative values of ticks to '1'. For unsigned integer overflow we will get sufficiently smaller timeout values than expected. Switch from callout_reset, which requires conversion to int based ticks to callout_reset_sbt to avoid this. Also correct isci to correctly resolve ccb timeout. This was based on the original work done by Eygene Ryabinkin <rea@freebsd.org> back in 5 Aug 2011 which used a macro to help avoid the overlow. Differential Revision: https://reviews.freebsd.org/D1157 Reviewed by: mav, davide MFC after: 1 month Sponsored by: Multiplay
2014-11-21 21:01:24 +00:00
callout_reset_sbt(&slot->timeout, SBT_1MS * ccb->ccb_h.timeout / 2,
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
0, ahci_timeout, slot, 0);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return;
}
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
/* Must be called with channel locked. */
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_timeout(struct ahci_channel *ch)
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
{
int i;
mtx_assert(&ch->mtx, MA_OWNED);
/* Handle the rest of commands. */
for (i = 0; i < ch->numslots; i++) {
/* Do we have a running request on slot? */
if (ch->slot[i].state < AHCI_SLOT_RUNNING)
continue;
ahci_end_transaction(&ch->slot[i], AHCI_ERR_TIMEOUT);
}
}
/* Must be called with channel locked. */
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_rearm_timeout(struct ahci_channel *ch)
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
{
int i;
mtx_assert(&ch->mtx, MA_OWNED);
for (i = 0; i < ch->numslots; i++) {
struct ahci_slot *slot = &ch->slot[i];
/* Do we have a running request on slot? */
if (slot->state < AHCI_SLOT_RUNNING)
continue;
if ((ch->toslots & (1 << i)) == 0)
continue;
Prevent overflow issues in timeout processing Previously, any timeout value for which (timeout * hz) will overflow the signed integer, will give weird results, since callout(9) routines will convert negative values of ticks to '1'. For unsigned integer overflow we will get sufficiently smaller timeout values than expected. Switch from callout_reset, which requires conversion to int based ticks to callout_reset_sbt to avoid this. Also correct isci to correctly resolve ccb timeout. This was based on the original work done by Eygene Ryabinkin <rea@freebsd.org> back in 5 Aug 2011 which used a macro to help avoid the overlow. Differential Revision: https://reviews.freebsd.org/D1157 Reviewed by: mav, davide MFC after: 1 month Sponsored by: Multiplay
2014-11-21 21:01:24 +00:00
callout_reset_sbt(&slot->timeout,
SBT_1MS * slot->ccb->ccb_h.timeout / 2, 0,
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
ahci_timeout, slot, 0);
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
}
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Locked by callout mechanism. */
static void
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
ahci_timeout(void *arg)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
struct ahci_slot *slot = arg;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = slot->ch;
device_t dev = ch->dev;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
uint32_t sstatus;
int ccs;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
int i;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
/* Check for stale timeout. */
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (slot->state < AHCI_SLOT_RUNNING)
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
return;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
/* Check if slot was not being executed last time we checked. */
if (slot->state < AHCI_SLOT_EXECUTING) {
/* Check if slot started executing. */
sstatus = ATA_INL(ch->r_mem, AHCI_P_SACT);
ccs = (ATA_INL(ch->r_mem, AHCI_P_CMD) & AHCI_P_CMD_CCS_MASK)
>> AHCI_P_CMD_CCS_SHIFT;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if ((sstatus & (1 << slot->slot)) != 0 || ccs == slot->slot ||
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
ch->fbs_enabled || ch->wrongccs)
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
slot->state = AHCI_SLOT_EXECUTING;
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
else if ((ch->rslots & (1 << ccs)) == 0) {
ch->wrongccs = 1;
slot->state = AHCI_SLOT_EXECUTING;
}
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
Prevent overflow issues in timeout processing Previously, any timeout value for which (timeout * hz) will overflow the signed integer, will give weird results, since callout(9) routines will convert negative values of ticks to '1'. For unsigned integer overflow we will get sufficiently smaller timeout values than expected. Switch from callout_reset, which requires conversion to int based ticks to callout_reset_sbt to avoid this. Also correct isci to correctly resolve ccb timeout. This was based on the original work done by Eygene Ryabinkin <rea@freebsd.org> back in 5 Aug 2011 which used a macro to help avoid the overlow. Differential Revision: https://reviews.freebsd.org/D1157 Reviewed by: mav, davide MFC after: 1 month Sponsored by: Multiplay
2014-11-21 21:01:24 +00:00
callout_reset_sbt(&slot->timeout,
SBT_1MS * slot->ccb->ccb_h.timeout / 2, 0,
Use a void * argument to callout handlers instead of timeout_t casts. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D22684
2019-12-05 18:47:29 +00:00
ahci_timeout, slot, 0);
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
return;
}
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
device_printf(dev, "Timeout on slot %d port %d\n",
slot->slot, slot->ccb->ccb_h.target_id & 0x0f);
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
device_printf(dev, "is %08x cs %08x ss %08x rs %08x tfd %02x "
"serr %08x cmd %08x\n",
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
ATA_INL(ch->r_mem, AHCI_P_IS), ATA_INL(ch->r_mem, AHCI_P_CI),
ATA_INL(ch->r_mem, AHCI_P_SACT), ch->rslots,
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
ATA_INL(ch->r_mem, AHCI_P_TFD), ATA_INL(ch->r_mem, AHCI_P_SERR),
ATA_INL(ch->r_mem, AHCI_P_CMD));
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
On command timeout handle frozen command first, to not run it inside XXX_end_transaction(). Submitted by: avg
2009-10-07 15:40:04 +00:00
/* Handle frozen command. */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ch->frozen) {
union ccb *fccb = ch->frozen;
ch->frozen = NULL;
fccb->ccb_h.status = CAM_REQUEUE_REQ | CAM_RELEASE_SIMQ;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (!(fccb->ccb_h.status & CAM_DEV_QFRZN)) {
xpt_freeze_devq(fccb->ccb_h.path, 1);
fccb->ccb_h.status |= CAM_DEV_QFRZN;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, fccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
if (!ch->fbs_enabled && !ch->wrongccs) {
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
/* Without FBS we know real timeout source. */
ch->fatalerr = 1;
/* Handle command with timeout. */
ahci_end_transaction(&ch->slot[slot->slot], AHCI_ERR_TIMEOUT);
/* Handle the rest of commands. */
for (i = 0; i < ch->numslots; i++) {
/* Do we have a running request on slot? */
if (ch->slot[i].state < AHCI_SLOT_RUNNING)
continue;
ahci_end_transaction(&ch->slot[i], AHCI_ERR_INNOCENT);
}
} else {
/* With FBS we wait for other commands timeout and pray. */
if (ch->toslots == 0)
xpt_freeze_simq(ch->sim, 1);
ch->toslots |= (1 << slot->slot);
if ((ch->rslots & ~ch->toslots) == 0)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_timeout(ch);
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
else
device_printf(dev, " ... waiting for slots %08x\n",
ch->rslots & ~ch->toslots);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
}
/* Must be called with channel locked. */
static void
ahci_end_transaction(struct ahci_slot *slot, enum ahci_err_type et)
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = slot->ch;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
union ccb *ccb = slot->ccb;
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
struct ahci_cmd_list *clp;
Fix panic, when due to some kind of congestion on FIS-based switching port multiplier some command triggers false positive timeout, but then completes normally. MFC after: 2 weeks
2010-09-16 12:39:50 +00:00
int lastto;
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
uint32_t sig;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
bus_dmamap_sync(ch->dma.work_tag, ch->dma.work_map,
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
BUS_DMASYNC_POSTREAD | BUS_DMASYNC_POSTWRITE);
clp = (struct ahci_cmd_list *)
(ch->dma.work + AHCI_CL_OFFSET + (AHCI_CL_SIZE * slot->slot));
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Read result registers to the result struct
* May be incorrect if several commands finished same time,
* so read only when sure or have to.
*/
if (ccb->ccb_h.func_code == XPT_ATA_IO) {
struct ata_res *res = &ccb->ataio.res;
if ((et == AHCI_ERR_TFE) ||
(ccb->ataio.cmd.flags & CAM_ATAIO_NEEDRESULT)) {
u_int8_t *fis = ch->dma.rfis + 0x40;
bus_dmamap_sync(ch->dma.rfis_tag, ch->dma.rfis_map,
BUS_DMASYNC_POSTREAD);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (ch->fbs_enabled) {
fis += ccb->ccb_h.target_id * 256;
res->status = fis[2];
res->error = fis[3];
} else {
uint16_t tfd = ATA_INL(ch->r_mem, AHCI_P_TFD);
res->status = tfd;
res->error = tfd >> 8;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
res->lba_low = fis[4];
res->lba_mid = fis[5];
res->lba_high = fis[6];
res->device = fis[7];
res->lba_low_exp = fis[8];
res->lba_mid_exp = fis[9];
res->lba_high_exp = fis[10];
res->sector_count = fis[12];
res->sector_count_exp = fis[13];
Marvell 88SE91xx controllers are known to report soft-reset completion without waiting for device readiness (or at least not updating FIS receive area in time). To workaround that, special quirk was added earlier to wait for the FIS receive area update. But it was found that under same PCI ID 0x91231b4b and revision 0x11 there are two completely different chip versions (firmware?): HBA and RAID. The problem is that RAID version in some cases, such as hot-plug, does not update FIS receive area at all! To workaround that, differentiate the chip versions by their capabilities, and, if RAID version found, skip FIS receive area update waiting and read device signature from the PxSIG register instead. This method doesn't work for HBA version when PMP attached, so keep using previous workaround there.
2011-05-26 09:23:01 +00:00
/*
* Some weird controllers do not return signature in
* FIS receive area. Read it from PxSIG register.
*/
if ((ch->quirks & AHCI_Q_ALTSIG) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_CONTROL) &&
(ccb->ataio.cmd.control & ATA_A_RESET) == 0) {
sig = ATA_INL(ch->r_mem, AHCI_P_SIG);
res->lba_high = sig >> 24;
res->lba_mid = sig >> 16;
res->lba_low = sig >> 8;
res->sector_count = sig;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else
bzero(res, sizeof(*res));
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
if ((ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA) == 0 &&
Disable NCQ for multiport Marvell 88SX61XX SATA controllers. Simultaneous active I/O to several disks (copying large file on ZFS) causes timeout after just a few seconds of run. Single port 88SX6111 seems like not affected. Skip reading transferred bytes count for these controllers. It works for 88SX6111, but 88SX6145 always returns zero there. Haven't tested others, but better to be safe.
2011-02-12 07:06:40 +00:00
(ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE &&
(ch->quirks & AHCI_Q_NOCOUNT) == 0) {
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
ccb->ataio.resid =
ccb->ataio.dxfer_len - le32toh(clp->bytecount);
}
} else {
Disable NCQ for multiport Marvell 88SX61XX SATA controllers. Simultaneous active I/O to several disks (copying large file on ZFS) causes timeout after just a few seconds of run. Single port 88SX6111 seems like not affected. Skip reading transferred bytes count for these controllers. It works for 88SX6111, but 88SX6145 always returns zero there. Haven't tested others, but better to be safe.
2011-02-12 07:06:40 +00:00
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE &&
(ch->quirks & AHCI_Q_NOCOUNT) == 0) {
Teach ahci(4), siis(4) and ATA_CAM ata(4) wrapper report to CAM residual I/O length on underruns, that often happens for some SCSI commands.
2010-11-08 15:36:15 +00:00
ccb->csio.resid =
ccb->csio.dxfer_len - le32toh(clp->bytecount);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
bus_dmamap_sync(ch->dma.data_tag, slot->dma.data_map,
(ccb->ccb_h.flags & CAM_DIR_IN) ?
BUS_DMASYNC_POSTREAD : BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ch->dma.data_tag, slot->dma.data_map);
}
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
if (et != AHCI_ERR_NONE)
ch->eslots |= (1 << slot->slot);
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
/* In case of error, freeze device for proper recovery. */
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if ((et != AHCI_ERR_NONE) && (!ch->recoverycmd) &&
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
!(ccb->ccb_h.status & CAM_DEV_QFRZN)) {
xpt_freeze_devq(ccb->ccb_h.path, 1);
ccb->ccb_h.status |= CAM_DEV_QFRZN;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Set proper result status. */
ccb->ccb_h.status &= ~CAM_STATUS_MASK;
switch (et) {
case AHCI_ERR_NONE:
ccb->ccb_h.status |= CAM_REQ_CMP;
if (ccb->ccb_h.func_code == XPT_SCSI_IO)
ccb->csio.scsi_status = SCSI_STATUS_OK;
break;
case AHCI_ERR_INVALID:
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
ch->fatalerr = 1;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status |= CAM_REQ_INVALID;
break;
case AHCI_ERR_INNOCENT:
ccb->ccb_h.status |= CAM_REQUEUE_REQ;
break;
case AHCI_ERR_TFE:
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
case AHCI_ERR_NCQ:
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ccb->ccb_h.func_code == XPT_SCSI_IO) {
ccb->ccb_h.status |= CAM_SCSI_STATUS_ERROR;
ccb->csio.scsi_status = SCSI_STATUS_CHECK_COND;
} else {
ccb->ccb_h.status |= CAM_ATA_STATUS_ERROR;
}
break;
case AHCI_ERR_SATA:
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
ch->fatalerr = 1;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (!ch->recoverycmd) {
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
xpt_freeze_simq(ch->sim, 1);
ccb->ccb_h.status &= ~CAM_STATUS_MASK;
ccb->ccb_h.status |= CAM_RELEASE_SIMQ;
}
ccb->ccb_h.status |= CAM_UNCOR_PARITY;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
break;
case AHCI_ERR_TIMEOUT:
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (!ch->recoverycmd) {
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
xpt_freeze_simq(ch->sim, 1);
ccb->ccb_h.status &= ~CAM_STATUS_MASK;
ccb->ccb_h.status |= CAM_RELEASE_SIMQ;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status |= CAM_CMD_TIMEOUT;
break;
default:
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
ch->fatalerr = 1;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status |= CAM_REQ_CMP_ERR;
}
/* Free slot. */
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
ch->oslots &= ~(1 << slot->slot);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->rslots &= ~(1 << slot->slot);
ch->aslots &= ~(1 << slot->slot);
slot->state = AHCI_SLOT_EMPTY;
slot->ccb = NULL;
/* Update channel stats. */
ch->numrslots--;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numrslotspd[ccb->ccb_h.target_id]--;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA)) {
ch->numtslots--;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numtslotspd[ccb->ccb_h.target_id]--;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Fix panic, when due to some kind of congestion on FIS-based switching port multiplier some command triggers false positive timeout, but then completes normally. MFC after: 2 weeks
2010-09-16 12:39:50 +00:00
/* Cancel timeout state if request completed normally. */
if (et != AHCI_ERR_TIMEOUT) {
lastto = (ch->toslots == (1 << slot->slot));
ch->toslots &= ~(1 << slot->slot);
if (lastto)
xpt_release_simq(ch->sim, TRUE);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* If it was first request of reset sequence and there is no error,
* proceed to second request. */
if ((ccb->ccb_h.func_code == XPT_ATA_IO) &&
(ccb->ataio.cmd.flags & CAM_ATAIO_CONTROL) &&
(ccb->ataio.cmd.control & ATA_A_RESET) &&
et == AHCI_ERR_NONE) {
ccb->ataio.cmd.control &= ~ATA_A_RESET;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_begin_transaction(ch, ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return;
}
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
/* If it was our READ LOG command - process it. */
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (ccb->ccb_h.recovery_type == RECOVERY_READ_LOG) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_read_log(ch, ccb);
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
/* If it was our REQUEST SENSE command - process it. */
} else if (ccb->ccb_h.recovery_type == RECOVERY_REQUEST_SENSE) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_request_sense(ch, ccb);
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
/* If it was NCQ or ATAPI command error, put result on hold. */
} else if (et == AHCI_ERR_NCQ ||
((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_SCSI_STATUS_ERROR &&
(ccb->ccb_h.flags & CAM_DIS_AUTOSENSE) == 0)) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->hold[slot->slot] = ccb;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numhslots++;
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
} else
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ccb);
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
/* If we have no other active commands, ... */
if (ch->rslots == 0) {
/* if there was fatal error - reset port. */
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
if (ch->toslots != 0 || ch->fatalerr) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
} else {
/* if we have slots in error, we can reinit port. */
if (ch->eslots != 0) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop(ch);
ahci_clo(ch);
ahci_start(ch, 1);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
}
/* if there commands on hold, we can do READ LOG. */
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (!ch->recoverycmd && ch->numhslots)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_issue_recovery(ch);
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
}
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
/* If all the rest of commands are in timeout - give them chance. */
} else if ((ch->rslots & ~ch->toslots) == 0 &&
et != AHCI_ERR_TIMEOUT)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_rearm_timeout(ch);
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
/* Unfreeze frozen command. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ch->frozen && !ahci_check_collision(ch, ch->frozen)) {
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
union ccb *fccb = ch->frozen;
ch->frozen = NULL;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_begin_transaction(ch, fccb);
According to SATA specification, when Serial ATA Enclosure Management Bridge (SEMB) is unable to communicate to Storage Enclosure Processor (SEP), in response to hard and soft resets it should among other things return value 0x7F in Status register. The weird side is that it means DRQ bit set, which tells that reset request is not completed. It would be fine if SEMB was the only device on port. But if SEMB connected to PMP or built into it, it may block access to other devices sharing same SATA port. Make some tunings/fixes to soft-reset handling to workaround the issue: - ahci(4): request CLO on the port after soft reset to ignore DRQ bit; - siis(4): gracefully reinitialize port after soft reset timeout (hardware doesn't detect reset request completion in this case); - mvs(4): if PMP is used, send dummy soft-reset to the PMP port to make it clear DRQ bit for us. For now this makes quirks in ata_pmp.c, hiding SEMB ports of SiI3726/SiI4726 PMPs, less important. Further, if hardware permit, I hope to implement real SEMB support.
2011-05-25 13:55:49 +00:00
xpt_release_simq(ch->sim, TRUE);
}
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
/* Start PM timer. */
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
if (ch->numrslots == 0 && ch->pm_level > 3 &&
(ch->curr[ch->pm_present ? 15 : 0].caps & CTS_SATA_CAPS_D_PMREQ)) {
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
callout_schedule(&ch->pm_timer,
(ch->pm_level == 4) ? hz / 1000 : hz / 8);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_issue_recovery(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
union ccb *ccb;
struct ccb_ataio *ataio;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
struct ccb_scsiio *csio;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
int i;
Fix some English grammar.
2011-04-19 10:57:40 +00:00
/* Find some held command. */
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
for (i = 0; i < ch->numslots; i++) {
if (ch->hold[i])
break;
}
ccb = xpt_alloc_ccb_nowait();
if (ccb == NULL) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "Unable to allocate recovery command\n");
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
completeall:
Fix some English grammar.
2011-04-19 10:57:40 +00:00
/* We can't do anything -- complete held commands. */
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
for (i = 0; i < ch->numslots; i++) {
if (ch->hold[i] == NULL)
continue;
ch->hold[i]->ccb_h.status &= ~CAM_STATUS_MASK;
ch->hold[i]->ccb_h.status |= CAM_RESRC_UNAVAIL;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ch->hold[i]);
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
ch->hold[i] = NULL;
ch->numhslots--;
}
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
return;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
ahci/siis/mvs: Fix panics after 3394d4239b. Full CCB header overwrites made frees go into wrong zones, causing kernel panics. Instead of copying full header use xpt_setup_ccb(), since the only field I see used from all the header is target_id. PR: 262263
2022-03-05 01:49:05 +00:00
xpt_setup_ccb(&ccb->ccb_h, ch->hold[i]->ccb_h.path,
ch->hold[i]->ccb_h.pinfo.priority);
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (ccb->ccb_h.func_code == XPT_ATA_IO) {
/* READ LOG */
ccb->ccb_h.recovery_type = RECOVERY_READ_LOG;
ccb->ccb_h.func_code = XPT_ATA_IO;
ccb->ccb_h.flags = CAM_DIR_IN;
ccb->ccb_h.timeout = 1000; /* 1s should be enough. */
ataio = &ccb->ataio;
ataio->data_ptr = malloc(512, M_AHCI, M_NOWAIT);
if (ataio->data_ptr == NULL) {
xpt_free_ccb(ccb);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev,
Fix some English grammar.
2011-04-19 10:57:40 +00:00
"Unable to allocate memory for READ LOG command\n");
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
goto completeall;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
}
ataio->dxfer_len = 512;
bzero(&ataio->cmd, sizeof(ataio->cmd));
ataio->cmd.flags = CAM_ATAIO_48BIT;
ataio->cmd.command = 0x2F; /* READ LOG EXT */
ataio->cmd.sector_count = 1;
ataio->cmd.sector_count_exp = 0;
ataio->cmd.lba_low = 0x10;
ataio->cmd.lba_mid = 0;
ataio->cmd.lba_mid_exp = 0;
} else {
/* REQUEST SENSE */
ccb->ccb_h.recovery_type = RECOVERY_REQUEST_SENSE;
ccb->ccb_h.recovery_slot = i;
ccb->ccb_h.func_code = XPT_SCSI_IO;
ccb->ccb_h.flags = CAM_DIR_IN;
ccb->ccb_h.status = 0;
ccb->ccb_h.timeout = 1000; /* 1s should be enough. */
csio = &ccb->csio;
csio->data_ptr = (void *)&ch->hold[i]->csio.sense_data;
csio->dxfer_len = ch->hold[i]->csio.sense_len;
csio->cdb_len = 6;
bzero(&csio->cdb_io, sizeof(csio->cdb_io));
csio->cdb_io.cdb_bytes[0] = 0x03;
csio->cdb_io.cdb_bytes[4] = csio->dxfer_len;
}
/* Freeze SIM while doing recovery. */
Properly handle memory allocation errors during error recovery.
2011-04-19 08:01:17 +00:00
ch->recoverycmd = 1;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
xpt_freeze_simq(ch->sim, 1);
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_begin_transaction(ch, ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_read_log(struct ahci_channel *ch, union ccb *ccb)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
uint8_t *data;
struct ata_res *res;
int i;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
ch->recoverycmd = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
data = ccb->ataio.data_ptr;
if ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_CMP &&
(data[0] & 0x80) == 0) {
for (i = 0; i < ch->numslots; i++) {
if (!ch->hold[i])
continue;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (ch->hold[i]->ccb_h.func_code != XPT_ATA_IO)
continue;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if ((data[0] & 0x1F) == i) {
res = &ch->hold[i]->ataio.res;
res->status = data[2];
res->error = data[3];
res->lba_low = data[4];
res->lba_mid = data[5];
res->lba_high = data[6];
res->device = data[7];
res->lba_low_exp = data[8];
res->lba_mid_exp = data[9];
res->lba_high_exp = data[10];
res->sector_count = data[12];
res->sector_count_exp = data[13];
} else {
ch->hold[i]->ccb_h.status &= ~CAM_STATUS_MASK;
ch->hold[i]->ccb_h.status |= CAM_REQUEUE_REQ;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ch->hold[i]);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->hold[i] = NULL;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numhslots--;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
} else {
if ((ccb->ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "Error while READ LOG EXT\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
else if ((data[0] & 0x80) == 0) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "Non-queued command error in READ LOG EXT\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
for (i = 0; i < ch->numslots; i++) {
if (!ch->hold[i])
continue;
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
if (ch->hold[i]->ccb_h.func_code != XPT_ATA_IO)
continue;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ch->hold[i]);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->hold[i] = NULL;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numhslots--;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
}
free(ccb->ataio.data_ptr, M_AHCI);
xpt_free_ccb(ccb);
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
xpt_release_simq(ch->sim, TRUE);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_process_request_sense(struct ahci_channel *ch, union ccb *ccb)
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
{
int i;
ch->recoverycmd = 0;
i = ccb->ccb_h.recovery_slot;
if ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_CMP) {
ch->hold[i]->ccb_h.status |= CAM_AUTOSNS_VALID;
} else {
ch->hold[i]->ccb_h.status &= ~CAM_STATUS_MASK;
ch->hold[i]->ccb_h.status |= CAM_AUTOSENSE_FAIL;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ch->hold[i]);
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
ch->hold[i] = NULL;
ch->numhslots--;
xpt_free_ccb(ccb);
xpt_release_simq(ch->sim, TRUE);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_start(struct ahci_channel *ch, int fbs)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int32_t cmd;
Add AHCI attachment code for Allwinner A10/A20 SoCs. The Allwinner SoC has an AHCI device on its internal main bus rather than the PCI bus. This SoC is somewhat underdocumented, and its SATA controller is no exception. The methods to support this chip were harvested from the Linux Allwinner SDK, and then constants invented to describe what's going on based on low-level constants contained in the SATA standard and guess work. This SoC requires a specific AHCI channel setup in order to start the operations on the channel properly. Clock setup and AHCI channel setup idea came from NetBSD. Tested on Cubieboard 2 and Banana pi (and attachment on Cubieboard by Pratik Singhal). Differential Revision: https://reviews.freebsd.org/D737 Submitted by: imp Reviewed by: imp, ganbold, mav, andrew
2015-07-03 14:11:01 +00:00
/* Run the channel start callback, if any. */
if (ch->start)
ch->start(ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Clear SATA error register */
ATA_OUTL(ch->r_mem, AHCI_P_SERR, 0xFFFFFFFF);
/* Clear any interrupts pending on this channel */
ATA_OUTL(ch->r_mem, AHCI_P_IS, 0xFFFFFFFF);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
/* Configure FIS-based switching if supported. */
if (ch->chcaps & AHCI_P_CMD_FBSCP) {
ch->fbs_enabled = (fbs && ch->pm_present) ? 1 : 0;
ATA_OUTL(ch->r_mem, AHCI_P_FBS,
ch->fbs_enabled ? AHCI_P_FBS_EN : 0);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Start operations on this channel */
cmd = ATA_INL(ch->r_mem, AHCI_P_CMD);
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
cmd &= ~AHCI_P_CMD_PMA;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_CMD, cmd | AHCI_P_CMD_ST |
(ch->pm_present ? AHCI_P_CMD_PMA : 0));
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int32_t cmd;
int timeout;
/* Kill all activity on this channel */
cmd = ATA_INL(ch->r_mem, AHCI_P_CMD);
ATA_OUTL(ch->r_mem, AHCI_P_CMD, cmd & ~AHCI_P_CMD_ST);
/* Wait for activity stop. */
timeout = 0;
do {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(10);
if (timeout++ > 50000) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "stopping AHCI engine failed\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
break;
}
} while (ATA_INL(ch->r_mem, AHCI_P_CMD) & AHCI_P_CMD_CR);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->eslots = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_clo(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int32_t cmd;
int timeout;
/* Issue Command List Override if supported */
if (ch->caps & AHCI_CAP_SCLO) {
cmd = ATA_INL(ch->r_mem, AHCI_P_CMD);
cmd |= AHCI_P_CMD_CLO;
ATA_OUTL(ch->r_mem, AHCI_P_CMD, cmd);
timeout = 0;
do {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(10);
if (timeout++ > 50000) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "executing CLO failed\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
break;
}
} while (ATA_INL(ch->r_mem, AHCI_P_CMD) & AHCI_P_CMD_CLO);
}
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop_fr(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int32_t cmd;
int timeout;
/* Kill all FIS reception on this channel */
cmd = ATA_INL(ch->r_mem, AHCI_P_CMD);
ATA_OUTL(ch->r_mem, AHCI_P_CMD, cmd & ~AHCI_P_CMD_FRE);
/* Wait for FIS reception stop. */
timeout = 0;
do {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(10);
if (timeout++ > 50000) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "stopping AHCI FR engine failed\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
break;
}
} while (ATA_INL(ch->r_mem, AHCI_P_CMD) & AHCI_P_CMD_FR);
}
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_start_fr(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int32_t cmd;
/* Start FIS reception on this channel */
cmd = ATA_INL(ch->r_mem, AHCI_P_CMD);
ATA_OUTL(ch->r_mem, AHCI_P_CMD, cmd | AHCI_P_CMD_FRE);
}
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_wait_ready(struct ahci_channel *ch, int t, int t0)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
int timeout = 0;
uint32_t val;
while ((val = ATA_INL(ch->r_mem, AHCI_P_TFD)) &
(ATA_S_BUSY | ATA_S_DRQ)) {
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
if (timeout > t) {
if (t != 0) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev,
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
"AHCI reset: device not ready after %dms "
"(tfd = %08x)\n",
MAX(t, 0) + t0, val);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (EBUSY);
}
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
DELAY(1000);
timeout++;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (bootverbose)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "AHCI reset: device ready after %dms\n",
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
timeout + t0);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return (0);
}
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
static void
ahci_reset_to(void *arg)
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_channel *ch = arg;
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
if (ch->resetting == 0)
return;
ch->resetting--;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_wait_ready(ch, ch->resetting == 0 ? -1 : 0,
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
(310 - ch->resetting) * 100) == 0) {
ch->resetting = 0;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_start(ch, 1);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
xpt_release_simq(ch->sim, TRUE);
return;
}
if (ch->resetting == 0) {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_clo(ch);
ahci_start(ch, 1);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
xpt_release_simq(ch->sim, TRUE);
return;
}
callout_schedule(&ch->reset_timer, hz / 10);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
struct ahci_controller *ctlr = device_get_softc(device_get_parent(ch->dev));
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
int i;
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
xpt_freeze_simq(ch->sim, 1);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (bootverbose)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "AHCI reset...\n");
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
/* Forget about previous reset. */
if (ch->resetting) {
ch->resetting = 0;
callout_stop(&ch->reset_timer);
xpt_release_simq(ch->sim, TRUE);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Requeue freezed command. */
if (ch->frozen) {
union ccb *fccb = ch->frozen;
ch->frozen = NULL;
fccb->ccb_h.status = CAM_REQUEUE_REQ | CAM_RELEASE_SIMQ;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
if (!(fccb->ccb_h.status & CAM_DEV_QFRZN)) {
xpt_freeze_devq(fccb->ccb_h.path, 1);
fccb->ccb_h.status |= CAM_DEV_QFRZN;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, fccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/* Kill the engine and requeue all running commands. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_stop(ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
for (i = 0; i < ch->numslots; i++) {
/* Do we have a running request on slot? */
if (ch->slot[i].state < AHCI_SLOT_RUNNING)
continue;
/* XXX; Commands in loading state. */
ahci_end_transaction(&ch->slot[i], AHCI_ERR_INNOCENT);
}
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
for (i = 0; i < ch->numslots; i++) {
if (!ch->hold[i])
continue;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ch->hold[i]);
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
ch->hold[i] = NULL;
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->numhslots--;
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
}
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
if (ch->toslots != 0)
xpt_release_simq(ch->sim, TRUE);
Add FIS-based switching support. If controller supports FBS, it allows several devices beyond Port Multiplier to work simultaneously, substantially increasing performance.
2010-01-28 17:54:47 +00:00
ch->eslots = 0;
MFp4: With FBS enabled, we have no idea what command caused timeout. Implement same logic as in siis(4) - wait for other commands complete or timeout and then give some more time.
2010-02-14 12:30:35 +00:00
ch->toslots = 0;
In some cases, at least on Marvell 88SE912x controllers, Current Command Slot field of the PxCMD register may point to an empty command slot. That breaks command timeout detection logic, making impossible to find what command actually caused timeout, and leading to infinite wait. Workaround that by checking whether pointed command slot is really used and can timeout in its time. And if not, fallback to the dumb algorithm used with FBS -- let all commands to time out and then fail all of them. Approved by: re (kib) MFC after: 1 week
2011-07-29 20:35:23 +00:00
ch->wrongccs = 0;
MFp4: - Handle timeouts and fatal errors with port hard-reset. The rest of recovery will be done by XPT on receiving async event. More gracefull per-device soft-reset recovery can be implemented later. - Add workaround for ATI SB600/SB700 PMP probe related bug, to speedup boot.
2009-11-03 11:47:07 +00:00
ch->fatalerr = 0;
MFp4: On error, freeze device queue, to allow periph driver to do proper recovery. Freeze SIM queue only in some cases, when it is needed to protect SIM. Implement better command timeout detection logic for non-queued commands. This fixes false positives when command with short timeout waiting for the long one. For example, when hald tastes CD during burning process. Read and clear SERR register on interrupt.
2009-10-21 12:42:25 +00:00
/* Tell the XPT about the event */
xpt_async(AC_BUS_RESET, ch->path, NULL);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Disable port interrupts */
ATA_OUTL(ch->r_mem, AHCI_P_IE, 0);
/* Reset and reconnect PHY, */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (!ahci_sata_phy_reset(ch)) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (bootverbose)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev,
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
"AHCI reset: device not found\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->devices = 0;
/* Enable wanted port interrupts */
ATA_OUTL(ch->r_mem, AHCI_P_IE,
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
(((ch->pm_level != 0) ? AHCI_P_IX_CPD | AHCI_P_IX_MP : 0) |
AHCI_P_IX_PRC | AHCI_P_IX_PC));
MFp4: Large set of CAM inprovements. - Unify bus reset/probe sequence. Whenever bus attached at boot or later, CAM will automatically reset and scan it. It allows to remove duplicate code from many drivers. - Any bus, attached before CAM completed it's boot-time initialization, will equally join to the process, delaying boot if needed. - New kern.cam.boot_delay loader tunable should help controllers that are still unable to register their buses in time (such as slow USB/ PCCard/ CardBus devices), by adding one more event to wait on boot. - To allow synchronization between different CAM levels, concept of requests priorities was extended. Priorities now split between several "run levels". Device can be freezed at specified level, allowing higher priority requests to pass. For example, no payload requests allowed, until PMP driver enable port. ATA XPT negotiate transfer parameters, periph driver configure caching and so on. - Frozen requests are no more counted by request allocation scheduler. It fixes deadlocks, when frozen low priority payload requests occupying slots, required by higher levels to manage theit execution. - Two last changes were holding proper ATA reinitialization and error recovery implementation. Now it is done: SATA controllers and Port Multipliers now implement automatic hot-plug and should correctly recover from timeouts and bus resets. - Improve SCSI error recovery for devices on buses without automatic sense reporting, such as ATAPI or USB. For example, it allows CAM to wait, while CD drive loads disk, instead of immediately return error status. - Decapitalize diagnostic messages and make them more readable and sensible. - Teach PMP driver to limit maximum speed on fan-out ports. - Make boot wait for PMP scan completes, and make rescan more reliable. - Fix pass driver, to return CCB to user level in case of error. - Increase number of retries in cd driver, as device may return several UAs.
2010-01-28 08:41:30 +00:00
xpt_release_simq(ch->sim, TRUE);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
return;
}
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
if (bootverbose)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
device_printf(ch->dev, "AHCI reset: device found\n");
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Wait for clearing busy status. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_wait_ready(ch, dumping ? 31000 : 0, 0)) {
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
if (dumping)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_clo(ch);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
else
ch->resetting = 310;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->devices = 1;
/* Enable wanted port interrupts */
ATA_OUTL(ch->r_mem, AHCI_P_IE,
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
(((ch->pm_level != 0) ? AHCI_P_IX_CPD | AHCI_P_IX_MP : 0) |
AHCI_P_IX_TFE | AHCI_P_IX_HBF |
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
AHCI_P_IX_HBD | AHCI_P_IX_IF | AHCI_P_IX_OF |
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
((ch->pm_level == 0) ? AHCI_P_IX_PRC : 0) | AHCI_P_IX_PC |
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
AHCI_P_IX_DP | AHCI_P_IX_UF | (ctlr->ccc ? 0 : AHCI_P_IX_SDB) |
AHCI_P_IX_DS | AHCI_P_IX_PS | (ctlr->ccc ? 0 : AHCI_P_IX_DHR)));
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
if (ch->resetting)
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
callout_reset(&ch->reset_timer, hz / 10, ahci_reset_to, ch);
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
else {
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_start(ch, 1);
Refactor hard-reset implementation in ahci(4). Instead of spinning in a tight loop for up to 15 seconds, polling for device readiness while it spins up, return reset completion just after PHY reports "connect well" or 100ms connection timeout. If device was found, use callout for checking device readiness with 100ms period up to full 31 second timeout. This fixes system freeze for 5-10 seconds on drives hot plug-in.
2011-04-12 20:50:57 +00:00
xpt_release_simq(ch->sim, TRUE);
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_setup_fis(struct ahci_channel *ch, struct ahci_cmd_tab *ctp, union ccb *ccb, int tag)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
u_int8_t *fis = &ctp->cfis[0];
We were returning 20 bytes as the FIS size to send, but only initializing 16. Initialize all 20 so we don't send garbage in the Auxiliary register. The SATA standard mandates a 5 dword length for the Host to Device FIS. Sponsored by: Netflix
2014-08-30 02:13:09 +00:00
bzero(fis, 20);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
fis[0] = 0x27; /* host to device */
fis[1] = (ccb->ccb_h.target_id & 0x0f);
if (ccb->ccb_h.func_code == XPT_SCSI_IO) {
fis[1] |= 0x80;
fis[2] = ATA_PACKET_CMD;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE &&
ch->curr[ccb->ccb_h.target_id].mode >= ATA_DMA)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
fis[3] = ATA_F_DMA;
else {
fis[5] = ccb->csio.dxfer_len;
fis[6] = ccb->csio.dxfer_len >> 8;
}
fis[7] = ATA_D_LBA;
fis[15] = ATA_A_4BIT;
bcopy((ccb->ccb_h.flags & CAM_CDB_POINTER) ?
ccb->csio.cdb_io.cdb_ptr : ccb->csio.cdb_io.cdb_bytes,
ctp->acmd, ccb->csio.cdb_len);
No need to erase all 64 bytes of CFIS area if we never use more then 16.
2013-03-24 16:51:21 +00:00
bzero(ctp->acmd + ccb->csio.cdb_len, 32 - ccb->csio.cdb_len);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else if ((ccb->ataio.cmd.flags & CAM_ATAIO_CONTROL) == 0) {
fis[1] |= 0x80;
fis[2] = ccb->ataio.cmd.command;
fis[3] = ccb->ataio.cmd.features;
fis[4] = ccb->ataio.cmd.lba_low;
fis[5] = ccb->ataio.cmd.lba_mid;
fis[6] = ccb->ataio.cmd.lba_high;
fis[7] = ccb->ataio.cmd.device;
fis[8] = ccb->ataio.cmd.lba_low_exp;
fis[9] = ccb->ataio.cmd.lba_mid_exp;
fis[10] = ccb->ataio.cmd.lba_high_exp;
fis[11] = ccb->ataio.cmd.features_exp;
Pass lower 3 bits of sector_count for FPDMA commands. When this code was written those bits were N/A, but now the lowest bit is Rebuild Assist Recovery Control (RARC). MFC after: 1 month
2020-10-22 03:30:39 +00:00
fis[12] = ccb->ataio.cmd.sector_count;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ccb->ataio.cmd.flags & CAM_ATAIO_FPDMA) {
Fix stupid math mistake in r366922. MFC after: 3 days
2020-11-21 00:51:39 +00:00
fis[12] &= 0x07;
Pass lower 3 bits of sector_count for FPDMA commands. When this code was written those bits were N/A, but now the lowest bit is Rebuild Assist Recovery Control (RARC). MFC after: 1 month
2020-10-22 03:30:39 +00:00
fis[12] |= tag << 3;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Add support for managing Shingled Magnetic Recording (SMR) drives. This change includes support for SCSI SMR drives (which conform to the Zoned Block Commands or ZBC spec) and ATA SMR drives (which conform to the Zoned ATA Command Set or ZAC spec) behind SAS expanders. This includes full management support through the GEOM BIO interface, and through a new userland utility, zonectl(8), and through camcontrol(8). This is now ready for filesystems to use to detect and manage zoned drives. (There is no work in progress that I know of to use this for ZFS or UFS, if anyone is interested, let me know and I may have some suggestions.) Also, improve ATA command passthrough and dispatch support, both via ATA and ATA passthrough over SCSI. Also, add support to camcontrol(8) for the ATA Extended Power Conditions feature set. You can now manage ATA device power states, and set various idle time thresholds for a drive to enter lower power states. Note that this change cannot be MFCed in full, because it depends on changes to the struct bio API that break compatilibity. In order to avoid breaking the stable API, only changes that don't touch or depend on the struct bio changes can be merged. For example, the camcontrol(8) changes don't depend on the new bio API, but zonectl(8) and the probe changes to the da(4) and ada(4) drivers do depend on it. Also note that the SMR changes have not yet been tested with an actual SCSI ZBC device, or a SCSI to ATA translation layer (SAT) that supports ZBC to ZAC translation. I have not yet gotten a suitable drive or SAT layer, so any testing help would be appreciated. These changes have been tested with Seagate Host Aware SATA drives attached to both SAS and SATA controllers. Also, I do not have any SATA Host Managed devices, and I suspect that it may take additional (hopefully minor) changes to support them. Thanks to Seagate for supplying the test hardware and answering questions. sbin/camcontrol/Makefile: Add epc.c and zone.c. sbin/camcontrol/camcontrol.8: Document the zone and epc subcommands. sbin/camcontrol/camcontrol.c: Add the zone and epc subcommands. Add auxiliary register support to build_ata_cmd(). Make sure to set the CAM_ATAIO_NEEDRESULT, CAM_ATAIO_DMA, and CAM_ATAIO_FPDMA flags as appropriate for ATA commands. Add a new get_ata_status() function to parse ATA result from SCSI sense descriptors (for ATA passthrough over SCSI) and ATA I/O requests. sbin/camcontrol/camcontrol.h: Update the build_ata_cmd() prototype Add get_ata_status(), zone(), and epc(). sbin/camcontrol/epc.c: Support for ATA Extended Power Conditions features. This includes support for all features documented in the ACS-4 Revision 12 specification from t13.org (dated February 18, 2016). The EPC feature set allows putting a drive into a power power mode immediately, or setting timeouts so that the drive will automatically enter progressively lower power states after various idle times. sbin/camcontrol/fwdownload.c: Update the firmware download code for the new build_ata_cmd() arguments. sbin/camcontrol/zone.c: Implement support for Shingled Magnetic Recording (SMR) drives via SCSI Zoned Block Commands (ZBC) and ATA Zoned Device ATA Command Set (ZAC). These specs were developed in concert, and are functionally identical. The primary differences are due to SCSI and ATA differences. (SCSI is big endian, ATA is little endian, for example.) This includes support for all commands defined in the ZBC and ZAC specs. sys/cam/ata/ata_all.c: Decode a number of additional ATA command names in ata_op_string(). Add a new CCB building function, ata_read_log(). Add ata_zac_mgmt_in() and ata_zac_mgmt_out() CCB building functions. These support both DMA and NCQ encapsulation. sys/cam/ata/ata_all.h: Add prototypes for ata_read_log(), ata_zac_mgmt_out(), and ata_zac_mgmt_in(). sys/cam/ata/ata_da.c: Revamp the ada(4) driver to support zoned devices. Add four new probe states to gather information needed for zone support. Add a new adasetflags() function to avoid duplication of large blocks of flag setting between the async handler and register functions. Add new sysctl variables that describe zone support and paramters. Add support for the new BIO_ZONE bio, and all of its subcommands: DISK_ZONE_OPEN, DISK_ZONE_CLOSE, DISK_ZONE_FINISH, DISK_ZONE_RWP, DISK_ZONE_REPORT_ZONES, and DISK_ZONE_GET_PARAMS. sys/cam/scsi/scsi_all.c: Add command descriptions for the ZBC IN/OUT commands. Add descriptions for ZBC Host Managed devices. Add a new function, scsi_ata_pass() to do ATA passthrough over SCSI. This will eventually replace scsi_ata_pass_16() -- it can create the 12, 16, and 32-byte variants of the ATA PASS-THROUGH command, and supports setting all of the registers defined as of SAT-4, Revision 5 (March 11, 2016). Change scsi_ata_identify() to use scsi_ata_pass() instead of scsi_ata_pass_16(). Add a new scsi_ata_read_log() function to facilitate reading ATA logs via SCSI. sys/cam/scsi/scsi_all.h: Add the new ATA PASS-THROUGH(32) command CDB. Add extended and variable CDB opcodes. Add Zoned Block Device Characteristics VPD page. Add ATA Return SCSI sense descriptor. Add prototypes for scsi_ata_read_log() and scsi_ata_pass(). sys/cam/scsi/scsi_da.c: Revamp the da(4) driver to support zoned devices. Add five new probe states, four of which are needed for ATA devices. Add five new sysctl variables that describe zone support and parameters. The da(4) driver supports SCSI ZBC devices, as well as ATA ZAC devices when they are attached via a SCSI to ATA Translation (SAT) layer. Since ZBC -> ZAC translation is a new feature in the T10 SAT-4 spec, most SATA drives will be supported via ATA commands sent via the SCSI ATA PASS-THROUGH command. The da(4) driver will prefer the ZBC interface, if it is available, for performance reasons, but will use the ATA PASS-THROUGH interface to the ZAC command set if the SAT layer doesn't support translation yet. As I mentioned above, ZBC command support is untested. Add support for the new BIO_ZONE bio, and all of its subcommands: DISK_ZONE_OPEN, DISK_ZONE_CLOSE, DISK_ZONE_FINISH, DISK_ZONE_RWP, DISK_ZONE_REPORT_ZONES, and DISK_ZONE_GET_PARAMS. Add scsi_zbc_in() and scsi_zbc_out() CCB building functions. Add scsi_ata_zac_mgmt_out() and scsi_ata_zac_mgmt_in() CCB/CDB building functions. Note that these have return values, unlike almost all other CCB building functions in CAM. The reason is that they can fail, depending upon the particular combination of input parameters. The primary failure case is if the user wants NCQ, but fails to specify additional CDB storage. NCQ requires using the 32-byte version of the SCSI ATA PASS-THROUGH command, and the current CAM CDB size is 16 bytes. sys/cam/scsi/scsi_da.h: Add ZBC IN and ZBC OUT CDBs and opcodes. Add SCSI Report Zones data structures. Add scsi_zbc_in(), scsi_zbc_out(), scsi_ata_zac_mgmt_out(), and scsi_ata_zac_mgmt_in() prototypes. sys/dev/ahci/ahci.c: Fix SEND / RECEIVE FPDMA QUEUED in the ahci(4) driver. ahci_setup_fis() previously set the top bits of the sector count register in the FIS to 0 for FPDMA commands. This is okay for read and write, because the PRIO field is in the only thing in those bits, and we don't implement that further up the stack. But, for SEND and RECEIVE FPDMA QUEUED, the subcommand is in that byte, so it needs to be transmitted to the drive. In ahci_setup_fis(), always set the the top 8 bits of the sector count register. We need it in both the standard and NCQ / FPDMA cases. sys/geom/eli/g_eli.c: Pass BIO_ZONE commands through the GELI class. sys/geom/geom.h: Add g_io_zonecmd() prototype. sys/geom/geom_dev.c: Add new DIOCZONECMD ioctl, which allows sending zone commands to disks. sys/geom/geom_disk.c: Add support for BIO_ZONE commands. sys/geom/geom_disk.h: Add a new flag, DISKFLAG_CANZONE, that indicates that a given GEOM disk client can handle BIO_ZONE commands. sys/geom/geom_io.c: Add a new function, g_io_zonecmd(), that handles execution of BIO_ZONE commands. Add permissions check for BIO_ZONE commands. Add command decoding for BIO_ZONE commands. sys/geom/geom_subr.c: Add DDB command decoding for BIO_ZONE commands. sys/kern/subr_devstat.c: Record statistics for REPORT ZONES commands. Note that the number of bytes transferred for REPORT ZONES won't quite match what is received from the harware. This is because we're necessarily counting bytes coming from the da(4) / ada(4) drivers, which are using the disk_zone.h interface to communicate up the stack. The structure sizes it uses are slightly different than the SCSI and ATA structure sizes. sys/sys/ata.h: Add many bit and structure definitions for ZAC, NCQ, and EPC command support. sys/sys/bio.h: Convert the bio_cmd field to a straight enumeration. This will yield more space for additional commands in the future. After change r297955 and other related changes, this is now possible. Converting to an enumeration will also prevent use as a bitmask in the future. sys/sys/disk.h: Define the DIOCZONECMD ioctl. sys/sys/disk_zone.h: Add a new API for managing zoned disks. This is very close to the SCSI ZBC and ATA ZAC standards, but uses integers in native byte order instead of big endian (SCSI) or little endian (ATA) byte arrays. This is intended to offer to the complete feature set of the ZBC and ZAC disk management without requiring the application developer to include SCSI or ATA headers. We also use one set of headers for ioctl consumers and kernel bio-level consumers. sys/sys/param.h: Bump __FreeBSD_version for sys/bio.h command changes, and inclusion of SMR support. usr.sbin/Makefile: Add the zonectl utility. usr.sbin/diskinfo/diskinfo.c Add disk zoning capability to the 'diskinfo -v' output. usr.sbin/zonectl/Makefile: Add zonectl makefile. usr.sbin/zonectl/zonectl.8 zonectl(8) man page. usr.sbin/zonectl/zonectl.c The zonectl(8) utility. This allows managing SCSI or ATA zoned disks via the disk_zone.h API. You can report zones, reset write pointers, get parameters, etc. Sponsored by: Spectra Logic Differential Revision: https://reviews.freebsd.org/D6147 Reviewed by: wblock (documentation)
2016-05-19 14:08:36 +00:00
fis[13] = ccb->ataio.cmd.sector_count_exp;
Add icc (Isochronous Command Completion) ccb_ataio field. MFC after: 1 week
2020-11-02 01:01:41 +00:00
if (ccb->ataio.ata_flags & ATA_FLAG_ICC)
fis[14] = ccb->ataio.icc;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
fis[15] = ATA_A_4BIT;
Add icc (Isochronous Command Completion) ccb_ataio field. MFC after: 1 week
2020-11-02 01:01:41 +00:00
if (ccb->ataio.ata_flags & ATA_FLAG_AUX) {
fis[16] = ccb->ataio.aux & 0xff;
fis[17] = (ccb->ataio.aux >> 8) & 0xff;
fis[18] = (ccb->ataio.aux >> 16) & 0xff;
fis[19] = (ccb->ataio.aux >> 24) & 0xff;
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else {
fis[15] = ccb->ataio.cmd.control;
}
return (20);
}
static int
ahci_sata_connect(struct ahci_channel *ch)
{
u_int32_t status;
ahci: increase timout For some devices, like Marvell 88SE9230, it takes more time to connect to the device. This patch introduces a special flag that extends the timeout from around 100ms to around 500ms. This change is based on the work of: Peter Eriksson <pen@lysator.liu.se> PR: 243401 Reviewed by: imp Tested by: dch MFC after: 3 days Sponsored by: Equinix Sponsored by: SkunkWerks, GmbH Sponsored by: Klara, Inc. Differential Revision: https://reviews.freebsd.org/D38413
2023-02-10 15:56:04 +00:00
int timeout, timeoutslot, found = 0;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ahci: increase timout For some devices, like Marvell 88SE9230, it takes more time to connect to the device. This patch introduces a special flag that extends the timeout from around 100ms to around 500ms. This change is based on the work of: Peter Eriksson <pen@lysator.liu.se> PR: 243401 Reviewed by: imp Tested by: dch MFC after: 3 days Sponsored by: Equinix Sponsored by: SkunkWerks, GmbH Sponsored by: Klara, Inc. Differential Revision: https://reviews.freebsd.org/D38413
2023-02-10 15:56:04 +00:00
/*
* Wait for "connect well", up to 100ms by default and
* up to 500ms for devices with the SLOWDEV quirk.
*/
timeoutslot = ((ch->quirks & AHCI_Q_SLOWDEV) ? 5000 : 1000);
for (timeout = 0; timeout < timeoutslot; timeout++) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
status = ATA_INL(ch->r_mem, AHCI_P_SSTS);
According to specification. device should respond to COMRESET with COMINIT in no more then 10ms. If we detected no device presence within that time, there is no reason to wait longer.
2011-04-19 10:51:19 +00:00
if ((status & ATA_SS_DET_MASK) != ATA_SS_DET_NO_DEVICE)
found = 1;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (((status & ATA_SS_DET_MASK) == ATA_SS_DET_PHY_ONLINE) &&
((status & ATA_SS_SPD_MASK) != ATA_SS_SPD_NO_SPEED) &&
((status & ATA_SS_IPM_MASK) == ATA_SS_IPM_ACTIVE))
break;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
if ((status & ATA_SS_DET_MASK) == ATA_SS_DET_PHY_OFFLINE) {
if (bootverbose) {
device_printf(ch->dev, "SATA offline status=%08x\n",
status);
}
return (0);
}
According to specification. device should respond to COMRESET with COMINIT in no more then 10ms. If we detected no device presence within that time, there is no reason to wait longer.
2011-04-19 10:51:19 +00:00
if (found == 0 && timeout >= 100)
break;
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(100);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
ahci: increase timout For some devices, like Marvell 88SE9230, it takes more time to connect to the device. This patch introduces a special flag that extends the timeout from around 100ms to around 500ms. This change is based on the work of: Peter Eriksson <pen@lysator.liu.se> PR: 243401 Reviewed by: imp Tested by: dch MFC after: 3 days Sponsored by: Equinix Sponsored by: SkunkWerks, GmbH Sponsored by: Klara, Inc. Differential Revision: https://reviews.freebsd.org/D38413
2023-02-10 15:56:04 +00:00
if (timeout >= timeoutslot || !found) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (bootverbose) {
According to specification. device should respond to COMRESET with COMINIT in no more then 10ms. If we detected no device presence within that time, there is no reason to wait longer.
2011-04-19 10:51:19 +00:00
device_printf(ch->dev,
"SATA connect timeout time=%dus status=%08x\n",
timeout * 100, status);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
return (0);
}
if (bootverbose) {
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
device_printf(ch->dev, "SATA connect time=%dus status=%08x\n",
timeout * 100, status);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
/* Clear SATA error register */
ATA_OUTL(ch->r_mem, AHCI_P_SERR, 0xffffffff);
return (1);
}
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_sata_phy_reset(struct ahci_channel *ch)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
{
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
int sata_rev;
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
uint32_t val, detval;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if (ch->listening) {
val = ATA_INL(ch->r_mem, AHCI_P_CMD);
val |= AHCI_P_CMD_SUD;
ATA_OUTL(ch->r_mem, AHCI_P_CMD, val);
ch->listening = 0;
}
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
sata_rev = ch->user[ch->pm_present ? 15 : 0].revision;
if (sata_rev == 1)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
val = ATA_SC_SPD_SPEED_GEN1;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
else if (sata_rev == 2)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
val = ATA_SC_SPD_SPEED_GEN2;
MFp4: Improve ATA mode/SATA revision control.
2009-11-26 08:49:46 +00:00
else if (sata_rev == 3)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
val = ATA_SC_SPD_SPEED_GEN3;
else
val = 0;
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
detval = ahci_ch_detval(ch, ATA_SC_DET_RESET);
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_SCTL,
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
detval | val |
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
ATA_SC_IPM_DIS_PARTIAL | ATA_SC_IPM_DIS_SLUMBER);
- Tune different wait loops to cut some more milliseconds from reset time. - Do not call ahci_start() before device signature received. It is required by the specification and caused non-fatal reset timeouts on AMD chipsets.
2011-04-18 13:34:31 +00:00
DELAY(1000);
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
detval = ahci_ch_detval(ch, ATA_SC_DET_IDLE);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_SCTL,
Implement ability to turn on/off PHYs for AHCI devices. As part of Chuck's work on fixing kernel crashes caused by disk I/O errors, it is useful to be able to trigger various kinds of errors. This patch allows causing an AHCI-attached disk to disappear, by having the driver keep the PHY disabled when the driver would otherwise enable the PHY. It also allows making the disk reappear by having the driver go back to setting the PHY enable/disable state as it normal would and simulating the hardware event that causes a bus rescan. Submitted by: Chuck Silvers Sponsored by: Netflix, Inc Differential Revision: https://reviews.freebsd.org/D16043
2018-11-03 00:37:51 +00:00
detval | val | ((ch->pm_level > 0) ? 0 :
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
(ATA_SC_IPM_DIS_PARTIAL | ATA_SC_IPM_DIS_SLUMBER)));
Disable PHY of unconnected ports when interface power management enabled. It allows to save a bit more power (about 0.5W on 2 unused ports of ICH8M).
2010-02-03 12:00:57 +00:00
if (!ahci_sata_connect(ch)) {
Some changes around hot-plug and interface power-management: - use ATA_SE_EXCHANGED (SError.DIAG.X) bit to detect hot-plug events when power-management enabled and ATA_SE_PHY_CHANGED (SError.DIAG.N) can't be trusted; - on controllers supporting staggered spin-up (SS) put unused channels into Listen state instead of Off. It should still save some power, but allow plug-in events to be detected; - on controllers supporting cold presence detection (CPD), when power management enabled, use CPD events to detect hot-plug in addition to PHY events.
2011-04-15 16:40:31 +00:00
if (ch->caps & AHCI_CAP_SSS) {
val = ATA_INL(ch->r_mem, AHCI_P_CMD);
val &= ~AHCI_P_CMD_SUD;
ATA_OUTL(ch->r_mem, AHCI_P_CMD, val);
ch->listening = 1;
} else if (ch->pm_level > 0)
Disable PHY of unconnected ports when interface power management enabled. It allows to save a bit more power (about 0.5W on 2 unused ports of ICH8M).
2010-02-03 12:00:57 +00:00
ATA_OUTL(ch->r_mem, AHCI_P_SCTL, ATA_SC_DET_DISABLE);
return (0);
}
return (1);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
static int
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_check_ids(struct ahci_channel *ch, union ccb *ccb)
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
{
if (ccb->ccb_h.target_id > ((ch->caps & AHCI_CAP_SPM) ? 15 : 0)) {
ccb->ccb_h.status = CAM_TID_INVALID;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ccb);
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return (-1);
}
if (ccb->ccb_h.target_lun != 0) {
ccb->ccb_h.status = CAM_LUN_INVALID;
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ccb);
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return (-1);
}
return (0);
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
static void
ahciaction(struct cam_sim *sim, union ccb *ccb)
{
struct ahci_channel *ch;
CAM_DEBUG(ccb->ccb_h.path, CAM_DEBUG_TRACE, ("ahciaction func_code=%x\n",
ccb->ccb_h.func_code));
ch = (struct ahci_channel *)cam_sim_softc(sim);
switch (ccb->ccb_h.func_code) {
/* Common cases first */
case XPT_ATA_IO: /* Execute the requested I/O operation */
case XPT_SCSI_IO:
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_check_ids(ch, ccb))
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return;
if (ch->devices == 0 ||
(ch->pm_present == 0 &&
ccb->ccb_h.target_id > 0 && ccb->ccb_h.target_id < 15)) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status = CAM_SEL_TIMEOUT;
break;
}
Implement automatic SCSI sense fetching for ahci(4).
2011-04-12 11:24:59 +00:00
ccb->ccb_h.recovery_type = RECOVERY_NONE;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Check for command collision. */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_check_collision(ch, ccb)) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
/* Freeze command. */
ch->frozen = ccb;
/* We have only one frozen slot, so freeze simq also. */
xpt_freeze_simq(ch->sim, 1);
return;
}
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_begin_transaction(ch, ccb);
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
case XPT_ABORT: /* Abort the specified CCB */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
break;
case XPT_SET_TRAN_SETTINGS:
{
struct ccb_trans_settings *cts = &ccb->cts;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
struct ahci_device *d;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_check_ids(ch, ccb))
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
if (cts->type == CTS_TYPE_CURRENT_SETTINGS)
d = &ch->curr[ccb->ccb_h.target_id];
else
d = &ch->user[ccb->ccb_h.target_id];
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_REVISION)
d->revision = cts->xport_specific.sata.revision;
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_MODE)
d->mode = cts->xport_specific.sata.mode;
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_BYTECOUNT)
d->bytecount = min(8192, cts->xport_specific.sata.bytecount);
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_TAGS)
d->tags = min(ch->numslots, cts->xport_specific.sata.tags);
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_PM)
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ch->pm_present = cts->xport_specific.sata.pm_present;
- Give ATA/SATA SIMs info about ATAPI packet size, supported by device. - Make ATA XPT to reject longer SCSI CDBs then supported by device, or any SCSI CDBs, if device doesn't support ATAPI.
2010-02-02 11:09:28 +00:00
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_ATAPI)
d->atapi = cts->xport_specific.sata.atapi;
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
if (cts->xport_specific.sata.valid & CTS_SATA_VALID_CAPS)
d->caps = cts->xport_specific.sata.caps;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
break;
}
case XPT_GET_TRAN_SETTINGS:
/* Get default/user set transfer settings for the target */
{
struct ccb_trans_settings *cts = &ccb->cts;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
struct ahci_device *d;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
uint32_t status;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
if (ahci_check_ids(ch, ccb))
Add Target/LUN ID checks and deny access to targets 1-14 when PMP absent. Enforce PMA bit clearing when PMP detached to avoid further scan timeouts.
2010-04-30 08:02:12 +00:00
return;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
if (cts->type == CTS_TYPE_CURRENT_SETTINGS)
d = &ch->curr[ccb->ccb_h.target_id];
else
d = &ch->user[ccb->ccb_h.target_id];
ATA/SATA controllers have no idea about protocol of the connected device until transport will do some probe actions (at least soft reset). Make ATA/SATA SIMs to not report bogus and confusing PROTO_ATA protocol. Make ATA/SATA transport to fill that gap by reporting protocol to SIM with XPT_SET_TRAN_SETTINGS and patching XPT_GET_TRAN_SETTINGS results if needed.
2012-06-06 06:52:51 +00:00
cts->protocol = PROTO_UNSPECIFIED;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
cts->protocol_version = PROTO_VERSION_UNSPECIFIED;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cts->transport = XPORT_SATA;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
cts->transport_version = XPORT_VERSION_UNSPECIFIED;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cts->proto_specific.valid = 0;
cts->xport_specific.sata.valid = 0;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
if (cts->type == CTS_TYPE_CURRENT_SETTINGS &&
(ccb->ccb_h.target_id == 15 ||
(ccb->ccb_h.target_id == 0 && !ch->pm_present))) {
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
status = ATA_INL(ch->r_mem, AHCI_P_SSTS) & ATA_SS_SPD_MASK;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
if (status & 0x0f0) {
cts->xport_specific.sata.revision =
(status & 0x0f0) >> 4;
cts->xport_specific.sata.valid |=
CTS_SATA_VALID_REVISION;
}
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
cts->xport_specific.sata.caps = d->caps & CTS_SATA_CAPS_D;
if (ch->pm_level) {
if (ch->caps & (AHCI_CAP_PSC | AHCI_CAP_SSC))
cts->xport_specific.sata.caps |= CTS_SATA_CAPS_H_PMREQ;
if (ch->caps2 & AHCI_CAP2_APST)
cts->xport_specific.sata.caps |= CTS_SATA_CAPS_H_APST;
}
if ((ch->caps & AHCI_CAP_SNCQ) &&
(ch->quirks & AHCI_Q_NOAA) == 0)
cts->xport_specific.sata.caps |= CTS_SATA_CAPS_H_DMAAA;
Improve SATA Asynchronous Notification feature support in CAM: - make SATA SIMs announce capabilities to handle SDB with Notification bit; - make PMP driver honor this SIMs capability; - make SATA XPT to negotiate and enable this feature for ATAPI devices. This feature allows supporting SATA ATAPI devices to inform system about some events happened, that may require attention. In my case this allows LG GH22LS50 SATA DVR-RW drive to report tray open/close events. Events reported to CAM in form of AC_SCSI_AEN async. Further they could be used as a hints for checking device status and reporting media change to upper layers, for example, via spoiling mechanism of GEOM.
2011-04-13 16:20:54 +00:00
cts->xport_specific.sata.caps |= CTS_SATA_CAPS_H_AN;
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
cts->xport_specific.sata.caps &=
ch->user[ccb->ccb_h.target_id].caps;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_CAPS;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
} else {
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
cts->xport_specific.sata.revision = d->revision;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_REVISION;
Make SATA XPT negotiate and enable some additional SATA features, such as: - device initiated power management (some devices support only this way); - Automatic Partial to Slumber Transition (more power saving); - DMA auto-activation (expected to slightly improve performance). More features could be added later, when hardware supports.
2010-05-02 12:07:47 +00:00
cts->xport_specific.sata.caps = d->caps;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_CAPS;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
cts->xport_specific.sata.mode = d->mode;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_MODE;
cts->xport_specific.sata.bytecount = d->bytecount;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_BYTECOUNT;
cts->xport_specific.sata.pm_present = ch->pm_present;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cts->xport_specific.sata.valid |= CTS_SATA_VALID_PM;
MFp4: - Extend XPT-SIM transfer settings control API. Now it allows to report to SATA SIM number of tags supported by each device, implement ATA mode and SATA revision negotiation for both SATA and PATA SIMs. - Make ahci(4) and siis(4) to use submitted maximum tag number, when scheduling requests. It allows to support NCQ on devices with lower tags count then controller supports. - Make PMP driver to report attached devices connection speeds. - Implement ATA mode negotiation between user settings, device and controller capabilities.
2009-11-24 12:47:58 +00:00
cts->xport_specific.sata.tags = d->tags;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_TAGS;
- Give ATA/SATA SIMs info about ATAPI packet size, supported by device. - Make ATA XPT to reject longer SCSI CDBs then supported by device, or any SCSI CDBs, if device doesn't support ATAPI.
2010-02-02 11:09:28 +00:00
cts->xport_specific.sata.atapi = d->atapi;
cts->xport_specific.sata.valid |= CTS_SATA_VALID_ATAPI;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
break;
}
case XPT_RESET_BUS: /* Reset the specified SCSI bus */
case XPT_RESET_DEV: /* Bus Device Reset the specified SCSI device */
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset(ch);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
break;
case XPT_TERM_IO: /* Terminate the I/O process */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
break;
case XPT_PATH_INQ: /* Path routing inquiry */
{
struct ccb_pathinq *cpi = &ccb->cpi;
cpi->version_num = 1; /* XXX??? */
MFp4: Check SNCQ HBA capability bit when reporting NCQ support to CAM.
2009-11-14 20:06:04 +00:00
cpi->hba_inquiry = PI_SDTR_ABLE;
if (ch->caps & AHCI_CAP_SNCQ)
cpi->hba_inquiry |= PI_TAG_ABLE;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
if (ch->caps & AHCI_CAP_SPM)
cpi->hba_inquiry |= PI_SATAPM;
cpi->target_sprt = 0;
Block FPDMA TRIM for ASMedia HBAs. Experiments show FPDMA TRIM command timeouts on ASMedia HBAs, while the same SSDs working fine on Intel HBAs. MFC after: 2 weeks
2017-04-18 08:04:15 +00:00
cpi->hba_misc = PIM_SEQSCAN | PIM_UNMAPPED;
if ((ch->quirks & AHCI_Q_NOAUX) == 0)
cpi->hba_misc |= PIM_ATA_EXT;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cpi->hba_eng_cnt = 0;
if (ch->caps & AHCI_CAP_SPM)
MFp4: Report real max_target = 15. SIM doesn't need to know that target 15 is PMP. It is XPT business.
2009-10-21 13:00:01 +00:00
cpi->max_target = 15;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
else
cpi->max_target = 0;
cpi->max_lun = 0;
cpi->initiator_id = 0;
cpi->bus_id = cam_sim_bus(sim);
cpi->base_transfer_speed = 150000;
Always null-terminate ccb_pathinq.(sim_vid|hba_vid|dev_name) The sim_vid, hba_vid, and dev_name fields of struct ccb_pathinq are fixed-length strings. AFAICT the only place they're read is in sbin/camcontrol/camcontrol.c, which assumes they'll be null-terminated. However, the kernel doesn't null-terminate them. A bunch of copy-pasted code uses strncpy to write them, and doesn't guarantee null-termination. For at least 4 drivers (mpr, mps, ciss, and hyperv), the hba_vid field actually overflows. You can see the result by doing "camcontrol negotiate da0 -v". This change null-terminates those fields everywhere they're set in the kernel. It also shortens a few strings to ensure they'll fit within the 16-character field. PR: 215474 Reported by: Coverity CID: 1009997 1010000 1010001 1010002 1010003 1010004 1010005 CID: 1331519 1010006 1215097 1010007 1288967 1010008 1306000 CID: 1211924 1010009 1010010 1010011 1010012 1010013 1010014 CID: 1147190 1010017 1010016 1010018 1216435 1010020 1010021 CID: 1010022 1009666 1018185 1010023 1010025 1010026 1010027 CID: 1010028 1010029 1010030 1010031 1010033 1018186 1018187 CID: 1010035 1010036 1010042 1010041 1010040 1010039 Reviewed by: imp, sephe, slm MFC after: 4 weeks Sponsored by: Spectra Logic Corp Differential Revision: https://reviews.freebsd.org/D9037 Differential Revision: https://reviews.freebsd.org/D9038
2017-01-04 20:26:42 +00:00
strlcpy(cpi->sim_vid, "FreeBSD", SIM_IDLEN);
strlcpy(cpi->hba_vid, "AHCI", HBA_IDLEN);
strlcpy(cpi->dev_name, cam_sim_name(sim), DEV_IDLEN);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cpi->unit_number = cam_sim_unit(sim);
cpi->transport = XPORT_SATA;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
cpi->transport_version = XPORT_VERSION_UNSPECIFIED;
Partially revert r236666: Return PROTO_ATA protocol in response to XPT_PATH_INQ. smartmontools uses it to identify ATA devices and I don't know any other place now where it is important. It could probably use XPT_GDEV_TYPE instead for more accurate protocol information, but let it live for now. Reported by: matthew MFC after: 3 days
2012-06-10 11:17:14 +00:00
cpi->protocol = PROTO_ATA;
MFp4: - Add Command Completion Coalescing support. - Add SNTF support. - Add two more power management modes (4, 5), implemented on driver level. - Fix interface mode setting. - Reduce interface reset time. - Do not report meaningless protocol/transport versions. - Report CAP2 register content. - Some performance optimizations.
2009-08-30 15:20:13 +00:00
cpi->protocol_version = PROTO_VERSION_UNSPECIFIED;
Make MAXPHYS tunable. Bump MAXPHYS to 1M. Replace MAXPHYS by runtime variable maxphys. It is initialized from MAXPHYS by default, but can be also adjusted with the tunable kern.maxphys. Make b_pages[] array in struct buf flexible. Size b_pages[] for buffer cache buffers exactly to atop(maxbcachebuf) (currently it is sized to atop(MAXPHYS)), and b_pages[] for pbufs is sized to atop(maxphys) + 1. The +1 for pbufs allow several pbuf consumers, among them vmapbuf(), to use unaligned buffers still sized to maxphys, esp. when such buffers come from userspace (*). Overall, we save significant amount of otherwise wasted memory in b_pages[] for buffer cache buffers, while bumping MAXPHYS to desired high value. Eliminate all direct uses of the MAXPHYS constant in kernel and driver sources, except a place which initialize maxphys. Some random (and arguably weird) uses of MAXPHYS, e.g. in linuxolator, are converted straight. Some drivers, which use MAXPHYS to size embeded structures, get private MAXPHYS-like constant; their convertion is out of scope for this work. Changes to cam/, dev/ahci, dev/ata, dev/mpr, dev/mpt, dev/mvs, dev/siis, where either submitted by, or based on changes by mav. Suggested by: mav (*) Reviewed by: imp, mav, imp, mckusick, scottl (intermediate versions) Tested by: pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D27225
2020-11-28 12:12:51 +00:00
cpi->maxio = ctob(AHCI_SG_ENTRIES - 1);
ATI SB600 can't handle 256 sectors transfers with FPDMA (NCQ). MFC after: 3 days
2009-09-03 12:37:17 +00:00
/* ATI SB600 can't handle 256 sectors with FPDMA (NCQ). */
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
if (ch->quirks & AHCI_Q_MAXIO_64K)
Round maxio for ATI SB600 to 64K. Submitted by: scottl@
2009-09-03 19:02:17 +00:00
cpi->maxio = min(cpi->maxio, 128 * 512);
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
cpi->hba_vendor = ch->vendorid;
cpi->hba_device = ch->deviceid;
cpi->hba_subvendor = ch->subvendorid;
cpi->hba_subdevice = ch->subdeviceid;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
cpi->ccb_h.status = CAM_REQ_CMP;
break;
}
default:
ccb->ccb_h.status = CAM_REQ_INVALID;
break;
}
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
ahci_done(ch, ccb);
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
static void
ahcipoll(struct cam_sim *sim)
{
struct ahci_channel *ch = (struct ahci_channel *)cam_sim_softc(sim);
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
uint32_t istatus;
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
Merge CAM locking changes from the projects/camlock branch to radically reduce lock congestion and improve SMP scalability of the SCSI/ATA stack, preparing the ground for the coming next GEOM direct dispatch support. Replace big per-SIM locks with bunch of smaller ones: - per-LUN locks to protect device and peripheral drivers state; - per-target locks to protect list of LUNs on target; - per-bus locks to protect reference counting; - per-send queue locks to protect queue of CCBs to be sent; - per-done queue locks to protect queue of completed CCBs; - remaining per-SIM locks now protect only HBA driver internals. While holding LUN lock it is allowed (while not recommended for performance reasons) to take SIM lock. The opposite acquisition order is forbidden. All the other locks are leaf locks, that can be taken anywhere, but should not be cascaded. Many functions, such as: xpt_action(), xpt_done(), xpt_async(), xpt_create_path(), etc. are no longer require (but allow) SIM lock to be held. To keep compatibility and solve cases where SIM lock can't be dropped, all xpt_async() calls in addition to xpt_done() calls are queued to completion threads for async processing in clean environment without SIM lock held. Instead of single CAM SWI thread, used for commands completion processing before, use multiple (depending on number of CPUs) threads. Load balanced between them using "hash" of the device B:T:L address. HBA drivers that can drop SIM lock during completion processing and have sufficient number of completion threads to efficiently scale to multiple CPUs can use new function xpt_done_direct() to avoid extra context switch. Make ahci(4) driver to use this mechanism depending on hardware setup. Sponsored by: iXsystems, Inc. MFC after: 2 months
2013-10-21 12:00:26 +00:00
/* Read interrupt statuses and process if any. */
istatus = ATA_INL(ch->r_mem, AHCI_P_IS);
if (istatus != 0)
ahci_ch_intr_main(ch, istatus);
Handle ready timeout during polled operation same as done in mvs(4) before.
2011-04-18 16:00:46 +00:00
if (ch->resetting != 0 &&
(--ch->resetpolldiv <= 0 || !callout_pending(&ch->reset_timer))) {
ch->resetpolldiv = 1000;
Bunch of microoptimizations to reduce dereferences and cache collisions.
2014-09-08 12:11:49 +00:00
ahci_reset_to(ch);
Handle ready timeout during polled operation same as done in mvs(4) before.
2011-04-18 16:00:46 +00:00
}
Separate the parallel scsi knowledge out of the core of the XPT, and modularize it so that new transports can be created. Add a transport for SATA Add a periph+protocol layer for ATA Add a driver for AHCI-compliant hardware. Add a maxio field to CAM so that drivers can advertise their max I/O capability. Modify various drivers so that they are insulated from the value of MAXPHYS. The new ATA/SATA code supports AHCI-compliant hardware, and will override the classic ATA driver if it is loaded as a module at boot time or compiled into the kernel. The stack now support NCQ (tagged queueing) for increased performance on modern SATA drives. It also supports port multipliers. ATA drives are accessed via 'ada' device nodes. ATAPI drives are accessed via 'cd' device nodes. They can all be enumerated and manipulated via camcontrol, just like SCSI drives. SCSI commands are not translated to their ATA equivalents; ATA native commands are used throughout the entire stack, including camcontrol. See the camcontrol manpage for further details. Testing this code may require that you update your fstab, and possibly modify your BIOS to enable AHCI functionality, if available. This code is very experimental at the moment. The userland ABI/API has changed, so applications will need to be recompiled. It may change further in the near future. The 'ada' device name may also change as more infrastructure is completed in this project. The goal is to eventually put all CAM busses and devices until newbus, allowing for interesting topology and management options. Few functional changes will be seen with existing SCSI/SAS/FC drivers, though the userland ABI has still changed. In the future, transports specific modules for SAS and FC may appear in order to better support the topologies and capabilities of these technologies. The modularization of CAM and the addition of the ATA/SATA modules is meant to break CAM out of the mold of being specific to SCSI, letting it grow to be a framework for arbitrary transports and protocols. It also allows drivers to be written to support discrete hardware without jeopardizing the stability of non-related hardware. While only an AHCI driver is provided now, a Silicon Image driver is also in the works. Drivers for ICH1-4, ICH5-6, PIIX, classic IDE, and any other hardware is possible and encouraged. Help with new transports is also encouraged. Submitted by: scottl, mav Approved by: re
2009-07-10 08:18:08 +00:00
}
Define a single instance of ahci_devclass and reference it from all the attachment code for various SOCs and busses. Remove all the static and should-have-been-static and named-differently instances of it. This should eliminate the recently-grown build warnings about multiple definitions when building arm kernels.
2017-10-02 02:58:28 +00:00
Separate out PCI attachment from the main AHCI driver. Move checks of PCI IDs into quirks, which mostly fit (though you'd get no argument from me that AHCI_Q_SATA1_UNIT0 is oddly specific). Set these quirks in the PCI attachment. Make some shared functions public so that PCI and possibly other bus attachments can use them. The split isn't perfect yet, but it is functional. The split will be perfected as other bus attachments for AHCI are written. Sponsored by: Netflix Reviewed by: kan, mav Differential Revision: https://reviews.freebsd.org/D699
2014-09-04 22:22:53 +00:00
MODULE_VERSION(ahci, 1);
MODULE_DEPEND(ahci, cam, 1, 1, 1);
Reference in New Issue Copy Permalink