d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
08977788d5
freebsd-nq/sys/dev/mlx/mlxreg.h

622 lines
19 KiB
C
Raw Normal View History

This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
/*-
* Copyright (c) 1999 Michael Smith
* 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.
* 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 AND CONTRIBUTORS ``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 OR CONTRIBUTORS 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.
*
* $FreeBSD$
*/
Move some hardware-related items to the hardware-related header. Shift to using the same queueing strategy that the amr driver uses. Some simple tests indicate that we use about 2% of the CPU at around 500tps with the controller completely saturated with I/O.
1999-10-26 23:20:43 +00:00
#define MLX_BLKSIZE 512 /* fixed feature */
Don't use PAGE_SIZE to calculate controller-specific attributes. PR: kern/21220 Submitted by: Dennis Lindroos MFC After: 1 week
2004-11-25 12:15:49 +00:00
#define MLX_PAGE_SIZE 4096 /* controller, not cpu, attribute */
Move some hardware-related items to the hardware-related header. Shift to using the same queueing strategy that the amr driver uses. Some simple tests indicate that we use about 2% of the CPU at around 500tps with the controller completely saturated with I/O.
1999-10-26 23:20:43 +00:00
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
/*
* Selected command codes.
*/
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_CMD_ENQUIRY_OLD 0x05
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
#define MLX_CMD_ENQUIRY 0x53
#define MLX_CMD_ENQUIRY2 0x1c
#define MLX_CMD_ENQSYSDRIVE 0x19
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_CMD_READSG 0xb6
#define MLX_CMD_WRITESG 0xb7
#define MLX_CMD_READSG_OLD 0x82
#define MLX_CMD_WRITESG_OLD 0x83
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
#define MLX_CMD_FLUSH 0x0a
#define MLX_CMD_LOGOP 0x72
#define MLX_CMD_REBUILDASYNC 0x16
#define MLX_CMD_CHECKASYNC 0x1e
#define MLX_CMD_REBUILDSTAT 0x0c
#define MLX_CMD_STOPCHANNEL 0x13
#define MLX_CMD_STARTCHANNEL 0x12
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_CMD_READ_CONFIG 0x4e
#define MLX_CMD_DIRECT_CDB 0x04
Add features required for basic userland management utility: - implement user-initiated background drive rebuild - implement user-initiated background consistency check - log controller-initiated background rebuild/check operations Try to fix the elusive "invalid log operation" bug, and panic if we do hit this one in the hopes of getting better information. Tidy up diagnostic messages. Try to use disk_create/disk_destroy correctly. This isn't working properly yet, but it's not clear whose fault that is.
2000-04-11 02:52:46 +00:00
#define MLX_CMD_DEVICE_STATE 0x50
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#ifdef _KERNEL
#define MLX_CFG_BASE0 0x10 /* first region */
#define MLX_CFG_BASE1 0x14 /* second region (type 3 only) */
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
/*
* Status values.
*/
#define MLX_STATUS_OK 0x0000
#define MLX_STATUS_RDWROFFLINE 0x0002 /* read/write claims drive is offline */
#define MLX_STATUS_WEDGED 0xdead /* controller not listening */
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
#define MLX_STATUS_LOST 0xbeef /* never came back */
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
#define MLX_STATUS_BUSY 0xffff /* command is in controller */
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
/*
* Accessor defines for the V3 interface.
*/
#define MLX_V3_MAILBOX 0x00
#define MLX_V3_STATUS_IDENT 0x0d
#define MLX_V3_STATUS 0x0e
#define MLX_V3_IDBR 0x40
#define MLX_V3_ODBR 0x41
#define MLX_V3_IER 0x43
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V3_FWERROR 0x3f
#define MLX_V3_FWERROR_PARAM1 0x00
#define MLX_V3_FWERROR_PARAM2 0x01
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
Add locking to mlx(4) to make it MPSAFE along with some other fixes: - Use callout(9) rather than timeout(9). - Add a mutex as an I/O lock that protects the adapter and is used for the I/O path. - Add an sx lock as a configuration lock that protects the relationship of configured volumes. - Freeze the request queue when a DMA load is deferred with EINPROGRESS and unfreeze the queue when the DMA callback is invoked. - Explicitly poll the hardware while waiting to submit a command to allow completed commands to free up slots in the command ring. - Remove driver-wide 'initted' variable from mlx_*_fw_handshake() routines. That state should be per-controller instead. Add it as an argument since the first caller knows when it is the first caller. - Remove explicit bus_space tag/handle and use bus_*() rather than bus_space_*(). - Move duplicated PCI device ID probing into a mlx_pci_match() routine. - Don't check for PCIM_CMD_MEMEN (the PCI bus will enable that when allocating the resource) and use pci_enable_busmaster() rather than manipulating the register directly. Tested by: no one despite multiple requests (hope it works)
2012-09-17 15:27:30 +00:00
#define MLX_V3_PUT_MAILBOX(sc, idx, val) bus_write_1(sc->mlx_mem, MLX_V3_MAILBOX + idx, val)
#define MLX_V3_GET_STATUS_IDENT(sc) bus_read_1 (sc->mlx_mem, MLX_V3_STATUS_IDENT)
#define MLX_V3_GET_STATUS(sc) bus_read_2 (sc->mlx_mem, MLX_V3_STATUS)
#define MLX_V3_GET_IDBR(sc) bus_read_1 (sc->mlx_mem, MLX_V3_IDBR)
#define MLX_V3_PUT_IDBR(sc, val) bus_write_1(sc->mlx_mem, MLX_V3_IDBR, val)
#define MLX_V3_GET_ODBR(sc) bus_read_1 (sc->mlx_mem, MLX_V3_ODBR)
#define MLX_V3_PUT_ODBR(sc, val) bus_write_1(sc->mlx_mem, MLX_V3_ODBR, val)
#define MLX_V3_PUT_IER(sc, val) bus_write_1(sc->mlx_mem, MLX_V3_IER, val)
#define MLX_V3_GET_FWERROR(sc) bus_read_1 (sc->mlx_mem, MLX_V3_FWERROR)
#define MLX_V3_PUT_FWERROR(sc, val) bus_write_1(sc->mlx_mem, MLX_V3_FWERROR, val)
#define MLX_V3_GET_FWERROR_PARAM1(sc) bus_read_1 (sc->mlx_mem, MLX_V3_FWERROR_PARAM1)
#define MLX_V3_GET_FWERROR_PARAM2(sc) bus_read_1 (sc->mlx_mem, MLX_V3_FWERROR_PARAM2)
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
#define MLX_V3_IDB_FULL (1<<0) /* mailbox is full */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V3_IDB_INIT_BUSY (1<<1) /* initialisation in progress */
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
#define MLX_V3_IDB_SACK (1<<1) /* acknowledge status read */
#define MLX_V3_ODB_SAVAIL (1<<0) /* status is available */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V3_FWERROR_PEND (1<<2) /* firmware error pending */
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
/*
* Accessor defines for the V4 interface.
*/
#define MLX_V4_MAILBOX 0x1000
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V4_MAILBOX_LENGTH 16
#define MLX_V4_STATUS_IDENT 0x1018
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
#define MLX_V4_STATUS 0x101a
#define MLX_V4_IDBR 0x0020
#define MLX_V4_ODBR 0x002c
#define MLX_V4_IER 0x0034
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V4_FWERROR 0x103f
#define MLX_V4_FWERROR_PARAM1 0x1000
#define MLX_V4_FWERROR_PARAM2 0x1001
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
/* use longword access? */
Add locking to mlx(4) to make it MPSAFE along with some other fixes: - Use callout(9) rather than timeout(9). - Add a mutex as an I/O lock that protects the adapter and is used for the I/O path. - Add an sx lock as a configuration lock that protects the relationship of configured volumes. - Freeze the request queue when a DMA load is deferred with EINPROGRESS and unfreeze the queue when the DMA callback is invoked. - Explicitly poll the hardware while waiting to submit a command to allow completed commands to free up slots in the command ring. - Remove driver-wide 'initted' variable from mlx_*_fw_handshake() routines. That state should be per-controller instead. Add it as an argument since the first caller knows when it is the first caller. - Remove explicit bus_space tag/handle and use bus_*() rather than bus_space_*(). - Move duplicated PCI device ID probing into a mlx_pci_match() routine. - Don't check for PCIM_CMD_MEMEN (the PCI bus will enable that when allocating the resource) and use pci_enable_busmaster() rather than manipulating the register directly. Tested by: no one despite multiple requests (hope it works)
2012-09-17 15:27:30 +00:00
#define MLX_V4_PUT_MAILBOX(sc, idx, val) bus_write_1(sc->mlx_mem, MLX_V4_MAILBOX + idx, val)
#define MLX_V4_GET_STATUS_IDENT(sc) bus_read_1 (sc->mlx_mem, MLX_V4_STATUS_IDENT)
#define MLX_V4_GET_STATUS(sc) bus_read_2 (sc->mlx_mem, MLX_V4_STATUS)
#define MLX_V4_GET_IDBR(sc) bus_read_4 (sc->mlx_mem, MLX_V4_IDBR)
#define MLX_V4_PUT_IDBR(sc, val) bus_write_4(sc->mlx_mem, MLX_V4_IDBR, val)
#define MLX_V4_GET_ODBR(sc) bus_read_4 (sc->mlx_mem, MLX_V4_ODBR)
#define MLX_V4_PUT_ODBR(sc, val) bus_write_4(sc->mlx_mem, MLX_V4_ODBR, val)
#define MLX_V4_PUT_IER(sc, val) bus_write_4(sc->mlx_mem, MLX_V4_IER, val)
#define MLX_V4_GET_FWERROR(sc) bus_read_1 (sc->mlx_mem, MLX_V4_FWERROR)
#define MLX_V4_PUT_FWERROR(sc, val) bus_write_1(sc->mlx_mem, MLX_V4_FWERROR, val)
#define MLX_V4_GET_FWERROR_PARAM1(sc) bus_read_1 (sc->mlx_mem, MLX_V4_FWERROR_PARAM1)
#define MLX_V4_GET_FWERROR_PARAM2(sc) bus_read_1 (sc->mlx_mem, MLX_V4_FWERROR_PARAM2)
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
#define MLX_V4_IDB_FULL (1<<0) /* mailbox is full */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V4_IDB_INIT_BUSY (1<<1) /* initialisation in progress */
Add support for the version 4 firmware interface and the DAC960 PG and PJ cards. This will probably also support the AcceleRAID and eXtremeRAID cards, but nobody has volunteered one for testing, so I haven't enabled their PCI device IDs. Slightly clean up communication between the disk devices and the controller device as per new practice, and move some more register- related items int mlxreg.h from mlxvar.h. Remove some unnecessary read-modify-write operations to the card control registers; they don't behave like that. Increase the status polling interval to 10 seconds. It's still possible to load the card up to the point where a status poll will find the previous poll still running, but this will reduce the incidence of complaints.
1999-10-14 02:54:06 +00:00
#define MLX_V4_IDB_HWMBOX_CMD (1<<0) /* posted hardware mailbox command */
#define MLX_V4_IDB_SACK (1<<1) /* acknowledge status read */
#define MLX_V4_IDB_MEMMBOX_CMD (1<<4) /* posted memory mailbox command */
#define MLX_V4_ODB_HWSAVAIL (1<<0) /* status is available for hardware mailbox */
#define MLX_V4_ODB_MEMSAVAIL (1<<1) /* status is available for memory mailbox */
#define MLX_V4_ODB_HWMBOX_ACK (1<<0) /* ack status read from hardware mailbox */
#define MLX_V4_ODB_MEMMBOX_ACK (1<<1) /* ack status read from memory mailbox */
#define MLX_V4_IER_MASK 0xfb /* message unit interrupt mask */
#define MLX_V4_IER_DISINT (1<<2) /* interrupt disable bit */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V4_FWERROR_PEND (1<<2) /* firmware error pending */
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
/*
* Accessor defines for the V5 interface
*/
#define MLX_V5_MAILBOX 0x50
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V5_MAILBOX_LENGTH 16
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
#define MLX_V5_STATUS_IDENT 0x5d
#define MLX_V5_STATUS 0x5e
#define MLX_V5_IDBR 0x60
#define MLX_V5_ODBR 0x61
#define MLX_V5_IER 0x34
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V5_FWERROR 0x63
#define MLX_V5_FWERROR_PARAM1 0x50
#define MLX_V5_FWERROR_PARAM2 0x51
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
Add locking to mlx(4) to make it MPSAFE along with some other fixes: - Use callout(9) rather than timeout(9). - Add a mutex as an I/O lock that protects the adapter and is used for the I/O path. - Add an sx lock as a configuration lock that protects the relationship of configured volumes. - Freeze the request queue when a DMA load is deferred with EINPROGRESS and unfreeze the queue when the DMA callback is invoked. - Explicitly poll the hardware while waiting to submit a command to allow completed commands to free up slots in the command ring. - Remove driver-wide 'initted' variable from mlx_*_fw_handshake() routines. That state should be per-controller instead. Add it as an argument since the first caller knows when it is the first caller. - Remove explicit bus_space tag/handle and use bus_*() rather than bus_space_*(). - Move duplicated PCI device ID probing into a mlx_pci_match() routine. - Don't check for PCIM_CMD_MEMEN (the PCI bus will enable that when allocating the resource) and use pci_enable_busmaster() rather than manipulating the register directly. Tested by: no one despite multiple requests (hope it works)
2012-09-17 15:27:30 +00:00
#define MLX_V5_PUT_MAILBOX(sc, idx, val) bus_write_1(sc->mlx_mem, MLX_V5_MAILBOX + idx, val)
#define MLX_V5_GET_STATUS_IDENT(sc) bus_read_1 (sc->mlx_mem, MLX_V5_STATUS_IDENT)
#define MLX_V5_GET_STATUS(sc) bus_read_2 (sc->mlx_mem, MLX_V5_STATUS)
#define MLX_V5_GET_IDBR(sc) bus_read_1 (sc->mlx_mem, MLX_V5_IDBR)
#define MLX_V5_PUT_IDBR(sc, val) bus_write_1(sc->mlx_mem, MLX_V5_IDBR, val)
#define MLX_V5_GET_ODBR(sc) bus_read_1 (sc->mlx_mem, MLX_V5_ODBR)
#define MLX_V5_PUT_ODBR(sc, val) bus_write_1(sc->mlx_mem, MLX_V5_ODBR, val)
#define MLX_V5_PUT_IER(sc, val) bus_write_1(sc->mlx_mem, MLX_V5_IER, val)
#define MLX_V5_GET_FWERROR(sc) bus_read_1 (sc->mlx_mem, MLX_V5_FWERROR)
#define MLX_V5_PUT_FWERROR(sc, val) bus_write_1(sc->mlx_mem, MLX_V5_FWERROR, val)
#define MLX_V5_GET_FWERROR_PARAM1(sc) bus_read_1 (sc->mlx_mem, MLX_V5_FWERROR_PARAM1)
#define MLX_V5_GET_FWERROR_PARAM2(sc) bus_read_1 (sc->mlx_mem, MLX_V5_FWERROR_PARAM2)
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
#define MLX_V5_IDB_EMPTY (1<<0) /* mailbox is empty */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V5_IDB_INIT_DONE (1<<1) /* initialisation has completed */
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
#define MLX_V5_IDB_HWMBOX_CMD (1<<0) /* posted hardware mailbox command */
#define MLX_V5_IDB_SACK (1<<1) /* acknowledge status read */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V5_IDB_RESET (1<<3) /* reset request */
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
#define MLX_V5_IDB_MEMMBOX_CMD (1<<4) /* posted memory mailbox command */
#define MLX_V5_ODB_HWSAVAIL (1<<0) /* status is available for hardware mailbox */
#define MLX_V5_ODB_MEMSAVAIL (1<<1) /* status is available for memory mailbox */
#define MLX_V5_ODB_HWMBOX_ACK (1<<0) /* ack status read from hardware mailbox */
#define MLX_V5_ODB_MEMMBOX_ACK (1<<1) /* ack status read from memory mailbox */
#define MLX_V5_IER_DISINT (1<<2) /* interrupt disable bit */
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#define MLX_V5_FWERROR_PEND (1<<2) /* firmware error pending */
#endif /* _KERNEL */
Major update to the Mylex DAC960 driver adding new hardware support and fixing some major bugs. - Add support for the v5 firmware interface, used by the DAC1164P (tested) and AcceleRAID 352 (untested but should work). We now cover all of the Mylex family's protocols except for v2 (used by EISA and Alpha-compatible cards). - Fix an accounting bug which resulted in endless 'poll still busy' messages. In situations of high controller load the count of poll commands could be incremented without actually successfully launching a command. This totally removes the accounting for status poll commnads; it was its own worst enemy. - Add some simple reentry prevention locks to processing of the waiting and completed command queues to prevent races which could result in I/O being done or completed twice (both are fatal). This highlights a need for simple locking primitives in both the UP and SMP kernels. - Streamline the handling of command completion to reduce the amount of redundant work being done. Remove the code which tests for commands that have gone missing in action; nobody has ever seen one of these and it wouldn't have worked properly anyhow. - Handle disconnection of drives from the controller in the detach, not shutdown method. This avoids problems flushing the cache in a panic when a drive is mounted. - Don't call bus_generic_detach when disconnecting drives; it doesn't actually do anything useful. - Increment the log message index regardless of whether we actually retrieved one or not. If we run into a message that we can't fetch, we don't want to spin endlessly complaining about the fact. - Don't assume that interrupts will work when we're flushing the controller. We may think they are enabled, but in eg. a panic situation the controller may not be able to deliver an interrupt.
1999-12-11 00:00:13 +00:00
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
/*
* Scatter-gather list format, type 1, kind 00.
*/
struct mlx_sgentry
{
u_int32_t sg_addr;
u_int32_t sg_count;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
/*
* Command result buffers, as placed in system memory by the controller.
*/
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_enquiry_old /* MLX_CMD_ENQUIRY_OLD */
{
u_int8_t me_num_sys_drvs;
u_int8_t res1[3];
u_int32_t me_drvsize[8];
u_int16_t me_flash_age;
u_int8_t me_status_flags;
u_int8_t me_free_state_change_count;
u_int8_t me_fwminor;
u_int8_t me_fwmajor;
u_int8_t me_rebuild_flag;
u_int8_t me_max_commands;
u_int8_t me_offline_sd_count;
u_int8_t res3;
u_int8_t me_critical_sd_count;
u_int8_t res4[3];
u_int8_t me_dead_count;
u_int8_t res5;
u_int8_t me_rebuild_count;
u_int8_t me_misc_flags;
struct
{
u_int8_t dd_targ;
u_int8_t dd_chan;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed me_dead[20];
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
struct mlx_enquiry /* MLX_CMD_ENQUIRY */
{
u_int8_t me_num_sys_drvs;
u_int8_t res1[3];
u_int32_t me_drvsize[32];
u_int16_t me_flash_age;
u_int8_t me_status_flags;
#define MLX_ENQ_SFLAG_DEFWRERR (1<<0) /* deferred write error indicator */
#define MLX_ENQ_SFLAG_BATTLOW (1<<1) /* battery low */
u_int8_t res2;
u_int8_t me_fwminor;
u_int8_t me_fwmajor;
u_int8_t me_rebuild_flag;
u_int8_t me_max_commands;
u_int8_t me_offline_sd_count;
u_int8_t res3;
u_int16_t me_event_log_seq_num;
u_int8_t me_critical_sd_count;
u_int8_t res4[3];
u_int8_t me_dead_count;
u_int8_t res5;
u_int8_t me_rebuild_count;
u_int8_t me_misc_flags;
#define MLX_ENQ_MISC_BBU (1<<3) /* battery backup present */
struct
{
u_int8_t dd_targ;
u_int8_t dd_chan;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed me_dead[20];
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
struct mlx_enquiry2 /* MLX_CMD_ENQUIRY2 */
{
u_int32_t me_hardware_id;
u_int32_t me_firmware_id;
u_int32_t res1;
u_int8_t me_configured_channels;
u_int8_t me_actual_channels;
u_int8_t me_max_targets;
u_int8_t me_max_tags;
u_int8_t me_max_sys_drives;
u_int8_t me_max_arms;
u_int8_t me_max_spans;
u_int8_t res2;
u_int32_t res3;
u_int32_t me_mem_size;
u_int32_t me_cache_size;
u_int32_t me_flash_size;
u_int32_t me_nvram_size;
u_int16_t me_mem_type;
u_int16_t me_clock_speed;
u_int16_t me_mem_speed;
u_int16_t me_hardware_speed;
Updates resulting from new documentation from Mylex and some cleaning: - Don't keep private copies of some of the data fields from the ENQUIRY and ENQUIRY2 commands. Instead, standardise on the ENQUIRY2 command for initial adapter information, and keep a copy of the entire structure. Refer to it where appropriate. - Move all of the controller description functionality into a new function. Print lots more controller data if bootverbose is set. Add knowledge of the DAC960 PR, PT, PTL0 and PRL controllers, rename the 960PTL -> PTL0 and 1100P -> 1100PVX. - Correctly terminate an error message. The controller interface procedures have been reviewed against the Mylex-supplied documentation; no changes appear necessary at this time.
1999-12-22 01:21:28 +00:00
u_int8_t res4[12];
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
u_int16_t me_max_commands;
u_int16_t me_max_sg;
u_int16_t me_max_dp;
u_int16_t me_max_iod;
u_int16_t me_max_comb;
u_int8_t me_latency;
u_int8_t res5;
u_int8_t me_scsi_timeout;
u_int8_t res6;
u_int16_t me_min_freelines;
u_int8_t res7[8];
u_int8_t me_rate_const;
u_int8_t res8[11];
u_int16_t me_physblk;
u_int16_t me_logblk;
u_int16_t me_maxblk;
u_int16_t me_blocking_factor;
u_int16_t me_cacheline;
u_int8_t me_scsi_cap;
u_int8_t res9[5];
Updates resulting from new documentation from Mylex and some cleaning: - Don't keep private copies of some of the data fields from the ENQUIRY and ENQUIRY2 commands. Instead, standardise on the ENQUIRY2 command for initial adapter information, and keep a copy of the entire structure. Refer to it where appropriate. - Move all of the controller description functionality into a new function. Print lots more controller data if bootverbose is set. Add knowledge of the DAC960 PR, PT, PTL0 and PRL controllers, rename the 960PTL -> PTL0 and 1100P -> 1100PVX. - Correctly terminate an error message. The controller interface procedures have been reviewed against the Mylex-supplied documentation; no changes appear necessary at this time.
1999-12-22 01:21:28 +00:00
u_int16_t me_firmware_build;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
u_int8_t me_fault_mgmt_type;
u_int8_t res10;
u_int32_t me_firmware_features;
u_int8_t res11[8];
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
struct mlx_enq_sys_drive /* MLX_CMD_ENQSYSDRIVE returns an array of 32 of these */
{
u_int32_t sd_size;
u_int8_t sd_state;
u_int8_t sd_raidlevel;
u_int16_t res1;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
struct mlx_eventlog_entry /* MLX_CMD_LOGOP/MLX_LOGOP_GET */
{
u_int8_t el_type;
u_int8_t el_length;
u_char el_target:5;
u_char el_channel:3;
u_char el_lun:6;
u_char res1:2;
u_int16_t el_seqno;
u_char el_errorcode:7;
u_char el_valid:1;
u_int8_t el_segment;
u_char el_sensekey:4;
u_char res2:1;
u_char el_ILI:1;
u_char el_EOM:1;
u_char el_filemark:1;
u_int8_t el_information[4];
u_int8_t el_addsense;
u_int8_t el_csi[4];
u_int8_t el_asc;
u_int8_t el_asq;
u_int8_t res3[12];
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
#define MLX_LOGOP_GET 0x00 /* operation codes for MLX_CMD_LOGOP */
#define MLX_LOGMSG_SENSE 0x00 /* log message contents codes */
struct mlx_rebuild_stat /* MLX_CMD_REBUILDSTAT */
{
u_int32_t rb_drive;
u_int32_t rb_size;
u_int32_t rb_remaining;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
This is a driver for the Mylex DAC960 family of integrated RAID controllers. It currently supports the P, PL, PD and PU variants, with more to be supported shortly.
1999-10-07 02:20:32 +00:00
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_config2
{
u_int16_t cf_flags1;
#define MLX_CF2_ACTV_NEG (1<<1)
#define MLX_CF2_NORSTRTRY (1<<7)
#define MLX_CF2_STRGWRK (1<<8)
#define MLX_CF2_HPSUPP (1<<9)
#define MLX_CF2_NODISCN (1<<10)
#define MLX_CF2_ARM (1<<13)
#define MLX_CF2_OFM (1<<15)
#define MLX_CF2_AEMI (MLX_CF2_ARM | MLX_CF2_OFM)
u_int8_t cf_oemid;
u_int8_t cf_oem_model;
u_int8_t cf_physical_sector;
u_int8_t cf_logical_sector;
u_int8_t cf_blockfactor;
u_int8_t cf_flags2;
#define MLX_CF2_READAH (1<<0)
#define MLX_CF2_BIOSDLY (1<<1)
#define MLX_CF2_REASS1S (1<<4)
#define MLX_CF2_FUAENABL (1<<6)
#define MLX_CF2_R5ALLS (1<<7)
u_int8_t cf_rcrate;
u_int8_t cf_res1;
u_int8_t cf_blocks_per_cache_line;
u_int8_t cf_blocks_per_stripe;
u_int8_t cf_scsi_param_0;
u_int8_t cf_scsi_param_1;
u_int8_t cf_scsi_param_2;
u_int8_t cf_scsi_param_3;
u_int8_t cf_scsi_param_4;
u_int8_t cf_scsi_param_5;
u_int8_t cf_scsi_initiator_id;
u_int8_t cf_res2;
u_int8_t cf_startup_mode;
u_int8_t cf_simultaneous_spinup_devices;
u_int8_t cf_delay_between_spinups;
u_int8_t cf_res3;
u_int16_t cf_checksum;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_sys_drv_span
{
u_int32_t sp_start_lba;
u_int32_t sp_nblks;
u_int8_t sp_arm[8];
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_sys_drv
{
u_int8_t sd_status;
u_int8_t sd_ext_status;
u_int8_t sd_mod1;
u_int8_t sd_mod2;
u_int8_t sd_raidlevel;
#define MLX_SYS_DRV_WRITEBACK (1<<7)
#define MLX_SYS_DRV_RAID0 0
#define MLX_SYS_DRV_RAID1 1
#define MLX_SYS_DRV_RAID3 3
#define MLX_SYS_DRV_RAID5 5
#define MLX_SYS_DRV_RAID6 6
#define MLX_SYS_DRV_JBOD 7
u_int8_t sd_valid_arms;
u_int8_t sd_valid_spans;
u_int8_t sd_init_state;
#define MLX_SYS_DRV_INITTED 0x81;
struct mlx_sys_drv_span sd_span[4];
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_phys_drv
{
u_int8_t pd_flags1;
#define MLX_PHYS_DRV_PRESENT (1<<0)
u_int8_t pd_flags2;
#define MLX_PHYS_DRV_OTHER 0x00
#define MLX_PHYS_DRV_DISK 0x01
#define MLX_PHYS_DRV_SEQUENTIAL 0x02
#define MLX_PHYS_DRV_CDROM 0x03
#define MLX_PHYS_DRV_FAST20 (1<<3)
#define MLX_PHYS_DRV_SYNC (1<<4)
#define MLX_PHYS_DRV_FAST (1<<5)
#define MLX_PHYS_DRV_WIDE (1<<6)
#define MLX_PHYS_DRV_TAG (1<<7)
u_int8_t pd_status;
#define MLX_PHYS_DRV_DEAD 0x00
#define MLX_PHYS_DRV_WRONLY 0x02
#define MLX_PHYS_DRV_ONLINE 0x03
#define MLX_PHYS_DRV_STANDBY 0x10
u_int8_t pd_res1;
u_int8_t pd_period;
u_int8_t pd_offset;
u_int32_t pd_config_size;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_core_cfg
{
u_int8_t cc_num_sys_drives;
u_int8_t cc_res1[3];
struct mlx_sys_drv cc_sys_drives[32];
struct mlx_phys_drv cc_phys_drives[5 * 16];
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
struct mlx_dcdb
{
u_int8_t dcdb_target:4;
u_int8_t dcdb_channel:4;
u_int8_t dcdb_flags;
#define MLX_DCDB_NO_DATA 0x00
#define MLX_DCDB_DATA_IN 0x01
#define MLX_DCDB_DATA_OUT 0x02
#define MLX_DCDB_EARLY_STATUS (1<<2)
#define MLX_DCDB_TIMEOUT_10S 0x10
#define MLX_DCDB_TIMEOUT_60S 0x20
#define MLX_DCDB_TIMEOUT_20M 0x30
#define MLX_DCDB_TIMEOUT_24H 0x40
#define MLX_DCDB_NO_AUTO_SENSE (1<<6)
#define MLX_DCDB_DISCONNECT (1<<7)
u_int16_t dcdb_datasize;
u_int32_t dcdb_physaddr;
u_int8_t dcdb_cdb_length:4;
u_int8_t dcdb_datasize_high:4;
u_int8_t dcdb_sense_length;
u_int8_t dcdb_cdb[12];
u_int8_t dcdb_sense[64];
u_int8_t dcdb_status;
u_int8_t res1;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
Add features required for basic userland management utility: - implement user-initiated background drive rebuild - implement user-initiated background consistency check - log controller-initiated background rebuild/check operations Try to fix the elusive "invalid log operation" bug, and panic if we do hit this one in the hopes of getting better information. Tidy up diagnostic messages. Try to use disk_create/disk_destroy correctly. This isn't working properly yet, but it's not clear whose fault that is.
2000-04-11 02:52:46 +00:00
struct mlx_bbtable_entry
{
u_int32_t bbt_block_number;
u_int8_t bbt_extent;
u_int8_t res1;
u_int8_t bbt_entry_type;
u_int8_t bbt_system_drive:5;
u_int8_t res2:3;
use __packed.
2002-09-23 18:54:32 +00:00
} __packed;
Add features required for basic userland management utility: - implement user-initiated background drive rebuild - implement user-initiated background consistency check - log controller-initiated background rebuild/check operations Try to fix the elusive "invalid log operation" bug, and panic if we do hit this one in the hopes of getting better information. Tidy up diagnostic messages. Try to use disk_create/disk_destroy correctly. This isn't working properly yet, but it's not clear whose fault that is.
2000-04-11 02:52:46 +00:00
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#ifdef _KERNEL
Move some hardware-related items to the hardware-related header. Shift to using the same queueing strategy that the amr driver uses. Some simple tests indicate that we use about 2% of the CPU at around 500tps with the controller completely saturated with I/O.
1999-10-26 23:20:43 +00:00
/*
* Inlines to build various command structures
*/
static __inline void
mlx_make_type1(struct mlx_command *mc,
u_int8_t code,
u_int16_t f1,
u_int32_t f2,
u_int8_t f3,
u_int32_t f4,
u_int8_t f5)
{
mc->mc_mailbox[0x0] = code;
mc->mc_mailbox[0x2] = f1 & 0xff;
mc->mc_mailbox[0x3] = (((f2 >> 24) & 0x3) << 6) | ((f1 >> 8) & 0x3f);
mc->mc_mailbox[0x4] = f2 & 0xff;
mc->mc_mailbox[0x5] = (f2 >> 8) & 0xff;
mc->mc_mailbox[0x6] = (f2 >> 16) & 0xff;
mc->mc_mailbox[0x7] = f3;
mc->mc_mailbox[0x8] = f4 & 0xff;
mc->mc_mailbox[0x9] = (f4 >> 8) & 0xff;
mc->mc_mailbox[0xa] = (f4 >> 16) & 0xff;
mc->mc_mailbox[0xb] = (f4 >> 24) & 0xff;
mc->mc_mailbox[0xc] = f5;
}
static __inline void
mlx_make_type2(struct mlx_command *mc,
u_int8_t code,
u_int8_t f1,
u_int8_t f2,
u_int8_t f3,
u_int8_t f4,
u_int8_t f5,
u_int8_t f6,
u_int32_t f7,
u_int8_t f8)
{
mc->mc_mailbox[0x0] = code;
mc->mc_mailbox[0x2] = f1;
mc->mc_mailbox[0x3] = f2;
mc->mc_mailbox[0x4] = f3;
mc->mc_mailbox[0x5] = f4;
mc->mc_mailbox[0x6] = f5;
mc->mc_mailbox[0x7] = f6;
mc->mc_mailbox[0x8] = f7 & 0xff;
mc->mc_mailbox[0x9] = (f7 >> 8) & 0xff;
mc->mc_mailbox[0xa] = (f7 >> 16) & 0xff;
mc->mc_mailbox[0xb] = (f7 >> 24) & 0xff;
mc->mc_mailbox[0xc] = f8;
}
static __inline void
mlx_make_type3(struct mlx_command *mc,
u_int8_t code,
u_int8_t f1,
u_int8_t f2,
u_int16_t f3,
u_int8_t f4,
u_int8_t f5,
u_int32_t f6,
u_int8_t f7)
{
mc->mc_mailbox[0x0] = code;
mc->mc_mailbox[0x2] = f1;
mc->mc_mailbox[0x3] = f2;
mc->mc_mailbox[0x4] = f3 & 0xff;
mc->mc_mailbox[0x5] = (f3 >> 8) & 0xff;
mc->mc_mailbox[0x6] = f4;
mc->mc_mailbox[0x7] = f5;
mc->mc_mailbox[0x8] = f6 & 0xff;
mc->mc_mailbox[0x9] = (f6 >> 8) & 0xff;
mc->mc_mailbox[0xa] = (f6 >> 16) & 0xff;
mc->mc_mailbox[0xb] = (f6 >> 24) & 0xff;
mc->mc_mailbox[0xc] = f7;
}
static __inline void
mlx_make_type4(struct mlx_command *mc,
u_int8_t code,
u_int16_t f1,
u_int32_t f2,
u_int32_t f3,
u_int8_t f4)
{
mc->mc_mailbox[0x0] = code;
mc->mc_mailbox[0x2] = f1 & 0xff;
mc->mc_mailbox[0x3] = (f1 >> 8) & 0xff;
mc->mc_mailbox[0x4] = f2 & 0xff;
mc->mc_mailbox[0x5] = (f2 >> 8) & 0xff;
mc->mc_mailbox[0x6] = (f2 >> 16) & 0xff;
mc->mc_mailbox[0x7] = (f2 >> 24) & 0xff;
mc->mc_mailbox[0x8] = f3 & 0xff;
mc->mc_mailbox[0x9] = (f3 >> 8) & 0xff;
mc->mc_mailbox[0xa] = (f3 >> 16) & 0xff;
mc->mc_mailbox[0xb] = (f3 >> 24) & 0xff;
mc->mc_mailbox[0xc] = f4;
}
static __inline void
mlx_make_type5(struct mlx_command *mc,
u_int8_t code,
u_int8_t f1,
u_int8_t f2,
u_int32_t f3,
u_int32_t f4,
u_int8_t f5)
{
mc->mc_mailbox[0x0] = code;
mc->mc_mailbox[0x2] = f1;
mc->mc_mailbox[0x3] = f2;
mc->mc_mailbox[0x4] = f3 & 0xff;
mc->mc_mailbox[0x5] = (f3 >> 8) & 0xff;
mc->mc_mailbox[0x6] = (f3 >> 16) & 0xff;
mc->mc_mailbox[0x7] = (f3 >> 24) & 0xff;
mc->mc_mailbox[0x8] = f4 & 0xff;
mc->mc_mailbox[0x9] = (f4 >> 8) & 0xff;
mc->mc_mailbox[0xa] = (f4 >> 16) & 0xff;
mc->mc_mailbox[0xb] = (f4 >> 24) & 0xff;
mc->mc_mailbox[0xc] = f5;
}
Update to the latest development version of the Mylex driver. Changes in this version include: - Support for version 2.x firmware (2.42 or later recommended). This means we are the only open-source driver supporting these adapters. This code has only been tested on a Digital KZPCA adapter in an Alpha system, but is believed to be correct. NOTE: EISA adapters are not yet supported. - Support the BIOS/Firmware initialisation handshake protocol. This allows the driver to bring the card up to operational state even if the BIOS can't be run (eg. if it's an x86 BIOS in an Alpha system). - A working command pasthrough interface. This allows a user-space configuration tool (under development) to issue arbitrary commands to the controller or to devices in the system.
2000-03-18 02:01:37 +00:00
#endif /* _KERNEL */
Reference in New Issue Copy Permalink