d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
227bb849d3
freebsd-skq/sys/dev/advansys/advansys.c

1410 lines
40 KiB
C
Raw Normal View History

Start each of the license/copyright comments with /*-, minor shuffle of lines
2005-01-06 01:43:34 +00:00
/*-
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
* Generic driver for the Advanced Systems Inc. SCSI controllers
* Product specific probe and attach routines can be found in:
*
* i386/isa/adv_isa.c ABP5140, ABP542, ABP5150, ABP842, ABP852
* i386/eisa/adv_eisa.c ABP742, ABP752
* pci/adv_pci.c ABP920, ABP930, ABP930U, ABP930UA, ABP940, ABP940U,
* ABP940UA, ABP950, ABP960, ABP960U, ABP960UA,
* ABP970, ABP970U
*
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
* Copyright (c) 1996-2000 Justin Gibbs.
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +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. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* 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.
*/
Start each of the license/copyright comments with /*-, minor shuffle of lines
2005-01-06 01:43:34 +00:00
/*-
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
* Ported from:
* advansys.c - Linux Host Driver for AdvanSys SCSI Adapters
*
* Copyright (c) 1995-1997 Advanced System Products, Inc.
* All Rights Reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that redistributions of source
* code retain the above copyright notice and this comment without
* modification.
*/
Use __FBSDID(). Also some minor style cleanups.
2003-08-24 17:55:58 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
#include <sys/param.h>
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
#include <sys/conf.h>
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
Mega busdma API commit. Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg. Lockfunc allows a driver to provide a function for managing its locking semantics while using busdma. At the moment, this is used for the asynchronous busdma_swi and callback mechanism. Two lockfunc implementations are provided: busdma_lock_mutex() performs standard mutex operations on the mutex that is specified from lockfuncarg. dftl_lock() is a panic implementation and is defaulted to when NULL, NULL are passed to bus_dma_tag_create(). The only time that NULL, NULL should ever be used is when the driver ensures that bus_dmamap_load() will not be deferred. Drivers that do not provide their own locking can pass busdma_lock_mutex,&Giant args in order to preserve the former behaviour. sparc64 and powerpc do not provide real busdma_swi functions, so this is largely a noop on those platforms. The busdma_swi on is64 is not properly locked yet, so warnings will be emitted on this platform when busdma callback deferrals happen. If anyone gets panics or warnings from dflt_lock() being called, please let me know right away. Reviewed by: tmm, gibbs
2003-07-01 15:52:06 +00:00
#include <sys/lock.h>
Add MODULE_DEPENDS for cam, pci, mca, eisa and isa where needed. PR: 106543 MFC after: 3 days
2006-12-11 18:28:31 +00:00
#include <sys/module.h>
Mega busdma API commit. Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg. Lockfunc allows a driver to provide a function for managing its locking semantics while using busdma. At the moment, this is used for the asynchronous busdma_swi and callback mechanism. Two lockfunc implementations are provided: busdma_lock_mutex() performs standard mutex operations on the mutex that is specified from lockfuncarg. dftl_lock() is a panic implementation and is defaulted to when NULL, NULL are passed to bus_dma_tag_create(). The only time that NULL, NULL should ever be used is when the driver ensures that bus_dmamap_load() will not be deferred. Drivers that do not provide their own locking can pass busdma_lock_mutex,&Giant args in order to preserve the former behaviour. sparc64 and powerpc do not provide real busdma_swi functions, so this is largely a noop on those platforms. The busdma_swi on is64 is not properly locked yet, so warnings will be emitted on this platform when busdma callback deferrals happen. If anyone gets panics or warnings from dflt_lock() being called, please let me know right away. Reviewed by: tmm, gibbs
2003-07-01 15:52:06 +00:00
#include <sys/mutex.h>
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
#include <machine/bus.h>
Newbusify adv driver. Reviewed by: imp
2000-04-07 11:32:42 +00:00
#include <machine/resource.h>
#include <sys/bus.h>
#include <sys/rman.h>
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
#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>
#include <cam/scsi/scsi_all.h>
#include <cam/scsi/scsi_message.h>
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/pmap.h>
#include <dev/advansys/advansys.h>
static void adv_action(struct cam_sim *sim, union ccb *ccb);
static void adv_execute_ccb(void *arg, bus_dma_segment_t *dm_segs,
int nsegments, int error);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
static void adv_intr_locked(struct adv_softc *adv);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
static void adv_poll(struct cam_sim *sim);
static void adv_run_doneq(struct adv_softc *adv);
static struct adv_ccb_info *
adv_alloc_ccb_info(struct adv_softc *adv);
static void adv_destroy_ccb_info(struct adv_softc *adv,
-Wunused cleanup. Fixed bug in advansys.c where we did not properly clean up ccb_info data structures in softc teardown. Submitted by: Poul-Henning Kamp <phk@freebsd.org>
1998-10-15 23:47:14 +00:00
struct adv_ccb_info *cinfo);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
static __inline struct adv_ccb_info *
adv_get_ccb_info(struct adv_softc *adv);
static __inline void adv_free_ccb_info(struct adv_softc *adv,
struct adv_ccb_info *cinfo);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
static __inline void adv_set_state(struct adv_softc *adv, adv_state state);
static __inline void adv_clear_state(struct adv_softc *adv, union ccb* ccb);
static void adv_clear_state_really(struct adv_softc *adv, union ccb* ccb);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
static __inline struct adv_ccb_info *
adv_get_ccb_info(struct adv_softc *adv)
{
struct adv_ccb_info *cinfo;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if ((cinfo = SLIST_FIRST(&adv->free_ccb_infos)) != NULL) {
SLIST_REMOVE_HEAD(&adv->free_ccb_infos, links);
} else {
cinfo = adv_alloc_ccb_info(adv);
}
return (cinfo);
}
static __inline void
adv_free_ccb_info(struct adv_softc *adv, struct adv_ccb_info *cinfo)
{
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
cinfo->state = ACCB_FREE;
SLIST_INSERT_HEAD(&adv->free_ccb_infos, cinfo, links);
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
static __inline void
adv_set_state(struct adv_softc *adv, adv_state state)
{
if (adv->state == 0)
xpt_freeze_simq(adv->sim, /*count*/1);
adv->state |= state;
}
static __inline void
adv_clear_state(struct adv_softc *adv, union ccb* ccb)
{
if (adv->state != 0)
adv_clear_state_really(adv, ccb);
}
static void
adv_clear_state_really(struct adv_softc *adv, union ccb* ccb)
{
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if ((adv->state & ADV_BUSDMA_BLOCK_CLEARED) != 0)
adv->state &= ~(ADV_BUSDMA_BLOCK_CLEARED|ADV_BUSDMA_BLOCK);
if ((adv->state & ADV_RESOURCE_SHORTAGE) != 0) {
int openings;
openings = adv->max_openings - adv->cur_active - ADV_MIN_FREE_Q;
if (openings >= adv->openings_needed) {
adv->state &= ~ADV_RESOURCE_SHORTAGE;
adv->openings_needed = 0;
}
}
if ((adv->state & ADV_IN_TIMEOUT) != 0) {
struct adv_ccb_info *cinfo;
cinfo = (struct adv_ccb_info *)ccb->ccb_h.ccb_cinfo_ptr;
if ((cinfo->state & ACCB_RECOVERY_CCB) != 0) {
struct ccb_hdr *ccb_h;
/*
* We now traverse our list of pending CCBs
* and reinstate their timeouts.
*/
ccb_h = LIST_FIRST(&adv->pending_ccbs);
while (ccb_h != NULL) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
cinfo = ccb_h->ccb_cinfo_ptr;
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(&cinfo->timer,
SBT_1MS * ccb_h->timeout, 0,
adv_timeout, ccb_h, 0);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
ccb_h = LIST_NEXT(ccb_h, sim_links.le);
}
adv->state &= ~ADV_IN_TIMEOUT;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev, "No longer in timeout\n");
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
}
}
if (adv->state == 0)
ccb->ccb_h.status |= CAM_RELEASE_SIMQ;
}
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
void
adv_map(void *arg, bus_dma_segment_t *segs, int nseg, int error)
{
bus_addr_t* physaddr;
physaddr = (bus_addr_t*)arg;
*physaddr = segs->ds_addr;
}
static void
adv_action(struct cam_sim *sim, union ccb *ccb)
{
struct adv_softc *adv;
CAM_DEBUG(ccb->ccb_h.path, CAM_DEBUG_TRACE, ("adv_action\n"));
adv = (struct adv_softc *)cam_sim_softc(sim);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
switch (ccb->ccb_h.func_code) {
/* Common cases first */
case XPT_SCSI_IO: /* Execute the requested I/O operation */
{
struct ccb_hdr *ccb_h;
struct ccb_scsiio *csio;
struct adv_ccb_info *cinfo;
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
int error;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb_h = &ccb->ccb_h;
csio = &ccb->csio;
cinfo = adv_get_ccb_info(adv);
if (cinfo == NULL)
panic("XXX Handle CCB info error!!!");
ccb_h->ccb_cinfo_ptr = cinfo;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
cinfo->ccb = ccb;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
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
error = bus_dmamap_load_ccb(adv->buffer_dmat,
cinfo->dmamap,
ccb,
adv_execute_ccb,
csio, /*flags*/0);
if (error == EINPROGRESS) {
/*
* So as to maintain ordering, freeze the controller
* queue until our mapping is returned.
*/
adv_set_state(adv, ADV_BUSDMA_BLOCK);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
break;
}
case XPT_RESET_DEV: /* Bus Device Reset the specified SCSI device */
case XPT_TARGET_IO: /* Execute target I/O request */
case XPT_ACCEPT_TARGET_IO: /* Accept Host Target Mode CDB */
case XPT_CONT_TARGET_IO: /* Continue Host Target I/O Connection*/
case XPT_EN_LUN: /* Enable LUN as a target */
case XPT_ABORT: /* Abort the specified CCB */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
#define IS_CURRENT_SETTINGS(c) (c->type == CTS_TYPE_CURRENT_SETTINGS)
#define IS_USER_SETTINGS(c) (c->type == CTS_TYPE_USER_SETTINGS)
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
case XPT_SET_TRAN_SETTINGS:
{
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
struct ccb_trans_settings_scsi *scsi;
struct ccb_trans_settings_spi *spi;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
struct ccb_trans_settings *cts;
target_bit_vector targ_mask;
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
struct adv_transinfo *tconf;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
u_int update_type;
cts = &ccb->cts;
targ_mask = ADV_TID_TO_TARGET_MASK(cts->ccb_h.target_id);
update_type = 0;
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
/*
* The user must specify which type of settings he wishes
* to change.
*/
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
if (IS_CURRENT_SETTINGS(cts) && !IS_USER_SETTINGS(cts)) {
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
tconf = &adv->tinfo[cts->ccb_h.target_id].current;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
update_type |= ADV_TRANS_GOAL;
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
} else if (IS_USER_SETTINGS(cts) && !IS_CURRENT_SETTINGS(cts)) {
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
tconf = &adv->tinfo[cts->ccb_h.target_id].user;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
update_type |= ADV_TRANS_USER;
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
} else {
ccb->ccb_h.status = CAM_REQ_INVALID;
break;
}
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
scsi = &cts->proto_specific.scsi;
spi = &cts->xport_specific.spi;
if ((update_type & ADV_TRANS_GOAL) != 0) {
if ((spi->valid & CTS_SPI_VALID_DISC) != 0) {
if ((spi->flags & CTS_SPI_FLAGS_DISC_ENB) != 0)
adv->disc_enable |= targ_mask;
else
adv->disc_enable &= ~targ_mask;
adv_write_lram_8(adv, ADVV_DISC_ENABLE_B,
adv->disc_enable);
}
if ((scsi->valid & CTS_SCSI_VALID_TQ) != 0) {
if ((scsi->flags & CTS_SCSI_FLAGS_TAG_ENB) != 0)
adv->cmd_qng_enabled |= targ_mask;
else
adv->cmd_qng_enabled &= ~targ_mask;
}
}
if ((update_type & ADV_TRANS_USER) != 0) {
if ((spi->valid & CTS_SPI_VALID_DISC) != 0) {
if ((spi->flags & CTS_SPI_VALID_DISC) != 0)
adv->user_disc_enable |= targ_mask;
else
adv->user_disc_enable &= ~targ_mask;
}
if ((scsi->valid & CTS_SCSI_VALID_TQ) != 0) {
if ((scsi->flags & CTS_SCSI_FLAGS_TAG_ENB) != 0)
adv->user_cmd_qng_enabled |= targ_mask;
else
adv->user_cmd_qng_enabled &= ~targ_mask;
}
}
/*
* If the user specifies either the sync rate, or offset,
* but not both, the unspecified parameter defaults to its
* current value in transfer negotiations.
*/
if (((spi->valid & CTS_SPI_VALID_SYNC_RATE) != 0)
|| ((spi->valid & CTS_SPI_VALID_SYNC_OFFSET) != 0)) {
/*
* If the user provided a sync rate but no offset,
* use the current offset.
*/
if ((spi->valid & CTS_SPI_VALID_SYNC_OFFSET) == 0)
spi->sync_offset = tconf->offset;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
/*
* If the user provided an offset but no sync rate,
* use the current sync rate.
*/
if ((spi->valid & CTS_SPI_VALID_SYNC_RATE) == 0)
spi->sync_period = tconf->period;
adv_period_offset_to_sdtr(adv, &spi->sync_period,
&spi->sync_offset,
cts->ccb_h.target_id);
adv_set_syncrate(adv, /*struct cam_path */NULL,
cts->ccb_h.target_id, spi->sync_period,
spi->sync_offset, update_type);
}
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
case XPT_GET_TRAN_SETTINGS:
/* Get default/user set transfer settings for the target */
{
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
struct ccb_trans_settings_scsi *scsi;
struct ccb_trans_settings_spi *spi;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
struct ccb_trans_settings *cts;
struct adv_transinfo *tconf;
target_bit_vector target_mask;
cts = &ccb->cts;
target_mask = ADV_TID_TO_TARGET_MASK(cts->ccb_h.target_id);
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
scsi = &cts->proto_specific.scsi;
spi = &cts->xport_specific.spi;
cts->protocol = PROTO_SCSI;
cts->protocol_version = SCSI_REV_2;
cts->transport = XPORT_SPI;
cts->transport_version = 2;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
scsi->flags &= ~CTS_SCSI_FLAGS_TAG_ENB;
spi->flags &= ~CTS_SPI_FLAGS_DISC_ENB;
if (cts->type == CTS_TYPE_CURRENT_SETTINGS) {
tconf = &adv->tinfo[cts->ccb_h.target_id].current;
if ((adv->disc_enable & target_mask) != 0)
spi->flags |= CTS_SPI_FLAGS_DISC_ENB;
if ((adv->cmd_qng_enabled & target_mask) != 0)
scsi->flags |= CTS_SCSI_FLAGS_TAG_ENB;
} else {
tconf = &adv->tinfo[cts->ccb_h.target_id].user;
if ((adv->user_disc_enable & target_mask) != 0)
spi->flags |= CTS_SPI_FLAGS_DISC_ENB;
if ((adv->user_cmd_qng_enabled & target_mask) != 0)
scsi->flags |= CTS_SCSI_FLAGS_TAG_ENB;
}
spi->sync_period = tconf->period;
spi->sync_offset = tconf->offset;
spi->bus_width = MSG_EXT_WDTR_BUS_8_BIT;
spi->valid = CTS_SPI_VALID_SYNC_RATE
| CTS_SPI_VALID_SYNC_OFFSET
| CTS_SPI_VALID_BUS_WIDTH
| CTS_SPI_VALID_DISC;
scsi->valid = CTS_SCSI_VALID_TQ;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
case XPT_CALC_GEOMETRY:
{
int extended;
extended = (adv->control & ADV_CNTL_BIOS_GT_1GB) != 0;
Merge common XPT_CALC_GEOMETRY functions into a single convenience function. Devices below may experience a change in geometry. * Due to a bug, aic(4) never used extended geometry. Changes all drives >1G to now use extended translation. * sbp(4) drives exactly 1 GB in size now no longer use extended geometry. * umass(4) drives exactly 1 GB in size now no longer use extended geometry. For all other controllers in this commit, this should be a no-op. Looked over by: scottl
2003-06-14 22:17:41 +00:00
cam_calc_geometry(&ccb->ccg, extended);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
xpt_done(ccb);
break;
}
case XPT_RESET_BUS: /* Reset the specified SCSI bus */
{
adv_stop_execution(adv);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_reset_bus(adv, /*initiate_reset*/TRUE);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_start_execution(adv);
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
case XPT_TERM_IO: /* Terminate the I/O process */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
case XPT_PATH_INQ: /* Path routing inquiry */
{
struct ccb_pathinq *cpi = &ccb->cpi;
cpi->version_num = 1; /* XXX??? */
cpi->hba_inquiry = PI_SDTR_ABLE|PI_TAG_ABLE;
cpi->target_sprt = 0;
cpi->hba_misc = 0;
cpi->hba_eng_cnt = 0;
cpi->max_target = 7;
cpi->max_lun = 7;
cpi->initiator_id = adv->scsi_id;
cpi->bus_id = cam_sim_bus(sim);
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
cpi->base_transfer_speed = 3300;
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, "Advansys", HBA_IDLEN);
strlcpy(cpi->dev_name, cam_sim_name(sim), DEV_IDLEN);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
cpi->unit_number = cam_sim_unit(sim);
cpi->ccb_h.status = CAM_REQ_CMP;
The first of 3 major steps to move the CAM layer forward to using the CAM_NEW_TRAN_CODE that has been in the tree for some years now. This first step consists solely of adding to or correcting CAM_NEW_TRAN_CODE pieces in the kernel source tree such that a both a GENERIC (at least on i386) and a LINT build with CAM_NEW_TRAN_CODE as an option will compile correctly and run (at least with some the h/w I have). After a short settle time, the other pieces (making CAM_NEW_TRAN_CODE the default and updating libcam and camcontrol) will be brought in. This will be an incompatible change in that the size of structures related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change in both size and content. However, basic system operation and basic system utilities work well enough with this change. Reviewed by: freebsd-scsi and specific stakeholders
2006-10-31 05:53:29 +00:00
cpi->transport = XPORT_SPI;
cpi->transport_version = 2;
cpi->protocol = PROTO_SCSI;
cpi->protocol_version = SCSI_REV_2;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
xpt_done(ccb);
break;
}
default:
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
}
}
/*
* Currently, the output of bus_dmammap_load suits our needs just
* fine, but should it change, we'd need to do something here.
*/
#define adv_fixup_dmasegs(adv, dm_segs) (struct adv_sg_entry *)(dm_segs)
static void
adv_execute_ccb(void *arg, bus_dma_segment_t *dm_segs,
int nsegments, int error)
{
struct ccb_scsiio *csio;
struct ccb_hdr *ccb_h;
struct cam_sim *sim;
struct adv_softc *adv;
struct adv_ccb_info *cinfo;
struct adv_scsi_q scsiq;
struct adv_sg_head sghead;
csio = (struct ccb_scsiio *)arg;
ccb_h = &csio->ccb_h;
sim = xpt_path_sim(ccb_h->path);
adv = (struct adv_softc *)cam_sim_softc(sim);
cinfo = (struct adv_ccb_info *)csio->ccb_h.ccb_cinfo_ptr;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
/*
* Setup our done routine to release the simq on
* the next ccb that completes.
*/
if ((adv->state & ADV_BUSDMA_BLOCK) != 0)
adv->state |= ADV_BUSDMA_BLOCK_CLEARED;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if ((ccb_h->flags & CAM_CDB_POINTER) != 0) {
if ((ccb_h->flags & CAM_CDB_PHYS) == 0) {
/* XXX Need phystovirt!!!! */
/* How about pmap_kenter??? */
scsiq.cdbptr = csio->cdb_io.cdb_ptr;
} else {
scsiq.cdbptr = csio->cdb_io.cdb_ptr;
}
} else {
scsiq.cdbptr = csio->cdb_io.cdb_bytes;
}
/*
* Build up the request
*/
scsiq.q1.status = 0;
scsiq.q1.q_no = 0;
scsiq.q1.cntl = 0;
scsiq.q1.sg_queue_cnt = 0;
scsiq.q1.target_id = ADV_TID_TO_TARGET_MASK(ccb_h->target_id);
scsiq.q1.target_lun = ccb_h->target_lun;
scsiq.q1.sense_len = csio->sense_len;
scsiq.q1.extra_bytes = 0;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
scsiq.q2.ccb_index = cinfo - adv->ccb_infos;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
scsiq.q2.target_ix = ADV_TIDLUN_TO_IX(ccb_h->target_id,
ccb_h->target_lun);
scsiq.q2.flag = 0;
scsiq.q2.cdb_len = csio->cdb_len;
if ((ccb_h->flags & CAM_TAG_ACTION_VALID) != 0)
scsiq.q2.tag_code = csio->tag_action;
else
scsiq.q2.tag_code = 0;
scsiq.q2.vm_id = 0;
if (nsegments != 0) {
Bring back bus_dmasync_op_t. It is now a typedef to an int, though the BUS_DMASYNC_ definitions remain as before. The does not change the ABI, and reverts the API to be a bit more compatible and flexible. This has survived a full 'make universe'. Approved by: re (bmah)
2003-05-27 04:59:59 +00:00
bus_dmasync_op_t op;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
scsiq.q1.data_addr = dm_segs->ds_addr;
scsiq.q1.data_cnt = dm_segs->ds_len;
if (nsegments > 1) {
scsiq.q1.cntl |= QC_SG_HEAD;
sghead.entry_cnt
= sghead.entry_to_copy
= nsegments;
sghead.res = 0;
sghead.sg_list = adv_fixup_dmasegs(adv, dm_segs);
scsiq.sg_head = &sghead;
} else {
scsiq.sg_head = NULL;
}
if ((ccb_h->flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_PREREAD;
else
op = BUS_DMASYNC_PREWRITE;
bus_dmamap_sync(adv->buffer_dmat, cinfo->dmamap, op);
} else {
scsiq.q1.data_addr = 0;
scsiq.q1.data_cnt = 0;
scsiq.sg_head = NULL;
}
Don't set the active flag on a transaction resource until it has been queued. Perform dma segment setup outside of splcam protection as this can take some time and the protection is not necessary. Inline a function. Clean up some whitespace.
1998-10-07 03:32:57 +00:00
/*
* Last time we need to check if this SCB needs to
* be aborted.
*/
if (ccb_h->status != CAM_REQ_INPROG) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if (nsegments != 0)
Don't set the active flag on a transaction resource until it has been queued. Perform dma segment setup outside of splcam protection as this can take some time and the protection is not necessary. Inline a function. Clean up some whitespace.
1998-10-07 03:32:57 +00:00
bus_dmamap_unload(adv->buffer_dmat, cinfo->dmamap);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_clear_state(adv, (union ccb *)csio);
Don't set the active flag on a transaction resource until it has been queued. Perform dma segment setup outside of splcam protection as this can take some time and the protection is not necessary. Inline a function. Clean up some whitespace.
1998-10-07 03:32:57 +00:00
adv_free_ccb_info(adv, cinfo);
xpt_done((union ccb *)csio);
return;
}
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if (adv_execute_scsi_queue(adv, &scsiq, csio->dxfer_len) != 0) {
/* Temporary resource shortage */
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_set_state(adv, ADV_RESOURCE_SHORTAGE);
if (nsegments != 0)
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dmamap_unload(adv->buffer_dmat, cinfo->dmamap);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
csio->ccb_h.status = CAM_REQUEUE_REQ;
adv_clear_state(adv, (union ccb *)csio);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_free_ccb_info(adv, cinfo);
xpt_done((union ccb *)csio);
return;
}
Don't set the active flag on a transaction resource until it has been queued. Perform dma segment setup outside of splcam protection as this can take some time and the protection is not necessary. Inline a function. Clean up some whitespace.
1998-10-07 03:32:57 +00:00
cinfo->state |= ACCB_ACTIVE;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb_h->status |= CAM_SIM_QUEUED;
LIST_INSERT_HEAD(&adv->pending_ccbs, ccb_h, sim_links.le);
/* Schedule our 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(&cinfo->timer, SBT_1MS * ccb_h->timeout, 0,
adv_timeout, csio, 0);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
static struct adv_ccb_info *
adv_alloc_ccb_info(struct adv_softc *adv)
{
int error;
struct adv_ccb_info *cinfo;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
cinfo = &adv->ccb_infos[adv->ccb_infos_allocated];
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
cinfo->state = ACCB_FREE;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
callout_init_mtx(&cinfo->timer, &adv->lock, 0);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
error = bus_dmamap_create(adv->buffer_dmat, /*flags*/0,
&cinfo->dmamap);
if (error != 0) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev, "Unable to allocate CCB info "
"dmamap - error %d\n", error);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
return (NULL);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv->ccb_infos_allocated++;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (cinfo);
}
static void
adv_destroy_ccb_info(struct adv_softc *adv, struct adv_ccb_info *cinfo)
{
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
callout_drain(&cinfo->timer);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dmamap_destroy(adv->buffer_dmat, cinfo->dmamap);
}
void
adv_timeout(void *arg)
{
union ccb *ccb;
struct adv_softc *adv;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
struct adv_ccb_info *cinfo, *cinfo2;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb = (union ccb *)arg;
adv = (struct adv_softc *)xpt_path_sim(ccb->ccb_h.path)->softc;
cinfo = (struct adv_ccb_info *)ccb->ccb_h.ccb_cinfo_ptr;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
xpt_print_path(ccb->ccb_h.path);
printf("Timed out\n");
/* Have we been taken care of already?? */
if (cinfo == NULL || cinfo->state == ACCB_FREE) {
return;
}
adv_stop_execution(adv);
if ((cinfo->state & ACCB_ABORT_QUEUED) == 0) {
struct ccb_hdr *ccb_h;
/*
* In order to simplify the recovery process, we ask the XPT
* layer to halt the queue of new transactions and we traverse
* the list of pending CCBs and remove their timeouts. This
* means that the driver attempts to clear only one error
* condition at a time. In general, timeouts that occur
* close together are related anyway, so there is no benefit
sys/dev: minor spelling fixes. Most affect comments, very few have user-visible effects.
2016-05-03 03:41:25 +00:00
* in attempting to handle errors in parallel. Timeouts will
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
* be reinstated when the recovery process ends.
*/
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_set_state(adv, ADV_IN_TIMEOUT);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* This CCB is the CCB representing our recovery actions */
cinfo->state |= ACCB_RECOVERY_CCB|ACCB_ABORT_QUEUED;
ccb_h = LIST_FIRST(&adv->pending_ccbs);
while (ccb_h != NULL) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
cinfo2 = ccb_h->ccb_cinfo_ptr;
callout_stop(&cinfo2->timer);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb_h = LIST_NEXT(ccb_h, sim_links.le);
}
/* XXX Should send a BDR */
/* Attempt an abort as our first tact */
xpt_print_path(ccb->ccb_h.path);
printf("Attempting abort\n");
adv_abort_ccb(adv, ccb->ccb_h.target_id,
ccb->ccb_h.target_lun, ccb,
CAM_CMD_TIMEOUT, /*queued_only*/FALSE);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
callout_reset(&cinfo->timer, 2 * hz, adv_timeout, ccb);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
} else {
/* Our attempt to perform an abort failed, go for a reset */
xpt_print_path(ccb->ccb_h.path);
printf("Resetting bus\n");
ccb->ccb_h.status &= ~CAM_STATUS_MASK;
ccb->ccb_h.status |= CAM_CMD_TIMEOUT;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_reset_bus(adv, /*initiate_reset*/TRUE);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv_start_execution(adv);
}
struct adv_softc *
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
adv_alloc(device_t dev, struct resource *res, long offset)
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
{
Newbusify adv driver. Reviewed by: imp
2000-04-07 11:32:42 +00:00
struct adv_softc *adv = device_get_softc(dev);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/*
* Allocate a storage area for us
*/
LIST_INIT(&adv->pending_ccbs);
SLIST_INIT(&adv->free_ccb_infos);
Newbusify adv driver. Reviewed by: imp
2000-04-07 11:32:42 +00:00
adv->dev = dev;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
adv->res = res;
adv->reg_off = offset;
mtx_init(&adv->lock, "adv", NULL, MTX_DEF);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return(adv);
}
void
adv_free(struct adv_softc *adv)
{
switch (adv->init_level) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 6:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
{
struct adv_ccb_info *cinfo;
while ((cinfo = SLIST_FIRST(&adv->free_ccb_infos)) != NULL) {
SLIST_REMOVE_HEAD(&adv->free_ccb_infos, links);
-Wunused cleanup. Fixed bug in advansys.c where we did not properly clean up ccb_info data structures in softc teardown. Submitted by: Poul-Henning Kamp <phk@freebsd.org>
1998-10-15 23:47:14 +00:00
adv_destroy_ccb_info(adv, cinfo);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
bus_dmamap_unload(adv->sense_dmat, adv->sense_dmamap);
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 5:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dmamem_free(adv->sense_dmat, adv->sense_buffers,
adv->sense_dmamap);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 4:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dma_tag_destroy(adv->sense_dmat);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 3:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dma_tag_destroy(adv->buffer_dmat);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 2:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
bus_dma_tag_destroy(adv->parent_dmat);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
case 1:
- Check if the ccb_infos in the softc is NULL before trying to free it in adv_free() as the ISA probe routine doesn't malloc() ccb_infos but does call adv_free(). - Release the ISA-only overrun DMA tags, bufs, and maps if the probe fails. Tested by: rwatson
2001-02-08 00:00:12 +00:00
if (adv->ccb_infos != NULL)
free(adv->ccb_infos, M_DEVBUF);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
case 0:
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_destroy(&adv->lock);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
break;
}
}
int
adv_init(struct adv_softc *adv)
{
struct adv_eeprom_config eeprom_config;
int checksum, i;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
int max_sync;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
u_int16_t config_lsw;
u_int16_t config_msw;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_lock(&adv->lock);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_lib_init(adv);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
/*
* Stop script execution.
*/
adv_write_lram_16(adv, ADV_HALTCODE_W, 0x00FE);
adv_stop_execution(adv);
if (adv_stop_chip(adv) == 0 || adv_is_chip_halted(adv) == 0) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_unlock(&adv->lock);
device_printf(adv->dev,
"Unable to halt adapter. Initialization failed\n");
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (1);
}
ADV_OUTW(adv, ADV_REG_PROG_COUNTER, ADV_MCODE_START_ADDR);
if (ADV_INW(adv, ADV_REG_PROG_COUNTER) != ADV_MCODE_START_ADDR) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_unlock(&adv->lock);
device_printf(adv->dev,
"Unable to set program counter. Initialization failed\n");
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (1);
}
config_msw = ADV_INW(adv, ADV_CONFIG_MSW);
config_lsw = ADV_INW(adv, ADV_CONFIG_LSW);
if ((config_msw & ADV_CFG_MSW_CLR_MASK) != 0) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
config_msw &= ~ADV_CFG_MSW_CLR_MASK;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/*
* XXX The Linux code flags this as an error,
* but what should we report to the user???
* It seems that clearing the config register
* makes this error recoverable.
*/
ADV_OUTW(adv, ADV_CONFIG_MSW, config_msw);
}
/* Suck in the configuration from the EEProm */
checksum = adv_get_eeprom_config(adv, &eeprom_config);
if (ADV_INW(adv, ADV_CHIP_STATUS) & ADV_CSW_AUTO_CONFIG) {
/*
* XXX The Linux code sets a warning level for this
* condition, yet nothing of meaning is printed to
* the user. What does this mean???
*/
if (adv->chip_version == 3) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if (eeprom_config.cfg_lsw != config_lsw)
eeprom_config.cfg_lsw = config_lsw;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if (eeprom_config.cfg_msw != config_msw) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
eeprom_config.cfg_msw = config_msw;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
}
}
if (checksum == eeprom_config.chksum) {
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* Range/Sanity checking */
if (eeprom_config.max_total_qng < ADV_MIN_TOTAL_QNG) {
eeprom_config.max_total_qng = ADV_MIN_TOTAL_QNG;
}
if (eeprom_config.max_total_qng > ADV_MAX_TOTAL_QNG) {
eeprom_config.max_total_qng = ADV_MAX_TOTAL_QNG;
}
if (eeprom_config.max_tag_qng > eeprom_config.max_total_qng) {
eeprom_config.max_tag_qng = eeprom_config.max_total_qng;
}
if (eeprom_config.max_tag_qng < ADV_MIN_TAG_Q_PER_DVC) {
eeprom_config.max_tag_qng = ADV_MIN_TAG_Q_PER_DVC;
}
adv->max_openings = eeprom_config.max_total_qng;
adv->user_disc_enable = eeprom_config.disc_enable;
adv->user_cmd_qng_enabled = eeprom_config.use_cmd_qng;
adv->isa_dma_speed = EEPROM_DMA_SPEED(eeprom_config);
adv->scsi_id = EEPROM_SCSIID(eeprom_config) & ADV_MAX_TID;
EEPROM_SET_SCSIID(eeprom_config, adv->scsi_id);
adv->control = eeprom_config.cntl;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
for (i = 0; i <= ADV_MAX_TID; i++) {
u_int8_t sync_data;
if ((eeprom_config.init_sdtr & (0x1 << i)) == 0)
sync_data = 0;
else
sync_data = eeprom_config.sdtr_data[i];
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_sdtr_to_period_offset(adv,
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
sync_data,
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
&adv->tinfo[i].user.period,
&adv->tinfo[i].user.offset,
i);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
}
config_lsw = eeprom_config.cfg_lsw;
eeprom_config.cfg_msw = config_msw;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
} else {
u_int8_t sync_data;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev, "Warning EEPROM Checksum mismatch. "
"Using default device parameters\n");
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* Set reasonable defaults since we can't read the EEPROM */
adv->isa_dma_speed = /*ADV_DEF_ISA_DMA_SPEED*/1;
adv->max_openings = ADV_DEF_MAX_TOTAL_QNG;
adv->disc_enable = TARGET_BIT_VECTOR_SET;
adv->user_disc_enable = TARGET_BIT_VECTOR_SET;
adv->cmd_qng_enabled = TARGET_BIT_VECTOR_SET;
adv->user_cmd_qng_enabled = TARGET_BIT_VECTOR_SET;
adv->scsi_id = 7;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv->control = 0xFFFF;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if (adv->chip_version == ADV_CHIP_VER_PCI_ULTRA_3050)
/* Default to no Ultra to support the 3030 */
adv->control &= ~ADV_CNTL_SDTR_ENABLE_ULTRA;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
sync_data = ADV_DEF_SDTR_OFFSET | (ADV_DEF_SDTR_INDEX << 4);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
for (i = 0; i <= ADV_MAX_TID; i++) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_sdtr_to_period_offset(adv, sync_data,
&adv->tinfo[i].user.period,
&adv->tinfo[i].user.offset,
i);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
}
config_lsw |= ADV_CFG_LSW_SCSI_PARITY_ON;
}
config_msw &= ~ADV_CFG_MSW_CLR_MASK;
config_lsw |= ADV_CFG_LSW_HOST_INT_ON;
if ((adv->type & (ADV_PCI|ADV_ULTRA)) == (ADV_PCI|ADV_ULTRA)
&& (adv->control & ADV_CNTL_SDTR_ENABLE_ULTRA) == 0)
/* 25ns or 10MHz */
max_sync = 25;
else
/* Unlimited */
max_sync = 0;
for (i = 0; i <= ADV_MAX_TID; i++) {
if (adv->tinfo[i].user.period < max_sync)
adv->tinfo[i].user.period = max_sync;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if (adv_test_external_lram(adv) == 0) {
if ((adv->type & (ADV_PCI|ADV_ULTRA)) == (ADV_PCI|ADV_ULTRA)) {
eeprom_config.max_total_qng =
ADV_MAX_PCI_ULTRA_INRAM_TOTAL_QNG;
eeprom_config.max_tag_qng =
ADV_MAX_PCI_ULTRA_INRAM_TAG_QNG;
} else {
eeprom_config.cfg_msw |= 0x0800;
config_msw |= 0x0800;
eeprom_config.max_total_qng =
ADV_MAX_PCI_INRAM_TOTAL_QNG;
eeprom_config.max_tag_qng = ADV_MAX_INRAM_TAG_QNG;
}
adv->max_openings = eeprom_config.max_total_qng;
}
ADV_OUTW(adv, ADV_CONFIG_MSW, config_msw);
ADV_OUTW(adv, ADV_CONFIG_LSW, config_lsw);
#if 0
/*
* Don't write the eeprom data back for now.
* I'd rather not mess up the user's card. We also don't
* fully sanitize the eeprom settings above for the write-back
* to be 100% correct.
*/
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if (adv_set_eeprom_config(adv, &eeprom_config) != 0)
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev,
"WARNING! Failure writing to EEPROM.\n");
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
#endif
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_set_chip_scsiid(adv, adv->scsi_id);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (adv_init_lram_and_mcode(adv)) {
mtx_unlock(&adv->lock);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (1);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
}
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv->disc_enable = adv->user_disc_enable;
adv_write_lram_8(adv, ADVV_DISC_ENABLE_B, adv->disc_enable);
for (i = 0; i <= ADV_MAX_TID; i++) {
/*
* Start off in async mode.
*/
adv_set_syncrate(adv, /*struct cam_path */NULL,
i, /*period*/0, /*offset*/0,
ADV_TRANS_CUR);
/*
* Enable the use of tagged commands on all targets.
* This allows the kernel driver to make up it's own mind
* as it sees fit to tag queue instead of having the
* firmware try and second guess the tag_code settins.
*/
adv_write_lram_8(adv, ADVV_MAX_DVC_QNG_BEG + i,
adv->max_openings);
}
adv_write_lram_8(adv, ADVV_USE_TAGGED_QNG_B, TARGET_BIT_VECTOR_SET);
adv_write_lram_8(adv, ADVV_CAN_TAGGED_QNG_B, TARGET_BIT_VECTOR_SET);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev,
"AdvanSys %s Host Adapter, SCSI ID %d, queue depth %d\n",
(adv->type & ADV_ULTRA) && (max_sync == 0)
? "Ultra SCSI" : "SCSI",
adv->scsi_id, adv->max_openings);
mtx_unlock(&adv->lock);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (0);
}
void
adv_intr(void *arg)
{
struct adv_softc *adv;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
adv = arg;
mtx_lock(&adv->lock);
adv_intr_locked(adv);
mtx_unlock(&adv->lock);
}
void
adv_intr_locked(struct adv_softc *adv)
{
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
u_int16_t chipstat;
u_int16_t saved_ram_addr;
u_int8_t ctrl_reg;
u_int8_t saved_ctrl_reg;
u_int8_t host_flag;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
chipstat = ADV_INW(adv, ADV_CHIP_STATUS);
/* Is it for us? */
if ((chipstat & (ADV_CSW_INT_PENDING|ADV_CSW_SCSI_RESET_LATCH)) == 0)
return;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ctrl_reg = ADV_INB(adv, ADV_CHIP_CTRL);
saved_ctrl_reg = ctrl_reg & (~(ADV_CC_SCSI_RESET | ADV_CC_CHIP_RESET |
ADV_CC_SINGLE_STEP | ADV_CC_DIAG |
ADV_CC_TEST));
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if ((chipstat & (ADV_CSW_SCSI_RESET_LATCH|ADV_CSW_SCSI_RESET_ACTIVE))) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev, "Detected Bus Reset\n");
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_reset_bus(adv, /*initiate_reset*/FALSE);
return;
}
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if ((chipstat & ADV_CSW_INT_PENDING) != 0) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
saved_ram_addr = ADV_INW(adv, ADV_LRAM_ADDR);
host_flag = adv_read_lram_8(adv, ADVV_HOST_FLAG_B);
adv_write_lram_8(adv, ADVV_HOST_FLAG_B,
host_flag | ADV_HOST_FLAG_IN_ISR);
adv_ack_interrupt(adv);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if ((chipstat & ADV_CSW_HALTED) != 0
&& (ctrl_reg & ADV_CC_SINGLE_STEP) != 0) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_isr_chip_halted(adv);
saved_ctrl_reg &= ~ADV_CC_HALT;
} else {
adv_run_doneq(adv);
}
ADV_OUTW(adv, ADV_LRAM_ADDR, saved_ram_addr);
#ifdef DIAGNOSTIC
if (ADV_INW(adv, ADV_LRAM_ADDR) != saved_ram_addr)
panic("adv_intr: Unable to set LRAM addr");
#endif
adv_write_lram_8(adv, ADVV_HOST_FLAG_B, host_flag);
}
ADV_OUTB(adv, ADV_CHIP_CTRL, saved_ctrl_reg);
}
Be consistent about "static" functions: if the function is marked static in its prototype, mark it static at the definition too. Inspired by: FlexeLint warning #512
2002-09-28 17:15:38 +00:00
static void
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_run_doneq(struct adv_softc *adv)
{
struct adv_q_done_info scsiq;
u_int doneq_head;
u_int done_qno;
doneq_head = adv_read_lram_16(adv, ADVV_DONE_Q_TAIL_W) & 0xFF;
done_qno = adv_read_lram_8(adv, ADV_QNO_TO_QADDR(doneq_head)
+ ADV_SCSIQ_B_FWD);
while (done_qno != ADV_QLINK_END) {
union ccb* ccb;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
struct adv_ccb_info *cinfo;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
u_int done_qaddr;
u_int sg_queue_cnt;
done_qaddr = ADV_QNO_TO_QADDR(done_qno);
/* Pull status from this request */
sg_queue_cnt = adv_copy_lram_doneq(adv, done_qaddr, &scsiq,
adv->max_dma_count);
/* Mark it as free */
adv_write_lram_8(adv, done_qaddr + ADV_SCSIQ_B_STATUS,
Correct compilation with option DIAGNOSTIC set. Upgrade to the latest firmware.
1998-09-20 05:04:05 +00:00
scsiq.q_status & ~(QS_READY|QS_ABORTED));
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* Process request based on retrieved info */
if ((scsiq.cntl & QC_SG_HEAD) != 0) {
u_int i;
/*
* S/G based request. Free all of the queue
* structures that contained S/G information.
*/
for (i = 0; i < sg_queue_cnt; i++) {
done_qno = adv_read_lram_8(adv, done_qaddr
+ ADV_SCSIQ_B_FWD);
#ifdef DIAGNOSTIC
Correct compilation with option DIAGNOSTIC set. Upgrade to the latest firmware.
1998-09-20 05:04:05 +00:00
if (done_qno == ADV_QLINK_END) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
panic("adv_qdone: Corrupted SG "
"list encountered");
}
#endif
Correct compilation with option DIAGNOSTIC set. Upgrade to the latest firmware.
1998-09-20 05:04:05 +00:00
done_qaddr = ADV_QNO_TO_QADDR(done_qno);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* Mark SG queue as free */
adv_write_lram_8(adv, done_qaddr
+ ADV_SCSIQ_B_STATUS, QS_FREE);
}
} else
sg_queue_cnt = 0;
#ifdef DIAGNOSTIC
Correct compilation with option DIAGNOSTIC set. Upgrade to the latest firmware.
1998-09-20 05:04:05 +00:00
if (adv->cur_active < (sg_queue_cnt + 1))
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
panic("adv_qdone: Attempting to free more "
"queues than are active");
#endif
adv->cur_active -= sg_queue_cnt + 1;
if ((scsiq.q_status != QS_DONE)
&& (scsiq.q_status & QS_ABORTED) == 0)
panic("adv_qdone: completed scsiq with unknown status");
scsiq.remain_bytes += scsiq.extra_bytes;
if ((scsiq.d3.done_stat == QD_WITH_ERROR) &&
(scsiq.d3.host_stat == QHSTA_M_DATA_OVER_RUN)) {
if ((scsiq.cntl & (QC_DATA_IN|QC_DATA_OUT)) == 0) {
scsiq.d3.done_stat = QD_NO_ERROR;
scsiq.d3.host_stat = QHSTA_NO_ERROR;
}
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
cinfo = &adv->ccb_infos[scsiq.d2.ccb_index];
ccb = cinfo->ccb;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->csio.resid = scsiq.remain_bytes;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_done(adv, ccb,
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
scsiq.d3.done_stat, scsiq.d3.host_stat,
scsiq.d3.scsi_stat, scsiq.q_no);
doneq_head = done_qno;
Don't set the active flag on a transaction resource until it has been queued. Perform dma segment setup outside of splcam protection as this can take some time and the protection is not necessary. Inline a function. Clean up some whitespace.
1998-10-07 03:32:57 +00:00
done_qno = adv_read_lram_8(adv, done_qaddr + ADV_SCSIQ_B_FWD);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv_write_lram_16(adv, ADVV_DONE_Q_TAIL_W, doneq_head);
}
void
adv_done(struct adv_softc *adv, union ccb *ccb, u_int done_stat,
u_int host_stat, u_int scsi_status, u_int q_no)
{
struct adv_ccb_info *cinfo;
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
if (!dumping)
mtx_assert(&adv->lock, MA_OWNED);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
cinfo = (struct adv_ccb_info *)ccb->ccb_h.ccb_cinfo_ptr;
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
LIST_REMOVE(&ccb->ccb_h, sim_links.le);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
callout_stop(&cinfo->timer);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
Bring back bus_dmasync_op_t. It is now a typedef to an int, though the BUS_DMASYNC_ definitions remain as before. The does not change the ABI, and reverts the API to be a bit more compatible and flexible. This has survived a full 'make universe'. Approved by: re (bmah)
2003-05-27 04:59:59 +00:00
bus_dmasync_op_t op;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_POSTREAD;
else
op = BUS_DMASYNC_POSTWRITE;
bus_dmamap_sync(adv->buffer_dmat, cinfo->dmamap, op);
bus_dmamap_unload(adv->buffer_dmat, cinfo->dmamap);
}
switch (done_stat) {
case QD_NO_ERROR:
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
if (host_stat == QHSTA_NO_ERROR) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->ccb_h.status = CAM_REQ_CMP;
break;
}
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
xpt_print_path(ccb->ccb_h.path);
printf("adv_done - queue done without error, "
"but host status non-zero(%x)\n", host_stat);
/*FALLTHROUGH*/
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
case QD_WITH_ERROR:
switch (host_stat) {
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
case QHSTA_M_TARGET_STATUS_BUSY:
case QHSTA_M_BAD_QUEUE_FULL_OR_BUSY:
/*
* Assume that if we were a tagged transaction
* the target reported queue full. Otherwise,
* report busy. The firmware really should just
* pass the original status back up to us even
* if it thinks the target was in error for
* returning this status as no other transactions
* from this initiator are in effect, but this
* ignores multi-initiator setups and there is
* evidence that the firmware gets its per-device
sys/dev: minor spelling fixes. Most affect comments, very few have user-visible effects.
2016-05-03 03:41:25 +00:00
* transaction counts screwed up occasionally.
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
*/
ccb->ccb_h.status |= CAM_SCSI_STATUS_ERROR;
if ((ccb->ccb_h.flags & CAM_TAG_ACTION_VALID) != 0
&& host_stat != QHSTA_M_TARGET_STATUS_BUSY)
scsi_status = SCSI_STATUS_QUEUE_FULL;
else
scsi_status = SCSI_STATUS_BUSY;
adv_abort_ccb(adv, ccb->ccb_h.target_id,
ccb->ccb_h.target_lun,
/*ccb*/NULL, CAM_REQUEUE_REQ,
/*queued_only*/TRUE);
/*FALLTHROUGH*/
case QHSTA_M_NO_AUTO_REQ_SENSE:
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
case QHSTA_NO_ERROR:
ccb->csio.scsi_status = scsi_status;
switch (scsi_status) {
case SCSI_STATUS_CHECK_COND:
case SCSI_STATUS_CMD_TERMINATED:
ccb->ccb_h.status |= CAM_AUTOSNS_VALID;
/* Structure copy */
ccb->csio.sense_data =
adv->sense_buffers[q_no - 1];
/* FALLTHROUGH */
case SCSI_STATUS_BUSY:
case SCSI_STATUS_RESERV_CONFLICT:
case SCSI_STATUS_QUEUE_FULL:
case SCSI_STATUS_COND_MET:
case SCSI_STATUS_INTERMED:
case SCSI_STATUS_INTERMED_COND_MET:
ccb->ccb_h.status |= CAM_SCSI_STATUS_ERROR;
break;
case SCSI_STATUS_OK:
ccb->ccb_h.status |= CAM_REQ_CMP;
break;
}
break;
case QHSTA_M_SEL_TIMEOUT:
ccb->ccb_h.status = CAM_SEL_TIMEOUT;
break;
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
case QHSTA_M_DATA_OVER_RUN:
ccb->ccb_h.status = CAM_DATA_RUN_ERR;
break;
case QHSTA_M_UNEXPECTED_BUS_FREE:
ccb->ccb_h.status = CAM_UNEXP_BUSFREE;
break;
case QHSTA_M_BAD_BUS_PHASE_SEQ:
ccb->ccb_h.status = CAM_SEQUENCE_FAIL;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
break;
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
case QHSTA_M_BAD_CMPL_STATUS_IN:
/* No command complete after a status message */
ccb->ccb_h.status = CAM_SEQUENCE_FAIL;
break;
case QHSTA_D_EXE_SCSI_Q_BUSY_TIMEOUT:
case QHSTA_M_WTM_TIMEOUT:
case QHSTA_M_HUNG_REQ_SCSI_BUS_RESET:
/* The SCSI bus hung in a phase */
ccb->ccb_h.status = CAM_SEQUENCE_FAIL;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_reset_bus(adv, /*initiate_reset*/TRUE);
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
break;
Handle the case when auto sense retrieval fails. Give automatic request sense operations a 5 second timeout.
1999-04-19 21:27:36 +00:00
case QHSTA_M_AUTO_REQ_SENSE_FAIL:
ccb->ccb_h.status = CAM_AUTOSENSE_FAIL;
break;
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
case QHSTA_D_QDONE_SG_LIST_CORRUPTED:
case QHSTA_D_ASC_DVC_ERROR_CODE_SET:
case QHSTA_D_HOST_ABORT_FAILED:
case QHSTA_D_EXE_SCSI_Q_FAILED:
case QHSTA_D_ASPI_NO_BUF_POOL:
case QHSTA_M_BAD_TAG_CODE:
case QHSTA_D_LRAM_CMP_ERROR:
case QHSTA_M_MICRO_CODE_ERROR_HALT:
default:
panic("%s: Unhandled Host status error %x",
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_get_nameunit(adv->dev), host_stat);
Beef up the error handling routine to handle more errors. Compensate for a bug in the AdvanSys firmware where a valid queue full condition can be reported via a different error code.
1999-04-07 22:59:12 +00:00
/* NOTREACHED */
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
break;
case QD_ABORTED_BY_HOST:
/* Don't clobber any, more explicit, error codes we've set */
if ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_INPROG)
ccb->ccb_h.status = CAM_REQ_ABORTED;
break;
default:
Correct the reporting of the queue full condition so that the XPT layer can properly throttle tags. Add diagnostic printfs for firmware result codes that we encounter but don't know anything about.
1998-10-29 17:41:34 +00:00
xpt_print_path(ccb->ccb_h.path);
printf("adv_done - queue done with unknown status %x:%x\n",
done_stat, host_stat);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->ccb_h.status = CAM_REQ_CMP_ERR;
break;
}
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv_clear_state(adv, ccb);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if ((ccb->ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP
&& (ccb->ccb_h.status & CAM_DEV_QFRZN) == 0) {
xpt_freeze_devq(ccb->ccb_h.path, /*count*/1);
ccb->ccb_h.status |= CAM_DEV_QFRZN;
}
adv_free_ccb_info(adv, cinfo);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
/*
* Null this out so that we catch driver bugs that cause a
* ccb to be completed twice.
*/
ccb->ccb_h.ccb_cinfo_ptr = NULL;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
ccb->ccb_h.status &= ~CAM_SIM_QUEUED;
xpt_done(ccb);
}
/*
* Function to poll for command completion when
* interrupts are disabled (crash dumps)
*/
static void
adv_poll(struct cam_sim *sim)
{
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
adv_intr_locked(cam_sim_softc(sim));
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
/*
* Attach all the sub-devices we can find
*/
int
adv_attach(adv)
struct adv_softc *adv;
{
struct ccb_setasync csa;
struct cam_devq *devq;
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
int max_sg;
/*
* Allocate an array of ccb mapping structures. We put the
* index of the ccb_info structure into the queue representing
* a transaction and use it for mapping the queue to the
* upper level SCSI transaction it represents.
*/
adv->ccb_infos = malloc(sizeof(*adv->ccb_infos) * adv->max_openings,
M_DEVBUF, M_NOWAIT);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
if (adv->ccb_infos == NULL)
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENOMEM);
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
adv->init_level++;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/*
* Create our DMA tags. These tags define the kinds of device
Fix typo (accessable --> accessible). PR: 18588 Submitted by: Anatoly Vorobey <mellon@pobox.com> Reviewed by: asmodai
2000-06-14 17:53:40 +00:00
* accessible memory allocations and memory mappings we will
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
* need to perform during normal operation.
*
* Unless we need to further restrict the allocation, we rely
* on the restrictions of the parent dmat, hence the common
* use of MAXADDR and MAXSIZE.
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
*
* The ASC boards use chains of "queues" (the transactional
* resources on the board) to represent long S/G lists.
* The first queue represents the command and holds a
* single address and data pair. The queues that follow
* can each hold ADV_SG_LIST_PER_Q entries. Given the
* total number of queues, we can express the largest
* transaction we can map. We reserve a few queues for
* error recovery. Take those into account as well.
*
* There is a way to take an interrupt to download the
* next batch of S/G entries if there are more than 255
* of them (the counter in the queue structure is a u_int8_t).
* We don't use this feature, so limit the S/G list size
* accordingly.
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
*/
adv_pci.c: Update list of supported products. Adjust probe message to include the ASC3030. advansys.c: Fix a long standing bug in the error recovery strategy. In order to keep recovery simple, we freeze the SIMQ, stopping the XPT from submitting new requests. Unfortunately, we also will freeze the SIMQ if bus_dmamap_load blocks or we run out of controller resources. On cards with limited resources it was possible to freeze the SIM a second time and never unfreeze it. Now we more carefully track our exception state so we never freeze the SIMQ more than once. Don't rely on pointers fitting in a 32bit field stored in the per-transaction data structures on the card. Use an index to an array of transaction mapping structures instead. This should allow this driver to work on the Alpha. Deal with the ASC3030 which is almost idistinguishable from the ASC3050. Unfortunately the ASC3030 does not work at Ultra speeds, so if we can't find an eeprom, we must assume that ultra is disabled. The SIIG cards using the 3030 do not have eeproms. As a side effect, we now honor the ultra disable bit in the eeprom if it is present. Don't bother attempting to write corrected eeprom data back to the eeprom. We can function just fine if the data is corrupted and I'd rather not risk messing up the user's eeprom. Modify the interrupt handler to catch latched external bus rests. Dynamically determine the maximum number of S/G elements we can map at a single time. The nature of the firmware interface for these cards makes this value dependent on the number of "queues" the card can support. advlib.c: advlib.h: advmcode.c: advmcode.h: Synchronize with the latest firmware image released in the Linux Advansys driver.
2000-01-14 03:33:38 +00:00
max_sg = (adv->max_openings - ADV_MIN_FREE_Q - 1) * ADV_SG_LIST_PER_Q;
if (max_sg > 255)
max_sg = 255;
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/* DMA tag for mapping buffers into device visible space. */
Clean up argument comments for bus_dma_tag_create() calls.
2003-03-29 09:46:10 +00:00
if (bus_dma_tag_create(
/* parent */ adv->parent_dmat,
/* alignment */ 1,
/* boundary */ 0,
/* lowaddr */ BUS_SPACE_MAXADDR,
/* highaddr */ BUS_SPACE_MAXADDR,
/* filter */ NULL,
/* filterarg */ 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
/* maxsize */ ADV_MAXPHYS,
Clean up argument comments for bus_dma_tag_create() calls.
2003-03-29 09:46:10 +00:00
/* nsegments */ max_sg,
/* maxsegsz */ BUS_SPACE_MAXSIZE_32BIT,
/* flags */ BUS_DMA_ALLOCNOW,
Mega busdma API commit. Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg. Lockfunc allows a driver to provide a function for managing its locking semantics while using busdma. At the moment, this is used for the asynchronous busdma_swi and callback mechanism. Two lockfunc implementations are provided: busdma_lock_mutex() performs standard mutex operations on the mutex that is specified from lockfuncarg. dftl_lock() is a panic implementation and is defaulted to when NULL, NULL are passed to bus_dma_tag_create(). The only time that NULL, NULL should ever be used is when the driver ensures that bus_dmamap_load() will not be deferred. Drivers that do not provide their own locking can pass busdma_lock_mutex,&Giant args in order to preserve the former behaviour. sparc64 and powerpc do not provide real busdma_swi functions, so this is largely a noop on those platforms. The busdma_swi on is64 is not properly locked yet, so warnings will be emitted on this platform when busdma callback deferrals happen. If anyone gets panics or warnings from dflt_lock() being called, please let me know right away. Reviewed by: tmm, gibbs
2003-07-01 15:52:06 +00:00
/* lockfunc */ busdma_lock_mutex,
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
/* lockarg */ &adv->lock,
Clean up argument comments for bus_dma_tag_create() calls.
2003-03-29 09:46:10 +00:00
&adv->buffer_dmat) != 0) {
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENXIO);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv->init_level++;
/* DMA tag for our sense buffers */
Clean up argument comments for bus_dma_tag_create() calls.
2003-03-29 09:46:10 +00:00
if (bus_dma_tag_create(
/* parent */ adv->parent_dmat,
/* alignment */ 1,
/* boundary */ 0,
/* lowaddr */ BUS_SPACE_MAXADDR,
/* highaddr */ BUS_SPACE_MAXADDR,
/* filter */ NULL,
/* filterarg */ NULL,
/* maxsize */ sizeof(struct scsi_sense_data) *
adv->max_openings,
/* nsegments */ 1,
/* maxsegsz */ BUS_SPACE_MAXSIZE_32BIT,
/* flags */ 0,
Mega busdma API commit. Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg. Lockfunc allows a driver to provide a function for managing its locking semantics while using busdma. At the moment, this is used for the asynchronous busdma_swi and callback mechanism. Two lockfunc implementations are provided: busdma_lock_mutex() performs standard mutex operations on the mutex that is specified from lockfuncarg. dftl_lock() is a panic implementation and is defaulted to when NULL, NULL are passed to bus_dma_tag_create(). The only time that NULL, NULL should ever be used is when the driver ensures that bus_dmamap_load() will not be deferred. Drivers that do not provide their own locking can pass busdma_lock_mutex,&Giant args in order to preserve the former behaviour. sparc64 and powerpc do not provide real busdma_swi functions, so this is largely a noop on those platforms. The busdma_swi on is64 is not properly locked yet, so warnings will be emitted on this platform when busdma callback deferrals happen. If anyone gets panics or warnings from dflt_lock() being called, please let me know right away. Reviewed by: tmm, gibbs
2003-07-01 15:52:06 +00:00
/* lockfunc */ busdma_lock_mutex,
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
/* lockarg */ &adv->lock,
Clean up argument comments for bus_dma_tag_create() calls.
2003-03-29 09:46:10 +00:00
&adv->sense_dmat) != 0) {
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENXIO);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv->init_level++;
/* Allocation for our sense buffers */
if (bus_dmamem_alloc(adv->sense_dmat, (void **)&adv->sense_buffers,
BUS_DMA_NOWAIT, &adv->sense_dmamap) != 0) {
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENOMEM);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
adv->init_level++;
/* And permanently map them */
bus_dmamap_load(adv->sense_dmat, adv->sense_dmamap,
adv->sense_buffers,
sizeof(struct scsi_sense_data)*adv->max_openings,
adv_map, &adv->sense_physbase, /*flags*/0);
adv->init_level++;
/*
* Fire up the chip
*/
if (adv_start_chip(adv) != 1) {
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
device_printf(adv->dev,
"Unable to start on board processor. Aborting.\n");
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENXIO);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
/*
* Create the device queue for our SIM.
*/
devq = cam_simq_alloc(adv->max_openings);
if (devq == NULL)
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENOMEM);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/*
* Construct our SIM entry.
*/
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
adv->sim = cam_sim_alloc(adv_action, adv_poll, "adv", adv,
device_get_unit(adv->dev), &adv->lock, 1, adv->max_openings, devq);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
if (adv->sim == NULL)
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENOMEM);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
/*
* Register the bus.
*
* XXX Twin Channel EISA Cards???
*/
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_lock(&adv->lock);
Prepare for future integration between CAM and newbus. xpt_bus_register now takes a device_t to be the parent of the bus that is being created. Most SIMs have been updated with a reasonable argument, but a few exceptions just pass NULL for now. This argument isn't used yet and the newbus integration likely won't be ready until after 7.0-RELEASE.
2007-06-17 05:55:54 +00:00
if (xpt_bus_register(adv->sim, adv->dev, 0) != CAM_SUCCESS) {
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
cam_sim_free(adv->sim, /*free devq*/TRUE);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_unlock(&adv->lock);
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENXIO);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
if (xpt_create_path(&adv->path, /*periph*/NULL, cam_sim_path(adv->sim),
CAM_TARGET_WILDCARD, CAM_LUN_WILDCARD)
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
!= CAM_REQ_CMP) {
xpt_bus_deregister(cam_sim_path(adv->sim));
cam_sim_free(adv->sim, /*free devq*/TRUE);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_unlock(&adv->lock);
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
return (ENXIO);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
}
Fixed return value of adv_attach function. Submitted by: Oleg Sharoiko <os@rsu.ru>
2000-05-23 10:12:42 +00:00
xpt_setup_ccb(&csa.ccb_h, adv->path, /*priority*/5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = AC_FOUND_DEVICE|AC_LOST_DEVICE;
csa.callback = advasync;
csa.callback_arg = adv;
xpt_action((union ccb *)&csa);
Add locking to adv(4) driver and mark it MPSAFE. - Disable the support for the second channel on twin-channel EISA cards as the current incarnation can't possibly work correctly (it hasn't worked since switching to new-bus where new-bus allocates the softc). If anyone bothers to test this again it can be fixed properly and brought back. - Use device_printf() and device_get_nameunit() instead of adv_name(). - Remove use of explicit bus space handles and tags. - Use PCI bus accessors and helper routines rather than accessing config registers directly. - Handle failures from adv_attach(). Tested by: no one (hope it works)
2012-10-12 21:31:44 +00:00
mtx_unlock(&adv->lock);
Advance Systems SCSI Host Adapter driver for CAM. Currently only support the 8bit SCSI AdvanSys products.
1998-09-15 07:03:43 +00:00
return (0);
}
Add MODULE_DEPENDS for cam, pci, mca, eisa and isa where needed. PR: 106543 MFC after: 3 days
2006-12-11 18:28:31 +00:00
MODULE_DEPEND(adv, cam, 1, 1, 1);
Reference in New Issue Copy Permalink