d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
fef5fd2315
freebsd-dev/sys/dev/aic7xxx/aic7xxx_freebsd.c

1903 lines
49 KiB
C
Raw Normal View History

Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
/*
* Bus independent FreeBSD shim for the aic7xxx based adaptec SCSI controllers
*
Update Copyright notices for new year. (should have been in last commit).
2001-01-27 20:54:24 +00:00
* Copyright (c) 1994-2001 Justin T. Gibbs.
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +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.
* 2. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* Alternatively, this software may be distributed under the terms of the
* GNU Public License ("GPL").
*
* 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.
*
* $Id$
*
* $FreeBSD$
*/
#include <dev/aic7xxx/aic7xxx_freebsd.h>
#include <dev/aic7xxx/aic7xxx_inline.h>
#ifndef AHC_TMODE_ENABLE
#define AHC_TMODE_ENABLE 0
#endif
#define ccb_scb_ptr spriv_ptr0
#ifdef AHC_DEBUG
static int ahc_debug = AHC_DEBUG;
#endif
#if UNUSED
static void ahc_dump_targcmd(struct target_cmd *cmd);
#endif
static void ahc_action(struct cam_sim *sim, union ccb *ccb);
static void ahc_get_tran_settings(struct ahc_softc *ahc,
int our_id, char channel,
struct ccb_trans_settings *cts);
static void ahc_async(void *callback_arg, uint32_t code,
struct cam_path *path, void *arg);
static void ahc_execute_scb(void *arg, bus_dma_segment_t *dm_segs,
int nsegments, int error);
static void ahc_poll(struct cam_sim *sim);
static void ahc_setup_data(struct ahc_softc *ahc, struct cam_sim *sim,
struct ccb_scsiio *csio, struct scb *scb);
static void ahc_abort_ccb(struct ahc_softc *ahc, struct cam_sim *sim,
union ccb *ccb);
static int ahc_create_path(struct ahc_softc *ahc,
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
char channel, u_int target, u_int lun,
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
struct cam_path **path);
static void ahc_set_recoveryscb(struct ahc_softc *ahc, struct scb *scb);
static int
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
ahc_create_path(struct ahc_softc *ahc, char channel, u_int target,
u_int lun, struct cam_path **path)
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
{
path_id_t path_id;
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
if (channel == 'B')
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
path_id = cam_sim_path(ahc->platform_data->sim_b);
else
path_id = cam_sim_path(ahc->platform_data->sim);
return (xpt_create_path(path, /*periph*/NULL,
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
path_id, target, lun));
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
}
/*
* Attach all the sub-devices we can find
*/
int
ahc_attach(struct ahc_softc *ahc)
{
char ahc_info[256];
struct ccb_setasync csa;
struct cam_devq *devq;
int bus_id;
int bus_id2;
struct cam_sim *sim;
struct cam_sim *sim2;
struct cam_path *path;
struct cam_path *path2;
long s;
int count;
int error;
count = 0;
sim = NULL;
sim2 = NULL;
ahc_controller_info(ahc, ahc_info);
printf("%s\n", ahc_info);
ahc_lock(ahc, &s);
/* Hook up our interrupt handler */
if ((error = bus_setup_intr(ahc->dev_softc, ahc->platform_data->irq,
Set the interrupt-harvesting bit for those hardware items that I can actually test. Nothing will happen until the sysadmin turns on intr-harvesting on her computer.
2001-02-18 18:02:15 +00:00
INTR_TYPE_CAM|INTR_ENTROPY, ahc_platform_intr, ahc,
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
&ahc->platform_data->ih)) != 0) {
device_printf(ahc->dev_softc, "bus_setup_intr() failed: %d\n",
error);
goto fail;
}
/*
* Attach secondary channel first if the user has
* declared it the primary channel.
*/
Limit CHANNEL_B_PRIMARY's effect to twin channel adapters where we can easily register our sub-busses in the reversed order. In the future, we may provide a hint to CAM on how to order the scans for multi-function adapters that also set this flag, but trying to do it the "twin channel" way will lead to a panic.
2001-02-17 01:53:22 +00:00
if ((ahc->features & AHC_TWIN) != 0
&& (ahc->flags & AHC_CHANNEL_B_PRIMARY) != 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
bus_id = 1;
bus_id2 = 0;
} else {
bus_id = 0;
bus_id2 = 1;
}
/*
* Create the device queue for our SIM(s).
*/
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
devq = cam_simq_alloc(AHC_MAX_QUEUE);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (devq == NULL)
goto fail;
/*
* Construct our first channel SIM entry
*/
Remove the last two uses of ahc->unit in the FreeBSD version of the driver. ahc->unit is depricated and will be going away as soon as the Linux driver catches up. In the FreeBSD case, it is always initialized to 0 and this caused some strangeness in registering multiple ahc controllers with CAM. Noticed by: Tor.Egge@fast.no
2000-09-20 04:46:15 +00:00
sim = cam_sim_alloc(ahc_action, ahc_poll, "ahc", ahc,
device_get_unit(ahc->dev_softc),
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
1, AHC_MAX_QUEUE, devq);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (sim == NULL) {
cam_simq_free(devq);
goto fail;
}
if (xpt_bus_register(sim, bus_id) != CAM_SUCCESS) {
cam_sim_free(sim, /*free_devq*/TRUE);
sim = NULL;
goto fail;
}
if (xpt_create_path(&path, /*periph*/NULL,
cam_sim_path(sim), CAM_TARGET_WILDCARD,
CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
xpt_bus_deregister(cam_sim_path(sim));
cam_sim_free(sim, /*free_devq*/TRUE);
sim = NULL;
goto fail;
}
xpt_setup_ccb(&csa.ccb_h, path, /*priority*/5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = AC_LOST_DEVICE;
csa.callback = ahc_async;
csa.callback_arg = sim;
xpt_action((union ccb *)&csa);
count++;
if (ahc->features & AHC_TWIN) {
sim2 = cam_sim_alloc(ahc_action, ahc_poll, "ahc",
Remove the last two uses of ahc->unit in the FreeBSD version of the driver. ahc->unit is depricated and will be going away as soon as the Linux driver catches up. In the FreeBSD case, it is always initialized to 0 and this caused some strangeness in registering multiple ahc controllers with CAM. Noticed by: Tor.Egge@fast.no
2000-09-20 04:46:15 +00:00
ahc, device_get_unit(ahc->dev_softc), 1,
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
AHC_MAX_QUEUE, devq);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (sim2 == NULL) {
printf("ahc_attach: Unable to attach second "
"bus due to resource shortage");
goto fail;
}
if (xpt_bus_register(sim2, bus_id2) != CAM_SUCCESS) {
printf("ahc_attach: Unable to attach second "
"bus due to resource shortage");
/*
* We do not want to destroy the device queue
* because the first bus is using it.
*/
cam_sim_free(sim2, /*free_devq*/FALSE);
goto fail;
}
if (xpt_create_path(&path2, /*periph*/NULL,
cam_sim_path(sim2),
CAM_TARGET_WILDCARD,
CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
xpt_bus_deregister(cam_sim_path(sim2));
cam_sim_free(sim2, /*free_devq*/FALSE);
sim2 = NULL;
goto fail;
}
xpt_setup_ccb(&csa.ccb_h, path2, /*priority*/5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = AC_LOST_DEVICE;
csa.callback = ahc_async;
csa.callback_arg = sim2;
xpt_action((union ccb *)&csa);
count++;
}
fail:
Limit CHANNEL_B_PRIMARY's effect to twin channel adapters where we can easily register our sub-busses in the reversed order. In the future, we may provide a hint to CAM on how to order the scans for multi-function adapters that also set this flag, but trying to do it the "twin channel" way will lead to a panic.
2001-02-17 01:53:22 +00:00
if ((ahc->features & AHC_TWIN) != 0
&& (ahc->flags & AHC_CHANNEL_B_PRIMARY) != 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc->platform_data->sim_b = sim;
ahc->platform_data->path_b = path;
ahc->platform_data->sim = sim2;
ahc->platform_data->path = path2;
} else {
ahc->platform_data->sim = sim;
ahc->platform_data->path = path;
ahc->platform_data->sim_b = sim2;
ahc->platform_data->path_b = path2;
}
ahc_unlock(ahc, &s);
if (count != 0)
/* We have to wait until after any system dumps... */
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
ahc->platform_data->eh =
EVENTHANDLER_REGISTER(shutdown_final, ahc_shutdown,
ahc, SHUTDOWN_PRI_DEFAULT);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
return (count);
}
/*
* Catch an interrupt from the adapter
*/
void
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
ahc_platform_intr(void *arg)
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
{
struct ahc_softc *ahc;
ahc = (struct ahc_softc *)arg;
ahc_intr(ahc);
}
/*
* We have an scb which has been processed by the
* adaptor, now we look to see how the operation
* went.
*/
void
ahc_done(struct ahc_softc *ahc, struct scb *scb)
{
union ccb *ccb;
CAM_DEBUG(scb->io_ctx->ccb_h.path, CAM_DEBUG_TRACE,
("ahc_done - scb %d\n", scb->hscb->tag));
ccb = scb->io_ctx;
LIST_REMOVE(scb, pending_links);
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
if ((scb->flags & SCB_UNTAGGEDQ) != 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
struct scb_tailq *untagged_q;
aic7xxx.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. Set the user and goal settings prior to setting the current settings. This allows the async update routine to filter out intermediate transfer negotiation updates that may be less than interesting. The Linux OSM uses this to reduce the amount of stuff printed to the console. aic7xxx.seq: Correct an issue with the aic7770 in twin channel mode. We could continually attempt to start a selection even though a selection was already occurring on one channel. This might have the side effect of hanging our selection or causing us to select the wrong device. While here, create a separate polling loop for when we have already started a selection. This should reduce the latency of our response to a (re)selection. The diffs look larger than they really are due to some code rearrangement to optimize out a jmp. aic7xxx_freebsd.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. aic7xxx_inline.h: Get back in sync with perforce revision ID. aic7xxx_pci.c: Identify adapters in ARO mode as such. Ensure that not only the subvendor ID is correct (9005) but also that the controller type field is valid before looking at other information in the subdevice id. Intel seems to have decided that their subdevice id of 8086 is more appropriate for some of their MBs with aic7xxx parts than Adaptec's sanctioned scheme. Add an exclusion entry for SISL (AAC on MB based adapters). Adapters in SISL mode are owned by the RAID controller, so even if a driver for the RAID controller is not present, it isn't safe for us to touch them.
2001-02-21 20:50:36 +00:00
int target_offset;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
aic7xxx.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. Set the user and goal settings prior to setting the current settings. This allows the async update routine to filter out intermediate transfer negotiation updates that may be less than interesting. The Linux OSM uses this to reduce the amount of stuff printed to the console. aic7xxx.seq: Correct an issue with the aic7770 in twin channel mode. We could continually attempt to start a selection even though a selection was already occurring on one channel. This might have the side effect of hanging our selection or causing us to select the wrong device. While here, create a separate polling loop for when we have already started a selection. This should reduce the latency of our response to a (re)selection. The diffs look larger than they really are due to some code rearrangement to optimize out a jmp. aic7xxx_freebsd.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. aic7xxx_inline.h: Get back in sync with perforce revision ID. aic7xxx_pci.c: Identify adapters in ARO mode as such. Ensure that not only the subvendor ID is correct (9005) but also that the controller type field is valid before looking at other information in the subdevice id. Intel seems to have decided that their subdevice id of 8086 is more appropriate for some of their MBs with aic7xxx parts than Adaptec's sanctioned scheme. Add an exclusion entry for SISL (AAC on MB based adapters). Adapters in SISL mode are owned by the RAID controller, so even if a driver for the RAID controller is not present, it isn't safe for us to touch them.
2001-02-21 20:50:36 +00:00
target_offset = SCB_GET_TARGET_OFFSET(ahc, scb);
untagged_q = &ahc->untagged_queues[target_offset];
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
TAILQ_REMOVE(untagged_q, scb, links.tqe);
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
scb->flags &= ~SCB_UNTAGGEDQ;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_run_untagged_queue(ahc, untagged_q);
}
untimeout(ahc_timeout, (caddr_t)scb, ccb->ccb_h.timeout_ch);
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
bus_dmasync_op_t op;
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_POSTREAD;
else
op = BUS_DMASYNC_POSTWRITE;
bus_dmamap_sync(ahc->buffer_dmat, scb->dmamap, op);
bus_dmamap_unload(ahc->buffer_dmat, scb->dmamap);
}
if (ccb->ccb_h.func_code == XPT_CONT_TARGET_IO) {
if (ahc_get_transaction_status(scb) == CAM_REQ_INPROG)
ccb->ccb_h.status |= CAM_REQ_CMP;
ccb->ccb_h.status &= ~CAM_SIM_QUEUED;
ahc_free_scb(ahc, scb);
xpt_done(ccb);
return;
}
/*
* If the recovery SCB completes, we have to be
* out of our timeout.
*/
if ((scb->flags & SCB_RECOVERY_SCB) != 0) {
struct scb *list_scb;
/*
* We were able to complete the command successfully,
* so reinstate the timeouts for all other pending
* commands.
*/
LIST_FOREACH(list_scb, &ahc->pending_scbs, pending_links) {
union ccb *ccb;
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
uint64_t time;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ccb = list_scb->io_ctx;
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
if (ccb->ccb_h.timeout == CAM_TIME_INFINITY)
continue;
time = ccb->ccb_h.timeout;
time *= hz;
time /= 1000;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ccb->ccb_h.timeout_ch =
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
timeout(ahc_timeout, list_scb, time);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
}
if (ahc_get_transaction_status(scb) == CAM_BDR_SENT
|| ahc_get_transaction_status(scb) == CAM_REQ_ABORTED)
ahc_set_transaction_status(scb, CAM_CMD_TIMEOUT);
ahc_print_path(ahc, scb);
printf("no longer in timeout, status = %x\n",
ccb->ccb_h.status);
}
/* Don't clobber any existing error state */
if (ahc_get_transaction_status(scb) == CAM_REQ_INPROG) {
ccb->ccb_h.status |= CAM_REQ_CMP;
} else if ((scb->flags & SCB_SENSE) != 0) {
/*
* We performed autosense retrieval.
*
* Zero any sense not transferred by the
* device. The SCSI spec mandates that any
* untransfered data should be assumed to be
* zero. Complete the 'bounce' of sense information
* through buffers accessible via bus-space by
* copying it into the clients csio.
*/
memset(&ccb->csio.sense_data, 0, sizeof(ccb->csio.sense_data));
memcpy(&ccb->csio.sense_data,
Index our "DMA safe" sense buffers by SCB rather than hscb. hscbs may be traded during error recovery due to the way we manage the qinfifo. This has the effect of changing the index to the sense buffer even though the request sense command references the original buffer. SCBs don't play this swapping game and so serve as a more consistent reference.
2001-01-09 00:40:38 +00:00
ahc_get_sense_buf(ahc, scb),
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
(scb->sg_list->len & AHC_SG_LEN_MASK)
- ccb->csio.sense_resid);
scb->io_ctx->ccb_h.status |= CAM_AUTOSNS_VALID;
}
ccb->ccb_h.status &= ~CAM_SIM_QUEUED;
ahc_free_scb(ahc, scb);
xpt_done(ccb);
}
static void
ahc_action(struct cam_sim *sim, union ccb *ccb)
{
struct ahc_softc *ahc;
struct tmode_lstate *lstate;
u_int target_id;
u_int our_id;
long s;
CAM_DEBUG(ccb->ccb_h.path, CAM_DEBUG_TRACE, ("ahc_action\n"));
ahc = (struct ahc_softc *)cam_sim_softc(sim);
target_id = ccb->ccb_h.target_id;
our_id = SIM_SCSI_ID(ahc, sim);
switch (ccb->ccb_h.func_code) {
/* Common cases first */
case XPT_ACCEPT_TARGET_IO: /* Accept Host Target Mode CDB */
case XPT_CONT_TARGET_IO:/* Continue Host Target I/O Connection*/
{
struct tmode_tstate *tstate;
cam_status status;
status = ahc_find_tmode_devs(ahc, sim, ccb, &tstate,
&lstate, TRUE);
if (status != CAM_REQ_CMP) {
if (ccb->ccb_h.func_code == XPT_CONT_TARGET_IO) {
/* Response from the black hole device */
tstate = NULL;
lstate = ahc->black_hole;
} else {
ccb->ccb_h.status = status;
xpt_done(ccb);
break;
}
}
if (ccb->ccb_h.func_code == XPT_ACCEPT_TARGET_IO) {
ahc_lock(ahc, &s);
SLIST_INSERT_HEAD(&lstate->accept_tios, &ccb->ccb_h,
sim_links.sle);
ccb->ccb_h.status = CAM_REQ_INPROG;
if ((ahc->flags & AHC_TQINFIFO_BLOCKED) != 0)
ahc_run_tqinfifo(ahc, /*paused*/FALSE);
ahc_unlock(ahc, &s);
break;
}
/*
* The target_id represents the target we attempt to
* select. In target mode, this is the initiator of
* the original command.
*/
our_id = target_id;
target_id = ccb->csio.init_id;
/* FALLTHROUGH */
}
case XPT_SCSI_IO: /* Execute the requested I/O operation */
case XPT_RESET_DEV: /* Bus Device Reset the specified SCSI device */
{
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
struct scb *scb;
struct hardware_scb *hscb;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
if ((ahc->flags & AHC_INITIATORROLE) == 0
&& (ccb->ccb_h.func_code == XPT_SCSI_IO
|| ccb->ccb_h.func_code == XPT_RESET_DEV)) {
ccb->ccb_h.status = CAM_PROVIDE_FAIL;
xpt_done(ccb);
}
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
/*
* get an scb to use.
*/
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
ahc_lock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if ((scb = ahc_get_scb(ahc)) == NULL) {
Close a race condition that doesn't really exist in -current. When a resource shortage occurs, freeze our queue and then set the resource shortage flag while the controller data structure is locked. The old code did these in the wrong order potentially allowing our interrupt handler to release the queue and clear the flag before the freeze ever occurred.
2000-10-09 19:54:54 +00:00
xpt_freeze_simq(sim, /*count*/1);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc->flags |= AHC_RESOURCE_SHORTAGE;
ahc_unlock(ahc, &s);
ahc_set_transaction_status() takes an SCB. This makes it difficult to use this helper function to report an error when SCB allocation fails.
2000-10-01 20:56:44 +00:00
ccb->ccb_h.status = CAM_REQUEUE_REQ;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
xpt_done(ccb);
return;
}
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
ahc_unlock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
hscb = scb->hscb;
CAM_DEBUG(ccb->ccb_h.path, CAM_DEBUG_SUBTRACE,
("start scb(%p)\n", scb));
scb->io_ctx = ccb;
/*
* So we can find the SCB when an abort is requested
*/
ccb->ccb_h.ccb_scb_ptr = scb;
/*
* Put all the arguments for the xfer in the scb
*/
hscb->control = 0;
hscb->scsiid = BUILD_SCSIID(ahc, sim, target_id, our_id);
hscb->lun = ccb->ccb_h.target_lun;
if (ccb->ccb_h.func_code == XPT_RESET_DEV) {
hscb->cdb_len = 0;
scb->flags |= SCB_DEVICE_RESET;
hscb->control |= MK_MESSAGE;
ahc_execute_scb(scb, NULL, 0, 0);
} else {
if (ccb->ccb_h.func_code == XPT_CONT_TARGET_IO) {
struct target_data *tdata;
tdata = &hscb->shared_data.tdata;
if (ahc->pending_device == lstate) {
scb->flags |= SCB_TARGET_IMMEDIATE;
ahc->pending_device = NULL;
}
hscb->control |= TARGET_SCB;
tdata->target_phases = IDENTIFY_SEEN;
if ((ccb->ccb_h.flags & CAM_SEND_STATUS) != 0) {
tdata->target_phases |= SPHASE_PENDING;
tdata->scsi_status =
ccb->csio.scsi_status;
}
tdata->initiator_tag = ccb->csio.tag_id;
}
if (ccb->ccb_h.flags & CAM_TAG_ACTION_VALID)
hscb->control |= ccb->csio.tag_action;
ahc_setup_data(ahc, sim, &ccb->csio, scb);
}
break;
}
case XPT_NOTIFY_ACK:
case XPT_IMMED_NOTIFY:
{
struct tmode_tstate *tstate;
struct tmode_lstate *lstate;
cam_status status;
status = ahc_find_tmode_devs(ahc, sim, ccb, &tstate,
&lstate, TRUE);
if (status != CAM_REQ_CMP) {
ccb->ccb_h.status = status;
xpt_done(ccb);
break;
}
SLIST_INSERT_HEAD(&lstate->immed_notifies, &ccb->ccb_h,
sim_links.sle);
ccb->ccb_h.status = CAM_REQ_INPROG;
ahc_send_lstate_events(ahc, lstate);
break;
}
case XPT_EN_LUN: /* Enable LUN as a target */
ahc_handle_en_lun(ahc, sim, ccb);
xpt_done(ccb);
break;
case XPT_ABORT: /* Abort the specified CCB */
{
ahc_abort_ccb(ahc, sim, ccb);
break;
}
case XPT_SET_TRAN_SETTINGS:
{
#ifdef AHC_NEW_TRAN_SETTINGS
struct ahc_devinfo devinfo;
struct ccb_trans_settings *cts;
struct ccb_trans_settings_scsi *scsi;
struct ccb_trans_settings_spi *spi;
struct ahc_initiator_tinfo *tinfo;
struct tmode_tstate *tstate;
uint16_t *discenable;
uint16_t *tagenable;
u_int update_type;
cts = &ccb->cts;
scsi = &cts->proto_specific.scsi;
spi = &cts->xport_specific.spi;
ahc_compile_devinfo(&devinfo, SIM_SCSI_ID(ahc, sim),
cts->ccb_h.target_id,
cts->ccb_h.target_lun,
SIM_CHANNEL(ahc, sim),
ROLE_UNKNOWN);
tinfo = ahc_fetch_transinfo(ahc, devinfo.channel,
devinfo.our_scsiid,
devinfo.target, &tstate);
update_type = 0;
if (cts->type == CTS_TYPE_CURRENT_SETTINGS) {
update_type |= AHC_TRANS_GOAL;
discenable = &tstate->discenable;
tagenable = &tstate->tagenable;
tinfo->current.protocol_version =
cts->protocol_version;
tinfo->current.transport_version =
cts->transport_version;
tinfo->goal.protocol_version =
cts->protocol_version;
tinfo->goal.transport_version =
cts->transport_version;
} else if (cts->type == CTS_TYPE_USER_SETTINGS) {
update_type |= AHC_TRANS_USER;
discenable = &ahc->user_discenable;
tagenable = &ahc->user_tagenable;
tinfo->user.protocol_version =
cts->protocol_version;
tinfo->user.transport_version =
cts->transport_version;
} else {
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
}
ahc_lock(ahc, &s);
if ((spi->valid & CTS_SPI_VALID_DISC) != 0) {
if ((spi->flags & CTS_SPI_FLAGS_DISC_ENB) != 0)
*discenable |= devinfo.target_mask;
else
*discenable &= ~devinfo.target_mask;
}
if ((scsi->valid & CTS_SCSI_VALID_TQ) != 0) {
if ((scsi->flags & CTS_SCSI_FLAGS_TAG_ENB) != 0)
*tagenable |= devinfo.target_mask;
else
*tagenable &= ~devinfo.target_mask;
}
if ((spi->valid & CTS_SPI_VALID_BUS_WIDTH) != 0) {
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_validate_width(ahc, /*tinfo limit*/NULL,
&spi->bus_width, ROLE_UNKNOWN);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_set_width(ahc, &devinfo, spi->bus_width,
update_type, /*paused*/FALSE);
}
if ((spi->valid & CTS_SPI_VALID_PPR_OPTIONS) == 0) {
if (update_type == AHC_TRANS_USER)
spi->ppr_options = tinfo->user.ppr_options;
else
spi->ppr_options = tinfo->goal.ppr_options;
}
if ((spi->valid & CTS_SPI_VALID_SYNC_OFFSET) == 0) {
if (update_type == AHC_TRANS_USER)
spi->sync_offset = tinfo->user.offset;
else
spi->sync_offset = tinfo->goal.offset;
}
if ((spi->valid & CTS_SPI_VALID_SYNC_RATE) == 0) {
if (update_type == AHC_TRANS_USER)
spi->sync_period = tinfo->user.period;
else
spi->sync_period = tinfo->goal.period;
}
if (((spi->valid & CTS_SPI_VALID_SYNC_RATE) != 0)
|| ((spi->valid & CTS_SPI_VALID_SYNC_OFFSET) != 0)) {
struct ahc_syncrate *syncrate;
u_int maxsync;
if ((ahc->features & AHC_ULTRA2) != 0)
maxsync = AHC_SYNCRATE_DT;
else if ((ahc->features & AHC_ULTRA) != 0)
maxsync = AHC_SYNCRATE_ULTRA;
else
maxsync = AHC_SYNCRATE_FAST;
syncrate = ahc_find_syncrate(ahc, &spi->sync_period,
&spi->ppr_options,
maxsync);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_validate_offset(ahc, /*tinfo limit*/NULL,
syncrate, &spi->sync_offset,
spi->bus_width, ROLE_UNKNOWN);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
/* We use a period of 0 to represent async */
if (spi->sync_offset == 0) {
spi->sync_period = 0;
spi->ppr_options = 0;
}
ahc_set_syncrate(ahc, &devinfo, syncrate,
spi->sync_period, spi->sync_offset,
spi->ppr_options, update_type,
/*paused*/FALSE);
}
ahc_unlock(ahc, &s);
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
#else
struct ahc_devinfo devinfo;
struct ccb_trans_settings *cts;
struct ahc_initiator_tinfo *tinfo;
struct tmode_tstate *tstate;
uint16_t *discenable;
uint16_t *tagenable;
u_int update_type;
long s;
cts = &ccb->cts;
ahc_compile_devinfo(&devinfo, SIM_SCSI_ID(ahc, sim),
cts->ccb_h.target_id,
cts->ccb_h.target_lun,
SIM_CHANNEL(ahc, sim),
ROLE_UNKNOWN);
tinfo = ahc_fetch_transinfo(ahc, devinfo.channel,
devinfo.our_scsiid,
devinfo.target, &tstate);
update_type = 0;
if ((cts->flags & CCB_TRANS_CURRENT_SETTINGS) != 0) {
update_type |= AHC_TRANS_GOAL;
discenable = &tstate->discenable;
tagenable = &tstate->tagenable;
} else if ((cts->flags & CCB_TRANS_USER_SETTINGS) != 0) {
update_type |= AHC_TRANS_USER;
discenable = &ahc->user_discenable;
tagenable = &ahc->user_tagenable;
} else {
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
}
ahc_lock(ahc, &s);
if ((cts->valid & CCB_TRANS_DISC_VALID) != 0) {
if ((cts->flags & CCB_TRANS_DISC_ENB) != 0)
*discenable |= devinfo.target_mask;
else
*discenable &= ~devinfo.target_mask;
}
if ((cts->valid & CCB_TRANS_TQ_VALID) != 0) {
if ((cts->flags & CCB_TRANS_TAG_ENB) != 0)
*tagenable |= devinfo.target_mask;
else
*tagenable &= ~devinfo.target_mask;
}
if ((cts->valid & CCB_TRANS_BUS_WIDTH_VALID) != 0) {
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_validate_width(ahc, /*tinfo limit*/NULL,
&cts->bus_width, ROLE_UNKNOWN);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_set_width(ahc, &devinfo, cts->bus_width,
update_type, /*paused*/FALSE);
}
if ((cts->valid & CCB_TRANS_SYNC_OFFSET_VALID) == 0) {
if (update_type == AHC_TRANS_USER)
cts->sync_offset = tinfo->user.offset;
else
cts->sync_offset = tinfo->goal.offset;
}
if ((cts->valid & CCB_TRANS_SYNC_RATE_VALID) == 0) {
if (update_type == AHC_TRANS_USER)
cts->sync_period = tinfo->user.period;
else
cts->sync_period = tinfo->goal.period;
}
if (((cts->valid & CCB_TRANS_SYNC_RATE_VALID) != 0)
|| ((cts->valid & CCB_TRANS_SYNC_OFFSET_VALID) != 0)) {
struct ahc_syncrate *syncrate;
u_int ppr_options;
u_int maxsync;
if ((ahc->features & AHC_ULTRA2) != 0)
maxsync = AHC_SYNCRATE_DT;
else if ((ahc->features & AHC_ULTRA) != 0)
maxsync = AHC_SYNCRATE_ULTRA;
else
maxsync = AHC_SYNCRATE_FAST;
ppr_options = 0;
if (cts->sync_period <= 9)
ppr_options = MSG_EXT_PPR_DT_REQ;
syncrate = ahc_find_syncrate(ahc, &cts->sync_period,
&ppr_options,
maxsync);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_validate_offset(ahc, /*tinfo limit*/NULL,
syncrate, &cts->sync_offset,
MSG_EXT_WDTR_BUS_8_BIT,
ROLE_UNKNOWN);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
/* We use a period of 0 to represent async */
if (cts->sync_offset == 0) {
cts->sync_period = 0;
ppr_options = 0;
}
if (ppr_options == MSG_EXT_PPR_DT_REQ
&& tinfo->user.transport_version >= 3) {
tinfo->goal.transport_version =
tinfo->user.transport_version;
tinfo->current.transport_version =
tinfo->user.transport_version;
}
ahc_set_syncrate(ahc, &devinfo, syncrate,
cts->sync_period, cts->sync_offset,
ppr_options, update_type,
/*paused*/FALSE);
}
ahc_unlock(ahc, &s);
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
#endif
break;
}
case XPT_GET_TRAN_SETTINGS:
/* Get default/user set transfer settings for the target */
{
ahc_lock(ahc, &s);
ahc_get_tran_settings(ahc, SIM_SCSI_ID(ahc, sim),
SIM_CHANNEL(ahc, sim), &ccb->cts);
ahc_unlock(ahc, &s);
xpt_done(ccb);
break;
}
case XPT_CALC_GEOMETRY:
{
struct ccb_calc_geometry *ccg;
uint32_t size_mb;
uint32_t secs_per_cylinder;
int extended;
ccg = &ccb->ccg;
size_mb = ccg->volume_size
/ ((1024L * 1024L) / ccg->block_size);
extended = SIM_IS_SCSIBUS_B(ahc, sim)
? ahc->flags & AHC_EXTENDED_TRANS_B
: ahc->flags & AHC_EXTENDED_TRANS_A;
if (size_mb > 1024 && extended) {
ccg->heads = 255;
ccg->secs_per_track = 63;
} else {
ccg->heads = 64;
ccg->secs_per_track = 32;
}
secs_per_cylinder = ccg->heads * ccg->secs_per_track;
ccg->cylinders = ccg->volume_size / secs_per_cylinder;
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
case XPT_RESET_BUS: /* Reset the specified SCSI bus */
{
int found;
ahc_lock(ahc, &s);
found = ahc_reset_channel(ahc, SIM_CHANNEL(ahc, sim),
/*initiate reset*/TRUE);
ahc_unlock(ahc, &s);
if (bootverbose) {
xpt_print_path(SIM_PATH(ahc, sim));
printf("SCSI bus reset delivered. "
"%d SCBs aborted.\n", found);
}
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;
if ((ahc->features & AHC_WIDE) != 0)
cpi->hba_inquiry |= PI_WIDE_16;
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
if ((ahc->features & AHC_TARGETMODE) != 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
cpi->target_sprt = PIT_PROCESSOR
| PIT_DISCONNECT
| PIT_TERM_IO;
} else {
cpi->target_sprt = 0;
}
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
cpi->hba_misc = 0;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
cpi->hba_eng_cnt = 0;
cpi->max_target = (ahc->features & AHC_WIDE) ? 15 : 7;
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
cpi->max_lun = AHC_NUM_LUNS - 1;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (SIM_IS_SCSIBUS_B(ahc, sim)) {
cpi->initiator_id = ahc->our_id_b;
if ((ahc->flags & AHC_RESET_BUS_B) == 0)
cpi->hba_misc |= PIM_NOBUSRESET;
} else {
cpi->initiator_id = ahc->our_id;
if ((ahc->flags & AHC_RESET_BUS_A) == 0)
cpi->hba_misc |= PIM_NOBUSRESET;
}
cpi->bus_id = cam_sim_bus(sim);
cpi->base_transfer_speed = 3300;
strncpy(cpi->sim_vid, "FreeBSD", SIM_IDLEN);
strncpy(cpi->hba_vid, "Adaptec", HBA_IDLEN);
strncpy(cpi->dev_name, cam_sim_name(sim), DEV_IDLEN);
cpi->unit_number = cam_sim_unit(sim);
#ifdef AHC_NEW_TRAN_SETTINGS
cpi->protocol = PROTO_SCSI;
cpi->protocol_version = SCSI_REV_2;
cpi->transport = XPORT_SPI;
cpi->transport_version = 2;
cpi->xport_specific.spi.ppr_options = SID_SPI_CLOCK_ST;
if ((ahc->features & AHC_DT) != 0) {
cpi->transport_version = 3;
cpi->xport_specific.spi.ppr_options =
SID_SPI_CLOCK_DT_ST;
}
#endif
cpi->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
default:
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ccb->ccb_h.status = CAM_PROVIDE_FAIL;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
xpt_done(ccb);
break;
}
}
static void
ahc_get_tran_settings(struct ahc_softc *ahc, int our_id, char channel,
struct ccb_trans_settings *cts)
{
#ifdef AHC_NEW_TRAN_SETTINGS
struct ahc_devinfo devinfo;
struct ccb_trans_settings_scsi *scsi;
struct ccb_trans_settings_spi *spi;
struct ahc_initiator_tinfo *targ_info;
struct tmode_tstate *tstate;
struct ahc_transinfo *tinfo;
scsi = &cts->proto_specific.scsi;
spi = &cts->xport_specific.spi;
ahc_compile_devinfo(&devinfo, our_id,
cts->ccb_h.target_id,
cts->ccb_h.target_lun,
channel, ROLE_UNKNOWN);
targ_info = ahc_fetch_transinfo(ahc, devinfo.channel,
devinfo.our_scsiid,
devinfo.target, &tstate);
if (cts->type == CTS_TYPE_CURRENT_SETTINGS)
tinfo = &targ_info->current;
else
tinfo = &targ_info->user;
scsi->flags &= ~CTS_SCSI_FLAGS_TAG_ENB;
spi->flags &= ~CTS_SPI_FLAGS_DISC_ENB;
if (cts->type == CTS_TYPE_USER_SETTINGS) {
if ((ahc->user_discenable & devinfo.target_mask) != 0)
spi->flags |= CTS_SPI_FLAGS_DISC_ENB;
if ((ahc->user_tagenable & devinfo.target_mask) != 0)
scsi->flags |= CTS_SCSI_FLAGS_TAG_ENB;
} else {
if ((tstate->discenable & devinfo.target_mask) != 0)
spi->flags |= CTS_SPI_FLAGS_DISC_ENB;
if ((tstate->tagenable & devinfo.target_mask) != 0)
scsi->flags |= CTS_SCSI_FLAGS_TAG_ENB;
}
cts->protocol_version = tinfo->protocol_version;
cts->transport_version = tinfo->transport_version;
spi->sync_period = tinfo->period;
spi->sync_offset = tinfo->offset;
spi->bus_width = tinfo->width;
spi->ppr_options = tinfo->ppr_options;
cts->protocol = PROTO_SCSI;
cts->transport = XPORT_SPI;
spi->valid = CTS_SPI_VALID_SYNC_RATE
| CTS_SPI_VALID_SYNC_OFFSET
| CTS_SPI_VALID_BUS_WIDTH
| CTS_SPI_VALID_PPR_OPTIONS;
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
if (cts->ccb_h.target_lun != CAM_LUN_WILDCARD) {
scsi->valid = CTS_SCSI_VALID_TQ;
spi->valid |= CTS_SPI_VALID_DISC;
} else {
scsi->valid = 0;
}
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
cts->ccb_h.status = CAM_REQ_CMP;
#else
struct ahc_devinfo devinfo;
struct ahc_initiator_tinfo *targ_info;
struct tmode_tstate *tstate;
struct ahc_transinfo *tinfo;
long s;
ahc_compile_devinfo(&devinfo, our_id,
cts->ccb_h.target_id,
cts->ccb_h.target_lun,
channel, ROLE_UNKNOWN);
targ_info = ahc_fetch_transinfo(ahc, devinfo.channel,
devinfo.our_scsiid,
devinfo.target, &tstate);
if ((cts->flags & CCB_TRANS_CURRENT_SETTINGS) != 0)
tinfo = &targ_info->current;
else
tinfo = &targ_info->user;
ahc_lock(ahc, &s);
cts->flags &= ~(CCB_TRANS_DISC_ENB|CCB_TRANS_TAG_ENB);
Fix single character typo in legacy transfer settings function thereby re-enable tagged queuing.
2000-10-06 19:34:40 +00:00
if ((cts->flags & CCB_TRANS_CURRENT_SETTINGS) == 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if ((ahc->user_discenable & devinfo.target_mask) != 0)
cts->flags |= CCB_TRANS_DISC_ENB;
if ((ahc->user_tagenable & devinfo.target_mask) != 0)
cts->flags |= CCB_TRANS_TAG_ENB;
} else {
if ((tstate->discenable & devinfo.target_mask) != 0)
cts->flags |= CCB_TRANS_DISC_ENB;
if ((tstate->tagenable & devinfo.target_mask) != 0)
cts->flags |= CCB_TRANS_TAG_ENB;
}
cts->sync_period = tinfo->period;
cts->sync_offset = tinfo->offset;
cts->bus_width = tinfo->width;
ahc_unlock(ahc, &s);
cts->valid = CCB_TRANS_SYNC_RATE_VALID
| CCB_TRANS_SYNC_OFFSET_VALID
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
| CCB_TRANS_BUS_WIDTH_VALID;
if (cts->ccb_h.target_lun != CAM_LUN_WILDCARD)
cts->valid |= CCB_TRANS_DISC_VALID|CCB_TRANS_TQ_VALID;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
cts->ccb_h.status = CAM_REQ_CMP;
#endif
}
static void
ahc_async(void *callback_arg, uint32_t code, struct cam_path *path, void *arg)
{
struct ahc_softc *ahc;
struct cam_sim *sim;
sim = (struct cam_sim *)callback_arg;
ahc = (struct ahc_softc *)cam_sim_softc(sim);
switch (code) {
case AC_LOST_DEVICE:
{
struct ahc_devinfo devinfo;
long s;
ahc_compile_devinfo(&devinfo, SIM_SCSI_ID(ahc, sim),
xpt_path_target_id(path),
xpt_path_lun_id(path),
SIM_CHANNEL(ahc, sim),
ROLE_UNKNOWN);
/*
* Revert to async/narrow transfers
* for the next device.
*/
ahc_lock(ahc, &s);
ahc_set_width(ahc, &devinfo, MSG_EXT_WDTR_BUS_8_BIT,
AHC_TRANS_GOAL|AHC_TRANS_CUR, /*paused*/FALSE);
ahc_set_syncrate(ahc, &devinfo, /*syncrate*/NULL,
/*period*/0, /*offset*/0, /*ppr_options*/0,
AHC_TRANS_GOAL|AHC_TRANS_CUR,
/*paused*/FALSE);
ahc_unlock(ahc, &s);
break;
}
default:
break;
}
}
static void
ahc_execute_scb(void *arg, bus_dma_segment_t *dm_segs, int nsegments,
int error)
{
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
struct scb *scb;
union ccb *ccb;
struct ahc_softc *ahc;
struct ahc_initiator_tinfo *tinfo;
struct tmode_tstate *tstate;
u_int mask;
long s;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
scb = (struct scb *)arg;
ccb = scb->io_ctx;
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
ahc = scb->ahc_softc;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (error != 0) {
if (error == EFBIG)
ahc_set_transaction_status(scb, CAM_REQ_TOO_BIG);
else
ahc_set_transaction_status(scb, CAM_REQ_CMP_ERR);
if (nsegments != 0)
bus_dmamap_unload(ahc->buffer_dmat, scb->dmamap);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_lock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_free_scb(ahc, scb);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_unlock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
xpt_done(ccb);
return;
}
if (nsegments != 0) {
struct ahc_dma_seg *sg;
bus_dma_segment_t *end_seg;
bus_dmasync_op_t op;
end_seg = dm_segs + nsegments;
/* Copy the segments into our SG list */
sg = scb->sg_list;
while (dm_segs < end_seg) {
sg->addr = dm_segs->ds_addr;
/* XXX Add in the 5th byte of the address later. */
sg->len = dm_segs->ds_len;
sg++;
dm_segs++;
}
/*
* Note where to find the SG entries in bus space.
* We also set the full residual flag which the
* sequencer will clear as soon as a data transfer
* occurs.
*/
scb->hscb->sgptr = scb->sg_list_phys | SG_FULL_RESID;
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_PREREAD;
else
op = BUS_DMASYNC_PREWRITE;
bus_dmamap_sync(ahc->buffer_dmat, scb->dmamap, op);
if (ccb->ccb_h.func_code == XPT_CONT_TARGET_IO) {
struct target_data *tdata;
tdata = &scb->hscb->shared_data.tdata;
tdata->target_phases |= DPHASE_PENDING;
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_OUT)
tdata->data_phase = P_DATAOUT;
else
tdata->data_phase = P_DATAIN;
/*
* If the transfer is of an odd length and in the
* "in" direction (scsi->HostBus), then it may
* trigger a bug in the 'WideODD' feature of
* non-Ultra2 chips. Force the total data-length
* to be even by adding an extra, 1 byte, SG,
* element. We do this even if we are not currently
* negotiated wide as negotiation could occur before
* this command is executed.
*/
if ((ahc->bugs & AHC_TMODE_WIDEODD_BUG) != 0
&& (ccb->csio.dxfer_len & 0x1) != 0
&& (ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN) {
nsegments++;
if (nsegments > AHC_NSEG) {
ahc_set_transaction_status(scb,
CAM_REQ_TOO_BIG);
bus_dmamap_unload(ahc->buffer_dmat,
scb->dmamap);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_lock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_free_scb(ahc, scb);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_unlock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
xpt_done(ccb);
return;
}
sg->addr = ahc->dma_bug_buf;
sg->len = 1;
sg++;
}
}
sg--;
sg->len |= AHC_DMA_LAST_SEG;
/* Copy the first SG into the "current" data pointer area */
scb->hscb->dataptr = scb->sg_list->addr;
scb->hscb->datacnt = scb->sg_list->len;
} else {
scb->hscb->sgptr = SG_LIST_NULL;
scb->hscb->dataptr = 0;
scb->hscb->datacnt = 0;
}
scb->sg_count = nsegments;
ahc_lock(ahc, &s);
/*
* Last time we need to check if this SCB needs to
* be aborted.
*/
if (ahc_get_transaction_status(scb) != CAM_REQ_INPROG) {
if (nsegments != 0)
bus_dmamap_unload(ahc->buffer_dmat,
scb->dmamap);
ahc_free_scb(ahc, scb);
ahc_unlock(ahc, &s);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
xpt_done(ccb);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
return;
}
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
tinfo = ahc_fetch_transinfo(ahc, SCSIID_CHANNEL(ahc, scb->hscb->scsiid),
SCSIID_OUR_ID(scb->hscb->scsiid),
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
SCSIID_TARGET(ahc, scb->hscb->scsiid),
&tstate);
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
mask = SCB_GET_TARGET_MASK(ahc, scb);
scb->hscb->scsirate = tinfo->scsirate;
scb->hscb->scsioffset = tinfo->current.offset;
if ((tstate->ultraenb & mask) != 0)
scb->hscb->control |= ULTRAENB;
if ((tstate->discenable & mask) != 0
&& (ccb->ccb_h.flags & CAM_DIS_DISCONNECT) == 0)
scb->hscb->control |= DISCENB;
if ((ccb->ccb_h.flags & CAM_NEGOTIATE) != 0
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
&& (tinfo->goal.width != 0
|| tinfo->goal.period != 0
|| tinfo->goal.ppr_options != 0)) {
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
scb->flags |= SCB_NEGOTIATE;
scb->hscb->control |= MK_MESSAGE;
}
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
LIST_INSERT_HEAD(&ahc->pending_scbs, scb, pending_links);
ccb->ccb_h.status |= CAM_SIM_QUEUED;
if (ccb->ccb_h.timeout != CAM_TIME_INFINITY) {
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
uint64_t time;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (ccb->ccb_h.timeout == CAM_TIME_DEFAULT)
ccb->ccb_h.timeout = 5 * 1000;
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
time = ccb->ccb_h.timeout;
time *= hz;
time /= 1000;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ccb->ccb_h.timeout_ch =
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
timeout(ahc_timeout, (caddr_t)scb, time);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
}
/*
* We only allow one untagged transaction
* per target in the initiator role unless
* we are storing a full busy target *lun*
* table in SCB space.
*/
if ((scb->hscb->control & (TARGET_SCB|TAG_ENB)) == 0
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
&& (ahc->flags & AHC_SCB_BTT) == 0) {
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
struct scb_tailq *untagged_q;
aic7xxx.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. Set the user and goal settings prior to setting the current settings. This allows the async update routine to filter out intermediate transfer negotiation updates that may be less than interesting. The Linux OSM uses this to reduce the amount of stuff printed to the console. aic7xxx.seq: Correct an issue with the aic7770 in twin channel mode. We could continually attempt to start a selection even though a selection was already occurring on one channel. This might have the side effect of hanging our selection or causing us to select the wrong device. While here, create a separate polling loop for when we have already started a selection. This should reduce the latency of our response to a (re)selection. The diffs look larger than they really are due to some code rearrangement to optimize out a jmp. aic7xxx_freebsd.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. aic7xxx_inline.h: Get back in sync with perforce revision ID. aic7xxx_pci.c: Identify adapters in ARO mode as such. Ensure that not only the subvendor ID is correct (9005) but also that the controller type field is valid before looking at other information in the subdevice id. Intel seems to have decided that their subdevice id of 8086 is more appropriate for some of their MBs with aic7xxx parts than Adaptec's sanctioned scheme. Add an exclusion entry for SISL (AAC on MB based adapters). Adapters in SISL mode are owned by the RAID controller, so even if a driver for the RAID controller is not present, it isn't safe for us to touch them.
2001-02-21 20:50:36 +00:00
int target_offset;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
aic7xxx.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. Set the user and goal settings prior to setting the current settings. This allows the async update routine to filter out intermediate transfer negotiation updates that may be less than interesting. The Linux OSM uses this to reduce the amount of stuff printed to the console. aic7xxx.seq: Correct an issue with the aic7770 in twin channel mode. We could continually attempt to start a selection even though a selection was already occurring on one channel. This might have the side effect of hanging our selection or causing us to select the wrong device. While here, create a separate polling loop for when we have already started a selection. This should reduce the latency of our response to a (re)selection. The diffs look larger than they really are due to some code rearrangement to optimize out a jmp. aic7xxx_freebsd.c: Use the target offset rather than the target Id to reference the untagged SCB array. The offset and id are identical save in the twin channel case. This should correct several issues with the 2742T. aic7xxx_inline.h: Get back in sync with perforce revision ID. aic7xxx_pci.c: Identify adapters in ARO mode as such. Ensure that not only the subvendor ID is correct (9005) but also that the controller type field is valid before looking at other information in the subdevice id. Intel seems to have decided that their subdevice id of 8086 is more appropriate for some of their MBs with aic7xxx parts than Adaptec's sanctioned scheme. Add an exclusion entry for SISL (AAC on MB based adapters). Adapters in SISL mode are owned by the RAID controller, so even if a driver for the RAID controller is not present, it isn't safe for us to touch them.
2001-02-21 20:50:36 +00:00
target_offset = SCB_GET_TARGET_OFFSET(ahc, scb);
untagged_q = &(ahc->untagged_queues[target_offset]);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
TAILQ_INSERT_TAIL(untagged_q, scb, links.tqe);
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
scb->flags |= SCB_UNTAGGEDQ;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (TAILQ_FIRST(untagged_q) != scb) {
ahc_unlock(ahc, &s);
return;
}
}
scb->flags |= SCB_ACTIVE;
if ((scb->flags & SCB_TARGET_IMMEDIATE) != 0) {
pause_sequencer(ahc);
if ((ahc->flags & AHC_PAGESCBS) == 0)
ahc_outb(ahc, SCBPTR, scb->hscb->tag);
ahc_outb(ahc, SCB_TAG, scb->hscb->tag);
ahc_outb(ahc, RETURN_1, CONT_MSG_LOOP);
unpause_sequencer(ahc);
} else {
ahc_queue_scb(ahc, scb);
}
ahc_unlock(ahc, &s);
}
static void
ahc_poll(struct cam_sim *sim)
{
ahc_intr(cam_sim_softc(sim));
}
static void
ahc_setup_data(struct ahc_softc *ahc, struct cam_sim *sim,
struct ccb_scsiio *csio, struct scb *scb)
{
struct hardware_scb *hscb;
struct ccb_hdr *ccb_h;
hscb = scb->hscb;
ccb_h = &csio->ccb_h;
if (ccb_h->func_code == XPT_SCSI_IO) {
hscb->cdb_len = csio->cdb_len;
if ((ccb_h->flags & CAM_CDB_POINTER) != 0) {
if (hscb->cdb_len > sizeof(hscb->cdb32)
|| (ccb_h->flags & CAM_CDB_PHYS) != 0) {
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
u_long s;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_set_transaction_status(scb,
CAM_REQ_INVALID);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_lock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_free_scb(ahc, scb);
Convert the driver to use a single DMA for fetching new commands instead of two (one to access the circular input fifo, the other to get the SCB). This costs us a command slot so the driver can now only queue 254 simultaneous commands. Have the kernel driver honor critical sections in sequencer code. When prefetching S/G segments only pull a cacheline's worth but never less than two elements. This reduces the impact of the prefetch on the main data transfer when compared to the 128 byte fetches the driver used to do. Add "bootverbose" logging for transfer negotiations. Correct a bug in ahc_set_syncrate() that would prevent an update of the sync parameters if only the ppr_options had changed. Correct locking for calls to ahc_free_scb(). ahc_free_scb() is no longer protected internally to simplify ports to other platforms. Make sure we unfreeze our SIMQ if a resource shortage has occurred and an SCB is been freed. ahc_pci.c: Turn on cacheline streaming for all controllers that support it. Clarify diagnostic messages about PCI interrupts.
2000-10-05 04:24:14 +00:00
ahc_unlock(ahc, &s);
xpt_done((union ccb *)csio);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
return;
}
if (hscb->cdb_len > 12) {
memcpy(hscb->cdb32,
csio->cdb_io.cdb_ptr,
hscb->cdb_len);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
scb->flags |= SCB_CDB32_PTR;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
} else {
memcpy(hscb->shared_data.cdb,
csio->cdb_io.cdb_ptr,
hscb->cdb_len);
}
} else {
if (hscb->cdb_len > 12) {
memcpy(hscb->cdb32, csio->cdb_io.cdb_bytes,
hscb->cdb_len);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
scb->flags |= SCB_CDB32_PTR;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
} else {
memcpy(hscb->shared_data.cdb,
csio->cdb_io.cdb_bytes,
hscb->cdb_len);
}
}
}
/* Only use S/G if there is a transfer */
if ((ccb_h->flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
if ((ccb_h->flags & CAM_SCATTER_VALID) == 0) {
/* We've been given a pointer to a single buffer */
if ((ccb_h->flags & CAM_DATA_PHYS) == 0) {
int s;
int error;
s = splsoftvm();
error = bus_dmamap_load(ahc->buffer_dmat,
scb->dmamap,
csio->data_ptr,
csio->dxfer_len,
ahc_execute_scb,
scb, /*flags*/0);
if (error == EINPROGRESS) {
/*
* So as to maintain ordering,
* freeze the controller queue
* until our mapping is
* returned.
*/
xpt_freeze_simq(sim,
/*count*/1);
scb->io_ctx->ccb_h.status |=
CAM_RELEASE_SIMQ;
}
splx(s);
} else {
struct bus_dma_segment seg;
/* Pointer to physical buffer */
if (csio->dxfer_len > AHC_MAXTRANSFER_SIZE)
panic("ahc_setup_data - Transfer size "
"larger than can device max");
seg.ds_addr = (bus_addr_t)csio->data_ptr;
seg.ds_len = csio->dxfer_len;
ahc_execute_scb(scb, &seg, 1, 0);
}
} else {
struct bus_dma_segment *segs;
if ((ccb_h->flags & CAM_DATA_PHYS) != 0)
panic("ahc_setup_data - Physical segment "
"pointers unsupported");
if ((ccb_h->flags & CAM_SG_LIST_PHYS) == 0)
panic("ahc_setup_data - Virtual segment "
"addresses unsupported");
/* Just use the segments provided */
segs = (struct bus_dma_segment *)csio->data_ptr;
ahc_execute_scb(scb, segs, csio->sglist_cnt, 0);
}
} else {
ahc_execute_scb(scb, NULL, 0, 0);
}
}
static void
ahc_set_recoveryscb(struct ahc_softc *ahc, struct scb *scb) {
if ((scb->flags & SCB_RECOVERY_SCB) == 0) {
struct scb *list_scb;
scb->flags |= SCB_RECOVERY_SCB;
/*
* Take all queued, but not sent SCBs out of the equation.
* Also ensure that no new CCBs are queued to us while we
* try to fix this problem.
*/
if ((scb->io_ctx->ccb_h.status & CAM_RELEASE_SIMQ) == 0) {
xpt_freeze_simq(SCB_GET_SIM(ahc, scb), /*count*/1);
scb->io_ctx->ccb_h.status |= CAM_RELEASE_SIMQ;
}
/*
* Go through all of our pending SCBs and remove
* any scheduled timeouts for them. We will reschedule
* them after we've successfully fixed this problem.
*/
LIST_FOREACH(list_scb, &ahc->pending_scbs, pending_links) {
union ccb *ccb;
ccb = list_scb->io_ctx;
untimeout(ahc_timeout, list_scb, ccb->ccb_h.timeout_ch);
}
}
}
void
ahc_timeout(void *arg)
{
struct scb *scb;
struct ahc_softc *ahc;
long s;
int found;
u_int last_phase;
int target;
int lun;
int i;
char channel;
scb = (struct scb *)arg;
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
ahc = (struct ahc_softc *)scb->ahc_softc;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_lock(ahc, &s);
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
ahc_pause_and_flushwork(ahc);
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if ((scb->flags & SCB_ACTIVE) == 0) {
/* Previous timeout took care of me already */
Store a pointer to our softc in the kernel's SCB structure. In the past we stored this data in the CCB and attained the CCB via a pointer in the SCB. In ahc_timeout(), however, the timedout SCB may have already been completed (inherent race), meaning that the CCB could have been recycled, and the ahc pointer reset. Clean up the logic in ahc_search_qinfifo that deals with the busy device table. For some reason it assumed that the only valid time to search to see if additional lun entries should be checked was if lun 0 matched. Now we properly itterate through the necessary luns. The busy device table is used to detect invalid reselections, so a device would have had to perform an unexpected reselection for this to cause problems. Further, all luns are collapsed to a single entry unless we have external ram with large SCBs (3940AU models) so the chance of this happening was rather remote. Clean up the logic for dealing with the untagged queues. We now set a flag in the SCB that indicates that it is on the untagged queue instead of inferring this from the type and setup of the CCB pased into us by CAM. In ahc_timeout(), don't print the path of the SCB until the controller is paused and we are sure that it has not completed yet. This, in conjunction with referencing the ahc pointer in the SCB rather than the CCB in the SCB avoids panics in the case of a timedout scb completing just before the timeout handler runs. This turns out to be guaranteed if interrupt delivery is failing, as we run our interrupt handler to flush any "just missed events" when a timeout occurs. Mention the likelyhood of broken interrupts if a timedout SCB is completed by our call to ahc_intr().
2000-10-11 23:46:34 +00:00
printf("%s: Timedout SCB already complete. "
"Interrupts may not be functioning.\n", ahc_name(ahc));
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
unpause_sequencer(ahc);
ahc_unlock(ahc, &s);
return;
}
target = SCB_GET_TARGET(ahc, scb);
channel = SCB_GET_CHANNEL(ahc, scb);
lun = SCB_GET_LUN(scb);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_print_path(ahc, scb);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
printf("SCB 0x%x - timed out ", scb->hscb->tag);
/*
* Take a snapshot of the bus state and print out
* some information so we can track down driver bugs.
*/
last_phase = ahc_inb(ahc, LASTPHASE);
for (i = 0; i < num_phases; i++) {
if (last_phase == phase_table[i].phase)
break;
}
printf("%s", phase_table[i].phasemsg);
printf(", SEQADDR == 0x%x\n",
ahc_inb(ahc, SEQADDR0) | (ahc_inb(ahc, SEQADDR1) << 8));
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
printf("STACK == 0x%x, 0x%x, 0x%x, 0x%x\n",
ahc_inb(ahc, STACK) | (ahc_inb(ahc, STACK) << 8),
ahc_inb(ahc, STACK) | (ahc_inb(ahc, STACK) << 8),
ahc_inb(ahc, STACK) | (ahc_inb(ahc, STACK) << 8),
ahc_inb(ahc, STACK) | (ahc_inb(ahc, STACK) << 8));
printf("SXFRCTL0 == 0x%x\n", ahc_inb(ahc, SXFRCTL0));
ahc_dump_card_state(ahc);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (scb->sg_count > 0) {
for (i = 0; i < scb->sg_count; i++) {
printf("sg[%d] - Addr 0x%x : Length %d\n",
i,
scb->sg_list[i].addr,
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
scb->sg_list[i].len & AHC_SG_LEN_MASK);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
}
}
if (scb->flags & (SCB_DEVICE_RESET|SCB_ABORT)) {
/*
* Been down this road before.
* Do a full bus reset.
*/
bus_reset:
ahc_set_transaction_status(scb, CAM_CMD_TIMEOUT);
found = ahc_reset_channel(ahc, channel, /*Initiate Reset*/TRUE);
printf("%s: Issued Channel %c Bus Reset. "
"%d SCBs aborted\n", ahc_name(ahc), channel, found);
} else {
/*
* If we are a target, transition to bus free and report
* the timeout.
*
* The target/initiator that is holding up the bus may not
* be the same as the one that triggered this timeout
* (different commands have different timeout lengths).
* If the bus is idle and we are actiing as the initiator
* for this request, queue a BDR message to the timed out
* target. Otherwise, if the timed out transaction is
* active:
* Initiator transaction:
* Stuff the message buffer with a BDR message and assert
* ATN in the hopes that the target will let go of the bus
* and go to the mesgout phase. If this fails, we'll
* get another timeout 2 seconds later which will attempt
* a bus reset.
*
* Target transaction:
* Transition to BUS FREE and report the error.
* It's good to be the target!
*/
u_int active_scb_index;
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
u_int saved_scbptr;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
saved_scbptr = ahc_inb(ahc, SCBPTR);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
active_scb_index = ahc_inb(ahc, SCB_TAG);
if (last_phase != P_BUSFREE
Sync Perforce IDs, add tranceiver state change support, and correct numerous error recovery buglets. Many thanks to Tor Egge for his assistance in diagnosing problems with the error recovery code. aic7xxx.c: Report missed bus free events using their own sequencer interrupt code to avoid confusion with other "bad phase" interrupts. Remove a delay used in debugging. This delay could only be hit in certain, very extreme, error recovery scenarios. Handle transceiver state changes correctly. You can now plug an SE device into a hot-plug LVD bus without hanging the controller. When stepping through a critical section, panic if we step more than a reasonable number of times. After a bus reset, disable bus reset interupts until we either our first attempt to (re)select another device, or another device attemps to select us. This removes the need to busy wait in kernel for the scsi reset line to fall yet still ensures we see any reset events that impact the state of either our initiator or target roles. Before this change, we had the potential of servicing a "storm" of reset interrupts if the reset line was held for a significant amount of time. Indicate the current sequencer address whenever we dump the card's state. aic7xxx.reg: Transceiver state change register definitions. Add the missed bussfree sequencer interrupt code. Re-enable the scsi reset interrupt if it has been disabled before every attempt to (re)select a device and when we have been selected as a target. When being (re)selected, check to see if the selection dissappeared just after we enabled our bus free interrupt. If the bus has gone free again, go back to the idle loop and wait for another selection. Note two locations where we should change our behavior if ATN is still raised. If ATN is raised during the presentation of a command complete or disconnect message, we should ignore the message and expect the target to put us in msgout phase. We don't currently do this as it requires some code re-arrangement so that critical sections can be properly placed around our handling of these two events. Otherwise, we cannot guarantee that the check of ATN is atomic relative to our acking of the message in byte (the kernel could assert ATN). Only set the IDENTIFY_SEEN flag after we have settled on the SCB for this transaction. The kernel looks at this flag before assuming that SCB_TAG is valid. This avoids confusion during certain types of error recovery. Add a critical section around findSCB. We cannot allow the kernel to remove an entry from the disconnected list while we are traversing it. Ditto for get_free_or_disc_scb. aic7xxx_freebsd.c: Only assume that SCB_TAG is accurate if IDENTIFY_SEEN is set in SEQ_FLAGS. Fix a typo that caused us to execute some code for the non-SCB paging case when paging SCBs. This only occurred during error recovery.
2000-11-10 20:13:41 +00:00
&& (ahc_inb(ahc, SEQ_FLAGS) & IDENTIFY_SEEN) != 0
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
&& (active_scb_index < ahc->scb_data->numscbs)) {
struct scb *active_scb;
/*
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
* If the active SCB is not us, assume that
* the active SCB has a longer timeout than
* the timedout SCB, and wait for the active
* SCB to timeout.
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
*/
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
active_scb = ahc_lookup_scb(ahc, active_scb_index);
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
if (active_scb != scb) {
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
struct ccb_hdr *ccbh;
uint64_t newtimeout;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_print_path(ahc, scb);
printf("Other SCB Timeout%s",
(scb->flags & SCB_OTHERTCL_TIMEOUT) != 0
? " again\n" : "\n");
scb->flags |= SCB_OTHERTCL_TIMEOUT;
newtimeout =
MAX(active_scb->io_ctx->ccb_h.timeout,
scb->io_ctx->ccb_h.timeout);
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
newtimeout *= hz;
newtimeout /= 1000;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ccbh = &scb->io_ctx->ccb_h;
scb->io_ctx->ccb_h.timeout_ch =
ahc_eisa.c: Initialize rid to 0. This doesn't seem to make any difference (the driver doesn't care what rid it gets and no-one seems to check rid's value), but follows standard conventions. Pass in our device_t to ahc_alloc(). We now use device_T softc storage, so passing NULL results in a panic. Set the unit number in our softc so that the driver core can retrieve it. ahc_pci.c: Set the unit number in our softc so that the driver core can retrieve it. aic7770.c: Insert our softc into the list of softcs when initialization is successful. aic7xxx.c: Remove a workaround for an aic7895 bug we will never trigger. Add additional diagnostic info to ahc_dump_card_state(). Always panic the system if a sequencer assertion fails. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Replace a hard coded number with a constant. Guard against looping forever in ahc_pause_and_flushwork(). A hot eject or card failure may make the intstat register return 0xFF, so limit the number of interrupts we'll process. Correct the code in ahc_search_qinfifo() that guarantees that the sequencer will see an abort collision if the qinfifo is modified when a DMA is in progress. We now do this fixup after modifying the queue. This guarantees that the HSCB we place at the head of the queue is not the same as the old head. Using "next hscb" (guaranteed not to be the same as the first SCB) before clearing the queue could free up the original head hscb to be used during a remove operation placing it again at the head of the qinfifo. aic7xxx.h: Reduce the maximum number of outstanding commands to 253 from 254. To handle our output queue correctly on machines that only support 32bit stores, we must clear the array 4 bytes at a time. To avoid colliding with a DMA write from the sequencer, we must be sure that 4 slots are empty when we write to clear the queue. This reduces us to 253 SCBs: 1 that just completed and the known three additional empty slots in the queue that preceed it. Yahoo was able to force this race on one of their systems. Interrupts were disabled for such a time that the entire output queue was filled (254 entries complete without any processing), and our 32bit write to clear the status clobbered one entry. Add a feature tag for devices that are removable. aic7xxx.reg: Never use the sequencer interrupt value of 0xF0. We need to guanrantee that an INTSTAT value of 0xFF can only occur during card failure or a hot-eject. Align the busy targets table with the begining of scratch space. This seems to appease a chip bug in the aic7895. aic7xxx.seq: Be sure to disable select-out after a bus free event that occurs early in a selection. If we don't disable select-out, we will believe that it is enabled even though a new selection will never occur. Move the clearing of SELDI to just before a jump. This appeases another chip bug of the aic7895. Make the target mode command loop a bit more efficient. AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. Properly cleanup the last SCB we tested against should we fail to properly find an SCB for a reselection. Add some additional sequencer debugging code. aic7xxx_freebsd.c: Limit the driver to 253 outstanding commands per adapter. Guard against overflow in timeout handling. aic7xxx_inline.h: AHC_SCB_BTT is a "flag" not a "feature". Check the right field in the softc. aic7xxx_pci.c: Set the removable feature for the apa1480 cardbus and the 29160C Compact PCI card. Don't report high byte termination information for narrow cards. Use a PCI read rather than a questionable delay when fetching/setting termination settings.
2001-01-22 21:03:48 +00:00
timeout(ahc_timeout, scb, newtimeout);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_unlock(ahc, &s);
return;
}
/* It's us */
if ((scb->hscb->control & TARGET_SCB) != 0) {
/*
* Send back any queued up transactions
* and properly record the error condition.
*/
ahc_freeze_devq(ahc, scb);
ahc_set_transaction_status(scb,
CAM_CMD_TIMEOUT);
ahc_freeze_scb(scb);
ahc_done(ahc, scb);
/* Will clear us from the bus */
restart_sequencer(ahc);
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
ahc_unlock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
return;
}
ahc_set_recoveryscb(ahc, active_scb);
Clean up error recovery code: aic7xxx.c: In target mode, reset the TQINPOS on every restart of the sequencer. In the past we did this only during a bus reset, but there are other reasons the sequencer might be reset. In ahc_clear_critical_section(), disable pausing chip interrupts while we step the sequencer out of a critical section. This avoids the possibility of getting a pausing interrupt (unexpected bus free, bus reset, etc.) that would prevent the sequencer from stepping. Send the correct async notifications in the case of a BDR or bus reset. In ahc_loadseq(), correct the calculation of our critical sections. In some cases, the sections would be larger than needed. aic7xxx.h: Remove an unused SCB flag. aic7xxx.seq: MK_MESSAGE is cleared by the kernel, there is no need to waste a sequencer instruction clearing it. aic7xxx_freebsd.c: Go through the host message loop instead of issuing a single byte message directly in the ahc_timeout() case where we are currently on the bus to the device. The effect is the same, but this way we get a nice printf saying that an expected BDR was delivered instead of an unexpected bus free. If we are requeuing an SCB for an error recovery action, be sure to set the DISCONNECTED flag in the in-core version of the SCB. This ensures that, in the SCB-paging case, the sequencer will still recognize the reselection as valid even if the version of the SCB with this flag set was never previously paged out to system memory. In the non-paging case, set the MK_MESSAGE flag in SCB_CONTROL directly. aic7xxx_pci.c: Enable the Memeory Write and Invalidate bug workaround for all aic7880 chips with revs < 1. This bug is rarely triggered in FreeBSD as most transfers end on cache-aligned boundaries, but a recheck of my references indicates that these chips are affected.
2000-10-09 01:46:01 +00:00
ahc_outb(ahc, MSG_OUT, HOST_MSG);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_outb(ahc, SCSISIGO, last_phase|ATNO);
ahc_print_path(ahc, active_scb);
printf("BDR message in message buffer\n");
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
active_scb->flags |= SCB_DEVICE_RESET;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
active_scb->io_ctx->ccb_h.timeout_ch =
timeout(ahc_timeout, (caddr_t)active_scb, 2 * hz);
unpause_sequencer(ahc);
} else {
int disconnected;
/* XXX Shouldn't panic. Just punt instead */
if ((scb->hscb->control & TARGET_SCB) != 0)
panic("Timed-out target SCB but bus idle");
if (last_phase != P_BUSFREE
&& (ahc_inb(ahc, SSTAT0) & TARGET) != 0) {
/* XXX What happened to the SCB? */
/* Hung target selection. Goto busfree */
printf("%s: Hung target selection\n",
ahc_name(ahc));
restart_sequencer(ahc);
Bring in a slew of fixes that were supposed to be in the last commit. In ahc_search_qinfifo, the SEARCH_REMOVE case must also handle an SCB that has been removed from the QINFIFO but not yet been fully dmaed to the card. Correct locking for ahc_get_scb() calls. Set SCB syncrate settings in ahc_execute_scb() to avoid a race condition that could allow a newly queued SCB to be missed by ahc_update_pending_syncrates(). When notifying the system of transfer negotiation updates, only set the valid bits for tagged queuing and disconnection if the path is fully qualified. Sync/Wide settins apply to all luns of a target, but tagged queuing and disconnection may change on a per-lun basis. Add missing ahc_unlock() calls in ahc_timeout() for the target mode case.
2000-10-06 04:01:06 +00:00
ahc_unlock(ahc, &s);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
return;
}
if (ahc_search_qinfifo(ahc, target, channel, lun,
scb->hscb->tag, ROLE_INITIATOR,
/*status*/0, SEARCH_COUNT) > 0) {
disconnected = FALSE;
} else {
disconnected = TRUE;
}
if (disconnected) {
ahc_set_recoveryscb(ahc, scb);
/*
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
* Actually re-queue this SCB in an attempt
* to select the device before it reconnects.
* In either case (selection or reselection),
* we will now issue a target reset to the
* timed-out device.
*
Clean up error recovery code: aic7xxx.c: In target mode, reset the TQINPOS on every restart of the sequencer. In the past we did this only during a bus reset, but there are other reasons the sequencer might be reset. In ahc_clear_critical_section(), disable pausing chip interrupts while we step the sequencer out of a critical section. This avoids the possibility of getting a pausing interrupt (unexpected bus free, bus reset, etc.) that would prevent the sequencer from stepping. Send the correct async notifications in the case of a BDR or bus reset. In ahc_loadseq(), correct the calculation of our critical sections. In some cases, the sections would be larger than needed. aic7xxx.h: Remove an unused SCB flag. aic7xxx.seq: MK_MESSAGE is cleared by the kernel, there is no need to waste a sequencer instruction clearing it. aic7xxx_freebsd.c: Go through the host message loop instead of issuing a single byte message directly in the ahc_timeout() case where we are currently on the bus to the device. The effect is the same, but this way we get a nice printf saying that an expected BDR was delivered instead of an unexpected bus free. If we are requeuing an SCB for an error recovery action, be sure to set the DISCONNECTED flag in the in-core version of the SCB. This ensures that, in the SCB-paging case, the sequencer will still recognize the reselection as valid even if the version of the SCB with this flag set was never previously paged out to system memory. In the non-paging case, set the MK_MESSAGE flag in SCB_CONTROL directly. aic7xxx_pci.c: Enable the Memeory Write and Invalidate bug workaround for all aic7880 chips with revs < 1. This bug is rarely triggered in FreeBSD as most transfers end on cache-aligned boundaries, but a recheck of my references indicates that these chips are affected.
2000-10-09 01:46:01 +00:00
* Set the MK_MESSAGE control bit indicating
* that we desire to send a message. We
* also set the disconnected flag since
* in the paging case there is no guarantee
* that our SCB control byte matches the
* version on the card. We don't want the
* sequencer to abort the command thinking
* an unsolicited reselection occurred.
*/
scb->hscb->control |= MK_MESSAGE|DISCONNECTED;
scb->flags |= SCB_DEVICE_RESET;
/*
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
* Remove any cached copy of this SCB in the
* disconnected list in preparation for the
* queuing of our abort SCB. We use the
* same element in the SCB, SCB_NEXT, for
* both the qinfifo and the disconnected list.
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
*/
Correct corruption of the qinfifo in ahc_search_qinififo() for all non-LVD controllers. We only need to take special action on the qinfifo if we have dectected the case of an SCB that has been removed from the qinfifo but has not been fully DMAed to the controller. A missing conditional caused this code to be executed every time an SCB was aborted from the queue Don't attempt to print the path of an SCB that has been freed. Clean up the traversal of the pending scb list in ahc_update_pending_syncrates(). This has no functional change. Correct ahc_timeout()'s requeing of a timedout SCB to effect a recovery action. We now use ahc_qinfifo_requeue() and a new function ahc_qinfifo_count() instead of performing the requeue inline. The old code did not conform to the new qinfifo method. Clear the timedout SCB from the disconnected list. This ensures that the SCB_NEXT field is free to be used for queuing us to the qinfifo.
2000-10-08 03:37:52 +00:00
ahc_search_disc_list(ahc, target, channel,
lun, scb->hscb->tag,
/*stop_on_first*/TRUE,
/*remove*/TRUE,
Clean up error recovery code: aic7xxx.c: In target mode, reset the TQINPOS on every restart of the sequencer. In the past we did this only during a bus reset, but there are other reasons the sequencer might be reset. In ahc_clear_critical_section(), disable pausing chip interrupts while we step the sequencer out of a critical section. This avoids the possibility of getting a pausing interrupt (unexpected bus free, bus reset, etc.) that would prevent the sequencer from stepping. Send the correct async notifications in the case of a BDR or bus reset. In ahc_loadseq(), correct the calculation of our critical sections. In some cases, the sections would be larger than needed. aic7xxx.h: Remove an unused SCB flag. aic7xxx.seq: MK_MESSAGE is cleared by the kernel, there is no need to waste a sequencer instruction clearing it. aic7xxx_freebsd.c: Go through the host message loop instead of issuing a single byte message directly in the ahc_timeout() case where we are currently on the bus to the device. The effect is the same, but this way we get a nice printf saying that an expected BDR was delivered instead of an unexpected bus free. If we are requeuing an SCB for an error recovery action, be sure to set the DISCONNECTED flag in the in-core version of the SCB. This ensures that, in the SCB-paging case, the sequencer will still recognize the reselection as valid even if the version of the SCB with this flag set was never previously paged out to system memory. In the non-paging case, set the MK_MESSAGE flag in SCB_CONTROL directly. aic7xxx_pci.c: Enable the Memeory Write and Invalidate bug workaround for all aic7880 chips with revs < 1. This bug is rarely triggered in FreeBSD as most transfers end on cache-aligned boundaries, but a recheck of my references indicates that these chips are affected.
2000-10-09 01:46:01 +00:00
/*save_state*/FALSE);
/*
* In the non-paging case, the sequencer will
* never re-reference the in-core SCB.
* To make sure we are notified during
* reslection, set the MK_MESSAGE flag in
* the card's copy of the SCB.
*/
Sync Perforce IDs, add tranceiver state change support, and correct numerous error recovery buglets. Many thanks to Tor Egge for his assistance in diagnosing problems with the error recovery code. aic7xxx.c: Report missed bus free events using their own sequencer interrupt code to avoid confusion with other "bad phase" interrupts. Remove a delay used in debugging. This delay could only be hit in certain, very extreme, error recovery scenarios. Handle transceiver state changes correctly. You can now plug an SE device into a hot-plug LVD bus without hanging the controller. When stepping through a critical section, panic if we step more than a reasonable number of times. After a bus reset, disable bus reset interupts until we either our first attempt to (re)select another device, or another device attemps to select us. This removes the need to busy wait in kernel for the scsi reset line to fall yet still ensures we see any reset events that impact the state of either our initiator or target roles. Before this change, we had the potential of servicing a "storm" of reset interrupts if the reset line was held for a significant amount of time. Indicate the current sequencer address whenever we dump the card's state. aic7xxx.reg: Transceiver state change register definitions. Add the missed bussfree sequencer interrupt code. Re-enable the scsi reset interrupt if it has been disabled before every attempt to (re)select a device and when we have been selected as a target. When being (re)selected, check to see if the selection dissappeared just after we enabled our bus free interrupt. If the bus has gone free again, go back to the idle loop and wait for another selection. Note two locations where we should change our behavior if ATN is still raised. If ATN is raised during the presentation of a command complete or disconnect message, we should ignore the message and expect the target to put us in msgout phase. We don't currently do this as it requires some code re-arrangement so that critical sections can be properly placed around our handling of these two events. Otherwise, we cannot guarantee that the check of ATN is atomic relative to our acking of the message in byte (the kernel could assert ATN). Only set the IDENTIFY_SEEN flag after we have settled on the SCB for this transaction. The kernel looks at this flag before assuming that SCB_TAG is valid. This avoids confusion during certain types of error recovery. Add a critical section around findSCB. We cannot allow the kernel to remove an entry from the disconnected list while we are traversing it. Ditto for get_free_or_disc_scb. aic7xxx_freebsd.c: Only assume that SCB_TAG is accurate if IDENTIFY_SEEN is set in SEQ_FLAGS. Fix a typo that caused us to execute some code for the non-SCB paging case when paging SCBs. This only occurred during error recovery.
2000-11-10 20:13:41 +00:00
if ((ahc->flags & AHC_PAGESCBS) == 0) {
Clean up error recovery code: aic7xxx.c: In target mode, reset the TQINPOS on every restart of the sequencer. In the past we did this only during a bus reset, but there are other reasons the sequencer might be reset. In ahc_clear_critical_section(), disable pausing chip interrupts while we step the sequencer out of a critical section. This avoids the possibility of getting a pausing interrupt (unexpected bus free, bus reset, etc.) that would prevent the sequencer from stepping. Send the correct async notifications in the case of a BDR or bus reset. In ahc_loadseq(), correct the calculation of our critical sections. In some cases, the sections would be larger than needed. aic7xxx.h: Remove an unused SCB flag. aic7xxx.seq: MK_MESSAGE is cleared by the kernel, there is no need to waste a sequencer instruction clearing it. aic7xxx_freebsd.c: Go through the host message loop instead of issuing a single byte message directly in the ahc_timeout() case where we are currently on the bus to the device. The effect is the same, but this way we get a nice printf saying that an expected BDR was delivered instead of an unexpected bus free. If we are requeuing an SCB for an error recovery action, be sure to set the DISCONNECTED flag in the in-core version of the SCB. This ensures that, in the SCB-paging case, the sequencer will still recognize the reselection as valid even if the version of the SCB with this flag set was never previously paged out to system memory. In the non-paging case, set the MK_MESSAGE flag in SCB_CONTROL directly. aic7xxx_pci.c: Enable the Memeory Write and Invalidate bug workaround for all aic7880 chips with revs < 1. This bug is rarely triggered in FreeBSD as most transfers end on cache-aligned boundaries, but a recheck of my references indicates that these chips are affected.
2000-10-09 01:46:01 +00:00
ahc_outb(ahc, SCBPTR, scb->hscb->tag);
ahc_outb(ahc, SCB_CONTROL,
ahc_inb(ahc, SCB_CONTROL)
| MK_MESSAGE);
}
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
/*
* Clear out any entries in the QINFIFO first
* so we are the next SCB for this target
* to run.
*/
ahc_search_qinfifo(ahc,
SCB_GET_TARGET(ahc, scb),
channel, SCB_GET_LUN(scb),
SCB_LIST_NULL,
ROLE_INITIATOR,
CAM_REQUEUE_REQ,
SEARCH_COMPLETE);
ahc_print_path(ahc, scb);
printf("Queuing a BDR SCB\n");
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
ahc_qinfifo_requeue_tail(ahc, scb);
ahc_outb(ahc, SCBPTR, saved_scbptr);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
scb->io_ctx->ccb_h.timeout_ch =
timeout(ahc_timeout, (caddr_t)scb, 2 * hz);
unpause_sequencer(ahc);
} else {
/* Go "immediatly" to the bus reset */
/* This shouldn't happen */
ahc_set_recoveryscb(ahc, scb);
ahc_print_path(ahc, scb);
printf("SCB %d: Immediate reset. "
"Flags = 0x%x\n", scb->hscb->tag,
scb->flags);
goto bus_reset;
}
}
}
ahc_unlock(ahc, &s);
}
static void
ahc_abort_ccb(struct ahc_softc *ahc, struct cam_sim *sim, union ccb *ccb)
{
union ccb *abort_ccb;
abort_ccb = ccb->cab.abort_ccb;
switch (abort_ccb->ccb_h.func_code) {
case XPT_ACCEPT_TARGET_IO:
case XPT_IMMED_NOTIFY:
case XPT_CONT_TARGET_IO:
{
struct tmode_tstate *tstate;
struct tmode_lstate *lstate;
struct ccb_hdr_slist *list;
cam_status status;
status = ahc_find_tmode_devs(ahc, sim, abort_ccb, &tstate,
&lstate, TRUE);
if (status != CAM_REQ_CMP) {
ccb->ccb_h.status = status;
break;
}
if (abort_ccb->ccb_h.func_code == XPT_ACCEPT_TARGET_IO)
list = &lstate->accept_tios;
else if (abort_ccb->ccb_h.func_code == XPT_IMMED_NOTIFY)
list = &lstate->immed_notifies;
else
list = NULL;
if (list != NULL) {
struct ccb_hdr *curelm;
int found;
curelm = SLIST_FIRST(list);
found = 0;
if (curelm == &abort_ccb->ccb_h) {
found = 1;
SLIST_REMOVE_HEAD(list, sim_links.sle);
} else {
while(curelm != NULL) {
struct ccb_hdr *nextelm;
nextelm =
SLIST_NEXT(curelm, sim_links.sle);
if (nextelm == &abort_ccb->ccb_h) {
found = 1;
SLIST_NEXT(curelm,
sim_links.sle) =
SLIST_NEXT(nextelm,
sim_links.sle);
break;
}
curelm = nextelm;
}
}
if (found) {
abort_ccb->ccb_h.status = CAM_REQ_ABORTED;
xpt_done(abort_ccb);
ccb->ccb_h.status = CAM_REQ_CMP;
} else {
aic7xxx.c: Filter incoming transfer negotiation requests to ensure they never exceed the settings specified by the user. In restart sequencer attempt to deal with a bug in the aic7895. If a third party reset occurs at just the right time, the stack register can lock up. When restarting the sequencer after handling the SCSI reset, poke SEQADDR1 before resting the sequencers program counter. When something strange happens, dump the card's transaction state via ahc_dump_card_state(). This should aid in debugging. Handle request sense transactions via the QINFIFO instead of attaching them to the waiting queue directly. The waiting queue consumes card SCB resources and, in the pathological case of every target on the bus beating our selection attemps and issuing a check condition, could have caused us to run out of SCBs. I have never seen this happen, and only early cards with 3 or 4 SCBs had any real chance of ever getting into this state. Add additional sequencer interrupt codes to support firmware diagnostics. The diagnostic code is enabled with the AHC_DEBUG_SEQUENCER kernel option. Make it possible to switch into and out of target mode on the fly. The card comes up by default as an initiator but will switch into target mode as soon as an enable lun operation is performed. As always, target mode behavior is gated by the AHC_TMODE_ENABLE kernel option so most users will not be affected by this change. In ahc_update_target_msg_request(), also issue a new request if the ppr_options have changed. Never issue a PPR as a target. It is forbidden by the spec. Correct a bug in ahc_parse_msg() that prevented us from responding to PPR messages as a target. Mark SCBs that are on the untagged queue with a flag instead of checking several fields in the SCB to see if the SCB should be on the queue. This makes it easier for things like automatic request sense requests to be queued without touching the untagged queues even though they are untagged requests. When dealing with ignore wide residue messages that occur in the middle of a transfer, reset HADDR, not SHADDR for non-ultra2 chips. Although SHADDR is where the firmware fetches the ending transfer address for a save data pointers request, it is readonly. Setting HADDR has the side effect of also updating SHADDR. Cleanup the output of ahc_dump_card_state() by nulling out the free scb list in the non-paging case. The free list is only used if we must page SCBs. Correct the transmission of cdbs > 12 bytes in length. When swapping HSCBs prior to notifing the sequencer of the new transaction, the bus address pointer for the cdb must also be recalculated to reflect its new location. We now defer the calculation of the cdb address until just before queing it to the card. When pulling transfer negotiation settings out of scratch ram, convert 5MHz/clock doubled settings to 10MHz. Add a new function ahc_qinfifo_requeue_tail() for use by error recovery actions and auto-request sense operations. These operations always occur when the sequencer is paused, so we can avoid the extra expense incurred in the normal SCB queue method. Use the BMOV instruction for all single byte moves on controllers that support it. The bmov instruction is twice as fast as an AND with an immediate of 0xFF as is used on older controllers. Correct a few bugs in ahc_dump_card_state(). If we have hardware assisted queue registers, use them to get the sequencer's idea of the head of the queue. When enumerating the untagged queue, it helps to use the correct index for the queue. aic7xxx.h: Indicate via a feature flag, which controllers can take on both the target and the initiator role at the same time. Add the AHC_SEQUENCER_DEBUG flag. Add the SCB_CDB32_PTR flag used for dealing with cdbs with lengths between 13 and 32 bytes. Add new prototypes. aic7xxx.reg: Allow the SCSIBUSL register to be written to. This is required to fix a selection timeout problem on the 7892/99. Cleanup the sequencer interrupt codes so that all debugging codes are grouped at the end of the list. Correct the definition of the ULTRA_ENB and DISC_DSB locations in scratch ram. This prevented the driver from properly honoring these settings when no serial eeprom was available. Remove an unused sequencer flag. aic7xxx.seq: Just before a potential select-out, clear the SCSIBUSL register. Occasionally, during a selection timeout, the contents of the register may be presented on the bus, causing much confusion. Add sequencer diagnostic code to detect software and or hardware bugs. The code attempts to verify most list operations so any corruption is caught before it occurs. We also track information about why a particular reconnection request was rejected. Don't clobber the digital REQ/ACK filter setting in SXFRCTL0 when clearing the channel. Fix a target mode bug that would cause us to return busy status instead of queue full in respnse to a tagged transaction. Cleanup the overrun case. It turns out that by simply butting the chip in bitbucket mode, it will ack any bytes until the phase changes. This drasticaly simplifies things. Prior to leaving the data phase, make sure that the S/G preload queue is empty. Remove code to place a request sense request on the waiting queue. This is all handled by the kernel now. Change the semantics of "findSCB". In the past, findSCB ensured that a freshly paged in SCB appeared on the disconnected list. The problem with this is that there is no guarantee that the paged in SCB is for a disconnected transation. We now defer any list manipulation to the caller who usually discards the SCB via the free list. Inline some busy target table operations. Add a critical section to protect adding an SCB to the disconnected list. aic7xxx_freebsd.c: Handle changes in the transfer negotiation setting API to filter incoming requests. No filtering is necessary for "goal" requests from the XPT. Set the SCB_CDB32_PTR flag when queing a transaction with a large cdb. In ahc_timeout, only take action if the active SCB is the timedout SCB. This deals with the case of two transactions to the same device with different timeout values. Use ahc_qinfifo_requeu_tail() instead of home grown version. aic7xxx_inline.h: Honor SCB_CDB32_PTR when queuing a new request. aic7xxx_pci.c: Use the maximum data fifo threshold for all chips.
2000-10-31 18:43:29 +00:00
xpt_print_path(abort_ccb->ccb_h.path);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
printf("Not found\n");
ccb->ccb_h.status = CAM_PATH_INVALID;
}
break;
}
/* FALLTHROUGH */
}
case XPT_SCSI_IO:
/* XXX Fully implement the hard ones */
ccb->ccb_h.status = CAM_UA_ABORT;
break;
default:
ccb->ccb_h.status = CAM_REQ_INVALID;
break;
}
xpt_done(ccb);
}
void
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
ahc_send_async(struct ahc_softc *ahc, char channel, u_int target,
u_int lun, ac_code code)
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
{
struct ccb_trans_settings cts;
struct cam_path *path;
void *arg;
int error;
arg = NULL;
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
error = ahc_create_path(ahc, channel, target, lun, &path);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (error != CAM_REQ_CMP)
return;
switch (code) {
case AC_TRANSFER_NEG:
#ifdef AHC_NEW_TRAN_SETTINGS
cts.type = CTS_TYPE_CURRENT_SETTINGS;
#else
cts.flags = CCB_TRANS_CURRENT_SETTINGS;
#endif
cts.ccb_h.path = path;
Add Perforce RCSIDs for easy revision correlation to my local tree. ahc_pci.c: Bring back the AHC_ALLOW_MEMIO option at least until the memory mapped I/O problem on the SuperMicro 370DR3 is better understood. aic7xxx.c: If we see a spurious SCSI interrupt, attempt to clear it and continue by unpausing the sequencer. Change the interface to ahc_send_async(). Some async messages need to be broadcast to all the luns of a target or all the targets of a bus. This is easier to achieve by passing explicit channel, target, and lun parameters instead of attempting to construct a device info struct to match. Filter the sync parameters for the PPR message in exactly the same way we do for an old fashioned SDTR message. Correct some typos and correct a panic message. Handle rejected PPR messages. In ahc_handle_msg_reject(), let ahc_build_transfer_msg() build any additional transfer messages instead of doing this inline. aic7xxx.h: Increase the size of both msgout_buf and msgin_buf to better accomodate PPR messages. aic7xxx_freebsd.c: Update for change in ahc_send_async() parameters. aic7xxx_freebsd.h Update for change in ahc_send_async() parameters. Honor AHC_ALLOW_MEMIO. aic7xxx_pci.c: Check the error register before going into full blown PCI interrupt handling. This avoids a few costly PCI configuration space reads when we run our PCI interrupt handler because another device sharing our interrupt line is more active than we are. Also unpause the sequencer after processing a PCI interrupt.
2000-09-22 22:18:05 +00:00
cts.ccb_h.target_id = target;
cts.ccb_h.target_lun = lun;
ahc_get_tran_settings(ahc, channel == 'A' ? ahc->our_id
: ahc->our_id_b,
channel, &cts);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
arg = &cts;
break;
case AC_SENT_BDR:
case AC_BUS_RESET:
break;
default:
panic("ahc_send_async: Unexpected async event");
}
xpt_async(code, path, arg);
aic7xxx.c: When restarting the sequencer, ensure that the SCBCNT register is 0. A non-zero count will prevent the setting of the CCSCBDIR bit in any future dma operations. The only time CCSCBCNT would be non-zero is if we happened to halt the dma during a reset, but even that should never happen. Better safe than sorry. When a command completes before the target responds to an ATN for a recovery command, we now notify the kernel so that any recovery operation requeued in the qinfifo can be removed safely. In the past, we did this in ahc_done(), but ahc_done() may be called without the card paused. This also avoids a recursive call to ahc_search_qinifo() which could have occurred if ahc_search_qinififo() happened to be the routine to complete a recovery action. Fix 8bit math used for adjusting the qinfifo. The index must be wrapped properly within the 256 entry array. We rely on the fact that qinfifonext is a uint8_t in most cases to handle this wrap, but we missed a few spots where the resultant calculation was promoted to an int. Change the way that we deal with aborting the first or second entry from the qinfifo. We now swap the first entry in the qinfifo with the "next queued scb" to force the sequencer to see an abort collision if we ever touch the qinififo while the sequencer is mid SCB dma. aic7xxx.reg: Add new MKMSG_FAILED sequencer interrupt. This displaced the BOGUS_TAG interrupt used in some previous sequencer code debugging. aic7xxx.seq: Increment our position in the qinfifo only once the dma is complete and we have verified that the queue has not been changed during our DMA. This simplifies code in the kernel. Protect against "instruction creep" when issuing a pausing sequencer interrupt. On at least the 7890/91/96/97, the sequencer will coast after issuing the interrupt for up to two instructions. In the past we delt with this by using carefully placed nops. Now we call a routine to issue the interrupt followed by a nop and a ret. Tell the kernel should an SCB complete with the MK_MESSAGE flag still set. This means the target ignored our ATN request. Clear the channel twice as we exit the data phase. On the aic7890/91, the S/G preload logic may require the second clearing to get the last S/G out of the FIFO. aic7xxx_freebsd.c: Don't bother searching the qinfifo for a doubly queued recovery scb in ahc_done. This case is handled by the core driver now. Free the path used to issue async callbacks after the callback is complete. aic7xxx_inline.h: Split the SCB queue routine into a routine that swaps the SCB with the "next queued SCB" and a routine that calls the swapping routine and notifies the card of the new SCB. The swapping routine is now also used by ahc_search_qinfifo.
2000-11-06 20:05:38 +00:00
xpt_free_path(path);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
}
void
ahc_platform_set_tags(struct ahc_softc *ahc,
struct ahc_devinfo *devinfo, int enable)
{
}
int
ahc_platform_alloc(struct ahc_softc *ahc, void *platform_arg)
{
Add the use of M_ZERO to some malloc calls. Submitted by: josh@zipperup.org Submitted by: Robert Drehmel <robd@gmx.net> Approved by: gibbs
2000-10-29 15:47:16 +00:00
ahc->platform_data = malloc(sizeof(struct ahc_platform_data), M_DEVBUF,
M_NOWAIT | M_ZERO);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
if (ahc->platform_data == NULL)
return (ENOMEM);
return (0);
}
void
ahc_platform_free(struct ahc_softc *ahc)
{
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
struct ahc_platform_data *pdata;
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
pdata = ahc->platform_data;
if (pdata != NULL) {
device_printf(ahc->dev_softc, "Platform free\n");
if (pdata->regs != NULL)
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
bus_release_resource(ahc->dev_softc,
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
pdata->regs_res_type,
pdata->regs_res_id,
pdata->regs);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
if (pdata->irq != NULL)
bus_release_resource(ahc->dev_softc,
pdata->irq_res_type,
0, pdata->irq);
if (pdata->sim_b != NULL) {
xpt_async(AC_LOST_DEVICE, pdata->path_b, NULL);
xpt_free_path(pdata->path_b);
xpt_bus_deregister(cam_sim_path(pdata->sim_b));
cam_sim_free(pdata->sim_b, /*free_devq*/TRUE);
}
if (pdata->sim != NULL) {
xpt_async(AC_LOST_DEVICE, pdata->path, NULL);
xpt_free_path(pdata->path);
xpt_bus_deregister(cam_sim_path(pdata->sim));
cam_sim_free(pdata->sim, /*free_devq*/TRUE);
}
if (pdata->eh != NULL)
EVENTHANDLER_DEREGISTER(shutdown_final, pdata->eh);
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
free(ahc->platform_data, M_DEVBUF);
}
}
int
ahc_softc_comp(struct ahc_softc *lahc, struct ahc_softc *rahc)
{
/* We don't sort softcs under FreeBSD so report equal always */
return (0);
}
ahc_eisa.c: ahc_pci.c: Add detach support. Make use of soft allocated on our behalf by newbus. For PCI devices, disable the mapping type we aren't using for extra protection from rogue code. aic7xxx_93cx6.c: aic7xxx_93cx6.h: Sync perforce IDs. aic7xxx_freebsd.c: Capture the eventhandle returned by EVENTHANDER_REGISTER so we can kill the handler off during detach. Use AHC_* constants instead of hard coded numbers in a few more places. Test PPR option state when deciding to "really" negotiate when the CAM_NEGOTIATE flag is passed in a CCB. Make use of core "ahc_pause_and_flushwork" routine in our timeout handler rather than re-inventing this code. Cleanup all of our resources (really!) in ahc_platform_free(). We should be all set to become a module now. Implement the core ahc_detach() routine shared by all of the FreeBSD front-ends. aic7xxx_freebsd.h: Softc storage for our event handler. Null implementation for the ahc_platform_flushwork() OSM callback. FreeBSD doesn't need this as XPT callbacks are safe from all contexts and are done directly in ahc_done(). aic7xxx_inline.h: Implement new lazy interrupt scheme. To avoid an extra PCI bus read, we first check our completion queues to see if any work has completed. If work is available, we assume that this is the source of the interrupt and skip reading INTSTAT. Any remaining interrupt status will be cleared by a second call to the interrupt handler should the interrupt line still be asserted. This drops the interrupt handler down to a single PCI bus read in the common case of I/O completion. This is the same overhead as in the not so distant past, but the extra sanity of perforning a PCI read after clearing the command complete interrupt and before running the completion queue to avoid missing command complete interrupts added a cycle. aic7xxx.c: During initialization, be sure to initialize all scratch ram locations before they are read to avoid parity errors. In this case, we use a new function, ahc_unbusy_tcl() to initialize the scratch ram busy target table. Replace instances of ahc_index_busy_tcl() used to unbusy a tcl without looking at the old value with ahc_unbusy_tcl(). Modify ahc_sent_msg so that it can find single byte messages. ahc_sent_msg is now used to determine if a transfer negotiation attempt resulted in a bus free. Be more careful in filtering out only the SCSI interrupts of interest in ahc_handle_scsiint. Rearrange interrupt clearing code to ensure that at least one PCI transaction occurrs after hitting CLRSINT1 and writting to CLRINT. CLRSINT1 writes take a bit to take effect, and the re-arrangement provides sufficient delay to ensure the write to CLRINT is effective. The old code might report a spurious interrupt on some "fast" chipsets. export ahc-update_target_msg_request for use by OSM code. If a target does not respond to our ATN request, clear it once we move to a non-message phase. This avoids sending a MSG_NOOP in some later message out phase. Use max lun and max target constants instead of hard-coded values. Use softc storage built into our device_t under FreeBSD. Fix a bug in ahc_free() that caused us to delete resources that were not allocated. Clean up any tstate/lstate info in ahc_free(). Clear the powerdown state in ahc_reset() so that registers can be accessed. Add a preliminary function for pausing the chip and processing any posted work. Add a preliminary suspend and resume functions. aic7xxx.h: Limit the number of supported luns to 64. We don't support information unit transfers, so this is the maximum that makes sense for these chips. Add a new flag AHC_ALL_INTERRUPTS that forces the processing of all interrupt state in a single invokation of ahc_intr(). When the flag is not set, we use the lazy interrupt handling scheme. Add data structures to store controller state while we are suspended. Use constants instead of hard coded values where appropriate. Correct some harmless "unsigned/signed" conflicts. aic7xxx.seq: Only perform the SCSIBUSL fix on ULTRA2 or newer controllers. Older controllers seem to be confused by this. In target mode, ignore PHASEMIS during data phases. This bit seems to be flakey on U160 controllers acting in target mode. aic7xxx_pci.c: Add support for the 29160C CPCI adapter. Add definitions for subvendor ID information available for devices with the "9005" vendor id. We currently use this information to determine if a multi-function device doesn't have the second channel hooked up on a board. Add rudimentary power mode code so we can put the controller into the D0 state. In the future this will be an OSM callback so that in FreeBSD we don't duplicate functionality provided by the PCI code. The powerstate code was added after I'd completed my regression tests on this code. Only capture "left over BIOS state" if the POWRDN setting is not set in HCNTRL. In target mode, don't bother sending incremental CRC data.
2000-12-20 01:11:37 +00:00
int
ahc_detach(device_t dev)
{
struct ahc_softc *ahc;
u_long s;
device_printf(dev, "detaching device\n");
ahc = device_get_softc(dev);
ahc_lock(ahc, &s);
bus_teardown_intr(dev, ahc->platform_data->irq, ahc->platform_data->ih);
ahc_unlock(ahc, &s);
ahc_free(ahc);
return (0);
}
Move aicasm to its own subdirectory. Separate our platform independent hooks from core driver functionality shared between platforms (FreeBSD and Linux at this time). Add sequencer workarounds for several chip->chipset interactions. Correct external SCB corruption problem on aic7895 based cards (3940AUW). Lots of cleanups resulting from the port to another OS.
2000-09-16 20:02:28 +00:00
#if UNUSED
static void
ahc_dump_targcmd(struct target_cmd *cmd)
{
uint8_t *byte;
uint8_t *last_byte;
int i;
byte = &cmd->initiator_channel;
/* Debugging info for received commands */
last_byte = &cmd[1].initiator_channel;
i = 0;
while (byte < last_byte) {
if (i == 0)
printf("\t");
printf("%#x", *byte++);
i++;
if (i == 8) {
printf("\n");
i = 0;
} else {
printf(", ");
}
}
}
#endif
Reference in New Issue Copy Permalink