Skip to contentSkip to conversation

ata: libata: Set DID_TIME_OUT for commands that actually timed out

PATCH Applied Niklas Cassel 2 participants 4 messages Oct 23, 2024 – Oct 24, 2024

Subsystems (2)
lore
Share & tools
lore

Niklas Cassel

author maintainer
#0 lore raw

PATCH 1/1 ata: libata: Set DID_TIME_OUT for commands that actually timed out

When ata_qc_complete() schedules a command for EH using ata_qc_schedule_eh(), blk_abort_request() will be called, which leads to req->q->mq_ops->timeout() / scsi_timeout() being called.

scsi_timeout(), if the LLDD has no abort handler (libata has no abort handler), will set host byte to DID_TIME_OUT, and then call scsi_eh_scmd_add() to add the command to EH.

Thus, when commands first enter libata's EH strategy_handler, all the commands that have been added to EH will have DID_TIME_OUT set.

Commit e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data") clears this bogus DID_TIME_OUT flag for all commands that reached libata's EH strategy_handler.

libata has its own flag (AC_ERR_TIMEOUT), that it sets for commands that have not received a completion at the time of entering EH.

ata_eh_worth_retry() has no special handling for AC_ERR_TIMEOUT, so by default timed out commands will get flag ATA_QCFLAG_RETRY set and will be retried after the port has been reset (ata_eh_link_autopsy() always triggers a port reset if any command has AC_ERR_TIMEOUT set).

For commands that have ATA_QCFLAG_RETRY set, but also has an error flag set (e.g. AC_ERR_TIMEOUT), ata_eh_finish() will not increment scmd->allowed, so the command will at most be retried (scmd->allowed number of times, which by default is set to 3).

However, scsi_eh_flush_done_q() will only retry commands for which scsi_noretry_cmd() returns false.

For commands that has DID_TIME_OUT set, if the command is either has FAILFAST or if the command is a passthrough command, scsi_noretry_cmd() will return true. Thus, such commands will never be retried.

Thus, make sure that libata sets SCSI's DID_TIME_OUT flag for commands that actually timed out (libata's AC_ERR_TIMEOUT flag), such that timed out commands will once again not be retried if they are also a FAILFAST or passthrough command.

Fixes e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data")
---
 drivers/ata/libata-eh.c | 1 +
 1 file changed, 1 insertion(+)
drivers/ata/libata-eh.c +1 -0
index fa41ea57a978..3b303d4ae37a 100644
@@ -651,6 +651,7 @@ void ata_scsi_cmd_error_handler(struct Scsi_Host *host, struct ata_port *ap,
 			/* the scmd has an associated qc */
 			if (!(qc->flags & ATA_QCFLAG_EH)) {
 				/* which hasn't failed yet, timeout */
+				set_host_byte(scmd, DID_TIME_OUT);
 				qc->err_mask |= AC_ERR_TIMEOUT;
 				qc->flags |= ATA_QCFLAG_EH;
 				nr_timedout++;
Signature · 1 line
2.47.0
author maintainer
#1 lore raw
Quoted 28 lines · Niklas Cassel

When ata_qc_complete() schedules a command for EH using ata_qc_schedule_eh(), blk_abort_request() will be called, which leads to req->q->mq_ops->timeout() / scsi_timeout() being called.

scsi_timeout(), if the LLDD has no abort handler (libata has no abort handler), will set host byte to DID_TIME_OUT, and then call scsi_eh_scmd_add() to add the command to EH.

Thus, when commands first enter libata's EH strategy_handler, all the commands that have been added to EH will have DID_TIME_OUT set.

Commit e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data") clears this bogus DID_TIME_OUT flag for all commands that reached libata's EH strategy_handler.

libata has its own flag (AC_ERR_TIMEOUT), that it sets for commands that have not received a completion at the time of entering EH.

ata_eh_worth_retry() has no special handling for AC_ERR_TIMEOUT, so by default timed out commands will get flag ATA_QCFLAG_RETRY set and will be retried after the port has been reset (ata_eh_link_autopsy() always triggers a port reset if any command has AC_ERR_TIMEOUT set).

For commands that have ATA_QCFLAG_RETRY set, but also has an error flag set (e.g. AC_ERR_TIMEOUT), ata_eh_finish() will not increment scmd->allowed, so the command will at most be retried (scmd->allowed number of times, which by default is set to 3).

Quoted 5 lines

However, scsi_eh_flush_done_q() will only retry commands for which scsi_noretry_cmd() returns false.

For commands that has DID_TIME_OUT set, if the command is either has FAILFAST or if the command is a passthrough command, scsi_noretry_cmd()

"if the command has either FAILFAST flag set or if ..."

I could fix up this sentence when applying.

Quoted 31 lines

will return true. Thus, such commands will never be retried.

Thus, make sure that libata sets SCSI's DID_TIME_OUT flag for commands that actually timed out (libata's AC_ERR_TIMEOUT flag), such that timed out commands will once again not be retried if they are also a FAILFAST or passthrough command.

Fixes e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data")
---
 drivers/ata/libata-eh.c | 1 +
 1 file changed, 1 insertion(+)
drivers/ata/libata-eh.c +1 -0
index fa41ea57a978..3b303d4ae37a 100644
@@ -651,6 +651,7 @@ void ata_scsi_cmd_error_handler(struct Scsi_Host *host, struct ata_port *ap,
 			/* the scmd has an associated qc */
 			if (!(qc->flags & ATA_QCFLAG_EH)) {
 				/* which hasn't failed yet, timeout */
+				set_host_byte(scmd, DID_TIME_OUT);
 				qc->err_mask |= AC_ERR_TIMEOUT;
 				qc->flags |= ATA_QCFLAG_EH;
 				nr_timedout++;
Signature · 1 line
2.47.0
maintainer Reviewed-by on v1 1/1
#2 lore raw
In reply to Niklas Cassel
Quoted 40 lines · Niklas Cassel

When ata_qc_complete() schedules a command for EH using ata_qc_schedule_eh(), blk_abort_request() will be called, which leads to req->q->mq_ops->timeout() / scsi_timeout() being called.

scsi_timeout(), if the LLDD has no abort handler (libata has no abort handler), will set host byte to DID_TIME_OUT, and then call scsi_eh_scmd_add() to add the command to EH.

Thus, when commands first enter libata's EH strategy_handler, all the commands that have been added to EH will have DID_TIME_OUT set.

Commit e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data") clears this bogus DID_TIME_OUT flag for all commands that reached libata's EH strategy_handler.

libata has its own flag (AC_ERR_TIMEOUT), that it sets for commands that have not received a completion at the time of entering EH.

ata_eh_worth_retry() has no special handling for AC_ERR_TIMEOUT, so by default timed out commands will get flag ATA_QCFLAG_RETRY set and will be retried after the port has been reset (ata_eh_link_autopsy() always triggers a port reset if any command has AC_ERR_TIMEOUT set).

For commands that have ATA_QCFLAG_RETRY set, but also has an error flag set (e.g. AC_ERR_TIMEOUT), ata_eh_finish() will not increment scmd->allowed, so the command will at most be retried (scmd->allowed number of times, which by default is set to 3).

However, scsi_eh_flush_done_q() will only retry commands for which scsi_noretry_cmd() returns false.

For commands that has DID_TIME_OUT set, if the command is either has FAILFAST or if the command is a passthrough command, scsi_noretry_cmd() will return true. Thus, such commands will never be retried.

Thus, make sure that libata sets SCSI's DID_TIME_OUT flag for commands that actually timed out (libata's AC_ERR_TIMEOUT flag), such that timed out commands will once again not be retried if they are also a FAILFAST or passthrough command.

Quoted 5 lines
Fixes e5dd410acb34 ("ata: libata: Clear DID_TIME_OUT for ATA PT commands with sense data")

Looks good.

Quoted 16 lines
---
 drivers/ata/libata-eh.c | 1 +
 1 file changed, 1 insertion(+)
drivers/ata/libata-eh.c +1 -0
index fa41ea57a978..3b303d4ae37a 100644
@@ -651,6 +651,7 @@ void ata_scsi_cmd_error_handler(struct Scsi_Host *host, struct ata_port *ap,
 			/* the scmd has an associated qc */
 			if (!(qc->flags & ATA_QCFLAG_EH)) {
 				/* which hasn't failed yet, timeout */
+				set_host_byte(scmd, DID_TIME_OUT);
 				qc->err_mask |= AC_ERR_TIMEOUT;
 				qc->flags |= ATA_QCFLAG_EH;
 				nr_timedout++;
Signature · 2 lines
Damien Le Moal
Western Digital Research
author maintainer Applied
#3 lore raw
In reply to Niklas Cassel
Quoted 9 lines · Niklas Cassel

When ata_qc_complete() schedules a command for EH using ata_qc_schedule_eh(), blk_abort_request() will be called, which leads to req->q->mq_ops->timeout() / scsi_timeout() being called.

scsi_timeout(), if the LLDD has no abort handler (libata has no abort handler), will set host byte to DID_TIME_OUT, and then call scsi_eh_scmd_add() to add the command to EH.

[...]

Applied to libata/linux.git (for-6.12-fixes), thanks!

[1/1] ata: libata: Set DID_TIME_OUT for commands that actually timed out

      https://git.kernel.org/libata/linux/c/8e59a2a5

Kind regards,
Niklas