lpfc: Fix regressions in 14.2.0.0 refactoring
Subsystems (3)
cover lpfc: Fix regressions in 14.2.0.0 refactoring
Two regressions were found in further testing of the refactoring work done for the 14.2.0.0 driver rev.
These 2 patches correct the regressions
James Smart (2):
lpfc: Fix broken sli4 abort path
lpfc: Fix locking for lpfc_sli_iocbq_lookup
drivers/scsi/lpfc/lpfc_scsi.c | 12 +++++++----- drivers/scsi/lpfc/lpfc_sli.c | 10 ++++++++-- 2 files changed, 15 insertions(+), 7 deletions(-)
Signature · 1 line
2.26.2
PATCH 1/2 lpfc: Fix broken sli4 abort path
There was a merge error in ther 14.2.0.0 patches that resulted in the sli4 path using the sli3 issue_abort_iotag routine. This resulted in txcmplq corruption.
Fix to use the sli4 routine when sli4.
--- drivers/scsi/lpfc/lpfc_scsi.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-)
drivers/scsi/lpfc/lpfc_scsi.c +7 -5
@@ -5929,13 +5929,15 @@ lpfc_abort_handler(struct scsi_cmnd *cmnd) } lpfc_cmd->waitq = &waitq; -if (phba->sli_rev ==LPFC_SLI_REV4 ) +if (phba->sli_rev ==LPFC_SLI_REV4 ) {spin_unlock (&pring_s4->ring_lock); -else + ret_val =lpfc_sli4_issue_abort_iotag (phba, iocb, + lpfc_sli_abort_fcp_cmpl); + }else { pring = &phba->sli.sli3_ring[LPFC_FCP_RING ]; - - ret_val =lpfc_sli_issue_abort_iotag (phba, pring, iocb, - lpfc_sli_abort_fcp_cmpl); + ret_val =lpfc_sli_issue_abort_iotag (phba, pring, iocb, + lpfc_sli_abort_fcp_cmpl); + }/* Make sure HBA is alive */ lpfc_issue_hb_tmo (phba);
Signature · 1 line
2.26.2
PATCH 2/2 lpfc: Fix locking for lpfc_sli_iocbq_lookup
In reply to James SmartThe rules changed for lpfc_sli_iocbq_lookup vs locking. Prior, the routine properly took out the lock. In newly refactored code, the locks must be held when calling the routine.
Fix lpfc_sli_process_sol_iocb() to take the locks before calling the routine.
Fix lpfc_sli_handle_fast_ring_event() to not release the locks to call the routine.
--- drivers/scsi/lpfc/lpfc_sli.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/scsi/lpfc/lpfc_sli.c +8 -2
@@ -3715,7 +3715,15 @@ lpfc_sli_process_sol_iocb(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,unsigned long iflag;u32 ulp_command, ulp_status, ulp_word4, ulp_context, iotag; +if (phba->sli_rev ==LPFC_SLI_REV4 ) +spin_lock_irqsave (&pring->ring_lock, iflag); +else +spin_lock_irqsave (&phba->hbalock, iflag); cmdiocbp =lpfc_sli_iocbq_lookup (phba, pring, saveq); +if (phba->sli_rev ==LPFC_SLI_REV4 ) +spin_unlock_irqrestore (&pring->ring_lock, iflag); +else +spin_unlock_irqrestore (&phba->hbalock, iflag); ulp_command =get_job_cmnd (phba, saveq); ulp_status =get_job_ulpstatus (phba, saveq);
@@ -4052,10 +4060,8 @@ lpfc_sli_handle_fast_ring_event(struct lpfc_hba *phba,break ; } -spin_unlock_irqrestore (&phba->hbalock, iflag); cmdiocbq =lpfc_sli_iocbq_lookup (phba, pring, &rspiocbq); -spin_lock_irqsave (&phba->hbalock, iflag);if (unlikely (!cmdiocbq))break ;if (cmdiocbq->cmd_flag &LPFC_DRIVER_ABORTED )
Signature · 1 line
2.26.2
Martin K. Petersen
AppliedJames,
Quoted 2 lines
Two regressions were found in further testing of the refactoring work done for the 14.2.0.0 driver rev.
Applied to 5.18/scsi-staging, thanks!
Signature · 1 line
Martin K. Petersen Oracle Linux Engineering
Martin K. Petersen
AppliedQuoted 10 lines · James Smart
Two regressions were found in further testing of the refactoring work done for the 14.2.0.0 driver rev.
These 2 patches correct the regressions
James Smart (2):
lpfc: Fix broken sli4 abort path
lpfc: Fix locking forlpfc_sli_iocbq_lookup[...]
Applied to 5.18/scsi-fixes, thanks!
[1/2] lpfc: Fix broken sli4 abort path
https://git.kernel.org/mkp/scsi/c/7294a9bcaa7e [2/2] lpfc: Fix locking for lpfc_sli_iocbq_lookup https://git.kernel.org/mkp/scsi/c/c26bd6602e1d
Signature · 1 line
Martin K. Petersen Oracle Linux Engineering