| Message ID | 20240203024521.2006455-1-ming.lei@redhat.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
| Series | scsi: core: move scsi_host_busy() out of host lock if it is for per-command | expand |
On 2/2/24 18:45, Ming Lei wrote: > Commit 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock for > waking up EH handler") is for fixing hard lockup issue triggered by EH. > > The core idea is to move scsi_host_busy() out of host lock if it is > called for per-command. However, the patch is merged as wrong and ^^^^^^^^^^^^^^^ I'm not sure what this means? > doesn't move scsi_host_busy() out of host lock, so fix it. Anyway: Reviewed-by: Bart Van Assche <bvanassche@acm.org>
On Sat, 03 Feb 2024 10:45:21 +0800, Ming Lei wrote: > Commit 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock for > waking up EH handler") is for fixing hard lockup issue triggered by EH. > > The core idea is to move scsi_host_busy() out of host lock if it is > called for per-command. However, the patch is merged as wrong and > doesn't move scsi_host_busy() out of host lock, so fix it. > > [...] Applied to 6.8/scsi-fixes, thanks! [1/1] scsi: core: move scsi_host_busy() out of host lock if it is for per-command https://git.kernel.org/mkp/scsi/c/4e6c90119907
diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index 4f455884fdc4..612489afe8d2 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -282,11 +282,12 @@ static void scsi_eh_inc_host_failed(struct rcu_head *head) { struct scsi_cmnd *scmd = container_of(head, typeof(*scmd), rcu); struct Scsi_Host *shost = scmd->device->host; + unsigned int busy = scsi_host_busy(shost); unsigned long flags; spin_lock_irqsave(shost->host_lock, flags); shost->host_failed++; - scsi_eh_wakeup(shost, scsi_host_busy(shost)); + scsi_eh_wakeup(shost, busy); spin_unlock_irqrestore(shost->host_lock, flags); } diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 1fb80eae9a63..df5ac03d5d6c 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -278,9 +278,11 @@ static void scsi_dec_host_busy(struct Scsi_Host *shost, struct scsi_cmnd *cmd) rcu_read_lock(); __clear_bit(SCMD_STATE_INFLIGHT, &cmd->state); if (unlikely(scsi_host_in_recovery(shost))) { + unsigned int busy = scsi_host_busy(shost); + spin_lock_irqsave(shost->host_lock, flags); if (shost->host_failed || shost->host_eh_scheduled) - scsi_eh_wakeup(shost, scsi_host_busy(shost)); + scsi_eh_wakeup(shost, busy); spin_unlock_irqrestore(shost->host_lock, flags); } rcu_read_unlock();
Commit 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock for waking up EH handler") is for fixing hard lockup issue triggered by EH. The core idea is to move scsi_host_busy() out of host lock if it is called for per-command. However, the patch is merged as wrong and doesn't move scsi_host_busy() out of host lock, so fix it. Fixes: 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock for waking up EH handler") Cc: Sathya Prakash Veerichetty <safhya.prakash@broadcom.com> Cc: Bart Van Assche <bvanassche@acm.org> Cc: Ewan D. Milne <emilne@redhat.com> Signed-off-by: Ming Lei <ming.lei@redhat.com> --- drivers/scsi/scsi_error.c | 3 ++- drivers/scsi/scsi_lib.c | 4 +++- 2 files changed, 5 insertions(+), 2 deletions(-)