Commit 537b604c authored by Michal Hocko's avatar Michal Hocko Committed by James Bottomley

scsi: fix scsi_error_handler vs. scsi_host_dev_release race

b9d5c6b7 ("[SCSI] cleanup setting task state in
scsi_error_handler()") has introduced a race between scsi_error_handler
and scsi_host_dev_release resulting in the hang when the device goes
away because scsi_error_handler might miss a wake up:

CPU0					CPU1
scsi_error_handler			scsi_host_dev_release
  					  kthread_stop()
  kthread_should_stop()
    test_bit(KTHREAD_SHOULD_STOP)
					    set_bit(KTHREAD_SHOULD_STOP)
					    wake_up_process()
					    wait_for_completion()

  set_current_state(TASK_INTERRUPTIBLE)
  schedule()

The most straightforward solution seems to be to invert the ordering of
the set_current_state and kthread_should_stop.

The issue has been noticed during reboot test on a 3.0 based kernel but
the current code seems to be affected in the same way.

[jejb: additional comment added]
Cc: <stable@vger.kernel.org> # 3.6+
Reported-and-debugged-by: default avatarMike Mayer <Mike.Meyer@teradata.com>
Signed-off-by: default avatarMichal Hocko <mhocko@suse.com>
Reviewed-by: default avatarDan Williams <dan.j.williams@intel.com>
Reviewed-by: default avatarHannes Reinecke <hare@suse.de>
Signed-off-by: default avatarJames Bottomley <JBottomley@Odin.com>
parent 420fa211
...@@ -2179,8 +2179,17 @@ int scsi_error_handler(void *data) ...@@ -2179,8 +2179,17 @@ int scsi_error_handler(void *data)
* We never actually get interrupted because kthread_run * We never actually get interrupted because kthread_run
* disables signal delivery for the created thread. * disables signal delivery for the created thread.
*/ */
while (!kthread_should_stop()) { while (true) {
/*
* The sequence in kthread_stop() sets the stop flag first
* then wakes the process. To avoid missed wakeups, the task
* should always be in a non running state before the stop
* flag is checked
*/
set_current_state(TASK_INTERRUPTIBLE); set_current_state(TASK_INTERRUPTIBLE);
if (kthread_should_stop())
break;
if ((shost->host_failed == 0 && shost->host_eh_scheduled == 0) || if ((shost->host_failed == 0 && shost->host_eh_scheduled == 0) ||
shost->host_failed != atomic_read(&shost->host_busy)) { shost->host_failed != atomic_read(&shost->host_busy)) {
SCSI_LOG_ERROR_RECOVERY(1, SCSI_LOG_ERROR_RECOVERY(1,
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment