rproc_report_crash() calls pm_stay_awake() unconditionally before
queue_work(). When multiple crash sources fire simultaneously (e.g.
watchdog and fatal interrupts), the second call enqueues nothing because
the work is already pending, but still increments the wakeup source
active count. The single pm_relax() that follows when the work runs
leaves the wakeup source with a non-zero count, preventing suspend.
Move pm_stay_awake() inside the queue_work() success path so the
wakeup source is only held when a work item is actually enqueued.
Fixes: a781e5aa5911 ("remoteproc: core: Prevent system suspend during
remoteproc recovery")
Cc: [email protected]
Signed-off-by: Mukesh Ojha <[email protected]>
---
drivers/remoteproc/remoteproc_core.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/remoteproc/remoteproc_core.c
b/drivers/remoteproc/remoteproc_core.c
index ee9354ff69f0..9d36e2f06348 100644
--- a/drivers/remoteproc/remoteproc_core.c
+++ b/drivers/remoteproc/remoteproc_core.c
@@ -2754,8 +2754,8 @@ void rproc_report_crash(struct rproc *rproc, enum
rproc_crash_type type)
}
/* Prevent suspend while the remoteproc is being recovered */
- pm_stay_awake(rproc->dev.parent);
- queue_work(rproc_recovery_wq, &rproc->crash_handler);
+ if (queue_work(rproc_recovery_wq, &rproc->crash_handler))
+ pm_stay_awake(rproc->dev.parent);
spin_unlock_irqrestore(&rproc->crash_handler_lock, flags);
dev_err(&rproc->dev, "crash detected in %s: type %s\n",
--
2.53.0
