All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling
@ 2026-04-02 12:55 Karol Wachowski
  2026-04-02 17:31 ` Lizhi Hou
  2026-04-02 19:27 ` Jeff Hugo
  0 siblings, 2 replies; 4+ messages in thread
From: Karol Wachowski @ 2026-04-02 12:55 UTC (permalink / raw)
  To: dri-devel
  Cc: oded.gabbay, jeff.hugo, maciej.falkowski, lizhi.hou,
	andrzej.kacprowski, Karol Wachowski

With OS scheduling mode the driver cannot determine which context
caused the timeout, so context abort cannot be used. Instead of
queuing context_abort_work, directly trigger full device recovery
when a job timeout (TDR) occurs in OS scheduling mode.

Fixes: ade00a6c903f ("accel/ivpu: Perform engine reset instead of device recovery on TDR")
Signed-off-by: Karol Wachowski <karol.wachowski@linux.intel.com>
---
 drivers/accel/ivpu/ivpu_pm.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c
index 83da9b297f37..c1ce8329790e 100644
--- a/drivers/accel/ivpu/ivpu_pm.c
+++ b/drivers/accel/ivpu/ivpu_pm.c
@@ -221,6 +221,12 @@ static void ivpu_job_timeout_work(struct work_struct *work)
 
 abort:
 	atomic_set(&vdev->job_timeout_counter, 0);
+
+	if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_OS) {
+		ivpu_pm_trigger_recovery(vdev, "Job timeout");
+		return;
+	}
+
 	ivpu_jsm_state_dump(vdev);
 	ivpu_dev_coredump(vdev);
 	queue_work(system_percpu_wq, &vdev->context_abort_work);
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling
  2026-04-02 12:55 [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling Karol Wachowski
@ 2026-04-02 17:31 ` Lizhi Hou
  2026-04-02 19:27 ` Jeff Hugo
  1 sibling, 0 replies; 4+ messages in thread
From: Lizhi Hou @ 2026-04-02 17:31 UTC (permalink / raw)
  To: Karol Wachowski, dri-devel
  Cc: oded.gabbay, jeff.hugo, maciej.falkowski, andrzej.kacprowski

Reviewed-by: Lizhi Hou <lizhi.hou@amd.com>

On 4/2/26 05:55, Karol Wachowski wrote:
> With OS scheduling mode the driver cannot determine which context
> caused the timeout, so context abort cannot be used. Instead of
> queuing context_abort_work, directly trigger full device recovery
> when a job timeout (TDR) occurs in OS scheduling mode.
>
> Fixes: ade00a6c903f ("accel/ivpu: Perform engine reset instead of device recovery on TDR")
> Signed-off-by: Karol Wachowski <karol.wachowski@linux.intel.com>
> ---
>   drivers/accel/ivpu/ivpu_pm.c | 6 ++++++
>   1 file changed, 6 insertions(+)
>
> diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c
> index 83da9b297f37..c1ce8329790e 100644
> --- a/drivers/accel/ivpu/ivpu_pm.c
> +++ b/drivers/accel/ivpu/ivpu_pm.c
> @@ -221,6 +221,12 @@ static void ivpu_job_timeout_work(struct work_struct *work)
>   
>   abort:
>   	atomic_set(&vdev->job_timeout_counter, 0);
> +
> +	if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_OS) {
> +		ivpu_pm_trigger_recovery(vdev, "Job timeout");
> +		return;
> +	}
> +
>   	ivpu_jsm_state_dump(vdev);
>   	ivpu_dev_coredump(vdev);
>   	queue_work(system_percpu_wq, &vdev->context_abort_work);

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling
  2026-04-02 12:55 [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling Karol Wachowski
  2026-04-02 17:31 ` Lizhi Hou
@ 2026-04-02 19:27 ` Jeff Hugo
  2026-04-02 19:32   ` Karol Wachowski
  1 sibling, 1 reply; 4+ messages in thread
From: Jeff Hugo @ 2026-04-02 19:27 UTC (permalink / raw)
  To: Karol Wachowski, dri-devel
  Cc: oded.gabbay, maciej.falkowski, lizhi.hou, andrzej.kacprowski

On 4/2/2026 6:55 AM, Karol Wachowski wrote:
> With OS scheduling mode the driver cannot determine which context
> caused the timeout, so context abort cannot be used. Instead of
> queuing context_abort_work, directly trigger full device recovery
> when a job timeout (TDR) occurs in OS scheduling mode.
> 
> Fixes: ade00a6c903f ("accel/ivpu: Perform engine reset instead of device recovery on TDR")
> Signed-off-by: Karol Wachowski <karol.wachowski@linux.intel.com>

Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling
  2026-04-02 19:27 ` Jeff Hugo
@ 2026-04-02 19:32   ` Karol Wachowski
  0 siblings, 0 replies; 4+ messages in thread
From: Karol Wachowski @ 2026-04-02 19:32 UTC (permalink / raw)
  To: Jeff Hugo, dri-devel
  Cc: oded.gabbay, maciej.falkowski, lizhi.hou, andrzej.kacprowski

On 4/2/2026 9:27 PM, Jeff Hugo wrote:
> On 4/2/2026 6:55 AM, Karol Wachowski wrote:
>> With OS scheduling mode the driver cannot determine which context
>> caused the timeout, so context abort cannot be used. Instead of
>> queuing context_abort_work, directly trigger full device recovery
>> when a job timeout (TDR) occurs in OS scheduling mode.
>>
>> Fixes: ade00a6c903f ("accel/ivpu: Perform engine reset instead of
>> device recovery on TDR")
>> Signed-off-by: Karol Wachowski <karol.wachowski@linux.intel.com>
> 
> Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>

Thanks, applied to drm-misc-next-fixes.

Karol

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2026-04-02 19:32 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-04-02 12:55 [PATCH] accel/ivpu: Trigger recovery on TDR with OS scheduling Karol Wachowski
2026-04-02 17:31 ` Lizhi Hou
2026-04-02 19:27 ` Jeff Hugo
2026-04-02 19:32   ` Karol Wachowski

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.