Hi Sebastian,
On Wed, Dec 18, 2024 at 07:55:30PM +0000, Sebastian Brzezinka wrote:
> Adding TAINT_WARN is expected when GPU cannot be restarted, informing
> about it will make it easier to find the source of the failure.
>
> Right now TAINT_WARN will cause CI machine to restart, leaving only err
> code EINTR.
you are describing here what you haven't done in your patch. If I
just read this I assume that you are adding the TAINT_WARN, which
is not the case.
Please describe in your commit log:
- The problem you are trying to solve.
- What's the the solution.
- How you implemented it.
The title is not very descriptive either.
> Signed-off-by: Sebastian Brzezinka <sebastian.brzezinka@xxxxxxxxx>
> ---
> drivers/gpu/drm/i915/gt/intel_reset.c | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c
> index c2fe3fc78e76..458c26706df4 100644
> --- a/drivers/gpu/drm/i915/gt/intel_reset.c
> +++ b/drivers/gpu/drm/i915/gt/intel_reset.c
> @@ -1113,6 +1113,7 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
> * Warn CI about the unrecoverable wedged condition.
> * Time for a reboot.
> */
> + gt_err(gt, "Unrecoverable wedged condition\n");
> add_taint_for_CI(gt->i915, TAINT_WARN);
> return false;
> }
> @@ -1264,9 +1265,10 @@ void intel_gt_reset(struct intel_gt *gt,
> }
>
> ret = resume(gt);
> - if (ret)
> + if (ret) {
> + gt_err(gt, "Failed to resume (%d)\n", ret);
> goto taint;
> -
> + }
> finish:
It looks much better if you leave one space before the error
labels, same as before, but now you are adding the brackets.
Andi
> reset_finish(gt, awake);
> unlock:
> @@ -1608,6 +1610,7 @@ void intel_gt_set_wedged_on_init(struct intel_gt *gt)
> set_bit(I915_WEDGED_ON_INIT, >->reset.flags);
>
> /* Wedged on init is non-recoverable */
> + gt_err(gt, "Non-recoverable wedged on init\n");
> add_taint_for_CI(gt->i915, TAINT_WARN);
> }
>
> --
> 2.34.1
