[Intel-gfx] [PATCH 042/190] drm/i915: Clean up GPU hang message
Arun Siluvery
arun.siluvery at linux.intel.com
Thu Feb 25 17:40:00 UTC 2016
On 11/01/2016 09:16, Chris Wilson wrote:
> Remove some redundant kernel messages as we deduce a hung GPU and
> capture the error state.
>
> v2: Fix "hang" vs "no progress" message whilst I was there
>
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/i915_irq.c | 21 +++++++--------------
> 1 file changed, 7 insertions(+), 14 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
> index d9757d227c86..ce52d7d9ad91 100644
> --- a/drivers/gpu/drm/i915/i915_irq.c
> +++ b/drivers/gpu/drm/i915/i915_irq.c
> @@ -3031,8 +3031,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
> struct drm_device *dev = dev_priv->dev;
> struct intel_engine_cs *ring;
> int i;
> - int busy_count = 0, rings_hung = 0;
> - bool stuck[I915_NUM_RINGS] = { 0 };
> + int busy_count = 0;
> #define BUSY 1
> #define KICK 5
> #define HUNG 20
> @@ -3108,7 +3107,6 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
> break;
> case HANGCHECK_HUNG:
> ring->hangcheck.score += HUNG;
> - stuck[i] = true;
> break;
> }
> }
> @@ -3134,17 +3132,12 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
> busy_count += busy;
> }
>
> - for_each_ring(ring, dev_priv, i) {
> - if (ring->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG) {
> - DRM_INFO("%s on %s\n",
> - stuck[i] ? "stuck" : "no progress",
> - ring->name);
> - rings_hung++;
this is required when engine resets are supported. I am converting this
to an engine_mask and send it directly to i915_handle_error().
regards
Arun
> - }
> - }
> -
> - if (rings_hung)
> - return i915_handle_error(dev, true, "Ring hung");
> + for_each_ring(ring, dev_priv, i)
> + if (ring->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG)
> + return i915_handle_error(dev, true,
> + "%s on %s",
> + ring->hangcheck.action == HANGCHECK_HUNG ? "Hang" : "No progress" ,
> + ring->name);
>
> /* Reset timer in case GPU hangs without another request being added */
> if (busy_count)
>
More information about the Intel-gfx
mailing list