drm/i915: detect hang using per ring hangcheck_score
Keep track of ring seqno progress and if there are no
progress detected, declare hang. Use actual head (acthd)
to distinguish between ring stuck and batchbuffer looping
situation. Stuck ring will be kicked to trigger progress.
This commit adds a hard limit for batchbuffer completion time.
If batchbuffer completion time is more than 4.5 seconds,
the gpu will be declared hung.
Review comment from Ben which nicely clarifies the semantic change:
"Maybe I'm just stating the functional changes of the patch, but in case
they were unintended here is what I see as potential issues:
1. "If ring B is waiting on ring A via semaphore, and ring A is making
progress, albeit slowly - the hangcheck will fire. The check will
determine that A is moving, however ring B will appear hung because
the ACTHD doesn't move. I honestly can't say if that's actually a
realistic problem to hit it probably implies the timeout value is too
low.
2. "There's also another corner case on the kick. If the seqno = 2
(though not stuck), and on the 3rd hangcheck, the ring is stuck, and
we try to kick it... we don't actually try to find out if the kick
helped"
v2: use atchd to detect stuck ring from loop (Ben Widawsky)
v3: Use acthd to check when ring needs kicking.
Declare hang on third time in order to give time for
kick_ring to take effect.
v4: Update commit msg
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: Paste in Ben's review comment.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 5ae5ca8..e88f173 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -683,7 +683,6 @@
wake_up_all(&ring->irq_queue);
if (i915_enable_hangcheck) {
- dev_priv->gpu_error.hangcheck_count = 0;
mod_timer(&dev_priv->gpu_error.hangcheck_timer,
round_jiffies_up(jiffies + DRM_I915_HANGCHECK_JIFFIES));
}
@@ -2422,61 +2421,76 @@
/**
* This is called when the chip hasn't reported back with completed
- * batchbuffers in a long time. The first time this is called we simply record
- * ACTHD. If ACTHD hasn't changed by the time the hangcheck timer elapses
- * again, we assume the chip is wedged and try to fix it.
+ * batchbuffers in a long time. We keep track per ring seqno progress and
+ * if there are no progress, hangcheck score for that ring is increased.
+ * Further, acthd is inspected to see if the ring is stuck. On stuck case
+ * we kick the ring. If we see no progress on three subsequent calls
+ * we assume chip is wedged and try to fix it by resetting the chip.
*/
void i915_hangcheck_elapsed(unsigned long data)
{
struct drm_device *dev = (struct drm_device *)data;
drm_i915_private_t *dev_priv = dev->dev_private;
struct intel_ring_buffer *ring;
- bool err = false, idle;
int i;
- u32 seqno[I915_NUM_RINGS];
- bool work_done;
+ int busy_count = 0, rings_hung = 0;
+ bool stuck[I915_NUM_RINGS];
if (!i915_enable_hangcheck)
return;
- idle = true;
for_each_ring(ring, dev_priv, i) {
- seqno[i] = ring->get_seqno(ring, false);
- idle &= i915_hangcheck_ring_idle(ring, seqno[i], &err);
- }
+ u32 seqno, acthd;
+ bool idle, err = false;
- /* If all work is done then ACTHD clearly hasn't advanced. */
- if (idle) {
- if (err) {
- if (i915_hangcheck_hung(dev))
- return;
+ seqno = ring->get_seqno(ring, false);
+ acthd = intel_ring_get_active_head(ring);
+ idle = i915_hangcheck_ring_idle(ring, seqno, &err);
+ stuck[i] = ring->hangcheck.acthd == acthd;
- goto repeat;
+ if (idle) {
+ if (err)
+ ring->hangcheck.score += 2;
+ else
+ ring->hangcheck.score = 0;
+ } else {
+ busy_count++;
+
+ if (ring->hangcheck.seqno == seqno) {
+ ring->hangcheck.score++;
+
+ /* Kick ring if stuck*/
+ if (stuck[i])
+ i915_hangcheck_ring_hung(ring);
+ } else {
+ ring->hangcheck.score = 0;
+ }
}
- dev_priv->gpu_error.hangcheck_count = 0;
- return;
+ ring->hangcheck.seqno = seqno;
+ ring->hangcheck.acthd = acthd;
}
- work_done = false;
for_each_ring(ring, dev_priv, i) {
- if (ring->hangcheck.seqno != seqno[i]) {
- work_done = true;
- ring->hangcheck.seqno = seqno[i];
+ if (ring->hangcheck.score > 2) {
+ rings_hung++;
+ DRM_ERROR("%s: %s on %s 0x%x\n", ring->name,
+ stuck[i] ? "stuck" : "no progress",
+ stuck[i] ? "addr" : "seqno",
+ stuck[i] ? ring->hangcheck.acthd & HEAD_ADDR :
+ ring->hangcheck.seqno);
}
}
- if (!work_done) {
- if (i915_hangcheck_hung(dev))
- return;
- } else {
- dev_priv->gpu_error.hangcheck_count = 0;
- }
+ if (rings_hung)
+ return i915_handle_error(dev, true);
-repeat:
- /* Reset timer case chip hangs without another request being added */
- mod_timer(&dev_priv->gpu_error.hangcheck_timer,
- round_jiffies_up(jiffies + DRM_I915_HANGCHECK_JIFFIES));
+ if (busy_count)
+ /* Reset timer case chip hangs without another request
+ * being added */
+ mod_timer(&dev_priv->gpu_error.hangcheck_timer,
+ round_jiffies_up(jiffies +
+ DRM_I915_HANGCHECK_JIFFIES));
}
/* drm_dma.h hooks