@@ -328,6 +328,8 @@ struct drm_i915_error_state {
u32 *active_bo_count, *pinned_bo_count;
struct intel_overlay_error_state *overlay;
struct intel_display_error_state *display;
+ int hangcheck_score[I915_NUM_RINGS];
+ enum intel_ring_hangcheck_action hangcheck_action[I915_NUM_RINGS];
};
struct intel_crtc_config;
@@ -213,6 +213,24 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
}
}
+static const char *hangcheck_action_to_str(enum intel_ring_hangcheck_action a)
+{
+ switch (a) {
+ case HANGCHECK_IDLE:
+ return "idle";
+ case HANGCHECK_WAIT:
+ return "wait";
+ case HANGCHECK_ACTIVE:
+ return "active";
+ case HANGCHECK_KICK:
+ return "kick";
+ case HANGCHECK_HUNG:
+ return "hung";
+ }
+
+ return "unknown";
+}
+
static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
struct drm_device *dev,
struct drm_i915_error_state *error,
@@ -253,6 +271,9 @@ static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
err_printf(m, " waiting: %s\n", yesno(error->waiting[ring]));
err_printf(m, " ring->head: 0x%08x\n", error->cpu_ring_head[ring]);
err_printf(m, " ring->tail: 0x%08x\n", error->cpu_ring_tail[ring]);
+ err_printf(m, " hangcheck: %s [%d]\n",
+ hangcheck_action_to_str(error->hangcheck_action[ring]),
+ error->hangcheck_score[ring]);
}
void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...)
@@ -718,6 +739,9 @@ static void i915_record_ring_state(struct drm_device *dev,
error->cpu_ring_head[ring->id] = ring->head;
error->cpu_ring_tail[ring->id] = ring->tail;
+
+ error->hangcheck_score[ring->id] = ring->hangcheck.score;
+ error->hangcheck_action[ring->id] = ring->hangcheck.action;
}
@@ -1975,6 +1975,8 @@ static void i915_hangcheck_elapsed(unsigned long data)
if (ring->hangcheck.seqno == seqno) {
if (ring_idle(ring, seqno)) {
+ ring->hangcheck.action = HANGCHECK_IDLE;
+
if (waitqueue_active(&ring->irq_queue)) {
/* Issue a wake-up to catch stuck h/w. */
DRM_ERROR("Hangcheck timer elapsed... %s idle\n",
@@ -2003,6 +2005,7 @@ static void i915_hangcheck_elapsed(unsigned long data)
acthd);
switch (ring->hangcheck.action) {
+ case HANGCHECK_IDLE:
case HANGCHECK_WAIT:
break;
case HANGCHECK_ACTIVE:
@@ -2018,6 +2021,8 @@ static void i915_hangcheck_elapsed(unsigned long data)
}
}
} else {
+ ring->hangcheck.action = HANGCHECK_ACTIVE;
+
/* Gradually reduce the count so that we catch DoS
* attempts across multiple batches.
*/
@@ -34,6 +34,7 @@ struct intel_hw_status_page {
#define I915_WRITE_IMR(ring, val) I915_WRITE(RING_IMR((ring)->mmio_base), val)
enum intel_ring_hangcheck_action {
+ HANGCHECK_IDLE = 0,
HANGCHECK_WAIT,
HANGCHECK_ACTIVE,
HANGCHECK_KICK,
Score and action reveals what all the rings were doing and why hang was declared. Add idle state so that we can distinguish between waiting and idle ring. v2: - add idle as a hangcheck action - consensed hangcheck status to single line (Chris) - mark active explicitly when we are making progress (Chris) Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com> --- drivers/gpu/drm/i915/i915_drv.h | 2 ++ drivers/gpu/drm/i915/i915_gpu_error.c | 24 ++++++++++++++++++++++++ drivers/gpu/drm/i915/i915_irq.c | 5 +++++ drivers/gpu/drm/i915/intel_ringbuffer.h | 1 + 4 files changed, 32 insertions(+)