drm/msm: Hangcheck progress detection

If the hangcheck timer expires, check if the fw's position in the
cmdstream has advanced (changed) since last timer expiration, and
allow it up to three additional "extensions" to it's alotted time.
The intention is to continue to catch "shader stuck in a loop" type
hangs quickly, but allow more time for things that are actually
making forward progress.

Because we need to sample the CP state twice to detect if there has
not been progress, this also cuts the the timer's duration in half.

v2: Fix typo (REG_A6XX_CP_CSQ_IB2_STAT), add comment
v3: Only halve hangcheck timer duration for generations which
    support progress detection (hdanton); removed unused a5xx
    progress (without knowing how to adjust for data buffered
    in ROQ it is too likely to report a false negative)
v4: Comment updates to better describe the total hangcheck
    duration when progress detection is applied

Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Tested-by: Chia-I Wu <olvaffe@gmail.com> # dEQP-GLES2.functional.flush_finish.wait
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/511584/
Link: https://lore.kernel.org/r/20221114193049.1533391-3-robdclark@gmail.com

1 files changed, 28 insertions, 0 deletions

diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.h b/drivers/gpu/drm/msm/msm_ringbuffer.h
index 2a5045abe46e..698b333abccd 100644
--- a/drivers/gpu/drm/msm/msm_ringbuffer.h
+++ b/drivers/gpu/drm/msm/msm_ringbuffer.h
@@ -35,6 +35,11 @@ struct msm_rbmemptrs {
 	volatile u64 ttbr0;
 };
 
+struct msm_cp_state {
+	uint64_t ib1_base, ib2_base;
+	uint32_t ib1_rem, ib2_rem;
+};
+
 struct msm_ringbuffer {
 	struct msm_gpu *gpu;
 	int id;
@@ -64,6 +69,29 @@ struct msm_ringbuffer {
 	uint64_t memptrs_iova;
 	struct msm_fence_context *fctx;
 
+	/**
+	 * hangcheck_progress_retries:
+	 *
+	 * The number of extra hangcheck duration cycles that we have given
+	 * due to it appearing that the GPU is making forward progress.
+	 *
+	 * For GPU generations which support progress detection (see.
+	 * msm_gpu_funcs::progress()), if the GPU appears to be making progress
+	 * (ie. the CP has advanced in the command stream, we'll allow up to
+	 * DRM_MSM_HANGCHECK_PROGRESS_RETRIES expirations of the hangcheck timer
+	 * before killing the job.  But to detect progress we need two sample
+	 * points, so the duration of the hangcheck timer is halved.  In other
+	 * words we'll let the submit run for up to:
+	 *
+	 * (DRM_MSM_HANGCHECK_DEFAULT_PERIOD / 2) * (DRM_MSM_HANGCHECK_PROGRESS_RETRIES + 1)
+	 */
+	int hangcheck_progress_retries;
+
+	/**
+	 * last_cp_state: The state of the CP at the last call to gpu->progress()
+	 */
+	struct msm_cp_state last_cp_state;
+
 	/*
 	 * preempt_lock protects preemption and serializes wptr updates against
 	 * preemption.  Can be aquired from irq context.