diff mbox series

[4/4] drm/msm/a5xx: workaround early ring-buffer emptiness check

Message ID 20240711100038.268803-5-vladimir.lypak@gmail.com (mailing list archive)
State New, archived
Headers show
Series fixes for Adreno A5Xx preemption | expand

Commit Message

Vladimir Lypak July 11, 2024, 10 a.m. UTC
There is another cause for soft lock-up of GPU in empty ring-buffer:
race between GPU executing last commands and CPU checking ring for
emptiness. On GPU side IRQ for retire is triggered by CACHE_FLUSH_TS
event and RPTR shadow (which is used to check ring emptiness) is updated
a bit later from CP_CONTEXT_SWITCH_YIELD. Thus if GPU is executing its
last commands slow enough or we check that ring too fast we will miss a
chance to trigger switch to lower priority ring because current ring isn't
empty just yet. This can escalate to lock-up situation described in
previous patch.
To work-around this issue we keep track of last submit sequence number
for each ring and compare it with one written to memptrs from GPU during
execution of CACHE_FLUSH_TS event.

Fixes: b1fc2839d2f9 ("drm/msm: Implement preemption for A5XX targets")
Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
---
 drivers/gpu/drm/msm/adreno/a5xx_gpu.c     | 4 ++++
 drivers/gpu/drm/msm/adreno/a5xx_gpu.h     | 1 +
 drivers/gpu/drm/msm/adreno/a5xx_preempt.c | 4 ++++
 3 files changed, 9 insertions(+)

Comments

Akhil P Oommen Aug. 5, 2024, 7:58 p.m. UTC | #1
On Thu, Jul 11, 2024 at 10:00:21AM +0000, Vladimir Lypak wrote:
> There is another cause for soft lock-up of GPU in empty ring-buffer:
> race between GPU executing last commands and CPU checking ring for
> emptiness. On GPU side IRQ for retire is triggered by CACHE_FLUSH_TS
> event and RPTR shadow (which is used to check ring emptiness) is updated
> a bit later from CP_CONTEXT_SWITCH_YIELD. Thus if GPU is executing its
> last commands slow enough or we check that ring too fast we will miss a
> chance to trigger switch to lower priority ring because current ring isn't
> empty just yet. This can escalate to lock-up situation described in
> previous patch.
> To work-around this issue we keep track of last submit sequence number
> for each ring and compare it with one written to memptrs from GPU during
> execution of CACHE_FLUSH_TS event.
> 

This is interesting! Is this just theoretical or are you able to hit
this race on your device (after picking other fixes in this series)?

-Akhil.

> Fixes: b1fc2839d2f9 ("drm/msm: Implement preemption for A5XX targets")
> Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
> ---
>  drivers/gpu/drm/msm/adreno/a5xx_gpu.c     | 4 ++++
>  drivers/gpu/drm/msm/adreno/a5xx_gpu.h     | 1 +
>  drivers/gpu/drm/msm/adreno/a5xx_preempt.c | 4 ++++
>  3 files changed, 9 insertions(+)
> 
> diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
> index 266744ee1d5f..001f11f5febc 100644
> --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
> +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
> @@ -65,6 +65,8 @@ void a5xx_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring,
>  
>  static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit)
>  {
> +	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
> +	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
>  	struct msm_ringbuffer *ring = submit->ring;
>  	struct drm_gem_object *obj;
>  	uint32_t *ptr, dwords;
> @@ -109,6 +111,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
>  		}
>  	}
>  
> +	a5xx_gpu->last_seqno[ring->id] = submit->seqno;
>  	a5xx_flush(gpu, ring, true);
>  	a5xx_preempt_trigger(gpu, true);
>  
> @@ -210,6 +213,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
>  	/* Write the fence to the scratch register */
>  	OUT_PKT4(ring, REG_A5XX_CP_SCRATCH_REG(2), 1);
>  	OUT_RING(ring, submit->seqno);
> +	a5xx_gpu->last_seqno[ring->id] = submit->seqno;
>  
>  	/*
>  	 * Execute a CACHE_FLUSH_TS event. This will ensure that the
> diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
> index 1120824853d4..7269eaab9a7a 100644
> --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
> +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
> @@ -34,6 +34,7 @@ struct a5xx_gpu {
>  	struct drm_gem_object *preempt_counters_bo[MSM_GPU_MAX_RINGS];
>  	struct a5xx_preempt_record *preempt[MSM_GPU_MAX_RINGS];
>  	uint64_t preempt_iova[MSM_GPU_MAX_RINGS];
> +	uint32_t last_seqno[MSM_GPU_MAX_RINGS];
>  
>  	atomic_t preempt_state;
>  	struct timer_list preempt_timer;
> diff --git a/drivers/gpu/drm/msm/adreno/a5xx_preempt.c b/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
> index f8d09a83c5ae..6bd92f9b2338 100644
> --- a/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
> +++ b/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
> @@ -55,6 +55,8 @@ static inline void update_wptr(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
>  /* Return the highest priority ringbuffer with something in it */
>  static struct msm_ringbuffer *get_next_ring(struct msm_gpu *gpu)
>  {
> +	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
> +	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
>  	unsigned long flags;
>  	int i;
>  
> @@ -64,6 +66,8 @@ static struct msm_ringbuffer *get_next_ring(struct msm_gpu *gpu)
>  
>  		spin_lock_irqsave(&ring->preempt_lock, flags);
>  		empty = (get_wptr(ring) == gpu->funcs->get_rptr(gpu, ring));
> +		if (!empty && ring == a5xx_gpu->cur_ring)
> +			empty = ring->memptrs->fence == a5xx_gpu->last_seqno[i];
>  		spin_unlock_irqrestore(&ring->preempt_lock, flags);
>  
>  		if (!empty)
> -- 
> 2.45.2
>
diff mbox series

Patch

diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
index 266744ee1d5f..001f11f5febc 100644
--- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c
@@ -65,6 +65,8 @@  void a5xx_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring,
 
 static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
 	struct msm_ringbuffer *ring = submit->ring;
 	struct drm_gem_object *obj;
 	uint32_t *ptr, dwords;
@@ -109,6 +111,7 @@  static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 		}
 	}
 
+	a5xx_gpu->last_seqno[ring->id] = submit->seqno;
 	a5xx_flush(gpu, ring, true);
 	a5xx_preempt_trigger(gpu, true);
 
@@ -210,6 +213,7 @@  static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 	/* Write the fence to the scratch register */
 	OUT_PKT4(ring, REG_A5XX_CP_SCRATCH_REG(2), 1);
 	OUT_RING(ring, submit->seqno);
+	a5xx_gpu->last_seqno[ring->id] = submit->seqno;
 
 	/*
 	 * Execute a CACHE_FLUSH_TS event. This will ensure that the
diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
index 1120824853d4..7269eaab9a7a 100644
--- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h
@@ -34,6 +34,7 @@  struct a5xx_gpu {
 	struct drm_gem_object *preempt_counters_bo[MSM_GPU_MAX_RINGS];
 	struct a5xx_preempt_record *preempt[MSM_GPU_MAX_RINGS];
 	uint64_t preempt_iova[MSM_GPU_MAX_RINGS];
+	uint32_t last_seqno[MSM_GPU_MAX_RINGS];
 
 	atomic_t preempt_state;
 	struct timer_list preempt_timer;
diff --git a/drivers/gpu/drm/msm/adreno/a5xx_preempt.c b/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
index f8d09a83c5ae..6bd92f9b2338 100644
--- a/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
+++ b/drivers/gpu/drm/msm/adreno/a5xx_preempt.c
@@ -55,6 +55,8 @@  static inline void update_wptr(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
 /* Return the highest priority ringbuffer with something in it */
 static struct msm_ringbuffer *get_next_ring(struct msm_gpu *gpu)
 {
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
 	unsigned long flags;
 	int i;
 
@@ -64,6 +66,8 @@  static struct msm_ringbuffer *get_next_ring(struct msm_gpu *gpu)
 
 		spin_lock_irqsave(&ring->preempt_lock, flags);
 		empty = (get_wptr(ring) == gpu->funcs->get_rptr(gpu, ring));
+		if (!empty && ring == a5xx_gpu->cur_ring)
+			empty = ring->memptrs->fence == a5xx_gpu->last_seqno[i];
 		spin_unlock_irqrestore(&ring->preempt_lock, flags);
 
 		if (!empty)