drm/etnaviv: expedited MMU fault handling

The GPU is halted when it hits a MMU exception, so there is no point in
waiting for the job timeout to expire or try to work out if the GPU is
still making progress in the timeout handler, as we know that the GPU
won't make any more progress.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>

Author: lynxeye-dev

drivers/gpu/drm/etnaviv/etnaviv_gpu.c

@@ -1539,6 +1539,8 @@ static irqreturn_t irq_handler(int irq, void *data)
 
 		if (intr & VIVS_HI_INTR_ACKNOWLEDGE_MMU_EXCEPTION) {
 			dump_mmu_fault(gpu);
+			gpu->state = ETNA_GPU_STATE_FAULT;
+			drm_sched_fault(&gpu->sched);
 			intr &= ~VIVS_HI_INTR_ACKNOWLEDGE_MMU_EXCEPTION;
 		}
 

drivers/gpu/drm/etnaviv/etnaviv_gpu.h

@@ -101,6 +101,7 @@ enum etnaviv_gpu_state {
 	ETNA_GPU_STATE_RESET,
 	ETNA_GPU_STATE_INITIALIZED,
 	ETNA_GPU_STATE_RUNNING,
+	ETNA_GPU_STATE_FAULT,
 };
 
 struct etnaviv_gpu {

drivers/gpu/drm/etnaviv/etnaviv_sched.c

@@ -55,8 +55,9 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
 	 */
 	dma_addr = gpu_read(gpu, VIVS_FE_DMA_ADDRESS);
 	change = dma_addr - gpu->hangcheck_dma_addr;
-	if (gpu->completed_fence != gpu->hangcheck_fence ||
-	    change < 0 || change > 16) {
+	if (gpu->state == ETNA_GPU_STATE_RUNNING &&
+	    (gpu->completed_fence != gpu->hangcheck_fence ||
+	     change < 0 || change > 16)) {
 		gpu->hangcheck_dma_addr = dma_addr;
 		gpu->hangcheck_fence = gpu->completed_fence;
 		goto out_no_timeout;