Sync Panthor with drm-tip (uAPI 1.5)

drivers/gpu/drm/amd/amdgpu/amdgpu_job.c

@@ -85,7 +85,7 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 exit:
 	adev->job_hang = false;
 	drm_dev_exit(idx);
-	return DRM_GPU_SCHED_STAT_NOMINAL;
+	return DRM_GPU_SCHED_STAT_RESET;
 }
 
 int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,

drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h

@@ -167,7 +167,6 @@ TRACE_EVENT(amdgpu_cs_ioctl,
 	    TP_PROTO(struct amdgpu_job *job),
 	    TP_ARGS(job),
 	    TP_STRUCT__entry(
-			     __field(uint64_t, sched_job_id)
 			     __string(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			     __field(unsigned int, context)
 			     __field(unsigned int, seqno)
@@ -177,23 +176,21 @@ TRACE_EVENT(amdgpu_cs_ioctl,
 			     ),
 
 	    TP_fast_assign(
-			   __entry->sched_job_id = job->base.id;
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job));
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
 			   __assign_str(ring, to_amdgpu_ring(job->base.sched)->name);
 			   __entry->num_ibs = job->num_ibs;
 			   ),
-	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
-		      __entry->sched_job_id, __get_str(timeline), __entry->context,
+	    TP_printk("timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
+		      __get_str(timeline), __entry->context,
 		      __entry->seqno, __get_str(ring), __entry->num_ibs)
 );
 
 TRACE_EVENT(amdgpu_sched_run_job,
 	    TP_PROTO(struct amdgpu_job *job),
 	    TP_ARGS(job),
 	    TP_STRUCT__entry(
-			     __field(uint64_t, sched_job_id)
 			     __string(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			     __field(unsigned int, context)
 			     __field(unsigned int, seqno)
@@ -202,15 +199,14 @@ TRACE_EVENT(amdgpu_sched_run_job,
 			     ),
 
 	    TP_fast_assign(
-			   __entry->sched_job_id = job->base.id;
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job));
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
 			   __assign_str(ring, to_amdgpu_ring(job->base.sched)->name);
 			   __entry->num_ibs = job->num_ibs;
 			   ),
-	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
-		      __entry->sched_job_id, __get_str(timeline), __entry->context,
+	    TP_printk("timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
+		      __get_str(timeline), __entry->context,
 		      __entry->seqno, __get_str(ring), __entry->num_ibs)
 );
 
@@ -519,21 +515,19 @@ TRACE_EVENT(amdgpu_ib_pipe_sync,
 	    TP_ARGS(sched_job, fence),
 	    TP_STRUCT__entry(
 			     __string(ring, sched_job->base.sched->name)
-			     __field(uint64_t, id)
 			     __field(struct dma_fence *, fence)
 			     __field(uint64_t, ctx)
 			     __field(unsigned, seqno)
 			     ),
 
 	    TP_fast_assign(
 			   __assign_str(ring, sched_job->base.sched->name);
-			   __entry->id = sched_job->base.id;
 			   __entry->fence = fence;
 			   __entry->ctx = fence->context;
 			   __entry->seqno = fence->seqno;
 			   ),
-	    TP_printk("job ring=%s, id=%llu, need pipe sync to fence=%p, context=%llu, seq=%u",
-		      __get_str(ring), __entry->id,
+	    TP_printk("job ring=%s need pipe sync to fence=%p, context=%llu, seq=%u",
+		      __get_str(ring),
 		      __entry->fence, __entry->ctx,
 		      __entry->seqno)
 );

drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c

@@ -531,7 +531,7 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 
 	ret = drm_sched_job_init(&submit->sched_job,
 				 &ctx->sched_entity[args->pipe],
-				 1, submit->ctx);
+				 1, submit->ctx, file->client_id);
 	if (ret)
 		goto err_submit_put;
 

drivers/gpu/drm/etnaviv/etnaviv_sched.c

@@ -72,11 +72,11 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
 	drm_sched_resubmit_jobs(&gpu->sched);
 
 	drm_sched_start(&gpu->sched, 0);
-	return DRM_GPU_SCHED_STAT_NOMINAL;
+	return DRM_GPU_SCHED_STAT_RESET;
 
 out_no_timeout:
 	list_add(&sched_job->list, &sched_job->sched->pending_list);
-	return DRM_GPU_SCHED_STAT_NOMINAL;
+	return DRM_GPU_SCHED_STAT_RESET;
 }
 
 static void etnaviv_sched_free_job(struct drm_sched_job *sched_job)

drivers/gpu/drm/lima/lima_gem.c

@@ -349,7 +349,7 @@ int lima_gem_submit(struct drm_file *file, struct lima_submit *submit)
 
 	err = lima_sched_task_init(
 		submit->task, submit->ctx->context + submit->pipe,
-		bos, submit->nr_bos, vm);
+		bos, submit->nr_bos, vm, file->client_id);
 	if (err)
 		goto err_out1;
 

drivers/gpu/drm/lima/lima_sched.c

@@ -112,7 +112,8 @@ static inline struct lima_sched_pipe *to_lima_pipe(struct drm_gpu_scheduler *sch
 int lima_sched_task_init(struct lima_sched_task *task,
 			 struct lima_sched_context *context,
 			 struct lima_bo **bos, int num_bos,
-			 struct lima_vm *vm)
+			 struct lima_vm *vm,
+			 u64 drm_client_id)
 {
 	int err, i;
 
@@ -123,7 +124,8 @@ int lima_sched_task_init(struct lima_sched_task *task,
 	for (i = 0; i < num_bos; i++)
 		drm_gem_object_get(&bos[i]->base.base);
 
-	err = drm_sched_job_init(&task->base, &context->base, 1, vm);
+	err = drm_sched_job_init(&task->base, &context->base, 1, vm,
+				 drm_client_id);
 	if (err) {
 		kfree(task->bos);
 		return err;
@@ -439,7 +441,7 @@ static enum drm_gpu_sched_stat lima_sched_timedout_job(struct drm_sched_job *job
 	drm_sched_resubmit_jobs(&pipe->base);
 	drm_sched_start(&pipe->base, 0);
 
-	return DRM_GPU_SCHED_STAT_NOMINAL;
+	return DRM_GPU_SCHED_STAT_RESET;
 }
 
 static void lima_sched_free_job(struct drm_sched_job *job)

drivers/gpu/drm/lima/lima_sched.h

@@ -88,7 +88,8 @@ struct lima_sched_pipe {
 int lima_sched_task_init(struct lima_sched_task *task,
 			 struct lima_sched_context *context,
 			 struct lima_bo **bos, int num_bos,
-			 struct lima_vm *vm);
+			 struct lima_vm *vm,
+			 u64 drm_client_id);
 void lima_sched_task_fini(struct lima_sched_task *task);
 
 int lima_sched_context_init(struct lima_sched_pipe *pipe,

drivers/gpu/drm/lima/lima_trace.h

@@ -14,21 +14,19 @@ DECLARE_EVENT_CLASS(lima_task,
 	TP_PROTO(struct lima_sched_task *task),
 	TP_ARGS(task),
 	TP_STRUCT__entry(
-		__field(uint64_t, task_id)
 		__field(unsigned int, context)
 		__field(unsigned int, seqno)
 		__string(pipe, task->base.sched->name)
 		),
 
 	TP_fast_assign(
-		__entry->task_id = task->base.id;
 		__entry->context = task->base.s_fence->finished.context;
 		__entry->seqno = task->base.s_fence->finished.seqno;
 		__assign_str(pipe, task->base.sched->name);
 		),
 
-	TP_printk("task=%llu, context=%u seqno=%u pipe=%s",
-		  __entry->task_id, __entry->context, __entry->seqno,
+	TP_printk("context=%u seqno=%u pipe=%s",
+		  __entry->context, __entry->seqno,
 		  __get_str(pipe))
 );
 

drivers/gpu/drm/panfrost/panfrost_drv.c

@@ -278,7 +278,7 @@ static int panfrost_ioctl_submit(struct drm_device *dev, void *data,
 
 	ret = drm_sched_job_init(&job->base,
 				 &file_priv->sched_entity[slot],
-				 1, NULL);
+				 1, NULL, file->client_id);
 	if (ret)
 		goto out_put_job;
 

drivers/gpu/drm/panfrost/panfrost_job.c

@@ -718,7 +718,7 @@ static enum drm_gpu_sched_stat panfrost_job_timedout(struct drm_sched_job
 	 * spurious. Bail out.
 	 */
 	if (dma_fence_is_signaled(job->done_fence))
-		return DRM_GPU_SCHED_STAT_NOMINAL;
+		return DRM_GPU_SCHED_STAT_RESET;
 
 	dev_err(pfdev->dev, "gpu sched timeout, js=%d, config=0x%x, status=0x%x, head=0x%x, tail=0x%x, sched_job=%p",
 		js,
@@ -733,7 +733,7 @@ static enum drm_gpu_sched_stat panfrost_job_timedout(struct drm_sched_job
 	atomic_set(&pfdev->reset.pending, 1);
 	panfrost_reset(pfdev, sched_job);
 
-	return DRM_GPU_SCHED_STAT_NOMINAL;
+	return DRM_GPU_SCHED_STAT_RESET;
 }
 
 static void panfrost_reset_work(struct work_struct *work)

drivers/gpu/drm/panthor/panthor_device.h

@@ -230,6 +230,24 @@ struct panthor_file {
 	/** @ptdev: Device attached to this file. */
 	struct panthor_device *ptdev;
 
+	/** @user_mmio: User MMIO related fields. */
+	struct {
+		/**
+		 * @offset: Offset used for user MMIO mappings.
+		 *
+		 * This offset should not be used to check the type of mapping
+		 * except in panthor_mmap(). After that point, MMIO mapping
+		 * offsets have been adjusted to match
+		 * DRM_PANTHOR_USER_MMIO_OFFSET and that macro should be used
+		 * instead.
+		 * Make sure this rule is followed at all times, because
+		 * userspace is in control of the offset, and can change the
+		 * value behind our back. Otherwise it can lead to erroneous
+		 * branching happening in kernel space.
+		 */
+		u64 offset;
+	} user_mmio;
+
 	/** @vms: VM pool attached to this file. */
 	struct panthor_vm_pool *vms;
 
@@ -439,4 +457,75 @@ static int panthor_request_ ## __name ## _irq(struct panthor_device *ptdev,			\
 
 extern struct workqueue_struct *panthor_cleanup_wq;
 
+static inline void gpu_write(struct panthor_device *ptdev, u32 reg, u32 data)
+{
+	writel(data, ptdev->iomem + reg);
+}
+
+static inline u32 gpu_read(struct panthor_device *ptdev, u32 reg)
+{
+	return readl(ptdev->iomem + reg);
+}
+
+static inline u32 gpu_read_relaxed(struct panthor_device *ptdev, u32 reg)
+{
+	return readl_relaxed(ptdev->iomem + reg);
+}
+
+static inline void gpu_write64(struct panthor_device *ptdev, u32 reg, u64 data)
+{
+	gpu_write(ptdev, reg, lower_32_bits(data));
+	gpu_write(ptdev, reg + 4, upper_32_bits(data));
+}
+
+static inline u64 gpu_read64(struct panthor_device *ptdev, u32 reg)
+{
+	return (gpu_read(ptdev, reg) | ((u64)gpu_read(ptdev, reg + 4) << 32));
+}
+
+static inline u64 gpu_read64_relaxed(struct panthor_device *ptdev, u32 reg)
+{
+	return (gpu_read_relaxed(ptdev, reg) |
+		((u64)gpu_read_relaxed(ptdev, reg + 4) << 32));
+}
+
+static inline u64 gpu_read64_counter(struct panthor_device *ptdev, u32 reg)
+{
+	u32 lo, hi1, hi2;
+	do {
+		hi1 = gpu_read(ptdev, reg + 4);
+		lo = gpu_read(ptdev, reg);
+		hi2 = gpu_read(ptdev, reg + 4);
+	} while (hi1 != hi2);
+	return lo | ((u64)hi2 << 32);
+}
+
+#define gpu_read_poll_timeout(dev, reg, val, cond, delay_us, timeout_us)	\
+	read_poll_timeout(gpu_read, val, cond, delay_us, timeout_us, false,	\
+			  dev, reg)
+
+#define gpu_read_poll_timeout_atomic(dev, reg, val, cond, delay_us,		\
+				     timeout_us)				\
+	read_poll_timeout_atomic(gpu_read, val, cond, delay_us, timeout_us,	\
+				 false, dev, reg)
+
+#define gpu_read64_poll_timeout(dev, reg, val, cond, delay_us, timeout_us)	\
+	read_poll_timeout(gpu_read64, val, cond, delay_us, timeout_us, false,	\
+			  dev, reg)
+
+#define gpu_read64_poll_timeout_atomic(dev, reg, val, cond, delay_us,		\
+				       timeout_us)				\
+	read_poll_timeout_atomic(gpu_read64, val, cond, delay_us, timeout_us,	\
+				 false, dev, reg)
+
+#define gpu_read_relaxed_poll_timeout_atomic(dev, reg, val, cond, delay_us,	\
+					     timeout_us)			\
+	read_poll_timeout_atomic(gpu_read_relaxed, val, cond, delay_us,		\
+				 timeout_us, false, dev, reg)
+
+#define gpu_read64_relaxed_poll_timeout(dev, reg, val, cond, delay_us,		\
+					timeout_us)				\
+	read_poll_timeout(gpu_read64_relaxed, val, cond, delay_us, timeout_us,	\
+			  false, dev, reg)
+
 #endif