Resguardo inicial

linux-7.2.3/drivers/gpu/drm/scheduler/tests/Makefile

@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0
+
+drm-sched-tests-y := \
+        mock_scheduler.o \
+        tests_basic.o \
+        tests_scheduler.o
+
+obj-$(CONFIG_DRM_SCHED_KUNIT_TEST) += drm-sched-tests.o

linux-7.2.3/drivers/gpu/drm/scheduler/tests/mock_scheduler.c

@@ -0,0 +1,370 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2025 Valve Corporation */
+
+#include "sched_tests.h"
+
+/*
+ * Here we implement the mock "GPU" (or the scheduler backend) which is used by
+ * the DRM scheduler unit tests in order to exercise the core functionality.
+ *
+ * Test cases are implemented in a separate file.
+ */
+
+/**
+ * drm_mock_sched_entity_new - Create a new mock scheduler entity
+ *
+ * @test: KUnit test owning the entity
+ * @priority: Scheduling priority
+ * @sched: Mock scheduler on which the entity can be scheduled
+ *
+ * Returns: New mock scheduler entity with allocation managed by the test
+ */
+struct drm_mock_sched_entity *
+drm_mock_sched_entity_new(struct kunit *test,
+			  enum drm_sched_priority priority,
+			  struct drm_mock_scheduler *sched)
+{
+	struct drm_mock_sched_entity *entity;
+	struct drm_gpu_scheduler *drm_sched;
+	int ret;
+
+	entity = kunit_kzalloc(test, sizeof(*entity), GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, entity);
+
+	drm_sched = &sched->base;
+	ret = drm_sched_entity_init(&entity->base,
+				    priority,
+				    &drm_sched, 1,
+				    NULL);
+	KUNIT_ASSERT_EQ(test, ret, 0);
+
+	entity->test = test;
+
+	return entity;
+}
+
+/**
+ * drm_mock_sched_entity_free - Destroys a mock scheduler entity
+ *
+ * @entity: Entity to destroy
+ *
+ * To be used from the test cases once done with the entity.
+ */
+void drm_mock_sched_entity_free(struct drm_mock_sched_entity *entity)
+{
+	drm_sched_entity_destroy(&entity->base);
+}
+
+static void drm_mock_sched_job_complete(struct drm_mock_sched_job *job)
+{
+	struct drm_mock_scheduler *sched =
+		drm_sched_to_mock_sched(job->base.sched);
+
+	lockdep_assert_held(&sched->lock);
+
+	job->flags |= DRM_MOCK_SCHED_JOB_DONE;
+	list_del(&job->link);
+	dma_fence_signal_locked(&job->hw_fence);
+	complete(&job->done);
+}
+
+static enum hrtimer_restart
+drm_mock_sched_job_signal_timer(struct hrtimer *hrtimer)
+{
+	struct drm_mock_sched_job *job =
+		container_of(hrtimer, typeof(*job), timer);
+	struct drm_mock_scheduler *sched =
+		drm_sched_to_mock_sched(job->base.sched);
+	struct drm_mock_sched_job *next;
+	ktime_t now = ktime_get();
+	unsigned long flags;
+	LIST_HEAD(signal);
+
+	spin_lock_irqsave(&sched->lock, flags);
+	list_for_each_entry_safe(job, next, &sched->job_list, link) {
+		if (!job->duration_us)
+			break;
+
+		if (ktime_before(now, job->finish_at))
+			break;
+
+		sched->hw_timeline.cur_seqno = job->hw_fence.seqno;
+		drm_mock_sched_job_complete(job);
+	}
+	spin_unlock_irqrestore(&sched->lock, flags);
+
+	return HRTIMER_NORESTART;
+}
+
+/**
+ * drm_mock_sched_job_new - Create a new mock scheduler job
+ *
+ * @test: KUnit test owning the job
+ * @entity: Scheduler entity of the job
+ *
+ * Returns: New mock scheduler job with allocation managed by the test
+ */
+struct drm_mock_sched_job *
+drm_mock_sched_job_new(struct kunit *test,
+		       struct drm_mock_sched_entity *entity)
+{
+	struct drm_mock_sched_job *job;
+	int ret;
+
+	job = kunit_kzalloc(test, sizeof(*job), GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, job);
+
+	ret = drm_sched_job_init(&job->base,
+				 &entity->base,
+				 1,
+				 NULL,
+				 1);
+	KUNIT_ASSERT_EQ(test, ret, 0);
+
+	job->test = test;
+
+	init_completion(&job->done);
+	INIT_LIST_HEAD(&job->link);
+	hrtimer_setup(&job->timer, drm_mock_sched_job_signal_timer,
+		      CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
+
+	return job;
+}
+
+static const char *drm_mock_sched_hw_fence_driver_name(struct dma_fence *fence)
+{
+	return "drm_mock_sched";
+}
+
+static const char *
+drm_mock_sched_hw_fence_timeline_name(struct dma_fence *fence)
+{
+	struct drm_mock_sched_job *job =
+		container_of(fence, typeof(*job), hw_fence);
+
+	return (const char *)job->base.sched->name;
+}
+
+static void drm_mock_sched_hw_fence_release(struct dma_fence *fence)
+{
+	struct drm_mock_sched_job *job =
+		container_of(fence, typeof(*job), hw_fence);
+
+	hrtimer_cancel(&job->timer);
+
+	/* Containing job is freed by the kunit framework */
+}
+
+static const struct dma_fence_ops drm_mock_sched_hw_fence_ops = {
+	.get_driver_name = drm_mock_sched_hw_fence_driver_name,
+	.get_timeline_name = drm_mock_sched_hw_fence_timeline_name,
+	.release = drm_mock_sched_hw_fence_release,
+};
+
+static struct dma_fence *mock_sched_run_job(struct drm_sched_job *sched_job)
+{
+	struct drm_mock_scheduler *sched =
+		drm_sched_to_mock_sched(sched_job->sched);
+	struct drm_mock_sched_job *job = drm_sched_job_to_mock_job(sched_job);
+
+	dma_fence_init(&job->hw_fence,
+		       &drm_mock_sched_hw_fence_ops,
+		       &sched->lock,
+		       sched->hw_timeline.context,
+		       atomic_inc_return(&sched->hw_timeline.next_seqno));
+
+	dma_fence_get(&job->hw_fence); /* Reference for the job_list */
+
+	spin_lock_irq(&sched->lock);
+	if (job->duration_us) {
+		ktime_t prev_finish_at = 0;
+
+		if (!list_empty(&sched->job_list)) {
+			struct drm_mock_sched_job *prev =
+				list_last_entry(&sched->job_list, typeof(*prev),
+						link);
+
+			prev_finish_at = prev->finish_at;
+		}
+
+		if (!prev_finish_at)
+			prev_finish_at = ktime_get();
+
+		job->finish_at = ktime_add_us(prev_finish_at, job->duration_us);
+	}
+	list_add_tail(&job->link, &sched->job_list);
+	if (job->finish_at)
+		hrtimer_start(&job->timer, job->finish_at, HRTIMER_MODE_ABS);
+	spin_unlock_irq(&sched->lock);
+
+	return &job->hw_fence;
+}
+
+/*
+ * Normally, drivers would take appropriate measures in this callback, such as
+ * killing the entity the faulty job is associated with, resetting the hardware
+ * and / or resubmitting non-faulty jobs.
+ *
+ * For the mock scheduler, there are no hardware rings to be resetted nor jobs
+ * to be resubmitted. Thus, this function merely ensures that
+ *   a) timedout fences get signaled properly and removed from the pending list
+ *   b) the mock scheduler framework gets informed about the timeout via a flag
+ *   c) The drm_sched_job, not longer needed, gets freed
+ */
+static enum drm_gpu_sched_stat
+mock_sched_timedout_job(struct drm_sched_job *sched_job)
+{
+	struct drm_mock_scheduler *sched = drm_sched_to_mock_sched(sched_job->sched);
+	struct drm_mock_sched_job *job = drm_sched_job_to_mock_job(sched_job);
+	unsigned long flags;
+
+	if (job->flags & DRM_MOCK_SCHED_JOB_DONT_RESET) {
+		job->flags |= DRM_MOCK_SCHED_JOB_RESET_SKIPPED;
+		return DRM_GPU_SCHED_STAT_NO_HANG;
+	}
+
+	spin_lock_irqsave(&sched->lock, flags);
+	if (!dma_fence_is_signaled_locked(&job->hw_fence)) {
+		list_del(&job->link);
+		job->flags |= DRM_MOCK_SCHED_JOB_TIMEDOUT;
+		dma_fence_set_error(&job->hw_fence, -ETIMEDOUT);
+		dma_fence_signal_locked(&job->hw_fence);
+	}
+	spin_unlock_irqrestore(&sched->lock, flags);
+
+	dma_fence_put(&job->hw_fence);
+	drm_sched_job_cleanup(sched_job);
+	/* Mock job itself is freed by the kunit framework. */
+
+	return DRM_GPU_SCHED_STAT_RESET;
+}
+
+static void mock_sched_free_job(struct drm_sched_job *sched_job)
+{
+	struct drm_mock_sched_job *job = drm_sched_job_to_mock_job(sched_job);
+
+	dma_fence_put(&job->hw_fence);
+	drm_sched_job_cleanup(sched_job);
+
+	/* Mock job itself is freed by the kunit framework. */
+}
+
+static void mock_sched_cancel_job(struct drm_sched_job *sched_job)
+{
+	struct drm_mock_scheduler *sched = drm_sched_to_mock_sched(sched_job->sched);
+	struct drm_mock_sched_job *job = drm_sched_job_to_mock_job(sched_job);
+	unsigned long flags;
+
+	hrtimer_cancel(&job->timer);
+
+	spin_lock_irqsave(&sched->lock, flags);
+	if (!dma_fence_is_signaled_locked(&job->hw_fence)) {
+		list_del(&job->link);
+		dma_fence_set_error(&job->hw_fence, -ECANCELED);
+		dma_fence_signal_locked(&job->hw_fence);
+	}
+	spin_unlock_irqrestore(&sched->lock, flags);
+
+	/*
+	 * The GPU Scheduler will call drm_sched_backend_ops.free_job(), still.
+	 * Mock job itself is freed by the kunit framework.
+	 */
+}
+
+static const struct drm_sched_backend_ops drm_mock_scheduler_ops = {
+	.run_job = mock_sched_run_job,
+	.timedout_job = mock_sched_timedout_job,
+	.free_job = mock_sched_free_job,
+	.cancel_job = mock_sched_cancel_job,
+};
+
+/**
+ * drm_mock_sched_new - Create a new mock scheduler
+ *
+ * @test: KUnit test owning the job
+ * @timeout: Job timeout to set
+ *
+ * Returns: New mock scheduler with allocation managed by the test
+ */
+struct drm_mock_scheduler *drm_mock_sched_new(struct kunit *test, long timeout)
+{
+	struct drm_sched_init_args args = {
+		.ops		= &drm_mock_scheduler_ops,
+		.num_rqs	= DRM_SCHED_PRIORITY_COUNT,
+		.credit_limit	= U32_MAX,
+		.hang_limit	= 1,
+		.timeout	= timeout,
+		.name		= "drm-mock-scheduler",
+	};
+	struct drm_mock_scheduler *sched;
+	int ret;
+
+	sched = kunit_kzalloc(test, sizeof(*sched), GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, sched);
+
+	ret = drm_sched_init(&sched->base, &args);
+	KUNIT_ASSERT_EQ(test, ret, 0);
+
+	sched->test = test;
+	sched->hw_timeline.context = dma_fence_context_alloc(1);
+	atomic_set(&sched->hw_timeline.next_seqno, 0);
+	INIT_LIST_HEAD(&sched->job_list);
+	spin_lock_init(&sched->lock);
+
+	return sched;
+}
+
+/**
+ * drm_mock_sched_fini - Destroys a mock scheduler
+ *
+ * @sched: Scheduler to destroy
+ *
+ * To be used from the test cases once done with the scheduler.
+ */
+void drm_mock_sched_fini(struct drm_mock_scheduler *sched)
+{
+	drm_sched_fini(&sched->base);
+}
+
+/**
+ * drm_mock_sched_advance - Advances the mock scheduler timeline
+ *
+ * @sched: Scheduler timeline to advance
+ * @num: By how many jobs to advance
+ *
+ * Advancing the scheduler timeline by a number of seqnos will trigger
+ * signalling of the hardware fences and unlinking the jobs from the internal
+ * scheduler tracking.
+ *
+ * This can be used from test cases which want complete control of the simulated
+ * job execution timing. For example submitting one job with no set duration
+ * would never complete it before test cases advances the timeline by one.
+ */
+unsigned int drm_mock_sched_advance(struct drm_mock_scheduler *sched,
+				    unsigned int num)
+{
+	struct drm_mock_sched_job *job, *next;
+	unsigned int found = 0;
+	unsigned long flags;
+	LIST_HEAD(signal);
+
+	spin_lock_irqsave(&sched->lock, flags);
+	if (WARN_ON_ONCE(sched->hw_timeline.cur_seqno + num <
+			 sched->hw_timeline.cur_seqno))
+		goto unlock;
+	sched->hw_timeline.cur_seqno += num;
+	list_for_each_entry_safe(job, next, &sched->job_list, link) {
+		if (sched->hw_timeline.cur_seqno < job->hw_fence.seqno)
+			break;
+
+		drm_mock_sched_job_complete(job);
+		found++;
+	}
+unlock:
+	spin_unlock_irqrestore(&sched->lock, flags);
+
+	return found;
+}
+
+MODULE_DESCRIPTION("DRM mock scheduler and tests");
+MODULE_LICENSE("GPL");

linux-7.2.3/drivers/gpu/drm/scheduler/tests/sched_tests.h

@@ -0,0 +1,224 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (c) 2025 Valve Corporation */
+
+#ifndef _SCHED_TESTS_H_
+#define _SCHED_TESTS_H_
+
+#include <kunit/test.h>
+#include <linux/atomic.h>
+#include <linux/completion.h>
+#include <linux/dma-fence.h>
+#include <linux/hrtimer.h>
+#include <linux/ktime.h>
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/types.h>
+
+#include <drm/gpu_scheduler.h>
+
+/*
+ * DOC: Mock DRM scheduler data structures
+ *
+ * drm_mock_* data structures are used to implement a mock "GPU".
+ *
+ * They subclass the core DRM scheduler objects and add their data on top, which
+ * enables tracking the submitted jobs and simulating their execution with the
+ * attributes as specified by the test case.
+ */
+
+/**
+ * struct drm_mock_scheduler - implements a trivial mock GPU execution engine
+ *
+ * @base: DRM scheduler base class
+ * @test: Backpointer to owning the kunit test case
+ * @lock: Lock to protect the simulated @hw_timeline and @job_list
+ * @job_list: List of jobs submitted to the mock GPU
+ * @hw_timeline: Simulated hardware timeline has a @context, @next_seqno and
+ *		 @cur_seqno for implementing a struct dma_fence signaling the
+ *		 simulated job completion.
+ *
+ * Trivial mock GPU execution engine tracks submitted jobs and enables
+ * completing them strictly in submission order.
+ */
+struct drm_mock_scheduler {
+	struct drm_gpu_scheduler base;
+
+	struct kunit		*test;
+
+	spinlock_t		lock;
+	struct list_head	job_list;
+
+	struct {
+		u64		context;
+		atomic_t	next_seqno;
+		unsigned int	cur_seqno;
+	} hw_timeline;
+};
+
+/**
+ * struct drm_mock_sched_entity - implements a mock GPU sched entity
+ *
+ * @base: DRM scheduler entity base class
+ * @test: Backpointer to owning the kunit test case
+ *
+ * Mock GPU sched entity is used by the test cases to submit jobs to the mock
+ * scheduler.
+ */
+struct drm_mock_sched_entity {
+	struct drm_sched_entity base;
+
+	struct kunit		*test;
+};
+
+/**
+ * struct drm_mock_sched_job - implements a mock GPU job
+ *
+ * @base: DRM sched job base class
+ * @done: Completion signaling job completion.
+ * @flags: Flags designating job state.
+ * @link: List head element used by job tracking by the drm_mock_scheduler
+ * @timer: Timer used for simulating job execution duration
+ * @duration_us: Simulated job duration in micro seconds, or zero if in manual
+ *		 timeline advance mode
+ * @finish_at: Absolute time when the jobs with set duration will complete
+ * @lock: Lock used for @hw_fence
+ * @hw_fence: Fence returned to DRM scheduler as the hardware fence
+ * @test: Backpointer to owning the kunit test case
+ *
+ * Mock GPU sched job is used by the test cases to submit jobs to the mock
+ * scheduler.
+ */
+struct drm_mock_sched_job {
+	struct drm_sched_job	base;
+
+	struct completion	done;
+
+#define DRM_MOCK_SCHED_JOB_DONE			0x1
+#define DRM_MOCK_SCHED_JOB_TIMEDOUT		0x2
+#define DRM_MOCK_SCHED_JOB_DONT_RESET		0x4
+#define DRM_MOCK_SCHED_JOB_RESET_SKIPPED	0x8
+	unsigned long		flags;
+
+	struct list_head	link;
+	struct hrtimer		timer;
+
+	unsigned int		duration_us;
+	ktime_t			finish_at;
+
+	struct dma_fence	hw_fence;
+
+	struct kunit		*test;
+};
+
+static inline struct drm_mock_scheduler *
+drm_sched_to_mock_sched(struct drm_gpu_scheduler *sched)
+{
+	return container_of(sched, struct drm_mock_scheduler, base);
+};
+
+static inline struct drm_mock_sched_entity *
+drm_sched_entity_to_mock_entity(struct drm_sched_entity *sched_entity)
+{
+	return container_of(sched_entity, struct drm_mock_sched_entity, base);
+};
+
+static inline struct drm_mock_sched_job *
+drm_sched_job_to_mock_job(struct drm_sched_job *sched_job)
+{
+	return container_of(sched_job, struct drm_mock_sched_job, base);
+};
+
+struct drm_mock_scheduler *drm_mock_sched_new(struct kunit *test,
+					      long timeout);
+void drm_mock_sched_fini(struct drm_mock_scheduler *sched);
+unsigned int drm_mock_sched_advance(struct drm_mock_scheduler *sched,
+				    unsigned int num);
+
+struct drm_mock_sched_entity *
+drm_mock_sched_entity_new(struct kunit *test,
+			  enum drm_sched_priority priority,
+			  struct drm_mock_scheduler *sched);
+void drm_mock_sched_entity_free(struct drm_mock_sched_entity *entity);
+
+struct drm_mock_sched_job *
+drm_mock_sched_job_new(struct kunit *test,
+		       struct drm_mock_sched_entity *entity);
+
+/**
+ * drm_mock_sched_job_submit - Arm and submit a job in one go
+ *
+ * @job: Job to arm and submit
+ */
+static inline void drm_mock_sched_job_submit(struct drm_mock_sched_job *job)
+{
+	drm_sched_job_arm(&job->base);
+	drm_sched_entity_push_job(&job->base);
+}
+
+/**
+ * drm_mock_sched_job_set_duration_us - Set a job duration
+ *
+ * @job: Job to set the duration for
+ * @duration_us: Duration in micro seconds
+ *
+ * Jobs with duration set will be automatically completed by the mock scheduler
+ * as the timeline progresses, unless a job without a set duration is
+ * encountered in the timelime in which case calling drm_mock_sched_advance()
+ * will be required to bump the timeline.
+ */
+static inline void
+drm_mock_sched_job_set_duration_us(struct drm_mock_sched_job *job,
+				   unsigned int duration_us)
+{
+	job->duration_us = duration_us;
+}
+
+/**
+ * drm_mock_sched_job_is_finished - Check if a job is finished
+ *
+ * @job: Job to check
+ *
+ * Returns: true if finished
+ */
+static inline bool
+drm_mock_sched_job_is_finished(struct drm_mock_sched_job *job)
+{
+	return job->flags & DRM_MOCK_SCHED_JOB_DONE;
+}
+
+/**
+ * drm_mock_sched_job_wait_finished - Wait until a job is finished
+ *
+ * @job: Job to wait for
+ * @timeout: Wait time in jiffies
+ *
+ * Returns: true if finished within the timeout provided, otherwise false
+ */
+static inline bool
+drm_mock_sched_job_wait_finished(struct drm_mock_sched_job *job, long timeout)
+{
+	if (job->flags & DRM_MOCK_SCHED_JOB_DONE)
+		return true;
+
+	return wait_for_completion_timeout(&job->done, timeout) != 0;
+}
+
+/**
+ * drm_mock_sched_job_wait_scheduled - Wait until a job is scheduled
+ *
+ * @job: Job to wait for
+ * @timeout: Wait time in jiffies
+ *
+ * Returns: true if scheduled within the timeout provided, otherwise false
+ */
+static inline bool
+drm_mock_sched_job_wait_scheduled(struct drm_mock_sched_job *job, long timeout)
+{
+	KUNIT_ASSERT_EQ(job->test, job->flags & DRM_MOCK_SCHED_JOB_DONE, 0);
+
+	return dma_fence_wait_timeout(&job->base.s_fence->scheduled,
+				      false,
+				      timeout) != 0;
+}
+
+#endif

linux-7.2.3/drivers/gpu/drm/scheduler/tests/tests_basic.c

@@ -0,0 +1,563 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2025 Valve Corporation */
+
+#include <linux/delay.h>
+
+#include "sched_tests.h"
+
+#define MOCK_TIMEOUT (HZ / 5)
+
+/*
+ * DRM scheduler basic tests should check the basic functional correctness of
+ * the scheduler, including some very light smoke testing. More targeted tests,
+ * for example focusing on testing specific bugs and other more complicated test
+ * scenarios, should be implemented in separate source units.
+ */
+
+static int drm_sched_basic_init(struct kunit *test)
+{
+	test->priv = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+
+	return 0;
+}
+
+static void drm_sched_basic_exit(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+
+	drm_mock_sched_fini(sched);
+}
+
+static int drm_sched_timeout_init(struct kunit *test)
+{
+	test->priv = drm_mock_sched_new(test, MOCK_TIMEOUT);
+
+	return 0;
+}
+
+static void drm_sched_basic_submit(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity;
+	struct drm_mock_sched_job *job;
+	unsigned int i;
+	bool done;
+
+	/*
+	 * Submit one job to the scheduler and verify that it gets scheduled
+	 * and completed only when the mock hw backend processes it.
+	 */
+
+	entity = drm_mock_sched_entity_new(test,
+					   DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+	job = drm_mock_sched_job_new(test, entity);
+
+	drm_mock_sched_job_submit(job);
+
+	done = drm_mock_sched_job_wait_scheduled(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	done = drm_mock_sched_job_wait_finished(job, HZ / 2);
+	KUNIT_ASSERT_FALSE(test, done);
+
+	i = drm_mock_sched_advance(sched, 1);
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	done = drm_mock_sched_job_wait_finished(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	drm_mock_sched_entity_free(entity);
+}
+
+struct drm_sched_basic_params {
+	const char *description;
+	unsigned int queue_depth;
+	unsigned int num_entities;
+	unsigned int job_us;
+	bool dep_chain;
+};
+
+static const struct drm_sched_basic_params drm_sched_basic_cases[] = {
+	{
+		.description = "A queue of jobs in a single entity",
+		.queue_depth = 100,
+		.job_us = 1000,
+		.num_entities = 1,
+	},
+	{
+		.description = "A chain of dependent jobs across multiple entities",
+		.queue_depth = 100,
+		.job_us = 1000,
+		.num_entities = 1,
+		.dep_chain = true,
+	},
+	{
+		.description = "Multiple independent job queues",
+		.queue_depth = 100,
+		.job_us = 1000,
+		.num_entities = 4,
+	},
+	{
+		.description = "Multiple inter-dependent job queues",
+		.queue_depth = 100,
+		.job_us = 1000,
+		.num_entities = 4,
+		.dep_chain = true,
+	},
+};
+
+static void
+drm_sched_basic_desc(const struct drm_sched_basic_params *params, char *desc)
+{
+	strscpy(desc, params->description, KUNIT_PARAM_DESC_SIZE);
+}
+
+KUNIT_ARRAY_PARAM(drm_sched_basic, drm_sched_basic_cases, drm_sched_basic_desc);
+
+static void drm_sched_basic_test(struct kunit *test)
+{
+	const struct drm_sched_basic_params *params = test->param_value;
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_job *job, *prev = NULL;
+	struct drm_mock_sched_entity **entity;
+	unsigned int i, cur_ent = 0;
+	bool done;
+
+	entity = kunit_kcalloc(test, params->num_entities, sizeof(*entity),
+			       GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, entity);
+
+	for (i = 0; i < params->num_entities; i++)
+		entity[i] = drm_mock_sched_entity_new(test,
+						      DRM_SCHED_PRIORITY_NORMAL,
+						      sched);
+
+	for (i = 0; i < params->queue_depth; i++) {
+		job = drm_mock_sched_job_new(test, entity[cur_ent++]);
+		cur_ent %= params->num_entities;
+		drm_mock_sched_job_set_duration_us(job, params->job_us);
+		if (params->dep_chain && prev)
+			drm_sched_job_add_dependency(&job->base,
+						     dma_fence_get(&prev->base.s_fence->finished));
+		drm_mock_sched_job_submit(job);
+		prev = job;
+	}
+
+	done = drm_mock_sched_job_wait_finished(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	for (i = 0; i < params->num_entities; i++)
+		drm_mock_sched_entity_free(entity[i]);
+}
+
+static void drm_sched_basic_entity_cleanup(struct kunit *test)
+{
+	struct drm_mock_sched_job *job, *mid, *prev = NULL;
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity[4];
+	const unsigned int qd = 100;
+	unsigned int i, cur_ent = 0;
+	bool done;
+
+	/*
+	 * Submit a queue of jobs across different entities with an explicit
+	 * chain of dependencies between them and trigger entity cleanup while
+	 * the queue is still being processed.
+	 */
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		entity[i] = drm_mock_sched_entity_new(test,
+						      DRM_SCHED_PRIORITY_NORMAL,
+						      sched);
+
+	for (i = 0; i < qd; i++) {
+		job = drm_mock_sched_job_new(test, entity[cur_ent++]);
+		cur_ent %= ARRAY_SIZE(entity);
+		drm_mock_sched_job_set_duration_us(job, 1000);
+		if (prev)
+			drm_sched_job_add_dependency(&job->base,
+						     dma_fence_get(&prev->base.s_fence->finished));
+		drm_mock_sched_job_submit(job);
+		if (i == qd / 2)
+			mid = job;
+		prev = job;
+	}
+
+	done = drm_mock_sched_job_wait_finished(mid, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	/* Exit with half of the queue still pending to be executed. */
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		drm_mock_sched_entity_free(entity[i]);
+}
+
+static struct kunit_case drm_sched_basic_tests[] = {
+	KUNIT_CASE(drm_sched_basic_submit),
+	KUNIT_CASE_PARAM(drm_sched_basic_test, drm_sched_basic_gen_params),
+	KUNIT_CASE(drm_sched_basic_entity_cleanup),
+	{}
+};
+
+static struct kunit_suite drm_sched_basic = {
+	.name = "drm_sched_basic_tests",
+	.init = drm_sched_basic_init,
+	.exit = drm_sched_basic_exit,
+	.test_cases = drm_sched_basic_tests,
+};
+
+static void drm_sched_basic_cancel(struct kunit *test)
+{
+	struct drm_mock_sched_entity *entity;
+	struct drm_mock_scheduler *sched;
+	struct drm_mock_sched_job *job;
+	bool done;
+
+	/*
+	 * Check that drm_sched_fini() uses the cancel_job() callback to cancel
+	 * jobs that are still pending.
+	 */
+
+	sched = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+	entity = drm_mock_sched_entity_new(test, DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+
+	job = drm_mock_sched_job_new(test, entity);
+
+	drm_mock_sched_job_submit(job);
+
+	done = drm_mock_sched_job_wait_scheduled(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	drm_mock_sched_entity_free(entity);
+	drm_mock_sched_fini(sched);
+
+	KUNIT_ASSERT_EQ(test, job->hw_fence.error, -ECANCELED);
+}
+
+static struct kunit_case drm_sched_cancel_tests[] = {
+	KUNIT_CASE(drm_sched_basic_cancel),
+	{}
+};
+
+static struct kunit_suite drm_sched_cancel = {
+	.name = "drm_sched_basic_cancel_tests",
+	.init = drm_sched_basic_init,
+	.exit = drm_sched_basic_exit,
+	.test_cases = drm_sched_cancel_tests,
+};
+
+static void drm_sched_basic_timeout(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity;
+	struct drm_mock_sched_job *job;
+	bool done;
+
+	/*
+	 * Submit a single job against a scheduler with the timeout configured
+	 * and verify that the timeout handling will run if the backend fails
+	 * to complete it in time.
+	 */
+
+	entity = drm_mock_sched_entity_new(test,
+					   DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+	job = drm_mock_sched_job_new(test, entity);
+
+	drm_mock_sched_job_submit(job);
+
+	done = drm_mock_sched_job_wait_scheduled(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	done = drm_mock_sched_job_wait_finished(job, MOCK_TIMEOUT / 2);
+	KUNIT_ASSERT_FALSE(test, done);
+
+	KUNIT_ASSERT_EQ(test,
+			job->flags & DRM_MOCK_SCHED_JOB_TIMEDOUT,
+			0);
+
+	done = drm_mock_sched_job_wait_finished(job, MOCK_TIMEOUT);
+	KUNIT_ASSERT_FALSE(test, done);
+
+	KUNIT_ASSERT_EQ(test,
+			job->flags & DRM_MOCK_SCHED_JOB_TIMEDOUT,
+			DRM_MOCK_SCHED_JOB_TIMEDOUT);
+
+	drm_mock_sched_entity_free(entity);
+}
+
+static void drm_sched_skip_reset(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity;
+	struct drm_mock_sched_job *job;
+	unsigned int i;
+	bool done;
+
+	/*
+	 * Submit a single job against a scheduler with the timeout configured
+	 * and verify that if the job is still running, the timeout handler
+	 * will skip the reset and allow the job to complete.
+	 */
+
+	entity = drm_mock_sched_entity_new(test,
+					   DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+	job = drm_mock_sched_job_new(test, entity);
+
+	job->flags = DRM_MOCK_SCHED_JOB_DONT_RESET;
+
+	drm_mock_sched_job_submit(job);
+
+	done = drm_mock_sched_job_wait_scheduled(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	done = drm_mock_sched_job_wait_finished(job, 2 * MOCK_TIMEOUT);
+	KUNIT_ASSERT_FALSE(test, done);
+
+	KUNIT_ASSERT_EQ(test,
+			job->flags & DRM_MOCK_SCHED_JOB_RESET_SKIPPED,
+			DRM_MOCK_SCHED_JOB_RESET_SKIPPED);
+
+	i = drm_mock_sched_advance(sched, 1);
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	done = drm_mock_sched_job_wait_finished(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	drm_mock_sched_entity_free(entity);
+}
+
+static struct kunit_case drm_sched_timeout_tests[] = {
+	KUNIT_CASE(drm_sched_basic_timeout),
+	KUNIT_CASE(drm_sched_skip_reset),
+	{}
+};
+
+static struct kunit_suite drm_sched_timeout = {
+	.name = "drm_sched_basic_timeout_tests",
+	.init = drm_sched_timeout_init,
+	.exit = drm_sched_basic_exit,
+	.test_cases = drm_sched_timeout_tests,
+};
+
+static void drm_sched_priorities(struct kunit *test)
+{
+	struct drm_mock_sched_entity *entity[DRM_SCHED_PRIORITY_COUNT];
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_job *job;
+	const unsigned int qd = 100;
+	unsigned int i, cur_ent = 0;
+	enum drm_sched_priority p;
+	bool done;
+
+	/*
+	 * Submit a bunch of jobs against entities configured with different
+	 * priorities.
+	 */
+
+	BUILD_BUG_ON(DRM_SCHED_PRIORITY_KERNEL > DRM_SCHED_PRIORITY_LOW);
+	BUILD_BUG_ON(ARRAY_SIZE(entity) != DRM_SCHED_PRIORITY_COUNT);
+
+	for (p = DRM_SCHED_PRIORITY_KERNEL; p <= DRM_SCHED_PRIORITY_LOW; p++)
+		entity[p] = drm_mock_sched_entity_new(test, p, sched);
+
+	for (i = 0; i < qd; i++) {
+		job = drm_mock_sched_job_new(test, entity[cur_ent++]);
+		cur_ent %= ARRAY_SIZE(entity);
+		drm_mock_sched_job_set_duration_us(job, 1000);
+		drm_mock_sched_job_submit(job);
+	}
+
+	done = drm_mock_sched_job_wait_finished(job, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		drm_mock_sched_entity_free(entity[i]);
+}
+
+static void drm_sched_change_priority(struct kunit *test)
+{
+	struct drm_mock_sched_entity *entity[DRM_SCHED_PRIORITY_COUNT];
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_job *job;
+	const unsigned int qd = 1000;
+	unsigned int i, cur_ent = 0;
+	enum drm_sched_priority p;
+
+	/*
+	 * Submit a bunch of jobs against entities configured with different
+	 * priorities and while waiting for them to complete, periodically keep
+	 * changing their priorities.
+	 *
+	 * We set up the queue-depth (qd) and job duration so the priority
+	 * changing loop has some time to interact with submissions to the
+	 * backend and job completions as they progress.
+	 */
+
+	for (p = DRM_SCHED_PRIORITY_KERNEL; p <= DRM_SCHED_PRIORITY_LOW; p++)
+		entity[p] = drm_mock_sched_entity_new(test, p, sched);
+
+	for (i = 0; i < qd; i++) {
+		job = drm_mock_sched_job_new(test, entity[cur_ent++]);
+		cur_ent %= ARRAY_SIZE(entity);
+		drm_mock_sched_job_set_duration_us(job, 1000);
+		drm_mock_sched_job_submit(job);
+	}
+
+	do {
+		drm_sched_entity_set_priority(&entity[cur_ent]->base,
+					      (entity[cur_ent]->base.priority + 1) %
+					      DRM_SCHED_PRIORITY_COUNT);
+		cur_ent++;
+		cur_ent %= ARRAY_SIZE(entity);
+		usleep_range(200, 500);
+	} while (!drm_mock_sched_job_is_finished(job));
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		drm_mock_sched_entity_free(entity[i]);
+}
+
+static struct kunit_case drm_sched_priority_tests[] = {
+	KUNIT_CASE(drm_sched_priorities),
+	KUNIT_CASE_SLOW(drm_sched_change_priority),
+	{}
+};
+
+static struct kunit_suite drm_sched_priority = {
+	.name = "drm_sched_basic_priority_tests",
+	.init = drm_sched_basic_init,
+	.exit = drm_sched_basic_exit,
+	.test_cases = drm_sched_priority_tests,
+};
+
+static void drm_sched_test_modify_sched(struct kunit *test)
+{
+	unsigned int i, cur_ent = 0, cur_sched = 0;
+	struct drm_mock_sched_entity *entity[13];
+	struct drm_mock_scheduler *sched[3];
+	struct drm_mock_sched_job *job;
+	const unsigned int qd = 1000;
+
+	/*
+	 * Submit a bunch of jobs against entities configured with different
+	 * schedulers and while waiting for them to complete, periodically keep
+	 * changing schedulers associated with each entity.
+	 *
+	 * We set up the queue-depth (qd) and job duration so the sched modify
+	 * loop has some time to interact with submissions to the backend and
+	 * job completions as they progress.
+	 *
+	 * For the number of schedulers and entities we use primes in order to
+	 * perturb the entity->sched assignments with less of a regular pattern.
+	 */
+
+	for (i = 0; i < ARRAY_SIZE(sched); i++)
+		sched[i] = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		entity[i] = drm_mock_sched_entity_new(test,
+						      DRM_SCHED_PRIORITY_NORMAL,
+						      sched[i % ARRAY_SIZE(sched)]);
+
+	for (i = 0; i < qd; i++) {
+		job = drm_mock_sched_job_new(test, entity[cur_ent++]);
+		cur_ent %= ARRAY_SIZE(entity);
+		drm_mock_sched_job_set_duration_us(job, 1000);
+		drm_mock_sched_job_submit(job);
+	}
+
+	do {
+		struct drm_gpu_scheduler *modify;
+
+		usleep_range(200, 500);
+		cur_ent++;
+		cur_ent %= ARRAY_SIZE(entity);
+		cur_sched++;
+		cur_sched %= ARRAY_SIZE(sched);
+		modify = &sched[cur_sched]->base;
+		drm_sched_entity_modify_sched(&entity[cur_ent]->base, &modify,
+					      1);
+	} while (!drm_mock_sched_job_is_finished(job));
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		drm_mock_sched_entity_free(entity[i]);
+
+	for (i = 0; i < ARRAY_SIZE(sched); i++)
+		drm_mock_sched_fini(sched[i]);
+}
+
+static struct kunit_case drm_sched_modify_sched_tests[] = {
+	KUNIT_CASE(drm_sched_test_modify_sched),
+	{}
+};
+
+static struct kunit_suite drm_sched_modify_sched = {
+	.name = "drm_sched_basic_modify_sched_tests",
+	.test_cases = drm_sched_modify_sched_tests,
+};
+
+static void drm_sched_test_credits(struct kunit *test)
+{
+	struct drm_mock_sched_entity *entity;
+	struct drm_mock_scheduler *sched;
+	struct drm_mock_sched_job *job[2];
+	bool done;
+	int i;
+
+	/*
+	 * Check that the configured credit limit is respected.
+	 */
+
+	sched = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+	sched->base.credit_limit = 1;
+
+	entity = drm_mock_sched_entity_new(test,
+					   DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+
+	job[0] = drm_mock_sched_job_new(test, entity);
+	job[1] = drm_mock_sched_job_new(test, entity);
+
+	drm_mock_sched_job_submit(job[0]);
+	drm_mock_sched_job_submit(job[1]);
+
+	done = drm_mock_sched_job_wait_scheduled(job[0], HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	done = drm_mock_sched_job_wait_scheduled(job[1], HZ);
+	KUNIT_ASSERT_FALSE(test, done);
+
+	i = drm_mock_sched_advance(sched, 1);
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	done = drm_mock_sched_job_wait_scheduled(job[1], HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	i = drm_mock_sched_advance(sched, 1);
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	done = drm_mock_sched_job_wait_finished(job[1], HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	drm_mock_sched_entity_free(entity);
+	drm_mock_sched_fini(sched);
+}
+
+static struct kunit_case drm_sched_credits_tests[] = {
+	KUNIT_CASE_SLOW(drm_sched_test_credits),
+	{}
+};
+
+static struct kunit_suite drm_sched_credits = {
+	.name = "drm_sched_basic_credits_tests",
+	.test_cases = drm_sched_credits_tests,
+};
+
+kunit_test_suites(&drm_sched_basic,
+		  &drm_sched_timeout,
+		  &drm_sched_cancel,
+		  &drm_sched_priority,
+		  &drm_sched_modify_sched,
+		  &drm_sched_credits);

linux-7.2.3/drivers/gpu/drm/scheduler/tests/tests_scheduler.c

@@ -0,0 +1,874 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2025 Valve Corporation */
+
+#include <linux/delay.h>
+#include <linux/kthread.h>
+#include <linux/ktime.h>
+#include <linux/math64.h>
+
+#include "sched_tests.h"
+
+/*
+ * DRM scheduler tests exercise load balancing decisions ie. entity selection
+ * logic.
+ */
+
+static int drm_sched_scheduler_init(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched;
+
+	sched = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+	sched->base.credit_limit = 1;
+
+	test->priv = sched;
+
+	return 0;
+}
+
+static int drm_sched_scheduler_init2(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched;
+
+	sched = drm_mock_sched_new(test, MAX_SCHEDULE_TIMEOUT);
+	sched->base.credit_limit = 2;
+
+	test->priv = sched;
+
+	return 0;
+}
+
+static void drm_sched_scheduler_exit(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+
+	drm_mock_sched_fini(sched);
+}
+
+static void drm_sched_scheduler_queue_overhead(struct kunit *test)
+{
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity;
+	const unsigned int job_us = 1000;
+	const unsigned int jobs = 1000;
+	const unsigned int total_us = jobs * job_us;
+	struct drm_mock_sched_job *job, *first;
+	ktime_t start, end;
+	bool done;
+	int i;
+
+	/*
+	 * Deep queue job at a time processing (single credit).
+	 *
+	 * This measures the overhead of picking and processing a job at a time
+	 * by comparing the ideal total "GPU" time of all submitted jobs versus
+	 * the time actually taken.
+	 */
+
+	KUNIT_ASSERT_EQ(test, sched->base.credit_limit, 1);
+
+	entity = drm_mock_sched_entity_new(test,
+					   DRM_SCHED_PRIORITY_NORMAL,
+					   sched);
+
+	for (i = 0; i <= jobs; i++) {
+		job = drm_mock_sched_job_new(test, entity);
+		if (i == 0)
+			first = job; /* Extra first job blocks the queue */
+		else
+			drm_mock_sched_job_set_duration_us(job, job_us);
+		drm_mock_sched_job_submit(job);
+	}
+
+	done = drm_mock_sched_job_wait_scheduled(first, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	start = ktime_get();
+	i = drm_mock_sched_advance(sched, 1); /* Release the queue */
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	/* Wait with a safe margin to avoid every failing. */
+	done = drm_mock_sched_job_wait_finished(job,
+						usecs_to_jiffies(total_us) * 5);
+	end = ktime_get();
+	KUNIT_ASSERT_TRUE(test, done);
+
+	pr_info("Expected %uus, actual %lldus\n",
+		total_us,
+		ktime_to_us(ktime_sub(end, start)));
+
+	drm_mock_sched_entity_free(entity);
+}
+
+static void drm_sched_scheduler_ping_pong(struct kunit *test)
+{
+	struct drm_mock_sched_job *job, *first, *prev = NULL;
+	struct drm_mock_scheduler *sched = test->priv;
+	struct drm_mock_sched_entity *entity[2];
+	const unsigned int job_us = 1000;
+	const unsigned int jobs = 1000;
+	const unsigned int total_us = jobs * job_us;
+	ktime_t start, end;
+	bool done;
+	int i;
+
+	/*
+	 * Two entitites in inter-dependency chain.
+	 *
+	 * This measures the overhead of picking and processing a job at a time,
+	 * where each job depends on the previous one from the diffferent
+	 * entity, by comparing the ideal total "GPU" time of all submitted jobs
+	 * versus the time actually taken.
+	 */
+
+	KUNIT_ASSERT_EQ(test, sched->base.credit_limit, 1);
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		entity[i] = drm_mock_sched_entity_new(test,
+						      DRM_SCHED_PRIORITY_NORMAL,
+						      sched);
+
+	for (i = 0; i <= jobs; i++) {
+		job = drm_mock_sched_job_new(test, entity[i & 1]);
+		if (i == 0)
+			first = job; /* Extra first job blocks the queue */
+		else
+			drm_mock_sched_job_set_duration_us(job, job_us);
+		if (prev)
+			drm_sched_job_add_dependency(&job->base,
+						     dma_fence_get(&prev->base.s_fence->finished));
+		drm_mock_sched_job_submit(job);
+		prev = job;
+	}
+
+	done = drm_mock_sched_job_wait_scheduled(first, HZ);
+	KUNIT_ASSERT_TRUE(test, done);
+
+	start = ktime_get();
+	i = drm_mock_sched_advance(sched, 1); /* Release the queue */
+	KUNIT_ASSERT_EQ(test, i, 1);
+
+	/* Wait with a safe margin to avoid every failing. */
+	done = drm_mock_sched_job_wait_finished(job,
+						usecs_to_jiffies(total_us) * 5);
+	end = ktime_get();
+	KUNIT_ASSERT_TRUE(test, done);
+
+	pr_info("Expected %uus, actual %lldus\n",
+		total_us,
+		ktime_to_us(ktime_sub(end, start)));
+
+	for (i = 0; i < ARRAY_SIZE(entity); i++)
+		drm_mock_sched_entity_free(entity[i]);
+}
+
+static struct kunit_case drm_sched_scheduler_overhead_tests[] = {
+	KUNIT_CASE_SLOW(drm_sched_scheduler_queue_overhead),
+	KUNIT_CASE_SLOW(drm_sched_scheduler_ping_pong),
+	{}
+};
+
+static struct kunit_suite drm_sched_scheduler_overhead = {
+	.name = "drm_sched_scheduler_overhead_tests",
+	.init = drm_sched_scheduler_init,
+	.exit = drm_sched_scheduler_exit,
+	.test_cases = drm_sched_scheduler_overhead_tests,
+};
+
+/*
+ * struct drm_sched_client_params - describe a workload emitted from a client
+ *
+ * A simulated client will create an entity with a scheduling @priority and emit
+ * jobs in a loop where each iteration will consist of:
+ *
+ * 1. Submit @job_cnt jobs, each with a set duration of @job_us.
+ * 2. If @sync is true wait for last submitted job to finish.
+ * 3. Sleep for @wait_us micro-seconds.
+ * 4. Repeat.
+ */
+struct drm_sched_client_params {
+	enum drm_sched_priority priority;
+	unsigned int job_cnt;
+	unsigned int job_us;
+	bool sync;
+	unsigned int wait_us;
+};
+
+struct drm_sched_test_params {
+	const char *description;
+	unsigned int num_clients;
+	struct drm_sched_client_params client[2];
+};
+
+static const struct drm_sched_test_params drm_sched_cases[] = {
+	{
+		.description = "Normal priority and normal priority",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+	},
+	{
+		.description = "Normal priority and low priority",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_LOW,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+	},
+	{
+		.description = "High priority and normal priority",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_HIGH,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+	},
+	{
+		.description = "High priority and low priority",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_HIGH,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_LOW,
+			.job_cnt = 1,
+			.job_us = 8000,
+			.wait_us = 0,
+			.sync = false,
+		},
+	},
+	{
+		.description = "50% and 50%",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 1500,
+			.wait_us = 1500,
+			.sync = true,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 2500,
+			.wait_us = 2500,
+			.sync = true,
+		},
+	},
+	{
+		.description = "50% and 50% low priority",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 1500,
+			.wait_us = 1500,
+			.sync = true,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_LOW,
+			.job_cnt = 1,
+			.job_us = 2500,
+			.wait_us = 2500,
+			.sync = true,
+		},
+	},
+	{
+		.description = "50% high priority and 50%",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_HIGH,
+			.job_cnt = 1,
+			.job_us = 1500,
+			.wait_us = 1500,
+			.sync = true,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 2500,
+			.wait_us = 2500,
+			.sync = true,
+		},
+	},
+	{
+		.description = "Low priority hog and interactive client",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_LOW,
+			.job_cnt = 3,
+			.job_us = 2500,
+			.wait_us = 500,
+			.sync = false,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 500,
+			.wait_us = 10000,
+			.sync = true,
+		},
+	},
+	{
+		.description = "Heavy rendering and interactive client",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 3,
+			.job_us = 2500,
+			.wait_us = 2500,
+			.sync = true,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 1000,
+			.wait_us = 9000,
+			.sync = true,
+		},
+	},
+	{
+		.description = "Very heavy rendering and interactive client",
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 4,
+			.job_us = 50000,
+			.wait_us = 1,
+			.sync = true,
+		},
+		.client[1] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 1,
+			.job_us = 1000,
+			.wait_us = 9000,
+			.sync = true,
+		},
+	},
+};
+
+static void
+drm_sched_desc(const struct drm_sched_test_params *params, char *desc)
+{
+	strscpy(desc, params->description, KUNIT_PARAM_DESC_SIZE);
+}
+
+KUNIT_ARRAY_PARAM(drm_sched_scheduler_two_clients,
+		  drm_sched_cases,
+		  drm_sched_desc);
+
+/*
+ * struct test_client_stats - track client stats
+ *
+ * For each client executing a simulated workload we track some timings for
+ * which we are interested in the minimum of all iterations (@min_us), maximum
+ * (@max_us) and the overall total for all iterations (@tot_us).
+ */
+struct test_client_stats {
+	unsigned int min_us;
+	unsigned int max_us;
+	unsigned long tot_us;
+};
+
+/*
+ * struct test_client - a simulated userspace client submitting scheduler work
+ *
+ * Each client executing a simulated workload is represented by one of these.
+ *
+ * Each of them instantiates a scheduling @entity and executes a workloads as
+ * defined in @params. Based on those @params the theoretical execution time of
+ * the client is calculated as @ideal_duration, while the actual wall time is
+ * tracked in @duration (calculated based on the @start and @end client time-
+ * stamps).
+ *
+ * Numerical @id is assigned to each for logging purposes.
+ *
+ * @worker and @work are used to provide an independent execution context from
+ * which scheduler jobs are submitted.
+ *
+ * During execution statistics on how long it took to submit and execute one
+ * iteration (whether or not synchronous) is kept in @cycle_time, while
+ * @latency_time tracks the @cycle_time minus the ideal duration of the one
+ * cycle.
+ *
+ * Once the client has completed the set number of iterations it will write the
+ * completion status into @done.
+ */
+struct test_client {
+	struct kunit	*test; /* Backpointer to the kunit test. */
+
+	struct drm_mock_sched_entity	*entity;
+	struct kthread_worker		*worker;
+	struct kthread_work		work;
+
+	struct drm_sched_client_params	params;
+
+	unsigned int	id;
+	ktime_t		duration;
+	ktime_t		ideal_duration;
+	unsigned int	cycles;
+	unsigned int	cycle;
+	ktime_t		start;
+	ktime_t		end;
+	bool		done;
+
+	struct test_client_stats	cycle_time;
+	struct test_client_stats	latency_time;
+};
+
+static void
+update_stats(struct test_client_stats *stats, unsigned int us)
+{
+	if (us > stats->max_us)
+		stats->max_us = us;
+	if (us < stats->min_us)
+		stats->min_us = us;
+	stats->tot_us += us;
+}
+
+static unsigned int
+get_stats_avg(struct test_client_stats *stats, unsigned int cycles)
+{
+	return div_u64(stats->tot_us, cycles);
+}
+
+static void drm_sched_client_work(struct kthread_work *work)
+{
+	struct test_client *client = container_of(work, typeof(*client), work);
+	const long sync_wait = MAX_SCHEDULE_TIMEOUT;
+	unsigned int cycle, work_us, period_us;
+	struct drm_mock_sched_job *job = NULL;
+
+	work_us = client->params.job_cnt * client->params.job_us;
+	period_us = work_us + client->params.wait_us;
+	client->cycles =
+		DIV_ROUND_UP((unsigned int)ktime_to_us(client->duration),
+			     period_us);
+	client->ideal_duration = us_to_ktime(client->cycles * period_us);
+
+	client->start = ktime_get();
+
+	for (cycle = 0; cycle < client->cycles; cycle++) {
+		ktime_t cycle_time;
+		unsigned int batch;
+		unsigned long us;
+
+		if (READ_ONCE(client->done))
+			break;
+
+		cycle_time = ktime_get();
+		for (batch = 0; batch < client->params.job_cnt; batch++) {
+			job = drm_mock_sched_job_new(client->test,
+						     client->entity);
+			drm_mock_sched_job_set_duration_us(job,
+							   client->params.job_us);
+			drm_mock_sched_job_submit(job);
+		}
+
+		if (client->params.sync)
+			drm_mock_sched_job_wait_finished(job, sync_wait);
+
+		cycle_time = ktime_sub(ktime_get(), cycle_time);
+		us = ktime_to_us(cycle_time);
+		update_stats(&client->cycle_time, us);
+		if (ktime_to_us(cycle_time) >= (long)work_us)
+			us = ktime_to_us(cycle_time) - work_us;
+		else if (WARN_ON_ONCE(client->params.sync)) /* GPU job took less than expected. */
+			us = 0;
+		update_stats(&client->latency_time, us);
+		WRITE_ONCE(client->cycle, cycle);
+
+		if (READ_ONCE(client->done))
+			break;
+
+		if (client->params.wait_us)
+			fsleep(client->params.wait_us);
+		else if (!client->params.sync)
+			cond_resched(); /* Do not hog the CPU if fully async. */
+	}
+
+	client->done = drm_mock_sched_job_wait_finished(job, sync_wait);
+	client->end = ktime_get();
+}
+
+static const char *prio_str(enum drm_sched_priority prio)
+{
+	switch (prio) {
+	case DRM_SCHED_PRIORITY_KERNEL:
+		return "kernel";
+	case DRM_SCHED_PRIORITY_LOW:
+		return "low";
+	case DRM_SCHED_PRIORITY_NORMAL:
+		return "normal";
+	case DRM_SCHED_PRIORITY_HIGH:
+		return "high";
+	default:
+		return "???";
+	}
+}
+
+static bool client_done(struct test_client *client)
+{
+	return READ_ONCE(client->done); /* READ_ONCE to document lockless read from a loop. */
+}
+
+static void drm_sched_scheduler_two_clients_test(struct kunit *test)
+{
+	const struct drm_sched_test_params *params = test->param_value;
+	struct drm_mock_scheduler *sched = test->priv;
+	struct test_client client[2] = { };
+	unsigned int prev_cycle[2] = { };
+	unsigned int i, j;
+	ktime_t start;
+
+	/*
+	 * Same job stream from two clients.
+	 */
+
+	for (i = 0; i < ARRAY_SIZE(client); i++)
+		client[i].entity =
+			drm_mock_sched_entity_new(test,
+						  params->client[i].priority,
+						  sched);
+
+	for (i = 0; i < ARRAY_SIZE(client); i++) {
+		client[i].test = test;
+		client[i].id = i;
+		client[i].duration = ms_to_ktime(1000);
+		client[i].params = params->client[i];
+		client[i].cycle_time.min_us = ~0U;
+		client[i].latency_time.min_us = ~0U;
+		client[i].worker =
+			kthread_create_worker(0, "%s-%u", __func__, i);
+		if (IS_ERR(client[i].worker)) {
+			for (j = 0; j < i; j++)
+				kthread_destroy_worker(client[j].worker);
+			KUNIT_FAIL(test, "Failed to create worker!\n");
+		}
+
+		kthread_init_work(&client[i].work, drm_sched_client_work);
+	}
+
+	for (i = 0; i < ARRAY_SIZE(client); i++)
+		kthread_queue_work(client[i].worker, &client[i].work);
+
+	/*
+	 * The clients (workers) can be a mix of async (deep submission queue),
+	 * sync (one job at a time), or something in between. Therefore it is
+	 * difficult to display a single metric representing their progress.
+	 *
+	 * Each struct drm_sched_client_params describes the actual submission
+	 * pattern which happens in the following steps:
+	 *  1. Submit N jobs
+	 *  2. Wait for last submitted job to finish
+	 *  3. Sleep for U micro-seconds
+	 *  4. Goto 1. for C cycles
+	 *
+	 * Where number of cycles is calculated to match the target client
+	 * duration from the respective struct drm_sched_test_params.
+	 *
+	 * To asses scheduling behaviour what we output for both clients is:
+	 *  - pct: Percentage progress of the jobs submitted
+	 *  - cps: "Cycles" per second (where one cycle is one complete
+	 *         iteration from the above)
+	 *  -  qd: Number of outstanding jobs in the client/entity
+	 */
+
+	pr_info(" [pct] - Job sumission progress\n"
+" [cps] - Cycles per second\n"
+"  [qd] - Number of outstanding jobs in the client/entity\n");
+	pr_info("%s:\n\t            pct1 cps1 qd1;  pct2 cps2 qd2\n",
+		params->description);
+	start = ktime_get();
+	while (!client_done(&client[0]) || !client_done(&client[1])) {
+		const unsigned int period_ms = 100;
+		const unsigned int frequency = 1000 / period_ms;
+		unsigned int pct[2], qd[2], cycle[2], cps[2];
+
+		for (i = 0; i < ARRAY_SIZE(client); i++) {
+			qd[i] = spsc_queue_count(&client[i].entity->base.job_queue);
+			cycle[i] = READ_ONCE(client[i].cycle);
+			cps[i] = DIV_ROUND_UP(100 * frequency *
+					      (cycle[i] - prev_cycle[i]),
+					      100);
+			if (client[i].cycles)
+				pct[i] = DIV_ROUND_UP(100 * (1 + cycle[i]),
+						      client[i].cycles);
+			else
+				pct[i] = 0;
+			prev_cycle[i] = cycle[i];
+		}
+
+		if (client_done(&client[0]))
+			pr_info("\t+%6lldms:               ; %3u %5u %4u\n",
+				ktime_to_ms(ktime_sub(ktime_get(), start)),
+				pct[1], cps[1], qd[1]);
+		else if (client_done(&client[1]))
+			pr_info("\t+%6lldms: %3u %5u %4u;\n",
+				ktime_to_ms(ktime_sub(ktime_get(), start)),
+				pct[0], cps[0], qd[0]);
+		else
+			pr_info("\t+%6lldms: %3u %5u %4u; %3u %5u %4u\n",
+				ktime_to_ms(ktime_sub(ktime_get(), start)),
+				pct[0], cps[0], qd[0],
+				pct[1], cps[1], qd[1]);
+
+		msleep(period_ms);
+	}
+
+	for (i = 0; i < ARRAY_SIZE(client); i++) {
+		kthread_flush_work(&client[i].work);
+		kthread_destroy_worker(client[i].worker);
+	}
+
+	for (i = 0; i < ARRAY_SIZE(client); i++)
+		KUNIT_ASSERT_TRUE(test, client[i].done);
+
+	for (i = 0; i < ARRAY_SIZE(client); i++) {
+		pr_info("    %u: prio=%s sync=%u elapsed_ms=%lldms (ideal_ms=%lldms) cycle_time(min,avg,max)=%u,%u,%u us latency_time(min,avg,max)=%u,%u,%u us",
+			i,
+			prio_str(params->client[i].priority),
+			params->client[i].sync,
+			ktime_to_ms(ktime_sub(client[i].end, client[i].start)),
+			ktime_to_ms(client[i].ideal_duration),
+			client[i].cycle_time.min_us,
+			get_stats_avg(&client[i].cycle_time, client[i].cycles),
+			client[i].cycle_time.max_us,
+			client[i].latency_time.min_us,
+			get_stats_avg(&client[i].latency_time, client[i].cycles),
+			client[i].latency_time.max_us);
+		drm_mock_sched_entity_free(client[i].entity);
+	}
+}
+
+static struct kunit_case drm_sched_scheduler_two_clients_tests[] = {
+	KUNIT_CASE_PARAM_ATTR(drm_sched_scheduler_two_clients_test,
+			      drm_sched_scheduler_two_clients_gen_params,
+			      { .speed = KUNIT_SPEED_SLOW }),
+	{}
+};
+
+static struct kunit_suite drm_sched_scheduler_two_clients1 = {
+	.name = "drm_sched_scheduler_two_clients_one_credit_tests",
+	.init = drm_sched_scheduler_init,
+	.exit = drm_sched_scheduler_exit,
+	.test_cases = drm_sched_scheduler_two_clients_tests,
+};
+
+static struct kunit_suite drm_sched_scheduler_two_clients2 = {
+	.name = "drm_sched_scheduler_two_clients_two_credits_tests",
+	.init = drm_sched_scheduler_init2,
+	.exit = drm_sched_scheduler_exit,
+	.test_cases = drm_sched_scheduler_two_clients_tests,
+};
+
+static const struct drm_sched_test_params drm_sched_many_cases[] = {
+	{
+		.description = "2 clients",
+		.num_clients = 2,
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 4,
+			.job_us = 1000,
+			.wait_us = 0,
+			.sync = true,
+		},
+	},
+	{
+		.description = "3 clients",
+		.num_clients = 3,
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 4,
+			.job_us = 1000,
+			.wait_us = 0,
+			.sync = true,
+		},
+	},
+	{
+		.description = "7 clients",
+		.num_clients = 7,
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 4,
+			.job_us = 1000,
+			.wait_us = 0,
+			.sync = true,
+		},
+	},
+	{
+		.description = "13 clients",
+		.num_clients = 13,
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 4,
+			.job_us = 1000,
+			.wait_us = 0,
+			.sync = true,
+		},
+	},
+	{
+		.description = "31 clients",
+		.num_clients = 31,
+		.client[0] = {
+			.priority = DRM_SCHED_PRIORITY_NORMAL,
+			.job_cnt = 2,
+			.job_us = 1000,
+			.wait_us = 0,
+			.sync = true,
+		},
+	},
+};
+
+KUNIT_ARRAY_PARAM(drm_sched_scheduler_many_clients,
+		  drm_sched_many_cases,
+		  drm_sched_desc);
+
+static void drm_sched_scheduler_many_clients_test(struct kunit *test)
+{
+	const struct drm_sched_test_params *params = test->param_value;
+	struct drm_mock_scheduler *sched = test->priv;
+	const unsigned int clients = params->num_clients;
+	unsigned int i, j, delta_total = 0, loops = 0;
+	struct test_client *client;
+	unsigned int *prev_cycle;
+	ktime_t start;
+	char *buf;
+
+	/*
+	 * Many clients with deep-ish async queues.
+	 */
+
+	buf = kunit_kmalloc(test, PAGE_SIZE, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, buf);
+	client = kunit_kcalloc(test, clients, sizeof(*client), GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, client);
+	prev_cycle = kunit_kcalloc(test, clients, sizeof(*prev_cycle),
+				   GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, prev_cycle);
+
+	for (i = 0; i < clients; i++)
+		client[i].entity =
+			drm_mock_sched_entity_new(test,
+						  DRM_SCHED_PRIORITY_NORMAL,
+						  sched);
+
+	for (i = 0; i < clients; i++) {
+		client[i].test = test;
+		client[i].id = i;
+		client[i].params = params->client[0];
+		client[i].duration = ms_to_ktime(1000 / clients);
+		client[i].cycle_time.min_us = ~0U;
+		client[i].latency_time.min_us = ~0U;
+		client[i].worker =
+			kthread_create_worker(0, "%s-%u", __func__, i);
+		if (IS_ERR(client[i].worker)) {
+			for (j = 0; j < i; j++)
+				kthread_destroy_worker(client[j].worker);
+			KUNIT_FAIL(test, "Failed to create worker!\n");
+		}
+
+		kthread_init_work(&client[i].work, drm_sched_client_work);
+	}
+
+	for (i = 0; i < clients; i++)
+		kthread_queue_work(client[i].worker, &client[i].work);
+
+	start = ktime_get();
+	pr_info("%u clients:\n\tt\t\tcycle:\t  min    avg    max : ...\n", clients);
+	for (;;) {
+		unsigned int min = ~0;
+		unsigned int max = 0;
+		unsigned int total = 0;
+		bool done = true;
+		char pbuf[16];
+
+		memset(buf, 0, PAGE_SIZE);
+		for (i = 0; i < clients; i++) {
+			unsigned int cycle, cycles;
+
+			/* Read current progress from the threaded worker. */
+			cycle = READ_ONCE(client[i].cycle);
+			cycles = READ_ONCE(client[i].cycles);
+
+			snprintf(pbuf, sizeof(pbuf), " %3d", cycle);
+			strncat(buf, pbuf, PAGE_SIZE);
+
+			total += cycle;
+			if (cycle < min)
+				min = cycle;
+			if (cycle > max)
+				max = cycle;
+
+			if (!min || (cycle + 1) < cycles)
+				done = false;
+		}
+
+		loops++;
+		delta_total += max - min;
+
+		pr_info("\t+%6lldms\t\t  %3u  %3u  %3u :%s\n",
+			ktime_to_ms(ktime_sub(ktime_get(), start)),
+			min, DIV_ROUND_UP(total, clients), max, buf);
+
+		if (done)
+			break;
+
+		msleep(100);
+	}
+
+	pr_info("    avg_max_min_delta(x100)=%u\n",
+		loops ? DIV_ROUND_UP(delta_total * 100, loops) : 0);
+
+	for (i = 0; i < clients; i++) {
+		kthread_flush_work(&client[i].work);
+		kthread_destroy_worker(client[i].worker);
+	}
+
+	for (i = 0; i < clients; i++)
+		drm_mock_sched_entity_free(client[i].entity);
+}
+
+static struct kunit_case drm_sched_scheduler_many_clients_tests[] = {
+	KUNIT_CASE_PARAM_ATTR(drm_sched_scheduler_many_clients_test,
+			      drm_sched_scheduler_many_clients_gen_params,
+			      { .speed = KUNIT_SPEED_SLOW }),
+	{}
+};
+
+static struct kunit_suite drm_sched_scheduler_many_clients = {
+	.name = "drm_sched_scheduler_many_clients_tests",
+	.init = drm_sched_scheduler_init2,
+	.exit = drm_sched_scheduler_exit,
+	.test_cases = drm_sched_scheduler_many_clients_tests,
+};
+
+kunit_test_suites(&drm_sched_scheduler_overhead,
+		  &drm_sched_scheduler_two_clients1,
+		  &drm_sched_scheduler_two_clients2,
+		  &drm_sched_scheduler_many_clients);