drm_sched_entity Struct Reference

#include <gpu_scheduler.h>

Collaboration diagram for drm_sched_entity:

Public Attributes
struct list_head	list
spinlock_t	lock
struct drm_sched_rq *	rq
struct drm_sched_entity_stats *	stats
struct drm_gpu_scheduler **	sched_list
unsigned int	num_sched_list
enum drm_sched_priority	priority
enum drm_sched_priority	rq_priority
ktime_t	rr_ts
struct spsc_queue	job_queue
atomic_t	fence_seq
uint64_t	fence_context
struct dma_fence *	dependency
struct dma_fence_cb	cb
atomic_t *	guilty
struct dma_fence __rcu *	last_scheduled
struct task_struct *	last_user
bool	stopped
struct completion	entity_idle
ktime_t	oldest_job_waiting
struct rb_node	rb_tree_node

Detailed Description

struct drm_sched_entity - A wrapper around a job queue (typically attached to the DRM file_priv).

Entities will emit jobs in order to their corresponding hardware ring, and the scheduler will alternate between entities based on scheduling policy.

Member Data Documentation

cb

struct dma_fence_cb drm_sched_entity::cb

@cb:

Callback for the dependency fence above.

dependency

struct dma_fence* drm_sched_entity::dependency

@dependency:

The dependency fence of the job which is on the top of the job queue.

entity_idle

struct completion drm_sched_entity::entity_idle

@entity_idle:

Signals when entity is not in use, used to sequence entity cleanup in drm_sched_entity_fini().

fence_context

uint64_t drm_sched_entity::fence_context

@fence_context:

A unique context for all the fences which belong to this entity. The &drm_sched_fence.scheduled uses the fence_context but &drm_sched_fence.finished uses fence_context + 1.

fence_seq

atomic_t drm_sched_entity::fence_seq

@fence_seq:

A linearly increasing seqno incremented with each new &drm_sched_fence which is part of the entity.

FIXME: Callers of drm_sched_job_arm() need to ensure correct locking, this doesn't need to be atomic.

guilty

atomic_t* drm_sched_entity::guilty

@guilty:

Points to entities' guilty.

job_queue

struct spsc_queue drm_sched_entity::job_queue

@job_queue: the list of jobs of this entity.

last_scheduled

struct dma_fence __rcu* drm_sched_entity::last_scheduled

@last_scheduled:

Points to the finished fence of the last scheduled job. Only written by drm_sched_entity_pop_job(). Can be accessed locklessly from drm_sched_job_arm() if the queue is empty.

last_user

struct task_struct* drm_sched_entity::last_user

@last_user: last group leader pushing a job into the entity.

list

struct list_head drm_sched_entity::list

@list:

Used to append this struct to the list of entities in the runqueue @rq under &drm_sched_rq.entities.

Protected by &drm_sched_rq.lock of @rq.

lock

spinlock_t drm_sched_entity::lock

@lock:

Lock protecting the run-queue (@rq) to which this entity belongs, @priority, the list of schedulers (@sched_list, @num_sched_list) and the @rr_ts field.

num_sched_list

unsigned int drm_sched_entity::num_sched_list

@num_sched_list:

Number of drm_gpu_schedulers in the @sched_list.

oldest_job_waiting

ktime_t drm_sched_entity::oldest_job_waiting

@oldest_job_waiting:

Marks earliest job waiting in SW queue

priority

enum drm_sched_priority drm_sched_entity::priority

@priority:

Priority of the entity. This can be modified by calling drm_sched_entity_set_priority(). Protected by @lock.

rb_tree_node

struct rb_node drm_sched_entity::rb_tree_node

@rb_tree_node:

The node used to insert this entity into time based priority queue

rq

struct drm_sched_rq* drm_sched_entity::rq

@rq:

Runqueue on which this entity is currently scheduled.

FIXME: Locking is very unclear for this. Writers are protected by @lock, but readers are generally lockless and seem to just race with not even a READ_ONCE.

rq_priority

enum drm_sched_priority drm_sched_entity::rq_priority

@rq_priority: Run-queue priority

rr_ts

ktime_t drm_sched_entity::rr_ts

@rr_ts:

Fake timestamp of the last popped job from the entity.

sched_list

struct drm_gpu_scheduler** drm_sched_entity::sched_list

@sched_list:

A list of schedulers (struct drm_gpu_scheduler). Jobs from this entity can be scheduled on any scheduler on this list.

This can be modified by calling drm_sched_entity_modify_sched(). Locking is entirely up to the driver, see the above function for more details.

This will be set to NULL if &num_sched_list equals 1 and @rq has been set already.

FIXME: This means priority changes through drm_sched_entity_set_priority() will be lost henceforth in this case.

stats

struct drm_sched_entity_stats* drm_sched_entity::stats

@stats: Stats object reference held by the entity and jobs.

stopped

bool drm_sched_entity::stopped

@stopped:

Marks the enity as removed from rq and destined for termination. This is set by calling drm_sched_entity_flush() and by drm_sched_fini().

---

The documentation for this struct was generated from the following file:

• XenevaOS/BaseHdr/drm/gpu_scheduler.h