[PATCH v5 1/6] drm/sched: Add internal job peek/pop API
Philipp Stanner
phasta at mailbox.org
Wed Feb 19 07:47:36 UTC 2025
On Tue, 2025-02-18 at 14:17 -0800, Matthew Brost wrote:
> On Tue, Feb 18, 2025 at 06:26:15PM +0000, Tvrtko Ursulin wrote:
> >
> > On 18/02/2025 12:26, Philipp Stanner wrote:
> > > Thx for the updated version. Overlooked it, I was out on Friday.
> > > See
> > > below
> > >
> > > On Fri, 2025-02-14 at 10:19 +0000, Tvrtko Ursulin wrote:
> > > > Idea is to add helpers for peeking and popping jobs from
> > > > entities
> > > > with
> > > > the goal of decoupling the hidden assumption in the code that
> > > > queue_node
> > > > is the first element in struct drm_sched_job.
> > > >
> > > > That assumption usually comes in the form of:
> > > >
> > > > while ((job = to_drm_sched_job(spsc_queue_pop(&entity-
> > > > > job_queue))))
> > > >
> > > > Which breaks if the queue_node is re-positioned due
> > > > to_drm_sched_job
> > > > being implemented with a container_of.
> > > >
> > > > This also allows us to remove duplicate definitions of
> > > > to_drm_sched_job.
> > > >
> > > > Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin at igalia.com>
> > > > Cc: Christian König <christian.koenig at amd.com>
> > > > Cc: Danilo Krummrich <dakr at kernel.org>
> > > > Cc: Matthew Brost <matthew.brost at intel.com>
> > > > Cc: Philipp Stanner <phasta at kernel.org>
> > > > ---
> > > > drivers/gpu/drm/scheduler/sched_entity.c | 11 +++---
> > > > drivers/gpu/drm/scheduler/sched_internal.h | 46
> > > > ++++++++++++++++++++++
> > > > drivers/gpu/drm/scheduler/sched_main.c | 7 ++--
> > > > 3 files changed, 54 insertions(+), 10 deletions(-)
> > > > create mode 100644 drivers/gpu/drm/scheduler/sched_internal.h
> > > >
> > > > diff --git a/drivers/gpu/drm/scheduler/sched_entity.c
> > > > b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > index 69bcf0e99d57..a171f05ad761 100644
> > > > --- a/drivers/gpu/drm/scheduler/sched_entity.c
> > > > +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > @@ -28,11 +28,10 @@
> > > > #include <drm/drm_print.h>
> > > > #include <drm/gpu_scheduler.h>
> > > > +#include "sched_internal.h"
> > > > +
> > > > #include "gpu_scheduler_trace.h"
> > > > -#define to_drm_sched_job(sched_job) \
> > > > - container_of((sched_job), struct
> > > > drm_sched_job,
> > > > queue_node)
> > > > -
> > > > /**
> > > > * drm_sched_entity_init - Init a context entity used by
> > > > scheduler
> > > > when
> > > > * submit to HW ring.
> > > > @@ -255,7 +254,7 @@ static void drm_sched_entity_kill(struct
> > > > drm_sched_entity *entity)
> > > > /* The entity is guaranteed to not be used by the
> > > > scheduler
> > > > */
> > > > prev = rcu_dereference_check(entity->last_scheduled,
> > > > true);
> > > > dma_fence_get(prev);
> > > > - while ((job = to_drm_sched_job(spsc_queue_pop(&entity-
> > > > > job_queue)))) {
> > > > + while ((job = drm_sched_entity_queue_pop(entity))) {
> > > > struct drm_sched_fence *s_fence = job-
> > > > >s_fence;
> > > > dma_fence_get(&s_fence->finished);
> > > > @@ -477,7 +476,7 @@ struct drm_sched_job
> > > > *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
> > > > {
> > > > struct drm_sched_job *sched_job;
> > > > - sched_job = to_drm_sched_job(spsc_queue_peek(&entity-
> > > > > job_queue));
> > > > + sched_job = drm_sched_entity_queue_peek(entity);
> > > > if (!sched_job)
> > > > return NULL;
> > > > @@ -513,7 +512,7 @@ struct drm_sched_job
> > > > *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
> > > > if (drm_sched_policy == DRM_SCHED_POLICY_FIFO) {
> > > > struct drm_sched_job *next;
> > > > - next =
> > > > to_drm_sched_job(spsc_queue_peek(&entity-
> > > > > job_queue));
> > > > + next = drm_sched_entity_queue_peek(entity);
> > > > if (next) {
> > > > struct drm_sched_rq *rq;
> > > > diff --git a/drivers/gpu/drm/scheduler/sched_internal.h
> > > > b/drivers/gpu/drm/scheduler/sched_internal.h
> > > > new file mode 100644
> > > > index 000000000000..815d384845a3
> > > > --- /dev/null
> > > > +++ b/drivers/gpu/drm/scheduler/sched_internal.h
> > > > @@ -0,0 +1,46 @@
> > > > +#ifndef _DRM_GPU_SCHEDULER_INTERNAL_H_
> > > > +#define _DRM_GPU_SCHEDULER_INTERNAL_H_
> > >
> > > DRM maintainer tools complain about a potentially missing license
> > > identifier:
> > >
> > > -:80: WARNING:SPDX_LICENSE_TAG: Missing or malformed SPDX-
> > > License-Identifier tag in line 1
> > > #80: FILE: drivers/gpu/drm/scheduler/sched_internal.h:1:
> > >
> > > The other scheduler files don't have one, either. Still, it might
> > > be
> > > good to add one for new files. So, shall we make it GPL?
> >
> > Ha, good question. And it is actually good I forgot to do this for
> > this
> > series (I was doing for unit tests last week, I mean adding SPDX
> > lines)
> > because, as sched_internal.h will take parts of gpu_scheduler.h
> > which is not
> > explicitly GPL, nor the other scheduler source files, apart from
> > MODULE_LICENSE which is "GPL and additional rights", question
> > indeed is what
> > copyright blurb to put there. IANAL so not sure. Surely there is
> > some
> > established practice for cases like this one just I don't know what
> > it is.
> >
>
> I think GPL or MIT would be the preference given [1]. Also ofc IANAL.
>
> Matt
>
> [1] https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/3852
Well, I briefly synced with Dave – since all our scheduler files do
contain the full MIT license header (with AMD's copyright), but no
SPDX, we should probably go for MIT with all SPDX identifiers.
P.
>
> > Regards,
> >
> > Tvrtko
> >
> > > Rest of the series looks good.
> > >
> > > P.
> > >
> > > > +
> > > > +/**
> > > > + * drm_sched_entity_queue_pop - Low level helper for popping
> > > > queued
> > > > jobs
> > > > + *
> > > > + * @entity: scheduler entity
> > > > + *
> > > > + * Low level helper for popping queued jobs.
> > > > + *
> > > > + * Returns: The job dequeued or NULL.
> > > > + */
> > > > +static inline struct drm_sched_job *
> > > > +drm_sched_entity_queue_pop(struct drm_sched_entity *entity)
> > > > +{
> > > > + struct spsc_node *node;
> > > > +
> > > > + node = spsc_queue_pop(&entity->job_queue);
> > > > + if (!node)
> > > > + return NULL;
> > > > +
> > > > + return container_of(node, struct drm_sched_job,
> > > > queue_node);
> > > > +}
> > > > +
> > > > +/**
> > > > + * drm_sched_entity_queue_peek - Low level helper for peeking
> > > > at the
> > > > job queue
> > > > + *
> > > > + * @entity: scheduler entity
> > > > + *
> > > > + * Low level helper for peeking at the job queue
> > > > + *
> > > > + * Returns: The job at the head of the queue or NULL.
> > > > + */
> > > > +static inline struct drm_sched_job *
> > > > +drm_sched_entity_queue_peek(struct drm_sched_entity *entity)
> > > > +{
> > > > + struct spsc_node *node;
> > > > +
> > > > + node = spsc_queue_peek(&entity->job_queue);
> > > > + if (!node)
> > > > + return NULL;
> > > > +
> > > > + return container_of(node, struct drm_sched_job,
> > > > queue_node);
> > > > +}
> > > > +
> > > > +#endif
> > > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c
> > > > b/drivers/gpu/drm/scheduler/sched_main.c
> > > > index 8c36a59afb72..c634993f1346 100644
> > > > --- a/drivers/gpu/drm/scheduler/sched_main.c
> > > > +++ b/drivers/gpu/drm/scheduler/sched_main.c
> > > > @@ -78,6 +78,8 @@
> > > > #include <drm/gpu_scheduler.h>
> > > > #include <drm/spsc_queue.h>
> > > > +#include "sched_internal.h"
> > > > +
> > > > #define CREATE_TRACE_POINTS
> > > > #include "gpu_scheduler_trace.h"
> > > > @@ -87,9 +89,6 @@ static struct lockdep_map
> > > > drm_sched_lockdep_map = {
> > > > };
> > > > #endif
> > > > -#define to_drm_sched_job(sched_job) \
> > > > - container_of((sched_job), struct
> > > > drm_sched_job,
> > > > queue_node)
> > > > -
> > > > int drm_sched_policy = DRM_SCHED_POLICY_FIFO;
> > > > /**
> > > > @@ -123,7 +122,7 @@ static bool drm_sched_can_queue(struct
> > > > drm_gpu_scheduler *sched,
> > > > {
> > > > struct drm_sched_job *s_job;
> > > > - s_job = to_drm_sched_job(spsc_queue_peek(&entity-
> > > > > job_queue));
> > > > + s_job = drm_sched_entity_queue_peek(entity);
> > > > if (!s_job)
> > > > return false;
> > >
> >
More information about the Intel-xe
mailing list