diff mbox series

[v3,4/4] drm/v3d: add multiple syncobjs support

Message ID ffd8b2e3dd2e0c686db441a0c0a4a0181ff85328.1633016479.git.mwen@igalia.com (mailing list archive)
State New, archived
Headers show
Series drm/v3d: add multiple in/out syncobjs support | expand

Commit Message

Melissa Wen Sept. 30, 2021, 4:19 p.m. UTC
Using the generic extension from the previous patch, a specific multisync
extension enables more than one in/out binary syncobj per job submission.
Arrays of syncobjs are set in struct drm_v3d_multisync, that also cares
of determining the stage for sync (wait deps) according to the job
queue.

v2:
- subclass the generic extension struct (Daniel)
- simplify adding dependency conditions to make understandable (Iago)

v3:
- fix conditions to consider single or multiples in/out_syncs (Iago)
- remove irrelevant comment (Iago)

Signed-off-by: Melissa Wen <mwen@igalia.com>
---
 drivers/gpu/drm/v3d/v3d_drv.c |   6 +-
 drivers/gpu/drm/v3d/v3d_drv.h |  24 +++--
 drivers/gpu/drm/v3d/v3d_gem.c | 185 ++++++++++++++++++++++++++++++----
 include/uapi/drm/v3d_drm.h    |  49 ++++++++-
 4 files changed, 232 insertions(+), 32 deletions(-)

Comments

Iago Toral Oct. 1, 2021, 7:29 a.m. UTC | #1
On Thu, 2021-09-30 at 17:19 +0100, Melissa Wen wrote:
> Using the generic extension from the previous patch, a specific
> multisync
> extension enables more than one in/out binary syncobj per job
> submission.
> Arrays of syncobjs are set in struct drm_v3d_multisync, that also
> cares
> of determining the stage for sync (wait deps) according to the job
> queue.
> 
> v2:
> - subclass the generic extension struct (Daniel)
> - simplify adding dependency conditions to make understandable (Iago)
> 
> v3:
> - fix conditions to consider single or multiples in/out_syncs (Iago)
> - remove irrelevant comment (Iago)
> 
> Signed-off-by: Melissa Wen <mwen@igalia.com>
> ---
>  drivers/gpu/drm/v3d/v3d_drv.c |   6 +-
>  drivers/gpu/drm/v3d/v3d_drv.h |  24 +++--
>  drivers/gpu/drm/v3d/v3d_gem.c | 185 ++++++++++++++++++++++++++++++
> ----
>  include/uapi/drm/v3d_drm.h    |  49 ++++++++-
>  4 files changed, 232 insertions(+), 32 deletions(-)
> 
> diff --git a/drivers/gpu/drm/v3d/v3d_drv.c
> b/drivers/gpu/drm/v3d/v3d_drv.c
> index 3d6b9bcce2f7..bd46396a1ae0 100644
> --- a/drivers/gpu/drm/v3d/v3d_drv.c
> +++ b/drivers/gpu/drm/v3d/v3d_drv.c
> @@ -96,6 +96,9 @@ static int v3d_get_param_ioctl(struct drm_device 

(...)

> 
> @@ -516,9 +536,11 @@ v3d_attach_fences_and_unlock_reservation(struct
> drm_file *file_priv,
>  					 struct v3d_job *job,
>  					 struct ww_acquire_ctx
> *acquire_ctx,
>  					 u32 out_sync,
> +					 struct v3d_submit_ext *se,
>  					 struct dma_fence *done_fence)
>  {
>  	struct drm_syncobj *sync_out;
> +	bool has_multisync = se && (se->flags & 

We always pass the 'se' pointer from a local variable allocated in the
stack of the caller so it is never going to be NULL.

I am happy to keep the NULL check if you want to protect against future
changes just in case, but if we do that then...

> DRM_V3D_EXT_ID_MULTI_SYNC);
>  	int i;
>  
>  	for (i = 0; i < job->bo_count; i++) {

(...)

> +static void
> +v3d_put_multisync_post_deps(struct v3d_submit_ext *se)
> +{
> +	unsigned int i;
> +
> +	if (!se->out_sync_count)

...we should also check for NULL here for consistency.

Also, I think there is another problem in the code: we always call
v3d_job_cleanup for failed jobs, but only call v3d_job_put for
successful jobs. However, reading the docs for drm_sched_job_init:

"Drivers must make sure drm_sched_job_cleanup() if this function
returns successfully, even when @job is aborted before
drm_sched_job_arm() is called."

So my understanding is that we should call v3d_job_cleanup instead of
v3d_job_put for successful jobs or we would be leaking sched resources
on every successful job, no?

Iago

> +		return;
> +
> +	for (i = 0; i < se->out_sync_count; i++)
> +		drm_syncobj_put(se->out_syncs[i].syncobj);
> +	kvfree(se->out_syncs);
> +}
> +
> +static int
> +v3d_get_multisync_post_deps(struct drm_file *file_priv,
> +			    struct v3d_submit_ext *se,
> +			    u32 count, u64 handles)
> +{
> +	struct drm_v3d_sem __user *post_deps;
> +	int i, ret;
> +
> +	if (!count)
> +		return 0;
> +
> +	se->out_syncs = (struct v3d_submit_outsync *)
> +			kvmalloc_array(count,
> +				       sizeof(struct
> v3d_submit_outsync),
> +				       GFP_KERNEL);
> +	if (!se->out_syncs)
> +		return -ENOMEM;
> +
> +	post_deps = u64_to_user_ptr(handles);
> +
> +	for (i = 0; i < count; i++) {
> +		struct drm_v3d_sem out;
> +
> +		ret = copy_from_user(&out, post_deps++, sizeof(out));
> +		if (ret) {
> +			DRM_DEBUG("Failed to copy post dep handles\n");
> +			goto fail;
> +		}
> +
> +		se->out_syncs[i].syncobj = drm_syncobj_find(file_priv,
> +							    out.handle)
> ;
> +		if (!se->out_syncs[i].syncobj) {
> +			ret = -EINVAL;
> +			goto fail;
> +		}
>  	}
> +	se->out_sync_count = count;
> +
> +	return 0;
> +
> +fail:
> +	for (i--; i >= 0; i--)
> +		drm_syncobj_put(se->out_syncs[i].syncobj);
> +	kvfree(se->out_syncs);
> +
> +	return ret;
> +}
> +
> +/* Get data for multiple binary semaphores synchronization. Parse
> syncobj
> + * to be signaled when job completes (out_sync).
> + */
> +static int
> +v3d_get_multisync_submit_deps(struct drm_file *file_priv,
> +			      struct drm_v3d_extension __user *ext,
> +			      void *data)
> +{
> +	struct drm_v3d_multi_sync multisync;
> +	struct v3d_submit_ext *se = data;
> +	int ret;
> +
> +	ret = copy_from_user(&multisync, ext, sizeof(multisync));
> +	if (ret)
> +		return ret;
> +
> +	if (multisync.pad)
> +		return -EINVAL;
> +
> +	ret = v3d_get_multisync_post_deps(file_priv, data,
> multisync.out_sync_count,
> +					  multisync.out_syncs);
> +	if (ret)
> +		return ret;
> +
> +	se->in_sync_count = multisync.in_sync_count;
> +	se->in_syncs = multisync.in_syncs;
> +	se->flags |= DRM_V3D_EXT_ID_MULTI_SYNC;
> +	se->wait_stage = multisync.wait_stage;
> +
> +	return 0;
>  }
>  
>  /* Whenever userspace sets ioctl extensions, v3d_get_extensions
> parses data
>   * according to the extension id (name).
>   */
>  static int
> -v3d_get_extensions(struct drm_file *file_priv, u64 ext_handles)
> +v3d_get_extensions(struct drm_file *file_priv,
> +		   u64 ext_handles,
> +		   void *data)
>  {
>  	struct drm_v3d_extension __user *user_ext;
> +	int ret;
>  
>  	user_ext = u64_to_user_ptr(ext_handles);
>  	while (user_ext) {
> @@ -555,7 +687,11 @@ v3d_get_extensions(struct drm_file *file_priv,
> u64 ext_handles)
>  		}
>  
>  		switch (ext.id) {
> -		case 0:
> +		case DRM_V3D_EXT_ID_MULTI_SYNC:
> +			ret = v3d_get_multisync_submit_deps(file_priv,
> user_ext, data);
> +			if (ret)
> +				return ret;
> +			break;
>  		default:
>  			DRM_DEBUG_DRIVER("Unknown extension id: %d\n",
> ext.id);
>  			return -EINVAL;
> @@ -586,6 +722,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  	struct v3d_dev *v3d = to_v3d_dev(dev);
>  	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
>  	struct drm_v3d_submit_cl *args = data;
> +	struct v3d_submit_ext se = {0};
>  	struct v3d_bin_job *bin = NULL;
>  	struct v3d_render_job *render = NULL;
>  	struct v3d_job *clean_job = NULL;
> @@ -606,7 +743,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  	}
>  
>  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> -		ret = v3d_get_extensions(file_priv, args->extensions);
> +		ret = v3d_get_extensions(file_priv, args->extensions,
> &se);
>  		if (ret) {
>  			DRM_DEBUG("Failed to get extensions.\n");
>  			return ret;
> @@ -614,7 +751,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  	}
>  
>  	ret = v3d_job_init(v3d, file_priv, (void *)&render,
> sizeof(*render),
> -			   v3d_render_job_free, args->in_sync_rcl,
> V3D_RENDER);
> +			   v3d_render_job_free, args->in_sync_rcl, &se,
> V3D_RENDER);
>  	if (ret)
>  		goto fail;
>  
> @@ -624,7 +761,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  
>  	if (args->bcl_start != args->bcl_end) {
>  		ret = v3d_job_init(v3d, file_priv, (void *)&bin,
> sizeof(*bin),
> -				   v3d_job_free, args->in_sync_bcl,
> V3D_BIN);
> +				   v3d_job_free, args->in_sync_bcl,
> &se, V3D_BIN);
>  		if (ret)
>  			goto fail;
>  
> @@ -638,7 +775,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  
>  	if (args->flags & DRM_V3D_SUBMIT_CL_FLUSH_CACHE) {
>  		ret = v3d_job_init(v3d, file_priv, (void *)&clean_job,
> sizeof(*clean_job),
> -				   v3d_job_free, 0, V3D_CACHE_CLEAN);
> +				   v3d_job_free, 0, 0,
> V3D_CACHE_CLEAN);
>  		if (ret)
>  			goto fail;
>  
> @@ -698,6 +835,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  						 last_job,
>  						 &acquire_ctx,
>  						 args->out_sync,
> +						 &se,
>  						 last_job->done_fence);
>  
>  	if (bin)
> @@ -716,6 +854,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> *data,
>  	v3d_job_cleanup((void *)bin);
>  	v3d_job_cleanup((void *)render);
>  	v3d_job_cleanup(clean_job);
> +	v3d_put_multisync_post_deps(&se);
>  
>  	return ret;
>  }
> @@ -735,6 +874,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> *data,
>  {
>  	struct v3d_dev *v3d = to_v3d_dev(dev);
>  	struct drm_v3d_submit_tfu *args = data;
> +	struct v3d_submit_ext se = {0};
>  	struct v3d_tfu_job *job = NULL;
>  	struct ww_acquire_ctx acquire_ctx;
>  	int ret = 0;
> @@ -747,7 +887,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> *data,
>  	}
>  
>  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> -		ret = v3d_get_extensions(file_priv, args->extensions);
> +		ret = v3d_get_extensions(file_priv, args->extensions,
> &se);
>  		if (ret) {
>  			DRM_DEBUG("Failed to get extensions.\n");
>  			return ret;
> @@ -755,7 +895,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> *data,
>  	}
>  
>  	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
> -			   v3d_job_free, args->in_sync, V3D_TFU);
> +			   v3d_job_free, args->in_sync, &se, V3D_TFU);
>  	if (ret)
>  		goto fail;
>  
> @@ -803,6 +943,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> *data,
>  	v3d_attach_fences_and_unlock_reservation(file_priv,
>  						 &job->base,
> &acquire_ctx,
>  						 args->out_sync,
> +						 &se,
>  						 job->base.done_fence);
>  
>  	v3d_job_put(&job->base);
> @@ -811,6 +952,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> *data,
>  
>  fail:
>  	v3d_job_cleanup((void *)job);
> +	v3d_put_multisync_post_deps(&se);
>  
>  	return ret;
>  }
> @@ -831,6 +973,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev, void
> *data,
>  	struct v3d_dev *v3d = to_v3d_dev(dev);
>  	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
>  	struct drm_v3d_submit_csd *args = data;
> +	struct v3d_submit_ext se = {0};
>  	struct v3d_csd_job *job = NULL;
>  	struct v3d_job *clean_job = NULL;
>  	struct ww_acquire_ctx acquire_ctx;
> @@ -852,7 +995,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev, void
> *data,
>  	}
>  
>  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> -		ret = v3d_get_extensions(file_priv, args->extensions);
> +		ret = v3d_get_extensions(file_priv, args->extensions,
> &se);
>  		if (ret) {
>  			DRM_DEBUG("Failed to get extensions.\n");
>  			return ret;
> @@ -860,12 +1003,12 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> void *data,
>  	}
>  
>  	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
> -			   v3d_job_free, args->in_sync, V3D_CSD);
> +			   v3d_job_free, args->in_sync, &se, V3D_CSD);
>  	if (ret)
>  		goto fail;
>  
>  	ret = v3d_job_init(v3d, file_priv, (void *)&clean_job,
> sizeof(*clean_job),
> -			   v3d_job_free, 0, V3D_CACHE_CLEAN);
> +			   v3d_job_free, 0, 0, V3D_CACHE_CLEAN);
>  	if (ret)
>  		goto fail;
>  
> @@ -904,6 +1047,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> void *data,
>  						 clean_job,
>  						 &acquire_ctx,
>  						 args->out_sync,
> +						 &se,
>  						 clean_job-
> >done_fence);
>  
>  	v3d_job_put(&job->base);
> @@ -918,6 +1062,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> void *data,
>  fail:
>  	v3d_job_cleanup((void *)job);
>  	v3d_job_cleanup(clean_job);
> +	v3d_put_multisync_post_deps(&se);
>  
>  	return ret;
>  }
> diff --git a/include/uapi/drm/v3d_drm.h b/include/uapi/drm/v3d_drm.h
> index 55b443ca6c0b..3dfc0af8756a 100644
> --- a/include/uapi/drm/v3d_drm.h
> +++ b/include/uapi/drm/v3d_drm.h
> @@ -73,6 +73,53 @@ struct drm_v3d_extension {
>  	__u32 flags; /* mbz */
>  };
>  
> +/* struct drm_v3d_sem - wait/signal semaphore
> + *
> + * If binary semaphore, it only takes syncobj handle and ignores
> flags and
> + * point fields. Point is defined for timeline syncobj feature.
> + */
> +struct drm_v3d_sem {
> +	__u32 handle; /* syncobj */
> +	/* rsv below, for future uses */
> +	__u32 flags;
> +	__u64 point;  /* for timeline sem support */
> +	__u64 mbz[2]; /* must be zero, rsv */
> +};
> +
> +/* Enum for each of the V3D queues. */
> +enum v3d_queue {
> +	V3D_BIN,
> +	V3D_RENDER,
> +	V3D_TFU,
> +	V3D_CSD,
> +	V3D_CACHE_CLEAN,
> +};
> +
> +/**
> + * struct drm_v3d_multi_sync - ioctl extension to add support
> multiples
> + * syncobjs for commands submission.
> + *
> + * When an extension of DRM_V3D_EXT_ID_MULTI_SYNC id is defined, it
> points to
> + * this extension to define wait and signal dependencies, instead of
> single
> + * in/out sync entries on submitting commands. The field flags is
> used to
> + * determine the stage to set wait dependencies.
> + */
> +struct drm_v3d_multi_sync {
> +	struct drm_v3d_extension base;
> +	/* Array of wait and signal semaphores */
> +	__u64 in_syncs;
> +	__u64 out_syncs;
> +
> +	/* Number of entries */
> +	__u32 in_sync_count;
> +	__u32 out_sync_count;
> +
> +	/* set the stage (v3d_queue) to sync */
> +	__u32 wait_stage;
> +
> +	__u32 pad; /* mbz */
> +};
> +
>  /**
>   * struct drm_v3d_submit_cl - ioctl argument for submitting commands
> to the 3D
>   * engine.
> @@ -228,6 +275,7 @@ enum drm_v3d_param {
>  	DRM_V3D_PARAM_SUPPORTS_CSD,
>  	DRM_V3D_PARAM_SUPPORTS_CACHE_FLUSH,
>  	DRM_V3D_PARAM_SUPPORTS_PERFMON,
> +	DRM_V3D_PARAM_SUPPORTS_MULTISYNC_EXT,
>  };
>  
>  struct drm_v3d_get_param {
> @@ -271,7 +319,6 @@ struct drm_v3d_submit_tfu {
>  
>  	/* Pointer to an array of ioctl extensions*/
>  	__u64 extensions;
> -
>  };
>  
>  /* Submits a compute shader for dispatch.  This job will block on
> any
Melissa Wen Oct. 1, 2021, 8:37 a.m. UTC | #2
On 10/01, Iago Toral wrote:
> On Thu, 2021-09-30 at 17:19 +0100, Melissa Wen wrote:
> > Using the generic extension from the previous patch, a specific
> > multisync
> > extension enables more than one in/out binary syncobj per job
> > submission.
> > Arrays of syncobjs are set in struct drm_v3d_multisync, that also
> > cares
> > of determining the stage for sync (wait deps) according to the job
> > queue.
> > 
> > v2:
> > - subclass the generic extension struct (Daniel)
> > - simplify adding dependency conditions to make understandable (Iago)
> > 
> > v3:
> > - fix conditions to consider single or multiples in/out_syncs (Iago)
> > - remove irrelevant comment (Iago)
> > 
> > Signed-off-by: Melissa Wen <mwen@igalia.com>
> > ---
> >  drivers/gpu/drm/v3d/v3d_drv.c |   6 +-
> >  drivers/gpu/drm/v3d/v3d_drv.h |  24 +++--
> >  drivers/gpu/drm/v3d/v3d_gem.c | 185 ++++++++++++++++++++++++++++++
> > ----
> >  include/uapi/drm/v3d_drm.h    |  49 ++++++++-
> >  4 files changed, 232 insertions(+), 32 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/v3d/v3d_drv.c
> > b/drivers/gpu/drm/v3d/v3d_drv.c
> > index 3d6b9bcce2f7..bd46396a1ae0 100644
> > --- a/drivers/gpu/drm/v3d/v3d_drv.c
> > +++ b/drivers/gpu/drm/v3d/v3d_drv.c
> > @@ -96,6 +96,9 @@ static int v3d_get_param_ioctl(struct drm_device 
> 
> (...)
> 
> > 
> > @@ -516,9 +536,11 @@ v3d_attach_fences_and_unlock_reservation(struct
> > drm_file *file_priv,
> >  					 struct v3d_job *job,
> >  					 struct ww_acquire_ctx
> > *acquire_ctx,
> >  					 u32 out_sync,
> > +					 struct v3d_submit_ext *se,
> >  					 struct dma_fence *done_fence)
> >  {
> >  	struct drm_syncobj *sync_out;
> > +	bool has_multisync = se && (se->flags & 
> 
> We always pass the 'se' pointer from a local variable allocated in the
> stack of the caller so it is never going to be NULL.
> 
> I am happy to keep the NULL check if you want to protect against future
> changes just in case, but if we do that then...
> 
> > DRM_V3D_EXT_ID_MULTI_SYNC);
> >  	int i;
> >  
> >  	for (i = 0; i < job->bo_count; i++) {
> 
> (...)
> 
> > +static void
> > +v3d_put_multisync_post_deps(struct v3d_submit_ext *se)
> > +{
> > +	unsigned int i;
> > +
> > +	if (!se->out_sync_count)
> 
> ...we should also check for NULL here for consistency.
yes, consistency makes sense here.
> 
> Also, I think there is another problem in the code: we always call
> v3d_job_cleanup for failed jobs, but only call v3d_job_put for
> successful jobs. However, reading the docs for drm_sched_job_init:
> 
> "Drivers must make sure drm_sched_job_cleanup() if this function
> returns successfully, even when @job is aborted before
> drm_sched_job_arm() is called."
> 
> So my understanding is that we should call v3d_job_cleanup instead of
> v3d_job_put for successful jobs or we would be leaking sched resources
> on every successful job, no?

When job_init is successful, v3d_job_cleanup is called by scheduler when
job is completed. drm_sched_job_cleanup describes how it works after
drm_sched_job_arm:

" After that point of no return @job is committed to be executed by the
scheduler, and this function should be called from the
&drm_sched_backend_ops.free_job callback."

On v3d_sched.c, .free_job points to v3d_sched_job_free(), which in turn
calls v3d_job_cleanup() (and then drm_sched_job_cleanup). So, it looks
ok.

Also, we can say that the very last v3d_job_put() is in charge to
decrement refcount initialized (set 1) in v3d_job_init(); while the
v3d_job_cleanup() from v3d_sched_job_free() callback decrements refcount
that was incremented when v3d_job_push() pushed the job to the
scheduler.

So, refcount pairs seem consistent, I mean, get and put. And about
drm_sched_job_cleanup, it is explicitly called when job_init fails or
after that by the scheduler.

Thanks,

Melissa

> 
> Iago
> 
> > +		return;
> > +
> > +	for (i = 0; i < se->out_sync_count; i++)
> > +		drm_syncobj_put(se->out_syncs[i].syncobj);
> > +	kvfree(se->out_syncs);
> > +}
> > +
> > +static int
> > +v3d_get_multisync_post_deps(struct drm_file *file_priv,
> > +			    struct v3d_submit_ext *se,
> > +			    u32 count, u64 handles)
> > +{
> > +	struct drm_v3d_sem __user *post_deps;
> > +	int i, ret;
> > +
> > +	if (!count)
> > +		return 0;
> > +
> > +	se->out_syncs = (struct v3d_submit_outsync *)
> > +			kvmalloc_array(count,
> > +				       sizeof(struct
> > v3d_submit_outsync),
> > +				       GFP_KERNEL);
> > +	if (!se->out_syncs)
> > +		return -ENOMEM;
> > +
> > +	post_deps = u64_to_user_ptr(handles);
> > +
> > +	for (i = 0; i < count; i++) {
> > +		struct drm_v3d_sem out;
> > +
> > +		ret = copy_from_user(&out, post_deps++, sizeof(out));
> > +		if (ret) {
> > +			DRM_DEBUG("Failed to copy post dep handles\n");
> > +			goto fail;
> > +		}
> > +
> > +		se->out_syncs[i].syncobj = drm_syncobj_find(file_priv,
> > +							    out.handle)
> > ;
> > +		if (!se->out_syncs[i].syncobj) {
> > +			ret = -EINVAL;
> > +			goto fail;
> > +		}
> >  	}
> > +	se->out_sync_count = count;
> > +
> > +	return 0;
> > +
> > +fail:
> > +	for (i--; i >= 0; i--)
> > +		drm_syncobj_put(se->out_syncs[i].syncobj);
> > +	kvfree(se->out_syncs);
> > +
> > +	return ret;
> > +}
> > +
> > +/* Get data for multiple binary semaphores synchronization. Parse
> > syncobj
> > + * to be signaled when job completes (out_sync).
> > + */
> > +static int
> > +v3d_get_multisync_submit_deps(struct drm_file *file_priv,
> > +			      struct drm_v3d_extension __user *ext,
> > +			      void *data)
> > +{
> > +	struct drm_v3d_multi_sync multisync;
> > +	struct v3d_submit_ext *se = data;
> > +	int ret;
> > +
> > +	ret = copy_from_user(&multisync, ext, sizeof(multisync));
> > +	if (ret)
> > +		return ret;
> > +
> > +	if (multisync.pad)
> > +		return -EINVAL;
> > +
> > +	ret = v3d_get_multisync_post_deps(file_priv, data,
> > multisync.out_sync_count,
> > +					  multisync.out_syncs);
> > +	if (ret)
> > +		return ret;
> > +
> > +	se->in_sync_count = multisync.in_sync_count;
> > +	se->in_syncs = multisync.in_syncs;
> > +	se->flags |= DRM_V3D_EXT_ID_MULTI_SYNC;
> > +	se->wait_stage = multisync.wait_stage;
> > +
> > +	return 0;
> >  }
> >  
> >  /* Whenever userspace sets ioctl extensions, v3d_get_extensions
> > parses data
> >   * according to the extension id (name).
> >   */
> >  static int
> > -v3d_get_extensions(struct drm_file *file_priv, u64 ext_handles)
> > +v3d_get_extensions(struct drm_file *file_priv,
> > +		   u64 ext_handles,
> > +		   void *data)
> >  {
> >  	struct drm_v3d_extension __user *user_ext;
> > +	int ret;
> >  
> >  	user_ext = u64_to_user_ptr(ext_handles);
> >  	while (user_ext) {
> > @@ -555,7 +687,11 @@ v3d_get_extensions(struct drm_file *file_priv,
> > u64 ext_handles)
> >  		}
> >  
> >  		switch (ext.id) {
> > -		case 0:
> > +		case DRM_V3D_EXT_ID_MULTI_SYNC:
> > +			ret = v3d_get_multisync_submit_deps(file_priv,
> > user_ext, data);
> > +			if (ret)
> > +				return ret;
> > +			break;
> >  		default:
> >  			DRM_DEBUG_DRIVER("Unknown extension id: %d\n",
> > ext.id);
> >  			return -EINVAL;
> > @@ -586,6 +722,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  	struct v3d_dev *v3d = to_v3d_dev(dev);
> >  	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
> >  	struct drm_v3d_submit_cl *args = data;
> > +	struct v3d_submit_ext se = {0};
> >  	struct v3d_bin_job *bin = NULL;
> >  	struct v3d_render_job *render = NULL;
> >  	struct v3d_job *clean_job = NULL;
> > @@ -606,7 +743,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  	}
> >  
> >  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> > -		ret = v3d_get_extensions(file_priv, args->extensions);
> > +		ret = v3d_get_extensions(file_priv, args->extensions,
> > &se);
> >  		if (ret) {
> >  			DRM_DEBUG("Failed to get extensions.\n");
> >  			return ret;
> > @@ -614,7 +751,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  	}
> >  
> >  	ret = v3d_job_init(v3d, file_priv, (void *)&render,
> > sizeof(*render),
> > -			   v3d_render_job_free, args->in_sync_rcl,
> > V3D_RENDER);
> > +			   v3d_render_job_free, args->in_sync_rcl, &se,
> > V3D_RENDER);
> >  	if (ret)
> >  		goto fail;
> >  
> > @@ -624,7 +761,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  
> >  	if (args->bcl_start != args->bcl_end) {
> >  		ret = v3d_job_init(v3d, file_priv, (void *)&bin,
> > sizeof(*bin),
> > -				   v3d_job_free, args->in_sync_bcl,
> > V3D_BIN);
> > +				   v3d_job_free, args->in_sync_bcl,
> > &se, V3D_BIN);
> >  		if (ret)
> >  			goto fail;
> >  
> > @@ -638,7 +775,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  
> >  	if (args->flags & DRM_V3D_SUBMIT_CL_FLUSH_CACHE) {
> >  		ret = v3d_job_init(v3d, file_priv, (void *)&clean_job,
> > sizeof(*clean_job),
> > -				   v3d_job_free, 0, V3D_CACHE_CLEAN);
> > +				   v3d_job_free, 0, 0,
> > V3D_CACHE_CLEAN);
> >  		if (ret)
> >  			goto fail;
> >  
> > @@ -698,6 +835,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  						 last_job,
> >  						 &acquire_ctx,
> >  						 args->out_sync,
> > +						 &se,
> >  						 last_job->done_fence);
> >  
> >  	if (bin)
> > @@ -716,6 +854,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
> > *data,
> >  	v3d_job_cleanup((void *)bin);
> >  	v3d_job_cleanup((void *)render);
> >  	v3d_job_cleanup(clean_job);
> > +	v3d_put_multisync_post_deps(&se);
> >  
> >  	return ret;
> >  }
> > @@ -735,6 +874,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> > *data,
> >  {
> >  	struct v3d_dev *v3d = to_v3d_dev(dev);
> >  	struct drm_v3d_submit_tfu *args = data;
> > +	struct v3d_submit_ext se = {0};
> >  	struct v3d_tfu_job *job = NULL;
> >  	struct ww_acquire_ctx acquire_ctx;
> >  	int ret = 0;
> > @@ -747,7 +887,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> > *data,
> >  	}
> >  
> >  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> > -		ret = v3d_get_extensions(file_priv, args->extensions);
> > +		ret = v3d_get_extensions(file_priv, args->extensions,
> > &se);
> >  		if (ret) {
> >  			DRM_DEBUG("Failed to get extensions.\n");
> >  			return ret;
> > @@ -755,7 +895,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> > *data,
> >  	}
> >  
> >  	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
> > -			   v3d_job_free, args->in_sync, V3D_TFU);
> > +			   v3d_job_free, args->in_sync, &se, V3D_TFU);
> >  	if (ret)
> >  		goto fail;
> >  
> > @@ -803,6 +943,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> > *data,
> >  	v3d_attach_fences_and_unlock_reservation(file_priv,
> >  						 &job->base,
> > &acquire_ctx,
> >  						 args->out_sync,
> > +						 &se,
> >  						 job->base.done_fence);
> >  
> >  	v3d_job_put(&job->base);
> > @@ -811,6 +952,7 @@ v3d_submit_tfu_ioctl(struct drm_device *dev, void
> > *data,
> >  
> >  fail:
> >  	v3d_job_cleanup((void *)job);
> > +	v3d_put_multisync_post_deps(&se);
> >  
> >  	return ret;
> >  }
> > @@ -831,6 +973,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev, void
> > *data,
> >  	struct v3d_dev *v3d = to_v3d_dev(dev);
> >  	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
> >  	struct drm_v3d_submit_csd *args = data;
> > +	struct v3d_submit_ext se = {0};
> >  	struct v3d_csd_job *job = NULL;
> >  	struct v3d_job *clean_job = NULL;
> >  	struct ww_acquire_ctx acquire_ctx;
> > @@ -852,7 +995,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev, void
> > *data,
> >  	}
> >  
> >  	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
> > -		ret = v3d_get_extensions(file_priv, args->extensions);
> > +		ret = v3d_get_extensions(file_priv, args->extensions,
> > &se);
> >  		if (ret) {
> >  			DRM_DEBUG("Failed to get extensions.\n");
> >  			return ret;
> > @@ -860,12 +1003,12 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> > void *data,
> >  	}
> >  
> >  	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
> > -			   v3d_job_free, args->in_sync, V3D_CSD);
> > +			   v3d_job_free, args->in_sync, &se, V3D_CSD);
> >  	if (ret)
> >  		goto fail;
> >  
> >  	ret = v3d_job_init(v3d, file_priv, (void *)&clean_job,
> > sizeof(*clean_job),
> > -			   v3d_job_free, 0, V3D_CACHE_CLEAN);
> > +			   v3d_job_free, 0, 0, V3D_CACHE_CLEAN);
> >  	if (ret)
> >  		goto fail;
> >  
> > @@ -904,6 +1047,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> > void *data,
> >  						 clean_job,
> >  						 &acquire_ctx,
> >  						 args->out_sync,
> > +						 &se,
> >  						 clean_job-
> > >done_fence);
> >  
> >  	v3d_job_put(&job->base);
> > @@ -918,6 +1062,7 @@ v3d_submit_csd_ioctl(struct drm_device *dev,
> > void *data,
> >  fail:
> >  	v3d_job_cleanup((void *)job);
> >  	v3d_job_cleanup(clean_job);
> > +	v3d_put_multisync_post_deps(&se);
> >  
> >  	return ret;
> >  }
> > diff --git a/include/uapi/drm/v3d_drm.h b/include/uapi/drm/v3d_drm.h
> > index 55b443ca6c0b..3dfc0af8756a 100644
> > --- a/include/uapi/drm/v3d_drm.h
> > +++ b/include/uapi/drm/v3d_drm.h
> > @@ -73,6 +73,53 @@ struct drm_v3d_extension {
> >  	__u32 flags; /* mbz */
> >  };
> >  
> > +/* struct drm_v3d_sem - wait/signal semaphore
> > + *
> > + * If binary semaphore, it only takes syncobj handle and ignores
> > flags and
> > + * point fields. Point is defined for timeline syncobj feature.
> > + */
> > +struct drm_v3d_sem {
> > +	__u32 handle; /* syncobj */
> > +	/* rsv below, for future uses */
> > +	__u32 flags;
> > +	__u64 point;  /* for timeline sem support */
> > +	__u64 mbz[2]; /* must be zero, rsv */
> > +};
> > +
> > +/* Enum for each of the V3D queues. */
> > +enum v3d_queue {
> > +	V3D_BIN,
> > +	V3D_RENDER,
> > +	V3D_TFU,
> > +	V3D_CSD,
> > +	V3D_CACHE_CLEAN,
> > +};
> > +
> > +/**
> > + * struct drm_v3d_multi_sync - ioctl extension to add support
> > multiples
> > + * syncobjs for commands submission.
> > + *
> > + * When an extension of DRM_V3D_EXT_ID_MULTI_SYNC id is defined, it
> > points to
> > + * this extension to define wait and signal dependencies, instead of
> > single
> > + * in/out sync entries on submitting commands. The field flags is
> > used to
> > + * determine the stage to set wait dependencies.
> > + */
> > +struct drm_v3d_multi_sync {
> > +	struct drm_v3d_extension base;
> > +	/* Array of wait and signal semaphores */
> > +	__u64 in_syncs;
> > +	__u64 out_syncs;
> > +
> > +	/* Number of entries */
> > +	__u32 in_sync_count;
> > +	__u32 out_sync_count;
> > +
> > +	/* set the stage (v3d_queue) to sync */
> > +	__u32 wait_stage;
> > +
> > +	__u32 pad; /* mbz */
> > +};
> > +
> >  /**
> >   * struct drm_v3d_submit_cl - ioctl argument for submitting commands
> > to the 3D
> >   * engine.
> > @@ -228,6 +275,7 @@ enum drm_v3d_param {
> >  	DRM_V3D_PARAM_SUPPORTS_CSD,
> >  	DRM_V3D_PARAM_SUPPORTS_CACHE_FLUSH,
> >  	DRM_V3D_PARAM_SUPPORTS_PERFMON,
> > +	DRM_V3D_PARAM_SUPPORTS_MULTISYNC_EXT,
> >  };
> >  
> >  struct drm_v3d_get_param {
> > @@ -271,7 +319,6 @@ struct drm_v3d_submit_tfu {
> >  
> >  	/* Pointer to an array of ioctl extensions*/
> >  	__u64 extensions;
> > -
> >  };
> >  
> >  /* Submits a compute shader for dispatch.  This job will block on
> > any
>
Iago Toral Oct. 1, 2021, 8:41 a.m. UTC | #3
On Fri, 2021-10-01 at 09:37 +0100, Melissa Wen wrote:
> On 10/01, Iago Toral wrote:
> > On Thu, 2021-09-30 at 17:19 +0100, Melissa Wen wrote:
> > > Using the generic extension from the previous patch, a specific
> > > multisync
> > > extension enables more than one in/out binary syncobj per job
> > > submission.
> > > Arrays of syncobjs are set in struct drm_v3d_multisync, that also
> > > cares
> > > of determining the stage for sync (wait deps) according to the
> > > job
> > > queue.
> > > 
> > > v2:
> > > - subclass the generic extension struct (Daniel)
> > > - simplify adding dependency conditions to make understandable
> > > (Iago)
> > > 
> > > v3:
> > > - fix conditions to consider single or multiples in/out_syncs
> > > (Iago)
> > > - remove irrelevant comment (Iago)
> > > 
> > > Signed-off-by: Melissa Wen <mwen@igalia.com>
> > > ---
> > >  drivers/gpu/drm/v3d/v3d_drv.c |   6 +-
> > >  drivers/gpu/drm/v3d/v3d_drv.h |  24 +++--
> > >  drivers/gpu/drm/v3d/v3d_gem.c | 185
> > > ++++++++++++++++++++++++++++++
> > > ----
> > >  include/uapi/drm/v3d_drm.h    |  49 ++++++++-
> > >  4 files changed, 232 insertions(+), 32 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/v3d/v3d_drv.c
> > > b/drivers/gpu/drm/v3d/v3d_drv.c
> > > index 3d6b9bcce2f7..bd46396a1ae0 100644
> > > --- a/drivers/gpu/drm/v3d/v3d_drv.c
> > > +++ b/drivers/gpu/drm/v3d/v3d_drv.c
> > > @@ -96,6 +96,9 @@ static int v3d_get_param_ioctl(struct
> > > drm_device 
> > 
> > (...)
> > 
> > > @@ -516,9 +536,11 @@
> > > v3d_attach_fences_and_unlock_reservation(struct
> > > drm_file *file_priv,
> > >  					 struct v3d_job *job,
> > >  					 struct ww_acquire_ctx
> > > *acquire_ctx,
> > >  					 u32 out_sync,
> > > +					 struct v3d_submit_ext *se,
> > >  					 struct dma_fence *done_fence)
> > >  {
> > >  	struct drm_syncobj *sync_out;
> > > +	bool has_multisync = se && (se->flags & 
> > 
> > We always pass the 'se' pointer from a local variable allocated in
> > the
> > stack of the caller so it is never going to be NULL.
> > 
> > I am happy to keep the NULL check if you want to protect against
> > future
> > changes just in case, but if we do that then...
> > 
> > > DRM_V3D_EXT_ID_MULTI_SYNC);
> > >  	int i;
> > >  
> > >  	for (i = 0; i < job->bo_count; i++) {
> > 
> > (...)
> > 
> > > +static void
> > > +v3d_put_multisync_post_deps(struct v3d_submit_ext *se)
> > > +{
> > > +	unsigned int i;
> > > +
> > > +	if (!se->out_sync_count)
> > 
> > ...we should also check for NULL here for consistency.
> yes, consistency makes sense here.
> > Also, I think there is another problem in the code: we always call
> > v3d_job_cleanup for failed jobs, but only call v3d_job_put for
> > successful jobs. However, reading the docs for drm_sched_job_init:
> > 
> > "Drivers must make sure drm_sched_job_cleanup() if this function
> > returns successfully, even when @job is aborted before
> > drm_sched_job_arm() is called."
> > 
> > So my understanding is that we should call v3d_job_cleanup instead
> > of
> > v3d_job_put for successful jobs or we would be leaking sched
> > resources
> > on every successful job, no?
> 
> When job_init is successful, v3d_job_cleanup is called by scheduler
> when
> job is completed. drm_sched_job_cleanup describes how it works after
> drm_sched_job_arm:
> 
> " After that point of no return @job is committed to be executed by
> the
> scheduler, and this function should be called from the
> &drm_sched_backend_ops.free_job callback."
> 
> On v3d_sched.c, .free_job points to v3d_sched_job_free(), which in
> turn
> calls v3d_job_cleanup() (and then drm_sched_job_cleanup). So, it
> looks
> ok.
> 
> Also, we can say that the very last v3d_job_put() is in charge to
> decrement refcount initialized (set 1) in v3d_job_init(); while the
> v3d_job_cleanup() from v3d_sched_job_free() callback decrements
> refcount
> that was incremented when v3d_job_push() pushed the job to the
> scheduler.
> 
> So, refcount pairs seem consistent, I mean, get and put. And about
> drm_sched_job_cleanup, it is explicitly called when job_init fails or
> after that by the scheduler.
> 

   A. Ah ok, thanks for explaining this!

With the consistency issue discussed above fixed, for the series:

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

Iago
Melissa Wen Oct. 1, 2021, 8:53 a.m. UTC | #4
On 10/01, Iago Toral wrote:
> On Fri, 2021-10-01 at 09:37 +0100, Melissa Wen wrote:
> > On 10/01, Iago Toral wrote:
> > > On Thu, 2021-09-30 at 17:19 +0100, Melissa Wen wrote:
> > > > Using the generic extension from the previous patch, a specific
> > > > multisync
> > > > extension enables more than one in/out binary syncobj per job
> > > > submission.
> > > > Arrays of syncobjs are set in struct drm_v3d_multisync, that also
> > > > cares
> > > > of determining the stage for sync (wait deps) according to the
> > > > job
> > > > queue.
> > > > 
> > > > v2:
> > > > - subclass the generic extension struct (Daniel)
> > > > - simplify adding dependency conditions to make understandable
> > > > (Iago)
> > > > 
> > > > v3:
> > > > - fix conditions to consider single or multiples in/out_syncs
> > > > (Iago)
> > > > - remove irrelevant comment (Iago)
> > > > 
> > > > Signed-off-by: Melissa Wen <mwen@igalia.com>
> > > > ---
> > > >  drivers/gpu/drm/v3d/v3d_drv.c |   6 +-
> > > >  drivers/gpu/drm/v3d/v3d_drv.h |  24 +++--
> > > >  drivers/gpu/drm/v3d/v3d_gem.c | 185
> > > > ++++++++++++++++++++++++++++++
> > > > ----
> > > >  include/uapi/drm/v3d_drm.h    |  49 ++++++++-
> > > >  4 files changed, 232 insertions(+), 32 deletions(-)
> > > > 
> > > > diff --git a/drivers/gpu/drm/v3d/v3d_drv.c
> > > > b/drivers/gpu/drm/v3d/v3d_drv.c
> > > > index 3d6b9bcce2f7..bd46396a1ae0 100644
> > > > --- a/drivers/gpu/drm/v3d/v3d_drv.c
> > > > +++ b/drivers/gpu/drm/v3d/v3d_drv.c
> > > > @@ -96,6 +96,9 @@ static int v3d_get_param_ioctl(struct
> > > > drm_device 
> > > 
> > > (...)
> > > 
> > > > @@ -516,9 +536,11 @@
> > > > v3d_attach_fences_and_unlock_reservation(struct
> > > > drm_file *file_priv,
> > > >  					 struct v3d_job *job,
> > > >  					 struct ww_acquire_ctx
> > > > *acquire_ctx,
> > > >  					 u32 out_sync,
> > > > +					 struct v3d_submit_ext *se,
> > > >  					 struct dma_fence *done_fence)
> > > >  {
> > > >  	struct drm_syncobj *sync_out;
> > > > +	bool has_multisync = se && (se->flags & 
> > > 
> > > We always pass the 'se' pointer from a local variable allocated in
> > > the
> > > stack of the caller so it is never going to be NULL.
> > > 
> > > I am happy to keep the NULL check if you want to protect against
> > > future
> > > changes just in case, but if we do that then...
> > > 
> > > > DRM_V3D_EXT_ID_MULTI_SYNC);
> > > >  	int i;
> > > >  
> > > >  	for (i = 0; i < job->bo_count; i++) {
> > > 
> > > (...)
> > > 
> > > > +static void
> > > > +v3d_put_multisync_post_deps(struct v3d_submit_ext *se)
> > > > +{
> > > > +	unsigned int i;
> > > > +
> > > > +	if (!se->out_sync_count)
> > > 
> > > ...we should also check for NULL here for consistency.
> > yes, consistency makes sense here.
> > > Also, I think there is another problem in the code: we always call
> > > v3d_job_cleanup for failed jobs, but only call v3d_job_put for
> > > successful jobs. However, reading the docs for drm_sched_job_init:
> > > 
> > > "Drivers must make sure drm_sched_job_cleanup() if this function
> > > returns successfully, even when @job is aborted before
> > > drm_sched_job_arm() is called."
> > > 
> > > So my understanding is that we should call v3d_job_cleanup instead
> > > of
> > > v3d_job_put for successful jobs or we would be leaking sched
> > > resources
> > > on every successful job, no?
> > 
> > When job_init is successful, v3d_job_cleanup is called by scheduler
> > when
> > job is completed. drm_sched_job_cleanup describes how it works after
> > drm_sched_job_arm:
> > 
> > " After that point of no return @job is committed to be executed by
> > the
> > scheduler, and this function should be called from the
> > &drm_sched_backend_ops.free_job callback."
> > 
> > On v3d_sched.c, .free_job points to v3d_sched_job_free(), which in
> > turn
> > calls v3d_job_cleanup() (and then drm_sched_job_cleanup). So, it
> > looks
> > ok.
> > 
> > Also, we can say that the very last v3d_job_put() is in charge to
> > decrement refcount initialized (set 1) in v3d_job_init(); while the
> > v3d_job_cleanup() from v3d_sched_job_free() callback decrements
> > refcount
> > that was incremented when v3d_job_push() pushed the job to the
> > scheduler.
> > 
> > So, refcount pairs seem consistent, I mean, get and put. And about
> > drm_sched_job_cleanup, it is explicitly called when job_init fails or
> > after that by the scheduler.
> > 
> 
>    A. Ah ok, thanks for explaining this!
nice!
> 
> With the consistency issue discussed above fixed, for the series:
> 
> Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

ok, thanks for reviewing and all improvement suggestions,

Melissa
> 
> Iago
>
diff mbox series

Patch

diff --git a/drivers/gpu/drm/v3d/v3d_drv.c b/drivers/gpu/drm/v3d/v3d_drv.c
index 3d6b9bcce2f7..bd46396a1ae0 100644
--- a/drivers/gpu/drm/v3d/v3d_drv.c
+++ b/drivers/gpu/drm/v3d/v3d_drv.c
@@ -96,6 +96,9 @@  static int v3d_get_param_ioctl(struct drm_device *dev, void *data,
 	case DRM_V3D_PARAM_SUPPORTS_PERFMON:
 		args->value = (v3d->ver >= 40);
 		return 0;
+	case DRM_V3D_PARAM_SUPPORTS_MULTISYNC_EXT:
+		args->value = 1;
+		return 0;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", args->param);
 		return -EINVAL;
@@ -135,9 +138,8 @@  v3d_postclose(struct drm_device *dev, struct drm_file *file)
 	struct v3d_file_priv *v3d_priv = file->driver_priv;
 	enum v3d_queue q;
 
-	for (q = 0; q < V3D_MAX_QUEUES; q++) {
+	for (q = 0; q < V3D_MAX_QUEUES; q++)
 		drm_sched_entity_destroy(&v3d_priv->sched_entity[q]);
-	}
 
 	v3d_perfmon_close_file(v3d_priv);
 	kfree(v3d_priv);
diff --git a/drivers/gpu/drm/v3d/v3d_drv.h b/drivers/gpu/drm/v3d/v3d_drv.h
index b900a050d5e2..b74b1351bfc8 100644
--- a/drivers/gpu/drm/v3d/v3d_drv.h
+++ b/drivers/gpu/drm/v3d/v3d_drv.h
@@ -19,15 +19,6 @@  struct reset_control;
 
 #define GMP_GRANULARITY (128 * 1024)
 
-/* Enum for each of the V3D queues. */
-enum v3d_queue {
-	V3D_BIN,
-	V3D_RENDER,
-	V3D_TFU,
-	V3D_CSD,
-	V3D_CACHE_CLEAN,
-};
-
 #define V3D_MAX_QUEUES (V3D_CACHE_CLEAN + 1)
 
 struct v3d_queue_state {
@@ -294,6 +285,21 @@  struct v3d_csd_job {
 	struct drm_v3d_submit_csd args;
 };
 
+struct v3d_submit_outsync {
+	struct drm_syncobj *syncobj;
+};
+
+struct v3d_submit_ext {
+	u32 flags;
+	u32 wait_stage;
+
+	u32 in_sync_count;
+	u64 in_syncs;
+
+	u32 out_sync_count;
+	struct v3d_submit_outsync *out_syncs;
+};
+
 /**
  * __wait_for - magic wait macro
  *
diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c
index 93f130fb3a13..d344eae36b57 100644
--- a/drivers/gpu/drm/v3d/v3d_gem.c
+++ b/drivers/gpu/drm/v3d/v3d_gem.c
@@ -454,11 +454,12 @@  v3d_job_add_deps(struct drm_file *file_priv, struct v3d_job *job,
 static int
 v3d_job_init(struct v3d_dev *v3d, struct drm_file *file_priv,
 	     void **container, size_t size, void (*free)(struct kref *ref),
-	     u32 in_sync, enum v3d_queue queue)
+	     u32 in_sync, struct v3d_submit_ext *se, enum v3d_queue queue)
 {
 	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
 	struct v3d_job *job;
-	int ret;
+	bool has_multisync = se && (se->flags & DRM_V3D_EXT_ID_MULTI_SYNC);
+	int ret, i;
 
 	*container = kcalloc(1, size, GFP_KERNEL);
 	if (!*container) {
@@ -479,9 +480,28 @@  v3d_job_init(struct v3d_dev *v3d, struct drm_file *file_priv,
 	if (ret)
 		goto fail_job;
 
-	ret = v3d_job_add_deps(file_priv, job, in_sync, 0);
-	if (ret)
-		goto fail_deps;
+	if (has_multisync) {
+		if (se->in_sync_count && se->wait_stage == queue) {
+			struct drm_v3d_sem __user *handle = u64_to_user_ptr(se->in_syncs);
+
+			for (i = 0; i < se->in_sync_count; i++) {
+				struct drm_v3d_sem in;
+
+				ret = copy_from_user(&in, handle++, sizeof(in));
+				if (ret) {
+					DRM_DEBUG("Failed to copy wait dep handle.\n");
+					goto fail_deps;
+				}
+				ret = v3d_job_add_deps(file_priv, job, in.handle, 0);
+				if (ret)
+					goto fail_deps;
+			}
+		}
+	} else {
+		ret = v3d_job_add_deps(file_priv, job, in_sync, 0);
+		if (ret)
+			goto fail_deps;
+	}
 
 	kref_init(&job->refcount);
 
@@ -516,9 +536,11 @@  v3d_attach_fences_and_unlock_reservation(struct drm_file *file_priv,
 					 struct v3d_job *job,
 					 struct ww_acquire_ctx *acquire_ctx,
 					 u32 out_sync,
+					 struct v3d_submit_ext *se,
 					 struct dma_fence *done_fence)
 {
 	struct drm_syncobj *sync_out;
+	bool has_multisync = se && (se->flags & DRM_V3D_EXT_ID_MULTI_SYNC);
 	int i;
 
 	for (i = 0; i < job->bo_count; i++) {
@@ -530,20 +552,130 @@  v3d_attach_fences_and_unlock_reservation(struct drm_file *file_priv,
 	drm_gem_unlock_reservations(job->bo, job->bo_count, acquire_ctx);
 
 	/* Update the return sync object for the job */
-	sync_out = drm_syncobj_find(file_priv, out_sync);
-	if (sync_out) {
-		drm_syncobj_replace_fence(sync_out, done_fence);
-		drm_syncobj_put(sync_out);
+	/* If it only supports a single signal semaphore*/
+	if (!has_multisync) {
+		sync_out = drm_syncobj_find(file_priv, out_sync);
+		if (sync_out) {
+			drm_syncobj_replace_fence(sync_out, done_fence);
+			drm_syncobj_put(sync_out);
+		}
+		return;
+	}
+
+	/* If multiple semaphores extension is supported */
+	if (se->out_sync_count) {
+		for (i = 0; i < se->out_sync_count; i++) {
+			drm_syncobj_replace_fence(se->out_syncs[i].syncobj,
+						  done_fence);
+			drm_syncobj_put(se->out_syncs[i].syncobj);
+		}
+		kvfree(se->out_syncs);
+	}
+}
+
+static void
+v3d_put_multisync_post_deps(struct v3d_submit_ext *se)
+{
+	unsigned int i;
+
+	if (!se->out_sync_count)
+		return;
+
+	for (i = 0; i < se->out_sync_count; i++)
+		drm_syncobj_put(se->out_syncs[i].syncobj);
+	kvfree(se->out_syncs);
+}
+
+static int
+v3d_get_multisync_post_deps(struct drm_file *file_priv,
+			    struct v3d_submit_ext *se,
+			    u32 count, u64 handles)
+{
+	struct drm_v3d_sem __user *post_deps;
+	int i, ret;
+
+	if (!count)
+		return 0;
+
+	se->out_syncs = (struct v3d_submit_outsync *)
+			kvmalloc_array(count,
+				       sizeof(struct v3d_submit_outsync),
+				       GFP_KERNEL);
+	if (!se->out_syncs)
+		return -ENOMEM;
+
+	post_deps = u64_to_user_ptr(handles);
+
+	for (i = 0; i < count; i++) {
+		struct drm_v3d_sem out;
+
+		ret = copy_from_user(&out, post_deps++, sizeof(out));
+		if (ret) {
+			DRM_DEBUG("Failed to copy post dep handles\n");
+			goto fail;
+		}
+
+		se->out_syncs[i].syncobj = drm_syncobj_find(file_priv,
+							    out.handle);
+		if (!se->out_syncs[i].syncobj) {
+			ret = -EINVAL;
+			goto fail;
+		}
 	}
+	se->out_sync_count = count;
+
+	return 0;
+
+fail:
+	for (i--; i >= 0; i--)
+		drm_syncobj_put(se->out_syncs[i].syncobj);
+	kvfree(se->out_syncs);
+
+	return ret;
+}
+
+/* Get data for multiple binary semaphores synchronization. Parse syncobj
+ * to be signaled when job completes (out_sync).
+ */
+static int
+v3d_get_multisync_submit_deps(struct drm_file *file_priv,
+			      struct drm_v3d_extension __user *ext,
+			      void *data)
+{
+	struct drm_v3d_multi_sync multisync;
+	struct v3d_submit_ext *se = data;
+	int ret;
+
+	ret = copy_from_user(&multisync, ext, sizeof(multisync));
+	if (ret)
+		return ret;
+
+	if (multisync.pad)
+		return -EINVAL;
+
+	ret = v3d_get_multisync_post_deps(file_priv, data, multisync.out_sync_count,
+					  multisync.out_syncs);
+	if (ret)
+		return ret;
+
+	se->in_sync_count = multisync.in_sync_count;
+	se->in_syncs = multisync.in_syncs;
+	se->flags |= DRM_V3D_EXT_ID_MULTI_SYNC;
+	se->wait_stage = multisync.wait_stage;
+
+	return 0;
 }
 
 /* Whenever userspace sets ioctl extensions, v3d_get_extensions parses data
  * according to the extension id (name).
  */
 static int
-v3d_get_extensions(struct drm_file *file_priv, u64 ext_handles)
+v3d_get_extensions(struct drm_file *file_priv,
+		   u64 ext_handles,
+		   void *data)
 {
 	struct drm_v3d_extension __user *user_ext;
+	int ret;
 
 	user_ext = u64_to_user_ptr(ext_handles);
 	while (user_ext) {
@@ -555,7 +687,11 @@  v3d_get_extensions(struct drm_file *file_priv, u64 ext_handles)
 		}
 
 		switch (ext.id) {
-		case 0:
+		case DRM_V3D_EXT_ID_MULTI_SYNC:
+			ret = v3d_get_multisync_submit_deps(file_priv, user_ext, data);
+			if (ret)
+				return ret;
+			break;
 		default:
 			DRM_DEBUG_DRIVER("Unknown extension id: %d\n", ext.id);
 			return -EINVAL;
@@ -586,6 +722,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	struct v3d_dev *v3d = to_v3d_dev(dev);
 	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
 	struct drm_v3d_submit_cl *args = data;
+	struct v3d_submit_ext se = {0};
 	struct v3d_bin_job *bin = NULL;
 	struct v3d_render_job *render = NULL;
 	struct v3d_job *clean_job = NULL;
@@ -606,7 +743,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
-		ret = v3d_get_extensions(file_priv, args->extensions);
+		ret = v3d_get_extensions(file_priv, args->extensions, &se);
 		if (ret) {
 			DRM_DEBUG("Failed to get extensions.\n");
 			return ret;
@@ -614,7 +751,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	}
 
 	ret = v3d_job_init(v3d, file_priv, (void *)&render, sizeof(*render),
-			   v3d_render_job_free, args->in_sync_rcl, V3D_RENDER);
+			   v3d_render_job_free, args->in_sync_rcl, &se, V3D_RENDER);
 	if (ret)
 		goto fail;
 
@@ -624,7 +761,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 
 	if (args->bcl_start != args->bcl_end) {
 		ret = v3d_job_init(v3d, file_priv, (void *)&bin, sizeof(*bin),
-				   v3d_job_free, args->in_sync_bcl, V3D_BIN);
+				   v3d_job_free, args->in_sync_bcl, &se, V3D_BIN);
 		if (ret)
 			goto fail;
 
@@ -638,7 +775,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 
 	if (args->flags & DRM_V3D_SUBMIT_CL_FLUSH_CACHE) {
 		ret = v3d_job_init(v3d, file_priv, (void *)&clean_job, sizeof(*clean_job),
-				   v3d_job_free, 0, V3D_CACHE_CLEAN);
+				   v3d_job_free, 0, 0, V3D_CACHE_CLEAN);
 		if (ret)
 			goto fail;
 
@@ -698,6 +835,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 						 last_job,
 						 &acquire_ctx,
 						 args->out_sync,
+						 &se,
 						 last_job->done_fence);
 
 	if (bin)
@@ -716,6 +854,7 @@  v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	v3d_job_cleanup((void *)bin);
 	v3d_job_cleanup((void *)render);
 	v3d_job_cleanup(clean_job);
+	v3d_put_multisync_post_deps(&se);
 
 	return ret;
 }
@@ -735,6 +874,7 @@  v3d_submit_tfu_ioctl(struct drm_device *dev, void *data,
 {
 	struct v3d_dev *v3d = to_v3d_dev(dev);
 	struct drm_v3d_submit_tfu *args = data;
+	struct v3d_submit_ext se = {0};
 	struct v3d_tfu_job *job = NULL;
 	struct ww_acquire_ctx acquire_ctx;
 	int ret = 0;
@@ -747,7 +887,7 @@  v3d_submit_tfu_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
-		ret = v3d_get_extensions(file_priv, args->extensions);
+		ret = v3d_get_extensions(file_priv, args->extensions, &se);
 		if (ret) {
 			DRM_DEBUG("Failed to get extensions.\n");
 			return ret;
@@ -755,7 +895,7 @@  v3d_submit_tfu_ioctl(struct drm_device *dev, void *data,
 	}
 
 	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
-			   v3d_job_free, args->in_sync, V3D_TFU);
+			   v3d_job_free, args->in_sync, &se, V3D_TFU);
 	if (ret)
 		goto fail;
 
@@ -803,6 +943,7 @@  v3d_submit_tfu_ioctl(struct drm_device *dev, void *data,
 	v3d_attach_fences_and_unlock_reservation(file_priv,
 						 &job->base, &acquire_ctx,
 						 args->out_sync,
+						 &se,
 						 job->base.done_fence);
 
 	v3d_job_put(&job->base);
@@ -811,6 +952,7 @@  v3d_submit_tfu_ioctl(struct drm_device *dev, void *data,
 
 fail:
 	v3d_job_cleanup((void *)job);
+	v3d_put_multisync_post_deps(&se);
 
 	return ret;
 }
@@ -831,6 +973,7 @@  v3d_submit_csd_ioctl(struct drm_device *dev, void *data,
 	struct v3d_dev *v3d = to_v3d_dev(dev);
 	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
 	struct drm_v3d_submit_csd *args = data;
+	struct v3d_submit_ext se = {0};
 	struct v3d_csd_job *job = NULL;
 	struct v3d_job *clean_job = NULL;
 	struct ww_acquire_ctx acquire_ctx;
@@ -852,7 +995,7 @@  v3d_submit_csd_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (args->flags & DRM_V3D_SUBMIT_EXTENSION) {
-		ret = v3d_get_extensions(file_priv, args->extensions);
+		ret = v3d_get_extensions(file_priv, args->extensions, &se);
 		if (ret) {
 			DRM_DEBUG("Failed to get extensions.\n");
 			return ret;
@@ -860,12 +1003,12 @@  v3d_submit_csd_ioctl(struct drm_device *dev, void *data,
 	}
 
 	ret = v3d_job_init(v3d, file_priv, (void *)&job, sizeof(*job),
-			   v3d_job_free, args->in_sync, V3D_CSD);
+			   v3d_job_free, args->in_sync, &se, V3D_CSD);
 	if (ret)
 		goto fail;
 
 	ret = v3d_job_init(v3d, file_priv, (void *)&clean_job, sizeof(*clean_job),
-			   v3d_job_free, 0, V3D_CACHE_CLEAN);
+			   v3d_job_free, 0, 0, V3D_CACHE_CLEAN);
 	if (ret)
 		goto fail;
 
@@ -904,6 +1047,7 @@  v3d_submit_csd_ioctl(struct drm_device *dev, void *data,
 						 clean_job,
 						 &acquire_ctx,
 						 args->out_sync,
+						 &se,
 						 clean_job->done_fence);
 
 	v3d_job_put(&job->base);
@@ -918,6 +1062,7 @@  v3d_submit_csd_ioctl(struct drm_device *dev, void *data,
 fail:
 	v3d_job_cleanup((void *)job);
 	v3d_job_cleanup(clean_job);
+	v3d_put_multisync_post_deps(&se);
 
 	return ret;
 }
diff --git a/include/uapi/drm/v3d_drm.h b/include/uapi/drm/v3d_drm.h
index 55b443ca6c0b..3dfc0af8756a 100644
--- a/include/uapi/drm/v3d_drm.h
+++ b/include/uapi/drm/v3d_drm.h
@@ -73,6 +73,53 @@  struct drm_v3d_extension {
 	__u32 flags; /* mbz */
 };
 
+/* struct drm_v3d_sem - wait/signal semaphore
+ *
+ * If binary semaphore, it only takes syncobj handle and ignores flags and
+ * point fields. Point is defined for timeline syncobj feature.
+ */
+struct drm_v3d_sem {
+	__u32 handle; /* syncobj */
+	/* rsv below, for future uses */
+	__u32 flags;
+	__u64 point;  /* for timeline sem support */
+	__u64 mbz[2]; /* must be zero, rsv */
+};
+
+/* Enum for each of the V3D queues. */
+enum v3d_queue {
+	V3D_BIN,
+	V3D_RENDER,
+	V3D_TFU,
+	V3D_CSD,
+	V3D_CACHE_CLEAN,
+};
+
+/**
+ * struct drm_v3d_multi_sync - ioctl extension to add support multiples
+ * syncobjs for commands submission.
+ *
+ * When an extension of DRM_V3D_EXT_ID_MULTI_SYNC id is defined, it points to
+ * this extension to define wait and signal dependencies, instead of single
+ * in/out sync entries on submitting commands. The field flags is used to
+ * determine the stage to set wait dependencies.
+ */
+struct drm_v3d_multi_sync {
+	struct drm_v3d_extension base;
+	/* Array of wait and signal semaphores */
+	__u64 in_syncs;
+	__u64 out_syncs;
+
+	/* Number of entries */
+	__u32 in_sync_count;
+	__u32 out_sync_count;
+
+	/* set the stage (v3d_queue) to sync */
+	__u32 wait_stage;
+
+	__u32 pad; /* mbz */
+};
+
 /**
  * struct drm_v3d_submit_cl - ioctl argument for submitting commands to the 3D
  * engine.
@@ -228,6 +275,7 @@  enum drm_v3d_param {
 	DRM_V3D_PARAM_SUPPORTS_CSD,
 	DRM_V3D_PARAM_SUPPORTS_CACHE_FLUSH,
 	DRM_V3D_PARAM_SUPPORTS_PERFMON,
+	DRM_V3D_PARAM_SUPPORTS_MULTISYNC_EXT,
 };
 
 struct drm_v3d_get_param {
@@ -271,7 +319,6 @@  struct drm_v3d_submit_tfu {
 
 	/* Pointer to an array of ioctl extensions*/
 	__u64 extensions;
-
 };
 
 /* Submits a compute shader for dispatch.  This job will block on any