2019-06-17 11:14:39

by Gerd Hoffmann

Subject: [PATCH 3/4] drm/virtio: simplify cursor updates

No need to do the reservation dance,
we can just wait on the fence directly.

Signed-off-by: Gerd Hoffmann <[email protected]>
---
drivers/gpu/drm/virtio/virtgpu_plane.c | 13 +++----------
1 file changed, 3 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 024c2aa0c929..4b805bf466d3 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -184,7 +184,6 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
struct virtio_gpu_framebuffer *vgfb;
struct virtio_gpu_object *bo = NULL;
uint32_t handle;
- int ret = 0;

if (plane->state->crtc)
output = drm_crtc_to_virtio_gpu_output(plane->state->crtc);
@@ -208,15 +207,9 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
cpu_to_le32(plane->state->crtc_w),
cpu_to_le32(plane->state->crtc_h),
0, 0, vgfb->fence);
- ret = virtio_gpu_object_reserve(bo, false);
- if (!ret) {
- reservation_object_add_excl_fence(bo->tbo.resv,
- &vgfb->fence->f);
- dma_fence_put(&vgfb->fence->f);
- vgfb->fence = NULL;
- virtio_gpu_object_unreserve(bo);
- virtio_gpu_object_wait(bo, false);
- }
+ dma_fence_wait(&vgfb->fence->f, true);
+ dma_fence_put(&vgfb->fence->f);
+ vgfb->fence = NULL;
}

if (plane->state->fb != old_state->fb) {
--
2.18.1

2019-06-17 14:18:33

by Daniel Vetter

Subject: Re: [PATCH 3/4] drm/virtio: simplify cursor updates

On Mon, Jun 17, 2019 at 01:14:05PM +0200, Gerd Hoffmann wrote:
> No need to do the reservation dance,
> we can just wait on the fence directly.
>
> Signed-off-by: Gerd Hoffmann <[email protected]>
> ---
> drivers/gpu/drm/virtio/virtgpu_plane.c | 13 +++----------
> 1 file changed, 3 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 024c2aa0c929..4b805bf466d3 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -184,7 +184,6 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
> struct virtio_gpu_framebuffer *vgfb;
> struct virtio_gpu_object *bo = NULL;
> uint32_t handle;
> - int ret = 0;
>
> if (plane->state->crtc)
> output = drm_crtc_to_virtio_gpu_output(plane->state->crtc);
> @@ -208,15 +207,9 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
> cpu_to_le32(plane->state->crtc_w),
> cpu_to_le32(plane->state->crtc_h),
> 0, 0, vgfb->fence);
> - ret = virtio_gpu_object_reserve(bo, false);
> - if (!ret) {
> - reservation_object_add_excl_fence(bo->tbo.resv,
> - &vgfb->fence->f);
> - dma_fence_put(&vgfb->fence->f);
> - vgfb->fence = NULL;
> - virtio_gpu_object_unreserve(bo);
> - virtio_gpu_object_wait(bo, false);
> - }
> + dma_fence_wait(&vgfb->fence->f, true);
> + dma_fence_put(&vgfb->fence->f);
> + vgfb->fence = NULL;

Even nicer would be to add the fence using
drm_atomic_set_fence_for_plane() in the prepare_fb hook. Assuming this
isn't necessary for correctness (but then I kinda have questions about
races and stuff).

But this gets the job done too I think, so:

Reviewed-by: Daniel Vetter <[email protected]>

> }
>
> if (plane->state->fb != old_state->fb) {
> --
> 2.18.1
>

--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

2019-06-18 07:43:44

by Gerd Hoffmann

Subject: Re: [PATCH 3/4] drm/virtio: simplify cursor updates

> Even nicer would be to add the fence using
> drm_atomic_set_fence_for_plane() in the prepare_fb hook. Assuming this
> isn't necessary for correctness (but then I kinda have questions about
> races and stuff).

I'll have a look. Maybe this way I can drop struct
virtio_gpu_framebuffer (where the fence is the only
thing beside struct drm_framebuffer).

cheers,
Gerd