Skip to contentSkip to conversation

virtio-gpu: update done only on the scanout associated with rect

PATCH In review Dongwon Kim 3 participants 5 messages May 5, 2022 – Jul 19, 2022

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 virtio-gpu: update done only on the scanout associated with rect

It only needs to update the scanouts containing the rect area coming with the resource-flush request from the guest.

Cc Gerd Hoffmann <[email protected]>
Cc Vivek Kasireddy <[email protected]>
---
 hw/display/virtio-gpu.c | 3 +++
 1 file changed, 3 insertions(+)
hw/display/virtio-gpu.c +3 -0
index 529b5246b2..165ecafd7a 100644
@@ -514,6 +514,9 @@ static void virtio_gpu_resource_flush(VirtIOGPU *g,
         for (i = 0; i < g->parent_obj.conf.max_outputs; i++) {
             scanout = &g->parent_obj.scanout[i];
             if (scanout->resource_id == res->resource_id &&
+                rf.r.x >= scanout->x && rf.r.y >= scanout->y &&
+                rf.r.x + rf.r.width <= scanout->x + scanout->width &&
+                rf.r.y + rf.r.height <= scanout->y + scanout->height &&
                 console_has_gl(scanout->con)) {
                 dpy_gl_update(scanout->con, 0, 0, scanout->width,
                               scanout->height);
Signature · 1 line
2.30.2
#1 lore raw

Hi

Quoted 4 lines · Dongwon Kim

It only needs to update the scanouts containing the rect area coming with the resource-flush request from the guest.

Cc Gerd Hoffmann <[email protected]>
Quoted 11 lines
Cc Vivek Kasireddy <[email protected]>
---
 hw/display/virtio-gpu.c | 3 +++
 1 file changed, 3 insertions(+)
hw/display/virtio-gpu.c +0 -0
index 529b5246b2..165ecafd7a 100644

hw/display/virtio-gpu.c static void virtio_gpu_resource_flush(VirtIOGPU *g, · line 520

Quoted 7 lines
hw/display/virtio-gpu.c +3 -0
@@ -514,6 +514,9 @@ static void virtio_gpu_resource_flush(VirtIOGPU *g,
         for (i = 0; i < g->parent_obj.conf.max_outputs; i++) {
             scanout = &g->parent_obj.scanout[i];
             if (scanout->resource_id == res->resource_id &&
+                rf.r.x >= scanout->x && rf.r.y >= scanout->y &&
+                rf.r.x + rf.r.width <= scanout->x + scanout->width &&
+                rf.r.y + rf.r.height <= scanout->y + scanout->height &&
 

That doesn't seem to handle intersections/overlapping, I think it should.

Quoted 8 lines
hw/display/virtio-gpu.c +0 -0
                 console_has_gl(scanout->con)) {
                 dpy_gl_update(scanout->con, 0, 0, scanout->width,
                               scanout->height);

--
2.30.2

Signature · 1 line
Marc-André Lureau
#2 lore raw
Quoted 15 lines · Marc-André Lureau

Hi

Quoted 4 lines · Dongwon Kim

It only needs to update the scanouts containing the rect area coming with the resource-flush request from the guest.

Cc Gerd Hoffmann <[email protected]>
Quoted 6 lines
Cc Vivek Kasireddy <[email protected]>
---
 hw/display/virtio-gpu.c | 3 +++
 1 file changed, 3 insertions(+)
Quoted 15 lines

hw/display/virtio-gpu.c static void virtio_gpu_resource_flush(VirtIOGPU *g, · line 520

Quoted 12 lines
hw/display/virtio-gpu.c +3 -0
index 529b5246b2..165ecafd7a 100644
@@ -514,6 +514,9 @@ static void virtio_gpu_resource_flush(VirtIOGPU *g,
         for (i = 0; i < g->parent_obj.conf.max_outputs; i++) {
             scanout = &g->parent_obj.scanout[i];
             if (scanout->resource_id == res->resource_id &&
+                rf.r.x >= scanout->x && rf.r.y >= scanout->y &&
+                rf.r.x + rf.r.width <= scanout->x + scanout->width &&
+                rf.r.y + rf.r.height <= scanout->y + scanout->height &&
 

That doesn't seem to handle intersections/overlapping, I think it should.

so set-scanouts and resource flushes are issued per scanout(CRTC/plane from guest's point of view). In case of intersections/overlapping, there will be two resource flushes (in case there are two scanouts) and each resource flush will take care of updating the scanout that covers partial damaged area.

The problem with the original code is that even if there is a resource flush request for a single scanout, it does update on both (or more) as well, which is unnecessary burden.

Thanks

Quoted 13 lines
Quoted 10 lines
hw/display/virtio-gpu.c +0 -0
                 console_has_gl(scanout->con)) {
                 dpy_gl_update(scanout->con, 0, 0, scanout->width,
                               scanout->height);

--
2.30.2

Signature · 1 line
Marc-André Lureau
#3 lore raw
Quoted 32 lines · Dongwon Kim
Quoted 15 lines · Marc-André Lureau

Hi

Quoted 4 lines · Dongwon Kim

It only needs to update the scanouts containing the rect area coming with the resource-flush request from the guest.

Cc Gerd Hoffmann <[email protected]>
Quoted 6 lines
Cc Vivek Kasireddy <[email protected]>
---
 hw/display/virtio-gpu.c | 3 +++
 1 file changed, 3 insertions(+)
Quoted 15 lines

hw/display/virtio-gpu.c static void virtio_gpu_resource_flush(VirtIOGPU *g, · line 520

Quoted 12 lines
hw/display/virtio-gpu.c +3 -0
index 529b5246b2..165ecafd7a 100644
@@ -514,6 +514,9 @@ static void virtio_gpu_resource_flush(VirtIOGPU *g,
         for (i = 0; i < g->parent_obj.conf.max_outputs; i++) {
             scanout = &g->parent_obj.scanout[i];
             if (scanout->resource_id == res->resource_id &&
+                rf.r.x >= scanout->x && rf.r.y >= scanout->y &&
+                rf.r.x + rf.r.width <= scanout->x + scanout->width &&
+                rf.r.y + rf.r.height <= scanout->y + scanout->height &&
 

That doesn't seem to handle intersections/overlapping, I think it should.

Quoted 5 lines

so set-scanouts and resource flushes are issued per scanout(CRTC/plane from guest's point of view). In case of intersections/overlapping, there will be two resource flushes (in case there are two scanouts) and each resource flush will take care of updating the scanout that covers partial damaged area.

Even though the linux kernel driver sends two flushes, one for each scanout, it is perfectly valid send a single flush for the complete resource.

So checking whenever the rectangle is completely within the scanout is not correct. When the scanout is covered partly you must update too. Only when the rectangle is completely outside the scanout it is valid to skip it.

take care,
Gerd

#4 lore raw
Quoted 39 lines · Gerd Hoffmann
Quoted 32 lines · Dongwon Kim
Quoted 15 lines · Marc-André Lureau

Hi

Quoted 4 lines · Dongwon Kim

It only needs to update the scanouts containing the rect area coming with the resource-flush request from the guest.

Cc Gerd Hoffmann <[email protected]>
Quoted 6 lines
Cc Vivek Kasireddy <[email protected]>
---
 hw/display/virtio-gpu.c | 3 +++
 1 file changed, 3 insertions(+)
Quoted 15 lines

hw/display/virtio-gpu.c static void virtio_gpu_resource_flush(VirtIOGPU *g, · line 520

Quoted 12 lines
hw/display/virtio-gpu.c +3 -0
index 529b5246b2..165ecafd7a 100644
@@ -514,6 +514,9 @@ static void virtio_gpu_resource_flush(VirtIOGPU *g,
         for (i = 0; i < g->parent_obj.conf.max_outputs; i++) {
             scanout = &g->parent_obj.scanout[i];
             if (scanout->resource_id == res->resource_id &&
+                rf.r.x >= scanout->x && rf.r.y >= scanout->y &&
+                rf.r.x + rf.r.width <= scanout->x + scanout->width &&
+                rf.r.y + rf.r.height <= scanout->y + scanout->height &&
 

That doesn't seem to handle intersections/overlapping, I think it should.

Quoted 5 lines

so set-scanouts and resource flushes are issued per scanout(CRTC/plane from guest's point of view). In case of intersections/overlapping, there will be two resource flushes (in case there are two scanouts) and each resource flush will take care of updating the scanout that covers partial damaged area.

Quoted 8 lines

Even though the linux kernel driver sends two flushes, one for each scanout, it is perfectly valid send a single flush for the complete resource.

So checking whenever the rectangle is completely within the scanout is not correct. When the scanout is covered partly you must update too. Only when the rectangle is completely outside the scanout it is valid to skip it.

Gerd,

I got your point. I will take a look into it.

Quoted 4 lines

take care,
Gerd