Update min Vulkan version to 1.0.8.0, and fix various bugs

Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

TBR=bsalomon@google.com

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1906623002

Committed: https://skia.googlesource.com/skia/+/58a8d9214a70e0f6c81c88a8b0b563c06bf0f70e

[egdaniel, 2016-04-20 18:43:37]

Description was changed from

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

BUG=skia:
==========

to

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

[egdaniel, 2016-04-20 18:51:53]

Description was changed from

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

to

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass
with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

[egdaniel, 2016-04-20 18:53:10]

egdaniel@google.com changed reviewers:
+ jvanverth@google.com

[jvanverth1, 2016-04-21 13:25:24]

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkBackendC...
File src/gpu/vk/GrVkBackendContext.cpp (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkBackendC...
src/gpu/vk/GrVkBackendContext.cpp:26: "VK_LAYER_GOOGLE_unique_objects",
Does this work with the WSI code? I commented it out because it wasn't before.

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp
File src/gpu/vk/GrVkGpu.cpp (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp#...
src/gpu/vk/GrVkGpu.cpp:1647: this->xferBarrier(pipeline.getRenderTarget(),
barrierType);
Why are we bothering to call this? It does nothing.

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp#...
src/gpu/vk/GrVkGpu.cpp:1664: fCurrentCmdBuffer->endRenderPass(this);
On the first iteration, it doesn't look like we do anything between
beginRenderPass and endRenderPass. Is there a way to avoid this?

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkImage.h
File src/gpu/vk/GrVkImage.h (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkImage.h#...
src/gpu/vk/GrVkImage.h:35: fImage(VK_NULL_HANDLE),
Nit: usually we put the commas before the variable, lined up with the colon.

[egdaniel, 2016-04-21 14:09:20]

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkBackendC...
File src/gpu/vk/GrVkBackendContext.cpp (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkBackendC...
src/gpu/vk/GrVkBackendContext.cpp:26: "VK_LAYER_GOOGLE_unique_objects",
On 2016/04/21 13:25:24, jvanverth1 wrote:
> Does this work with the WSI code? I commented it out because it wasn't before.

works with new SDK

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp
File src/gpu/vk/GrVkGpu.cpp (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp#...
src/gpu/vk/GrVkGpu.cpp:1647: this->xferBarrier(pipeline.getRenderTarget(),
barrierType);
On 2016/04/21 13:25:24, jvanverth1 wrote:
> Why are we bothering to call this? It does nothing.

gone

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkGpu.cpp#...
src/gpu/vk/GrVkGpu.cpp:1664: fCurrentCmdBuffer->endRenderPass(this);
On 2016/04/21 13:25:24, jvanverth1 wrote:
> On the first iteration, it doesn't look like we do anything between
> beginRenderPass and endRenderPass. Is there a way to avoid this?

Well we should never get here on the first iteration since we set the primitive
type to be that of the first mesh. In reality we actually never end up in this
if statement since all our flushes currently use the same primitive type for all
meshes. But the api does allow for changing the primitive type so this is here
for future proofing.

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkImage.h
File src/gpu/vk/GrVkImage.h (right):

https://codereview.chromium.org/1906623002/diff/20001/src/gpu/vk/GrVkImage.h#...
src/gpu/vk/GrVkImage.h:35: fImage(VK_NULL_HANDLE),
On 2016/04/21 13:25:24, jvanverth1 wrote:
> Nit: usually we put the commas before the variable, lined up with the colon.

Woops no idea why my mind did this opposite here.

[jvanverth1, 2016-04-21 14:18:14]

lgtm

[egdaniel, 2016-04-21 14:23:45]

The CQ bit was checked by egdaniel@google.com

[egdaniel, 2016-04-21 14:23:45]

The patchset sent to the CQ was uploaded after l-g-t-m from jvanverth@google.com
Link to the patchset: https://codereview.chromium.org/1906623002/#ps70001
(title: "rebase")

[commit-bot: I haz the power, 2016-04-21 14:23:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1906623002/70001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1906623002/70001

[commit-bot: I haz the power, 2016-04-21 14:25:24]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-04-21 14:25:25]

Try jobs failed on following builders:
  skia_presubmit-Trybot on client.skia.fyi (JOB_FAILED,
http://build.chromium.org/p/client.skia.fyi/builders/skia_presubmit-Trybot/bu...)

[egdaniel, 2016-04-21 15:02:06]

Description was changed from

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass
with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

to

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass
with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

TBR=bsalomon@google.com

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

[egdaniel, 2016-04-21 15:02:10]

The CQ bit was checked by egdaniel@google.com

[commit-bot: I haz the power, 2016-04-21 15:02:15]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1906623002/70001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1906623002/70001

[commit-bot: I haz the power, 2016-04-21 15:03:12]

Description was changed from

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass
with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

TBR=bsalomon@google.com

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

to

==========
Update min Vulkan version to 1.0.8.0, and fix various bugs

With updating the SDK, the debug layers also showed multiple bugs.
I have fixed those as well in this CL. These include:

1. Incorrectly tracking the allocated descriptor sets from the descriptor pools

2. Using MemoryBarriers inside render passes.

3. Correctly setting the Stencil Image layout anytime we are using a render pass
with a stencil attachment

4. Setting the correct aspect mask for Depth/Stencil in a barrier.

TBR=bsalomon@google.com

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...

Committed:
https://skia.googlesource.com/skia/+/58a8d9214a70e0f6c81c88a8b0b563c06bf0f70e
==========

[commit-bot: I haz the power, 2016-04-21 15:03:13]

Committed patchset #5 (id:70001) as
https://skia.googlesource.com/skia/+/58a8d9214a70e0f6c81c88a8b0b563c06bf0f70e