Handle pixel (un)pack buffers with non-zero offset.

Handle pixel (un)pack buffers with non-zero offset.

The passthrough command buffer (Compressed)Tex(Sub)Image and ReadPixels
handlers didn't handle the case where shm_id == 0 but shm_offset != 0
which can happen when a PBO is bound.

This patch also adds a dataSize argument to the CompressedTex(Sub)Image
doers that will be needed to have a robust version of them when "data"
can be an offset in a PBO.

BUG=chromium:602688
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2811693005
Cr-Commit-Position: refs/heads/master@{#464137}
Committed: https://chromium.googlesource.com/chromium/src/+/95f938225bec04fe8e3d87bcc692f67c6388cf06

[Corentin Wallez, 2017-04-11 19:09:38]

Description was changed from

==========
Handle pixel (un)pack buffers with non-zero offset.

The passthrough command buffer (Compressed)Tex(Sub)Image and ReadPixels
handlers didn't handle the case where shm_id == 0 but shm_offset != 0
which can happen when a PBO is bound.

This patch also adds a dataSize argument to the CompressedTex(Sub)Image
doers that will be needed to have a robust version of them when "data"
can be an offset in a PBO.

BUG=chromium:602688
==========

to

==========
Handle pixel (un)pack buffers with non-zero offset.

The passthrough command buffer (Compressed)Tex(Sub)Image and ReadPixels
handlers didn't handle the case where shm_id == 0 but shm_offset != 0
which can happen when a PBO is bound.

This patch also adds a dataSize argument to the CompressedTex(Sub)Image
doers that will be needed to have a robust version of them when "data"
can be an offset in a PBO.

BUG=chromium:602688
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
==========

[Corentin Wallez, 2017-04-11 19:10:12]

cwallez@chromium.org changed reviewers:
+ geofflang@chromium.org, zmo@chromium.org

[Corentin Wallez, 2017-04-11 19:10:12]

PTAL, fixes some contexts lost in the WebGL 2 CTS with the passthrough cmd
decoder.

[Corentin Wallez, 2017-04-11 19:10:16]

The CQ bit was checked by cwallez@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-11 19:10:44]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Zhenyao Mo, 2017-04-11 19:56:54]

lgtm with a few nits

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
File gpu/command_buffer/service/gles2_cmd_decoder_passthrough_doers.cc (right):

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_doers.cc:470: GLsizei
dataSize,
nit: this is actually against chromium coding style.

imageSize -> image_size, dataSize -> data_size

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
File gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc
(right):

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2532:
unsigned int data_size = 0;
nit: use GLsizei to be consistent.

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2589:
unsigned int data_size = 0;
nit: use GLsizei.

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2644:
unsigned int data_size = 0;
nit: use GLsizei.

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2705:
unsigned int data_size = 0;
nit: use GLsizei.

[commit-bot: I haz the power, 2017-04-11 20:15:45]

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

[commit-bot: I haz the power, 2017-04-11 20:15:46]

Dry run: This issue passed the CQ dry run.

[Corentin Wallez, 2017-04-11 22:06:34]

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
File gpu/command_buffer/service/gles2_cmd_decoder_passthrough_doers.cc (right):

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_doers.cc:470: GLsizei
dataSize,
On 2017/04/11 at 19:56:54, Zhenyao Mo wrote:
> nit: this is actually against chromium coding style.
> 
> imageSize -> image_size, dataSize -> data_size

Done here and in other places in this file and the header.

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
File gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc
(right):

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2532:
unsigned int data_size = 0;
On 2017/04/11 at 19:56:54, Zhenyao Mo wrote:
> nit: use GLsizei to be consistent.

data_size has to be unsigned because GetSharedMemoryAndSizeAs expects an
unsigned int*.

Are you concerned about overflows and clamping of data_size? ANGLE checks
data_size >= 0 early and return an error if < 0.

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2589:
unsigned int data_size = 0;
On 2017/04/11 at 19:56:54, Zhenyao Mo wrote:
> nit: use GLsizei.

See comment above

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2644:
unsigned int data_size = 0;
On 2017/04/11 at 19:56:54, Zhenyao Mo wrote:
> nit: use GLsizei.

See comment above

https://codereview.chromium.org/2811693005/diff/1/gpu/command_buffer/service/...
gpu/command_buffer/service/gles2_cmd_decoder_passthrough_handlers.cc:2705:
unsigned int data_size = 0;
On 2017/04/11 at 19:56:54, Zhenyao Mo wrote:
> nit: use GLsizei.

See comment above

[Corentin Wallez, 2017-04-12 17:40:02]

The CQ bit was checked by cwallez@chromium.org

[Corentin Wallez, 2017-04-12 17:40:02]

Thanks for the review!

[Corentin Wallez, 2017-04-12 17:40:03]

The patchset sent to the CQ was uploaded after l-g-t-m from zmo@chromium.org
Link to the patchset: https://codereview.chromium.org/2811693005/#ps20001
(title: "Address comments")

[commit-bot: I haz the power, 2017-04-12 17:40:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-12 20:38:17]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1492018802073630,
"parent_rev": "950fbc86a6623aebe69b471dfa156504b8f4467b", "commit_rev":
"95f938225bec04fe8e3d87bcc692f67c6388cf06"}

[commit-bot: I haz the power, 2017-04-12 20:39:15]

Description was changed from

==========
Handle pixel (un)pack buffers with non-zero offset.

The passthrough command buffer (Compressed)Tex(Sub)Image and ReadPixels
handlers didn't handle the case where shm_id == 0 but shm_offset != 0
which can happen when a PBO is bound.

This patch also adds a dataSize argument to the CompressedTex(Sub)Image
doers that will be needed to have a robust version of them when "data"
can be an offset in a PBO.

BUG=chromium:602688
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
==========

to

==========
Handle pixel (un)pack buffers with non-zero offset.

The passthrough command buffer (Compressed)Tex(Sub)Image and ReadPixels
handlers didn't handle the case where shm_id == 0 but shm_offset != 0
which can happen when a PBO is bound.

This patch also adds a dataSize argument to the CompressedTex(Sub)Image
doers that will be needed to have a robust version of them when "data"
can be an offset in a PBO.

BUG=chromium:602688
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2811693005
Cr-Commit-Position: refs/heads/master@{#464137}
Committed:
https://chromium.googlesource.com/chromium/src/+/95f938225bec04fe8e3d87bcc692...
==========

[commit-bot: I haz the power, 2017-04-12 20:39:16]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/95f938225bec04fe8e3d87bcc692...