Distinguish ES2/ES3 contexts in GPU command buffer

Distinguish ES2/ES3 contexts in GPU command buffer

BUG=463675
TEST=bots, WebGL2 demos
R=kbr@chromium.org,bajones@chromium.org

Committed: https://crrev.com/0ee1586af8c85b0b6ab241bb7c8e6b8fd565848f
Cr-Commit-Position: refs/heads/master@{#319016}

[Zhenyao Mo, 2015-03-03 22:25:47]

Please take a look.

[bajones, 2015-03-03 22:33:22]

On 2015/03/03 at 22:25:47, zmo wrote:
> Please take a look.

LGTM! Thank you!

[Zhenyao Mo, 2015-03-03 22:36:39]

zmo@chromium.org changed reviewers:
+ piman@chromium.org, sievers@chromium.org

[Zhenyao Mo, 2015-03-03 22:37:28]

also adding piman/sievers for OWNER approval

[no sievers, 2015-03-03 22:48:22]

On 2015/03/03 22:37:28, Zhenyao Mo wrote:
> also adding piman/sievers for OWNER approval

lgtm

[Zhenyao Mo, 2015-03-03 22:56:43]

Forgot to mention: this CL depends on an ANGLE roll:
https://codereview.chromium.org/970363002

Will land once the ANGLE rolled in.

[Ken Russell (switch to Gerrit), 2015-03-03 23:14:49]

LGTM overall.

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
File gpu/command_buffer/common/gles2_cmd_utils.cc (right):

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
gpu/command_buffer/common/gles2_cmd_utils.cc:897: const int32
kES3ContextRequired = 0x10003;
Might want to think about how this will evolve once ES 3.1 comes along. Maybe
two optional integer parameters specifying the minimum major and minor version
of ES? This seems fine for now though.

[Zhenyao Mo, 2015-03-04 00:24:03]

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
File gpu/command_buffer/common/gles2_cmd_utils.cc (right):

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
gpu/command_buffer/common/gles2_cmd_utils.cc:897: const int32
kES3ContextRequired = 0x10003;
On 2015/03/03 23:14:49, Ken Russell wrote:
> Might want to think about how this will evolve once ES 3.1 comes along. Maybe
> two optional integer parameters specifying the minimum major and minor version
> of ES? This seems fine for now though.

I was thinking about that, but the blink side attributes defines WebGL version
as one integer, so I feel no need to take the extra step in command buffer for
now.  What you suggest sounds good to me when we get to that point.

[piman, 2015-03-04 00:44:50]

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
File gpu/command_buffer/common/gles2_cmd_utils.cc (right):

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
gpu/command_buffer/common/gles2_cmd_utils.cc:897: const int32
kES3ContextRequired = 0x10003;
On 2015/03/04 00:24:03, Zhenyao Mo wrote:
> On 2015/03/03 23:14:49, Ken Russell wrote:
> > Might want to think about how this will evolve once ES 3.1 comes along.
Maybe
> > two optional integer parameters specifying the minimum major and minor
version
> > of ES? This seems fine for now though.
> 
> I was thinking about that, but the blink side attributes defines WebGL version
> as one integer, so I feel no need to take the extra step in command buffer for
> now.  What you suggest sounds good to me when we get to that point.

Actually I think we should get rid of the EGL-type int-array attrib lists, and
instead explicitly serialize a struct over IPC. It would be much easier to
manage and evolve.

[piman, 2015-03-04 00:45:02]

lgtm

[Zhenyao Mo, 2015-03-04 00:49:45]

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
File gpu/command_buffer/common/gles2_cmd_utils.cc (right):

https://codereview.chromium.org/978593003/diff/20001/gpu/command_buffer/commo...
gpu/command_buffer/common/gles2_cmd_utils.cc:897: const int32
kES3ContextRequired = 0x10003;
On 2015/03/04 00:44:50, piman (Very slow to review) wrote:
> On 2015/03/04 00:24:03, Zhenyao Mo wrote:
> > On 2015/03/03 23:14:49, Ken Russell wrote:
> > > Might want to think about how this will evolve once ES 3.1 comes along.
> Maybe
> > > two optional integer parameters specifying the minimum major and minor
> version
> > > of ES? This seems fine for now though.
> > 
> > I was thinking about that, but the blink side attributes defines WebGL
version
> > as one integer, so I feel no need to take the extra step in command buffer
for
> > now.  What you suggest sounds good to me when we get to that point.
> 
> Actually I think we should get rid of the EGL-type int-array attrib lists, and
> instead explicitly serialize a struct over IPC. It would be much easier to
> manage and evolve.

I agree.  Will follow up on this.

[Zhenyao Mo, 2015-03-04 00:59:31]

Patchset #3 (id:40001) has been deleted

[Zhenyao Mo, 2015-03-04 00:59:41]

Patchset #3 (id:60001) has been deleted

[Zhenyao Mo, 2015-03-04 01:17:06]

The CQ bit was checked by zmo@chromium.org

[Zhenyao Mo, 2015-03-04 01:17:07]

The patchset sent to the CQ was uploaded after l-g-t-m from
bajones@chromium.org, kbr@chromium.org, sievers@chromium.org, piman@chromium.org
Link to the patchset: https://codereview.chromium.org/978593003/#ps80001 (title:
"rebase")

[commit-bot: I haz the power, 2015-03-04 01:17:56]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/978593003/80001

[commit-bot: I haz the power, 2015-03-04 03:50:28]

Committed patchset #3 (id:80001)

[commit-bot: I haz the power, 2015-03-04 03:51:09]

Patchset 3 (id:??) landed as
https://crrev.com/0ee1586af8c85b0b6ab241bb7c8e6b8fd565848f
Cr-Commit-Position: refs/heads/master@{#319016}