Move GL headers from gpu to third_party/khronos

Move GL headers from gpu to third_party/khronos

This change moves the OpenGL ES and EGL header files from gpu/ to a new third_party/khronos/ directory. The headers are also rebased to the latest upstream versions.

BUG=102824
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=111170

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=111523

[Paweł Hajdan Jr., 2011-11-02 17:46:47]

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
File tools/checklicenses/checklicenses.py (left):

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
tools/checklicenses/checklicenses.py:117: 'UNKNOWN',
Excellent, thank you!

One little thing: could you please compare run time of this tool with and
without your modification to licensecheck.pl? Please make sure to always run
with empty caches, "echo 3 > /proc/sys/vm/drop_caches".

[apatrick_chromium, 2011-11-02 17:57:04]

I have a feeling this might break NaCl since it references those GL headers in
gpu/. There is a nacl_integration step on the bots but I think it is disabled.

[Sami, 2011-11-02 18:06:46]

On 2011/11/02 17:57:04, apatrick_chromium wrote:
> I have a feeling this might break NaCl since it references those GL headers in
> gpu/. There is a nacl_integration step on the bots but I think it is disabled.

Yeah, I'm currently looking into a NaCl failure on the mac try bot:

http://build.chromium.org/p/tryserver.chromium/builders/mac/builds/640/steps/...

Any suggestions would be appreciated since I don't have a local mac build
environment.

[Sami, 2011-11-02 18:24:25]

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
File tools/checklicenses/checklicenses.py (left):

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
tools/checklicenses/checklicenses.py:117: 'UNKNOWN',
Sure, here are the results.

Without the changes:

0m36.454s
0m34.651s
0m37.063s

With the changes:

0m36.915s
0m34.467s
0m37.470s

On 2011/11/02 17:46:47, Paweł Hajdan Jr. wrote:
> Excellent, thank you!
> 
> One little thing: could you please compare run time of this tool with and
> without your modification to licensecheck.pl? Please make sure to always run
> with empty caches, "echo 3 > /proc/sys/vm/drop_caches".

[Paweł Hajdan Jr., 2011-11-02 18:25:04]

One suggestion I can have is: please split out the licensecheck change. It seems
independent from the other part.

[Paweł Hajdan Jr., 2011-11-02 18:28:31]

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
File tools/checklicenses/checklicenses.py (left):

http://codereview.chromium.org/8423038/diff/1/tools/checklicenses/checklicens...
tools/checklicenses/checklicenses.py:117: 'UNKNOWN',
On 2011/11/02 18:24:25, Sami wrote:
> Sure, here are the results.
> 
> Without the changes:
> 
> 0m36.454s
> 0m34.651s
> 0m37.063s
> 
> With the changes:
> 
> 0m36.915s
> 0m34.467s
> 0m37.470s

Awesome, the speed doesn't seem affected much. Thank you for doing this test. If
you could separate out this change, I'd be happy to give the review stamp.

[Sami, 2011-11-02 20:31:45]

On 2011/11/02 18:28:31, Paweł Hajdan Jr. wrote:
> Awesome, the speed doesn't seem affected much. Thank you for doing this test.
If
> you could separate out this change, I'd be happy to give the review stamp.

Thanks. Good idea about splitting the changes:
http://codereview.chromium.org/8351079

[Sami, 2011-11-10 12:13:34]

Anyone up for a review?

Thanks,
- Sami

[greggman, 2011-11-10 17:32:08]

My first question. Why are we doing this? It will make updating GPU code 3-4x
more tedious. 

Instead of 1 step 

1) Add extension needed for perf

now it will be

1) Add extension to third_party/gl2ext.h
2) Update DEPS
3) check it in
4) wait for it to be pushed to LGKR
5) write code waiting on steps 1-4

Are we doing this just because of some arbitrary rule someone wrote or because
there is a real need?

[Sami, 2011-11-11 11:04:30]

On 2011/11/10 17:32:08, greggman wrote:
> My first question. Why are we doing this? It will make updating GPU code 3-4x
> more tedious. 
> 
> Instead of 1 step 
> 
> 1) Add extension needed for perf
> 
> now it will be
> 
> 1) Add extension to third_party/gl2ext.h
> 2) Update DEPS
> 3) check it in
> 4) wait for it to be pushed to LGKR
> 5) write code waiting on steps 1-4
> 
> Are we doing this just because of some arbitrary rule someone wrote or because
> there is a real need?

This is really just motivated by the statement in
http://www.chromium.org/developers/adding-3rd-party-libraries that all third
party code must live in third_party. There is no functional difference either
way. However, I do think it is a little more organized to have all the headers
in a single place instead of sprinkling them around the source tree.

I'm not sure why you would need to update DEPS when touching gl2ext.h after this
change, since the headers are still in the same repository as before but just
under a different path.

If you think this is a waste of time and we should leave things as is, that's
fine too.

[greggman, 2011-11-11 18:00:00]

okay, thanks for clarifying that.

lgtm

[commit-bot: I haz the power, 2011-11-14 16:59:51]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@google.com/8423038/13001

[commit-bot: I haz the power, 2011-11-14 17:00:00]

Presubmit check for 8423038-13001 failed and returned exit status 1.

Running presubmit commit checks ...

** Presubmit ERRORS **
Missing LGTM from an OWNER for:
content/renderer/gpu/webgraphicscontext3d_command_buffer_impl.cc,ui/gfx/gl/generate_bindings.py,ui/gfx/gl/gl.gyp,ppapi/native_client/src/shared/ppapi_proxy/nacl.scons,ppapi/native_client/src/shared/ppapi_proxy/ppapi_proxy.gyp

Presubmit checks took 2.5s to calculate.

Was the presubmit check useful? Please send feedback & hate mail to
maruel@chromium.org!

[Sami (do not use), 2011-11-14 17:09:14]

On 2011/11/11 18:00:00, greggman wrote:
> okay, thanks for clarifying that.
> 
> lgtm

Cheers, thanks for taking a look.

apatrick and piman, would you mind commenting on the gpu/gl and ppapi parts?

[piman, 2011-11-14 17:32:04]

lgtm

[apatrick_chromium, 2011-11-18 17:38:24]

LGTM

[commit-bot: I haz the power, 2011-11-21 13:53:07]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@google.com/8423038/24001

[commit-bot: I haz the power, 2011-11-21 13:53:19]

Presubmit check for 8423038-24001 failed and returned exit status 1.

Running presubmit commit checks ...

** Presubmit ERRORS **
Missing LGTM from an OWNER for:
webkit/DEPS,webkit/glue/gl_bindings_skia_cmd_buffer.cc

Presubmit checks took 2.8s to calculate.

[tony, 2011-11-21 21:27:37]

webkit/glue LGTM (but you still need darin's approval for webkit/DEPS)

[darin (slow to review), 2011-11-21 22:45:09]

LGTM

[commit-bot: I haz the power, 2011-11-22 15:44:06]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@google.com/8423038/31001

[commit-bot: I haz the power, 2011-11-22 17:26:25]

Change committed as 111170

[commit-bot: I haz the power, 2011-11-24 12:09:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@google.com/8423038/36001

[commit-bot: I haz the power, 2011-11-24 13:29:39]

Change committed as 111523