Handle compressed textures allocated via TexStorage2D.

Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel

Committed: https://crrev.com/5e216d9b62275101c30f4a5cb2174bb68970ef3c
Cr-Commit-Position: refs/heads/master@{#391139}

[Ken Russell (switch to Gerrit), 2016-04-28 14:15:32]

Description was changed from

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
==========

to

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.win:win_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel
==========

[Ken Russell (switch to Gerrit), 2016-04-28 14:15:32]

Description was changed from

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.win:win_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel
==========

to

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel
==========

[Ken Russell (switch to Gerrit), 2016-04-28 14:25:16]

kbr@chromium.org changed reviewers:
+ zmo@chromium.org

[Ken Russell (switch to Gerrit), 2016-04-28 14:25:17]

Mo: please review.

The test changes in https://github.com/KhronosGroup/WebGL/pull/1641 cover the
new code paths.

Once that pull request is rolled in and the WebGL 1.0 tests are run as part of
the WebGL 2.0 conformance suite, this will be covered on the waterfall and CQ.

[Zhenyao Mo, 2016-04-28 17:57:13]

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/comm...
File gpu/command_buffer/common/gles2_cmd_utils.h (right):

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/comm...
gpu/command_buffer/common/gles2_cmd_utils.h:149: static bool
IsCompressedTextureFormat(int internal_format);
You don't have to redefined this function.  We already have
FeatureInfo::validators()->compressed_texture_format, which contains all the
allowed compressed texture formats. You can just use it in the decoder. It's
better if in the future when we add new extensions, we don't have to also update
the duplicated table here.

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10435: // allocated via
TexStorage2D.
nit: can you also add a comment here that although we do expose TexStorage2D
internally for ES2, but compressed formats are not part of it?

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10450:
glBindTexture(texture->target(), texture->service_id());
You also need to unbind the PIXEL_UNPACK_BUFFER if one is bound, and recover the
binding in the end of this function.

[Ken Russell (switch to Gerrit), 2016-04-29 13:10:06]

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/comm...
File gpu/command_buffer/common/gles2_cmd_utils.h (right):

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/comm...
gpu/command_buffer/common/gles2_cmd_utils.h:149: static bool
IsCompressedTextureFormat(int internal_format);
On 2016/04/28 17:57:12, Zhenyao Mo wrote:
> You don't have to redefined this function.  We already have
> FeatureInfo::validators()->compressed_texture_format, which contains all the
> allowed compressed texture formats. You can just use it in the decoder. It's
> better if in the future when we add new extensions, we don't have to also
update
> the duplicated table here.

Thanks for pointing this out. It's necessary to call this both from the
GLES2Decoder as well as the TextureManager, but since the TextureManager already
accesses the decoder, it's best to just put it in the decoder. Done.

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10435: // allocated via
TexStorage2D.
On 2016/04/28 17:57:13, Zhenyao Mo wrote:
> nit: can you also add a comment here that although we do expose TexStorage2D
> internally for ES2, but compressed formats are not part of it?

Done.

https://codereview.chromium.org/1925093002/diff/20001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10450:
glBindTexture(texture->target(), texture->service_id());
On 2016/04/28 17:57:13, Zhenyao Mo wrote:
> You also need to unbind the PIXEL_UNPACK_BUFFER if one is bound, and recover
the
> binding in the end of this function.

Thanks for pointing that out. Done.

[Zhenyao Mo, 2016-04-29 17:15:16]

https://codereview.chromium.org/1925093002/diff/40001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/1925093002/diff/40001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10573:
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
This should be bound_buffer->service_id()

[Zhenyao Mo, 2016-04-29 17:22:17]

lgtm once the above mentioned bug is fixed.

[Ken Russell (switch to Gerrit), 2016-05-02 20:50:55]

https://codereview.chromium.org/1925093002/diff/40001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/1925093002/diff/40001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:10573:
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
On 2016/04/29 17:15:16, Zhenyao Mo wrote:
> This should be bound_buffer->service_id()

Thank you for catching that and apologies for not having written tests covering
this specific code path. Fixed.

[Ken Russell (switch to Gerrit), 2016-05-02 20:51:01]

The CQ bit was checked by kbr@chromium.org

[Ken Russell (switch to Gerrit), 2016-05-02 20:51:01]

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/1925093002/#ps60001
(title: "Fixed bug pointed out by zmo.")

[commit-bot: I haz the power, 2016-05-02 20:51:20]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1925093002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1925093002/60001

[commit-bot: I haz the power, 2016-05-02 21:28:37]

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

[commit-bot: I haz the power, 2016-05-02 21:28:38]

Try jobs failed on following builders:
  cast_shell_linux on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/cast_shell_linu...)

[Ken Russell (switch to Gerrit), 2016-05-02 22:02:49]

The CQ bit was checked by kbr@chromium.org

[Ken Russell (switch to Gerrit), 2016-05-02 22:02:49]

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/1925093002/#ps80001
(title: "Fixed gpu_unittests after last refactor.")

[commit-bot: I haz the power, 2016-05-02 22:03:37]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1925093002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1925093002/80001

[commit-bot: I haz the power, 2016-05-03 01:15:27]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-05-03 01:18:07]

Description was changed from

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel
==========

to

==========
Handle compressed textures allocated via TexStorage2D.

Compressed internal formats should be allowed as arguments to
TexStorage2D. Textures allocated in this way are also cleared lazily,
either just before the first CompressedTexSubImage2D if it does not
cover the entire mipmap level, or if they're sampled.

https://github.com/KhronosGroup/WebGL/pull/1641 expands the
webgl-compressed-texture-s3tc.html conformance test to cover this
behavior.

BUG=605129
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel

Committed: https://crrev.com/5e216d9b62275101c30f4a5cb2174bb68970ef3c
Cr-Commit-Position: refs/heads/master@{#391139}
==========

[commit-bot: I haz the power, 2016-05-03 01:18:08]

Patchset 5 (id:??) landed as
https://crrev.com/5e216d9b62275101c30f4a5cb2174bb68970ef3c
Cr-Commit-Position: refs/heads/master@{#391139}