gpu: Make sure external texture destination is initialized

gpu: Make sure external texture destination is initialized

When copying a texture from the GL_TEXTURE_EXTERNAL_OES target
(typically a video frame), we expect that the target size also defines
the source frame size. Add a check to make sure the target texture is
initialized when we do this.

BUG=225781, 239276
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=199256

[greggman, 2013-05-09 16:15:11]

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
gpu/command_buffer/service/gles2_cmd_decoder.cc:9630: if
(source_texture->target() == GL_TEXTURE_EXTERNAL_OES) {
Doesn't this code just need to move to about line 9665? The code at line 9641
will make the level exist

[Sami, 2013-05-09 16:18:42]

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
gpu/command_buffer/service/gles2_cmd_decoder.cc:9630: if
(source_texture->target() == GL_TEXTURE_EXTERNAL_OES) {
On 2013/05/09 16:15:11, greggman wrote:
> Doesn't this code just need to move to about line 9665? The code at line 9641
> will make the level exist

The problem is that for GL_TEXTURE_EXTERNAL, source_width and source_height are
both undefined so we can't really initialize the destination level directly. See
the comment on line 9579.

[greggman, 2013-05-09 16:41:23]

lgtm

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
gpu/command_buffer/service/gles2_cmd_decoder.cc:9631: if (!dest_level_defined) {
Then please add a TODO: here that this coude should be removed. The contact for
CopyTextureCHROMIUM is that it creates the level. The client shouldn't have to
know the size. The comment above even says they need to provide a way to get the
width/height

[Sami, 2013-05-09 16:46:24]

Thanks.

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/14858022/diff/1/gpu/command_buffer/service/gl...
gpu/command_buffer/service/gles2_cmd_decoder.cc:9631: if (!dest_level_defined) {
On 2013/05/09 16:41:23, greggman wrote:
> Then please add a TODO: here that this coude should be removed. The contact
for
> CopyTextureCHROMIUM is that it creates the level. The client shouldn't have to
> know the size. The comment above even says they need to provide a way to get
the
> width/height

Right, this is just a temporary safeguard while this code exist. I've duplicated
the TODO from above here.

[commit-bot: I haz the power, 2013-05-09 16:46:46]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/14858022/7002

[commit-bot: I haz the power, 2013-05-09 17:03:04]

Retried try job too often on ios_dbg_simulator for step(s) googleurl_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=ios_dbg_si...

[commit-bot: I haz the power, 2013-05-09 17:09:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/14858022/7002

[commit-bot: I haz the power, 2013-05-09 17:16:50]

Retried try job too often on ios_dbg_simulator for step(s) googleurl_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=ios_dbg_si...

[commit-bot: I haz the power, 2013-05-09 17:19:49]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/14858022/7002

[commit-bot: I haz the power, 2013-05-09 17:29:34]

Retried try job too often on ios_dbg_simulator for step(s) googleurl_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=ios_dbg_si...

[commit-bot: I haz the power, 2013-05-09 18:06:09]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/14858022/7002

[commit-bot: I haz the power, 2013-05-09 18:14:28]

Retried try job too often on ios_dbg_simulator for step(s) googleurl_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=ios_dbg_si...

[commit-bot: I haz the power, 2013-05-09 18:52:23]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/14858022/7002

[commit-bot: I haz the power, 2013-05-09 19:25:40]

Change committed as 199256