GitLab

trying to write to a socket.

Our old behavior caused network failures on servers that hang
up right after sending a file, but don't use
"connection: close".

BUG=105824

Review URL: http://codereview.chromium.org/8775044

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113496 0039d316-1c4b-4281-b951-d872f2087c98

The "Languages and Settings" menu item should not be present in the first place in the spellcheck context menu on OS X.

I'll remove the menu item in a followup CL but in the meantime we can remove the code that specifically disables the command on Mac. In the interim having it enabled doesn't do any harm.

BUG=69944
TEST=None


Review URL: http://codereview.chromium.org/8850006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113495 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: http://codereview.chromium.org/8571023

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113494 0039d316-1c4b-4281-b951-d872f2087c98

Works fine now when I tried 20 iterations.

R=dennisjeffrey@chromium.org
BUG=91225
TEST=

Review URL: http://codereview.chromium.org/8863005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113493 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
TEST=

Review URL: http://codereview.chromium.org/8863003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113492 0039d316-1c4b-4281-b951-d872f2087c98

This change has tendrils stretching throughout the code, but mostly this lets us delete a bunch of stuff.

This also does a slight refactor to put the PPB_Var implementations in 1 place, since they were already practically identical, and I didn't want to do the backwards-compat code in 2 places.

BUG=106596
TEST=N/A
TBR=darin,tony

darin,tony TBR for rubber-stamp of webkit/glue/webkit_glue.gypi

Review URL: http://codereview.chromium.org/8826011

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113491 0039d316-1c4b-4281-b951-d872f2087c98

BUG=106043
TEST=
R=rnk@chromium.org


Review URL: http://codereview.chromium.org/8806028

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113488 0039d316-1c4b-4281-b951-d872f2087c98

Currently we assume that any profile picture URL that contains "AAAAAAAAAAI/AAAAAAAAAAA" represents a default profile picture.

It turns out that this is incorrect. There are valid profile pictures with that URL.

With the OAuth user info API we can simply detect default profile pictures by the existance of the "picture" field in the returned data. Accounts with a default profile picture don't have this field.

BUG=106444
TEST=Compiled and verified that unit tests passed.


Review URL: http://codereview.chromium.org/8801023

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113487 0039d316-1c4b-4281-b951-d872f2087c98

BUG=chromium-os:23764
TEST=Log in as guest, attempt to lock screen, nothing should happen.


Review URL: http://codereview.chromium.org/8842002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113485 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113484 0039d316-1c4b-4281-b951-d872f2087c98

Revision: 904
Author: apatrick@chromium.org
Date: 11:11:33 AM, Wednesday, December 07, 2011
Message:
Update version to 904.
Review URL: http://codereview.appspot.com/5449113
----
Modified : /trunk/src/common/version.h

Revision: 903
Author: apatrick@chromium.org
Date: 11:10:16 AM, Wednesday, December 07, 2011
Message:
Fixed leak of render target associated with ColorBuffer.

Following r853, I am assuming that any render target returned by FrameBuffer::getRenderTarget has to be released. Updating the usage of getRenderTarget to reflect that.

This is not reproducible in Chrome at the moment but it will be once I start rendering to pbuffers.

I ran the WebGL conformance tests. No regressions or crashes.
Review URL: http://codereview.appspot.com/5451110
----
Modified : /trunk/src/libGLESv2/Context.cpp

Revision: 902
Author: apatrick@chromium.org
Date: 6:05:22 PM, Monday, December 05, 2011
Message:
Implemented EGL_NV_post_sub_buffer.

The change to the sample is just to demonstrate how I tested it. I won't check in the stuff under samples/.

It's not clear to me what eglPostSubBufferNV should do if EGL_POST_SUB_BUFFER_SUPPORTED_NV is false. At the moment it fails silently as though it was called on the wrong surface type:

 "If <surface> is a back-buffered surface, then the requested portion
    of the color buffer is copied to the native window associated with
    that surface. If <surface> is a single-buffered window, pixmap, or
    pbuffer surface, eglSwapBuffers and eglPostSubBufferNV have no
    effect."
Review URL: http://codereview.appspot.com/5345050
----
Modified : /trunk/samples/build_samples.gyp
Modified : /trunk/samples/gles2_book/Common/esUtil.c
Modified : /trunk/samples/gles2_book/Common/esUtil.h
Added : /trunk/samples/gles2_book/PostSubBuffer
Added : /trunk/samples/gles2_book/PostSubBuffer/PostSubBuffer.c
Added : /trunk/samples/gles2_book/PostSubBuffer/PostSubBuffer.vcproj
Modified : /trunk/samples/samples.sln
Modified : /trunk/src/libEGL/Display.cpp
Modified : /trunk/src/libEGL/Surface.cpp
Modified : /trunk/src/libEGL/Surface.h
Modified : /trunk/src/libEGL/libEGL.cpp

Revision: 901
Author: daniel@transgaming.com
Date: 11:42:32 AM, Tuesday, November 29, 2011
Message:
Creates helper function for common subimage param checking

TRAC #18802
Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
Author: Shannon Woods
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/libGLESv2/libGLESv2.cpp

Revision: 900
Author: daniel@transgaming.com
Date: 11:42:28 AM, Tuesday, November 29, 2011
Message:
Adds missing parameter checks to texture functions

TRAC #18802
Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
Author: Shannon Woods
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/libGLESv2/libGLESv2.cpp

Revision: 899
Author: daniel@transgaming.com
Date: 11:42:23 AM, Tuesday, November 29, 2011
Message:
Adds a level parameter to texture getWidth and getHeight

TRAC #18802
Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
Author: Shannon Woods
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/libGLESv2/Renderbuffer.cpp
Modified : /trunk/src/libGLESv2/Texture.cpp
Modified : /trunk/src/libGLESv2/Texture.h
Modified : /trunk/src/libGLESv2/libGLESv2.cpp

Revision: 898
Author: daniel@transgaming.com
Date: 9:23:51 AM, Tuesday, November 29, 2011
Message:
Checks parameters of matrices for under-fullness in shader constructor check.

TRAC #18906
Issue=218
Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/compiler/ParseHelper.cpp

Revision: 897
Author: daniel@transgaming.com
Date: 9:21:42 AM, Tuesday, November 29, 2011
Message:
Fix post-build steps when spaces appear in directory names

TRAC #18901
Issue=245
Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
----
Modified : /trunk/samples/gles2_book/MultiTexture/MultiTexture.vcproj
Modified : /trunk/samples/gles2_book/ParticleSystem/ParticleSystem.vcproj
Modified : /trunk/src/common/version.h

Revision: 895
Author: daniel@transgaming.com
Date: 2:34:26 PM, Thursday, November 24, 2011
Message:
Update major version.
----
Modified : /trunk/src/common/version.h

Revision: 894
Author: daniel@transgaming.com
Date: 2:34:22 PM, Thursday, November 24, 2011
Message:
Add another contributor
----
Modified : /trunk/CONTRIBUTORS
Modified : /trunk/src/common/version.h

Revision: 893
Author: daniel@transgaming.com
Date: 2:34:18 PM, Thursday, November 24, 2011
Message:
Update contributors
----
Modified : /trunk/CONTRIBUTORS
Modified : /trunk/src/common/version.h

Revision: 892
Author: daniel@transgaming.com
Date: 2:34:14 PM, Thursday, November 24, 2011
Message:
Fix rendering for the closing line of line loops.

Trac #19035
Signed-off-by: Nicolas Capens

Line loops via DrawArrays were incorrect because the vertex buffer already
has the vertex offset factored in.
Line loops via DrawElements were incorrect because the vertex buffer was
offset by the minIndex. We need to use the same minIndex when rendering
the closing loop.
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/libGLESv2/Context.cpp
Modified : /trunk/src/libGLESv2/Context.h

Revision: 891
Author: daniel@transgaming.com
Date: 2:34:09 PM, Thursday, November 24, 2011
Message:
Avoid rendering the last pixel of lines.

Trac #19035
Issue=257
Author: Yuriy O'Donnell

Signed-off-by: Nicolas Capens
Signed-off-by: Daniel Koch
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/libEGL/Display.cpp

Revision: 890
Author: daniel@transgaming.com
Date: 2:34:04 PM, Thursday, November 24, 2011
Message:
Fix HLSL translation for mod(vec2,vec2)

ANGLEBUG=258
Signed-off-by: Daniel Koch
Signed-off-by: Nicolas Capens
Author: Sam Hocevar
----
Modified : /trunk/src/common/version.h
Modified : /trunk/src/compiler/OutputHLSL.cpp
Modified : /trunk/src/compiler/OutputHLSL.h


Review URL: http://codereview.chromium.org/8856011

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113483 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: http://codereview.chromium.org/8836001

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113482 0039d316-1c4b-4281-b951-d872f2087c98

I had fixed it in r111201

R=dennisjeffrey@chromium.org
BUG=104886
TEST=

Review URL: http://codereview.chromium.org/8866001

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113481 0039d316-1c4b-4281-b951-d872f2087c98

The old ProfileManagerObserver was hard to use since it required a new class or subclassing. Using callbacks is easier.

BUG=100461
TEST=

Review URL: http://codereview.chromium.org/8840003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113480 0039d316-1c4b-4281-b951-d872f2087c98

This reverts commit bd45bd25.

TEST=ran webkit tests in DRT and webgl tests in chrome
BUG=103989

Review URL: http://codereview.chromium.org/8758026

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113479 0039d316-1c4b-4281-b951-d872f2087c98

I think that r113138 and r113167 combined made it too light

BUG=106669
TEST=manual

Review URL: http://codereview.chromium.org/8833006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113478 0039d316-1c4b-4281-b951-d872f2087c98

BUG=88782
TEST=none


Review URL: http://codereview.chromium.org/8834003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113477 0039d316-1c4b-4281-b951-d872f2087c98

Change did not help.

TBR=dtseng

BUG=
TEST=

Review URL: http://codereview.chromium.org/8824009

TBR=dtseng@chromium.org
Review URL: http://codereview.chromium.org/8863001

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113476 0039d316-1c4b-4281-b951-d872f2087c98

to be as wide as possible to cover all variants of this leak.

Remove the suppression for the leak of bug 95718 because it is a
variant of the above leak.

R=cbentzel@chromium.org
BUG=77990
TEST=none

Review URL: http://codereview.chromium.org/8823019

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113474 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: http://codereview.chromium.org/8803022

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113473 0039d316-1c4b-4281-b951-d872f2087c98

The first bit is designed to ensure that the correctness of callback_forward.h is tested by the callback unittests. The second bit is designed to increase the likelihood of developers actually using callback_forward.h as intended.

BUG=None
TEST=None


Review URL: http://codereview.chromium.org/8846003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113472 0039d316-1c4b-4281-b951-d872f2087c98

These were fixed by my earlier r112315

BUG=102767
TEST=None

R=dennisjeffrey@chromium.org

Review URL: http://codereview.chromium.org/8856008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113471 0039d316-1c4b-4281-b951-d872f2087c98

Makes tests either use mock compositor or mock WebGraphicsContext3D depending upon which compositor we're running. This is needed to enable ui tests on the bots.

BUG=104360
TEST=none

Review URL: http://codereview.chromium.org/8805033

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113470 0039d316-1c4b-4281-b951-d872f2087c98

HTMLConstructionSite::insertText changed its signature.

BUG=105961
TBR=mikelawther@chromium.org

Review URL: http://codereview.chromium.org/8852010

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113469 0039d316-1c4b-4281-b951-d872f2087c98

Also move NTPPromoShown to notification_promo.cc

BUG=106743
TEST=NONE.

Review URL: http://codereview.chromium.org/8833009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113468 0039d316-1c4b-4281-b951-d872f2087c98

Return buffers that are in flight during shutdown and make sure to issue 
defered stop.

BUG=106369
Review URL: http://codereview.chromium.org/8819010

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113467 0039d316-1c4b-4281-b951-d872f2087c98

these tests to using manifest_version 2, but it looks like more effort than
it's worth.

BUG=62897
Review URL: http://codereview.chromium.org/8825015

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113466 0039d316-1c4b-4281-b951-d872f2087c98

Don't attempt to forward declare StringPiece. #include the declaration, or remove the forward declaration where it was not actually used.

1) Forward declaring StringPiece is generally discouraged because it prevents callers from benefiting from automatic coersion from string/char* types.
2) A follow-up CL (http://codereview.chromium.org/8659047/) will make StringPiece a template, and thus awkward to forward declare. The very small number of places that were appropriately forward declaring it do not justify writing a 'string_piece_forward.h'.

This particular CL would be one of the rare ones where a forward declaration would be 'appropriate', but it does not warrant forward-declaring both the template and the typedef, and the number of similar cases does not justify creating a _forward.h header.

BUG=87634
R=tony@chromium.org

Review URL: http://codereview.chromium.org/8819022

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113465 0039d316-1c4b-4281-b951-d872f2087c98

BUG=43460
R=mirandac@chromium.org

Review URL: http://codereview.chromium.org/8848003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113464 0039d316-1c4b-4281-b951-d872f2087c98

This is cleaner and won't cause spurious crash reports.


BUG=
TEST=


Review URL: http://codereview.chromium.org/8823005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113462 0039d316-1c4b-4281-b951-d872f2087c98

BUG=104783
TBR=acolwell

Review URL: http://codereview.chromium.org/8849009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113461 0039d316-1c4b-4281-b951-d872f2087c98

BUG=82717
TBR=ananta@chromium.org

Review URL: http://codereview.chromium.org/8849008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113459 0039d316-1c4b-4281-b951-d872f2087c98

Fixing background mode manager to correctly display the name of the profile in the status icon. Previously it was not being updated when Chrome first started up and it was also using the incorrect name. Now it gets the correct name from the profile info cache and also updates based on changes to the profile info cache.

BUG=101837
TEST=start chrome with background apps and the status icon will have names populated correctly, also BackgroundModeManagerTest::ProfileInfoCacheObserver

Review URL: http://codereview.chromium.org/8802013

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113458 0039d316-1c4b-4281-b951-d872f2087c98

This CL removes Hunspell-related options from the spelling context menu on the Mac and tries to bring the menu more inline with the native platform menu.

Notably:
* Spelling submenu observer is split into Hunspell and Mac implementation files.
* strings are split into mac/non-mac variants (saving a bit of space on non-mac platforms?) and the titles are brought more in line with OS X (probably need more work to make this match on all OS X versions).
* Menu is reordered on Mac to more closely match the Cocoa spelling submenu. 

BUG=69944
TEST=Spelling submenu should continue to work on non-Mac platforms. On Mac it should contain 3 items - "Show Spelling and Grammar", "Check Spelling while you Type" and "Language Settings..."


Review URL: http://codereview.chromium.org/8823009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113457 0039d316-1c4b-4281-b951-d872f2087c98

Set mask after sending modifier keys.
Print unknown keysym in hex and add warning for unknown keycode.

BUG=106630
TEST=All BrowserKeyEventsTest except for AccessKey will pass


Review URL: http://codereview.chromium.org/8836007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113456 0039d316-1c4b-4281-b951-d872f2087c98

The largest sources of flake should now be fixed or supressed.

BUG= none
TEST= none

Review URL: http://codereview.chromium.org/8823014

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113455 0039d316-1c4b-4281-b951-d872f2087c98

When diff-ing two snapshots, do the merge *before* doing the diff. In practice this doesn't usually yield different results, but it is possible. Merging before diffing makes more sense though, and avoids some duplicated work.

BUG=105301
Review URL: http://codereview.chromium.org/8676023

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113454 0039d316-1c4b-4281-b951-d872f2087c98

Older gcc versions apparently do not support this when used with STL.

Review URL: http://codereview.chromium.org/8771061

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113451 0039d316-1c4b-4281-b951-d872f2087c98

BUG=chromium-os:23871
TEST=None

Review URL: http://codereview.chromium.org/8858001

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113448 0039d316-1c4b-4281-b951-d872f2087c98

BUG=chromium-os:21845
TEST=Shut down option should be present at the bottom of the wrench menu on any cros device.

Review URL: http://codereview.chromium.org/8834006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@113447 0039d316-1c4b-4281-b951-d872f2087c98