- 29 Jan, 2014 1 commit
-
-
Jamie Madill authored
With certain selection statements with a vertex input in the condition and side-effects in the else-block, we'd run in to a D3D9 compiler bug which would cause incorrect results. We can work around this bug in D3D9 by selectively rewriting these statements to not use an else clause. BUG=322794 Change-Id: I93c96fb201ff4959c00d9a36321faac7e0343278 Reviewed-on: https://chromium-review.googlesource.com/184021Reviewed-by:
Nicolas Capens <nicolascapens@chromium.org> Commit-Queue: Nicolas Capens <nicolascapens@chromium.org> Tested-by:
Jamie Madill <jmadill@chromium.org>
-
- 27 Jan, 2014 1 commit
-
-
Nicolas Capens authored
Remove the ANGLE_COMPILE_OPTIMIZATION_LEVEL precompiler definition override to default to D3DCOMPILE_OPTIMIZATION_LEVEL3. BUG=322795 R=geofflang@chromium.org Change-Id: I7d8a2676c3510d26e9f0876aa30c6d8b4ae89f24 Reviewed-on: https://chromium-review.googlesource.com/183995Tested-by:
Nicolas Capens <nicolascapens@google.com> Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Nicolas Capens <nicolascapens@google.com>
-
- 23 Jan, 2014 1 commit
-
-
Geoff Lang authored
Change-Id: Ia0b8b0e67a0af5eeeec6e3fc7bfdbfdc70f519f8 Reviewed-on: https://chromium-review.googlesource.com/183161Reviewed-by:
Shannon Woods <shannonwoods@chromium.org> Reviewed-by:
-
- 16 Jan, 2014 2 commits
-
-
Geoff Lang authored
Change-Id: I424cb538808ec64137c85af3bbe8742043e1c573 Reviewed-on: https://chromium-review.googlesource.com/182840Reviewed-by:
-
Geoff Lang authored
The gyp files already define NOMINMAX and WIN32_LEAN_AND_MEAN. windows.h is also included via the D3D includes. This ends up speeding up the build process significantly by not having windows.h included in nearly every file. Also fixes issues on non-windows systems by not having stdarg.h included in debug.cpp. Change-Id: Id32d9aac37a5a29df832a062dd8024302a798a1c Reviewed-on: https://chromium-review.googlesource.com/182810Reviewed-by:
-
- 15 Jan, 2014 3 commits
-
-
Geoff Lang authored
Changed ANGLE_DISABLE_* to ANGLE_ENABLE_* for perf and tracing defines so they are disabled by default. Updated the gyp files to only turn on perf by default for windows debug builds. Change-Id: I71706674e6d12fbf4208acc8f100d963b82c7674 Reviewed-on: https://chromium-review.googlesource.com/182146Reviewed-by:
-
Geoff Lang authored
Change-Id: I73dbb08796c3e3a7a95f36dbf378b37eb9f936a7 Reviewed-on: https://chromium-review.googlesource.com/182145Reviewed-by:
-
Geoff Lang authored
Change-Id: I8cf650073ead5b52bac44e31c8c3c894bc31e01c Reviewed-on: https://chromium-review.googlesource.com/182156Reviewed-by:
-
- 13 Jan, 2014 1 commit
-
-
Geoff Lang authored
Change-Id: I7ecdb049352af0dd94e8fbaede5953dfad9a75ef Reviewed-on: https://chromium-review.googlesource.com/181903Reviewed-by:
-
- 10 Jan, 2014 2 commits
-
-
Geoff Lang authored
Change-Id: Id8bf2b9cf44f9fcd79eff774db64cb59b9aebd73 Reviewed-on: https://chromium-review.googlesource.com/178786Reviewed-by:
-
Geoff Lang authored
BUG=angle:537 Change-Id: I146cb04eb69fc8df152f8e6cd9f78f5e87e72a7e Reviewed-on: https://chromium-review.googlesource.com/178785Reviewed-by:
-
- 06 Jan, 2014 2 commits
-
-
Jamie Madill authored
The GL spec states that by default a vertex attribute has a size of 4. Change-Id: Ic4c977fb936cc0b232ed4742eeb061d6ba6843be Reviewed-on: https://chromium-review.googlesource.com/181522Tested-by:
-
Shannon Woods authored
BUG=angle:428 Change-Id: I282a6bf138b4e2a2c17baf428cbe1e636f43bdd6 Reviewed-on: https://chromium-review.googlesource.com/181640Reviewed-by:
-
- 03 Jan, 2014 2 commits
-
-
Nicolas Capens authored
Create a fake variable for function identifiers used as a variable identifier to allow continuing compilation after reporting the error. BUG=chromium:330222 Change-Id: I553804b3db95cd5dc946e94193fc9181d6b2c032 Reviewed-on: https://chromium-review.googlesource.com/180977Reviewed-by:
-
Nicolas Capens authored
The next patch requires a parser grammar change, but Cygwin's Bison version got updated so this separates the changes caused by that update from the ones for the actual bug fix. Change-Id: I1a2d58433a65488ffe0b94e1760c716e14496e8f Reviewed-on: https://chromium-review.googlesource.com/180976Reviewed-by:
-
- 02 Jan, 2014 1 commit
-
-
Brendan Long authored
BUG=536 Change-Id: I72287f5ce9736cd48fc241022e09446db717cf92 Reviewed-on: https://chromium-review.googlesource.com/179260Reviewed-by:
-
- 18 Dec, 2013 2 commits
-
-
Jamie Madill authored
A missing condition would produce a sort order where sometimes -1 was shifted to the front of the semantic array, instead of the end. BUG=angle:527 Change-Id: I69b2e5ccc03f6523771601cd59293d6cd325be2f Reviewed-on: https://chromium-review.googlesource.com/180660Reviewed-by:
-
Jamie Madill authored
We would print non-casted unsigned char output, giving an an ascii code on output, instead of printing the integer value of the size field. BUG=angle:539 Change-Id: I963c297d01df6283825d684ada7a34fd535f2550 Reviewed-on: https://chromium-review.googlesource.com/180650Reviewed-by:
-
- 12 Dec, 2013 1 commit
-
-
Jamie Madill authored
This fixes device remove handling on AMD systems, as well as notifying the app layer there may have been an internal problem during Present. BUG=313210 Change-Id: I843ffb3946dd9ef34fbd24b3a46a7b97e3c0fd16 Reviewed-on: https://chromium-review.googlesource.com/179140Reviewed-by:
-
- 10 Dec, 2013 1 commit
-
-
Jamie Madill authored
We were using the first type as the type for each declaration, which could cause erros for arrays, eg "float one[4], two;" BUG=angle:431 TEST=WebGL CTS 1.0.2 Change-Id: I104d6701cc019ae3b3831a2d568ffe2b656b45a2 Reviewed-on: https://chromium-review.googlesource.com/179350Reviewed-by:
-
- 09 Dec, 2013 2 commits
-
-
Jamie Madill authored
When an app would load libGLESv2.dll via LoadModule, then would call methods on a thread that was already created, we could creash because of uninitialized thread local storage. BUG=angle:488 Change-Id: I9b05de462232a16d639c58ec07386b7c38a01793 Reviewed-on: https://chromium-review.googlesource.com/179197Reviewed-by:
-
Jamie Madill authored
When an app would load libEGL.dll via LoadModule, then would call methods on a thread that was already created, we could creash because of uninitialized thread local storage. BUG=angle:488 Change-Id: If3f73fb59dcc5f0713f5b6fb9ca9f8535ee23c73 Reviewed-on: https://chromium-review.googlesource.com/179196Reviewed-by:
-
- 06 Dec, 2013 1 commit
-
-
Geoff Lang authored
bug=angle:535 Change-Id: I5b86874cddbd3b90fe141e94085f5a4afb9f3db3 Reviewed-on: https://chromium-review.googlesource.com/178891Reviewed-by:
-
- 03 Dec, 2013 3 commits
-
-
Jamie Madill authored
BUG=angle:519 Change-Id: I3d3d0746b2b3027d04c91cc5fdcd301930e3b22a Reviewed-on: https://chromium-review.googlesource.com/178455Reviewed-by:
-
Jamie Madill authored
BUG=angle:534 Change-Id: I893e552628c7deba4bc67baae6e5e28b6161c5d4 Reviewed-on: https://chromium-review.googlesource.com/178454Reviewed-by:
-
Jamie Madill authored
This could happen because we reset the dirty swap interval flag before we called makeCurrent. BUG=angle:481 Change-Id: I5013265b9669ed5055ba8d777d0fcd5588f30b10 Reviewed-on: https://chromium-review.googlesource.com/178582Reviewed-by:
-
- 26 Nov, 2013 1 commit
-
-
Jamie Madill authored
immediately allow recreating a new device. On Intel systems there's a cool-down period when you must wait before recreating the device. BUG=313210 Change-Id: I22cdf41eecf0eae5058b59e1b0863173d604d606
-
- 25 Nov, 2013 1 commit
-
-
Jamie Madill authored
for D3D9. This prevents D3D errors which we could see when using the Debug runtime. BUG=321834 Change-Id: I1496366117c7642199a941fbccf94dee04c78d8b
-
- 22 Nov, 2013 1 commit
-
-
Shannon Woods authored
Change-Id: I8e61d31d180dcb4e79e276525b2afd1e17e70ba5
-
- 20 Nov, 2013 3 commits
-
-
Jamie Madill authored
-
Geoff Lang authored
Change-Id: Id1bb71d997e710028759a9e9a2e0b24e35f92fc4
-
Geoff Lang authored
Change-Id: I49845a0a4e660027dd817b7a783163c1aa5d04d7
-
- 19 Nov, 2013 5 commits
-
-
Jamie Madill authored
Add a build step to copy the D3DCompiler_46 DLL from the Windows SDK folder to the build products path. BUG=315387 R=zmo@chromium.org,shannonwoods@chromium.org Change-Id: If9a149a6f780e78451c11008b0f7bfd4dd39b582
-
Nicolas Capens authored
BUG=angle:450 Change-Id: If7be5f72e1037d92256edf45e1cfc7fedbf0b4bd
-
Ehsan Akhgari authored
This will make it possible to build angle in unified mode for Mozilla. Change-Id: Iba4b971392b8abae8b5ff8f96e9172582f43de90
-
Jamie Madill authored
BUG=313210 Change-Id: I6b7e34466e89e7854c8797cbdcffc4a24c2c32f5
-
Ehsan Akhgari authored
Change-Id: Ied066feb2b8954dc7f0398704cd3429780549989
-
- 07 Nov, 2013 1 commit
-
-
Jamie Madill authored
The old modifiesState method really checked if an operator was an assignment, so restored that behaviour and use the new side effects detection only for the new code. ANGLEBUG=486 BUG= R=nicolascapens@chromium.org, zmo@chromium.org Review URL: https://codereview.appspot.com/22130043
-
- 04 Nov, 2013 1 commit
-
-
Jamie Madill authored
short-circuiting conditionals for expressions with potential side-effects. Conservatively assume aggreate and selection operators have side effects for now. BUG= ANGLEBUG=486 R=geofflang@chromium.org, kbr@chromium.org, nicolas@transgaming.com, shannonwoods@chromium.org Review URL: https://codereview.appspot.com/14441075
-
- 01 Nov, 2013 1 commit
-
-
Jamie Madill authored
InfoSink.h needs stdlib.h for abs(int) and free() in the global namespace. ExpressionParser needs malloc.h, because bison needs malloc and free in the global namespace, but "#include <cassert>" will put it only in the std:: namespace on QNX. BUG=500 R=geofflang@chromium.org, jmadill@chromium.org, shannonwoods@chromium.org, zmo@chromium.org Review URL: https://codereview.appspot.com/19330044
-
