Punted on bison parser. It was turning out to be more complicated than necessary. Manual parsing combined with a flex lexer is easier.

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1012 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/5059048

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1011 736b8ea6-26fd-11df-bfd4-992fa37f6226

Issue=156
TRAC #16816
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1010 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20059
ISSUE=290
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1009 736b8ea6-26fd-11df-bfd4-992fa37f6226

I think the assumption that getDisplay() returns a valid display in the Fence destructor is wrong.

I'm trying to fix a crash in the field that looks like this:

Thread 0 *CRASHED* ( EXCEPTION_ACCESS_VIOLATION_READ @ 0x00000000 )

0x69582e38	 [libglesv2.dll	 - fence.cpp:27	gl::Fence::~Fence()
0x69582f29	 [libglesv2.dll	 + 0x00022f29]	gl::Fence::`scalar deleting destructor'(unsigned int)
0x6958077d	 [libglesv2.dll	 - context.cpp:1020	gl::Context::deleteFence(unsigned int)
0x69582b9b	 [libglesv2.dll	 - context.cpp:195	gl::Context::~Context()
0x69582dcb	 [libglesv2.dll	 + 0x00022dcb]	gl::Context::`scalar deleting destructor'(unsigned int)
0x69582df2	 [libglesv2.dll	 - context.cpp:4259	glDestroyContext
0x73166ab8	 [libegl.dll	 - display.cpp:768	egl::Display::destroyContext(gl::Context *)
0x73168393	 [libegl.dll	 - libegl.cpp:861	eglDestroyContext
0x6e18f1db	 [chrome.dll	 - gl_context_egl.cc:76	gfx::GLContextEGL::Destroy()
0x6e18f40d	 [chrome.dll	 - gl_context_egl.cc:43	gfx::GLContextEGL::~GLContextEGL()

Here's the disassembly:

69582E21  push        esi  
69582E22  mov         esi,ecx 
69582E24  cmp         dword ptr [esi+4],0 
69582E28  mov         dword ptr [esi],695CBBE0h 
69582E2E  je          69582E3F 
69582E30  call        695743F5 // this is getDisplay()
69582E35  push        dword ptr [esi+4] 
69582E38  mov         edx,dword ptr [eax] // crashes here because EAX is zero
69582E3A  mov         ecx,eax 
69582E3C  call        dword ptr [edx+24h] // this is freeEventQuery()
69582E3F  pop         esi  
69582E40  ret   

It looks like getDisplay() returns null.

http://code.google.com/p/chromium/issues/detail?id=117817
Review URL: https://codereview.appspot.com/5875044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1008 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20055
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1007 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20057
ISSUE=285
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1006 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20056
ISSUE=256
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1005 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20056
ISSUE=256
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1004 736b8ea6-26fd-11df-bfd4-992fa37f6226

Detect when declaring previously undeclared variables as invariant, and avoid using an uninitialized type.

TRAC #20042
ISSUE=241
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1002 736b8ea6-26fd-11df-bfd4-992fa37f6226

Start a new scope for conditional statements even when there are no curly brackets. Also fix the DO WHILE grammar.

TRAC #20052
ISSUE=252,255
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1001 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #20052
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@1000 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19197
ISSUE=260
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@999 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19988
Issue=304
Signed-off-by: Daniel Koch

git-svn-id: https://angleproject.googlecode.com/svn/trunk@998 736b8ea6-26fd-11df-bfd4-992fa37f6226

This is a quick fix for gclient runhooks error on Mac/Linux.

BUG=
TEST=gclient runhooks --force on Mac/Linux
Review URL: https://codereview.appspot.com/5732047

git-svn-id: https://angleproject.googlecode.com/svn/trunk@997 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19989
Issue=302
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@996 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19988
Issue=304
Signed-off-by: Daniel Koch

git-svn-id: https://angleproject.googlecode.com/svn/trunk@995 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@994 736b8ea6-26fd-11df-bfd4-992fa37f6226

This could hurt performance quite a bit.

BUG=296
TEST=

Review URL: https://codereview.appspot.com/5675086

git-svn-id: https://angleproject.googlecode.com/svn/trunk@993 736b8ea6-26fd-11df-bfd4-992fa37f6226

Trac #19900
Issue=291

- Move compiler initialization to a separate function
- all this function to ensure the compiler is loaded when a compile is attempted
(previously if the compiler was released, you'd have to create a new shader object
to get it reloaded).

Signed-off-by: Shannon Woods
Signed-off-by: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@992 736b8ea6-26fd-11df-bfd4-992fa37f6226

Trac# 19900
Issue=291

- add an 'uncompile' function to clear all results from any previous compile attempts
- allow compiling to be done multiple times on the same shader
- allow a shader to be compiled with no source specified (to produce an appropriate error).

Signed-off-by: Shannon Woods
Signed-off-by: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@991 736b8ea6-26fd-11df-bfd4-992fa37f6226

Trac #19900
Issue=291

Signed-off-by: Shannon Woods
Signed-off-by: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@990 736b8ea6-26fd-11df-bfd4-992fa37f6226

Trac 19900
Issue=291

Moved a bunch of member variables from 'protected' to 'private'

Signed-off-by: Shannon Woods
Signed-off-by: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@989 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@986 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/5639055

git-svn-id: https://angleproject.googlecode.com/svn/trunk@985 736b8ea6-26fd-11df-bfd4-992fa37f6226

This is because the ninja build system, unlike make, builds translator_hlsl and fails on the clang warnings.
Review URL: https://codereview.appspot.com/5647048

git-svn-id: https://angleproject.googlecode.com/svn/trunk@984 736b8ea6-26fd-11df-bfd4-992fa37f6226

This is because some GPUs / drivers are unable to create the device once the Chromium GPU process sandbox has been locked down and waiting for the first context / surface to be created is too late.
Review URL: https://codereview.appspot.com/5647046

git-svn-id: https://angleproject.googlecode.com/svn/trunk@983 736b8ea6-26fd-11df-bfd4-992fa37f6226

PIX doesn't seem to work with share handles, so disable them when we detect that pix is enabled.

BUG=
TEST=

Review URL: https://codereview.appspot.com/5625048

git-svn-id: https://angleproject.googlecode.com/svn/trunk@982 736b8ea6-26fd-11df-bfd4-992fa37f6226

Issue=155
Signed-off-by: Nicolas Capens
Author: Pierre Leveille

git-svn-id: https://angleproject.googlecode.com/svn/trunk@981 736b8ea6-26fd-11df-bfd4-992fa37f6226

I also fixed Config.cpp so it compiles without the warning.

This is because this is now a warning as error in Chromium.
Review URL: https://codereview.appspot.com/5600047

git-svn-id: https://angleproject.googlecode.com/svn/trunk@980 736b8ea6-26fd-11df-bfd4-992fa37f6226

%z was actually just a modifier.
Review URL: https://codereview.appspot.com/5569082

git-svn-id: https://angleproject.googlecode.com/svn/trunk@979 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/5569078

git-svn-id: https://angleproject.googlecode.com/svn/trunk@978 736b8ea6-26fd-11df-bfd4-992fa37f6226

%Iu is used for size_t printf arguments in MSVC. The GCC equivalent is %z. MSVC does not support %z though.
Review URL: https://codereview.appspot.com/5578050

git-svn-id: https://angleproject.googlecode.com/svn/trunk@977 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@976 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@975 736b8ea6-26fd-11df-bfd4-992fa37f6226

Work around a bug observed with ATI graphics cards where instanced draw calls fail when all attributes are non-instanced, by repeating the draw call.

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@974 736b8ea6-26fd-11df-bfd4-992fa37f6226

Re-implement line loops by constructing a 'looping' index buffer, so it is compatible with instanced draws.

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@973 736b8ea6-26fd-11df-bfd4-992fa37f6226

Implement glDrawArraysInstanced by constructing a 'counting' index buffer since D3D9 only supports instancing on indexed draw calls.

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@972 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@971 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #19489
Signed-off-by: Daniel Koch
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@970 736b8ea6-26fd-11df-bfd4-992fa37f6226