R=alokp@chromium.org

Review URL: https://codereview.appspot.com/9436044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2425 736b8ea6-26fd-11df-bfd4-992fa37f6226

TBR=kbr@chromium.org

Review URL: https://codereview.appspot.com/10088043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2424 736b8ea6-26fd-11df-bfd4-992fa37f6226

R=kbr@chromium.org

Review URL: https://codereview.appspot.com/9866043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2423 736b8ea6-26fd-11df-bfd4-992fa37f6226

The version committed in Issue 9827044 was generated in a bad environment and doesn't compile againts top-of-tree. This version does.

BUG=429
R=alokp@chromium.org

Review URL: https://codereview.appspot.com/9738044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2249 736b8ea6-26fd-11df-bfd4-992fa37f6226

This change also required that support be added for associating built-in
variables with an extension, similar to how functions could be associated with
extensions previously.

R=alokp@chromium.org

Review URL: https://codereview.appspot.com/9827044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2248 736b8ea6-26fd-11df-bfd4-992fa37f6226

Also removed support for max-array-size. You can only define a fixed-size array.

TEST=WebGL conformance tests

Review URL: https://codereview.appspot.com/9697044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2247 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/9685045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2246 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2245 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2244 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2243 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2242 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #22937
Signed-off-by: Shannon Woods
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2214 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/9295045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2212 736b8ea6-26fd-11df-bfd4-992fa37f6226

BUG=crbug 179653
R=aedla@chromium.org, kbr@chromium.org

Review URL: https://codereview.appspot.com/8834048

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2211 736b8ea6-26fd-11df-bfd4-992fa37f6226

This is so Chrome can create another D3D device on the same adapter that can share resources with ANGLE's D3D device.
Review URL: https://codereview.appspot.com/9225046

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2210 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #23127
Signed-off-by: Geoff Lang
Signed-off-by: Jamie Madill
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2209 736b8ea6-26fd-11df-bfd4-992fa37f6226

TRAC #23127
Signed-off-by: Geoff Lang
Signed-off-by: Jamie Madill
Author: Nicolas Capens

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2208 736b8ea6-26fd-11df-bfd4-992fa37f6226

Fixed error that allowed negative index for accessing
vector, matrix, and array. Now we report compile error
and clamp the index to 0.

Re-arranged code around it to handle negative index
at the one location.

BUG=crbug.com/239411
TEST=bug test case
R=aedla@chromium.org, kbr@chromium.org

Review URL: https://codereview.appspot.com/9193045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2207 736b8ea6-26fd-11df-bfd4-992fa37f6226

R=kbr@chromium.org

Review URL: https://codereview.appspot.com/9078046

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2202 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/9309044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2200 736b8ea6-26fd-11df-bfd4-992fa37f6226

We had a TTypeLine to represent a structure field,
which simply encapsulated a TType and line number.
The line number was only used during parsing for error reporting.
There is no need to store a line number because it is already
available in the parser token.

TEST=WebGL conformance tests
R=kbr@chromium.org

Review URL: https://codereview.appspot.com/9223045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2198 736b8ea6-26fd-11df-bfd4-992fa37f6226

R=kbr@chromium.org

Review URL: https://codereview.appspot.com/9242043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2197 736b8ea6-26fd-11df-bfd4-992fa37f6226

Saves 20% memory for global symbol initialization

R=alokp@chromium.org

Review URL: https://codereview.appspot.com/9152043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2196 736b8ea6-26fd-11df-bfd4-992fa37f6226

Identifier is sufficient to parse struct fieldname.

TEST=WebGL conformance tests.

Review URL: https://codereview.appspot.com/8709048

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2195 736b8ea6-26fd-11df-bfd4-992fa37f6226

It is handled by the preprocessor. Added an assert.

TEST=WebGL conformance tests.
TBR=kbr@chromium.org

Review URL: https://codereview.appspot.com/8653052

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2194 736b8ea6-26fd-11df-bfd4-992fa37f6226

TEST=WebGL conformance tests
TBR=kbr@chromium.org

Review URL: https://codereview.appspot.com/8983043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2193 736b8ea6-26fd-11df-bfd4-992fa37f6226

This reverts commit 12e9eebd.

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2192 736b8ea6-26fd-11df-bfd4-992fa37f6226

Identifier is sufficient to parse struct fieldname.
Note that we do not need stacked lexer state anymore.

TEST=WebGL conformance tests.

Review URL: https://codereview.appspot.com/8565045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2191 736b8ea6-26fd-11df-bfd4-992fa37f6226

To distinguish the tokens we used to track the lexer context,
which is fragile. This patch implements a better way -
combine the two tokens into a common parser rule and let
the parser context decide if the declaration is valid,
which it already does by checking the naming conflicts.

TEST=WebGL conformance tests.

Review URL: https://codereview.appspot.com/8797044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2190 736b8ea6-26fd-11df-bfd4-992fa37f6226

Delete dead-code for handling comments and invalid octal integers. These are already handled by the preprocessor.

TEST=WebGL conformance tests.

Review URL: https://codereview.appspot.com/8632053

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2189 736b8ea6-26fd-11df-bfd4-992fa37f6226

TBR=kbr@chromium.org

Review URL: https://codereview.appspot.com/8927045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2188 736b8ea6-26fd-11df-bfd4-992fa37f6226

TEST=WebGL conformance tests.
TBR=kbr@chromium.org

Review URL: https://codereview.appspot.com/8925044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2187 736b8ea6-26fd-11df-bfd4-992fa37f6226

Note that I am planning to get rid of PoolAllocator entirely.

BUG=crbug 179654

Review URL: https://codereview.appspot.com/8662046

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2186 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/8834045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2185 736b8ea6-26fd-11df-bfd4-992fa37f6226

Review URL: https://codereview.appspot.com/8833047

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2184 736b8ea6-26fd-11df-bfd4-992fa37f6226

Some AMD drivers fail with D3DERR_NOTAVAILABLE after the machine has been in sleep mode and never seem to recover for a given IDirect3D9 object. I reproduced with Catalyst 9.12.0.0 and a Radeon 7700.

Instead, record the display mode at the time the it was initialized and thereafter use its format as the argument to CheckDeviceFormat etc. Previously, CheckDeviceFormat was passed an uninitialized value as the format if GetAdapterDisplayMode failed.

See http://crbug.com/229950
Review URL: https://codereview.appspot.com/8677044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2183 736b8ea6-26fd-11df-bfd4-992fa37f6226

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2046 736b8ea6-26fd-11df-bfd4-992fa37f6226

It isn't used anymore.
Review URL: https://codereview.appspot.com/8539043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2043 736b8ea6-26fd-11df-bfd4-992fa37f6226

The call to IDirect3DDevice9::Present in testDeviceLost sometimes never returns on AMD, resulting in a hang.

Also, on AMD, device lost is sometimes never reported at all, even through present occluded or display mode changed. I'll need to do something more robust at the app level. It has something to do with the Chrome GPU process sandbox. These issues only happen when the sandbox is enabled.

Cleanup for Surface::recreateAdditionalSwapChains to follow once I determine this gets us back to a stable place.
Review URL: https://codereview.appspot.com/8038044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2040 736b8ea6-26fd-11df-bfd4-992fa37f6226

This code hasn't been used by Chrome since it switched to using shared textures for its presentation path on Vista and later.

It will also resolve this:
https://codereview.appspot.com/7816046/
Review URL: https://codereview.appspot.com/7872045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@2002 736b8ea6-26fd-11df-bfd4-992fa37f6226