The patch adds support for GLSL ES 3.1 image types.

Internal format layout qualifiers for images are added.

Support for the readonly and writeonly qualifiers are added. The other
memory qualifiers are omitted as to make the patch simpler.

Tests are added which check for correct and incorrect usage of images,
internal format layout and memory qualifiers.

BUG=angleproject:1442

TEST=angle_unittests
TEST=angle_end2end_tests

Change-Id: Ie4d3acb2a195de11b405ad54110a04c4c1de0b7e
Reviewed-on: https://chromium-review.googlesource.com/378855Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

This should fix the weird gn problem with git cl format.

BUG=None

Change-Id: I35d4cf110bda302b6336ef412613a80fcde8208b
Reviewed-on: https://chromium-review.googlesource.com/402253Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This ASSERT was being overly restrictive, when the code would work
correctly when reading from system memory storage as well as staging.

BUG=angleproject:1555

Change-Id: Ib45185d8c127a1fb2ce54db9e96fe41d68d5ff2e
Reviewed-on: https://chromium-review.googlesource.com/402269Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=none

Change-Id: I9ac61281dbfbde6dc285bc687734f6a0fb67034b
Reviewed-on: https://chromium-review.googlesource.com/402230Reviewed-by: Jamie Madill <jmadill@chromium.org>

BUG=None

Change-Id: I32cbcd4daa1f386654ea5e8f0a25d8635b5df731
Reviewed-on: https://chromium-review.googlesource.com/402016Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=512090

Change-Id: I9413d6f5c13b9ea59ab9c923dc6c5d157f344166
Reviewed-on: https://chromium-review.googlesource.com/291652
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

WebGL 2.0 spec section 5.25 specifies this.

BUG=angleproject:1542
TEST=WebGL 2 conformance tests

Change-Id: Ic5758bc0e767a8e6dd032e42c0a69bcae06381c4
Reviewed-on: https://chromium-review.googlesource.com/399682Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

BUG=

Change-Id: I4135f4ae939043efa8ccab65ef1a5eb441370828
Reviewed-on: https://chromium-review.googlesource.com/401673Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: I408d0e78064aa4e493d6ca66565384ac585040e1
Reviewed-on: https://chromium-review.googlesource.com/401400Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1354

Change-Id: Id7dd8438224adb1e2729bcdc18a306e5dfc83a3b
Reviewed-on: https://chromium-review.googlesource.com/401399
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: Ib25fe4905fdd6c532c67de1d4ca1ba37932f8e0b
Reviewed-on: https://chromium-review.googlesource.com/401149Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Empty declarations are possible in loop init expressions. Make
PruneEmptyDeclarations take this into account.

BUG=angleproject:1550
TEST=angle_unittests

Change-Id: If407babf9b6f7a26dfcf73ff345493d3e2af3f9a
Reviewed-on: https://chromium-review.googlesource.com/401147Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

BUG=

Change-Id: I5f36d5c51aa3eb6d498ff21733e67de418c26386
Reviewed-on: https://chromium-review.googlesource.com/401164Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1549

Change-Id: I8f003257b6f39282cb52ccdbacb73860ab37c919
Reviewed-on: https://chromium-review.googlesource.com/390271Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Shannon Woods <shannonwoods@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: I979c8768975380dba7f0b256b2729b2147ff7984
Reviewed-on: https://chromium-review.googlesource.com/399042
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: I2e8051910f50dc040f6be4922142364d36788c4e
Reviewed-on: https://chromium-review.googlesource.com/399041
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: Ied04fa28387f3bef9f2bead7b60878b1a353c8bd
Reviewed-on: https://chromium-review.googlesource.com/399039
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: Ic1aaab56d2313a19fb0362f280dbfbf804fb3fcc
Reviewed-on: https://chromium-review.googlesource.com/399040
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

ADD_AND_TRUE_TO_LOOP_CONDITION breaks loop unrolling because LoopInfo
expects the loop condition to have been validated to be "symbol < constant"
and stayed that way.

BUG=chromium:653454
BUG=chromium:653694

Change-Id: I2057fde3a308623bd2dd7571ee878dd42b1fdac3
Reviewed-on: https://chromium-review.googlesource.com/395306Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Change-Id: I192c4223e3fdefb7cecf96b50bc70529519ca650
Reviewed-on: https://chromium-review.googlesource.com/399678Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1354

Change-Id: I204962a7178d47a43034d0213c81b82d8f6a25be
Reviewed-on: https://chromium-review.googlesource.com/399043Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Previously part of the code saw it as [begin, end) while others saw it
as [begin, end].

BUG=angleproject:1052
BUG=chromium:656485

Change-Id: Id8e9e26b167e02a07dfc5341569df53a14d0623d
Reviewed-on: https://chromium-review.googlesource.com/399565Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Comma nodes always have just two parameters. If there's an expression
with several commas in the middle, it's parsed as a tree of comma
operations. It makes more sense to represent it as a binary node
rather than an aggregate node.

After this patch, TIntermAggregate is still used for function
prototypes, function parameter lists, function calls, and variable and
invariant declarations.

BUG=angleproject:1490
TEST=angle_unittests, angle_end2end_tests

Change-Id: I66be10624bf27bcf25987b4d93958d4a07600771
Reviewed-on: https://chromium-review.googlesource.com/397320Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

After this all node types are printed correctly in intermediate
output.

BUG=angleproject:1544
TEST=shader_translator -i

Change-Id: I5c120b0ff69f66b00db88e8423a1c698d339af22
Reviewed-on: https://chromium-review.googlesource.com/397319Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

BUG=angleproject:1354

Change-Id: Ic71392555ae4212c4144bc67d7c5e53874af4d53
Reviewed-on: https://chromium-review.googlesource.com/398198Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

This makes the code easier to understand. Function definition nodes
always have just two children, the parameters node and the function
body node, so there was no proper reason why they should be aggregate
nodes.

As a part of this change, intermediate output is modified to print
symbol table ids of functions so that debugging function id related
functionality will be easier in the future.

After this patch, TIntermAggregate is still used for function
prototypes, function parameter lists, function calls, variable and
invariant declarations and the comma (sequence) operator.

BUG=angleproject:1490
TEST=angle_unittests, angle_end2end_tests

Change-Id: Ib88b4ca5d21abd5f126836ca5900d0baecabd19e
Reviewed-on: https://chromium-review.googlesource.com/394707
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

It looks like the problem was not making the OS Window visible.
It turned out the previous version of the test succeeded only
when I ran angle_end2end_tests.exe with
--gtest_filter=EGLSyncControlTest.SyncValuesTest and failed
when I ran all tests, especially with --single-process-tests.

I tested the modified version of the test which repeated the
measurement 50 times and couldn't get a reliable test execution
until I figured out that back buffers were queueing and not
swapping successfully with the window being invisible.

This version works reliably when running all angle_end2end_tests.exe
on my workstation and on a test laptop, both running Windows 10,
with or without --single-process-tests.

BUG=angleproject:1402

Change-Id: I97764a86b20119cc586ea2a4fd1d94558c4e737e
Reviewed-on: https://chromium-review.googlesource.com/397739Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Stanislav Chiknavaryan <stanisc@chromium.org>

On some Mac drivers with shader version 4.1, they will
treat unused std140 and shared uniform blocks' members as inactive. However,
WebGL2.0 based on OpenGL ES3.0.4 requires all members of a named uniform block
declared with a shared or std140 layout qualifier to be considered active.
The uniform block itself is also considered active.

This workaround is to reference all members of unused std140 and shared uniform blocks
at the beginning of the vertex/fragment shader's main().

BUG=chromium:618464
TEST=UniformBufferTest.ActiveUniformBlockNumber

Change-Id: I18da4e2b61b0170068bf5ea38ce54667b0737780
Reviewed-on: https://chromium-review.googlesource.com/395648Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1492

Change-Id: I6c210243dffb775efd61c33160dabcbf7406a757
Reviewed-on: https://chromium-review.googlesource.com/397798Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: I3aa2dcb8603a839f9c07cd9dd41cb695d2e699f2
Reviewed-on: https://chromium-review.googlesource.com/395529
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

This change will make it easier to split types of TIntermAggregate
nodes representing functions and function calls into different node
classes.

BUG=angleproject:1490
TEST=angle_unittests

Change-Id: I730aa7858fe31fda86218fc685980c6ad486f5e0
Reviewed-on: https://chromium-review.googlesource.com/394706Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

The __clang__ macro cannot be directly used with an #if directive if it
has not been defined. This causes a compilation error under GCC 5.4
Linux.

Change-Id: Ieca752cd9f0e90c82febc35e19e1c1f3c6096de1
Reviewed-on: https://chromium-review.googlesource.com/397220Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Now TSymbol objects always get their id when they are constructed. The
id cannot be changed after the TSymbol is created.

This makes it simpler to insert both mangled and unmangled versions of
a function to the symbol table. These can now both share the same
TSymbol object, unlike before, when inserting the same symbol twice
would have changed its symbol id.

This requires changes to function definition parsing: function
definition nodes now share any symbol created by previous prototype
declarations of the function. The parameters on the symbol get set to
the parameters in the function definition header.

BUG=angleproject:1490
TEST=angle_unittests

Change-Id: I8e600e9b5e5de27d64b85c5042cfd23ff02abe63
Reviewed-on: https://chromium-review.googlesource.com/396838Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

The local dirty bits were sometimes not synced because gl::Texture doesn't
know that they exist.  Also fix calculation of when the workaround dirty
bits need to be set.

BUG=angleproject:1386

Change-Id: I3d9d1a01e5441be783190422093c485ea5da7aec
Reviewed-on: https://chromium-review.googlesource.com/396542Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1354

Change-Id: I07dccc470c57aadd41cae8fce327e4aca58a0bea
Reviewed-on: https://chromium-review.googlesource.com/395113Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Right-shifting a negative number should sign-extend according to the
ESSL 3.00.6 spec. Implement sign-extending right shift so that it
doesn't hit any undefined behavior in the C++ spec. Negative lhs
operands are now allowed for bit-shift right.

Also implement bit-shift left via conversion to unsigned integer, so
that it does not hit signed integer overflow. Negative lhs operands
are now allowed also for bit-shift left as well.

BUG=chromium:654103
TEST=angle_unittests

Change-Id: Iee241de9fd0d74c2f8a88219bddec690bb8e4db2
Reviewed-on: https://chromium-review.googlesource.com/395688
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Fixed the compiling error in Mac by adding missing override identifier
for getSequence.

BUG=angleproject:1490
TEST=compile on macOS Sierra 10.12

Change-Id: I8f8d556c82885bbee1712e1b6e5f439b67906023
Reviewed-on: https://chromium-review.googlesource.com/396403Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=angleproject:1354

Change-Id: I994b7a596b5b67fadbd91ad0d64054c792523200
Reviewed-on: https://chromium-review.googlesource.com/395111
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

These are distinct from the renderer-level workarounds.

Add the first workaround, loseContextOnOutOfMemory. This is enabled
when the application enables reset notifications via KHR_robustness,
and is intended for more robust handling of errors in key APIs for
Chromium's correctness (in particular, sync objects).

Change Context::getResetStatus to persistently return the "lost" status
if it was set by calling Context::markContextLost. Previously, if
markContextLost was called but the implementation hadn't actually
received an error like a lost device, the "lost" reset status would be
dropped. Returning it only once to the caller is also fragile. Tested
this by manually injecting a failure in FenceSync11::clientWait and
ensuring that Chromium detected it as expected.

BUG=chromium:650138

Change-Id: Ie53069eacd1754ad5d64936e3fef315af24605fa
Reviewed-on: https://chromium-review.googlesource.com/394233
Commit-Queue: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1490
TEST=build on Linux GCC

Change-Id: I64f53d636d4acee0469b6e9ba1201322087d5f65
Reviewed-on: https://chromium-review.googlesource.com/395732Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>