The redesigned code places the internal buffers and input elements for
instance point sprite emulation (FL9_3) at the start of the arrays,
instead of swapping with the first non-instanced element. This makes
the tracking logic of the caching somewhat cleaner.

This facilitates the work of implementing instancing-on-instancing for
FL9_3 conformance.

BUG=angleproject:1279

Change-Id: Ifb030816a313b1e8b916c57ef05915914443312a
Reviewed-on: https://chromium-review.googlesource.com/325090Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Fails on the Windows builders:

https://build.chromium.org/p/chromium.gpu.fyi/builders/GPU%20Win%20Builder/builds/42201

e:\b\build\slave\gpu_win_builder\build\src\third_party\angle\src\libangle\renderer\d3d\d3d11\load_functions_table_autogen.cpp(787) : error C2440: 'return' : cannot convert from 'const std::map<GLenum,rx::LoadImageFunction,std::less<_Kty>,std::allocator<std::pair<const _Kty,_Ty>>>' to 'const std::map<GLenum,rx::d3d11::LoadImageFunctionInfo,std::less<_Kty>,std::allocator<std::pair<const _Kty,_Ty>>> &'
        with
        [
            _Kty=GLuint
,            _Ty=rx::LoadImageFunction
        ]
        and
        [
            _Kty=GLuint
,            _Ty=rx::d3d11::LoadImageFunctionInfo
        ]
        Reason: cannot convert from 'const std::map<GLenum,rx::LoadImageFunction,std::less<_Kty>,std::allocator<std::pair<const _Kty,_Ty>>>' to 'const std::map<GLenum,rx::d3d11::LoadImageFunctionInfo,std::less<_Kty>,std::allocator<std::pair<const _Kty,_Ty>>>'
        with
        [
            _Kty=GLuint
,            _Ty=rx::LoadImageFunction
        ]
        and
        [
            _Kty=GLuint
,            _Ty=rx::d3d11::LoadImageFunctionInfo
        ]
        No user-defined-conversion operator available that can perform this conversion, or the operator cannot be called

This reverts commit 52d3e43b.

Change-Id: I83774ea09ccbb2b92ff609714e1c7201beb775bc
Reviewed-on: https://chromium-review.googlesource.com/326540Reviewed-by: Jamie Madill <jmadill@chromium.org>

Pass the FBO attachment to the Buffer packing method so we can
retrieve the layer of the attachment when doing an asynchronous
readback. Also take advantage of the TextureHelper11 class to
remove some redundant code.

BUG=angleproject:1290

Change-Id: I26bb21a03e0ff7a42aab4eee75f3c3d12915f398
Reviewed-on: https://chromium-review.googlesource.com/324021Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This is essentially a reland of 9cf9bcbe

This code always allocates a new memory buffer for the texture memory,
picks a load function to copy/convert the input data into it, and the
uploads.

In the case where the input format matches the upload format we should
be able to skip the allocation and copy and be much happier.

Change-Id: I3c1a0ab34be91c7d39eb2d0866029d01352b6056
Reviewed-on: https://chromium-review.googlesource.com/325526
Tryjob-Request: John Bauman <jbauman@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

BUG=580045

Change-Id: I917f62dca047ca163bfdbcc5437b2fc1d7a1e020
Reviewed-on: https://chromium-review.googlesource.com/326401Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Timer queries without context virtualization are now supported in the
Direct3D11 backend. Only the elapsed time portion of the GL extension is
supported though due to limitations of D3D11 preventing a reliable
implementation of the timestamp functionality. As a result of this, the
counter bits for the timestamp is set to 0 and any queries of the
timestamp will always return 0.

BUG=angleproject:657

Change-Id: I51ca1a1a6bd6bc13155cebeacdca414b764db168
Reviewed-on: https://chromium-review.googlesource.com/325780Reviewed-by: Jamie Madill <jmadill@chromium.org>

This allows ANGLE to render directly onto a D3D swapchain in the correct
orientation when using the D3D11 renderer.

The trick is to add an extra uniform to each shader which takes either
the value +1.0 or -1.0. When rendering to a texture, ANGLE sets this
value to -1.0. When rendering to the default framebuffer, ANGLE sets
this value to +1.0. ANGLE multiplies vertex positions by this value in
the VS to invert rendering when appropriate. It also corrects other
state (e.g. viewport/scissor rect) and shader built-in values
(e.g. gl_FragCoord).

This saves a substantial amount of GPU time and lowers power
consumption. For example, the old method (where ANGLE renders all
content onto an offscreen texture, and then copies/inverts this onto the
swapchain at eglSwapBuffers() time) uses about 20% of the GPU each frame
on a Lumia 630.

Verification:
+ dEQP GL ES2 tests pass when "present path fast" is enabled
+ all ANGLE_end2end_tests pass when "present path fast" is enabled

BUG=angleproject:1219

Change-Id: Ib6eeea46bafa6ebce4adada0ae9db3a433b8fc4c
Reviewed-on: https://chromium-review.googlesource.com/321360Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tryjob-Request: Austin Kinross <aukinros@microsoft.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

They are no longer needed by Chromium.

BUG=522967

Change-Id: I187eed52378a58f3ea24c3e21082c6a93e56da7f
Reviewed-on: https://chromium-review.googlesource.com/325330Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Adds a new extension to allow transcode ETC formats to suitable BC formats.
This commit implements ETC1 to BC1. More formats will be supported soon.

BUG=angleproject:1285

Change-Id: Iacbfbc2248dfe1aebf24b92696249a9404e331cd
Reviewed-on: https://chromium-review.googlesource.com/316511Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Pbuffers created using eglCreatePbufferFromClientBuffer from share handles
created using keyed mutexes can now have their keyed mutex queried. All other
pbuffers will continue to be created without keyed mutexes.

BUG=337606

Change-Id: Id9039064dd8c89fe1706cc8ecb31111c994761fc
Reviewed-on: https://chromium-review.googlesource.com/324253Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

This reverts commit 3026829e.

Change-Id: I0b4c3f5b1453b993b149423bb1ce407f4918cf54
Reviewed-on: https://chromium-review.googlesource.com/325435Reviewed-by: Geoff Lang <geofflang@chromium.org>

The base level is passed to shaders in a uniform block created
specifically for passing sampler metadata. This is done on feature levels
above 9_3, which treat samplers as indices to sampler arrays in shaders.

BUG=angleproject:596
TEST=angle_end2end_tests

Change-Id: I846f2fc195ab1fd884052824ffd3c1d65083c0fb
Reviewed-on: https://chromium-review.googlesource.com/322122Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

Non-zero layer attachments are FBO attachments of 3D textures bound
from a layer other than zero. These haven't ever worked AFAIK. Fix
them by retrieving the correct layer from the FBO attachment.

Note: 3D attachments are still broken with PBO ReadPixels. The fix
for those will come in a subsequent patch.

BUG=angleproject:1290

Change-Id: I5417e7374188dd320e1209d006723ce070f98561
Reviewed-on: https://chromium-review.googlesource.com/323472
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Use the TextureHelper class to abstractly handle 2D and 3d textures.

Also refactor the Image11 copy methods to be a bit cleaner and not
use the copy conversion path when unnecessary.

This patch does not yet fix layer attachments - the fix for that will
come up in a subsequent patch.

BUG=angleproject:1290

Change-Id: If8b7aa8848ca4260e0dde690e7a99e115a97fabb
Reviewed-on: https://chromium-review.googlesource.com/323442
Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This makes debugging using graphics tools easier on Windows.

BUG=angleproject:1290
BUG=angleproject:1299

Change-Id: Ia3889c10fd8e23486f0122e0b567c470517bc9e0
Reviewed-on: https://chromium-review.googlesource.com/324821Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1219

Change-Id: I1e303ec30ffe877aebf00720620b818c335fac76
Reviewed-on: https://chromium-review.googlesource.com/310652Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Austin Kinross <aukinros@microsoft.com>

EXT_disjoint_timer_query is feature complete with the WebGL tests passing
with Chromium using ANGLE as a backend. There is some flakiness in the
timestamp query test on WebGL, but investigation revealed a bug on
Chromium's end and a fix is being made there. Since the extension is
feature complete, it is now enabled by default on OpenGL so that it can be
regression tested.

BUG=angleproject:1265

Change-Id: If018b7e3ae84aff7e40c73ff8e672a86689ae6c4
Reviewed-on: https://chromium-review.googlesource.com/324580
Tryjob-Request: Ian Ewell <ewell@google.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Ian Ewell <ewell@google.com>

This won't be mirrored on the waterfall for the time being due to lack
of hardware.

BUG=542370

Change-Id: I2ecfde30f66d3dcdde2d05144d647c2b4b5c83d3
Reviewed-on: https://chromium-review.googlesource.com/325133Tested-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>

Without this, gn-built isolates do not include the test expectations
file in the isolate, causing the test suite to bail out at startup.
Also rename target_name to api_name as target_name is a special
variable name when inside a target definition.

BUG=580045

Change-Id: I20be39c7e50e24757a0828f9056613e171f0002f
Reviewed-on: https://chromium-review.googlesource.com/325040Tested-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

Translation of samplers to HLSL on D3D11 is changed as follows:

Instead of passing around HLSL sampler and HLSL texture references in
shaders, all references to ESSL samplers are converted to constant
indices within the shader body. Each ESSL sampler is identified by an
unique index. In the code generated to implement ESSL texture functions,
these indices are used to index arrays of HLSL samplers and HLSL
textures to get the sampler and texture to use.

HLSL textures and samplers are grouped into arrays by their types. Each
unique combination of a HLSL texture type + HLSL sampler type gets its
own array. To convert a unique sampler index to an index to one of these
arrays, a constant offset is applied. In the most common case of a 2D
texture and a regular (non-comparison) sampler, the index offset is
always zero and is omitted.

The end goal of this refactoring is to make adding extra metadata for
samplers easier. The unique sampler index can be used in follow-up
changes to index an array of metadata passed in uniforms, which can
contain such things as the base level of the texture.

This does not solve the issues with samplers in structs.

The interface from the point of view of libANGLE is still exactly the
same, the only thing that changes is how samplers are handled inside the
shader.

On feature level 9_3, the D3D compiler has a bug where it can report that
the maximum sampler index is exceeded when in fact it is not. This can
happen when an array of samplers is declared in the shader. Because of
this the new approach can't be used on D3D11 feature level 9_3, but it
will continue using the old approach instead.

BUG=angleproject:1261
TEST=angle_end2end_tests,
     dEQP-GLES3.functional.shaders.texture_functions.* (no regressions)
     dEQP-GLES3.functional.texture.units.* (no regressions)

Change-Id: I5fbb0c4280000202dc2795a628b56bd8194ef96f
Reviewed-on: https://chromium-review.googlesource.com/320571Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Tryjob-Request: Olli Etuaho <oetuaho@nvidia.com>

Using BindFramebuffer(1) then GenFramebuffers would return 1.
This leads to a memory leak and was something that was obscuring
debugging a bug in my ReadPixels fix for ES3.

This also fixes a bug where running the texture tests along produces
some random failures.

BUG=angleproject:1290
BUG=angleproject:1299

Change-Id: If11e8c743d2ddde725b12749ac012f670cd290e1
Reviewed-on: https://chromium-review.googlesource.com/324820Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

BUG=

Change-Id: I9e67ba2165ac87959f319fd2485fe385c60a8ea0
Reviewed-on: https://chromium-review.googlesource.com/324461Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Unknown when these started passing, but they seem to be OK locally
and on the bot.

BUG=angleproject:1095
BUG=angleproject:1097
BUG=angleproject:1101

Change-Id: Icf6102792b356cd23bf39414f75e796fe7e81879
Reviewed-on: https://chromium-review.googlesource.com/324940
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1304

Change-Id: If0ec8682b1d519bc6460cd6ed82f42c4c1ceb952
Reviewed-on: https://chromium-review.googlesource.com/324930
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

BUG=580045

Change-Id: Id19f843e8a04d38938b6827a7a64970eb088bcc6
Reviewed-on: https://chromium-review.googlesource.com/324760Reviewed-by: Kenneth Russell <kbr@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

This is to fix failures in the automated testing while we work on them.

BUG=angleproject:1303

Change-Id: I0813b5cf1dca1e329504704c70b9d2b0c34d003b
Reviewed-on: https://chromium-review.googlesource.com/324460Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Basic timer queries are supported and tested in the OpenGL backend
but are not enabled by default. A good portion of the existing query
code was also refactored for improved validation - specifically for
validating that the appropriate extensions are available.

BUG=angleproject:1265

Change-Id: Iebae994cd7a8d3ed3e9fc3776fe2f3d99caa9237
Reviewed-on: https://chromium-review.googlesource.com/323450Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tryjob-Request: Ian Ewell <ewell@google.com>
Tested-by: Ian Ewell <ewell@google.com>

We might want UNUSED_ASSERTION as a special case but the unused trace
variable macro was being abused in places.

BUG=angleproject:1290

Change-Id: I5f9252b3ff0484d2e859b9fcf68d5d060b28f1fc
Reviewed-on: https://chromium-review.googlesource.com/324290Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1145

Change-Id: I9f3a8926eaad88076026f58922289487f90518c7
Reviewed-on: https://chromium-review.googlesource.com/324270Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

An obvious bug crept in setVisible where setVisible(true) twice would
end up hiding the window.

BUG=angleproject:1145

Change-Id: Iea8ca3de4ab8c05e6da66fb04c8d377824547b41
Reviewed-on: https://chromium-review.googlesource.com/324090
Tryjob-Request: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>

BUG=522967

Change-Id: I9a5c1793b089920273c74ab43fd5464d0aea7adb
Reviewed-on: https://chromium-review.googlesource.com/324022Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Change-Id: I9f733055ba3c22950eeedc17e756fe50af272a16
Reviewed-on: https://chromium-review.googlesource.com/324280Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Tested-by: Andrii Shyshkalov <tandrii@chromium.org>

BUG=angleproject:1145

Change-Id: I3072c31f2f743943b2ed9a3b1a020529112d5f71
Reviewed-on: https://chromium-review.googlesource.com/324091Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

BUG=540829

Change-Id: I3595f316afa1897224ad95d6f9a3da3685aa0831
Reviewed-on: https://chromium-review.googlesource.com/324032
Tryjob-Request: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>

Split the test into several categories based on what fixture types
they need.

BUG=angleproject:1290

Change-Id: Ia25a022d90a0571e03d86c8e38f6243189968149
Reviewed-on: https://chromium-review.googlesource.com/324020Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

This re-lands https://chromium-review.googlesource.com/#/c/323690/. It didn't break tryjobs, they actually didn't work.

Change-Id: Id3da7658515f59e8dae550816d82b1fcce1e5f7a
Reviewed-on: https://chromium-review.googlesource.com/324220Reviewed-by: Sergiy Byelozyorov <sergiyb@chromium.org>
Tested-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Andrii Shyshkalov <tandrii@chromium.org>

 * IsAttachmentEnabled was checking the wrong draw buffer state.  Instead
   of checking that drawbuffer[colorAttachment] is in use, it should have
   been scanning for a drawbuffer state that points to colorAttachment.
 * Allow for maxDrawBuffer != maxColorAttachments.  Tested by the GL
   backend on some systems that don't have the draw buffers extension.

Fixed by updating the helpers and adding a new getDrawBuffer helper.

BUG=angleproject:1121

Change-Id: Idd1b0a9ec4a3f944d332c708364408bf5d59e1fd
Reviewed-on: https://chromium-review.googlesource.com/292740Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tryjob-Request: Geoff Lang <geofflang@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>

This avoids sandbox issues when calling SetPixelFormat on windows from
other processes while in a sandboxed process.

BUG=540829

Change-Id: I012168531d8c2e5b189671533cea9f57ec713538
Reviewed-on: https://chromium-review.googlesource.com/296340Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tryjob-Request: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>

Previously our output would spam multiple lines by checking equality
for each color component. Instead, use a helper struct with a stream
operator to clean up the output. This makes it much more legible.

New output style:

  Expected: (0, 255, 0, 255)
  Actual: angle::MakeGLColor(255, 0, 127, 255)
  Which is: (255, 0, 127, 255)

BUG=angleproject:1290

Change-Id: Ieea018ab6c149a20b5fd74d74c972118b63a4b5e
Reviewed-on: https://chromium-review.googlesource.com/323441Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

These standard functions are much more powerful than the C random()
routines. Use them to improve the random utils, and use a class
to clean things up further.

This fixes a problem I was having using random_utils where I was
having trouble generating random 32 bit unsigned integers.

BUG=angleproject:1290

Change-Id: I5081764053d0667a4e323553b7dea531256aa778
Reviewed-on: https://chromium-review.googlesource.com/323440Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>