Generate a synthetic file containing the spirv disassembly, and allow the debugger to single line step over these instructions, along with inspection of the SSA values.

All of this is no-op unless ENABLE_VK_DEBUGGER is defined at compile time, and the VK_DEBUG_PORT env var is set at run time.

Bug: b/145351270
Change-Id: Iba71717d78f7213ba586a1632b44f5fe08addf08
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/38915
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Change-Id: I0595a5d7b1bfe20a063eb98caf7497a36e9daa2c
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40088
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

Change-Id: Ibe0a658c20ccf2fba47d5d0badb0764150e7a8a9
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40048
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Bug: b/144825072
Change-Id: If620d4ce920dad1d929c8e1ad3ea87d8de8f9b1e
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40030Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

A2B10G10R10 (RGBA) was already supported by
SwiftShader, but A2R10G10B10 (BGRA) was not. Most of
this cl is trivial, as it simply adds an equivalent
version of the new formats from the code used for the
already supported formats, with the R and B channels
swapped.

The only new piece of code is rounding for 1010102
formats at the top of the PixelRoutine::writeColor()
function. There was already rounding for 8 bit formats,
but not 1010102, which led to potential off by 1 errors
in the output, which is fairly large when it happens
on the 2 bit alpha channel. This fixes one of the
dEQP-VK.pipeline.blend.*a2r10g10b10* tests.

Tests: dEQP-VK.*a2r10g10b10*
Bug b/142661203

Change-Id: Ifcae17aecafab3ea7967fdc755391ddd5e651ca5
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40008Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

The logic of PhysicalDevice::getImageFormatProperties(), as well as
testing with ANGLE, shows that multisampling of integer formats is
supported and works.  The value of
VkPhysicalDeviceLimits::sampledImageIntegerSampleCounts should therefore
be set to:

    VK_SAMPLE_COUNT_1_BIT | VK_SAMPLE_COUNT_4_BIT

Bug: b/147438583
Change-Id: Iaa7bd22789fd20b4e7975d3d2ffee2a8f90b006b
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40068Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ian Elliott <ianelliott@google.com>

The primary purpose of doing this is to have all the PowerVR SDK
examples in one place and remain up-to-date. This change also lets us
simplify a bit of our CMakeLists.txt as we only need to use
add_subdirectory to build everything we need.

The PowerVR SDK examples are built by explicitly enabling
SWIFTSHADER_BUILD_PVR_EXAMPLES (either through a cmake gui, or
-DSWIFTSHADER_BUILD_PVR_EXAMPLES=ON on the command line). You must also specify
a window system for the examples to use with something like
"-DPVR_WINDOW_SYSTEM=X11".

Several cmake option flags were renamed to start with "SWIFTSHADER_" to avoid
name collisions with subdirectories.

Bug: b/141361933
Change-Id: I6d2fac579f62ff4bc211d5fd5127fb670eb9b4bf
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39909Tested-by: Sean Risser <srisser@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

This is a follow-up to this cl:
https://swiftshader-review.googlesource.com/c/SwiftShader/+/39748
where filtering was enabled for depth only formats.

According to the Vulkan spec (about
VK_FORMAT_FEATURE_SAMPLED_IMAGE_FILTER_LINEAR_BIT):
"If the format being queried is a depth/stencil format, this bit only
 specifies that the depth aspect (not the stencil aspect) of an image
 of this format supports linear filtering, and that linear filtering
 of the depth aspect is supported whether depth compare is enabled in
 the sampler or not. If this bit is not present, linear filtering
 with depth compare disabled is unsupported and linear filtering with
 depth compare enabled is supported, but may compute the filtered
 value in an implementation-dependent manner which differs from the
 normal rules of linear filtering. The resulting value must be in the
 range [0,1] and should be proportional to, or a weighted average of,
 the number of comparison passes or failures."

So VK_FORMAT_D32_SFLOAT_S8_UINT can be added to the list of
filterable formats, since VK_FORMAT_D32_SFLOAT is already supported.

Change-Id: Ida83c2630c920f3dc2c3e17853dfe91c835a4b7f
Tests: dEQP-VK.*d32_sfloat_s8_uint*
Bug: b/146563038
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/40028
Presubmit-Ready: Alexis Hétu <sugoi@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

This is an attempt to speed up build times and avoid the flaky timeouts we're seeing with the LLVM debug builds.

Bug: b/147355576
Change-Id: I8dc83dd40a570849d02fe95822de534f0faae595
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39956Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

Addresses the review comment:
https://swiftshader-review.googlesource.com/c/SwiftShader/+/39885/3/src/Pipeline/SpirvShader.cpp#422

Bug: b/145351270
Change-Id: I9a5ceefe048e63e0f52b7e9a15bd84f6bac16629
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39955Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Bug: b/147359661
Change-Id: I20e3b90b02327250d77a1006cea712022ad0f9d5
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39951Reviewed-by: Alexis Hétu <sugoi@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Change-Id: I3ed2e5239fb255174fc523b59e7d038fae0ff984
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39954Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Better to have a descriptive compiler error message than weird missing symbols.

Bug: b/147359661
Change-Id: I66c0e5beb1e94f79210528867c990ab3e0f8b0ae
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39949Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Bug: b/147359661
Change-Id: Ieb2de4d93ca617b190977e258023992f4a6d34c9
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39948Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

GCC 4.8 doesn't even understand --std=c++14

Bug: b/147359661
Change-Id: I5189d24638007ffe40e8903a382bccab9ef1c0fc
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39950Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Move `setActiveLaneMask()` from `SpirvShader::EmitState` to `SpirvShader`. Also capitalize the function name, as this is the more common style in this class.

The debugger will want to expose which lanes are active, so we'll need to perform additional work whenever this is called.

This is a no-op change.

Bug: b/145351270
Change-Id: I294c1d7cda08faaa6b784db148b70a7c4d84d431
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39886
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Previously `OpExtInst` and `OpExtInstImport` was hard-coded to only support `GLSL.std.450`.
We will want to support `OpenCL.Debug.100`, so put in the plumbing to properly support other extensions.

Bug: b/145351270
Change-Id: I60fbb8c45bb57b747067437641676946129b1251
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39885Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

Strings are used for debug instructions.

Bug: b/145351270
Change-Id: I6bc711c6c8b320bef6e5abbe9c4bda189fdafc0b
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39884Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Instead of declaring a bunch of private forward declarations in SpirvShader, create a single Impl struct, which can hold all the forward declarations. The names of these forward declarations can then directly match the SpirvShader file name suffixes.

The Impl struct can also be used later to store pointers to any internal state.

This was proposed here: https://swiftshader-review.googlesource.com/c/SwiftShader/+/38929/1/src/Pipeline/SpirvShaderGroup.cpp#23

Bug: b/145351270
Change-Id: Ie0de5e651fce9969fe0c24f51ae57652cade54d1
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39883Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

Previously Frame was an Thread-internally managed structure, with special accessors for the variable containers of the top most frame.

With this change, the Frame can be modified by calling `Thread::update()`. This allows for more powerful edits of the frame, while still preserving a mutex lock over the structure.

Bug: b/145351270
Change-Id: I81e87884e26bdd6f22b47e8f71b37bbfc91c7830
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39882
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

B10G11R11_UFLOAT blending is required in order for SwANGLE
to expose GL_EXT_color_buffer_float.

In this cl:
- I added a minor readability improvement by storing
  "*Pointer<Int>(data + OFFSET(DrawData, colorPitchB[index]))"
  in a variable, "pitchB" since it's used in every case.
- Added a constant for B10G11R11 masking
- Added clamping in the floating point blending code for
  unsigned floating point formats
- Fixed the VK_FORMAT_B10G11R11_UFLOAT_PACK32 output format
  which now takes the coverage mask into account
- Included a minor followup cleanup in ShaderCore

Bug: b/146223877
Tests: dEQP-VK.*b10g11r11*
Change-Id: Ifb95f34a10cdbee9d185bc25feba0aeaca0d9e70
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39929Reviewed-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Presubmit-Ready: Alexis Hétu <sugoi@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

`vk::dbg::Reference` is a `vk::dbg::Value` that reads and writes from a value stored in memory.

Used for displaying variables.

Bug: b/145351270
Change-Id: I64146cb11b13267bef15d727af82a2256a9f25ce
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39881Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Presubmit-Ready: Ben Clayton <bclayton@google.com>

This is a step toward being able to claim support for
storageImageExtendedFormats.

Bug: b/146579770
Test: dEQP-VK.*r16g16*
Change-Id: Iba658f39271f7e181c10c117e53c8e3dbdf37595
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39769
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

SpirvShader doesn't actually do anything with this (yet).
Extracts the dull plumbing from the meat of the SpirvShader debugger implementation.

Bug: b/145351270
Change-Id: I1a648d8e463c428a232f711d0cc5dd0ecf94ad1e
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/38914Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

Add a debugger context and server to vk::Device, along with a getter for acquiring the debugger context.

Generate a synthetic file containing all the Vulkan commands in the command buffer, and allow the debugger to single line step over these.

All of this is no-op unless ENABLE_VK_DEBUGGER is defined at compile time, and the VK_DEBUGGER_PORT env var is set at run time.

Bug: b/145351270
Change-Id: I8bea398a6c08d4cb23d76172dbca2a08ae109a6d
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/38913
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

In file included from ../../third_party/swiftshader/third_party/llvm-subzero/lib/Support/Signals.cpp:183:
    ../../third_party/swiftshader/third_party/llvm-subzero/lib/Support/Windows/Signals.inc(220,13): error: unused function 'printStackTraceWithLLVMSymbolizer' [-Werror,-Wunused-function]
    static bool printStackTraceWithLLVMSymbolizer(llvm::raw_ostream &OS,
    ^
    1 error generated.

Bug: chromium build fix
Change-Id: I9ed9ac680fb5286fda0685788c6dbf76a4e34369
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39908Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

The following tests are failing due to invalid SPIR-V merge blocks:

  dEQP-VK.spirv_assembly.instruction.graphics.loop.break_frag
  dEQP-VK.spirv_assembly.instruction.graphics.loop.break_vert
  dEQP-VK.spirv_assembly.instruction.graphics.loop.continue_frag
  dEQP-VK.spirv_assembly.instruction.graphics.loop.continue_vert
  dEQP-VK.spirv_assembly.instruction.graphics.loop.multi_block_loop_construct_frag
  dEQP-VK.spirv_assembly.instruction.graphics.loop.multi_block_loop_construct_vert
  dEQP-VK.spirv_assembly.instruction.graphics.loop.return_frag
  dEQP-VK.spirv_assembly.instruction.graphics.loop.return_vert
  dEQP-VK.spirv_assembly.instruction.graphics.opphi.out_of_order_frag
  dEQP-VK.spirv_assembly.instruction.graphics.opphi.out_of_order_vert
  dEQP-VK.spirv_assembly.instruction.graphics.switch_block_order.out_of_order_frag
  dEQP-VK.spirv_assembly.instruction.graphics.switch_block_order.out_of_order_vert

This fix is currently working its way through the dEQP branches.
Once it lands in 1.1.6, this change will be reverted and the dEQP SHA will be rolled forward.

Change-Id: Ib69aa6aeb09c89519ff669281ef839a8555946bc
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39878Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

LLVM has custom crash handling and stack trace printing code which may
interfere with the application's error handling. Specifically,
AddressSanitizer makes calls to SymInitialize() and SymSetOptions().

This change disables the config options, and adds a missing preprocessor
conditional for ENABLE_BACKTRACES on Windows to match the Unix behavior.

Bug: chromium:1033484
Change-Id: I3aa35d418212448c15eba4b0c3cc5b55c2da006e
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39888
Presubmit-Ready: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Bug: 140421726

Change-Id: I375d937c92d1e88d30720cb0deca52c079fab134
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39249Tested-by: David Turner <digit@google.com>
Kokoro-Presubmit: David Turner <digit@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

This CL simplifies the implementation of vk::Semaphore by getting
rid of vk::Semaphore::Impl (moving its fields into vk::Semaphore
itself).

This requires a minor change to the vk::Semaphore constructor which
now takes an allocator parameter. The latter is used to allocate
the "External" instance on demand. Before the CL, said instance
was 'allocated' from a fixed storage area inside the Impl class.

Note that this doesn't change the external semaphore's behaviour.
In particular, only one implementation can live in the source tree
at the moment, while the Vulkan specification makes it clear that
it should be possible to export a single VkSemaphore to several types
of platform-specific handles (though this doesn't seem to be tested
by the Vulkan-CTS). This last issue will be addressed in a future CL.

Bug: 140421736
Change-Id: I3610d9e7e8cb8e49368b658d157408cbd23ee6db
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39052
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: David Turner <digit@google.com>

Mirror the rather significant changes in Android.mk from:
https://github.com/KhronosGroup/SPIRV-Tools/commit/64f36ea5296ec023f65f75d72196d3ff7ae7c7b2?diff=split

Bug: b/123642959
Change-Id: Iaf82487e768dd0062a97cfe8ebff4e7f85816141
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39876Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Memory offset calculations in 32-bit SIMD elements limit us to
addressing at most 4 GiB. Signed arithmetic further restricts it to
2 GiB.

The legacy OpenGL ES implementation limits image allocations to 1 GiB,
to discourage excessive memory usage which wouldn't typically succeed
on other (mobile) implementations anyway. So until we have a strong use
case which requires more, let's impose the same limit.

Bug: b/146515574
Change-Id: Iaa7bd22789fd20b4e7975d3d2ffee2a8f90b006a
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39828
Presubmit-Ready: Nicolas Capens <nicolascapens@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Addresses comments on:
https://swiftshader-review.googlesource.com/c/SwiftShader/+/38891/1/src/Vulkan/Debug/ID.hpp#34

Bug: b/145351270
Change-Id: If1aabf0b2d2ac7e992bfd59a19eeeba57cb7f84e
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39248Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>

... even if they're not directly used by the SPIR-V shader code.
This will let the debugger display these values in the locals window.

Bug: b/145351270
Change-Id: I1d390e95bf38e89b072ef374e6a35cd8222a45c6
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/38912Tested-by: Ben Clayton <bclayton@google.com>
Presubmit-Ready: Ben Clayton <bclayton@google.com>
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Move the `.clang-format` rule up one level from `./src` to `.`.
Add a `.clang-format` file to `./third_party` that disables formatting.
Update `./src/clang-format-all.sh` to include `./tests`.
Format all source files in `./tests`.

Bug: b/144825072
Change-Id: I01c7ae37a7b1fe8d63ee143107da8acefba20e76
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39873
Kokoro-Presubmit: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

* changes:
  Update SPIRV-Tools to 8aa423930
  Squashed 'third_party/SPIRV-Tools/' changes from c3f22f7cb..8aa423930
  VulkanUnitTests: Fix SPIR-V validation issue
  Update SPIRV-Headers to 204cd131c
  Squashed 'third_party/SPIRV-Headers/' changes from af64a9e82..204cd131c

Making sure we extend 1010102 to the full 16 bit range
fixes 51 of the 53 failures found in ToT dEQP-VK.*a2b10*

Also added a utility function to OR all elements of an
int vector, which allows us to use more vector operations
(as opposed to scalar) while improving the readability.

Bug: b/146633956
Change-Id: If8b946c45cf27f5868d7a97166e21dba565ed72f
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/39768Tested-by: Alexis Hétu <sugoi@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Contains initial support for OpenCL.DebugInfo.100 extended instruction set

Changes:
  8aa423930 Avoid pessimizing std::move (#3124)
  fad92f1e7 Fix typo in validation message (#3122)
  578c5ac13 Change default version for CreatInstBindlessCheckPass to 2 (#3119)
  9215c1b7d Fix convert-relax-to-half invalid code (#3099) (#3106)
  64f36ea52 Support OpenCL.DebugInfo.100 extended instruction set (#3080)
  e01bc6d4e spirv-fuzz: Always add new globals to entry point interfaces (#3113)
  dcb7169bb spirv-fuzz: Transformation to add a new function to a module (#3114)
  2e41d5ece spirv-fuzz: Avoid passing access chains as parameters (#3112)
  e70b009b0 Add support for SPV_KHR_non_semantic_info (#3110)
  38d7fbaad spirv-fuzz: Transformations to add types, constants and variables (#3101)
  fccbc00ac Make Instrumentation format version 2 the default (Step 1) (#3096)
  96354f504 spirv-fuzz: Fuzzer pass to merge blocks (#3097)
  5c019b592 Start SPIRV-Tools v2020.1
  c413b982c Finalize SPIRV-Tools v2019.5
  2afbe9051 Update CHANGES
  00ca4e5bd Don't crash when folding construct of empty struct (#3092)
  0a2b38d08 spirv-fuzz: function outlining fuzzer pass (#3078)
  983b5b4fc spirv-fuzz: Use validator to check break/continue dominance conditions (#3089)
  e82a42860 WebGPU: Array size at most max signed int + 1 (#3077)
  0a5d99d02 Permit the debug instructions in WebGPU SPIR-V - remove from the optimizer (#3083)
  af7410597 graphics robust access: use signed clamp (#3073)
  3ed458604 Folding: perform add and sub on mismatched integer types (#3084)
  47f3eb426 spirv-fuzz: Fix invalid tests (#3079)
  b334829a9 Validate nested constructs (#3068)
  52e9cc930 spirv-fuzz: Improve debugging facilities (#3074)
  54385458c Handle unreachable block when computing register pressure (#3070)
  868ca3954 Improve RegisterSizePasses (#3059)
  f31f26f73 utils/vscode: Add install.bat (#3071)
  03957e8a9 build: cmake: Add support for Fuchsia. (#3062)
  a62012ced Add test with explicit example of stripping reflection info (#3064)
  8312c523e Permit the debug instructions in WebGPU SPIR-V (#3063)
  85f3e93d1 Respect CMAKE_INSTALL_LIBDIR in installed CMake files (#3054)
  45dde9ad6 Add missing dealloc (#3061)
  2ee9aaa28 Initialize binary for use as guard later (#3058)
  0391d0823 Handle OpPhi with no in operands in value numbering (#3056)
  ca703c887 Kill the id-to-func map after wrap-opkill (#3055)
  57b4cb40b Convert stderr and stdout in status to strings on assignment (#3049)

Commands:
  git subtree pull --prefix third_party/SPIRV-Tools https://github.com/KhronosGroup/SPIRV-Tools master --squash

Bug: b/123642959
Change-Id: Id1120acb6126f62c3928774c985fcc09d2619584

8aa423930 Avoid pessimizing std::move (#3124)
fad92f1e7 Fix typo in validation message (#3122)
578c5ac13 Change default version for CreatInstBindlessCheckPass to 2 (#3119)
9215c1b7d Fix convert-relax-to-half invalid code (#3099) (#3106)
64f36ea52 Support OpenCL.DebugInfo.100 extended instruction set (#3080)
e01bc6d4e spirv-fuzz: Always add new globals to entry point interfaces (#3113)
dcb7169bb spirv-fuzz: Transformation to add a new function to a module (#3114)
2e41d5ece spirv-fuzz: Avoid passing access chains as parameters (#3112)
e70b009b0 Add support for SPV_KHR_non_semantic_info (#3110)
38d7fbaad spirv-fuzz: Transformations to add types, constants and variables (#3101)
fccbc00ac Make Instrumentation format version 2 the default (Step 1) (#3096)
96354f504 spirv-fuzz: Fuzzer pass to merge blocks (#3097)
5c019b592 Start SPIRV-Tools v2020.1
c413b982c Finalize SPIRV-Tools v2019.5
2afbe9051 Update CHANGES
00ca4e5bd Don't crash when folding construct of empty struct (#3092)
0a2b38d08 spirv-fuzz: function outlining fuzzer pass (#3078)
983b5b4fc spirv-fuzz: Use validator to check break/continue dominance conditions (#3089)
e82a42860 WebGPU: Array size at most max signed int + 1 (#3077)
0a5d99d02 Permit the debug instructions in WebGPU SPIR-V - remove from the optimizer (#3083)
af7410597 graphics robust access: use signed clamp (#3073)
3ed458604 Folding: perform add and sub on mismatched integer types (#3084)
47f3eb426 spirv-fuzz: Fix invalid tests (#3079)
b334829a9 Validate nested constructs (#3068)
52e9cc930 spirv-fuzz: Improve debugging facilities (#3074)
54385458c Handle unreachable block when computing register pressure (#3070)
868ca3954 Improve RegisterSizePasses (#3059)
f31f26f73 utils/vscode: Add install.bat (#3071)
03957e8a9 build: cmake: Add support for Fuchsia. (#3062)
a62012ced Add test with explicit example of stripping reflection info (#3064)
8312c523e Permit the debug instructions in WebGPU SPIR-V (#3063)
85f3e93d1 Respect CMAKE_INSTALL_LIBDIR in installed CMake files (#3054)
45dde9ad6 Add missing dealloc (#3061)
2ee9aaa28 Initialize binary for use as guard later (#3058)
0391d0823 Handle OpPhi with no in operands in value numbering (#3056)
ca703c887 Kill the id-to-func map after wrap-opkill (#3055)
57b4cb40b Convert stderr and stdout in status to strings on assignment (#3049)

git-subtree-dir: third_party/SPIRV-Tools
git-subtree-split: 8aa423930db37e37086665efcc55944d577c06e5

`LoopDivergentMergePhi` was attempting to use the same merge block for a `OpLoopMerge` and a `OpSelectionMerge`. This is not legal, and was caught by improvements to the SPIR-V tools validator.

The fix is simple: Add a new merge block for the inner `OpSelectionMerge` that immediately jumps to the merge block for the `OpLoopMerge`.

Bug: b/123642959
Bug: b/133440380
Bug: b/133481698
Change-Id: I8b5da04a1cd2db3e44bb5b50f1619a8565e5bc9d