The ENABLE_DBG_MSGS flag controls printing of verbose debug messages to
stdout as each SPIR-V instruction is executed. This is very handy for
performing text diffs when the thread count is reduced to 1 and execution
is deterministic.

Bug: b/140287657
Change-Id: I3b62f0f9f3017087cf7f2786e1c30497cfa05a20
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/31488Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

There's no point writing these to disk, we just overwrite them with the next test.
Unfortunately there's no way to actually disable writing these files.

Think green! Save a SSD!

Change-Id: Ib4f7f9417839acf491eac1828bbae7b067988e06
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44015Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

I missed these when factoring out the root CMakeLists into multiple
files.

Bug: b/145758253
Change-Id: I9c5869b8bab01d823ae064ef7906566afe1e5933
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44108Tested-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Replace set_target_properties(... LINK_FLAGS ...) with
target_link_options. This allows us to keep our link flags variables.
Note that this requires CMake 3.13.

Bug: b/145758253
Change-Id: I8425f7a8e6db087d4000a721971a0dadfed6c3a5
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44110Tested-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

A future change will make use of target_link_options, which was added in
CMake 3.13.

Bug: b/145758253
Change-Id: Icac9609c5d2a9df6039a1d3defeb8db9d3cfd61d
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44109Tested-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

This is so that we can require CMake 3.13 in order to use target_link_options in
a future change.

Bug: b/145758253
Change-Id: Ia01743395453160ae658cc507499a68a79d4eede
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44088Tested-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Going above this causes mayhem on my 3990x.
Given that this is pretty much top end hardware, I don't think anything else in existance could cope with more than 100.

Change-Id: I8b6e4ad09312655c59084aaf5185bb99ee459bdb
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44016Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

vk::DescriptorSet has a header which has a pointer to the descriptor set
layout, which is needed for vkUpdateDescriptorSets. This information
should not be available to shaders during execution.

Bug: b/152227757
Change-Id: I088828a28ab4ed75ae84fb3f103455f70a8bc051
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43968
Presubmit-Ready: Nicolas Capens <nicolascapens@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Only the descriptor binding type is needed. It's an implementation
detail that we currently store the entire original
VkDescriptorSetLayoutBinding structures, which might change in the
future.

This also avoids specializing generated code for things which should
not be treated as pipeline immutable state.

Bug: b/152227757
Change-Id: I20ca7de53e972e77649dec20fcf623f818d4f53c
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43972
Presubmit-Ready: Nicolas Capens <nicolascapens@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Install gcc with aptitude instead of apt-get.

Mirroring workaround from https://github.com/google/amber/pull/844

Bug: none. Fixing kokoro.
Change-Id: I4b57653a539ef78bf037d508dedc47f339c5257d
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44130
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Fix an issue where tryWait() would abort when called on an
unsignaled semaphore. Due to the fact that zx_object_wait_one()
will return ZX_ERR_TIMED_OUT in this case (because the expiration
time was set with zx_clock_get_monotonic() instead of
ZX_TIME_INFINITE).

This fixes an issue that appears when vkWaitForIdle() is called
by certain Fuchsia Vulkan graphics tests.

+ Replace ABORT() calls with DABORT() and proper returns to
  avoid crashing if DCHECK_ALWAYS_ON is not defined at
  build time.

Bug: None
Change-Id: Idaf14ed580c0c3ab326705842bcfe5e23eaad2b8
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44018
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: David Turner <digit@google.com>

In my refactoring of the astc-encoder (https://swiftshader-review.googlesource.com/c/SwiftShader/+/43269),
I had incorrectly set the dependency on astc-encoder in the Vulkan
module, rather than Device. Vulkan depends on Device, and we make use of
astc-encoder public macros in both modules. Thanks to bclayton@ for
finding and reporting this problem.

Fixes: b/154593186
Fixes: b/154593187
Fixes: b/154589883
Fixes: b/154591494
Fixes: b/154591735
Fixes: b/154593191
Fixes: b/154591498
Fixes: b/154593113
Change-Id: I8a71a6471ef85783110afb4f034a611116892b05
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44148
Presubmit-Ready: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Use llvm::VectorType::getNumElements() instead by casting to the
expected subtype.

Bug: b/152339534
Change-Id: I29d2b8d60872f6f4eff4ae2bb38d1738515d11dc
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44068
Presubmit-Ready: Nicolas Capens <nicolascapens@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

When JIT_IN_SEPARATE_THREAD is defined we generate Reactor routines on
LLVM in a new thread. Since concurrent code generation was enabled by
making ::jit thread-local, we need to assign it the pointer of the
thread that creates the new thread.

Bug: b/153803432
Change-Id: Ia7454ac5823d61a1fb139b337d2af2de3a5104af
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44048Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Switch from std::mutex and std::unique_lock to marl::mutex and marl::lock.
The marl versions are TSA-friendly replacements, and do not involve the
marl scheduler.

Bug: b/153194656
Change-Id: Ief3758fb0d52b8fcdaa4208d2c25ad04c2d40108
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43436Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

This CL adds -Wno-deprecated-copy to the CLang compile options, to
avoid multiple errors with Fuchsia's Clang toolchain.

Also updates the package name used to access Fuchsia libaries from
FuchsiaSdk to FuchsiaLibraries. This reflects changes performed in
https://fuchsia-review.googlesource.com/c/fuchsia/+/379673/

Bug: None
Change-Id: Ie97a69a2bd9adaa043b3be2abdbf2e8a6dcd1936
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44017
Presubmit-Ready: David Turner <digit@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: David Turner <digit@google.com>

* changes:
  Update Marl to 748d3c161
  Squashed 'third_party/marl/' changes from 539094011..748d3c161

LLVMReactor used to have a Big Fat Global Mutex over the entire
lifetime of the Nucleus object. This was required as LLVMReactor
used global variables for storing builder state.

Over the past year, there has been significant code cleanup and
global state has been reduced to a couple of globals that can now
be marked thread_local. With all state now being immutable global
or thread local, we are now able to remove the mutex.

ASAN and TSAN checks for our unittests are clean.

Bug: b/153803432
Change-Id: Ibe4019fb783f86e734387db431539e915369b488
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/33484Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

TargetMachine have internal fields mutated during compilation.
In order to support concurrent implementation, each compile needs a new TargetMachine.

Bug: b/153803432
Change-Id: I1a46d8e0193e6176c777d82ce10008afca91ec94
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44013Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Required for making Reactor thread safe.

Bug: b/153803432
Change-Id: I6d36346e22e4ba7a629e280baaa8d293969c366f
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44012Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

https://swiftshader-review.googlesource.com/c/SwiftShader/+/43693 added an assert to SpirvShader::Operand that checks the object is a constant or an intermediate.

SpirvShader::Impl::Debugger::exposeVariable() was constructing an Operand for pointer object types (but was not using it), firing this assert.

Bug: b/129000021
Bug: b/148401179
Change-Id: Id22a92e28b1bb16781b98979d26180b1c022db96
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/44014
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

This changes a CMake FATAL_ERROR message into a WARNING to allow
building SwiftShader for arm64 with an LLVM backend properly.

Otherwise, the CMake configuration aborts prematurely even if
Subzero is not used in the final binaries.

Bug: b/154080675
Change-Id: Ia9ea9896d793071c6d275864955a89a1c262c5d6
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43853
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: David Turner <digit@google.com>
Presubmit-Ready: David Turner <digit@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Changes:
    748d3c161 Add usage recommendations to README.md
    2939fcfed Replace marl::parallelize()
    4b4600208 Fix stupid typo in front page code samples.
    742bba9c8 Fix clang-tidy and go-vet warnings.

Commands:
    ./third_party/update-marl.sh --squash

Bug: b/140546382
Change-Id: Ia8226c11bafbf40907a4622cfdfae3cdcdc303ba

748d3c161 Add usage recommendations to README.md
2939fcfed Replace marl::parallelize()
4b4600208 Fix stupid typo in front page code samples.
742bba9c8 Fix clang-tidy and go-vet warnings.

git-subtree-dir: third_party/marl
git-subtree-split: 748d3c16165c7dcff018273cfaf9d8f26a33333a

Enable `-Wextra` and `-Wunreachable-code-loop-increment`.
Disable `-Wno-unused-parameter`.

Detects the build warning / breakage fixed by https://swiftshader-review.googlesource.com/c/SwiftShader/+/43989.

Bug: none. Reducing chance of future build breakages.
Change-Id: I3001f0ded71f2f71f817581996cc727e345845a3
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43991
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Caught with clang -Wextra.

Bug: b/150155499
Change-Id: I80bdcaeb7f11954e57fd6e27d8747625eba66694
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43990
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Presubmit-Ready: Ben Clayton <bclayton@google.com>

These were generated by running the LLVM CMake with a Fuchsia toolchain,
from the release/10.x branch of https://github.com/ then copying the
include/llvm/Config directory here.

The content is exactly the same as the linux one, except for config.h
due to platform differences.

The generated stripped Fuchsia/x64 binary is 25 MiB (compared to 22 MiB
for LLVM 7.0).

  git clone https://github.com/llvm/llvm-project.git --depth=1
  git checkout release/10.x
  cd llvm-project
  mkdir build && cd build
  cmake ../llvm -DCMAKE_TOOLCHAIN_FILE=/tmp/cmake-sdk/FuchsiaSdkToolchain.cmake -GNinja
  ninja -j32
  mkdir -p $SWIFTSHADER_DIR/third_party/llvm-10.0/configs/fuchsia/include/llvm
  cp -r include/llvm/Config $SWIFTSHADER_DIR/third_party/llvm-10.0/configs/fuchsia/include/llvm/

Bug: b/154082485
Change-Id: Ife30265cf77ee79f14face1f0de8e9b9f8d639b1
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43851
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: David Turner <digit@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

Moved the copy to bin custom command into Vulkan/CMakeLists.txt.
We always copy, just like for the GL targets.

Bug: b/145758253
Change-Id: Idf3af01b453e202610156eff3b302c41fb0ac292
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43974Tested-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>

If a project that uses SwiftShader already imports the Vulkan::Headers
target, and these are different from the version in SwiftShader, we
fail to build because we have a static_assert to match the exact
version. Replace compile-time assert with a runtime assert to make
sure our code still works with different versions of the headers.

This fixes Amber building against SwiftShader.

Bug: b/154215163
Change-Id: I39e32370b51c702ab6d0a9790d06c31537f27472
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43973Tested-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

If parent project sets CMAKE_BUILD_TYPE as a non-cache variable
(e.g. like https://github.com/google/amber does), CMake emits an error
because we're trying to set it as a cache variable.

Bug: b/145758253
Change-Id: Ie169c45dd11674be9e40610f5dd711b8fd5830f8
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43952
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Pass "-C ${SWIFTSHADER_DIR}" to the git submodule command so that it
executes from the SwiftShader directory, and not CMAKE_CURRENT_SOURCE,
which will be an outer directory. Also, make sure to always check if one
target from the submodule already exists to allow parent projects to add
their own version.

Also:
* Refactor submodule init code into function(InitSubmodule)
* Only download libbacktrace if REACTOR_EMIT_DEBUG_INFO is enabled
* Add downloads of json and cppdap if SWIFTSHADER_BUILD_CPPDAP is enabled

Bug: b/145758253
Change-Id: I3bd3f3fcb7f7190f052d2e6f144468bffa0c79da
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43951
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Essentially, do not assume CMAKE_SOURCE_DIR is the SwiftShader root
directory. Instead, we now set SWIFTSHADER_DIR in the root CMakeLists,
and use that everywhere.

Bug: b/145758253
Change-Id: I4968d0dc06ebd8e45791f7cd956e5aeeefddd54f
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43949
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Bug: b/145758253
Change-Id: I15f02fc3f619aac398564f4aacfd30d1dadbe5ef
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43948
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Should have done this in my previous changes. These options aren't meant
to be exposed to dependee targets. This also allows us to disable/remove
these options on a given target, without depender targets adding them
back.

Bug: b/145758253
Change-Id: Ieaba6824554d74e65d313409916bdf9ce4fd6182
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43928
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

marl::blocking_call() calls the passed function on another thread.
Attempting to unlock a mutex on a different thread that locked it is
illegal.

Also lock for destruction.

Bug: b/140421726
Change-Id: I87226176bc3a1a454d9c79a139f4d42e51f66814
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43908Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Change-Id: I47b09ed307bb2cf4159b375171de0d393a017d3d
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43994Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

"ProcessState contains information about an exited process, available after a call to Wait or Run"

As ProcessState is only assigned after `Wait()` has completed, and this is run on another goroutine, it may be nil.
Check before using it.

Change-Id: I319f0703b3547186b53caca8d9d70bc5cd738d34
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43993Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

These used a static for caching the computation, but didn't use a
static initializer. This meant there was a race between the check
for zero and the assignment.

Bug: b/153803432
Change-Id: Id5b91050ced0b8d4811eb32ffed24885816ade7a
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43909Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

This will also be used by the pipelines for dynamic recompilation.

Bug: b/154158882
Change-Id: I0fe16dbc3bc4eed8ffdd90afcd62f255ca8e4505
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43848
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

Fix warning raised by the compiler being too smart for its own good:

  error: loop will run at most once (loop increment never executed) [-Werror,-Wunreachable-code-loop-increment]

Bug: b/153194656
Change-Id: Idca543c13213c6e0bd2d1ff2fdbaa48348f53bf2
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43989Tested-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>