This was tested by using __writeeflags(__readeflags() | 0x40000) to
enable alignment checks on x86.

Bug: b/169957911
Change-Id: Ie97b2fda281548fac94b13abe93213a9a1495b0c
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/48929Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Sean Risser <srisser@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

https://llvm.org/docs/LangRef.html#alloca-instruction does not specify
the type of the 'NumElements' parameter explicitly, but it is multiplied
by sizeof(<type>), which itself is of type size_t and thus 64-bit on
a 64-bit platform. This assumption is made by MemorySanitizer when it
emits instructions corresponding to this multiplication, which get
constant folded and requires both operands to be of the same bit size.

Bug: b/155148722
Change-Id: I432e69bfcea4d40ce7b593637fd242bc1a25a714
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49888Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Runtime symbols referenced by MemorySanitizer instrumentation (e.g.
__msan_warning_noreturn) reside in the executable and are exported from
it (i.e. __attribute__((visibility("default"))). They can be
resolved using dlsym(RTLD_DEFAULT). This works for either an executable
or shared library using Reactor.

Note that this differs from our regular symbol Resolver approach which
lists pairs of symbol names and addresses. There is no public header
which declares these MSan symbols, and if we list symbols only available
in a newer version of LLVM's compiler-rt, the executable won't link or
the shared library won't load.

This change has no effect at this point. A later change will enable
MemorySanitizer intrumentation for Reactor routines.

Bug: b/155148722
Change-Id: I4601e499a2bf7073dbdd68679b74518500ebe217
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49848Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Most of Chromium uses Subzero as the JIT backend for Reactor, but
Fuchsia on ARM64 uses LLVM, which previously broke when switching to
ORCv2 due to LLVM_ENABLE_ABI_BREAKING_CHECKS being set inconsistently
https://swiftshader-review.googlesource.com/c/SwiftShader/+/49908

Bug: b/171897980
Bug: chromium:1140854
Change-Id: I4ab3422387fa966fb76fb90b27532630de69594a
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49852
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Bug:  b/149490161
Change-Id: I15f009d69ddb0725f8171f728195fbf94059c6e7
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49928Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

Reactor backend: LLVM

Change-Id: I26387d86e55d9f56589298a1f364ff090ab9f49a
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49209Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

LLVM_ENABLE_ABI_BREAKING_CHECKS was disabled everywhere except for
Fuchsia. It triggered llvm::Error::assertIsChecked() in the new ORCv2
code because we're supposed to check Expected<> didn't return an error
by casting it to a bool, which we only do in Debug builds.

Since this option costs performance in Release builds and we don't want
to abort on errors (especially benign ones), it should only be enable
for Debug builds. This is also the default behavior for standalone
builds of LLVM:
https://llvm.org/docs/ProgrammersManual.html#abi-breaking-checks

Bug: b/171236524
Bug: chromium:1140854
Change-Id: If886fe9ef61d0f5392949d34a63ed3d7de7aa916
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49908Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Since there does not seem to be any elegant way of
checking whether a class is the first base class in
c++ (given that the 2009 c++ proposal n2965 was
rejected), this python script does it by scanning
all source files for illegal patterns.

The python file was implemented in a way to enable
easy addition of further checks, should need arise.

Bug: b/152253831
Change-Id: I25701a3302579700ccb6222141843d27f38d9069
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49788
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Jari Komppa <jari@siru.fi>

The target triple is used by MemorySanitizer to determine where to map
the shadow and origin data, which is OS specific. When not set, the
MemorySanitizer::initializeModule() method, which inserts an __msan_init
call to the module's constructor list, produces an
"unsupported operating system" error.

Note that JITs like Reactor have no use for a module constructor list
(__msan_init gets called during executable startup), but there is
currently no way to skip this initialization. Leaving it unused is fine
though. Likewise, setting the target triple has not other known JIT uses
at the moment, but it can't hurt.

Bug: b/155148722
Change-Id: Ia31d8c71ee495aadd91530a8d1c5fdc93add8e01
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49853
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

* changes:
  Update SPIR-V Tools to f7da52775
  Squashed 'third_party/SPIRV-Tools/' changes from 82b378d671..f7da527757

We were mixing uses of ::llvm:: and llvm::
Since it is a well-known namespace, there's virtually no risk of
unintentionally reusing the same name.
This follows the recommendation of go/totw/151.

Bug: consistency cleanup
Change-Id: I914f7a2d7b76f21f1f0758506319f2dd47cc1832
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49828Tested-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

MemorySanitizer doesn't support instrumenting masked_gather and
masked_scatter LLVM intrinsics. Its visitIntrinsicInst() method ends up
calling handleUnknownIntrinsic(), which silently doesn't handle it and
subsequently visitInstruction() checks all operands for poisoned bits.
In the case of a scatter, a 0 bit in the mask means the corresponding
element doesn't get written, so it doesn't matter if it's uninitialized
data. The current implementation leads to false positives.

Work around it by emulating gather and scatter as element-wise loads and
stores. This can be correctly instrumented by MemorySanitizer.

Note this change has no effect currently since we don't support MSan
instrumentation for Reactor yet. We just unpoison all stores. Previously
we did that in element-wise manner after the intrinsic executes. Now
it's done as part of the element stores.

Bug: b/155148722
Change-Id: I9058cd926667fb6df5d9626bc87fb2d0a596771b
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49809Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

A statement like "If(Bool(b)) {}" would lead to interpreting part of its
maco expansion "IfElseData ifElse__(Bool(b));" as a function declaration.

See https://en.wikipedia.org/wiki/Most_vexing_parse

Fix it by using C++11 universal initialization syntax.

Bug: b/155148722
Change-Id: I64d0a28f64a4735a61776a44dfb7560d2854cde3
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49808
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Changes:
    f7da527757 Temporarily add EmptyPass to prevent glslang from failing (#4004)

Commands:
    ./third_party/update-spirvtools.sh

Bug: b/123642959
Change-Id: If776b87d8c72e12a71e864ecd9a0ce4d61f9ddbc

f7da527757 Temporarily add EmptyPass to prevent glslang from failing (#4004)

git-subtree-dir: third_party/SPIRV-Tools
git-subtree-split: f7da527757140ae701be58274ce6db2f4234d9ff

UBSAN pointed out that memcpy cannot have nullptr dest or source,
which I've also fixed; however, this also made me realize a latent bug:
the dest and source args needed to be swapped. It seems this bug hasn't
bit us yet because the normal use-case is to resize the bitvector, and
then call 'set' to set all bits to 0 or 1. In any case, still worth
fixing.

Bug: b/171581990
Change-Id: Iaaa7e019b53ae661f49c97195f2d144b5715f164
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49528
Presubmit-Ready: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

This change adds third_party directory to the ignore list
for the pre-submit script.

Bug: b/171873846
Change-Id: Ic1a6fee89c76e8ca3526c95e29335eacf2e505f1
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49768
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Jari Komppa <jari@siru.fi>

* changes:
  Update SPIR-V Tools to 82b378d67
  Squashed 'third_party/SPIRV-Tools/' changes from 8a5500656..82b378d67
  Update SPIR-V Headers to 7845730ca
  Squashed 'third_party/SPIRV-Headers/' changes from 979924c8b..7845730ca

This change makes sure the memset template is not
used with classes that contain virtual members.

Bug: b/152253831
Change-Id: I07d4777f33a9bd55c8f99d8b50a80c13df5d5805
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49648
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Jari Komppa <jari@siru.fi>

Changes:
    82b378d67 spirv-opt: Add support to prevent functions from being inlined if they have DontInline flag (#3858)
    56d0f5035 Propagate OpLine to all applied instructions in spirv-opt (#3951)
    7403dfafd CMake: Add SPIRV_TOOLS_BUILD_STATIC flag (#3910)
    25ee27576 Avoid copying a ref in a loop (#4000)
    5edb328e8 spirv-val: Allow the ViewportIndex and Layer built-ins on SPIR-V 1.5 (#3986)
    cbd1fa6c4 Simplify logic to decide whether CCP modified the IR (#3997)
    df4198e50 Add DebugValue for DebugDecl invisible to value assignment (#3973)
    34ae8a475 Fix bounds check instrumentation to handle 16-bit values (#3983)
    abe2eff36 spirv-fuzz: Add expand vector reduction transformation (#3869)
    99ad4f1e2 spirv-fuzz: Don't replace irrelevant indices in OpAccessChain (#3988)
    895dafcc1 spirv-fuzz: Add FuzzerPassAddCompositeExtract (#3904)
    69f07da41 spirv-fuzz: Fix mismatch with shrinker step limit (#3985)
    9223493f4 spirv-fuzz: Fix off-by-one error in replayer (#3982)
    4f5423187 spirv-fuzz: Get order right for OpSelect arguments (#3974)
    88f7bcb6a spirv-fuzz: Do not add synonym-creating loops in dead blocks (#3975)
    64eaa9832 spirv-fuzz: Skip OpTypeSampledImage when propagating up (#3976)
    53f2a69c1 spirv-fuzz: Pass OpUndef in function call if needed (#3978)
    de2c0ba20 spirv-fuzz: Fix off-by-one in TransformationCompositeConstruct (#3979)
    dc4a18228 spirv-fuzz: Tolerate absent ids in data synonym fact management (#3966)
    8496780f5 spirv-fuzz: Fix to id availability (#3971)
    3b7aebca4 spirv-fuzz: Fix operand types (#3962)
    6fe34cdf1 Update SPIRV-Headers revision in DEPS file (#3961)
    a8d7062fe spirv-fuzz: Don't flatten conditional if condition is irrelevant (#3944)
    a3d5378df spirv-fuzz: Do not produce OpPhis of type OpTypeSampledImage (#3964)
    7edd0525b spirv-fuzz: Restrict fuzzer pass to reachable blocks (#3970)
    f03779a1a spirv-fuzz: Handle more types when extending OpPhi instructions (#3969)
    c794b7046 spirv-fuzz: Skip early terminator wrappers when merging returns (#3968)
    dc9c6407d spirv-fuzz: Avoid irrelevant constants in synonym-creating loops (#3967)
    26954c281 spirv-fuzz: Skip dead blocks in FuzzerPassAddOpPhiSynonyms (#3965)
    5600fb85b spirv-fuzz: Avoid the type manager when looking for struct types (#3963)
    ba15b5886 spirv-fuzz: Fix to TransformationDuplicateRegionWithSelection (#3941)
    598435004 spirv-fuzz: Skip OpFunction when replacing irrelevant ids (#3932)
    8362eae55 spirv-fuzz: Use component-wise selectors when flattening conditional branches (#3921)
    ebe0ea09f Add SPV_EXT_shader_image_int64 (#3852)
    a1d38174b Support SPV_KHR_fragment_shading_rate (#3943)
    6fac705e7 spirv-val: Fix validation of OpPhi instructions (#3919)
    b1350659b spirv-fuzz: Avoid void struct member when outlining functions (#3936)
    fd0f295da spirv-fuzz: Do not allow Block-decorated structs when adding parameters (#3931)
    5d7893b37 spirv-fuzz: Fix to operand id type (#3937)
    cd1d3b6e0 spirv-fuzz: Handle dead blocks in TransformationEquationInstruction (#3933)
    6cdae9da7 spirv-fuzz: Do not allow sampled image load when flattening conditionals (#3930)
    7e1825a59 spirv-fuzz: Take care of OpPhi instructions when inlining (#3939)
    502e98295 spirv-fuzz: Fix to TransformationInlineFunction (#3913)
    bf1a11dab spirv-fuzz: Wrap early terminators before merging returns (#3925)
    fd3948e16 Add DebugValue for function param regardless of scope (#3923)
    663d050a9 Temporary fix to make GoogleTest compile. (#3922)
    dd534e877 spirv-fuzz: Lower probability of adding bit instruction synonyms (#3917)
    53aeba10c spirv-fuzz: Fix handling of OpPhi in FlattenConditionalBranch (#3916)
    5c64374dd spirv-fuzz: Avoid creating blocks without parents (#3908)
    57b3723c5 spirv-fuzz: Do not allow creation of constants of block-decorated structs (#3903)
    12ca825a6 spirv-fuzz: Fixes related to irrelevant ids (#3901)
    2e6cf706e spirv-fuzz: Fix to transformation that adds a synonym via a loop (#3898)
    4b884928d spirv-fuzz: Fix to duplicate region with selection (#3896)
    c2553a315 spirv-fuzz: Do not expose synonym facts for non-existent ids (#3891)
    360228785 spirv-fuzz: Do not add synonyms involving irrelevant ids (#3890)
    d52f79122 spirv-fuzz: Do not replace irrelevant ids that are not in blocks (#3892)
    e02265992 spirv-fuzz: Wrap OpKill and similar in function calls (#3884)
    11d592422 Update val to handle reversed instruction sections. (#3887)
    fc8264854 spirv-fuzz: Replace dead-block terminators with OpKill etc. (#3882)
    63cc22d64 spirv-fuzz: TransformationPropagateInstructionDown (#3692)
    65b2a9e81 spirv-fuzz: Transformation to add wrappers for OpKill and similar (#3881)
    624b16cd0 spirv-fuzz: Use overflow ids when duplicating regions (#3878)
    bd0dd9cef spirv-fuzz: Fix rvalue references (#3883)
    f15133788 spirv-fuzz: Avoid using block-decorated structs in transformations (#3877)
    0e8553072 spirv-fuzz: Use overflow ids when inlining functions (#3880)
    67f8e2edd Debug info preservation in convert-local-access-chains pass (#3835)
    d91afd8de spirv-fuzz: Refactor fact manager tests (#3879)
    b920b620a spirv-fuzz: Integrate spirv-reduce with shrinker (#3849)
    74a711a76 spirv-fuzz: Only recommend passes when a pass had an effect (#3863)
    fc7860e2d spirv-fuzz: Merge the return instructions in a function (#3838)
    57abfd88c Debug info preservation in redundancy-elimination pass (#3839)
    e24603836 Debug info preservation in if-conversion pass (#3861)
    16cc197c8 spirv-fuzz: Refactor conditions in the fact manager (#3867)
    615fbe6cb spirv-fuzz: TransformationWrapRegionInSelection (#3674)
    f2b8a4ee5 spirv-fuzz: Enable some passes with high probability (#3860)
    50e04f612 spirv-fuzz: Do not add constants for Block-decorated structs (#3862)
    8edd79ddc spirv-fuzz: Fix to FuzzerPassFlattenConditionalBranches (#3865)
    719bade4f spirv-fuzz: Fix TransformationRecordSynonymousConstants (#3868)
    009facc97 spirv-fuzz: Fix flatten conditional branch transformation (#3859)
    fcb22ecf0 spirv-fuzz: Report fresh ids in transformations (#3856)
    c6ca885c0 [spirv-dis] Add some context comments to disassembly. (#3847)
    446adb05f Fix use-after-move in val/validate.cpp (#3848)
    4b07d50cd spirv-fuzz: Fix bug in TransformationDuplicateRegionWithSelection (#3819)
    fec56146a spirv-fuzz: Support OpNot bit instruction case (#3841)
    9e17b9d07 spirv-fuzz: Return IR and transformation context after replay (#3846)
    e12087d6c spirv-fuzz: fix test (#3845)
    9edeeafdb spirv-fuzz: Use unique_ptr<FactManager> in TransformationContext (#3844)
    50dacda55 Start SPIRV-Tools v2020.6
    b27e039c6 Finalize SPIRV-Tools v2020.5
    a5903a969 Update CHANGES
    330c72549 spirv-fuzz: Support dead blocks in TransformationAddSynonym (#3832)
    36185f8b0 spirv-fuzz: Move IRContext parameter into constructor (#3837)
    0e7fe4d35 Add missing backticks around <result-id> (#3840)
    d1bb98fd4 Validate SPIRV Version number when parsing binary header (#3834)
    67525bded spirv-fuzz: Create synonym of int constant using a loop (#3790)
    7cc4b4d2c Fix compiler error on macOS with XCode12 (#3836)
    5a5b750aa spirv-fuzz: Handle OpPhis in TransformationInlineFunction (#3833)
    0a1fb588c Update CHANGES
    125b64241 spirv-fuzz: Refactor fuzzer, replayer and shrinker (#3818)
    60ce96e2f spirv-fuzz: Add pass recommendations (#3757)
    2945963cc spirv-fuzz: Consider all ids from dead blocks irrelevant (#3795)
    50ae4c5f4 Fix header guard macros (#3811)
    296e9c7bc spirv-fuzz: Fix TransformationDuplicateRegionWithSelection (#3815)
    937a757f0 spirv-val: Add DeviceIndex (#3812)
    34ef0c3fd Fix missed modification flagging (#3814)
    748edbf8c spirv-fuzz: Use an irrelevant id for the unused components (#3810)
    8d49fb2f4 spirv-fuzz: Improvements to random number generation (#3809)
    7e28d809c Add buffer oob check to bindless instrumentation (#3800)
    8fc504110 spirv-fuzz: Remove CanFindOrCreateZeroConstant (#3807)
    e8ce4355a spirv-fuzz: Add bit instruction synonym transformation (#3775)
    e7c84feda spirv-fuzz: Skip unreachable blocks (#3729)
    f20b523cb Fix build errors (#3804)
    3131686d2 spirv-fuzz: Handle invalid ids in fact manager (#3742)
    4c239bd81 spirv-fuzz: Support memory instructions MoveInstructionDown (#3700)
    1e1c308de spirv-fuzz: Pass submanagers to other submanagers when necessary (#3796)
    f62357e7b spirv-fuzz: Transformation to flatten conditional branch (#3667)
    5df930054 spirv-val: Add BaseInstance, BaseVertex, DrawIndex, and ViewIndex (#3782)
    286b3095d Properly mark IR changed if instruction folder creates more than one constant. (#3799)
    726af6f78 Add missing file to BUILD.gn (#3798)
    244e6c1be spirv-fuzz: Add TransformationDuplicateRegionWithSelection (#3773)
    5dcb576b6 spirv-reduce: Support reducing a specific function (#3774)
    de7d57984 spirv-reduce: Refactoring (#3793)
    ed9863e46 Favour 'integrity' over 'coherence' as a replacement for 'sanity'. (#3619)
    8743d385f spirv-fuzz: Fix header guards in transformations/fuzzer passes (#3784)
    2de7d3af0 spirv-fuzz: Add SPIRV_FUZZ_PROTOC_COMMAND (#3789)
    e589d0d57 Add missing include (#3788)
    a715b1b40 Improve spirv-fuzz CMake code (#3781)
    a187dd58a Allow SPV_KHR_8bit_storage extension. (#3780)
    1ab52e54a spirv-opt: Add function to compute nesting depth of a block (#3771)
    fd05605be spirv-fuzz: Transformation to convert OpSelect to conditional branch (#3681)
    2c60d16a6 spirv-val: Add Vulkan VUID labels to BuiltIn (#3756)
    c341f7a6c spirv-fuzz: Add support for BuiltIn decoration (#3736)
    c278dada9 spirv-fuzz: Fix GetIdEquivalenceClasses (#3767)
    788468408 spirv-fuzz: Replace id in OpPhi coming from a dead predecessor (#3744)
    3daabd321 spirv-fuzz: Transformation to replace the use of an irrelevant id (#3697)
    d7f078f27 spirv-fuzz: TransformationMutatePointer (#3737)
    43a518601 spirv-fuzz: Compute interprocedural loop nesting depth of blocks (#3753)
    8a0ebd40f Correctly replace debug lexical scope of instruction (#3718)
    f428aa39c spirv-fuzz: Remove opaque pointer design pattern (#3755)
    08291a3a9 spirv-fuzz: Create synonym via OpPhi and existing synonyms (#3701)
    7e4948b2a Add LoopNestingDepth function to StructuredCFGAnalysis (#3754)
    50cf38b8c spirv-fuzz: Do not make synonyms of void result ids (#3747)
    bceab9fab Do not register DebugFunction for functions optimized away. (#3749)
    e02f178a7 Handle DebugScope in compact-ids pass (#3724)
    9e26ae045 spirv-fuzz: Overflow ids (#3734)
    2205254cf Fix DebugNoScope to not output InlinedAt operand. (#3748)
    230f363e6 spirv-fuzz: Split the fact manager into multiple files (#3699)
    5adc5ae64 spirv-fuzz: Add inline function transformation (#3517)
    1341b58a8 spirv-fuzz: Fix MaybeGetZeroConstant (#3740)
    12df3cafe Fix SSA-rewrite to remove DebugDeclare for variables without loads (#3719)
    3f8501de9 Add undef for inlined void function (#3720)
    4dd122392 spirv-fuzz: Add words instead of logical operands (#3728)
    b79773a35 CCP should mark IR changed if it created new constants. (#3732)
    a711c594b spirv-fuzz: add FuzzerPassAddCompositeInserts (#3606)
    582c276d4 spirv-fuzz: Support pointer types in FuzzerPassAddParameters (#3627)
    3434cb0b0 Let ADCE pass check DebugScope (#3703)
    ee7f0c882 spirv-opt: Implement opt::Function::HasEarlyReturn function (#3711)
    e28436f2b spirv-fuzz: Check termination instructions when donating modules (#3710)
    1023dd7a0 Fix -Wrange-loop-analysis warning (#3712)
    82f4bf128 spirv-fuzz: Check header dominance when adding dead block (#3694)
    b8de4f57e Allow DebugTypeTemplate for Type operand (#3702)
    c20995ef8 spirv-fuzz: Improve code coverage of tests (#3686)
    eade36db2 spirv-fuzz: Fuzzer pass to randomly apply loop preheaders (#3668)
    72ea7bec4 spirv-fuzz: Support identical predecessors in TransformationPropagateInstructionUp (#3689)
    b4c4da3e7 Improve non-semantic instruction handling in the optimizer (#3693)
    948577c5d Fix the bug (#3680)
    df859f77d spirv-fuzz: Check integer and float width capabilities (#3670)
    2641d3351 spirv-fuzz: consider additional access chain instructions (#3672)
    5e5929455 spirv-fuzz: Ignore specialization constants (#3664)
    1435e427d Fix the bug (#3683)
    be099cde1 spirv-fuzz: Fix width in FuzzerPassAddEquationInstructions (#3685)
    f0ca96d12 Preserve debug info in dead-insert-elim pass (#3652)
    0d629b903 Validate more OpenCL.DebugInfo.100 instructions (#3684)
    13a65b1ae Only validation locations for appropriate execution models (#3656)
    fd3cabd8b spirv-fuzz: Fix in operand type assertion (#3666)
    f50553867 spirv-opt: Add spvOpcodeIsAccessChain (#3682)
    b7056e7e0 spirv-fuzz: FuzzerPassPropagateInstructionsUp (#3478)
    8e1380996 Handle no index access chain in local access chain convert (#3678)
    bdeeae78a Roll 2 dependencies (#3677)
    2990a2192 Avoid using /MP4 for clang on windows. (#3662)
    7b2dd11dd spirv-fuzz: TransformationReplaceAddSubMulWithCarryingExtended (#3598)
    6d7f34fbf spirv-fuzz: Add TransformationMakeVectorOperationDynamic (#3597)
    d29eac95a spirv-fuzz: iterate over blocks in replace linear algebra pass (#3654)
    efc85ff66 spirv-fuzz: make outliner pass use additional transformations (#3604)
    5fd92a7e0 OpenCL.DebugInfo.100 DebugTypeArray with variable size (#3549)
    3f33a9aa5 spirv-opt: Improve the code of the Instruction class (#3610)
    0419751b0 spirv-fuzz: Handle OpPhis in livesafe functions (#3642)
    a10e76059 spirv-fuzz: Handle OpPhi during constant obfuscation (#3640)
    28f32ca53 spirv-fuzz: Fix FuzzerPassCopyObjects (#3638)
    8bc27a1cf spirv-fuzz: Remove OpFunctionCall operands in correct order (#3630)
    d9c73ebd9 spirv-fuzz: Handle capabilities during module donation (#3651)
    9f2223602 spirv-fuzz: Refactor boilerplate in TransformationAddParameter (#3625)
    92a71657f spirv-fuzz: TransformationMoveInstructionDown (#3477)
    b78f4b151 Remove DebugDeclare only for target variables in ssa-rewrite (#3511)
    91cea06ab Fix typo in ASAN CI build (#3623)
    2aaa8653d spirv-fuzz: Transformation to add loop preheader (#3599)
    96bcc8274 spirv-fuzz: Pass to replace int operands with ints of opposite signedness (#3612)
    ebaefda66 Debug info preservation in loop-unroll pass (#3548)
    50300450a Validator support for non-semantic clspv reflection (#3618)
    ab4fe12a4 spirv-fuzz: Fix memory bugs (#3622)
    c6e6597c4 spirv-fuzz: Implement the OpOuterProduct linear algebra case (#3617)
    054f034ea spirv-fuzz: Compute corollary facts from OpBitcast (#3538)
    a1ea15c90 Update some language usage. (#3611)
    863b8e3d3 spirv-fuzz: Relax type constraints in DataSynonym facts (#3602)
    7e75fea9e spirv-fuzz: Remove non-deterministic behaviour (#3608)
    f9b088fe0 Avoid use of 'sanity' and 'sanity check' in the code base (#3585)
    150be20d4 spirv-fuzz: Add condition to make functions livesafe (#3587)
    ce16ccf38 Rolling 4 dependencies (#3601)
    1dfc6fc7e spirv-fuzz: Implement the OpTranspose linear algebra case (#3589)
    b63f0e5ed Fix SyntaxWarning in Python 3.8 (#3388)
    6aed7ffbc CMake: Enable building with BUILD_SHARED_LIBS=1 (#3490)
    31c821393 Avoid operand type range checks (#3379)
    6a3eb679b Preserve debug info in scalar replacement pass (#3461)
    2796840d2 Update OpenCL capabilities validation (#3149)
    b25ee93c2 build(deps): bump lodash from 4.17.15 to 4.17.19 in /tools/sva (#3596)

Commands:
    ./third_party/update-spirvtools.sh

Bug: b/123642959
Change-Id: Idde3643775d090b11dcb37ec2b3bd5a93c5e56fc

82b378d67 spirv-opt: Add support to prevent functions from being inlined if they have DontInline flag (#3858)
56d0f5035 Propagate OpLine to all applied instructions in spirv-opt (#3951)
7403dfafd CMake: Add SPIRV_TOOLS_BUILD_STATIC flag (#3910)
25ee27576 Avoid copying a ref in a loop (#4000)
5edb328e8 spirv-val: Allow the ViewportIndex and Layer built-ins on SPIR-V 1.5 (#3986)
cbd1fa6c4 Simplify logic to decide whether CCP modified the IR (#3997)
df4198e50 Add DebugValue for DebugDecl invisible to value assignment (#3973)
34ae8a475 Fix bounds check instrumentation to handle 16-bit values (#3983)
abe2eff36 spirv-fuzz: Add expand vector reduction transformation (#3869)
99ad4f1e2 spirv-fuzz: Don't replace irrelevant indices in OpAccessChain (#3988)
895dafcc1 spirv-fuzz: Add FuzzerPassAddCompositeExtract (#3904)
69f07da41 spirv-fuzz: Fix mismatch with shrinker step limit (#3985)
9223493f4 spirv-fuzz: Fix off-by-one error in replayer (#3982)
4f5423187 spirv-fuzz: Get order right for OpSelect arguments (#3974)
88f7bcb6a spirv-fuzz: Do not add synonym-creating loops in dead blocks (#3975)
64eaa9832 spirv-fuzz: Skip OpTypeSampledImage when propagating up (#3976)
53f2a69c1 spirv-fuzz: Pass OpUndef in function call if needed (#3978)
de2c0ba20 spirv-fuzz: Fix off-by-one in TransformationCompositeConstruct (#3979)
dc4a18228 spirv-fuzz: Tolerate absent ids in data synonym fact management (#3966)
8496780f5 spirv-fuzz: Fix to id availability (#3971)
3b7aebca4 spirv-fuzz: Fix operand types (#3962)
6fe34cdf1 Update SPIRV-Headers revision in DEPS file (#3961)
a8d7062fe spirv-fuzz: Don't flatten conditional if condition is irrelevant (#3944)
a3d5378df spirv-fuzz: Do not produce OpPhis of type OpTypeSampledImage (#3964)
7edd0525b spirv-fuzz: Restrict fuzzer pass to reachable blocks (#3970)
f03779a1a spirv-fuzz: Handle more types when extending OpPhi instructions (#3969)
c794b7046 spirv-fuzz: Skip early terminator wrappers when merging returns (#3968)
dc9c6407d spirv-fuzz: Avoid irrelevant constants in synonym-creating loops (#3967)
26954c281 spirv-fuzz: Skip dead blocks in FuzzerPassAddOpPhiSynonyms (#3965)
5600fb85b spirv-fuzz: Avoid the type manager when looking for struct types (#3963)
ba15b5886 spirv-fuzz: Fix to TransformationDuplicateRegionWithSelection (#3941)
598435004 spirv-fuzz: Skip OpFunction when replacing irrelevant ids (#3932)
8362eae55 spirv-fuzz: Use component-wise selectors when flattening conditional branches (#3921)
ebe0ea09f Add SPV_EXT_shader_image_int64 (#3852)
a1d38174b Support SPV_KHR_fragment_shading_rate (#3943)
6fac705e7 spirv-val: Fix validation of OpPhi instructions (#3919)
b1350659b spirv-fuzz: Avoid void struct member when outlining functions (#3936)
fd0f295da spirv-fuzz: Do not allow Block-decorated structs when adding parameters (#3931)
5d7893b37 spirv-fuzz: Fix to operand id type (#3937)
cd1d3b6e0 spirv-fuzz: Handle dead blocks in TransformationEquationInstruction (#3933)
6cdae9da7 spirv-fuzz: Do not allow sampled image load when flattening conditionals (#3930)
7e1825a59 spirv-fuzz: Take care of OpPhi instructions when inlining (#3939)
502e98295 spirv-fuzz: Fix to TransformationInlineFunction (#3913)
bf1a11dab spirv-fuzz: Wrap early terminators before merging returns (#3925)
fd3948e16 Add DebugValue for function param regardless of scope (#3923)
663d050a9 Temporary fix to make GoogleTest compile. (#3922)
dd534e877 spirv-fuzz: Lower probability of adding bit instruction synonyms (#3917)
53aeba10c spirv-fuzz: Fix handling of OpPhi in FlattenConditionalBranch (#3916)
5c64374dd spirv-fuzz: Avoid creating blocks without parents (#3908)
57b3723c5 spirv-fuzz: Do not allow creation of constants of block-decorated structs (#3903)
12ca825a6 spirv-fuzz: Fixes related to irrelevant ids (#3901)
2e6cf706e spirv-fuzz: Fix to transformation that adds a synonym via a loop (#3898)
4b884928d spirv-fuzz: Fix to duplicate region with selection (#3896)
c2553a315 spirv-fuzz: Do not expose synonym facts for non-existent ids (#3891)
360228785 spirv-fuzz: Do not add synonyms involving irrelevant ids (#3890)
d52f79122 spirv-fuzz: Do not replace irrelevant ids that are not in blocks (#3892)
e02265992 spirv-fuzz: Wrap OpKill and similar in function calls (#3884)
11d592422 Update val to handle reversed instruction sections. (#3887)
fc8264854 spirv-fuzz: Replace dead-block terminators with OpKill etc. (#3882)
63cc22d64 spirv-fuzz: TransformationPropagateInstructionDown (#3692)
65b2a9e81 spirv-fuzz: Transformation to add wrappers for OpKill and similar (#3881)
624b16cd0 spirv-fuzz: Use overflow ids when duplicating regions (#3878)
bd0dd9cef spirv-fuzz: Fix rvalue references (#3883)
f15133788 spirv-fuzz: Avoid using block-decorated structs in transformations (#3877)
0e8553072 spirv-fuzz: Use overflow ids when inlining functions (#3880)
67f8e2edd Debug info preservation in convert-local-access-chains pass (#3835)
d91afd8de spirv-fuzz: Refactor fact manager tests (#3879)
b920b620a spirv-fuzz: Integrate spirv-reduce with shrinker (#3849)
74a711a76 spirv-fuzz: Only recommend passes when a pass had an effect (#3863)
fc7860e2d spirv-fuzz: Merge the return instructions in a function (#3838)
57abfd88c Debug info preservation in redundancy-elimination pass (#3839)
e24603836 Debug info preservation in if-conversion pass (#3861)
16cc197c8 spirv-fuzz: Refactor conditions in the fact manager (#3867)
615fbe6cb spirv-fuzz: TransformationWrapRegionInSelection (#3674)
f2b8a4ee5 spirv-fuzz: Enable some passes with high probability (#3860)
50e04f612 spirv-fuzz: Do not add constants for Block-decorated structs (#3862)
8edd79ddc spirv-fuzz: Fix to FuzzerPassFlattenConditionalBranches (#3865)
719bade4f spirv-fuzz: Fix TransformationRecordSynonymousConstants (#3868)
009facc97 spirv-fuzz: Fix flatten conditional branch transformation (#3859)
fcb22ecf0 spirv-fuzz: Report fresh ids in transformations (#3856)
c6ca885c0 [spirv-dis] Add some context comments to disassembly. (#3847)
446adb05f Fix use-after-move in val/validate.cpp (#3848)
4b07d50cd spirv-fuzz: Fix bug in TransformationDuplicateRegionWithSelection (#3819)
fec56146a spirv-fuzz: Support OpNot bit instruction case (#3841)
9e17b9d07 spirv-fuzz: Return IR and transformation context after replay (#3846)
e12087d6c spirv-fuzz: fix test (#3845)
9edeeafdb spirv-fuzz: Use unique_ptr<FactManager> in TransformationContext (#3844)
50dacda55 Start SPIRV-Tools v2020.6
b27e039c6 Finalize SPIRV-Tools v2020.5
a5903a969 Update CHANGES
330c72549 spirv-fuzz: Support dead blocks in TransformationAddSynonym (#3832)
36185f8b0 spirv-fuzz: Move IRContext parameter into constructor (#3837)
0e7fe4d35 Add missing backticks around <result-id> (#3840)
d1bb98fd4 Validate SPIRV Version number when parsing binary header (#3834)
67525bded spirv-fuzz: Create synonym of int constant using a loop (#3790)
7cc4b4d2c Fix compiler error on macOS with XCode12 (#3836)
5a5b750aa spirv-fuzz: Handle OpPhis in TransformationInlineFunction (#3833)
0a1fb588c Update CHANGES
125b64241 spirv-fuzz: Refactor fuzzer, replayer and shrinker (#3818)
60ce96e2f spirv-fuzz: Add pass recommendations (#3757)
2945963cc spirv-fuzz: Consider all ids from dead blocks irrelevant (#3795)
50ae4c5f4 Fix header guard macros (#3811)
296e9c7bc spirv-fuzz: Fix TransformationDuplicateRegionWithSelection (#3815)
937a757f0 spirv-val: Add DeviceIndex (#3812)
34ef0c3fd Fix missed modification flagging (#3814)
748edbf8c spirv-fuzz: Use an irrelevant id for the unused components (#3810)
8d49fb2f4 spirv-fuzz: Improvements to random number generation (#3809)
7e28d809c Add buffer oob check to bindless instrumentation (#3800)
8fc504110 spirv-fuzz: Remove CanFindOrCreateZeroConstant (#3807)
e8ce4355a spirv-fuzz: Add bit instruction synonym transformation (#3775)
e7c84feda spirv-fuzz: Skip unreachable blocks (#3729)
f20b523cb Fix build errors (#3804)
3131686d2 spirv-fuzz: Handle invalid ids in fact manager (#3742)
4c239bd81 spirv-fuzz: Support memory instructions MoveInstructionDown (#3700)
1e1c308de spirv-fuzz: Pass submanagers to other submanagers when necessary (#3796)
f62357e7b spirv-fuzz: Transformation to flatten conditional branch (#3667)
5df930054 spirv-val: Add BaseInstance, BaseVertex, DrawIndex, and ViewIndex (#3782)
286b3095d Properly mark IR changed if instruction folder creates more than one constant. (#3799)
726af6f78 Add missing file to BUILD.gn (#3798)
244e6c1be spirv-fuzz: Add TransformationDuplicateRegionWithSelection (#3773)
5dcb576b6 spirv-reduce: Support reducing a specific function (#3774)
de7d57984 spirv-reduce: Refactoring (#3793)
ed9863e46 Favour 'integrity' over 'coherence' as a replacement for 'sanity'. (#3619)
8743d385f spirv-fuzz: Fix header guards in transformations/fuzzer passes (#3784)
2de7d3af0 spirv-fuzz: Add SPIRV_FUZZ_PROTOC_COMMAND (#3789)
e589d0d57 Add missing include (#3788)
a715b1b40 Improve spirv-fuzz CMake code (#3781)
a187dd58a Allow SPV_KHR_8bit_storage extension. (#3780)
1ab52e54a spirv-opt: Add function to compute nesting depth of a block (#3771)
fd05605be spirv-fuzz: Transformation to convert OpSelect to conditional branch (#3681)
2c60d16a6 spirv-val: Add Vulkan VUID labels to BuiltIn (#3756)
c341f7a6c spirv-fuzz: Add support for BuiltIn decoration (#3736)
c278dada9 spirv-fuzz: Fix GetIdEquivalenceClasses (#3767)
788468408 spirv-fuzz: Replace id in OpPhi coming from a dead predecessor (#3744)
3daabd321 spirv-fuzz: Transformation to replace the use of an irrelevant id (#3697)
d7f078f27 spirv-fuzz: TransformationMutatePointer (#3737)
43a518601 spirv-fuzz: Compute interprocedural loop nesting depth of blocks (#3753)
8a0ebd40f Correctly replace debug lexical scope of instruction (#3718)
f428aa39c spirv-fuzz: Remove opaque pointer design pattern (#3755)
08291a3a9 spirv-fuzz: Create synonym via OpPhi and existing synonyms (#3701)
7e4948b2a Add LoopNestingDepth function to StructuredCFGAnalysis (#3754)
50cf38b8c spirv-fuzz: Do not make synonyms of void result ids (#3747)
bceab9fab Do not register DebugFunction for functions optimized away. (#3749)
e02f178a7 Handle DebugScope in compact-ids pass (#3724)
9e26ae045 spirv-fuzz: Overflow ids (#3734)
2205254cf Fix DebugNoScope to not output InlinedAt operand. (#3748)
230f363e6 spirv-fuzz: Split the fact manager into multiple files (#3699)
5adc5ae64 spirv-fuzz: Add inline function transformation (#3517)
1341b58a8 spirv-fuzz: Fix MaybeGetZeroConstant (#3740)
12df3cafe Fix SSA-rewrite to remove DebugDeclare for variables without loads (#3719)
3f8501de9 Add undef for inlined void function (#3720)
4dd122392 spirv-fuzz: Add words instead of logical operands (#3728)
b79773a35 CCP should mark IR changed if it created new constants. (#3732)
a711c594b spirv-fuzz: add FuzzerPassAddCompositeInserts (#3606)
582c276d4 spirv-fuzz: Support pointer types in FuzzerPassAddParameters (#3627)
3434cb0b0 Let ADCE pass check DebugScope (#3703)
ee7f0c882 spirv-opt: Implement opt::Function::HasEarlyReturn function (#3711)
e28436f2b spirv-fuzz: Check termination instructions when donating modules (#3710)
1023dd7a0 Fix -Wrange-loop-analysis warning (#3712)
82f4bf128 spirv-fuzz: Check header dominance when adding dead block (#3694)
b8de4f57e Allow DebugTypeTemplate for Type operand (#3702)
c20995ef8 spirv-fuzz: Improve code coverage of tests (#3686)
eade36db2 spirv-fuzz: Fuzzer pass to randomly apply loop preheaders (#3668)
72ea7bec4 spirv-fuzz: Support identical predecessors in TransformationPropagateInstructionUp (#3689)
b4c4da3e7 Improve non-semantic instruction handling in the optimizer (#3693)
948577c5d Fix the bug (#3680)
df859f77d spirv-fuzz: Check integer and float width capabilities (#3670)
2641d3351 spirv-fuzz: consider additional access chain instructions (#3672)
5e5929455 spirv-fuzz: Ignore specialization constants (#3664)
1435e427d Fix the bug (#3683)
be099cde1 spirv-fuzz: Fix width in FuzzerPassAddEquationInstructions (#3685)
f0ca96d12 Preserve debug info in dead-insert-elim pass (#3652)
0d629b903 Validate more OpenCL.DebugInfo.100 instructions (#3684)
13a65b1ae Only validation locations for appropriate execution models (#3656)
fd3cabd8b spirv-fuzz: Fix in operand type assertion (#3666)
f50553867 spirv-opt: Add spvOpcodeIsAccessChain (#3682)
b7056e7e0 spirv-fuzz: FuzzerPassPropagateInstructionsUp (#3478)
8e1380996 Handle no index access chain in local access chain convert (#3678)
bdeeae78a Roll 2 dependencies (#3677)
2990a2192 Avoid using /MP4 for clang on windows. (#3662)
7b2dd11dd spirv-fuzz: TransformationReplaceAddSubMulWithCarryingExtended (#3598)
6d7f34fbf spirv-fuzz: Add TransformationMakeVectorOperationDynamic (#3597)
d29eac95a spirv-fuzz: iterate over blocks in replace linear algebra pass (#3654)
efc85ff66 spirv-fuzz: make outliner pass use additional transformations (#3604)
5fd92a7e0 OpenCL.DebugInfo.100 DebugTypeArray with variable size (#3549)
3f33a9aa5 spirv-opt: Improve the code of the Instruction class (#3610)
0419751b0 spirv-fuzz: Handle OpPhis in livesafe functions (#3642)
a10e76059 spirv-fuzz: Handle OpPhi during constant obfuscation (#3640)
28f32ca53 spirv-fuzz: Fix FuzzerPassCopyObjects (#3638)
8bc27a1cf spirv-fuzz: Remove OpFunctionCall operands in correct order (#3630)
d9c73ebd9 spirv-fuzz: Handle capabilities during module donation (#3651)
9f2223602 spirv-fuzz: Refactor boilerplate in TransformationAddParameter (#3625)
92a71657f spirv-fuzz: TransformationMoveInstructionDown (#3477)
b78f4b151 Remove DebugDeclare only for target variables in ssa-rewrite (#3511)
91cea06ab Fix typo in ASAN CI build (#3623)
2aaa8653d spirv-fuzz: Transformation to add loop preheader (#3599)
96bcc8274 spirv-fuzz: Pass to replace int operands with ints of opposite signedness (#3612)
ebaefda66 Debug info preservation in loop-unroll pass (#3548)
50300450a Validator support for non-semantic clspv reflection (#3618)
ab4fe12a4 spirv-fuzz: Fix memory bugs (#3622)
c6e6597c4 spirv-fuzz: Implement the OpOuterProduct linear algebra case (#3617)
054f034ea spirv-fuzz: Compute corollary facts from OpBitcast (#3538)
a1ea15c90 Update some language usage. (#3611)
863b8e3d3 spirv-fuzz: Relax type constraints in DataSynonym facts (#3602)
7e75fea9e spirv-fuzz: Remove non-deterministic behaviour (#3608)
f9b088fe0 Avoid use of 'sanity' and 'sanity check' in the code base (#3585)
150be20d4 spirv-fuzz: Add condition to make functions livesafe (#3587)
ce16ccf38 Rolling 4 dependencies (#3601)
1dfc6fc7e spirv-fuzz: Implement the OpTranspose linear algebra case (#3589)
b63f0e5ed Fix SyntaxWarning in Python 3.8 (#3388)
6aed7ffbc CMake: Enable building with BUILD_SHARED_LIBS=1 (#3490)
31c821393 Avoid operand type range checks (#3379)
6a3eb679b Preserve debug info in scalar replacement pass (#3461)
2796840d2 Update OpenCL capabilities validation (#3149)
b25ee93c2 build(deps): bump lodash from 4.17.15 to 4.17.19 in /tools/sva (#3596)

git-subtree-dir: third_party/SPIRV-Tools
git-subtree-split: 82b378d671836b51343b010ca9ec32db14485147

Changes:
    7845730ca Bump revision to 4, for SPIR-V 1.5.
    05836bdba Add SPV_EXT_shader_image_int64 (#170)
    be32cb6c8 Added SPV_KHR_fragment_shading_rate (#172)
    c43a43c7c  Register the Xenia emulator as a generator (#171)
    d4e76fb32 Register the Messiah SPIR-V CodeGen (#169)
    060627f0b Register the ANGLE compiler (#168)
    03a842f95 Rebuild of latest headers, which slightly moves OpTerminateInvocation
    3fdabd0da Reserve SPIR-V token range for upcoming Intel extensions. (#165)
    5538bf438 Update BUILD.bazel and BUILD.gn (#166)
    96013f32b Publish the headers for the clspv embedded reflection non-semantic extended instruction set (#164)
    76afa6397 Update the registry in spir-v.xml to modernize and split out opcodes. (#156)

Commands:
    ./third_party/update-spirvheaders.sh

Bug: b/123642959
Change-Id: I68027fa8e2770cac37af6d648fbb655b2112ee88

7845730ca Bump revision to 4, for SPIR-V 1.5.
05836bdba Add SPV_EXT_shader_image_int64 (#170)
be32cb6c8 Added SPV_KHR_fragment_shading_rate (#172)
c43a43c7c  Register the Xenia emulator as a generator (#171)
d4e76fb32 Register the Messiah SPIR-V CodeGen (#169)
060627f0b Register the ANGLE compiler (#168)
03a842f95 Rebuild of latest headers, which slightly moves OpTerminateInvocation
3fdabd0da Reserve SPIR-V token range for upcoming Intel extensions. (#165)
5538bf438 Update BUILD.bazel and BUILD.gn (#166)
96013f32b Publish the headers for the clspv embedded reflection non-semantic extended instruction set (#164)
76afa6397 Update the registry in spir-v.xml to modernize and split out opcodes. (#156)

git-subtree-dir: third_party/SPIRV-Headers
git-subtree-split: 7845730cab6ebbdeb621e7349b7dc1a59c3377be

llvm::PointerType::isValidElementType() checks that the pointee type is
not void. Use a byte pointer instead.

Bug: b/155148722
Change-Id: Ibf2fd23ce0a19a6947a6817f0b73753cfa170832
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49568Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

thread_local variables in shared libraries are initialized at load-time,
but this is not observed by MemorySanitizer if the loader itself was not
instrumented, leading to false-positive unitialized variable errors.

Bug: b/155148722
Bug: b/157525646
Change-Id: Ib6b865a44678af83d94499d3764e5134607328c4
Signed-off-by: Nicolas Capens <capn@google.com># Please enter the commit message for your changes. Lines starting
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49548Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Changes:
    3285a2dfb Workaround MSAN false-positive
    6bf128cab Assert that threads are joined before destruction
    06b72af16 Scheduler: Add a couple of asserts

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

Bug: b/140546382
Change-Id: I8df3d233e0dd6bf2e9ebfe85bbd7630c902a6678

3285a2dfb Workaround MSAN false-positive
6bf128cab Assert that threads are joined before destruction
06b72af16 Scheduler: Add a couple of asserts

git-subtree-dir: third_party/marl
git-subtree-split: 3285a2dfb966fda7639391c43ec823fa88d70267

LLVM's ORCv2 JIT API, used by the LLVMJIT.cpp file, is causing some
issues for Fuchsia and potentially some Android build. Temporarily bring
back the ORCv1 version of LLVMJIT to unblock them.

Bug: b/171897980
Bug: chromium:1140854
Change-Id: I3465bba20fe36fb9d276acfa993e3593ef10a07d
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49688Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

As LLVM Orc v2 API is in flux, changes are being made in google3. This
CL applies those changes, while also continuing to make our backend
work against LLVM 10.

Bug: b/171402030
Change-Id: If7be93bfe495e9d40a6e69415978a622ed3afb50
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49608
Presubmit-Ready: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>

Make sure session is initialized before it is used by objectLayer.
Fixes assert related to objectLayer using an invalid
std::recursive_mutex in the session instance.

Also added an explicit resource removal to fix assert in
~RTDyldObjectLinkingLayer().

Bug: b/171402030
Change-Id: I4f628a9ed4afd0126462858f8f0e3e42f90e21b3
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49508
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Bug: b/155148722
Change-Id: Ic7cbd90f696d3a59a11a9fb54fcdef6325b99197
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49168Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

This extension allows subgroup operations to use 8-bit, 16-bit, and
64-bit integers as well as 16-bit floating point types if the
implementation supports those types. So no work is required other
than enabling the extension.

Bug: b/169606946
Tests: dEQP-VK.subgroups.arithmetic.*
Change-Id: I824ad65a1c9d5f9fb98bc5769f97920de5815d0f
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49448Tested-by: Sean Risser <srisser@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>

Rather than occasionally updating VkStringify's table to match
vulkan_core.h, we can use vulkan.hpp's to_string functions to get us the
name of structure types.

vulkan.hpp is such a burdensome file that it makes more sense to only
include it in VkStringify.cpp instead of including it everywhere that
looks at a structure extension list. And since vulkan.hpp uses its own
equivalent equivalent types for every Vk* object, every place that calls
vkhpp::to_string would have to do some clunky static_cast.

I think in the long run, it's easier to maintain vulkan.hpp if we don't
use it directly in our core files.

Bug: b/170962162
Change-Id: Ib0380eaba4d05d9d5127fb0569efa6ef88d0156b
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49368
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Sean Risser <srisser@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

We were calling `new MemoryMapper` and passing it into the `llvm::SectionMemoryManager`, but the `SectionMemoryManager` does not take ownership, resulting in a leak.

Given that the `MemoryMapper` is entirely stateless, just use a pointer to a single static instance.

Bug: b/171402030
Change-Id: Ic3a118a5f4aeea6b6b60861399993e81827c9d48
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49488Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Ben Clayton <bclayton@google.com>

... by only using the AHardwareBuffer_* functions which will
use the device's gralloc implementation.

Note: this change does not fix the 5 existing deqp failures
in the suballocated.* tests that are mentioned in b/169796031.

Note: ANGLE doesn't seem to have support for
VK_FORMAT_G8_B8_R8_3PLANE_420_UNORM yet: b/171302758.

Bug: b/141698760
Bug: b/147316305
Test: launch_cvd && cts -m CtsDeqpTestCases
  dEQP-VK.api.external.memory.android_hardware_buffer.*
Change-Id: Ia551e78198c20d6c40ffa9448418f73829da1880
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49248
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Jason Macnak <natsu@google.com>
Reviewed-by: Trevor Black <vantablack@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

LLVM changes certain function return types to Expected<T> instead of T.
This change adds an Unwrap that supports both cases, allowing the code
to build in both contexts.

Bug: b/171236524
Change-Id: I461ab440797695ca1e92d4827d6e6d8965dcf6b2
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49468Tested-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Bug: b/171236524
Change-Id: Ia72dd0544ed46e047aa85246fd4e3f75c20da25f
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49450Tested-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>

See the referenced bug for the gory details. For now, this makes it so
that Subzero's codegen doesn't crash even in the simple ReactorUnitTest
cases.

Bug: b/171222930
Change-Id: Ie22a91d7e3e4f3d940b7b9120067469ac4264020
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49388
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Also fix crashes when building with `REACTOR_EMIT_DEBUG_INFO`, but
with a no-debug-info build.

Change-Id: I183473a78235b938b44bf4a017f5239c78827285
Bug: b/171236524
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/43811
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

A developer can change SWIFTSHADER_LOG_LEVEL to Debug or Verbose to make
LOG_TRAP abort when a debugger is attached. This keeps debugging easy in
most circumstances where devs don't care about unsupported struct
extensions. But it still allows devs to go right to where an unsupported
structure is being used if they so desire.

Bug: b/169868497
Change-Id: I916506ee557f8de01516fe4efb33044c2367db77
Reviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/49288Reviewed-by: Alexis Hétu <sugoi@google.com>
Kokoro-Result: kokoro <noreply+kokoro@google.com>
Tested-by: Sean Risser <srisser@google.com>