platform/upstream/llvm.git
4 years ago[DSE,MSSA] Delete instructions after printing it.
Florian Hahn [Mon, 15 Jun 2020 14:59:09 +0000 (15:59 +0100)]
[DSE,MSSA] Delete instructions after printing it.

Also enables a now-passing test case, that exposed a crash caused by the
wrong order.

4 years ago[X86][SSE] Add LowerVectorAllZero helper for checking if all bits of a vector are...
Simon Pilgrim [Mon, 15 Jun 2020 14:54:12 +0000 (15:54 +0100)]
[X86][SSE] Add LowerVectorAllZero helper for checking if all bits of a vector are zero.

Pull the lowering code out of LowerVectorAllZeroTest (and rename it MatchVectorAllZeroTest).

We should be able to reuse this in combineVectorSizedSetCCEquality as well.

Another cleanup to simplify D81547.

4 years ago[PowerPC] Do not add the relocation addend to the instruction encoding
Stefan Pintilie [Mon, 15 Jun 2020 14:40:02 +0000 (09:40 -0500)]
[PowerPC] Do not add the relocation addend to the instruction encoding

We should not be adding the relocation addend to the instruction encoding.
This patch removes that and sets those bits to zero.

Differential Revision: https://reviews.llvm.org/D81082

4 years ago[DSE,MSSA] Add additional merging test cases (NFC).
Florian Hahn [Mon, 15 Jun 2020 11:34:12 +0000 (12:34 +0100)]
[DSE,MSSA] Add additional merging test cases (NFC).

Additional tests added ahead of partial overlapping store merging.

4 years ago[clang-tidy] Add diagnostics level to YAML output
Dmitry Polukhin [Sat, 2 May 2020 09:35:41 +0000 (02:35 -0700)]
[clang-tidy] Add diagnostics level to YAML output

Summary:
Also added BuildDirectory for completness and removed unused `Fix`.

Test Plan: check-all

Reviewers: alexfh, gribozavr2

Subscribers: xazax.hun, cfe-commits

Tags: #clang-tools-extra, #clang

Differential Revision: https://reviews.llvm.org/D79285

4 years ago[NFC] Remove unnecessary require global-isel from tests
Dominik Montada [Mon, 15 Jun 2020 14:35:18 +0000 (16:35 +0200)]
[NFC] Remove unnecessary require global-isel from tests

4 years ago[NFC] Add braces to if-statement in MachineVerifier
Dominik Montada [Mon, 15 Jun 2020 14:33:56 +0000 (16:33 +0200)]
[NFC] Add braces to if-statement in MachineVerifier

4 years ago[X86][SSE] LowerVectorAllZeroTest - add support for >256-bit vectors
Simon Pilgrim [Mon, 15 Jun 2020 14:30:06 +0000 (15:30 +0100)]
[X86][SSE] LowerVectorAllZeroTest - add support for >256-bit vectors

Reduce by splitting the vector until we reach the target size for PTEST/MOVMSK_PCMPEQ. There might be some cases where AVX512 can perform this with 512-bit vectors but so far I haven't encountered any such pattern that reaches LowerVectorAllZeroTest.

Prep work for D81547

4 years agoRevert "[X86] Separate imm from relocImm handling."
Hans Wennborg [Mon, 15 Jun 2020 14:08:25 +0000 (16:08 +0200)]
Revert "[X86] Separate imm from relocImm handling."

> relocImm was a complexPattern that handled both ConstantSDNode
> and X86Wrapper. But it was only applied selectively because using
> it would cause patterns to be not importable into FastISel or
> GlobalISel. So it only got applied to flag setting instructions,
> stores, RMW arithmetic instructions, and rotates.
>
> Most of the test changes are a result of making patterns available
> to GlobalISel or FastISel. The absolute-cmp.ll change is due to
> this fixing a pattern ordering issue to make an absolute symbol
> match to an 8-bit immediate before trying a 32-bit immediate.
>
> I tried to use PatFrags to reduce the repetition, but I was getting
> errors from TableGen.

This caused "Invalid EmitNode" assertions, see the llvm-commits thread for
discussion.

4 years ago[X86][SSE] LowerVectorAllZeroTest - remove unnecessary bitcasts
Simon Pilgrim [Mon, 15 Jun 2020 14:12:32 +0000 (15:12 +0100)]
[X86][SSE] LowerVectorAllZeroTest - remove unnecessary bitcasts

matchScalarReduction should return all its source vectors with the same type, so we can safely perform the OR reduction with the original type.

So we just need to bitcast for PTEST/PCMPEQB with the final reduced vector.

4 years ago[ARM][MachineOutliner] Fix no-lr-save testcase.
Yvan Roux [Mon, 15 Jun 2020 14:03:56 +0000 (16:03 +0200)]
[ARM][MachineOutliner] Fix no-lr-save testcase.

Now that saving LR into a register is handled, some register constraints
are needed to keep machine-outliner-no-lr-save.mir meaningful.

4 years ago[strictfp] Replace dangling strictfp attrs with nobuiltin
Kevin P. Neal [Thu, 14 May 2020 18:57:19 +0000 (11:57 -0700)]
[strictfp] Replace dangling strictfp attrs with nobuiltin

In preparation for a patch that will enforce new rules for the usage of
the strictfp attribute, this patch introduces auto-upgrade behavior that
will replace the strictfp attribute on callsites with nobuiltin if the
enclosing function declaration doesn't also have the strictfp attribute.

This auto-upgrade isn't being performed on .ll files because that would
prevent us from writing a test for the forthcoming verifier behavior.

Differential Revision: https://reviews.llvm.org/D70096

4 years ago[clang] Fix or emit diagnostic for checked arithmetic builtins with
Jeff Mott [Mon, 15 Jun 2020 13:08:58 +0000 (06:08 -0700)]
[clang] Fix or emit diagnostic for checked arithmetic builtins with
_ExtInt types

- Fix computed size for _ExtInt types passed to checked arithmetic
  builtins.
- Emit diagnostic when signed _ExtInt larger than 128-bits is passed
    to __builtin_mul_overflow.
- Change Sema checks for builtins to accept placeholder types.

Differential Revision: https://reviews.llvm.org/D81420

4 years ago[mlir] Fixed GCC compile issues and linking problems using SHARED_LIBS.
Marcel Koester [Mon, 15 Jun 2020 11:51:53 +0000 (13:51 +0200)]
[mlir] Fixed GCC compile issues and linking problems using SHARED_LIBS.

Differential Revision: https://reviews.llvm.org/D81839

4 years ago[Analyzer][StreamChecker] Add check for pointer escape.
Balázs Kéri [Mon, 15 Jun 2020 12:59:54 +0000 (14:59 +0200)]
[Analyzer][StreamChecker] Add check for pointer escape.

Summary:
After an escaped FILE* stream handle it is not possible to make
reliable checks on it because any function call can have effect
on it.

Reviewers: Szelethus, baloghadamsoftware, martong, NoQ

Reviewed By: NoQ

Subscribers: NoQ, rnkovacs, xazax.hun, baloghadamsoftware, szepet, a.sidorin, mikhail.ramalho, Szelethus, donat.nagy, dkrupp, gamesh411, Charusso, martong, ASDenysPetrov, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D80699

4 years ago[ARM][MachineOutliner] Add LR RegSave mode.
Yvan Roux [Mon, 15 Jun 2020 13:22:08 +0000 (15:22 +0200)]
[ARM][MachineOutliner] Add LR RegSave mode.

Outline chunks of code which need to save and restore the link register
when a spare register can be used to it.

Differential Revision: https://reviews.llvm.org/D80127

4 years ago[AArch64] Fix BTI instruction emission.
Daniel Kiss [Mon, 15 Jun 2020 13:02:38 +0000 (15:02 +0200)]
[AArch64] Fix BTI instruction emission.

Summary:
SCTLR_EL1.BT[01] controls the PACI[AB]SP compatibility with PBYTE 11
(see [1])
This bit will be set to zero so PACI[AB]SP are equal to BTI C
instruction only.

[1] https://developer.arm.com/docs/ddi0595/b/aarch64-system-registers/sctlr_el1

Reviewers: chill, tamas.petz, pbarrio, ostannard

Reviewed By: tamas.petz, ostannard

Subscribers: kristof.beyls, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D81746

4 years ago[mlir] Remove OperandAdaptor
Jacques Pienaar [Mon, 15 Jun 2020 13:01:31 +0000 (06:01 -0700)]
[mlir] Remove OperandAdaptor

Use ::Adaptor alias instead uniformly. Makes the naming more consistent as
adaptor can refer to attributes now too.

Differential Revision: https://reviews.llvm.org/D81789

4 years agoAMDGPU/GlobalISel: Workaround some load/store type selection patterns
Matt Arsenault [Wed, 10 Jun 2020 00:05:05 +0000 (20:05 -0400)]
AMDGPU/GlobalISel: Workaround some load/store type selection patterns

The logic is written for what loads/stores should be selectable. There
are a set of cases that should be selectable, but due to missing MVTs
and/or selection patterns, will fail to select. I think eventually
load/store select patterns should ignore the type and only look at the
value size, but until that happens, bitcast these to equivalent i32
vectors.

4 years agoAMDGPU/GlobalISel: Use less artifical example to avoid abort=0
Matt Arsenault [Tue, 9 Jun 2020 20:21:07 +0000 (16:21 -0400)]
AMDGPU/GlobalISel: Use less artifical example to avoid abort=0

These were failing due to an unlegalizable G_CONCAT_VECTORS due to
registers with types that are naturally illegal.

4 years agoGlobalISel: Support lowering vector->vector G_BITCAST
Matt Arsenault [Tue, 9 Jun 2020 15:19:12 +0000 (11:19 -0400)]
GlobalISel: Support lowering vector->vector G_BITCAST

Extract subvectors and cast to the result element type before
remerging.

4 years agoFix clang/test/Modules/empty.modulemap by increasing limit.
Sander de Smalen [Mon, 15 Jun 2020 08:47:49 +0000 (09:47 +0100)]
Fix clang/test/Modules/empty.modulemap by increasing limit.

This fixes a buildbot failure on Builder llvm-clang-win-x-aarch64.

The size of the module increased to just over 30kb due to new
Decl TypeDefs being added to the module after D81459 that
adds new ACLE types.

4 years agoattempt to fix failing buildbots after 3bab88b7baa20b276faaee0aa7ca87f636c91877
Tyker [Sun, 14 Jun 2020 12:39:14 +0000 (14:39 +0200)]
attempt to fix failing buildbots after 3bab88b7baa20b276faaee0aa7ca87f636c91877

Prevent IR-gen from emitting consteval declarations

Summary: with this patch instead of emitting calls to consteval function. the IR-gen will emit a store of the already computed result.

4 years ago[DebugInfo] Report errors for truncated debug line standard opcode
James Henderson [Tue, 9 Jun 2020 14:37:55 +0000 (15:37 +0100)]
[DebugInfo] Report errors for truncated debug line standard opcode

Standard opcodes usually have ULEB128 arguments, so it is generally not
possible to recover from such errors. This patch causes the parser to
stop parsing the table in such situations.

Also don't emit the operands or add data to the table if there is an
error reading these opcodes.

Reviewed by: JDevlieghere

Differential Revision: https://reviews.llvm.org/D81470

4 years ago[MLIR][Standard] Simplify complex operand unpacking
Frederik Gossen [Mon, 15 Jun 2020 10:45:03 +0000 (10:45 +0000)]
[MLIR][Standard] Simplify complex operand unpacking

Apply post-commit suggestions (see https://reviews.llvm.org/D79479).

Differential Revision: https://reviews.llvm.org/D81731

4 years ago[MLIR][Standard] Add documentation for `std.dim` and fix test cases
Frederik Gossen [Mon, 15 Jun 2020 10:39:05 +0000 (10:39 +0000)]
[MLIR][Standard] Add documentation for `std.dim` and fix test cases

Apply post-commit suggestions (see https://reviews.llvm.org/D81551).
Add documentation, simplify, and fix test cases.

Differential Revision: https://reviews.llvm.org/D81722

4 years agoRevert "[lld-macho] No need to explicitly specify -arch in tests"
Kirill Bobyrev [Mon, 15 Jun 2020 10:27:30 +0000 (12:27 +0200)]
Revert "[lld-macho] No need to explicitly specify -arch in tests"

This reverts commit 51c5baacf36f5d5d0dbed00590a9e56a5f4a0d73 and also
337fb8c767ea93058517cb9b0be5daf21b448b71 - "[lld-macho] Set REQUIRES:
x86 on more tests".

These patches cause test crashes:

http://lab.llvm.org:8011/builders/ppc64le-lld-multistage-test/builds/10054

4 years ago[mlir] Extended BufferPlacement to support more sophisticated scenarios in which
Marcel Koester [Wed, 6 May 2020 13:06:19 +0000 (15:06 +0200)]
[mlir] Extended BufferPlacement to support more sophisticated scenarios in which
allocations cannot be moved freely and can remain in divergent control flow.

The current BufferPlacement pass does not support allocation nodes that carry
additional dependencies (like in the case of dynamic shaped types). These
allocations can often not be moved freely and in turn might remain in divergent
control-flow branches. This requires a different strategy with respect to block
arguments and aliases. This CL adds additinal functionality to support
allocation nodes in divergent control flow while avoiding memory leaks.

Differential Revision: https://reviews.llvm.org/D79850

4 years agoRevert "Prevent IR-gen from emitting consteval declarations"
Kirill Bobyrev [Mon, 15 Jun 2020 10:14:15 +0000 (12:14 +0200)]
Revert "Prevent IR-gen from emitting consteval declarations"

This reverts commit 3bab88b7baa20b276faaee0aa7ca87f636c91877.

This patch causes test failures:
http://lab.llvm.org:8011/builders/clang-cmake-armv7-quick/builds/17260

4 years ago[yaml2obj] - Introduce the "NoHeaders" key for "SectionHeaderTable"
Georgii Rymar [Thu, 11 Jun 2020 13:46:56 +0000 (16:46 +0300)]
[yaml2obj] - Introduce the "NoHeaders" key for "SectionHeaderTable"

We have an issue currently. The following YAML piece just ignores the `Excluded` key.

```
SectionHeaderTable:
  Sections: []
  Excluded:
    - Name: .foo
```

Currently the meaning is: exclude the whole table.

The code checks that the `Sections` key is empty and doesn't catch/check
invalid/duplicated/missed `Excluded` entries.

Also there is no way to exclude all sections except the first null section,
because `Sections: []` currently just excludes the whole the sections header table.

To fix it, I suggest a change of the behavior.

1) A new `NoHeaders` key is added. It provides an explicit syntax to drop the whole table.
2) The meaning of the following is changed:

```
SectionHeaderTable:
  Sections: []
  Excluded:
    - Name: .foo

```
Assuming there are 2 sections in the object (a null section and `.foo`), with this patch it
means: exclude the `.foo` section, keep the null section. The null section is an implicit
section and I think it is reasonable to make "Sections: []" to mean it is implicitly added.
It will be consistent with the global "Sections" tag that is used to describe sections.

3) `SectionHeaderTable->Sections` is now optional. No `Sections` is the same as
   `Sections: []` (I think it avoids a confusion).
4) Using of `NoHeaders` together with `Sections`/`Excluded` is not allowed.
5) It is possible to use the `Excluded` key without the `Sections` key now (in this case
   `Excluded` must contain all sections).
6) `SectionHeaderTable:` or `SectionHeaderTable: []` is not allowed.
7) When the `SectionHeaderTable` key is present, we still require all sections to be
   present in `Sections` and `Excluded` lists. No changes here, we are still strict.

Differential revision: https://reviews.llvm.org/D81655

4 years ago[X86][SSE] Add tests for and/or reduction results compared to zero
Simon Pilgrim [Mon, 15 Jun 2020 09:40:27 +0000 (10:40 +0100)]
[X86][SSE] Add tests for and/or reduction results compared to zero

These should fold to memcmp/ptest/movmsk+cmpeq patterns

4 years ago[clangd] Turn on RecoveryAST for clangd by default.
Haojian Wu [Fri, 12 Jun 2020 14:05:38 +0000 (16:05 +0200)]
[clangd] Turn on RecoveryAST for clangd by default.

Reviewers: sammccall

Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, kadircet, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81739

4 years ago[VE] Support relocation information in MC layer
Kazushi (Jam) Marukawa [Mon, 15 Jun 2020 09:21:54 +0000 (11:21 +0200)]
[VE] Support relocation information in MC layer

Summary:
Change VEAsmParser to support identification with relocation information
in assmebler.  Change VEAsmBackend to support relocation information in
MC layer.  Change VEDisassembler and VEMCCodeEmitter to support binary
generation of branch target operands.  Add REFLONG fixup and variant kind
to support new R_VE_REFLONG ELF symbol.  And, add regression test in both
MC and CodeGen to check binary genaration with relocation information.

Differential Revision: https://reviews.llvm.org/D81553

4 years ago[MachineVerifier][GlobalISel] Check that branches have a MBB operand or are declared...
Dominik Montada [Wed, 10 Jun 2020 15:15:09 +0000 (17:15 +0200)]
[MachineVerifier][GlobalISel] Check that branches have a MBB operand or are declared indirect. Add missing properties to G_BRJT, G_BRINDIRECT

Summary:
Teach MachineVerifier to check branches for MBB operands if they are not declared indirect.

Add `isBarrier`, `isIndirectBranch` to `G_BRINDIRECT` and `G_BRJT`.
Without these, `MachineInstr.isConditionalBranch()` was giving a
false-positive for those instructions.

Reviewers: aemerson, qcolombet, dsanders, arsenm

Reviewed By: dsanders

Subscribers: hiraditya, wdng, simoncook, s.egerton, arsenm, rovka, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D81587

4 years ago[Test] Add an example of unprofitable PR Phi insertion
Max Kazantsev [Mon, 15 Jun 2020 08:56:06 +0000 (15:56 +0700)]
[Test] Add an example of unprofitable PR Phi insertion

This test demonstrates weird behavior of SimplifyCFG: seems that bigger
size of block leads to worse optimization choice.

4 years agoPrevent IR-gen from emitting consteval declarations
Tyker [Sun, 14 Jun 2020 12:39:14 +0000 (14:39 +0200)]
Prevent IR-gen from emitting consteval declarations

Summary: with this patch instead of emitting calls to consteval function. the IR-gen will emit a store of the already computed result.

Reviewers: rsmith

Reviewed By: rsmith

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D76420

4 years agoNFC: Make sure function arguments have the same name in declaration and definition
Kirill Bobyrev [Mon, 15 Jun 2020 08:45:08 +0000 (10:45 +0200)]
NFC: Make sure function arguments have the same name in declaration and definition

This code generates Clang-Tidy warnings otherwise.

4 years ago[mlir][shape] Lower Shape `ConstSizeOp` to Standard `ConstantOp`.
Alexander Belyaev [Fri, 12 Jun 2020 13:50:03 +0000 (15:50 +0200)]
[mlir][shape] Lower Shape `ConstSizeOp` to Standard `ConstantOp`.

Differential Revision: https://reviews.llvm.org/D81735

4 years ago[CostModel] getCFInstrCost in getUserCost.
Sam Parker [Mon, 15 Jun 2020 07:58:25 +0000 (08:58 +0100)]
[CostModel] getCFInstrCost in getUserCost.

Have BasicTTI call the base implementation so that both agree on the
default behaviour, which the default being a cost of '1'. This has
required an X86 specific implementation as it seems to be very
reliant on those instructions being free. Changes are also made to
AMDGPU so that their implementations distinguish between cost kinds,
so that the unrolling isn't affected. PowerPC also has its own
implementation to prevent changes to the reg-usage vectorizer test.

The cost model test changes now reflect that ret instructions are not
generally free.

Differential Revision: https://reviews.llvm.org/D79164

4 years ago[lldb] Handle all Clang::Type::Builtin enums
Kirill Bobyrev [Mon, 15 Jun 2020 08:18:59 +0000 (10:18 +0200)]
[lldb] Handle all Clang::Type::Builtin enums

Cleanup after https://reviews.llvm.org/D81459

4 years ago[CMake][runtimes] Skip adding 2nd set of the same variables for a generic target
Kristina Bessonova [Wed, 10 Jun 2020 14:18:30 +0000 (16:18 +0200)]
[CMake][runtimes] Skip adding 2nd set of the same variables for a generic target

No need to parse and add the same variables twice if runtimes is being
built for a generic target (i.e. w/o multilib).

Reviewed By: phosek

Differential Revision: https://reviews.llvm.org/D81574

4 years ago[NFCI][CostModel] Unify FNeg cost
Sam Parker [Thu, 11 Jun 2020 08:59:12 +0000 (09:59 +0100)]
[NFCI][CostModel] Unify FNeg cost

Enable TTIImpl::getUserCost to handle FNeg so that
getInstructionThroughput can call that instead. This means we can
remove the code in the AMDGPU backend too.

Differential Revision: https://reviews.llvm.org/D81635

4 years ago[IR] Prefer hasFnAttribute() where possible (NFC)
Nikita Popov [Sun, 14 Jun 2020 21:52:39 +0000 (23:52 +0200)]
[IR] Prefer hasFnAttribute() where possible (NFC)

When checking for an enum function attribute, use hasFnAttribute()
rather than hasAttribute() at FunctionIndex, because it is
significantly faster (and more concise to boot).

4 years ago[CostModel] Unify ExtractElement cost.
Sam Parker [Mon, 15 Jun 2020 07:27:14 +0000 (08:27 +0100)]
[CostModel] Unify ExtractElement cost.

Move the cost modelling, with the reduction pattern matching, from
getInstructionThroughput into generic TTIImpl::getUserCost. The
modelling in the AMDGPU backend can now be removed.

Differential Revision: https://reviews.llvm.org/D81643

4 years ago[NFC] Bail early simplifying unconditional branches
Max Kazantsev [Mon, 15 Jun 2020 06:58:54 +0000 (13:58 +0700)]
[NFC] Bail early simplifying unconditional branches

4 years ago[msan] Fix comment of __msan::Origin::isHeapOrigin
Fangrui Song [Mon, 15 Jun 2020 06:00:04 +0000 (23:00 -0700)]
[msan] Fix comment of __msan::Origin::isHeapOrigin

4 years agoRevert "Return "[InstCombine] Simplify compare of Phi with constant inputs against...
Sam Parker [Mon, 15 Jun 2020 06:46:28 +0000 (07:46 +0100)]
Revert "Return "[InstCombine] Simplify compare of Phi with constant inputs against a constant""

This reverts commit 23291b9863c8af7ad348c4a7d85d8d784df88eb1.

This caused performance regressions.

4 years ago[SVE] Ensure proper mangling of ACLE tuple types
Sander de Smalen [Mon, 15 Jun 2020 06:20:55 +0000 (07:20 +0100)]
[SVE] Ensure proper mangling of ACLE tuple types

The AAPCS specifies that the tuple types such as `svint32x2_t`
should use their `arm_sve.h` names when mangled instead of their
builtin names.

This patch also renames the internal types for the tuples to
be prefixed with `__clang_`, so they are not misinterpreted as
specified internal types like the non-tuple types which *are* defined
in the AAPCS. Using a builtin type for the tuples is a purely
a choice of the Clang implementation.

Reviewers: rsandifo-arm, c-rhodes, efriedma, rengolin

Reviewed By: efriedma

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81721

4 years ago[SveEmitter] Add SVE tuple types and builtins for svundef.
Sander de Smalen [Mon, 15 Jun 2020 06:20:43 +0000 (07:20 +0100)]
[SveEmitter] Add SVE tuple types and builtins for svundef.

This patch adds new SVE types to Clang that describe tuples of SVE
vectors. For example `svint32x2_t` which maps to the twice-as-wide
vector `<vscale x 8 x i32>`. Similarly, `svint32x3_t` will map to
`<vscale x 12 x i32>`.

It also adds builtins to return an `undef` vector for a given
SVE type.

Reviewers: c-rhodes, david-arm, ctetreau, efriedma, rengolin

Reviewed By: c-rhodes

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81459

4 years ago[SafeStack,NFC] Make StackColoring read-only
Vitaly Buka [Mon, 15 Jun 2020 06:03:13 +0000 (23:03 -0700)]
[SafeStack,NFC] Make StackColoring read-only

Move core which removes markers out of StackColoring.

4 years ago[SafeStack,NFC] Remove unneded branch
Vitaly Buka [Mon, 15 Jun 2020 04:28:34 +0000 (21:28 -0700)]
[SafeStack,NFC] Remove unneded branch

4 years ago[SafeStack,NFC] Fix naming style
Vitaly Buka [Mon, 15 Jun 2020 02:04:47 +0000 (19:04 -0700)]
[SafeStack,NFC] Fix naming style

4 years ago[SafeStack,NFC] Cleanup LiveRange interface
Vitaly Buka [Mon, 15 Jun 2020 01:59:23 +0000 (18:59 -0700)]
[SafeStack,NFC] Cleanup LiveRange interface

4 years ago[SafeStack,NFC] "const" cleanup
Vitaly Buka [Sun, 14 Jun 2020 23:31:09 +0000 (16:31 -0700)]
[SafeStack,NFC] "const" cleanup

4 years ago[SafeStack,NFC] Add BlockLifetimeInfo constructor
Vitaly Buka [Sun, 14 Jun 2020 23:06:07 +0000 (16:06 -0700)]
[SafeStack,NFC] Add BlockLifetimeInfo constructor

4 years ago[SafeStack,NFC] Use IntrinsicInst instead of Instruction
Vitaly Buka [Sun, 14 Jun 2020 22:58:10 +0000 (15:58 -0700)]
[SafeStack,NFC] Use IntrinsicInst instead of Instruction

4 years ago[SafeStack,NFC] Move ClColoring into SafeStack.cpp
Vitaly Buka [Sun, 14 Jun 2020 22:32:07 +0000 (15:32 -0700)]
[SafeStack,NFC] Move ClColoring into SafeStack.cpp

This allows to reuse the code in other components.

4 years ago[SafeStack,NFC] Move unconditional code into constructor
Vitaly Buka [Sun, 14 Jun 2020 21:35:34 +0000 (14:35 -0700)]
[SafeStack,NFC] Move unconditional code into constructor

Prepare to move ClColoring from SafeStackCode to SafeStackLayout.
This will allow to reuse the code in other components.

4 years ago[Test] Update test with check script, add two more motivating cases
Max Kazantsev [Mon, 15 Jun 2020 05:36:33 +0000 (12:36 +0700)]
[Test] Update test with check script, add two more motivating cases

4 years ago[PowerPC] fma chain break to expose more ILP
Chen Zheng [Mon, 15 Jun 2020 01:34:52 +0000 (21:34 -0400)]
[PowerPC] fma chain break to expose more ILP

This patch tries to reassociate two patterns related to FMA to expose
more ILP on PowerPC.

// Pattern 1:
//   A =  FADD X,  Y          (Leaf)
//   B =  FMA  A,  M21,  M22  (Prev)
//   C =  FMA  B,  M31,  M32  (Root)
// -->
//   A =  FMA  X,  M21,  M22
//   B =  FMA  Y,  M31,  M32
//   C =  FADD A,  B

// Pattern 2:
//   A =  FMA  X,  M11,  M12  (Leaf)
//   B =  FMA  A,  M21,  M22  (Prev)
//   C =  FMA  B,  M31,  M32  (Root)
// -->
//   A =  FMUL M11,  M12
//   B =  FMA  X,  M21,  M22
//   D =  FMA  A,  M31,  M32
//   C =  FADD B,  D

Reviewed By: jsji

Differential Revision: https://reviews.llvm.org/D80175

4 years ago[NewPM] Avoid redundant CGSCC run for updated SCC
Wenlei He [Tue, 26 May 2020 20:53:35 +0000 (13:53 -0700)]
[NewPM] Avoid redundant CGSCC run for updated SCC

Summary:
When an SCC got split due to inlining, we have two mechanisms for reprocessing the updated SCC, first is UR.UpdatedC
that repeatedly rerun the new, current SCC; second is a worklist for all newly split SCCs. We can avoid rerun of
the same SCC when the SCC is set to be processed by both mechanisms *back to back*. In pathological cases, such redundant
rerun could cause exponential size growth due to inlining along cycles, even when there's no SCC mutation and hence
convergence is not a problem.

Note that it's ok to have SCC updated and rerun immediately, and also in the work list if we have actually moved an SCC
to be topologically "below" the current one due to merging. In that case, we will need to revisit the current SCC after
those moved SCCs. For that reason, the redundant avoidance here only targets back to back rerun of the same SCC - the
case described by the now removed FIXME comment.

Reviewers: chandlerc, wmi

Subscribers: llvm-commits, hoy

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80589

4 years ago[PowerPC] Add some InstAlias for mtspr/mfspr instructions
Kang Zhang [Mon, 15 Jun 2020 02:43:13 +0000 (02:43 +0000)]
[PowerPC] Add some InstAlias for mtspr/mfspr instructions

Summary:

We have defined MTSPR/MFSPR and MTSPR8/MFSPR8, but we only defined
mtspr/mfspr InstAlias for some MTSPR/MFSPR.
This patch is to add the InstAlias definitions for MTSPR8/MFSPR8,
and add the some new mtspr/mfspr InstAlias we may use.

Reviewed By: steven.zhang

Differential Revision: https://reviews.llvm.org/D77531

4 years ago[lld-macho] Set REQUIRES: x86 on more tests
Jez Ng [Mon, 15 Jun 2020 01:59:28 +0000 (18:59 -0700)]
[lld-macho] Set REQUIRES: x86 on more tests

Summary: Fixes the build break caused by D81802.

4 years ago[PowerPC] fold a bug for rlwinm folding when with full mask.
Chen Zheng [Mon, 15 Jun 2020 01:22:50 +0000 (21:22 -0400)]
[PowerPC] fold a bug for rlwinm folding when with full mask.

Reviewed By: steven.zhang

Differential Revision: https://reviews.llvm.org/D81006

4 years ago[lld-macho] Properly handle & validate relocation r_length
Jez Ng [Sun, 14 Jun 2020 02:52:20 +0000 (19:52 -0700)]
[lld-macho] Properly handle & validate relocation r_length

Summary:
We should be reading / writing our addends / relocated addresses based on
r_length, and not just based on the type of the relocation. But since only
some r_length values are valid for a given reloc type, I've also added some
validation.

ld64 has code to allow for r_length = 0 in X86_64_RELOC_BRANCH relocs, but I'm
not sure how to create such a relocation...

Reviewed By: smeenai

Differential Revision: https://reviews.llvm.org/D80854

4 years ago[lld-macho] No need to explicitly specify -arch in tests
Jez Ng [Sun, 14 Jun 2020 04:45:29 +0000 (21:45 -0700)]
[lld-macho] No need to explicitly specify -arch in tests

Summary: After {D81326} landed, some tests started failing if they did
not have `-arch` specified. I think one of the reasons happened was due
to the fact that we were taking a reference to a temporary value that
was freed too early. Fixing that got the error to go away on my local
Linux machine.

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D81802

4 years ago[X86][SSE] Fold BITOP(MOVMSK(X),MOVMSK(Y)) -> MOVMSK(BITOP(X,Y))
Simon Pilgrim [Sun, 14 Jun 2020 16:53:33 +0000 (17:53 +0100)]
[X86][SSE] Fold BITOP(MOVMSK(X),MOVMSK(Y)) -> MOVMSK(BITOP(X,Y))

Reduce XMM->GPR traffic by performing bitops on the vectors, and using a single MOVMSK call.

This requires us to use vectors of the same size and element width, but we can mix fp/int type equivalents with suitable bitcasting.

4 years ago[IR] Remove unused IndexAttrPair typedef (NFC)
Nikita Popov [Sun, 14 Jun 2020 20:27:17 +0000 (22:27 +0200)]
[IR] Remove unused IndexAttrPair typedef (NFC)

This was part of an older attributes implementation.

4 years ago[IR] Support efficient AssertingVH/PoisoningVH lookup
Nikita Popov [Sat, 13 Jun 2020 19:10:22 +0000 (21:10 +0200)]
[IR] Support efficient AssertingVH/PoisoningVH lookup

Currently, there doesn't seem to be any way to look up a Value*
in a map/set indexed by AssertingVH/PoisoningVH, without creating
a value handle -- which is fairly expensive, because it involves
adding the value handle to the use list and immediately removing
it again. Using find_as(Value *) does not work (and is in fact
worse than just using find(Value *)), because it will end up
creating multiple value handles during the lookup itself.

For AssertingVH, address this by simply using DenseMapInfo<T *>
instead of manually implementing something. The AssertingVH<T>
will now get coerced to T*, rather than the other way around.

For PoisoningVH, add extra overloads of getHashValue() and
isEqual() that accept a T* argument.

This allows using find_as(Value *) to perform efficient lookups
in assertion-enabled builds.

Differential Revision: https://reviews.llvm.org/D81793

4 years ago[LAA] Do not set CanDoRT to false for AS that do not need RT checks.
Florian Hahn [Sat, 13 Jun 2020 13:24:50 +0000 (14:24 +0100)]
[LAA] Do not set CanDoRT to false for AS that do not need RT checks.

Alternative approach to D80570.

canCheckPtrAtRT already contains checks the figure out for which alias
sets runtime checks are needed. But it currently sets CanDoRT to false
for alias sets for which we cannot do RT checks but also do not need
any.

If we know that we do not need RT checks based on the number of
reads/writes in the alias set, we can skip processing the AS.

This patch also adds an assertion to ensure that DepCands does not
contain more than one write from the alias set.

Reviewers: Ayal, anemet, hfinkel, dmgreen

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D80622

4 years ago[LoopUnroll] Allow loops with multiple exiting blocks where loop latch
Whitney Tsang [Sun, 14 Jun 2020 18:42:24 +0000 (18:42 +0000)]
[LoopUnroll] Allow loops with multiple exiting blocks where loop latch
is not necessary one of them.

Summary: Currently LoopUnrollPass already allow loops with multiple
exiting blocks, but it is only allowed when the loop latch is one of the
exiting blocks.
When the loop latch is not an exiting block, then only single exiting
block is supported.
When possible, the single loop latch or the single exiting block
terminator is optimized to an unconditional branch in the unrolled loop.

This patch allows loops with multiple exiting blocks even if the loop
latch is not one of them. However, the optimization of exiting block
terminator to unconditional branch is not done when there exists more
than one exiting block.
Reviewer: dmgreen, Meinersbur, etiotto, fhahn, efriedma, bmahjour
Reviewed By: efriedma
Subscribers: hiraditya, zzheng, llvm-commits
Tag: LLVM
Differential Revision: https://reviews.llvm.org/D81053

4 years agoAMDGPU: Add some baseline immediate encoding test changes
Matt Arsenault [Sun, 14 Jun 2020 16:16:38 +0000 (12:16 -0400)]
AMDGPU: Add some baseline immediate encoding test changes

Add some encoding checks and add a few new cases.

4 years agoAMDGPU: Do not bundle inline asm
Matt Arsenault [Thu, 11 Jun 2020 20:09:41 +0000 (16:09 -0400)]
AMDGPU: Do not bundle inline asm

Fixes bug 46285

4 years agoGlobalISel: Add some basic getters to GISelKnownBits
Matt Arsenault [Wed, 10 Jun 2020 13:40:51 +0000 (09:40 -0400)]
GlobalISel: Add some basic getters to GISelKnownBits

4 years agoAMDGPU/GlobalISel: Select general case for G_PTRMASK
Matt Arsenault [Sat, 23 May 2020 21:10:58 +0000 (17:10 -0400)]
AMDGPU/GlobalISel: Select general case for G_PTRMASK

4 years agoAMDGPU: Fix spill/restore of 192-bit registers
Matt Arsenault [Fri, 12 Jun 2020 01:40:21 +0000 (21:40 -0400)]
AMDGPU: Fix spill/restore of 192-bit registers

I tried to use an IR inline asm test, but that doesn't work since the
inline asm handling asserts without an MVT to use.

4 years ago[X86][SSE] Add tests for missing BITOP(MOVMSK(X),MOVMSK(Y)) -> MOVMSK(BITOP(X,Y)...
Simon Pilgrim [Sun, 14 Jun 2020 15:24:18 +0000 (16:24 +0100)]
[X86][SSE] Add tests for missing BITOP(MOVMSK(X),MOVMSK(Y)) -> MOVMSK(BITOP(X,Y)) fold

This would help reduce XMM->GPR traffic for some reduction cases.

4 years ago[PowerPC] Support constrained rounding operations
Qiu Chaofan [Sun, 14 Jun 2020 15:43:31 +0000 (23:43 +0800)]
[PowerPC] Support constrained rounding operations

This patch adds handling of constrained FP intrinsics about round,
truncate and extend for PowerPC target, with necessary tests.

Reviewed By: steven.zhang

Differential Revision: https://reviews.llvm.org/D64193

4 years ago[PowerPC] Exploit vnmsubfp instruction
Qiu Chaofan [Sun, 14 Jun 2020 15:19:17 +0000 (23:19 +0800)]
[PowerPC] Exploit vnmsubfp instruction

On PowerPC, we have vnmsubfp Altivec instruction for fnmsub operation on
v4f32 type. Default pattern for this instruction never works since we
don't have legal fneg for v4f32 when VSX disabled.

Reviewed By: steven.zhang

Differential Revision: https://reviews.llvm.org/D80617

4 years ago[DAGCombiner] Require ninf for division estimation
Qiu Chaofan [Sun, 14 Jun 2020 14:58:22 +0000 (22:58 +0800)]
[DAGCombiner] Require ninf for division estimation

Current implementation of division estimation isn't correct for some
cases like 1.0/0.0 (result is nan, not expected inf).

And this change exposes a potential infinite loop: we use
isConstOrConstSplatFP in combineRepeatedFPDivisors to look up if the
divisor is some constant. But it doesn't work after legalized on some
platforms. This patch restricts the method to act before LegalDAG.

Reviewed By: spatel

Differential Revision: https://reviews.llvm.org/D80542

4 years ago[PassManager] restore early-cse to vector cleanup
Sanjay Patel [Sun, 14 Jun 2020 13:50:10 +0000 (09:50 -0400)]
[PassManager] restore early-cse to vector cleanup

As noted in D80236 - the early-cse pass was included here before:
D75145 / rG71a316883d50
But it got moved outside of the "extra" option there, then it
got dropped while adjusting -vector-combine:
rG6438ea45e053
rG57bb4787d72f

So this is restoring the behavior and adding a test to prevent
accidental changes again. I don't see an equivalent option for
the new pass manager.

4 years ago[OpenMP][Tool] Extend reuse of OMPT testing
Joachim Protze [Wed, 11 Mar 2020 16:02:19 +0000 (17:02 +0100)]
[OpenMP][Tool] Extend reuse of OMPT testing

This patch allows to specify a prefix (default:empty) to be included into print-out
written by callback.h.
Also adding a cmake target to find the header file from other tests.

Reviewed by: jdoerfert

Differential Revision: https://reviews.llvm.org/D76008

4 years ago[OpenMP] support alloc of serialized tasks
Joachim Protze [Wed, 10 Jun 2020 14:44:52 +0000 (16:44 +0200)]
[OpenMP] support alloc of serialized tasks

Reviewed by: AndreyChurbanov

Differential Revision: https://reviews.llvm.org/D81497

4 years ago[LVI] Fix class indentation (NFC)
Nikita Popov [Sun, 14 Jun 2020 13:42:27 +0000 (15:42 +0200)]
[LVI] Fix class indentation (NFC)

This class uses a mix of different indentation levels, normalize it.

4 years ago[LVI] Cache lookup of experimental.guard intrinsic (NFC)
Nikita Popov [Sat, 13 Jun 2020 13:36:55 +0000 (15:36 +0200)]
[LVI] Cache lookup of experimental.guard intrinsic (NFC)

When LVI is performing assume intersections, it also checks for
llvm.experimental.guard intrinsics. To avoid unnecessary block
scans, it first checks whether this intrinsic is declared in the
module at all. I've noticed that we end up spending quite a lot
of time looking up that function again and again...

Avoid this by only looking it up once when LazyValueInfo is
constructed. This of course assumes that we don't introduce new
guard intrinsics (which is the case for all existing uses of LVI --
and even if it weren't, it would not introduce miscompiles, just
potentially lose optimization power.)

Differential Revision: https://reviews.llvm.org/D81796

4 years ago[ARM] Additional cast cost tests.
David Green [Sun, 14 Jun 2020 11:05:20 +0000 (12:05 +0100)]
[ARM] Additional cast cost tests.

This adds additional cast cpst tests useful for MVE, notably around half
types.

4 years ago[InstCombine] reassociate FP diff of sums into sum of diffs
Sanjay Patel [Sun, 14 Jun 2020 13:09:03 +0000 (09:09 -0400)]
[InstCombine] reassociate FP diff of sums into sum of diffs

(a[0] + a[1] + a[2] + a[3]) - (b[0] + b[1] + b[2] +b[3]) -->
(a[0] - b[0]) + (a[1] - b[1]) + (a[2] - b[2]) + (a[3] - b[3])

This should be the last step in solving PR43953:
https://bugs.llvm.org/show_bug.cgi?id=43953

We started emitting reduction intrinsics with:
D80867/ rGe50059f6b6b3
So it's a relatively easy pattern match now to re-order those ops.
Also, I have not seen any complaints for the switch to intrinsics
yet, so I'll propose to remove the "experimental" tag from the
intrinsics soon.

Differential Revision: https://reviews.llvm.org/D81491

4 years ago[InstCombine] allow undef elements when comparing vector constants for min/max bailout
Sanjay Patel [Sun, 14 Jun 2020 13:01:33 +0000 (09:01 -0400)]
[InstCombine] allow undef elements when comparing vector constants for min/max bailout

This is a hacky, but low-risk fix to avoid the infinite loop in PR46271:
https://bugs.llvm.org/show_bug.cgi?id=46271

As discussed there, the problem is that FoldOpIntoSelect() can get into a conflict
with a transform that wants to pull a 'not' op through min/max via
SimplifyDemandedVectorElts(). We need to relax our matching of min/max to include
undefined elements in vector constants to avoid that. Alternatively, we could
improve or cripple the demanded elements analysis, but that could create even
more problems.

The likely better, safer alternative will be to create min/max intrinsics, so
we can remove all of the hacks related to min/max matching in instcombine.

Differential Revision: https://reviews.llvm.org/D81698

4 years ago[X86][SSE] LowerVectorAllZeroTest - add support for pre-SSE41 targets
Simon Pilgrim [Sun, 14 Jun 2020 12:10:33 +0000 (13:10 +0100)]
[X86][SSE] LowerVectorAllZeroTest - add support for pre-SSE41 targets

Even without PTEST, we can still efficiently perform an OR reduction as PMOVMSKB(PCMPEQB(X,0)) == 0, avoiding xmm->gpr extractions.

4 years ago[MLIR][NFC] Update vim syntax file
Uday Bondhugula [Sun, 14 Jun 2020 12:29:06 +0000 (17:59 +0530)]
[MLIR][NFC] Update vim syntax file

Add a few more commonly used ops and missing keywords.

4 years ago[clangd] Fix windows builds failing on check-clangd
njames93 [Sun, 14 Jun 2020 12:29:17 +0000 (13:29 +0100)]
[clangd] Fix windows builds failing on check-clangd

4 years ago[X86][SSE] Add non-SSE41 target PTEST tests
Simon Pilgrim [Sun, 14 Jun 2020 11:20:48 +0000 (12:20 +0100)]
[X86][SSE] Add non-SSE41 target PTEST tests

Ensure codegen is still reasonable - ideally we'd make use of MOVMSK for this.

4 years ago[NFC] mv llvm/test/tools/obj2yaml/macho-DWARF-debug-ranges.yaml llvm/test/ObjectYAML...
Xing GUO [Sun, 14 Jun 2020 08:39:15 +0000 (16:39 +0800)]
[NFC] mv llvm/test/tools/obj2yaml/macho-DWARF-debug-ranges.yaml llvm/test/ObjectYAML/MachO/DWARF-debug_ranges.yaml

4 years ago[ObjectYAML][DWARF] Let the target address size be inferred from FileHeader.
Xing GUO [Sun, 14 Jun 2020 04:00:46 +0000 (12:00 +0800)]
[ObjectYAML][DWARF] Let the target address size be inferred from FileHeader.

This patch adds a new field `bool Is64bit` in `DWARFYAML::Data` to indicate the address size of target. It's helpful for inferring the `AddrSize` in some DWARF sections.

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D81709

4 years ago[IteratedDominanceFrontier] Decrease number of SmallPtrSet::insert and delete unneede...
Fangrui Song [Sun, 14 Jun 2020 02:48:50 +0000 (19:48 -0700)]
[IteratedDominanceFrontier] Decrease number of SmallPtrSet::insert and delete unneeded SmallVector::clear

Also, fix the argument name to be consistent with the declaration.

4 years ago[X86] Add mayLoad flag to FARCALL*m/FARJMP memory instrutions. Add 'm' to the end...
Craig Topper [Sat, 13 Jun 2020 22:40:51 +0000 (15:40 -0700)]
[X86] Add mayLoad flag to FARCALL*m/FARJMP memory instrutions. Add 'm' to the end of FARJMP64/FARCALL64 instruction names.

We never codegen them so this doesn't matter in practice. But
sometimes someone comes along and tries to use these flags
for something else. LIke the Load Value Inject inline assembly
handling.

4 years ago[X86] Automatically harden inline assembly RET instructions against Load Value Inject...
Craig Topper [Sat, 13 Jun 2020 22:16:00 +0000 (15:16 -0700)]
[X86] Automatically harden inline assembly RET instructions against Load Value Injection (LVI)

Previously, the X86AsmParser would issue a warning whenever a ret instruction is encountered. This patch changes the behavior to automatically transform each ret instruction in an inline assembly stream into:

shlq $0, (%rsp)
lfence
ret

which is secure, according to https://software.intel.com/security-software-guidance/insights/deep-dive-load-value-injection#specialinstructions.

Patch by Scott Constable with some minor changes by Craig Topper.

4 years ago[X86] Teach combineBitcastvxi1 to prefer movmsk on avx512 in more cases
Craig Topper [Sat, 13 Jun 2020 21:50:06 +0000 (14:50 -0700)]
[X86] Teach combineBitcastvxi1 to prefer movmsk on avx512 in more cases

If the input to the bitcast is a sign bit test, it makes sense to
directly use vpmovmskb or vmovmskps/pd. This removes the need to
copy the sign bits to a k-register and then to a GPR.

Fixes PR46200.

Differential Revision: https://reviews.llvm.org/D81327

4 years ago[X86] Move -x86-use-vzeroupper command line flag into runOnMachineFunction for the...
Craig Topper [Sat, 13 Jun 2020 21:40:56 +0000 (14:40 -0700)]
[X86] Move -x86-use-vzeroupper command line flag into runOnMachineFunction for the pass itself rather than the pass pipeline construction

This pass has no dependencies on other passes so conditionally
including it in the pipeline doens't do much. Just move it the
pass itself to keep it isolated.

4 years ago[NFCI][AggressiveInstCombiner] Add `STATISTIC()`s for transforms
Roman Lebedev [Sat, 13 Jun 2020 12:24:53 +0000 (15:24 +0300)]
[NFCI][AggressiveInstCombiner] Add `STATISTIC()`s for transforms