OSDN Git Service

android-x86/external-llvm.git
5 years ago[IR] Add WithOverflowInst class
Nikita Popov [Tue, 16 Apr 2019 18:55:16 +0000 (18:55 +0000)]
[IR] Add WithOverflowInst class

This adds a WithOverflowInst class with a few helper methods to get
the underlying binop, signedness and nowrap type and makes use of it
where sensible. There will be two more uses in D60650/D60656.

The refactorings are all NFC, though I left some TODOs where things
could be improved. In particular we have two places where add/sub are
handled but mul isn't.

Differential Revision: https://reviews.llvm.org/D60668

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358512 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Tests] Add branch_weights to latches so that test is not effected by future profitab...
Philip Reames [Tue, 16 Apr 2019 16:32:59 +0000 (16:32 +0000)]
[Tests] Add branch_weights to latches so that test is not effected by future profitability patch to LoopPredication

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358506 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Hexagon] Remove indeterministic traversal order
Krzysztof Parzyszek [Tue, 16 Apr 2019 16:05:07 +0000 (16:05 +0000)]
[Hexagon] Remove indeterministic traversal order

Patch by Sergei Larin.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358505 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Test tabs in disassemble-align.s with a more visible character
Fangrui Song [Tue, 16 Apr 2019 15:58:42 +0000 (15:58 +0000)]
[llvm-objdump] Test tabs in disassemble-align.s with a more visible character

Summary: Apply rupprecht's suggestion in D60376

Reviewers: rupprecht

Reviewed By: rupprecht

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60777

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358504 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DAGCombiner] Add missing flag to addressing mode check
Luis Marques [Tue, 16 Apr 2019 15:09:18 +0000 (15:09 +0000)]
[DAGCombiner] Add missing flag to addressing mode check

The checks in `canFoldInAddressingMode` tested for addressing modes that have a
base register but didn't set the `HasBaseReg` flag to true (it's false by
default). This patch fixes that. Although the omission of the flag was
technically incorrect it had no known observable impact, so no tests were
changed by this patch.

Differential Revision:  https://reviews.llvm.org/D60314

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358502 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[OCaml] Update api to account for FNeg and CallBr instructions
whitequark [Tue, 16 Apr 2019 15:00:19 +0000 (15:00 +0000)]
[OCaml] Update api to account for FNeg and CallBr instructions

Summary:
This diff adds minimal support for the recent FNeg and CallBr
instructions to the OCaml bindings.

Reviewers: whitequark

Reviewed By: whitequark

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60680

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358501 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[RISCV] Custom lower SHL_PARTS, SRA_PARTS, SRL_PARTS
Luis Marques [Tue, 16 Apr 2019 14:38:32 +0000 (14:38 +0000)]
[RISCV] Custom lower SHL_PARTS, SRA_PARTS, SRL_PARTS

When not optimizing for minimum size (-Oz) we custom lower wide shifts
(SHL_PARTS, SRA_PARTS, SRL_PARTS) instead of expanding to a libcall.

Differential Revision: https://reviews.llvm.org/D59477

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358498 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[SystemZ] Add missing intrinsics to intrinsics-immarg.ll
Ulrich Weigand [Tue, 16 Apr 2019 14:35:18 +0000 (14:35 +0000)]
[SystemZ] Add missing intrinsics to intrinsics-immarg.ll

As of r356091, support for the ImmArg intrinsics was added,
including a SystemZ test case.  However, that test case doesn't
actually verify all SystemZ intrinsics with immediate arguments,
only a subset.  The rest of them actually works correctly, there's
just no test for them.  This patch add all missing intrinsics.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358495 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm][Support] Provide interface to set thread priorities
Kadir Cetinkaya [Tue, 16 Apr 2019 14:32:43 +0000 (14:32 +0000)]
[llvm][Support] Provide interface to set thread priorities

Summary:
We have a multi-platform thread priority setting function(last piece
landed with D58683), I wanted to make this available to all llvm community,
there seem to be other users of such functionality with portability fixmes:
lib/Support/CrashRecoveryContext.cpp
tools/clang/tools/libclang/CIndex.cpp

Reviewers: gribozavr, ioeric

Subscribers: krytarowski, jfb, kristina, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D59130

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358494 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Consistently use "return nullptr" in functions returning pointers
Nico Weber [Tue, 16 Apr 2019 14:24:42 +0000 (14:24 +0000)]
llvm-undname: Consistently use "return nullptr" in functions returning pointers

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358492 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Fix nullptr deref on invalid structor names in template args
Nico Weber [Tue, 16 Apr 2019 14:10:34 +0000 (14:10 +0000)]
llvm-undname: Fix nullptr deref on invalid structor names in template args

Similar to r358421: A StructorIndentifierNode has a Class field which
is read when printing it, but if the StructorIndentifierNode appears in
a template argument then demangleFullyQualifiedSymbolName() which sets
Class isn't called. Since StructorIndentifierNodes are always leaf
names, we can just reject them as well.

Found by oss-fuzz.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358491 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Tweak arena allocator
Nico Weber [Tue, 16 Apr 2019 13:52:30 +0000 (13:52 +0000)]
llvm-undname: Tweak arena allocator

- Make `allocUnalignedBuffer` look more like `allocArray` and `alloc`.
  No behavior change.
- Change `Head->Used < Head->Capacity` to `Head->Used <= Head->Capacity`
  in `allocArray` and `alloc`. No intended behavior change, might be a
  minuscule memory usage improvement. Noticed this since it was the logic
  used in `allocUnalignedBuffer`.
- Don't let `allocArray` alloc too small buffers for names that have
  more than 512 levels of nesting (in 64-bit builds). Fixes a heap
  buffer overflow found by oss-fuzz.

Differential Revision: https://reviews.llvm.org/D60774

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358489 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: add a missing CHECK: to a passing test
Nico Weber [Tue, 16 Apr 2019 13:30:50 +0000 (13:30 +0000)]
llvm-undname: add a missing CHECK: to a passing test

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358488 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoFix llvm-undname tests after r358485
Nico Weber [Tue, 16 Apr 2019 13:18:51 +0000 (13:18 +0000)]
Fix llvm-undname tests after r358485

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358487 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agogn build: Merge r358422
Nico Weber [Tue, 16 Apr 2019 12:54:43 +0000 (12:54 +0000)]
gn build: Merge r358422

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358486 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Add a -raw-file flag to pass a raw buffer to microsoftDemangle
Nico Weber [Tue, 16 Apr 2019 12:51:40 +0000 (12:51 +0000)]
llvm-undname: Add a -raw-file flag to pass a raw buffer to microsoftDemangle

The default handling splits input into lines. Since
llvm-microsoft-demangle-fuzzer doesn't do this, oss-fuzz produces inputs
that only trigger crashes if the input isn't split into lines. This adds
a hidden flag -raw-file which passes file contents to microsoftDemangle() in
the same way the fuzzer does, for reproducing oss-fuzz reports.

Also change llvm-undname to have a non-0 exit code for invalid symbols.

Differential Revision: https://reviews.llvm.org/D60771

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358485 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRe-commit r357452: SimplifyCFG SinkCommonCodeFromPredecessors: Also sink function...
Hans Wennborg [Tue, 16 Apr 2019 12:13:25 +0000 (12:13 +0000)]
Re-commit r357452: SimplifyCFG SinkCommonCodeFromPredecessors: Also sink function calls without used results (PR41259)

The original commit caused false positives from AddressSanitizer's
use-after-scope checks, which have now been fixed in r358478.

> The code was previously checking that candidates for sinking had exactly
> one use or were a store instruction (which can't have uses). This meant
> we could sink call instructions only if they had a use.
>
> That limitation seemed a bit arbitrary, so this patch changes it to
> "instruction has zero or one use" which seems more natural and removes
> the need to special-case stores.
>
> Differential revision: https://reviews.llvm.org/D59936

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358483 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRemoved CMake cache upgrade code from 2011
Dmitri Gribenko [Tue, 16 Apr 2019 09:46:02 +0000 (09:46 +0000)]
Removed CMake cache upgrade code from 2011

Summary:
This code was added in r141266 to make a breaking change to CMake, but
still be compatible with existing cache files.  The cache files from
2011 are irrelevant today in 2019.

Subscribers: mgorny, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60711

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358482 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoAsan use-after-scope: don't poison allocas if there were untraced lifetime intrinsics...
Hans Wennborg [Tue, 16 Apr 2019 07:54:20 +0000 (07:54 +0000)]
Asan use-after-scope: don't poison allocas if there were untraced lifetime intrinsics in the function (PR41481)

If there are any intrinsics that cannot be traced back to an alloca, we
might have missed the start of a variable's scope, leading to false
error reports if the variable is poisoned at function entry. Instead, if
there are some intrinsics that can't be traced, fail safe and don't
poison the variables in that function.

Differential revision: https://reviews.llvm.org/D60686

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358478 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoUse native llvm JSON library for time profiler output
Anton Afanasyev [Tue, 16 Apr 2019 06:35:07 +0000 (06:35 +0000)]
Use native llvm JSON library for time profiler output

Summary: Replace plain json text output with llvm JSON library wrapper using.

Reviewers: takuto.ikuta, lebedev.ri

Subscribers: hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60609

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358476 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Align instructions to a tab stop in disassembly output
Fangrui Song [Tue, 16 Apr 2019 03:56:55 +0000 (03:56 +0000)]
[llvm-objdump] Align instructions to a tab stop in disassembly output

This relands D60376/rL358405, with the difference: sed 'y/\t/ /' -> tr '\t' ' '
BSD sed doesn't support escape characters for the 'y' command.
I didn't use it in rL358405 because it was not listed at
https://llvm.org/docs/GettingStarted.html#software but it
should be available.

Original description:

In GNU objdump, -w/--wide aligns instructions in the disassembly output.
This patch does the same to llvm-objdump. However, we always use the
wide format (-w/--wide is ignored), because the narrow format
(instructions are misaligned) is probably not very useful.

In llvm-readobj, we made a similar decision: always use the wide format,
accept but ignore -W/--wide.

To save some columns, we change the tab before hex bytes (controlled by
--[no-]show-raw-insn) to a space.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358474 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Simplify PrintHelpMessage() logic
Fangrui Song [Tue, 16 Apr 2019 02:37:29 +0000 (02:37 +0000)]
[llvm-objdump] Simplify PrintHelpMessage() logic

This relands rL358418. It missed one test that should also use -macho
Note, all the other -private-header -exports-trie tests are used
together with -macho.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358472 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[CodeExtractor] Add a few debug lines to understand why a region is not extracted
Quentin Colombet [Tue, 16 Apr 2019 02:12:05 +0000 (02:12 +0000)]
[CodeExtractor] Add a few debug lines to understand why a region is not extracted

The CodeExtractor is not smart enough to compute which basic block is
the entry of a region. Instead it relies on the order of the list
of basic blocks that is handed to it and assumes that the entry
is the first block in the list.

Without the additional debug information, it is hard to understand
why a valid region does not get extracted, because we would miss
that the order of in the list just doesn't match what the CodeExtractor
wants.

NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358471 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRemove some more unused headers from MachineFunction.h and friends.
Eric Christopher [Tue, 16 Apr 2019 01:06:03 +0000 (01:06 +0000)]
Remove some more unused headers from MachineFunction.h and friends.

Most of these headers are still included via transitive includes and
so won't likely show any problems or improvements in incremental
rebuild time.

Differential Revision: https://reviews.llvm.org/D60741

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358468 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoReapply [Support] Fix recursive response file expansion guard
Shoaib Meenai [Tue, 16 Apr 2019 00:18:50 +0000 (00:18 +0000)]
Reapply [Support] Fix recursive response file expansion guard

The test in the dependent revision has been fixed for Windows.

Original commit message:

Response file expansion limits the amount of expansion to prevent
potential infinite recursion. However, the current logic assumes that
any argument beginning with @ is a response file, which is not true for
e.g. `-Xlinker -rpath -Xlinker @executable_path/../lib` on Darwin.
Having too many of these non-response file arguments beginning with @
prevents actual response files from being expanded. Instead, limit based
on the number of successful response file expansions, which should still
prevent infinite recursion but also avoid false positives.

Differential Revision: https://reviews.llvm.org/D60631

llvm-svn: 358452

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358466 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoReapply [Support] Add a test for recursive response file expansion
Shoaib Meenai [Tue, 16 Apr 2019 00:18:47 +0000 (00:18 +0000)]
Reapply [Support] Add a test for recursive response file expansion

Use the appropriate tokenizer to fix the test on Windows.

Original commit message:

I'm going to be modifying the logic to avoid infinitely recursing on
self-referential response files, so add a unit test to verify the
expected behavior.

Differential Revision: https://reviews.llvm.org/D60630

llvm-svn: 358451

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358465 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert [Support] Add a test for recursive response file expansion
Shoaib Meenai [Mon, 15 Apr 2019 22:51:54 +0000 (22:51 +0000)]
Revert [Support] Add a test for recursive response file expansion

This reverts r358451 (git commit c8497467edc5766ae81ffbde58159f8c6af50803)

The test breaks a Windows buildbot:
http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/17016/steps/test-check-all/logs/stdio

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358461 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert [Support] Fix recursive response file expansion guard
Shoaib Meenai [Mon, 15 Apr 2019 22:51:53 +0000 (22:51 +0000)]
Revert [Support] Fix recursive response file expansion guard

This reverts r358452 (git commit c8df4fb9c3865eac52a99602c26bbc070098c3d4)

A dependent commit breaks the Windows buildbots.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358460 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert r358405: "[llvm-objdump] Align instructions to a tab stop in disassembly output"
Alex Lorenz [Mon, 15 Apr 2019 22:36:12 +0000 (22:36 +0000)]
Revert r358405: "[llvm-objdump] Align instructions to a tab stop in disassembly output"

The test fails on darwin due to a sed error:

sed: 1: "y/\t/ /": transform strings are not the same length

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358459 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[AArch64][GlobalISel] Don't do extending loads combine for non-pow-2 types.
Amara Emerson [Mon, 15 Apr 2019 22:34:08 +0000 (22:34 +0000)]
[AArch64][GlobalISel] Don't do extending loads combine for non-pow-2 types.

Since non-pow-2 types are going to get split up into multiple loads anyway,
don't do the [SZ]EXTLOAD combine for those and save us trouble later in
legalization.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358458 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[LSR] Rewrite misses some fixup locations if it splits critical edge
Quentin Colombet [Mon, 15 Apr 2019 22:23:46 +0000 (22:23 +0000)]
[LSR] Rewrite misses some fixup locations if it splits critical edge

If LSR split critical edge during rewriting phi operands and
phi node has other pending fixup operands, we need to
update those pending fixups. Otherwise formulae will not be
implemented completely and some instructions will not be eliminated.

llvm.org/PR41445

Differential Revision: https://reviews.llvm.org/D60645

Patch by: Denis Bakhvalov <denis.bakhvalov@intel.com>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358457 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoOnly use argv[0] as the main executable name if it exists.
Sean Silva [Mon, 15 Apr 2019 22:07:56 +0000 (22:07 +0000)]
Only use argv[0] as the main executable name if it exists.

Under some environments, argv[0] doesn't hold a valid file name, but
sys::fs::getMainExecutable will find the main executable properly.

This patch tweaks the logic to fall back to sys::fs::getMainExecutable
in more situations.

Differential Revision: https://reviews.llvm.org/D60730

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358455 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[EarlyCSE] add more tests for double-negated select condition; NFC
Sanjay Patel [Mon, 15 Apr 2019 21:51:51 +0000 (21:51 +0000)]
[EarlyCSE] add more tests for double-negated select condition; NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358454 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Support] Fix recursive response file expansion guard
Shoaib Meenai [Mon, 15 Apr 2019 21:31:28 +0000 (21:31 +0000)]
[Support] Fix recursive response file expansion guard

Response file expansion limits the amount of expansion to prevent
potential infinite recursion. However, the current logic assumes that
any argument beginning with @ is a response file, which is not true for
e.g. `-Xlinker -rpath -Xlinker @executable_path/../lib` on Darwin.
Having too many of these non-response file arguments beginning with @
prevents actual response files from being expanded. Instead, limit based
on the number of successful response file expansions, which should still
prevent infinite recursion but also avoid false positives.

Differential Revision: https://reviews.llvm.org/D60631

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358452 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Support] Add a test for recursive response file expansion
Shoaib Meenai [Mon, 15 Apr 2019 21:31:25 +0000 (21:31 +0000)]
[Support] Add a test for recursive response file expansion

I'm going to be modifying the logic to avoid infinitely recursing on
self-referential response files, so add a unit test to verify the
expected behavior.

Differential Revision: https://reviews.llvm.org/D60630

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358451 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Limit the 'x' inline assembly constraint to zmm0-15 when used for a 512 type.
Craig Topper [Mon, 15 Apr 2019 21:06:32 +0000 (21:06 +0000)]
[X86] Limit the 'x' inline assembly constraint to zmm0-15 when used for a 512 type.

The 'v' constraint is used to select zmm0-31. This makes 512 bit consistent with 128/256-bit.a

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358450 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Fix a stack folding test to have a full xmm2-31 clobber list instead of stoppin...
Craig Topper [Mon, 15 Apr 2019 21:06:23 +0000 (21:06 +0000)]
[X86] Fix a stack folding test to have a full xmm2-31 clobber list instead of stopping at xmm15. Add an additional dependency to keep instruction below inline asm block.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358449 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoTime profiler: small fixes and optimizations
Anton Afanasyev [Mon, 15 Apr 2019 21:02:47 +0000 (21:02 +0000)]
Time profiler: small fixes and optimizations

Summary: Fixes from Roman's review here: https://reviews.llvm.org/D58675#1465336

Reviewers: lebedev.ri

Subscribers: hiraditya, mgrang, cfe-commits, llvm-commits

Tags: #clang, #llvm

Differential Revision: https://reviews.llvm.org/D60663

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358448 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoAMDGPU: Fix unreachable when counting register usage of SGPR96
Matt Arsenault [Mon, 15 Apr 2019 20:51:12 +0000 (20:51 +0000)]
AMDGPU: Fix unreachable when counting register usage of SGPR96

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358447 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoAMDGPU: Fix printed format of SReg_96
Matt Arsenault [Mon, 15 Apr 2019 20:42:18 +0000 (20:42 +0000)]
AMDGPU: Fix printed format of SReg_96

These are artificial, so I think this should only come up with inline
asm comments.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358446 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[EarlyCSE] add test for select condition double-negation; NFC
Sanjay Patel [Mon, 15 Apr 2019 20:25:31 +0000 (20:25 +0000)]
[EarlyCSE] add test for select condition double-negation; NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358444 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert r358418: "[llvm-objdump] Simplify PrintHelpMessage() logic"
Alex Lorenz [Mon, 15 Apr 2019 20:16:19 +0000 (20:16 +0000)]
Revert r358418: "[llvm-objdump] Simplify PrintHelpMessage() logic"

This reverts commit r358418 as it broke `test/Object/objdump-export-list`
on Darwin.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358443 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Tests] Add a few more tests for LoopPredication w/invariant loads
Philip Reames [Mon, 15 Apr 2019 19:45:27 +0000 (19:45 +0000)]
[Tests] Add a few more tests for LoopPredication w/invariant loads

Making sure to cover an important legality cornercase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358439 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Block i32/i64 for 'k' and 'Yk' in getRegForInlineAsmConstraint without avx512bw.
Craig Topper [Mon, 15 Apr 2019 18:39:45 +0000 (18:39 +0000)]
[X86] Block i32/i64 for 'k' and 'Yk' in getRegForInlineAsmConstraint without avx512bw.

32 and 64 bit k-registers require avx512bw. If we don't block this properly, it leads to a crash.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358436 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[LoopPred] Stop passing around builders [NFC]
Philip Reames [Mon, 15 Apr 2019 18:15:08 +0000 (18:15 +0000)]
[LoopPred] Stop passing around builders [NFC]

This is a preparatory patch for D60093. This patch itself is NFC, but while preparing this I noticed and committed a small hoisting change in rL358419.

The basic structure of the new scheme is that we pass around the guard ("the using instruction"), and select an optimal insert point by examining operands at each construction point. This seems conceptually a bit cleaner to start with as it isolates the knowledge about insertion safety at the actual insertion point.

Note that the non-hoisting path is not actually used at the moment. That's not exercised until D60093 is rebased on this one.

Differential Revision: https://reviews.llvm.org/D60718

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358434 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[x86] update test checks; NFC
Sanjay Patel [Mon, 15 Apr 2019 17:38:47 +0000 (17:38 +0000)]
[x86] update test checks; NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358432 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DEBUGINFO] Prevent Instcombine from dropping debuginfo when removing zexts
Wolfgang Pieb [Mon, 15 Apr 2019 17:36:29 +0000 (17:36 +0000)]
[DEBUGINFO] Prevent Instcombine from dropping debuginfo when removing zexts

Zexts can be treated like no-op casts when it comes to assessing whether their
removal affects debug info.

Reviewer: aprantl

Differential Revision: https://reviews.llvm.org/D60641

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358431 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoAdd explicit dependency to MCDwarf.h in ARC backend.
Pete Couperus [Mon, 15 Apr 2019 17:36:19 +0000 (17:36 +0000)]
Add explicit dependency to MCDwarf.h in ARC backend.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358430 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoFix header guard typo and build error.
Hiroshi Yamauchi [Mon, 15 Apr 2019 17:29:21 +0000 (17:29 +0000)]
Fix header guard typo and build error.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358429 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[CommandLineParser] Add DefaultOption flag
Don Hinton [Mon, 15 Apr 2019 17:18:10 +0000 (17:18 +0000)]
[CommandLineParser] Add DefaultOption flag

Summary: Add DefaultOption flag to CommandLineParser which provides a
default option or alias, but allows users to override it for some
other purpose as needed.

Also, add `-h` as a default alias to `-help`, which can be seamlessly
overridden by applications like llvm-objdump and llvm-readobj which
use `-h` as an alias for other options.

(relanding after revert, r358414)
Added DefaultOptions.clear() to reset().

Reviewers: alexfh, klimek

Reviewed By: klimek

Subscribers: kristina, MaskRay, mehdi_amini, inglorion, dexonsmith, hiraditya, llvm-commits, jhenderson, arphaman, cfe-commits

Tags: #clang, #llvm

Differential Revision: https://reviews.llvm.org/D59746

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358428 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Restore the pavg intrinsics.
Craig Topper [Mon, 15 Apr 2019 17:17:35 +0000 (17:17 +0000)]
[X86] Restore the pavg intrinsics.

The pattern we replaced these with may be too hard to match as demonstrated by
PR41496 and PR41316.

This patch restores the intrinsics and then we can start focusing
on the optimizing the intrinsics.

I've mostly reverted the original patch that removed them. Though I modified
the avx512 intrinsics to not have masking built in.

Differential Revision: https://reviews.llvm.org/D60674

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358427 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoAdd slbfee instruction.
Sean Fertile [Mon, 15 Apr 2019 17:08:43 +0000 (17:08 +0000)]
Add slbfee instruction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358425 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[PGO] Profile guided code size optimization.
Hiroshi Yamauchi [Mon, 15 Apr 2019 16:49:00 +0000 (16:49 +0000)]
[PGO] Profile guided code size optimization.

Summary:
Enable some of the existing size optimizations for cold code under PGO.

A ~5% code size saving in big internal app under PGO.

The way it gets BFI/PSI is discussed in the RFC thread

http://lists.llvm.org/pipermail/llvm-dev/2019-March/130894.html

Note it doesn't currently touch loop passes.

Reviewers: davidxl, eraman

Reviewed By: eraman

Subscribers: mgorny, javed.absar, smeenai, mehdi_amini, eraman, zzheng, steven_wu, dexonsmith, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D59514

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358422 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Fix nullptr deref on invalid conversion operator names in template...
Nico Weber [Mon, 15 Apr 2019 16:42:44 +0000 (16:42 +0000)]
llvm-undname: Fix nullptr deref on invalid conversion operator names in template args

A ConversionOperatorIdentifierNode has a TargetType which is read when
printing it, but if the ConversionOperatorIdentifierNode appears in a
template argument there's nothing that can provide the TargetType.
Normally the COIN is a symbol (leaf) name and takes its TargetType from the
symbol's type, but in a template argument context the COIN can only be
either a non-leaf name piece or a type, and must hence be invalid.

Similar to the COIN check in demangleDeclarator().

Found by oss-fuzz.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358421 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[EarlyCSE] add tests for selects with commuted operands (PR41101); NFC
Sanjay Patel [Mon, 15 Apr 2019 16:01:05 +0000 (16:01 +0000)]
[EarlyCSE] add tests for selects with commuted operands (PR41101); NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358420 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[LoopPred] Hoist and of predicated checks where legal
Philip Reames [Mon, 15 Apr 2019 15:53:25 +0000 (15:53 +0000)]
[LoopPred] Hoist and of predicated checks where legal

If we have multiple range checks which can be predicated, hoist the and of the results outside the loop.  This minorly cleans up the resulting IR, but the main motivation is as a building block for D60093.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358419 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Simplify PrintHelpMessage() logic
Fangrui Song [Mon, 15 Apr 2019 15:52:32 +0000 (15:52 +0000)]
[llvm-objdump] Simplify PrintHelpMessage() logic

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358418 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Wrap things in namespace llvm
Fangrui Song [Mon, 15 Apr 2019 15:31:42 +0000 (15:31 +0000)]
[llvm-objdump] Wrap things in namespace llvm

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358417 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Delete unused forward declarations
Fangrui Song [Mon, 15 Apr 2019 15:08:01 +0000 (15:08 +0000)]
[llvm-objdump] Delete unused forward declarations

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358416 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Reorganize cl::opt variables and move Mach-O specifics to MachODump.cpp
Fangrui Song [Mon, 15 Apr 2019 15:00:10 +0000 (15:00 +0000)]
[llvm-objdump] Reorganize cl::opt variables and move Mach-O specifics to MachODump.cpp

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358415 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert r358337: "[CommandLineParser] Add DefaultOption flag"
Ilya Biryukov [Mon, 15 Apr 2019 14:43:50 +0000 (14:43 +0000)]
Revert r358337: "[CommandLineParser] Add DefaultOption flag"

The change causes test failures under asan. Reverting to unbreak our
integrate.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358414 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[EarlyCSE] regenerate test checks; NFC
Sanjay Patel [Mon, 15 Apr 2019 14:02:37 +0000 (14:02 +0000)]
[EarlyCSE] regenerate test checks; NFC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358407 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-objdump] Align instructions to a tab stop in disassembly output
Fangrui Song [Mon, 15 Apr 2019 13:32:41 +0000 (13:32 +0000)]
[llvm-objdump] Align instructions to a tab stop in disassembly output

Summary:
In GNU objdump, -w/--wide aligns instructions in the disassembly output.
This patch does the same to llvm-objdump. However, we always use the
wide format (-w/--wide is ignored), because the narrow format
(instructions are misaligned) is probably not very useful.

In llvm-readobj, we made a similar decision: always use the wide format,
accept but ignore -W/--wide.

To save some columns, we change the tab before hex bytes (controlled by
--[no-]show-raw-insn) to a space.

Reviewers: rupprecht, jhenderson, grimar

Reviewed By: jhenderson

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60376

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358405 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[InstCombine] canonicalize fdiv after fmul if reassociation is allowed
Sanjay Patel [Mon, 15 Apr 2019 13:23:38 +0000 (13:23 +0000)]
[InstCombine] canonicalize fdiv after fmul if reassociation is allowed

  (X / Y) * Z --> (X * Z) / Y

This can allow other optimizations/reassociations as shown in the test diffs.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358404 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoDAG: propagate ConsecutiveRegs flags to returns too.
Tim Northover [Mon, 15 Apr 2019 12:04:10 +0000 (12:04 +0000)]
DAG: propagate ConsecutiveRegs flags to returns too.

Arguments already have a flag to inform backends when they have been split up.
The AArch64 arm64_32 ABI makes use of these on return types too, so that code
emitted for armv7k can be ABI-compliant.

There should be no CodeGen changes yet, just making more information available.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358399 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoDAG: propagate whether an arg is a pointer for CallingConv decisions.
Tim Northover [Mon, 15 Apr 2019 12:03:54 +0000 (12:03 +0000)]
DAG: propagate whether an arg is a pointer for CallingConv decisions.

The arm64_32 ABI specifies that pointers (despite being 32-bits) should be
zero-extended to 64-bits when passed in registers for efficiency reasons. This
means that the SelectionDAG needs to be able to tell the backend that an
argument was originally a pointer, which is implmented here.

Additionally, some memory intrinsics need to be declared as taking an i8*
instead of an iPTR.

There should be no CodeGen change yet, but it will be triggered when AArch64
backend support for ILP32 is added.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358398 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agogn build: Merge r358390
Nico Weber [Mon, 15 Apr 2019 11:25:23 +0000 (11:25 +0000)]
gn build: Merge r358390

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358397 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-readelf] Correctly dump symbols whose section id is SHN_XINDEX
Eugene Leviant [Mon, 15 Apr 2019 11:21:47 +0000 (11:21 +0000)]
[llvm-readelf] Correctly dump symbols whose section id is SHN_XINDEX

Differential revision: https://reviews.llvm.org/D60614

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358396 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[llvm-readobj] Reapply: Improve error message for --string-dump
Stephen Tozer [Mon, 15 Apr 2019 11:17:48 +0000 (11:17 +0000)]
[llvm-readobj] Reapply: Improve error message for --string-dump

This is a resubmission of a previous patch that caused test failures,
with the fixes for the relevant tests included.

Fixes bug 40630: https://bugs.llvm.org/show_bug.cgi?id=40630

This patch changes the error message when the section specified by
--string-dump cannot be found by including the name of the section in
the error message and changing the prefix text to not imply that the
file itself was invalid. As part of this change some uses of
std::error_code have been replaced with the llvm Error class to better
encapsulate the error info (rather than passing File strings around),
and the WithColor class replaces string literal error prefixes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358395 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[AMDGPU] Fixed incorrect test in vcnd/vcmp optimization
Tim Renouf [Mon, 15 Apr 2019 10:36:24 +0000 (10:36 +0000)]
[AMDGPU] Fixed incorrect test in vcnd/vcmp optimization

This fixes a test I introduced in change D59191 (that added src0 and
src1 modifiers to the v_cndmask instruction for disassembly purposes).

Spotted by David Binderman in bug 41488.

Differential Revision: https://reviews.llvm.org/D60652

Change-Id: I6ac95e66cd84e812ed3359ad57bcd0e13198ba0c

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358392 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Docs] Switch a code block from LLVM to text
Jeremy Morse [Mon, 15 Apr 2019 10:23:22 +0000 (10:23 +0000)]
[Docs] Switch a code block from LLVM to text

While I can't replicate this locally, it looks like the buildbots don't
recognize the IR block in r358385 l764 as IR. Downgrade it to being just
text while I look into it.

http://lab.llvm.org:8011/builders/llvm-sphinx-docs/builds/30132

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358391 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoFileCheck [1/12]: Move variable table in new object
Thomas Preud'homme [Mon, 15 Apr 2019 10:10:11 +0000 (10:10 +0000)]
FileCheck [1/12]: Move variable table in new object

Summary:
This patch is part of a patch series to add support for FileCheck
numeric expressions. This specific patch adds a new class to hold
pattern matching global state.

The table holding the values of FileCheck variable constitutes some sort
of global state for the matching phase, yet is passed as parameters of
all functions using it. This commit create a new FileCheckPatternContext
class pointed at from FileCheckPattern. While it increases the line
count, it separates local data from global state. Later commits build
on that to add numeric expression global state to that class.

Copyright:
    - Linaro (changes up to diff 183612 of revision D55940)
    - GraphCore (changes in later versions of revision D55940 and
                 in new revision created off D55940)

Reviewers: jhenderson, chandlerc, jdenny, probinson, grimar, arichardson, rnk

Subscribers: hiraditya, llvm-commits, probinson, dblaikie, grimar, arichardson, tra, rnk, kristina, hfinkel, rogfer01, JonChesterfield

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60381

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358390 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[TableGen] Include schedule model name in diagnostic.
Simon Tatham [Mon, 15 Apr 2019 10:06:26 +0000 (10:06 +0000)]
[TableGen] Include schedule model name in diagnostic.

If you have more than one schedule model in your TableGen target
definitions, then the diagnostic "No schedule information for
instruction 'foo'" is rather unhelpful, because it doesn't tell you
_which_ schedule model is missing the necessary information (or, as it
might be, missing the UnsupportedFeatures definition that would stop
it thinking it needed it).

Extended the message to include the name of the schedule model that
it's complaining about.

Reviewers: nhaehnle, hfinkel, javedabsar, efriedma, javed.absar

Reviewed By: javed.absar

Subscribers: javed.absar, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60559

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358389 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Docs] Correct some indentation muppetry that trips buildbots
Jeremy Morse [Mon, 15 Apr 2019 10:04:52 +0000 (10:04 +0000)]
[Docs] Correct some indentation muppetry that trips buildbots

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358388 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DWARF] Fix DWARFVerifier::DieRangeInfo::contains
Fangrui Song [Mon, 15 Apr 2019 10:02:36 +0000 (10:02 +0000)]
[DWARF] Fix DWARFVerifier::DieRangeInfo::contains

It didn't handle empty LHS correctly. If two ranges of LHS were
contiguous and jointly contained one range of RHS, it could also be incorrect.

DWARFAddressRange::contains can be removed and its tests can be merged into DWARFVerifier::DieRangeInfo::contains

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358387 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DebugInfo][Docs] Document variable location metadata transformations
Jeremy Morse [Mon, 15 Apr 2019 09:42:38 +0000 (09:42 +0000)]
[DebugInfo][Docs] Document variable location metadata transformations

This patch adds documentation explaining how variable location information is
compiled from the IR representation down to the end of the codegen pipeline,
but avoiding discussion of file formats and encoding.

This should make it clearer how the dbg.value / dbg.declare etc intrinsics
are transformed and arranged into DBG_VALUE instructions, and their meaning.

Differential Revision: https://reviews.llvm.org/D59790

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358385 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoCodegen: Fixed perf branch_weights in couple of tests. NFC.
Yevgeny Rouban [Mon, 15 Apr 2019 09:30:31 +0000 (09:30 +0000)]
Codegen: Fixed perf branch_weights in couple of tests. NFC.

This is need to pass future checks of perf branch_weights metadata.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358384 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agoRevert "[DWARF] Delete redundant check in DWARFAddressRange::intersects"
Fangrui Song [Mon, 15 Apr 2019 09:01:10 +0000 (09:01 +0000)]
Revert "[DWARF] Delete redundant check in DWARFAddressRange::intersects"

This reverts rL358377. It isn't redundant, though the semantic seems weird.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358381 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Transforms][ASan] Move findAllocaForValue() to Utils/Local.cpp. NFC
Alexander Potapenko [Mon, 15 Apr 2019 08:59:56 +0000 (08:59 +0000)]
[Transforms][ASan] Move findAllocaForValue() to Utils/Local.cpp. NFC

Summary:
Factor out findAllocaForValue() from ASan so that we can use it in
MSan to handle lifetime intrinsics.

Reviewers: eugenis, pcc

Subscribers: hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60615

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358380 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[NewPM] Add Option handling for SimplifyCFG
Serguei Katkov [Mon, 15 Apr 2019 08:57:53 +0000 (08:57 +0000)]
[NewPM] Add Option handling for SimplifyCFG

This patch enables passing options to SimplifyCFGPass via the passes pipeline.

Reviewers: chandlerc, fedor.sergeev, leonardchan, philip.pfaffe
Reviewed By: fedor.sergeev
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D60675

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358379 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DWARF] Delete redundant check in DWARFAddressRange::intersects
Fangrui Song [Mon, 15 Apr 2019 08:36:21 +0000 (08:36 +0000)]
[DWARF] Delete redundant check in DWARFAddressRange::intersects

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358377 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DWARF] Fix DWARFVerifier::DieRangeInfo::intersects
Fangrui Song [Mon, 15 Apr 2019 08:30:10 +0000 (08:30 +0000)]
[DWARF] Fix DWARFVerifier::DieRangeInfo::intersects

It was incorrect if RHS had more than 1 ranges and one of the ranges interacted with *this

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358376 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[DWARF] Make DWARFDebugLine::ParsingState::RowNumber a local variable
Fangrui Song [Mon, 15 Apr 2019 07:40:30 +0000 (07:40 +0000)]
[DWARF] Make DWARFDebugLine::ParsingState::RowNumber a local variable

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358374 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[SelectionDAG] Use KnownBits::computeForAddSub/computeForAddCarry
Bjorn Pettersson [Mon, 15 Apr 2019 07:19:11 +0000 (07:19 +0000)]
[SelectionDAG] Use KnownBits::computeForAddSub/computeForAddCarry

Summary:
Use KnownBits::computeForAddSub/computeForAddCarry
in SelectionDAG::computeKnownBits when doing value
tracking for addition/subtraction.

This should improve the precision of the known bits,
as we only used to make a simple estimate of known
zeroes. The KnownBits support functions are also
able to deduce bits that are known to be one in the
result.

Reviewers: spatel, RKSimon, nikic, lebedev.ri

Reviewed By: nikic

Subscribers: nikic, javed.absar, lebedev.ri, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60460

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358372 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Regenerate checks for domain-reassignment.mir
Craig Topper [Mon, 15 Apr 2019 05:22:47 +0000 (05:22 +0000)]
[X86] Regenerate checks for domain-reassignment.mir

Apparently there are some stray IMPLICIT_DEF operations that weren't in the
checks. Not sure if they've always been there or something changed at some
point.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358371 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Sparc] Fix typo. NFC.
Jim Lin [Mon, 15 Apr 2019 05:16:46 +0000 (05:16 +0000)]
[Sparc] Fix typo. NFC.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358370 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[GlobalISel] Enable CSE in the IRTranslator & legalizer for -O0 with constants only.
Amara Emerson [Mon, 15 Apr 2019 05:04:20 +0000 (05:04 +0000)]
[GlobalISel] Enable CSE in the IRTranslator & legalizer for -O0 with constants only.

Other opcodes shouldn't be CSE'd until we can be sure debug info quality won't
be degraded.

This change also improves the IRTranslator so that in most places, but not all,
it creates constants using the MIRBuilder directly instead of first creating a
new destination vreg and then creating a constant. By doing this, the
buildConstant() method can just return the vreg of an existing G_CONSTANT
instead of having to create a COPY from it.

I measured a 0.2% improvement in compile time and a 0.9% improvement in code
size at -O0 ARM64.

Compile time:
Program                                        base   cse    diff
test-suite...ark/tramp3d-v4/tramp3d-v4.test     9.04   9.12  0.8%
test-suite...Mark/mafft/pairlocalalign.test     2.68   2.66 -0.7%
test-suite...-typeset/consumer-typeset.test     5.53   5.51 -0.4%
test-suite :: CTMark/lencod/lencod.test         5.30   5.28 -0.3%
test-suite :: CTMark/Bullet/bullet.test        25.82  25.76 -0.2%
test-suite...:: CTMark/ClamAV/clamscan.test     6.92   6.90 -0.2%
test-suite...TMark/7zip/7zip-benchmark.test    34.24  34.17 -0.2%
test-suite :: CTMark/SPASS/SPASS.test           6.25   6.24 -0.1%
test-suite...:: CTMark/sqlite3/sqlite3.test     1.66   1.66 -0.1%
test-suite :: CTMark/kimwitu++/kc.test         13.61  13.60 -0.0%
Geomean difference                                          -0.2%

Code size:
Program                                        base     cse      diff
test-suite...-typeset/consumer-typeset.test    1315632  1266480 -3.7%
test-suite...:: CTMark/ClamAV/clamscan.test    1313892  1297508 -1.2%
test-suite :: CTMark/lencod/lencod.test        1439504  1423112 -1.1%
test-suite...TMark/7zip/7zip-benchmark.test    2936980  2904172 -1.1%
test-suite :: CTMark/Bullet/bullet.test        3478276  3445460 -0.9%
test-suite...ark/tramp3d-v4/tramp3d-v4.test    8082868  8033492 -0.6%
test-suite :: CTMark/kimwitu++/kc.test         3870380  3853972 -0.4%
test-suite :: CTMark/SPASS/SPASS.test          1434904  1434896 -0.0%
test-suite...Mark/mafft/pairlocalalign.test    764528   764528   0.0%
test-suite...:: CTMark/sqlite3/sqlite3.test    782092   782092   0.0%
Geomean difference                                              -0.9%

Differential Revision: https://reviews.llvm.org/D60580

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358369 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[GlobalISel] Introduce a CSEConfigBase class to allow targets to define their own...
Amara Emerson [Mon, 15 Apr 2019 04:53:46 +0000 (04:53 +0000)]
[GlobalISel] Introduce a CSEConfigBase class to allow targets to define their own CSE configs.

Because CodeGen can't depend on GlobalISel, we need a way to encapsulate the CSE
configs that can be passed between TargetPassConfig and the targets' custom
pass configs. This CSEConfigBase allows targets to create custom CSE configs
which is then used by the GISel passes for the CSEMIRBuilder.

This support will be used in a follow up commit to allow constant-only CSE for
-O0 compiles in D60580.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358368 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Fix oss-fuzz-foudn crash-on-invalid with incomplete special table nodes
Nico Weber [Sun, 14 Apr 2019 23:32:37 +0000 (23:32 +0000)]
llvm-undname: Fix oss-fuzz-foudn crash-on-invalid with incomplete special table nodes

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358367 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agogn build: Merge r358297
Nico Weber [Sun, 14 Apr 2019 23:23:14 +0000 (23:23 +0000)]
gn build: Merge r358297

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358366 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agogn build: Merge r358243
Nico Weber [Sun, 14 Apr 2019 23:21:58 +0000 (23:21 +0000)]
gn build: Merge r358243

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358365 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agogn build: Merge r358272
Nico Weber [Sun, 14 Apr 2019 23:19:32 +0000 (23:19 +0000)]
gn build: Merge r358272

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358364 91177308-0d34-0410-b5e6-96231b3b80d8

5 years agollvm-undname: Fix another crash-on-invalid found by oss-fuzz
Nico Weber [Sun, 14 Apr 2019 23:08:12 +0000 (23:08 +0000)]
llvm-undname: Fix another crash-on-invalid found by oss-fuzz

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358363 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Redefine KUNPCK instructions to take a narrower source register class than...
Craig Topper [Sun, 14 Apr 2019 20:52:42 +0000 (20:52 +0000)]
[X86] Redefine KUNPCK instructions to take a narrower source register class than destination register class. Remove copies from the isel output pattern.

There's no reason for the inputs to be the destination register class. This just
forces an unnecessary copy in the output patterns.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358362 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Put the locked mi8 instrutions above the locked mi/mi32 so they will be prefered.
Craig Topper [Sun, 14 Apr 2019 19:00:00 +0000 (19:00 +0000)]
[X86] Put the locked mi8 instrutions above the locked mi/mi32 so they will be prefered.

We want 64mi8 to be prefered over 64mi32. The order for 16mi/32mi doesn't
really matter.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358361 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Change IMUL with immediate instruction order to ri8 instructions come before...
Craig Topper [Sun, 14 Apr 2019 18:59:57 +0000 (18:59 +0000)]
[X86] Change IMUL with immediate instruction order to ri8 instructions come before ri/ri32 instructions.

This will ensure IMUL64ri8 is tried before IMUL64ri32. For IMUL32 and IMUL16 the
order doesn't really matter because only the ri8 versions use a predicate. That
automatically gives them priority.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358360 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Move VPTESTM matching from the isel table to custom code in X86ISelDAGToDAG.
Craig Topper [Sun, 14 Apr 2019 18:26:11 +0000 (18:26 +0000)]
[X86] Move VPTESTM matching from the isel table to custom code in X86ISelDAGToDAG.

We had many tablegen patterns for these instructions. And due to the
commutability of the patterns, tablegen expands them to even more patterns. All
together VPTESTMD patterns accounted for more the 50K of the 610K isel table.
This had gotten bad when we stopped canonicalizing AND to vXi64. This required
a pattern for every combination of bitcast input type.

This change moves the matching to custom code where it is easier to look through
the bitcasts without being concerned with the specific types.

The test changes are because we are now stricter with one use checks as its
required to make load folding legal. We now require the AND and any BITCAST to
only have a single use. This prevents forming VPTESTM and a VPAND with the same
inputs.

We now support broadcast loads for 128/256 patterns without VLX. We'll widen to
512-bit like and still fold the broadcast since the amount of memory read
doesn't change.

There are a few tests that got slightly longer because are now prefering
load + VPTESTM over XOR+VPCMPEQ for (seteq (load), allzeros). Previously we were
able to share the XOR with multiple VPTESTM instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358359 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[X86] Don't form masked vpcmp/vcmp/vptestm operations if the setcc node has more...
Craig Topper [Sun, 14 Apr 2019 18:26:06 +0000 (18:26 +0000)]
[X86] Don't form masked vpcmp/vcmp/vptestm operations if the setcc node has more than one use.

We're better of emitting a single compare + kand rather than a compare for the
other use and a masked compare.

I'm looking into using custom instruction selection for VPTESTM to reduce the
ridiculous number of permutations of patterns in the isel table. Putting a one
use check on all masked compare folding makes load fold matching in the custom
code easier.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358358 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[ConstantRange] Simplify unittests after getSetSize was removed
Fangrui Song [Sun, 14 Apr 2019 09:19:15 +0000 (09:19 +0000)]
[ConstantRange] Simplify unittests after getSetSize was removed

Reviewers: lebedev.ri, nikic

Reviewed By: nikic

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60662

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358354 91177308-0d34-0410-b5e6-96231b3b80d8

5 years ago[Mem2Reg] Delete unused PointerAllocaValues
Fangrui Song [Sun, 14 Apr 2019 07:28:29 +0000 (07:28 +0000)]
[Mem2Reg] Delete unused PointerAllocaValues

It is unused after AliasSetTracker support was removed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@358352 91177308-0d34-0410-b5e6-96231b3b80d8