platform/upstream/llvm.git
5 years ago[WebAssembly] Fix typo in relocation checking
Sam Clegg [Tue, 23 Apr 2019 14:49:38 +0000 (14:49 +0000)]
[WebAssembly] Fix typo in relocation checking

Runtime relocation are generated for relocations of type
R_WASM_MEMORY_ADDR_I32 when in PIC mode (either -shared or -pie).

Followup on https://reviews.llvm.org/D60882.

Differential Revision: https://reviews.llvm.org/D60992

llvm-svn: 358995

5 years ago[RISCV] Support assembling %tls_{ie,gd}_pcrel_hi modifiers
Lewis Revill [Tue, 23 Apr 2019 14:46:13 +0000 (14:46 +0000)]
[RISCV] Support assembling %tls_{ie,gd}_pcrel_hi modifiers

This patch adds support for parsing and assembling the %tls_ie_pcrel_hi
and %tls_gd_pcrel_hi modifiers.

Differential Revision: https://reviews.llvm.org/D55342

llvm-svn: 358994

5 years agogn build: Merge r358944
Nico Weber [Tue, 23 Apr 2019 14:32:18 +0000 (14:32 +0000)]
gn build: Merge r358944

llvm-svn: 358993

5 years ago[AMDGPU] Fix hidden argument metadata duplication for V3
Scott Linder [Tue, 23 Apr 2019 14:31:17 +0000 (14:31 +0000)]
[AMDGPU] Fix hidden argument metadata duplication for V3

Essentially complete a proper rebase of the V3 metadata change over
https://reviews.llvm.org/D49096.

Minimize the diff between the V2 and V3 variants of the relevant lit
tests, and clean up some trailing whitespace.

llvm-svn: 358992

5 years agogn build: Merge r358949
Nico Weber [Tue, 23 Apr 2019 14:31:15 +0000 (14:31 +0000)]
gn build: Merge r358949

llvm-svn: 358991

5 years ago[LLD][ELF] - Remove dynamic-section-sh_size.elf binary, convert test to yaml. NFCI.
George Rimar [Tue, 23 Apr 2019 14:21:31 +0000 (14:21 +0000)]
[LLD][ELF] - Remove dynamic-section-sh_size.elf binary, convert test to yaml. NFCI.

dynamic-section-sh_size.elf was introduced in D25090.
Now it is possible to use yaml2obj instead.
That is what this patch does.

Also I added one more case of a possibly broken .dynamic
section just in case.

llvm-svn: 358990

5 years ago[X86] Pull out collectConcatOps helper. NFCI.
Simon Pilgrim [Tue, 23 Apr 2019 14:07:49 +0000 (14:07 +0000)]
[X86] Pull out collectConcatOps helper. NFCI.

Create collectConcatOps helper that returns all the subvector ops for CONCAT_VECTORS or a INSERT_SUBVECTOR series.

llvm-svn: 358989

5 years ago[libc++] Remove redundant conditionals for Apple platforms
Louis Dionne [Tue, 23 Apr 2019 14:05:04 +0000 (14:05 +0000)]
[libc++] Remove redundant conditionals for Apple platforms

Summary:
In a bunch of places, we used to check whether LIBCXX_BUILDING_LIBCXXABI
is defined OR we're building for an Apple platform. This used to
be necessary in a time when Apple's build script did NOT define
LIBCXX_BUILDING_LIBCXXABI. However this is not relevant anymore
since Apple's build does define LIBCXX_BUILDING_LIBCXXABI.

Reviewers: EricWF

Subscribers: christof, jkorous, dexonsmith, libcxx-commits

Tags: #libc

Differential Revision: https://reviews.llvm.org/D60842

llvm-svn: 358988

5 years agoARM: disallow add/sub to sp unless Rn is also sp.
Tim Northover [Tue, 23 Apr 2019 13:50:13 +0000 (13:50 +0000)]
ARM: disallow add/sub to sp unless Rn is also sp.

The manual says that Thumb2 add/sub instructions are only allowed to modify sp
if the first source is also sp. This is slightly different from the usual rGPR
restriction since it's context-sensitive, so implement it in C++.

llvm-svn: 358987

5 years ago[Docs] ReleaseNotes: fixup markup in memcmp()->bcmp() entry
Roman Lebedev [Tue, 23 Apr 2019 13:46:18 +0000 (13:46 +0000)]
[Docs] ReleaseNotes: fixup markup in memcmp()->bcmp() entry

llvm-svn: 358986

5 years ago[LLD][ELF] - Remove file-class.a binary from inputs and improve the test case.
George Rimar [Tue, 23 Apr 2019 13:27:54 +0000 (13:27 +0000)]
[LLD][ELF] - Remove file-class.a binary from inputs and improve the test case.

file-class.a was used to diagnose the "corrupted ELF file: invalid file class"
error when the object was fetched from the archive.

file-class.a contained an object of 16 bytes size. I replaced it with
an echo call (because it is impossible to use yaml2obj for that, and I am
not sure it is worth to support), and also increased its size to 18 bytes.
That allowed to also test a case when such object is a regular input and not an
archive member (we have a bit different logic for these cases).

llvm-svn: 358985

5 years ago[DAGCombiner] generalize binop-of-splats scalarization
Sanjay Patel [Tue, 23 Apr 2019 13:16:41 +0000 (13:16 +0000)]
[DAGCombiner] generalize binop-of-splats scalarization

If we only match build vectors, we can miss some patterns
that use shuffles as seen in the affected tests.

Note that the underlying calls within getSplatSourceVector()
have the potential for compile-time explosion because of
exponential recursion looking through binop opcodes, but
currently the list of supported opcodes is very limited.
Both of those problems should be addressed in follow-up
patches.

llvm-svn: 358984

5 years agoAMDGPU: Fix LCSSA phi lowering in SILowerI1Copies
Nicolai Haehnle [Tue, 23 Apr 2019 13:12:52 +0000 (13:12 +0000)]
AMDGPU: Fix LCSSA phi lowering in SILowerI1Copies

Summary:
When an LCSSA phi survives through instruction selection, the pass
ends up removing that phi entirely because it is dominated by the
logic that does the lanemask merging.

This then used to trigger an assertion when processing a dependent
phi instruction.

Change-Id: Id4949719f8298062fe476a25718acccc109113b6

Reviewers: llvm-commits

Subscribers: kzhuravl, jvesely, wdng, yaxunl, t-tye, tpr, dstuttard, rtaylor, arsenm

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60999

llvm-svn: 358983

5 years ago[CallSite removal] move InlineCost to CallBase usage
Fedor Sergeev [Tue, 23 Apr 2019 12:43:27 +0000 (12:43 +0000)]
[CallSite removal] move InlineCost to CallBase usage

Converting InlineCost interface and its internals into CallBase usage.
Inliners themselves are still not converted.

Reviewed By: reames
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60636

llvm-svn: 358982

5 years ago[ELF] Change default output section type to SHT_PROGBITS
Andrew Ng [Tue, 23 Apr 2019 12:38:52 +0000 (12:38 +0000)]
[ELF] Change default output section type to SHT_PROGBITS

This fixes an issue where a symbol only section at the start of a
PT_LOAD segment, causes incorrect alignment of the file offset for the
start of the segment which results in the output of an invalid ELF.

SHT_PROGBITS was the default output section type in the past.

Differential Revision: https://reviews.llvm.org/D60131

llvm-svn: 358981

5 years ago[LLD][COFF] Fix /linkrepro with output options that take a filename or path
Alexandre Ganea [Tue, 23 Apr 2019 12:30:49 +0000 (12:30 +0000)]
[LLD][COFF] Fix /linkrepro with output options that take a filename or path

The following options: /pdb, /out or /implib now emit in the repro.tar/response.txt only a filename stripped from its path, to avoid non-existent paths on the reproducer's machine.

Differential Revision: https://reviews.llvm.org/D59530

llvm-svn: 358980

5 years ago[ELF] Change findOrphanPos to only consider live sections
Andrew Ng [Tue, 23 Apr 2019 12:17:15 +0000 (12:17 +0000)]
[ELF] Change findOrphanPos to only consider live sections

This patch changes the behaviour of findOrphanPos to only consider live
sections when placing orphan sections. This used to be how it behaved in
the past.

Differential Revision: https://reviews.llvm.org/D60273

llvm-svn: 358979

5 years agoRemoving the explicit specifier from some default constructors; NFC.
Aaron Ballman [Tue, 23 Apr 2019 12:16:28 +0000 (12:16 +0000)]
Removing the explicit specifier from some default constructors; NFC.

llvm-svn: 358978

5 years ago[ARM] Update check for CBZ in Ifcvt
David Green [Tue, 23 Apr 2019 12:11:26 +0000 (12:11 +0000)]
[ARM] Update check for CBZ in Ifcvt

The check for creating CBZ in constant island pass recently obtained the
ability to search backwards to find a Cmp instruction. The code in IfCvt should
mirror this to allow more conversions to the smaller form. The common code has
been pulled out into a separate function to be shared between the two places.

Differential Revision: https://reviews.llvm.org/D60090

llvm-svn: 358977

5 years agoMove postfix expression code out of the NativePDB plugin
Pavel Labath [Tue, 23 Apr 2019 11:50:07 +0000 (11:50 +0000)]
Move postfix expression code out of the NativePDB plugin

Summary:
The NativePDB plugin contains code to convert "programs" describing the
layout of function frames into dwarf (for easier interaction with the
rest of lldb). This functionality is useful for the Breakpad plugin too,
as it contains the same kind of expressions (because breakpad info is
generated from pdb files).

In this patch, I move the core classes of this code into a common place,
where it can be used from both files. Previously, these were the details
of the implementation, but here I am exposing them (instead of just a
single "string->string" conversion function), as breakpad will need to
use these in a slightly different way. The reason for that is that
breakpad files generated from dwarf expressions use a slightly different
syntax, although most of the core code can be reused with a bit of
thought.

This is also the reason why I am not moving the parsing or dwarf
generation bits, as they will need to be generalized a bit before
they're usable for both scenarios.

This patch should be NFC, modulo renaming the moved entities to more
neutral names.

The reason I am moving this to the "Symbol" library, is because both
customers will be "Symbol"Files, and also the unwinding code lives in
the Symbol library. From a purely dependency standpoint this code will
probably be standalone, and so it could be moved all the way to Utility,
but that seems too low for this kind of functionality.

Reviewers: jasonmolenda, amccarth, clayborg, JDevlieghere, aleksandr.urakov

Subscribers: aprantl, markmentovai, lldb-commits

Differential Revision: https://reviews.llvm.org/D60599

llvm-svn: 358976

5 years ago[PPC][PPC64] Improve some llvm-objdump -d -D tests
Fangrui Song [Tue, 23 Apr 2019 11:47:28 +0000 (11:47 +0000)]
[PPC][PPC64] Improve some llvm-objdump -d -D tests

Various improvement:

Some offsets in disassembly are incorrect after several layout adjustment. Fix them.
llvm-objdump -D should not be used. -D dumps unrelated non-text sections. Replace them with llvm-objdump -d, llvm-readelf -x, etc
Many llvm-objdump -d tests use {{.*}} . Add the option --no-show-raw-insn to avoid check hex bytes.

ppc64-long-branch.s does not need a shared object. Delete it.
Make ppc64-ifunc.s check 2 ifuncs.

Reviewers: ruiu, espindola

Subscribers: emaste, nemanjai, arichardson, kbarton, jsji, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60998

llvm-svn: 358975

5 years ago[ARM] Don't replicate instructions in Ifcvt at minsize
David Green [Tue, 23 Apr 2019 11:46:58 +0000 (11:46 +0000)]
[ARM] Don't replicate instructions in Ifcvt at minsize

Ifcvt can replicate instructions as it converts them to be predicated. This
stops that from happening on thumb2 targets at minsize where an extra IT
instruction is likely needed.

Differential Revision: https://reviews.llvm.org/D60089

llvm-svn: 358974

5 years agoFix "-Wimplicit-fallthrough" warning. NFCI.
Simon Pilgrim [Tue, 23 Apr 2019 11:45:28 +0000 (11:45 +0000)]
Fix "-Wimplicit-fallthrough" warning. NFCI.

llvm-svn: 358973

5 years ago[LLD][ELF] - Fix mips-invalid-options-descriptor.s test case and convert to YAML.
George Rimar [Tue, 23 Apr 2019 11:19:51 +0000 (11:19 +0000)]
[LLD][ELF] - Fix mips-invalid-options-descriptor.s test case and convert to YAML.

It was initially introduced in D25229 to report the "zero option descriptor size"
error message. In following commits it was broken and did not report this
error anymore. I think that happened because elf object was a result of fuzzing
and it was broken in many ways.

This patch converts this test to a YAML, removes a binary and hence fixes the
original intention.

llvm-svn: 358972

5 years ago[Analyzer] Second fix for last commit for IteratorChecker
Adam Balogh [Tue, 23 Apr 2019 11:18:50 +0000 (11:18 +0000)]
[Analyzer] Second fix for last commit for IteratorChecker

A variable was redeclared instead of assigned in an internal
block, leaving the original uninitialized. This is fixed now.

llvm-svn: 358971

5 years agoFix MSVC "32-bit shift implicitly converted to 64 bits" warning. NFCI.
Simon Pilgrim [Tue, 23 Apr 2019 11:16:16 +0000 (11:16 +0000)]
Fix MSVC "32-bit shift implicitly converted to 64 bits" warning. NFCI.

llvm-svn: 358970

5 years agoFix MSVC "32-bit shift implicitly converted to 64 bits" warning. NFCI.
Simon Pilgrim [Tue, 23 Apr 2019 11:11:34 +0000 (11:11 +0000)]
Fix MSVC "32-bit shift implicitly converted to 64 bits" warning. NFCI.

llvm-svn: 358969

5 years ago[analyzer][CrossTU] Extend CTU to VarDecls with initializer
Rafael Stahl [Tue, 23 Apr 2019 11:04:41 +0000 (11:04 +0000)]
[analyzer][CrossTU] Extend CTU to VarDecls with initializer

Summary:
The existing CTU mechanism imports `FunctionDecl`s where the definition is available in another TU. This patch extends that to VarDecls, to bind more constants.

- Add VarDecl importing functionality to CrossTranslationUnitContext
- Import Decls while traversing them in AnalysisConsumer
- Add VarDecls to CTU external mappings generator
- Name changes from "external function map" to "external definition map"

Reviewers: NoQ, dcoughlin, xazax.hun, george.karpenkov, martong

Reviewed By: xazax.hun

Subscribers: Charusso, baloghadamsoftware, mikhail.ramalho, Szelethus, donat.nagy, dkrupp, george.karpenkov, mgorny, whisperity, szepet, rnkovacs, a.sidorin, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D46421

llvm-svn: 358968

5 years agomodify-python-lldb: Remove \a-removing code
Pavel Labath [Tue, 23 Apr 2019 10:04:22 +0000 (10:04 +0000)]
modify-python-lldb: Remove \a-removing code

instead, remove \a directly from the interface files.

llvm-svn: 358967

5 years ago[LLD][ELF] - Remove multiple-eh-relocs.elf binary from the inputs. NFCI.
George Rimar [Tue, 23 Apr 2019 10:03:09 +0000 (10:03 +0000)]
[LLD][ELF] - Remove multiple-eh-relocs.elf binary from the inputs. NFCI.

Introduced multiple-relocations-sections.test based on YAML instead.

llvm-svn: 358966

5 years ago[DAGCombiner] Combine OR as ADD when no common bits are set
Bjorn Pettersson [Tue, 23 Apr 2019 10:01:08 +0000 (10:01 +0000)]
[DAGCombiner] Combine OR as ADD when no common bits are set

Summary:
The DAGCombiner is rewriting (canonicalizing) an ISD::ADD
with no common bits set in the operands as an ISD::OR node.

This could sometimes result in "missing out" on some
combines that normally are performed for ADD. To be more
specific this could happen if we already have rewritten an
ADD into OR, and later (after legalizations or combines)
we expose patterns that could have been optimized if we
had seen the OR as an ADD (e.g. reassociations based on ADD).

To make the DAG combiner less sensitive to if ADD or OR is
used for these "no common bits set" ADD/OR operations we
now apply most of the ADD combines also to an OR operation,
when value tracking indicates that the operands have no
common bits set.

Reviewers: spatel, RKSimon, craig.topper, kparzysz

Reviewed By: spatel

Subscribers: arsenm, rampitec, lebedev.ri, jvesely, nhaehnle, hiraditya, javed.absar, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D59758

llvm-svn: 358965

5 years agoFuncUnwinders: remove "current_offset" from function arguments
Pavel Labath [Tue, 23 Apr 2019 09:57:14 +0000 (09:57 +0000)]
FuncUnwinders: remove "current_offset" from function arguments

Summary:
This argument was added back in 2010 (r118882) to support the ability to unwind
from functions whose eh_frame entry does not cover the entire range of
the function.

However, due to the caching happening in FuncUnwinders, this solution is
very fragile. FuncUnwinders will cache the plan it got from eh_frame
regardless of the value of the current_offset, so our ability to unwind
from a given function depended what was the value of "current_offset" the
first time that this function was called.

Furthermore, since the "image show-unwind" command did not know what's
the right offset to pass, this created an unfortunate situation where
"image show-unwind" would show no valid plans for a function, even
though they were available and being used.

In this patch I implement the feature slightly differently. Instead of
giving just a base address to the eh_frame unwinder, I give it the
entire range we are interested in. Then, I change the unwinder to return
the first plan that covers (even partially) that range. This way even a
partial plan will be returned, regardless of the address in the function
where we are stopped at.

This solution is still not 100% correct, as it will not handle a
function which is covered by two independent fde entries. However, I
don't expect anybody will write this kind of functions, and this wasn't
handled by the previous implementation either. If this is ever needed in
the future. The eh_frame unwinder can be extended to return "composite"
unwind plans created by merging sevelar fde entries.

I also create a test which triggers this scenario. As doing this is
virtually impossible without hand-written assembly, the test only works
on x86 linux.

Reviewers: jasonmolenda, clayborg

Subscribers: lldb-commits

Differential Revision: https://reviews.llvm.org/D60829

llvm-svn: 358964

5 years ago[AArch64] Add support for MTE intrinsics
Javed Absar [Tue, 23 Apr 2019 09:39:58 +0000 (09:39 +0000)]
[AArch64] Add support for MTE intrinsics

This patch provides intrinsics support for Memory Tagging Extension (MTE),
which was introduced with the Armv8.5-a architecture.
The intrinsics are described in detail in the latest
ACLE Q1 2019 documentation: https://developer.arm.com/docs/101028/latest
Reviewed by: David Spickett
Differential Revision: https://reviews.llvm.org/D60486

llvm-svn: 358963

5 years ago[ARM][FIX] Add missing f16.lane.vldN/vstN lowering
Diogo N. Sampaio [Tue, 23 Apr 2019 09:36:39 +0000 (09:36 +0000)]
[ARM][FIX] Add missing f16.lane.vldN/vstN lowering

Summary:
Add missing D and Q lane VLDSTLane lowering
for fp16 elements.

Reviewers: efriedma, kosarev, SjoerdMeijer, ostannard

Reviewed By: efriedma

Subscribers: javed.absar, kristof.beyls, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60874

llvm-svn: 358962

5 years ago[LLD][ELF] - Update test case after llvm-mc change.
George Rimar [Tue, 23 Apr 2019 09:30:26 +0000 (09:30 +0000)]
[LLD][ELF] - Update test case after llvm-mc change.

rL358960 "[llvm-mc] - Properly set the the address align field of the compressed sections."
changed the sh_addralign field of the "zlib" compressed debug sections.

llvm-svn: 358961

5 years ago[llvm-mc] - Properly set the the address align field of the compressed sections.
George Rimar [Tue, 23 Apr 2019 09:16:53 +0000 (09:16 +0000)]
[llvm-mc] - Properly set the the address align field of the compressed sections.

About the compressed sections spec says:
(https://docs.oracle.com/cd/E37838_01/html/E36783/section_compression.html)
sh_addralign fields of the section header for a compressed section
reflect the requirements of the compressed section.

Currently, llvm-mc always puts uncompressed section alignment to sh_addralign.
It is not correct. zlib styled section contains an Elfxx_Chdr header,
so we should either use 4 or 8 values depending on the target
(Uncompressed section alignment is stored in ch_addralign field of the compression header).

GNU assembler version 2.31.1 also has this issue,
but in 2.32.51 it was already fixed. This is how it was found
during debugging of the https://bugs.llvm.org/show_bug.cgi?id=40482
actually.

Differential revision: https://reviews.llvm.org/D60965

llvm-svn: 358960

5 years agoUnwindPlan: pretty-print dwarf expressions
Pavel Labath [Tue, 23 Apr 2019 09:16:51 +0000 (09:16 +0000)]
UnwindPlan: pretty-print dwarf expressions

Summary:
Previously we were printing the dwarf expressions in unwind rules simply
as "dwarf-expr". This patch uses the existing dwarf-printing
capabilities in lldb to enhance this dump output, and print the full
decoded dwarf expression.

Reviewers: jasonmolenda, clayborg

Subscribers: aprantl, lldb-commits

Differential Revision: https://reviews.llvm.org/D60949

llvm-svn: 358959

5 years ago[LSR] Limit the recursion for setup cost
David Green [Tue, 23 Apr 2019 08:52:21 +0000 (08:52 +0000)]
[LSR] Limit the recursion for setup cost

In some circumstances we can end up with setup costs that are very complex to
compute, even though the scevs are not very complex to create. This can also
lead to setupcosts that are calculated to be exactly -1, which LSR treats as an
invalid cost. This patch puts a limit on the recursion depth for setup cost to
prevent them taking too long.

Thanks to @reames for the report and test case.

Differential Revision: https://reviews.llvm.org/D60944

llvm-svn: 358958

5 years agoyamlify TestMiniDumpUUID binaries
Pavel Labath [Tue, 23 Apr 2019 08:49:39 +0000 (08:49 +0000)]
yamlify TestMiniDumpUUID binaries

Summary:
Instead of checking in raw minidump binaries, check in their yaml form,
and call yaml2obj in the test.

Reviewers: clayborg

Subscribers: javed.absar, lldb-commits

Differential Revision: https://reviews.llvm.org/D60948

llvm-svn: 358957

5 years ago[Analyzer] Fix for previous commit
Adam Balogh [Tue, 23 Apr 2019 07:45:10 +0000 (07:45 +0000)]
[Analyzer] Fix for previous commit

A compilation warning was in my previous commit which broke the buildbot
because it is using `-Werror` for compilation. This patch fixes this
issue.

llvm-svn: 358955

5 years ago[PowerPC] Fix test with -fno-discard-value-names after rC358949
Fangrui Song [Tue, 23 Apr 2019 07:39:23 +0000 (07:39 +0000)]
[PowerPC] Fix test with -fno-discard-value-names after rC358949

For the clang driver, -DLLVM_ENABLE_ASSERTIONS=off builds default to discard value names.

llvm-svn: 358953

5 years ago[Analyzer] Instead of recording comparisons in interator checkers do an eager state...
Adam Balogh [Tue, 23 Apr 2019 07:15:55 +0000 (07:15 +0000)]
[Analyzer] Instead of recording comparisons in interator checkers do an eager state split

Currently iterator checkers record comparison of iterator positions
and process them for keeping track the distance between them (e.g.
whether a position is the same as the end position). However this
makes some processing unnecessarily complex and it is not needed at
all: we only need to keep track between the abstract symbols stored
in these iterator positions. This patch changes this and opens the
path to comparisons to the begin() and end() symbols between the
container (e.g. size, emptiness) which are stored as symbols, not
iterator positions. The functionality of the checker is unchanged.

Differential Revision: https://reviews.llvm.org/D53701

llvm-svn: 358951

5 years ago[PPC64] Allow R_PPC64_DTPREL* to preemptable local-dynamic symbols
Fangrui Song [Tue, 23 Apr 2019 06:31:44 +0000 (06:31 +0000)]
[PPC64] Allow R_PPC64_DTPREL* to preemptable local-dynamic symbols

Similar to D60945.

Differential Revision: https://reviews.llvm.org/D60994

llvm-svn: 358950

5 years ago[PowerPC] [Clang] Port MMX intrinsics and basic test cases to Power
Qiu Chaofan [Tue, 23 Apr 2019 05:50:24 +0000 (05:50 +0000)]
[PowerPC] [Clang] Port MMX intrinsics and basic test cases to Power

Port mmintrin.h which include x86 MMX intrinsics implementation to PowerPC platform (using Altivec).

To make the include process correct, PowerPC's toolchain class is overrided to insert new headers directory (named ppc_wrappers) into the path. Basic test cases for several intrinsic functions are added.

The header is mainly developed by Steven Munroe, with contributions from Paul Clarke, Bill Schmidt, Jinsong Ji and Zixuan Wu.

Reviewed By: Jinsong Ji

Differential Revision: https://reviews.llvm.org/D59924

llvm-svn: 358949

5 years ago[WebAssembly] Bail out of fastisel earlier when computing PIC addresses
Sam Clegg [Tue, 23 Apr 2019 03:43:26 +0000 (03:43 +0000)]
[WebAssembly] Bail out of fastisel earlier when computing PIC addresses

This change partially reverts https://reviews.llvm.org/D54647 in favor
of bailing out during computeAddress instead.

This catches the condition earlier and handles more cases.

Differential Revision: https://reviews.llvm.org/D60986

llvm-svn: 358948

5 years ago[analyzer] Unbreak body farms in presence of multiple declarations.
Artem Dergachev [Tue, 23 Apr 2019 02:56:00 +0000 (02:56 +0000)]
[analyzer] Unbreak body farms in presence of multiple declarations.

When growing a body on a body farm, it's essential to use the same redeclaration
of the function that's going to be used during analysis. Otherwise our
ParmVarDecls won't match the ones that are used to identify argument regions.

This boils down to trusting the reasoning in AnalysisDeclContext. We shouldn't
canonicalize the declaration before farming the body because it makes us not
obey the sophisticated decision-making process of AnalysisDeclContext.

Differential Revision: https://reviews.llvm.org/D60899

llvm-svn: 358946

5 years ago[analyzer] PR41335: Fix crash when no-store event is in a body-farmed function.
Artem Dergachev [Tue, 23 Apr 2019 02:50:38 +0000 (02:50 +0000)]
[analyzer] PR41335: Fix crash when no-store event is in a body-farmed function.

Stuffing invalid source locations (such as those in functions produced by
body farms) into path diagnostics causes crashes.

Fix a typo in a nearby function name.

Differential Revision: https://reviews.llvm.org/D60808

llvm-svn: 358945

5 years ago[analyzer] PR41269: Add a bit of C++ smart pointer modeling.
Artem Dergachev [Tue, 23 Apr 2019 02:45:42 +0000 (02:45 +0000)]
[analyzer] PR41269: Add a bit of C++ smart pointer modeling.

Implement cplusplus.SmartPtrModeling, a new checker that doesn't
emit any warnings but models methods of smart pointers more precisely.

For now the only thing it does is make `(bool) P` return false when `P`
is a freshly moved pointer. This addresses a false positive in the
use-after-move-checker.

Differential Revision: https://reviews.llvm.org/D60796

llvm-svn: 358944

5 years agoUse llvm::stable_sort
Fangrui Song [Tue, 23 Apr 2019 02:42:06 +0000 (02:42 +0000)]
Use llvm::stable_sort

Make some small adjustment while touching the code: make parameters
const, use less_first(), etc.

Differential Revision: https://reviews.llvm.org/D60989

llvm-svn: 358943

5 years agoadd Qiu Chaofan (qiucf@cn.ibm.com) to the CREDITS.txt
Qiu Chaofan [Tue, 23 Apr 2019 02:37:48 +0000 (02:37 +0000)]
add Qiu Chaofan (qiucf@cn.ibm.com) to the CREDITS.txt

llvm-svn: 358942

5 years agoRevert "Use const DebugLoc&"
Chandler Carruth [Tue, 23 Apr 2019 01:42:07 +0000 (01:42 +0000)]
Revert "Use const DebugLoc&"

This reverts r358910 (git commit 2b744665308fc8d30a3baecb4947f2bd81aa7d30)

While this patch *seems* trivial and safe and correct, it is not. The
copies are actually load bearing copies. You can observe this with MSan
or other ways of checking for use-after-destroy, but otherwise this may
result in ... difficult to debug inexplicable behavior.

I suspect the issue is that the debug location is used after the
original reference to it is removed. The metadata backing it gets
destroyed as its last references goes away, and then we reference it
later through these const references.

llvm-svn: 358940

5 years agoOne small tweak to LocateExecutableScriptingResources - I
Jason Molenda [Tue, 23 Apr 2019 01:13:27 +0000 (01:13 +0000)]
One small tweak to LocateExecutableScriptingResources - I
was still stat'ing the possibly-dSYM FileSpec before I
(more cheaply) checked the filepath for telltale dSYM
components.
<rdar://problem/50086007>

llvm-svn: 358939

5 years agoAdd a small check to PlatformDarwin::LoadScriptingResourceForModule
Jason Molenda [Tue, 23 Apr 2019 01:02:51 +0000 (01:02 +0000)]
Add a small check to PlatformDarwin::LoadScriptingResourceForModule
which reads the python files in a dSYM bundle, to check that the
SymbolFile is actually a dSYM bundle filepath; delay any fetching
of the ScriptInterpreter until after we've done that check.

When debugging a binary without a dSYM on darwin systems, the
SymbolFile we fetch is actually the ObjectFile -- so we would do
an unnecessary trip into Python land and stat around the filesystem
looking for a python file to read in.  There's no reason to do any
of this unless the SymbolFile's file path includes the .dSYM bundle
telltale path components.

<rdar://problem/50065315>

llvm-svn: 358938

5 years ago[libc++abi] Don't use a .sh.cpp test for uncaught_exception
Louis Dionne [Tue, 23 Apr 2019 00:03:34 +0000 (00:03 +0000)]
[libc++abi] Don't use a .sh.cpp test for uncaught_exception

Otherwise, we don't seem to get the DYLD_LIBRARY_PATH set up correctly
and the tests are run against the system libc++abi dylib.

llvm-svn: 358937

5 years ago[CMake] Replace the sanitizer support in runtimes build with multilib
Petr Hosek [Mon, 22 Apr 2019 23:31:39 +0000 (23:31 +0000)]
[CMake] Replace the sanitizer support in runtimes build with multilib

This is a more generic solution; while the sanitizer support can be used
only for sanitizer instrumented builds, the multilib support can be used
to build other variants such as noexcept which is what we would like to use
in Fuchsia.

The name CMake target name uses the target name, same as for the regular
runtimes build and the name of the multilib, concatenated with '+'. The
libraries are installed in a subdirectory named after the multilib.

Differential Revision: https://reviews.llvm.org/D60926

llvm-svn: 358935

5 years agoFully qualify llvm::Optional, some compilers complain otherwise.
Adrian Prantl [Mon, 22 Apr 2019 22:51:34 +0000 (22:51 +0000)]
Fully qualify llvm::Optional, some compilers complain otherwise.

llvm-svn: 358933

5 years ago[c++2a] Implement semantic restrictions for 'export' declarations.
Richard Smith [Mon, 22 Apr 2019 22:50:11 +0000 (22:50 +0000)]
[c++2a] Implement semantic restrictions for 'export' declarations.

llvm-svn: 358932

5 years agoDebugInfo: Emit only one kind of accelerated access/name table
David Blaikie [Mon, 22 Apr 2019 22:45:11 +0000 (22:45 +0000)]
DebugInfo: Emit only one kind of accelerated access/name table

Currently to opt in to debug_names in DWARFv5, the IR must contain
'nameTableKind: Default' which also enables debug_pubnames.

Instead, only allow one of {debug_names, apple_names, debug_pubnames,
debug_gnu_pubnames}.

nameTableKind: Default gives debug_names in DWARFv5 and greater,
debug_pubnames in v4 and earlier - and apple_names when tuning for lldb
on MachO.
nameTableKind: GNU always gives gnu_pubnames

llvm-svn: 358931

5 years ago[SelectionDAG] move splat util functions up from x86 lowering
Sanjay Patel [Mon, 22 Apr 2019 22:43:36 +0000 (22:43 +0000)]
[SelectionDAG] move splat util functions up from x86 lowering

This was supposed to be NFC, but the change in SDLoc
definitions causes instruction scheduling changes.

There's nothing x86-specific in this code, and it can
likely be used from DAGCombiner's simplifyVBinOp().

llvm-svn: 358930

5 years agoFix a bug in my change to ModulesDidLoad in r357955.
Jason Molenda [Mon, 22 Apr 2019 22:42:29 +0000 (22:42 +0000)]
Fix a bug in my change to ModulesDidLoad in r357955.
In the process of hoisting the LoadScriptingResourceForModule
out of Target::ModuleAdded and into Target::ModulesDidLoad,
I had ModulesDidLoad fetching the Target's entire image list
and look for scripting resources in those -- instead of only
looking for scripting resources in the modules that had
been added to the target's image list.

<rdar://problem/50065315>

llvm-svn: 358929

5 years ago[Docs] Move API docs to the front page
Jonas Devlieghere [Mon, 22 Apr 2019 22:41:55 +0000 (22:41 +0000)]
[Docs] Move API docs to the front page

This moves the links to the C++ and Python API docs up to the main page.
As of now the links are still broken [1], but at least this will prevent
the additional frustration of searching for the links only to find out
they're broken.

[1] http://lists.llvm.org/pipermail/lldb-dev/2019-April/014992.html

llvm-svn: 358928

5 years agoTry to work around compile errors with older versions of GCC.
Adrian Prantl [Mon, 22 Apr 2019 22:40:37 +0000 (22:40 +0000)]
Try to work around compile errors with older versions of GCC.

llvm-svn: 358927

5 years agoRelax test to check for a valid number instead of a specific number.
Douglas Yung [Mon, 22 Apr 2019 22:31:57 +0000 (22:31 +0000)]
Relax test to check for a valid number instead of a specific number.

llvm-svn: 358926

5 years agoRename C++ TestGlobalVariables.py to have a distinct name from C version.
Adrian Prantl [Mon, 22 Apr 2019 22:16:33 +0000 (22:16 +0000)]
Rename C++ TestGlobalVariables.py to have a distinct name from C version.

llvm-svn: 358924

5 years ago[AMDGPU] Fix an issue in `op_sel_hi` skipping.
Michael Liao [Mon, 22 Apr 2019 22:05:49 +0000 (22:05 +0000)]
[AMDGPU] Fix an issue in `op_sel_hi` skipping.

Summary:
- Only apply packed literal `op_sel_hi` skipping on operands requiring
  packed literals. Even an instruction is `packed`, it may have operand
  requiring non-packed literal, such as `v_dot2_f32_f16`.

Reviewers: rampitec, arsenm, kzhuravl

Subscribers: jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60978

llvm-svn: 358922

5 years ago[dsymutil] Collect parseable Swift interfaces in the .dSYM bundle.
Adrian Prantl [Mon, 22 Apr 2019 21:33:22 +0000 (21:33 +0000)]
[dsymutil] Collect parseable Swift interfaces in the .dSYM bundle.

When a Swift module built with debug info imports a library without
debug info from a textual interface, the textual interface is
necessary to reconstruct types defined in the library's interface. By
recording the Swift interface files in DWARF dsymutil can collect them
and LLDB can find them.

This patch teaches dsymutil to look for DW_TAG_imported_modules and
records all references to parseable Swift ingterfrace files and copies
them to

  a.out.dSYM/Contents/Resources/<Arch>/<ModuleName>.swiftinterface

<rdar://problem/49751748>

llvm-svn: 358921

5 years ago[InstCombine] Eliminate stores to constant memory
Philip Reames [Mon, 22 Apr 2019 20:28:19 +0000 (20:28 +0000)]
[InstCombine] Eliminate stores to constant memory

If we have a store to a piece of memory which is known constant, then we know the store must be storing back the same value. As a result, the store (or memset, or memmove) must either be down a dead path, or a noop. In either case, it is valid to simply remove the store.

The motivating case for this involves a memmove to a buffer which is constant down a path which is dynamically dead.

Note that I'm choosing to implement the less aggressive of two possible semantics here. We could simply say that the store *is undefined*, and prune the path. Consensus in the review was that the more aggressive form might be a good follow on change at a later date.

Differential Revision: https://reviews.llvm.org/D60659

llvm-svn: 358919

5 years ago[EditLineTest] Not always TERM is available, e.g. on some bots.
Davide Italiano [Mon, 22 Apr 2019 20:27:10 +0000 (20:27 +0000)]
[EditLineTest] Not always TERM is available, e.g. on some bots.

llvm-svn: 358918

5 years ago[VerifyDiagnosticConsumer] Document -verify=<prefixes> in doxygen
Joel E. Denny [Mon, 22 Apr 2019 20:25:06 +0000 (20:25 +0000)]
[VerifyDiagnosticConsumer] Document -verify=<prefixes> in doxygen

Previously, it was only documented by `-cc1 -help`, so people weren't
aware of it, as discussed in D60732.

Reviewed By: Charusso, NoQ

Differential Revision: https://reviews.llvm.org/D60845

llvm-svn: 358917

5 years ago[Reproducers] Fix lifetime issue
Jonas Devlieghere [Mon, 22 Apr 2019 20:05:02 +0000 (20:05 +0000)]
[Reproducers] Fix lifetime issue

Deallocating the data recorder in during the ::Keep() operation causes
problems down the line when exiting the debugger. The command
interpreter still holds a pointer to the now deallocated object and has
no way to know it no longer exists. This is exactly what the m_record
flag was meant for, although it wasn't hooked up properly either.

llvm-svn: 358916

5 years ago[libcxx] Update gen_link_script.py to support different input and output
Petr Hosek [Mon, 22 Apr 2019 19:51:33 +0000 (19:51 +0000)]
[libcxx] Update gen_link_script.py to support different input and output

This enables the use of this script from other build systems like
GN which don't support post-build actions as well as for static
archives.

Differential Revision: https://reviews.llvm.org/D60309

llvm-svn: 358915

5 years ago[Support] unflake TempFileCollisions test
Bob Haarman [Mon, 22 Apr 2019 19:46:25 +0000 (19:46 +0000)]
[Support] unflake TempFileCollisions test

Summary:
This test was added to verify that createUniqueEntity() does
not enter an infinite loop when all possible names are taken. However,
it also checked that all possible names are generated, which is flaky
(because the names are generated randomly). This change increases the
number of attempts we make to make flakes exceedingly
unlikely (3.88e-62).

Reviewers: fedor.sergeev, rsmith

Reviewed By: fedor.sergeev

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D56336

llvm-svn: 358914

5 years ago[InstSimplify] Move masked.gather w/no active lanes handling to InstSimplify from...
Philip Reames [Mon, 22 Apr 2019 19:30:01 +0000 (19:30 +0000)]
[InstSimplify] Move masked.gather w/no active lanes handling to InstSimplify from InstCombine

In the process, use the existing masked.load combine which is slightly stronger, and handles a mix of zero and undef elements in the mask.

llvm-svn: 358913

5 years agogn build: Merge r358869
Nico Weber [Mon, 22 Apr 2019 19:25:40 +0000 (19:25 +0000)]
gn build: Merge r358869

llvm-svn: 358912

5 years agoUse correct way to test for MIPS arch after rOMP355687
Dimitry Andric [Mon, 22 Apr 2019 19:20:46 +0000 (19:20 +0000)]
Use correct way to test for MIPS arch after rOMP355687

Summary:
I ran into some issues after rOMP355687, where __atomic_fetch_add was
being used incorrectly on x86, and this turns out to be caused by the
following added conditionals:

```
#if defined(KMP_ARCH_MIPS)
```

The problem is, these macros are always defined, and are either 0 or 1
depending on the architecture.  E.g. the correct way to test for MIPS
is:

```
#if KMP_ARCH_MIPS
```

Reviewers: petarj, jlpeyton, Hahnfeld, AndreyChurbanov

Reviewed By: petarj, AndreyChurbanov

Subscribers: AndreyChurbanov, sdardis, arichardson, atanasyan, jfb, jdoerfert, openmp-commits, llvm-commits

Tags: #openmp

Differential Revision: https://reviews.llvm.org/D60938

llvm-svn: 358911

5 years agoUse const DebugLoc&
Matt Arsenault [Mon, 22 Apr 2019 19:14:27 +0000 (19:14 +0000)]
Use const DebugLoc&

llvm-svn: 358910

5 years agoAMDGPU: Skip debug instructions in assert
Matt Arsenault [Mon, 22 Apr 2019 19:14:26 +0000 (19:14 +0000)]
AMDGPU: Skip debug instructions in assert

These are inserted after branch relaxation, and for some reason it's
decided to put them in the long branch expansion block. It's probably
not great to rely on the source block address, so this should probably
be switched to being PC relative instead of relying on the block
address

llvm-svn: 358909

5 years ago[libc++][test] Update some wstring_convert tests for MSVC quirks
Casey Carter [Mon, 22 Apr 2019 19:08:13 +0000 (19:08 +0000)]
[libc++][test] Update some wstring_convert tests for MSVC quirks

Due to MSVC's decision to encode `wchar_t` as UTF-16, it rejects wide
character/string literals that expect a character value greater than
`\xffff`. UTF-16 `wchar_t` is clearly non-conforming, given that the
standard requires wchar_t to be capable of representing all characters
in the supported wide character execution sets, but rejecting e.g.
`\x40003` is a reasonably sane compromise given that encoding choice:
there's an expectation that `\xFOO` produces a single character in the
resulting literal. Consequently `L'\x40003'`/`L"\x40003"` are ill-formed
literals on MSVC. `L'\U00040003'` is a high surrogate (and produces a
warning about ignoring the "second character" in a multi-character
literal), and `L"\U00040003"` is a perfectly-valid `const wchar_t[3]`.

This change updates these tests to use universal-character-names instead
of raw values for the intended character values, which technically makes
them portable even to implementations that don't use a unicode
transformation format encoding for their wide character execution
character set. The two-character literal `L"\u1005e"` is awkward - the
`e` looks like part of the UCN's hex encoding - but necessary to compile
in '03 mode since '03 didn't allow UCNs to be used for members of the
basic execution character set even in character/string literals.

I've also eliminated the extraneous `\x00` "bonus null-terminator" in
some of the string literals which doesn't affect the tested behavior.

I'm sorry about using `*L"\U00040003"` in `conversions.string/to_bytes.pass.cpp`,
but it's correct for platforms with 32-bit wchar_t, *and* doesn't
trigger narrowing warnings as did the prior `CharT(0x40003)`.

Differential Revision: https://reviews.llvm.org/D60950

llvm-svn: 358908

5 years ago[Tests] Revise a test as requested by reviewer in D59703
Philip Reames [Mon, 22 Apr 2019 18:51:58 +0000 (18:51 +0000)]
[Tests] Revise a test as requested by reviewer in D59703

llvm-svn: 358907

5 years ago[Tests] Add a negative test for masked.gather part of D59703
Philip Reames [Mon, 22 Apr 2019 18:28:44 +0000 (18:28 +0000)]
[Tests] Add a negative test for masked.gather part of D59703

llvm-svn: 358906

5 years ago[ASTMatchers] Introduce Objective-C matchers `isClassMessage`, `isClassMethod`, and...
Ben Hamilton [Mon, 22 Apr 2019 17:54:11 +0000 (17:54 +0000)]
[ASTMatchers] Introduce Objective-C matchers `isClassMessage`, `isClassMethod`, and `isInstanceMethod`

Summary:
isClassMessage is an equivalent to isInstanceMessage for ObjCMessageExpr, but matches message expressions to classes.

isClassMethod and isInstanceMethod check whether a method declaration (or definition) is for a class method or instance method (respectively).

Contributed by @mywman!

Reviewers: benhamilton, klimek, mwyman

Reviewed By: benhamilton, mwyman

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D60920

llvm-svn: 358904

5 years ago[IPSCCP] Add missing `AssumptionCacheTracker` dependency
Justin Bogner [Mon, 22 Apr 2019 17:38:29 +0000 (17:38 +0000)]
[IPSCCP] Add missing `AssumptionCacheTracker` dependency

Back in August, r340525 introduced a dependency on the assumption
cache tracker in the ipsccp pass, but that commit missed a call to
INITIALIZE_PASS_DEPENDENCY, which leaves the assumption cache
improperly registered if SCCP is the only thing that pulls it in.

llvm-svn: 358903

5 years ago[OPENMP][NVPTX] Fix the test, NFC.
Alexey Bataev [Mon, 22 Apr 2019 17:25:31 +0000 (17:25 +0000)]
[OPENMP][NVPTX] Fix the test, NFC.

Fix the test to run it really in SPMD mode without runtime. Previously
it was run in SPMD + full runtime mode and does not allow to cehck the
functionality correctly.

llvm-svn: 358902

5 years ago[LPM/BPI] Preserve BPI through trivial loop pass pipeline (e.g. LCSSA, LoopSimplify)
Philip Reames [Mon, 22 Apr 2019 17:13:43 +0000 (17:13 +0000)]
[LPM/BPI] Preserve BPI through trivial loop pass pipeline (e.g. LCSSA, LoopSimplify)

Currently, we do not expose BPI to loop passes at all. In the old pass manager, we appear to have been ignoring the fact that LCSSA and/or LoopSimplify didn't preserve BPI, and making it available to the following loop passes anyways.  In the new one, it's invalidated before running any loop pass if either LCSSA or LoopSimplify actually make changes. If they don't make changes, then BPI is valid and available.  So, we go ahead and teach LCSSA and LoopSimplify how to preserve BPI for consistency between old and new pass managers.

This patch avoids an invalidation between the two requires in the following trivial pass pipeline:
opt -passes="requires<branch-prob>,loop(no-op-loop),requires<branch-prob>"
(when the input file is one which requires either LCSSA or LoopSimplify to canonicalize the loops)

Differential Revision: https://reviews.llvm.org/D60790

llvm-svn: 358901

5 years ago[PGO/SamplePGO][NFC] Move the function updateProfWeight from Instruction
Wei Mi [Mon, 22 Apr 2019 17:04:51 +0000 (17:04 +0000)]
[PGO/SamplePGO][NFC] Move the function updateProfWeight from Instruction
to CallInst.

The issue was raised here: https://reviews.llvm.org/D60903#1472783

The function Instruction::updateProfWeight is only used for CallInst in
profile update. From the current interface, it is very easy to think that
the function can also be used for branch instruction. However, Branch
instruction does't need the scaling the function provides for
branch_weights and VP (value profile), in addition, scaling may introduce
inaccuracy for branch probablity.

The patch moves the function updateProfWeight from Instruction class to
CallInst to remove the confusion. The patch also changes the scaling of
branch_weights from a loop to a block because we know that ProfileData
for branch_weights of CallInst will only have two operands at most.

Differential Revision: https://reviews.llvm.org/D60911

llvm-svn: 358900

5 years ago[WebAssembly] Error on relocations against undefined data symbols.
Sam Clegg [Mon, 22 Apr 2019 16:12:54 +0000 (16:12 +0000)]
[WebAssembly] Error on relocations against undefined data symbols.

We can't (currently) meaningfully resolve certain types of relocations
against undefined data symbols.  Previously when `--allow-undefined` was
used we were treating such relocation much like weak data symbols and
simply inserting zeros.  This change turns such use cases in to an
error.

This means that `--allow-undefined` is no longer effective for data
symbols.

Fixes https://bugs.llvm.org/show_bug.cgi?id=40364

Differential Revision: https://reviews.llvm.org/D60882

llvm-svn: 358899

5 years ago[sema][objc] Minor refactor to OverrideSearch. NFCI.
Matt Davis [Mon, 22 Apr 2019 16:04:44 +0000 (16:04 +0000)]
[sema][objc] Minor refactor to OverrideSearch. NFCI.

Summary:
* Removed a member that was only used during construction.
* Use range-based for iteration when accessing the result of the search.
* Require an `ObjCMethodDecl` reference upon construction of an
* Constify.

Reviewers: rjmccall

Reviewed By: rjmccall

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D60850

llvm-svn: 358898

5 years agoUse llvm::stable_sort. NFC
Fangrui Song [Mon, 22 Apr 2019 15:53:43 +0000 (15:53 +0000)]
Use llvm::stable_sort. NFC

llvm-svn: 358897

5 years ago[NFC] Fix typo in debug log
Louis Dionne [Mon, 22 Apr 2019 15:40:50 +0000 (15:40 +0000)]
[NFC] Fix typo in debug log

llvm-svn: 358896

5 years agoRemove spurious semicolons; NFC.
Aaron Ballman [Mon, 22 Apr 2019 15:31:09 +0000 (15:31 +0000)]
Remove spurious semicolons; NFC.

llvm-svn: 358895

5 years agoAMDGPU/GlobalISel: Fix non-power-of-2 G_EXTRACT sources
Matt Arsenault [Mon, 22 Apr 2019 15:22:46 +0000 (15:22 +0000)]
AMDGPU/GlobalISel: Fix non-power-of-2 G_EXTRACT sources

llvm-svn: 358894

5 years agoSTLExtras: add stable_sort wrappers
Fangrui Song [Mon, 22 Apr 2019 15:19:13 +0000 (15:19 +0000)]
STLExtras: add stable_sort wrappers

llvm-svn: 358893

5 years agoGlobalISel: Legalize scalar G_EXTRACT sources
Matt Arsenault [Mon, 22 Apr 2019 15:10:42 +0000 (15:10 +0000)]
GlobalISel: Legalize scalar G_EXTRACT sources

llvm-svn: 358892

5 years agollvm-undname: Fix an assert-on-invalid, found by oss-fuzz
Nico Weber [Mon, 22 Apr 2019 15:05:18 +0000 (15:05 +0000)]
llvm-undname: Fix an assert-on-invalid, found by oss-fuzz

llvm-svn: 358891

5 years agoAMDGPU: Fix not checking for copy when looking at copy src
Matt Arsenault [Mon, 22 Apr 2019 14:54:39 +0000 (14:54 +0000)]
AMDGPU: Fix not checking for copy when looking at copy src

Effectively reverts r356956. The check for isFullCopy was excessive,
but there still needs to be a check that this is a copy.

llvm-svn: 358890

5 years ago[LLD][ELF] - Remove a binary from the inputs. NFCI.
George Rimar [Mon, 22 Apr 2019 14:53:32 +0000 (14:53 +0000)]
[LLD][ELF] - Remove a binary from the inputs. NFCI.

section-index.elf was removed and the corresponding test
was replaced with a yaml2obj based test.

llvm-svn: 358889

5 years ago[AMDGPU][MC] Corrected parsing of SP3 'neg' modifier
Dmitry Preobrazhensky [Mon, 22 Apr 2019 14:35:47 +0000 (14:35 +0000)]
[AMDGPU][MC] Corrected parsing of SP3 'neg' modifier

See bug 41156: https://bugs.llvm.org/show_bug.cgi?id=41156

Reviewers: artem.tamazov, arsenm

Differential Revision: https://reviews.llvm.org/D60624

llvm-svn: 358888

5 years ago[TargetLowering][AMDGPU][X86] Improve SimplifyDemandedBits bitcast handling
Simon Pilgrim [Mon, 22 Apr 2019 14:04:35 +0000 (14:04 +0000)]
[TargetLowering][AMDGPU][X86] Improve SimplifyDemandedBits bitcast handling

This patch adds support for BigBitWidth -> SmallBitWidth bitcasts, splitting the DemandedBits/Elts accordingly.

The AMDGPU backend needed an extra  (srl (and x, c1 << c2), c2) -> (and (srl(x, c2), c1) combine to encourage BFE creation, I investigated putting this in DAGCombine but it caused a lot of noise on other targets - some improvements, some regressions.

The X86 changes are all definite wins.

Differential Revision: https://reviews.llvm.org/D60462

llvm-svn: 358887

5 years ago[DAGCombiner] make variable name less ambiguous; NFC
Sanjay Patel [Mon, 22 Apr 2019 13:42:50 +0000 (13:42 +0000)]
[DAGCombiner] make variable name less ambiguous; NFC

llvm-svn: 358886

5 years ago[LLD][ELF] - Do not forget to use ch_addralign field after decompressing the sections.
George Rimar [Mon, 22 Apr 2019 13:40:42 +0000 (13:40 +0000)]
[LLD][ELF] - Do not forget to use ch_addralign field after decompressing the sections.

LLD did not use ELF::Chdr::ch_addralign for decompressed sections.
This resulted in a broken output.

Fixes https://bugs.llvm.org/show_bug.cgi?id=40482.

Differential revision: https://reviews.llvm.org/D60959

llvm-svn: 358885