1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-24 03:33:20 +01:00
Commit Graph

139211 Commits

Author SHA1 Message Date
Quentin Colombet
f515b3c80c [AArch64][InstructionSelector] Teach how to select FP load/store.
This patch allows to select 32 and 64-bit FP load and store.

llvm-svn: 283832
2016-10-11 00:21:14 +00:00
Quentin Colombet
4a90fcde10 [AArch64][InstructionSelector] Teach the selector how to handle vector OR.
This only adds the support for 64-bit vector OR. Adding more sizes is
not difficult, but it requires a bigger refactoring because ORs work on
any size, not necessarly the ones that match the width of the register
width. Right now, this is not expressed in the legalization, so don't
bother pushing the refactoring yet.

llvm-svn: 283831
2016-10-11 00:21:11 +00:00
Quentin Colombet
453627f3cf [AArch64][MachineLegalizer] Mark v2s32 G_LOAD as legal.
Actually every 64-bit loads are legal, but right now the API does not
offer a simple way to express that.

llvm-svn: 283829
2016-10-11 00:21:08 +00:00
Rui Ueyama
96b98f169a Revert r283824 and r283823: Define DbiStreamBuilder::addDbgStream to add stream.
This reverts commit r283824 and r283823 to fix buildbots.

llvm-svn: 283828
2016-10-11 00:15:50 +00:00
Rui Ueyama
be776e9ace Fix a bug in DbiStreamBuilder::addDbgStream.
This feature will be tested in LLD unit tests.

llvm-svn: 283824
2016-10-10 23:44:04 +00:00
Rui Ueyama
b3845e5ad0 Define DbiStreamBuilder::addDbgStream to add stream.
Previously, there is no way to create a stream other than pre-defined
special stream such as DBI or IPI. This patch adds a new method,
addDbgStream, to add a debug stream to a PDB file.

Differential Revision: https://reviews.llvm.org/D25356

llvm-svn: 283823
2016-10-10 23:35:36 +00:00
Chris Matthews
2fa8e84ed6 Fix issue which cases lit installed with setup.py to not resolve main
llvm-svn: 283818
2016-10-10 23:22:11 +00:00
NAKAMURA Takumi
f7d2c2c56c Fix llvm-lit.in corresponding to r283710.
Traceback (most recent call last):
    File "bin/llvm-lit", line 44, in <module>
      lit.main(builtin_parameters)
  AttributeError: 'module' object has no attribute 'main'

Suggested by Artem Belevich.

llvm-svn: 283816
2016-10-10 23:02:42 +00:00
Peter Collingbourne
861bb221e9 Revert r283690, "MC: Remove unused entities."
llvm-svn: 283814
2016-10-10 22:49:37 +00:00
Sanjay Patel
c75ab18833 [x86] auto-generate checks
llvm-svn: 283812
2016-10-10 22:04:12 +00:00
Sanjay Patel
e0b0718d9e [x86] auto-generate checks
llvm-svn: 283811
2016-10-10 22:01:42 +00:00
Tim Northover
7cb88e7053 GlobalISel: select G_GLOBAL_VALUE uses on AArch64.
llvm-svn: 283809
2016-10-10 21:50:00 +00:00
Tim Northover
94d146decd GlobalISel: allow G_GLOBAL_VALUEs in AArch64 legalization.
llvm-svn: 283808
2016-10-10 21:49:53 +00:00
Tim Northover
de665d6f28 GlobalISel: support selecting G_GEP instructions.
They're basically just an alias for G_ADD on AArch64.

llvm-svn: 283807
2016-10-10 21:49:49 +00:00
Tim Northover
c4bdf87acf GlobalISel: support selecting constants on AArch64.
llvm-svn: 283806
2016-10-10 21:49:42 +00:00
Dehao Chen
f2fd9e1ece Rename isHotFunction/isColdFunction to isFunctionEntryHot/isFunctionEntryCold. (NFC)
This is in preparation for https://reviews.llvm.org/D25048

llvm-svn: 283805
2016-10-10 21:47:28 +00:00
Zachary Turner
14d2c535cb Revert "Disallow ArrayRef assignment from temporaries."
This reverts commit r283798, as it causes static asserts on
MSVC 2015 with the following errors:

ArrayRefTest.cpp(38): error C2338: Assigning from single prvalue element
ArrayRefTest.cpp(41): error C2338: Assigning from single xvalue element
ArrayRefTest.cpp(47): error C2338: Assigning from an initializer list

llvm-svn: 283803
2016-10-10 21:36:23 +00:00
Zachary Turner
80d8c682d7 Rename llvm::apply -> llvm::apply_tuple.
llvm::cl already has a function called llvm::apply() so this is
causing an ODR violation.  The STLExtras version should win the
vote on which one gets to be called apply() since it is named
after the equivalent STL function, but since renaiming the cl
version is more difficult, let's do this for now to get the
bots green.

llvm-svn: 283800
2016-10-10 21:24:34 +00:00
Jordan Rose
b52b42ffac Disallow ArrayRef assignment from temporaries.
Without this, the following statements will create ArrayRefs that
refer to temporary storage that goes out of scope by the end of the
line:

  someArrayRef = getSingleElement();
  someArrayRef = {elem1, elem2};

Note that the constructor still has this problem:

  ArrayRef<Element> someArrayRef = getSingleElement();
  ArrayRef<Element> someArrayRef = {elem1, elem2};

but that's a little harder to get rid of because we want to be able to
use this in calls:

  takesArrayRef(getSingleElement());
  takesArrayRef({elem1, elem2});

Part of rdar://problem/16375365. Reviewed by Duncan Exon Smith.

llvm-svn: 283798
2016-10-10 20:57:33 +00:00
Hal Finkel
621305e718 [SelectionDAGBuilder] Support llvm.flt.rounds on targets where i32 is not legal
Add integer expansion for FLT_ROUNDS_ for targets where i32 is not a legal
type.

Patch by Edward Jones, thanks!

Differential Revision: https://reviews.llvm.org/D24459

llvm-svn: 283797
2016-10-10 20:45:15 +00:00
Justin Lebar
f2afb57e75 [ADT] Use () instead of {} in an attempt to work around MSVC 2012 ICEs.
llvm-svn: 283796
2016-10-10 20:18:02 +00:00
Justin Lebar
96d7bc1342 [ADT] Don't use make_pointee_iterator in IteratorTest.
llvm-svn: 283794
2016-10-10 19:56:52 +00:00
Mehdi Amini
3003721ba0 Use StringRef in TableGen generated Intrinsics.gen file (NFC)
llvm-svn: 283792
2016-10-10 19:31:09 +00:00
Justin Lebar
14c600c960 [ADT] Remove make_pointe{e,r}_iterator, because it seems to crash MSVC 2015.
llvm-svn: 283791
2016-10-10 19:29:37 +00:00
Adrian Prantl
95da712509 Teach llvm::StripDebugInfo() about global variable !dbg attachments.
This is a regression introduced by the global variable ownership
reversal performed in r281284.

rdar://problem/28448075

llvm-svn: 283784
2016-10-10 17:53:33 +00:00
Justin Lebar
86b122fcd7 [ADT] Attempt to fix MSVC 2015 ICE via judicious addition of std::decay to make_pointe{r,e}_iterator.
llvm-svn: 283783
2016-10-10 17:18:45 +00:00
Mehdi Amini
f29c5dbf50 Update documentation after r283671 ("Turn cl::values() (for enum) from a vararg function to using C++ variadic template")
llvm-svn: 283782
2016-10-10 17:13:14 +00:00
Zachary Turner
4022c2b021 Add llvm::apply to STLExtras.
This is equivalent to the C++14 std::apply().  Since we are not
using C++14 yet, this allows us to still make use of apply anyway.

Differential revision: https://reviews.llvm.org/D25100

llvm-svn: 283779
2016-10-10 16:44:09 +00:00
Justin Lebar
408de1dad4 Use unique_ptr in LLVMContextImpl's constant maps.
Reviewers: timshen

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D25419

llvm-svn: 283767
2016-10-10 16:26:13 +00:00
Justin Lebar
02ac32407b [ADT] Add make_pointe{e,r}_iterator.
Reviewers: timshen

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D25418

llvm-svn: 283765
2016-10-10 16:26:03 +00:00
Justin Lebar
8960a1706e [ADT] Let MapVector handle non-copyable values.
Summary: The keys must still be copyable, because we store two copies of them.

Reviewers: timshen

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D25404

llvm-svn: 283764
2016-10-10 16:25:59 +00:00
Alexandros Lamprineas
7a09ca4165 [ARM] Fix invalid VLDM/VSTM access when targeting Big Endian with NEON
The instructions VLDM/VSTM can only access word-aligned memory
locations and produce alignment fault if the condition is not met.

The compiler currently generates VLDM/VSTM for v2f64 load/store
regardless the alignment of the memory access. Instead, if a v2f64
load/store is not word-aligned, the compiler should generate
VLD1/VST1. For each non double-word-aligned VLD1/VST1, a VREV
instruction should be generated when targeting Big Endian.

Differential Revision: https://reviews.llvm.org/D25281

llvm-svn: 283763
2016-10-10 16:01:54 +00:00
Nirav Dave
a67f7f63e3 Add return type for checkForValidSection parsing function. NFC Intended.
llvm-svn: 283761
2016-10-10 15:24:54 +00:00
Zvi Rackover
8755b41ec0 [X86] Prefer rotate by 1 over rotate by imm
Summary:
Rotate by 1 is translated to 1 micro-op, while rotate with imm8 is translated to 2 micro-ops.

Fixes pr30644.

Reviewers: delena, igorb, craig.topper, spatel, RKSimon

Differential Revision: https://reviews.llvm.org/D25399

llvm-svn: 283758
2016-10-10 14:43:55 +00:00
Simon Pilgrim
4df8e37155 [SLPVectorizer][X86] Add 512-bit sitofp/uitofp tests
llvm-svn: 283756
2016-10-10 14:28:06 +00:00
Simon Pilgrim
1dbb87da59 [SLPVectorizer][X86] Add avx512 sitofp/uitofp tests
llvm-svn: 283751
2016-10-10 14:14:31 +00:00
Simon Pilgrim
8047ad9428 [SLPVectorizer][X86] Fixed alignments of scalar loads in sitofp/uitofp tests
Fixed copy+paste vector alignment to correct for per-element scalar loads

Increased to 512-bit data sizes in preparation of avx512 tests

llvm-svn: 283748
2016-10-10 14:10:41 +00:00
Simon Pilgrim
d3fc0ef15e Fixed windows stdout/stderr redirection in inline asm constraint tests
llvm-svn: 283741
2016-10-10 11:11:27 +00:00
George Rimar
dbc4420f77 [Object/ELF] - Do not crash on invalid Header->e_shoff value.
sections_begin() may return unalignment pointer when Header->e_shoff isinvalid.
That may result in a crash in clients, for example we have one in LLD:

assert((PtrWord & ~PointerBitMask) == 0 &&
       "Pointer is not sufficiently aligned");
fails when trying to push_back Elf_Shdr* (unaligned) into TinyPtrVector.

Patch forces check for alignment of Header->e_shoff.

Differential revision: https://reviews.llvm.org/D25368

llvm-svn: 283740
2016-10-10 10:51:38 +00:00
Chris Dewhurst
17365ee380 This pass, fixing an erratum in some LEON 2 processors ensures that the SDIV instruction is not issued, but replaced by SDIVcc instead, which does not exhibit the error. Unit test included.
Differential Review: https://reviews.llvm.org/D24660

llvm-svn: 283727
2016-10-10 08:53:06 +00:00
Daniel Jasper
0d0a9b6265 Fix WebAssembly build after r283702.
llvm-svn: 283723
2016-10-10 06:49:55 +00:00
Craig Topper
47e80c7323 [AVX-512] Add missing pattern sext or zext from bytes to quad words with a 128-bit load as input.
llvm-svn: 283720
2016-10-10 06:25:48 +00:00
Craig Topper
af87c866dd [AVX-512] Add test cases for AVX512 sign/zero extend instructions derived from the sse41 and avx2 test cases. Code will be improved in future commits.
llvm-svn: 283719
2016-10-10 06:25:45 +00:00
Craig Topper
2038c8e9eb [AVX-512] Add an AVX512VL/BW command line to sse41-pmovxrm.ll and avx2-pmovxrm.ll. Also disable peephole so we really test pattern matching.
llvm-svn: 283718
2016-10-10 06:25:42 +00:00
Michael Zuckerman
78eb7f43a4 [x86][inline-asm][llvm] accept 'v' constraint
Commit in the name of:Coby Tayree
1.'v' constraint for (x86) non-avx arch imitates the already implemented 'x' constraint, i.e. allows XMM{0-15} & YMM{0-15} depending on the apparent arch & mode (32/64).
2.for the avx512 arch it allows [X,Y,Z]MM{0-31} (mode dependent)

This patch applies the needed changes to clang
 clang patch: https://reviews.llvm.org/D25004

Differential Revision: D25005
 

llvm-svn: 283717
2016-10-10 05:48:56 +00:00
Dylan McKay
01431d3493 [AVR] Enable generation of the TableGen assembly writer tables
This also changes the order of the statements in CMakeLists.txt to be
alphabetical.

llvm-svn: 283711
2016-10-10 01:28:45 +00:00
Brian Gesiak
c06864a754 [lit] Remove (or allow specific) unused imports
Summary:
Using Python linter flake8 on the utils/lit reveals several linter
warnings designated "F401: Unused import". Fix or silence these
warnings.

Some of these unused imports are legitimate, while some are part of lit's API.
For example, users of lit expect to be able to access `lit.formats.ShTest` in
their `lit.cfg`, despite the module hierarchy for that symbol actually being
`lit.formats.shtest.ShTest`. To silence linter errors for these lines,
include a "noqa" directive.

Reviewers: echristo, delcypher, beanz, ddunbar

Subscribers: mehdi_amini, llvm-commits

Differential Revision: https://reviews.llvm.org/D25407

llvm-svn: 283710
2016-10-10 01:22:06 +00:00
Brian Gesiak
beb38d584e [lit] Remove unused TestingProgressDisplay attr
Summary:
`TestingProgressDisplay` initializes its `current` attribute to `None`, but
never reads or writes the value again. Remove it.

Reviewers: echristo, delcypher, beanz, ddunbar

Subscribers: llvm-commits, mehdi_amini

Differential Revision: https://reviews.llvm.org/D25415

llvm-svn: 283709
2016-10-10 01:20:43 +00:00
Brian Gesiak
8421c838ab [lit] Fix undefined symbol ArgumentError
Summary:
`ArgumentError` is not defined by the Python standard library.
Executing this line of code would throw a exception, but not the
intended one. It would throw a `NameError` exception, since `ArgumentError`
is undefined.

Use `ValueError` instead, which is defined by the Python standard
library.

Reviewers: echristo, delcypher, beanz, ddunbar

Subscribers: llvm-commits, mehdi_amini

Differential Revision: https://reviews.llvm.org/D25410

llvm-svn: 283708
2016-10-10 01:19:27 +00:00
Brian Gesiak
eb592dbc2b [lit] Remove semicolons in Python code
Summary:
Semicolons aren't necessary as statement terminators in Python, and
each of these uses are superfluous as they appear at the end of a line.
The convention is to not use semicolons where not needed, so remove them.

Reviewers: echristo, delcypher, beanz, ddunbar

Subscribers: llvm-commits, mehdi_amini

Differential Revision: https://reviews.llvm.org/D25409

llvm-svn: 283707
2016-10-10 01:18:14 +00:00