Justin Lebar [Fri, 9 Sep 2016 21:07:26 +0000 (21:07 +0000)]
[NVPTX] Implement llvm.fabs.f32, llvm.max.f32, etc.
Summary:
Previously these only worked via NVPTX-specific intrinsics.
This change will allow us to convert these target-specific intrinsics
into the general LLVM versions, allowing existing LLVM passes to reason
about their behavior.
It also gets us some minor codegen improvements as-is, from situations
where we canonicalize code into one of these llvm intrinsics.
Reviewers: majnemer
Subscribers: llvm-commits, jholewinski, tra
Differential Revision: https://reviews.llvm.org/D24300
llvm-svn: 281092
Justin Lebar [Fri, 9 Sep 2016 21:01:02 +0000 (21:01 +0000)]
[StreamExecutor] Make SE work with an in-tree LLVM build.
Summary:
With these changes, we can put parallel-libs within llvm/projects and
build as normal.
This is kind of the minimal change I could figure out how to make while
still making us compatible with llvm's build system. Some things I'm
not thrilled about include:
* The creation of a CoreTests directory (the macros really seemed to
want this)
* Pulling SimpleHostPlatformDevice.h into CoreTests. It seems to me
this should live inside unittests/include, or maybe tests/include,
but I didn't want to make that change in this patch.
One important piece of work that remains to be done is to make
$ ninja check-streamexecutor
run all the tests. Right now the only way I've figured out to run the
tests is
$ ninja projects/parallel-libs/streamexecutor/unittests/StreamExecutorUnitTests
$ projects/parallel-libs/streamexecutor/unittests/CoreTests/CoreTests
Reviewers: jhen
Subscribers: beanz, parallel_libs-commits, jprice
Differential Revision: https://reviews.llvm.org/D24368
llvm-svn: 281091
Filipe Cabecinhas [Fri, 9 Sep 2016 20:43:20 +0000 (20:43 +0000)]
[asan] Add a new AddressDescription structure, which can describe any type of address.
Summary:
This is useful for inclusion in the Error* structures, to describe an
arbitrary address.
Remove the old struct since it's used only once. This removes one level of
indirection, and moves all *AddressDescription to be one of the recently
introduced structures.
This merges differential revisions: D24131 and D24132
Reviewers: kcc, eugenis, vitalybuka
Subscribers: kubabrecka, llvm-commits
Differential Revision: https://reviews.llvm.org/D24131
llvm-svn: 281090
Justin Lebar [Fri, 9 Sep 2016 20:35:43 +0000 (20:35 +0000)]
[CUDA] Make __GCC_ATOMIC_XXX_LOCK_FREE macros the same on host/device.
Summary:
This fixes a bug where we were unable to compile the following CUDA
file with libstdc++ (didn't try libc++):
#include <future>
void foo() { std::shared_future<int> x; }
The problem is that <future> only defines std::shared_future if
__GCC_ATOMIC_INT_LOCK_FREE > 1. When we compiled this file for device,
the macro was set to 1, and then the class didn't exist at all.
Reviewers: tra
Subscribers: cfe-commits, jhen
Differential Revision: https://reviews.llvm.org/D24407
llvm-svn: 281089
Saleem Abdulrasool [Fri, 9 Sep 2016 20:11:31 +0000 (20:11 +0000)]
ARM: move the builtins libcall CC setup
Move the target specific setup into the target specific lowering setup. As
pointed out by Anton, the initial change was moving this too high up the stack
resulting in a violation of the layering (the target generic code path setup
target specific bits). Sink this into the ARM specific setup. NFC.
llvm-svn: 281088
Michael Gottesman [Fri, 9 Sep 2016 19:48:22 +0000 (19:48 +0000)]
[cmake] Fix a stale comment from an earlier version of r281085. NFC.
llvm-svn: 281086
Michael Gottesman [Fri, 9 Sep 2016 19:45:34 +0000 (19:45 +0000)]
[cmake] Export gtest/gtest_main and its dependencies via a special build tree only cmake exports file.
Previously, gtest/gtest_main were not exported via cmake. The intention here was
to ensure that users whom are linking against the LLVM install tree would not
get the gtest/gtest_main targets. This prevents downstream projects that link
against the LLVM build tree (i.e. Swift) from getting this dependency
information in their cmake builds. Without such dependency information, linker
issues can result on linux due to LLVMSupport being put before gtest on the
linker command line.
This commit preserves behavior that we want for the install tree, while adding
support for the build tree by:
1. The special casing for gtest/gtest_main in the add_llvm_library code is
removed in favor of a flag called "BUILDTREE_ONLY". If this is set, then the
library is communicating that it is only meant to be exported into the build
tree and is not meant to be installed or exported via the install tree. This
part is just a tweak to remove the special case, the underlying code is the
same.
2. The cmake code that exports cmake targets for the build tree has special code
to import an additional targets file called
LLVMBuildTreeOnlyExports.cmake. Additionally the extra targets are added to the
LLVMConfig.cmake's LLVM_EXPORTED_TARGETS variable. In contrast, the
"installation" cmake file uses the normal LLVM_EXPORTS_TARGETS as before and
does not include the extra exports file. This is implemented by
defining/undefining variables when performing a configure of the build/install
tree LLVMConfig.cmake files.
llvm-svn: 281085
Rafael Espindola [Fri, 9 Sep 2016 19:42:11 +0000 (19:42 +0000)]
Compact InputSectionData from 64 to 48 bytes. NFC.
llvm-svn: 281084
Saleem Abdulrasool [Fri, 9 Sep 2016 19:41:46 +0000 (19:41 +0000)]
test: adjust the GUID
Although the GUID seems to be stable across test runs now, it seems to be
unstable across hosts. Lets be a bit more lax about the reading of the RSDS
record.
llvm-svn: 281083
Rafael Espindola [Fri, 9 Sep 2016 19:32:36 +0000 (19:32 +0000)]
Add a lower level zlib::uncompress.
SmallVectors are convenient, but they don't cover every use case.
In particular, they are fairly large (3 pointers + one element) and
there is no way to take ownership of the buffer to put it somewhere
else. This patch then adds a lower lever interface that works with
any buffer.
llvm-svn: 281082
Wei Ding [Fri, 9 Sep 2016 19:31:51 +0000 (19:31 +0000)]
AMDGPU : Fix mqsad_u32_u8 instruction incorrect data type.
Differential Revision: http://reviews.llvm.org/D23700
llvm-svn: 281081
Tom Stellard [Fri, 9 Sep 2016 19:28:00 +0000 (19:28 +0000)]
AMDGPU/SI: Make sure llvm.amdgcn.implicitarg.ptr() is 8-byte aligned for HSA
Reviewers: arsenm
Subscribers: arsenm, wdng, nhaehnle, llvm-commits
Differential Revision: https://reviews.llvm.org/D24405
llvm-svn: 281080
Saleem Abdulrasool [Fri, 9 Sep 2016 19:26:03 +0000 (19:26 +0000)]
COFF: make builds more reproducible
Change the way we calculate the build id to use MD5 to give reproducible build
ids. Previously we would generate random bytes for the build id GUID.
llvm-svn: 281079
Manman Ren [Fri, 9 Sep 2016 19:03:07 +0000 (19:03 +0000)]
Modules: revert r280728.
In post-commit review, Richard suggested a better way to fix this.
rdar://
27926200
llvm-svn: 281078
Zachary Turner [Fri, 9 Sep 2016 19:00:49 +0000 (19:00 +0000)]
[pdb] Print out some more info when dumping a raw stream.
We have various command line options that print the type of a
stream, the size of a stream, etc but nowhere that it can all be
viewed together.
Since a previous patch introduced the ability to dump the bytes
of a stream, this seems like a good place to present a full view
of the stream's properties including its size, what kind of data
it represents, and the blocks it occupies. So I added the
ability to print that information to the -stream-data command
line option.
llvm-svn: 281077
Vedant Kumar [Fri, 9 Sep 2016 18:44:40 +0000 (18:44 +0000)]
[llvm-cov] Remove some asserts in the html renderer (NFC)
These asserts are making tests fragile. The renderer does not enter an
invalid state when they fail, however, it may spit out a garbled
coverage report because the source text no longer matches the provided
coverage mapping.
Another follow-up to r281072.
llvm-svn: 281076
Evgeniy Stepanov [Fri, 9 Sep 2016 18:43:24 +0000 (18:43 +0000)]
[asan] Disable handle_abort in Android tests.
The same thing is already done on Mac. handle_abort slows down tests
significantly because it triggers tombstone collection on Android;
also, it changes failed test outcome from "not-crash" to "crash" (as
in "bin/not --crash").
This change adds handle_abort=0 to asan options on android (test
only!), and also tweaks android_run.py to semi-correctly pass the
crash/no-crash status to the caller.
llvm-svn: 281075
Dehao Chen [Fri, 9 Sep 2016 18:42:35 +0000 (18:42 +0000)]
Do not widen load for different variable in GVN.
Summary:
Widening load in GVN is too early because it will block other optimizations like PRE, LICM.
https://llvm.org/bugs/show_bug.cgi?id=29110
The SPECCPU2006 benchmark impact of this patch:
Reference: o2_nopatch
(1): o2_patched
Benchmark Base:Reference (1)
-------------------------------------------------------
spec/2006/fp/C++/444.namd 25.2 -0.08%
spec/2006/fp/C++/447.dealII 45.92 +1.05%
spec/2006/fp/C++/450.soplex 41.7 -0.26%
spec/2006/fp/C++/453.povray 35.65 +1.68%
spec/2006/fp/C/433.milc 23.79 +0.42%
spec/2006/fp/C/470.lbm 41.88 -1.12%
spec/2006/fp/C/482.sphinx3 47.94 +1.67%
spec/2006/int/C++/471.omnetpp 22.46 -0.36%
spec/2006/int/C++/473.astar 21.19 +0.24%
spec/2006/int/C++/483.xalancbmk 36.09 -0.11%
spec/2006/int/C/400.perlbench 33.28 +1.35%
spec/2006/int/C/401.bzip2 22.76 -0.04%
spec/2006/int/C/403.gcc 32.36 +0.12%
spec/2006/int/C/429.mcf 41.04 -0.41%
spec/2006/int/C/445.gobmk 26.94 +0.04%
spec/2006/int/C/456.hmmer 24.5 -0.20%
spec/2006/int/C/458.sjeng 28 -0.46%
spec/2006/int/C/462.libquantum 55.25 +0.27%
spec/2006/int/C/464.h264ref 45.87 +0.72%
geometric mean +0.23%
For most benchmarks, it's a wash, but we do see stable improvements on some benchmarks, e.g. 447,453,482,400.
Reviewers: davidxl, hfinkel, dberlin, sanjoy, reames
Subscribers: gberry, junbuml
Differential Revision: https://reviews.llvm.org/D24096
llvm-svn: 281074
Rui Ueyama [Fri, 9 Sep 2016 18:37:08 +0000 (18:37 +0000)]
Fix another -Wunused-variable for non-assert build.
llvm-svn: 281073
Vedant Kumar [Fri, 9 Sep 2016 18:34:43 +0000 (18:34 +0000)]
[llvm-cov] Try to fix the native_separators.c test some more
It's still breaking this bot (though, it looks like it always had been):
http://lab.llvm.org:8011/builders/clang-x86-windows-msvc2015
This time, add quotes around llvm-{cov,config} so that lit won't expand
them.
Thanks to Reid for suggesting the patch!
llvm-svn: 281072
Douglas Katzman [Fri, 9 Sep 2016 18:20:49 +0000 (18:20 +0000)]
Myriad: nominally "support" ASAN.
Doesn't work, but needs to be enabled in order to get there.
llvm-svn: 281071
Zachary Turner [Fri, 9 Sep 2016 18:17:52 +0000 (18:17 +0000)]
[pdb] Add command line options for dumping individual streams and blocks
I ran into a situation where I wanted to print out the contents of
page 6 of a PDB as a binary blob, and there was no straightforward
way to do that.
In addition to adding that, this patch also adds the ability to dump
a stream by index as a binary blob, and it will stitch together all
the blocks and dump the whole thing as one seemingly contiguous
sequence of bytes.
llvm-svn: 281070
Rui Ueyama [Fri, 9 Sep 2016 18:07:33 +0000 (18:07 +0000)]
Fix -Wunused-variable for non-assert build.
llvm-svn: 281069
George Rokos [Fri, 9 Sep 2016 18:04:23 +0000 (18:04 +0000)]
[OPENMP] ppc64le recognized as big-endian
There is a bug in CMakeLists which causes powerpc64le systems to be recognized as big-endian. This patch fixes the issue.
Differential Revision: https://reviews.llvm.org/D23626
llvm-svn: 281068
Zachary Turner [Fri, 9 Sep 2016 18:03:39 +0000 (18:03 +0000)]
[pdb] Pass CVRecord's through the visitor as non-const references.
This simplifies a lot of code, and will actually be necessary for
an upcoming patch to serialize TPI record hash values.
The idea before was that visitors should be examining records, not
modifying them. But this is no longer true with a visitor that
constructs a CVRecord from Yaml. To handle this until now, we
were doing some fixups on CVRecord objects at a higher level, but
the code is really awkward, and it makes sense to just have the
visitor write the bytes into the CVRecord. In doing so I uncovered
a few bugs related to `Data` and `RawData` and fixed those.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D24362
llvm-svn: 281067
Kostya Serebryany [Fri, 9 Sep 2016 18:00:04 +0000 (18:00 +0000)]
[libFuzzer] one more puzzle, value_profile cracks it in a second
llvm-svn: 281066
George Rokos [Fri, 9 Sep 2016 17:55:26 +0000 (17:55 +0000)]
[OPENMP] Implementation of omp_get_default_device and omp_set_default_device
Implementation of missing OpenMP 4.0 API functions omp_get_default_device and omp_set_default_device.
Also, added support for the environment variable OMP_DEFAULT_DEVICE.
Differential Revision: https://reviews.llvm.org/D23587
llvm-svn: 281065
Eric Liu [Fri, 9 Sep 2016 17:50:49 +0000 (17:50 +0000)]
Also cleanup comments around redundant colons/commas in format::cleanup.
Reviewers: djasper
Subscribers: klimek, cfe-commits
Differential Revision: https://reviews.llvm.org/D24400
llvm-svn: 281064
Zachary Turner [Fri, 9 Sep 2016 17:46:17 +0000 (17:46 +0000)]
[pdb] Write PDB TPI Stream from Yaml.
This writes the full sequence of type records described in
Yaml to the TPI stream of the PDB file.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D24316
llvm-svn: 281063
Vedant Kumar [Fri, 9 Sep 2016 17:37:11 +0000 (17:37 +0000)]
[llvm-cov] Handle native paths correctly in the text index
Treat filenames the same way in the text index as we do in the html
index. This is a follow-up to r281008 (an attempt to unbreak the
native_separators.c test on Windows).
Patch by Maggie Yi!
llvm-svn: 281062
Aaron Ballman [Fri, 9 Sep 2016 17:30:59 +0000 (17:30 +0000)]
Move the ocaml_all target into the Misc folder for cleanliness; NFC.
llvm-svn: 281061
Reid Kleckner [Fri, 9 Sep 2016 17:29:36 +0000 (17:29 +0000)]
[codeview] Don't assert if the array element type is incomplete
This can happen when the frontend knows the debug info will be emitted
somewhere else. Usually this happens for dynamic classes with out of
line constructors or key functions, but it can also happen when modules
are enabled.
llvm-svn: 281060
Vedant Kumar [Fri, 9 Sep 2016 17:24:31 +0000 (17:24 +0000)]
[Bitcode] Add compatibility test for the 3.9 release
Fork off compatibility.ll for the 3.9 release. The *.bc file in this
commit was produced using a Release build of the release_39 branch.
llvm-svn: 281059
Todd Fiala [Fri, 9 Sep 2016 17:07:15 +0000 (17:07 +0000)]
xfail DarwinLog "filter message by regex" tests
These tests are not working reliably. I'm marking them
xfail until I resolve the issue.
Tracked by:
llvm.org/pr30299
llvm-svn: 281058
Reid Kleckner [Fri, 9 Sep 2016 17:03:53 +0000 (17:03 +0000)]
[DebugInfo] Ensure complete type is emitted with -fstandalone-debug
The logic for upgrading a class from a forward decl to a complete type
was not checking the debug info emission level before applying the
vtable optimization. This meant we ended up without debug info for a
class which was required to be complete. I noticed it because it
triggered an assertion during CodeView emission, but that's a separate
issue.
llvm-svn: 281057
Reid Kleckner [Fri, 9 Sep 2016 16:42:50 +0000 (16:42 +0000)]
Make -fstandalone-debug and -flimit-debug-info available in clang-cl
Our limited debug info optimizations are breaking down at DLL
boundaries, so we're going to evaluate the size impact of these
settings, and possibly change the default.
Users should be able to override our settings, though.
llvm-svn: 281056
Sanjay Patel [Fri, 9 Sep 2016 16:35:20 +0000 (16:35 +0000)]
[InstCombine] add tests to show pattern matching failures due to commutation
I was looking to fix a bug in getComplexity(), and these cases showed up as
obvious failures. I'm not sure how to find these in general though.
llvm-svn: 281055
Justin Lebar [Fri, 9 Sep 2016 16:29:13 +0000 (16:29 +0000)]
Attempt to fix static_asserts in SelectionDAGNodes.h that MSVC chokes on.
These were added in r281051, which, I am embarrassed to admit, has an
incomplete commit message that I forgot to update before pushing. You
can ignore element (2) in that list.
llvm-svn: 281054
Reid Kleckner [Fri, 9 Sep 2016 16:27:04 +0000 (16:27 +0000)]
[codeview] Extend the heuristic for detecting classes imported from DLLs
If a dynamic class contains a dllimport method, then assume the class
may not be constructed in this DLL, and therefore the vtable will live
in a different PDB.
This heuristic is still incomplete, and will miss things like abstract
base classes that are only constructed on one side of the DLL interface.
That said, this heuristic does detect some cases that are currently
problematic, and may be useful to other projects that don't use many
DLLs.
llvm-svn: 281053
Tobias Grosser [Fri, 9 Sep 2016 16:11:26 +0000 (16:11 +0000)]
FlattenAlgo: Ensure we _really_ obtain a param space
This resolves "isl_space.c:1775: not a parameter space" errors I have seen
on two systems.
llvm-svn: 281052
Justin Lebar [Fri, 9 Sep 2016 16:07:06 +0000 (16:07 +0000)]
[SelectionDAG] Fix two issues with SDNode::getRawSubclassData().
1) On some platforms, sizeof(SDNodeBits) == 1, so we were only copying
one byte out of the bitfield when we wanted to copy two, and we were
leaving half of the return value of getRawSubclassData() undefined.
2) Something something bitfields, not sure exactly what the issue or fix
is, yet. (TODO)
Summary:
Previously we were assuming that SDNodeBits covered all of SDNode's
anonymous subclass data bitfield union. But that's not right; it might
have size 1, in which it clearly doesn't.
This patch adds a field that does cover the whole union and adds
static_asserts to ensure it stays correct.
Reviewers: ahatanak, chandlerc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D24223
llvm-svn: 281051
Sam Kolton [Fri, 9 Sep 2016 14:44:04 +0000 (14:44 +0000)]
AMDGPU] Assembler: better support for immediate literals in assembler.
Summary:
Prevously assembler parsed all literals as either 32-bit integers or 32-bit floating-point values. Because of this we couldn't support f64 literals.
E.g. in instruction "v_fract_f64 v[0:1], 0.5", literal 0.5 was encoded as 32-bit literal 0x3f000000, which is incorrect and will be interpreted as 3.
0517578125E-5 instead of 0.5. Correct encoding is inline constant 240 (optimal) or 32-bit literal 0x3FE00000 at least.
With this change the way immediate literals are parsed is changed. All literals are always parsed as 64-bit values either integer or floating-point. Then we convert parsed literals to correct form based on information about type of operand parsed (was literal floating or binary) and type of expected instruction operands (is this f32/64 or b32/64 instruction).
Here are rules how we convert literals:
- We parsed fp literal:
- Instruction expects 64-bit operand:
- If parsed literal is inlinable (e.g. v_fract_f64_e32 v[0:1], 0.5)
- then we do nothing this literal
- Else if literal is not-inlinable but instruction requires to inline it (e.g. this is e64 encoding, v_fract_f64_e64 v[0:1], 1.5)
- report error
- Else literal is not-inlinable but we can encode it as additional 32-bit literal constant
- If instruction expect fp operand type (f64)
- Check if low 32 bits of literal are zeroes (e.g. v_fract_f64 v[0:1], 1.5)
- If so then do nothing
- Else (e.g. v_fract_f64 v[0:1], 3.1415)
- report warning that low 32 bits will be set to zeroes and precision will be lost
- set low 32 bits of literal to zeroes
- Instruction expects integer operand type (e.g. s_mov_b64_e32 s[0:1], 1.5)
- report error as it is unclear how to encode this literal
- Instruction expects 32-bit operand:
- Convert parsed 64 bit fp literal to 32 bit fp. Allow lose of precision but not overflow or underflow
- Is this literal inlinable and are we required to inline literal (e.g. v_trunc_f32_e64 v0, 0.5)
- do nothing
- Else report error
- Do nothing. We can encode any other 32-bit fp literal (e.g. v_trunc_f32 v0,
10000000.0)
- Parsed binary literal:
- Is this literal inlinable (e.g. v_trunc_f32_e32 v0, 35)
- do nothing
- Else, are we required to inline this literal (e.g. v_trunc_f32_e64 v0, 35)
- report error
- Else, literal is not-inlinable and we are not required to inline it
- Are high 32 bit of literal zeroes or same as sign bit (32 bit)
- do nothing (e.g. v_trunc_f32 v0, 0xdeadbeef)
- Else
- report error (e.g. v_trunc_f32 v0, 0x123456789abcdef0)
For this change it is required that we know operand types of instruction (are they f32/64 or b32/64). I added several new register operands (they extend previous register operands) and set operand types to corresponding types:
'''
enum OperandType {
OPERAND_REG_IMM32_INT,
OPERAND_REG_IMM32_FP,
OPERAND_REG_INLINE_C_INT,
OPERAND_REG_INLINE_C_FP,
}
'''
This is not working yet:
- Several tests are failing
- Problems with predicate methods for inline immediates
- LLVM generated assembler parts try to select e64 encoding before e32.
More changes are required for several AsmOperands.
Reviewers: vpykhtin, tstellarAMD
Subscribers: arsenm, kzhuravl, artem.tamazov
Differential Revision: https://reviews.llvm.org/D22922
llvm-svn: 281050
George Rimar [Fri, 9 Sep 2016 14:35:36 +0000 (14:35 +0000)]
Recommit [ELF] - Versionscript: do not treat non-wildcarded names as wildcards.
Fixed code that was not checked before on windows for me, because of testcases that are
disabled on that platform atm.
Inital commit message:
"[ELF] - Versionscript: do not treat non-wildcarded names as wildcards."
Previously we incorrectly handled cases when symbol name in extern c++ tag
was enclosed in quotes. Next case was treated as wildcard:
GLIBCXX_3.4 {
extern "C++" {
"aaa*"
}
But it should have not. Quotes around aaa here means that we should have do exact
name matching.
That is PR30268 which has name with pointer is interpreted as wildcard by lld:
extern "C++" {
"operator delete[](void*)";
Patch fixes the issue.
Differential revision: https://reviews.llvm.org/D24229
llvm-svn: 281049
Filipe Cabecinhas [Fri, 9 Sep 2016 14:18:22 +0000 (14:18 +0000)]
[asan] Since r280945 fixed the OS X abort() problem, merge scariness_score_test.cc and make it a general test.
Summary: Merges back both scariness_score_test.cc files, since the Linux-specific version shouldn't be needed any more.
Reviewers: kcc, eugenis, vitalybuka
Subscribers: llvm-commits, kubabrecka
Differential Revision: https://reviews.llvm.org/D24347
llvm-svn: 281048
Chris Dewhurst [Fri, 9 Sep 2016 14:16:51 +0000 (14:16 +0000)]
[Sparc][LEON] Removed the parts of the errata fixes implemented using inline assembly as this is not the desired behaviour for end-users. Small change to a unit test to implement this without requiring the inline assembly.
llvm-svn: 281047
George Rimar [Fri, 9 Sep 2016 14:16:00 +0000 (14:16 +0000)]
Revert r281045, it broke BB.
Broken BB:
http://lab.llvm.org:8011/builders/lld-x86_64-darwin13/builds/27211
llvm-svn: 281046
George Rimar [Fri, 9 Sep 2016 13:58:07 +0000 (13:58 +0000)]
Recommit [ELF] - Versionscript: do not treat non-wildcarded names as wildcards.
Fixed code that was not checked by testcases that are disabled on windows.
Inital commit message:
"[ELF] - Versionscript: do not treat non-wildcarded names as wildcards."
Previously we incorrectly handled cases when symbol name in extern c++ tag
was enclosed in quotes. Next case was treated as wildcard:
GLIBCXX_3.4 {
extern "C++" {
"aaa*"
}
But it should have not. Quotes around aaa here means that we should have do exact
name matching.
That is PR30268 which has name with pointer is interpreted as wildcard by lld:
extern "C++" {
"operator delete[](void*)";
Patch fixes the issue.
Differential revision: https://reviews.llvm.org/D24229
llvm-svn: 281045
James Molloy [Fri, 9 Sep 2016 13:35:36 +0000 (13:35 +0000)]
[ARM] ADD with a negative offset can become SUB for free
So model that directly in TTI::getIntImmCost().
llvm-svn: 281044
James Molloy [Fri, 9 Sep 2016 13:35:28 +0000 (13:35 +0000)]
[ARM] icmp %x, -C can be lowered to a simple ADDS or CMN
Tell TargetTransformInfo about this so ConstantHoisting is informed.
llvm-svn: 281043
Simon Pilgrim [Fri, 9 Sep 2016 13:31:52 +0000 (13:31 +0000)]
[SelectionDAG] Ensure DAG::getZeroExtendInReg is called with a scalar type
Fixes issue with rL280927 identified by Mikael Holmén
llvm-svn: 281042
George Rimar [Fri, 9 Sep 2016 12:56:55 +0000 (12:56 +0000)]
[ELF] - Revert r281038 to see if that unbreaks build bot.
http://lab.llvm.org:8011/builders/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/builds/19703
llvm-svn: 281041
James Molloy [Fri, 9 Sep 2016 12:52:24 +0000 (12:52 +0000)]
[Thumb] Select (CMPZ X, -C) -> (CMPZ (ADDS X, C), 0)
The CMPZ #0 disappears during peepholing, leaving just a tADDi3, tADDi8 or t2ADDri. This avoids having to materialize the expensive negative constant in Thumb-1, and allows a shrinking from a 32-bit CMN to a 16-bit ADDS in Thumb-2.
llvm-svn: 281040
Tobias Grosser [Fri, 9 Sep 2016 12:31:38 +0000 (12:31 +0000)]
Add namespace specifier before nullptr_t
This fixes the following compile time errors:
error: unknown type name 'nullptr_t'; did you mean 'std::nullptr_t'
llvm-svn: 281039
George Rimar [Fri, 9 Sep 2016 12:22:28 +0000 (12:22 +0000)]
[ELF] - Versionscript: do not treat non-wildcarded names as wildcards.
Previously we incorrectly handled cases when symbol name in extern c++ tag
was enclosed in quotes. Next case was treated as wildcard:
GLIBCXX_3.4 {
extern "C++" {
"aaa*"
}
But it should have not. Quotes around aaa here means that we should have do exact
name matching.
That is PR30268 which has name with pointer is interpreted as wildcard by lld:
extern "C++" {
"operator delete[](void*)";
Patch fixes the issue.
Differential revision: https://reviews.llvm.org/D24229
llvm-svn: 281038
Tim Northover [Fri, 9 Sep 2016 11:47:31 +0000 (11:47 +0000)]
GlobalISel: remove G_TYPE and G_PHI
These instructions were only necessary when type information was stored in the
MachineInstr (because only generic MachineInstrs possessed a type). Now that
it's in MachineRegisterInfo, COPY and PHI work fine.
llvm-svn: 281037
Tim Northover [Fri, 9 Sep 2016 11:46:58 +0000 (11:46 +0000)]
GlobalISel: fix comments and add assertions for valid instructions.
llvm-svn: 281036
Tim Northover [Fri, 9 Sep 2016 11:46:34 +0000 (11:46 +0000)]
GlobalISel: move type information to MachineRegisterInfo.
We want each register to have a canonical type, which means the best place to
store this is in MachineRegisterInfo rather than on every MachineInstr that
happens to use or define that register.
Most changes following from this are pretty simple (you need an MRI anyway if
you're going to be doing any transformations, so just check the type there).
But legalization doesn't really want to check redundant operands (when, for
example, a G_ADD only ever has one type) so I've made use of MCInstrDesc's
operand type field to encode these constraints and limit legalization's work.
As an added bonus, more validation is possible, both in MachineVerifier and
MachineIRBuilder (coming soon).
llvm-svn: 281035
Tobias Grosser [Fri, 9 Sep 2016 11:16:50 +0000 (11:16 +0000)]
IslNodeBuilder: Add missing __isl_take annotation
llvm-svn: 281034
Simon Dardis [Fri, 9 Sep 2016 11:06:01 +0000 (11:06 +0000)]
Revert "[mips] Fix c.<cc>.<fmt> instruction definition."
This reverts commit r281022. Mips buildbot broke, due to unhandled register
class FCC.
llvm-svn: 281033
Nitesh Jain [Fri, 9 Sep 2016 10:46:25 +0000 (10:46 +0000)]
[LLDB][MIPS] Fix Emulation for JALR64 Instruction
Subscribers: jaydeep, bhushan, slthakur, sdardis, lldb-commits
llvm-svn: 281032
Nitesh Jain [Fri, 9 Sep 2016 10:20:08 +0000 (10:20 +0000)]
[LLDB][MIPS] Fix TestEhFrameUnwind.py for MIPS
Reviewers: clayborg, labath
Subscribers: jaydeep, bhushan, slthakur, lldb-commits
Differential Revision: https://reviews.llvm.org/D24122
llvm-svn: 281031
Dimitar Vlahovski [Fri, 9 Sep 2016 10:14:11 +0000 (10:14 +0000)]
Fixing a build breakage caused from a change in LLVM rL281019
Summary:
LLVM guys did some clean-up of the Attribute getters/setters
and because of that the build was failing.
Reviewers: ldrumm
Subscribers: lldb-commits
Differential Revision: https://reviews.llvm.org/D24382
llvm-svn: 281030
Pavel Labath [Fri, 9 Sep 2016 10:12:57 +0000 (10:12 +0000)]
NFC: Reformat ABISysV_i386 register context into something readable
add clang-format guards so it does not reformat it again.
llvm-svn: 281029
Sam Kolton [Fri, 9 Sep 2016 10:08:02 +0000 (10:08 +0000)]
[AMDGPU] Assembler: rename amd_kernel_code_t asm names according to spec
Summary:
Also removed duplicate code from AMDGPUTargetAsmStreamer.
This change only change how amd_kernel_code_t is parsed and printed. No variable names are changed.
Reviewers: vpykhtin, tstellarAMD
Subscribers: arsenm, wdng, nhaehnle
Differential Revision: https://reviews.llvm.org/D24296
llvm-svn: 281028
James Molloy [Fri, 9 Sep 2016 09:51:06 +0000 (09:51 +0000)]
[Thumb1] Teach optimizeCompareInstr about thumb1 compares
This avoids us doing a completely unneeded "cmp r0, #0" after a flag-setting instruction if we only care about the Z or C flags.
Add LSL/LSR to the whitelist while we're here and add testing. This code could really do with a spring clean.
llvm-svn: 281027
Nitesh Jain [Fri, 9 Sep 2016 09:50:33 +0000 (09:50 +0000)]
[LLDB][MIPS] Fix TestLldbGdbServer.py failure
Subscribers: jaydeep, bhushan, slthakur, lldb-commits
llvm-svn: 281026
Pavel Labath [Fri, 9 Sep 2016 09:49:54 +0000 (09:49 +0000)]
Fix new gdb-remote client unit test for windows
The behaviour of FileSpec differed between host OS versions. Hardcode the path
syntax to posix, as we don't care about that in this test.
llvm-svn: 281025
Eugene Leviant [Fri, 9 Sep 2016 09:46:16 +0000 (09:46 +0000)]
Linker script: implement AT [ (address) ] for PHDR
Differential revision: https://reviews.llvm.org/D24340
llvm-svn: 281024
Sam Kolton [Fri, 9 Sep 2016 09:37:51 +0000 (09:37 +0000)]
[AMDGPU] Assembler: match e32 VOP instructions before e64.
Summary:
Split assembler match table in 4 tables with assembler variants:
Default - all instructions except VOP3, SDWA and DPP
- VOP3
- SDWA
- DPP
First match Default table then VOP3, SDWA and DPP.
Reviewers: tstellarAMD, artem.tamazov, vpykhtin
Subscribers: arsenm, wdng, nhaehnle, AMDGPU
Differential Revision: https://reviews.llvm.org/D24252
llvm-svn: 281023
Simon Dardis [Fri, 9 Sep 2016 09:22:52 +0000 (09:22 +0000)]
[mips] Fix c.<cc>.<fmt> instruction definition.
As part of this effort, remove MipsFCmp nodes and use tablegen
patterns rather than custom lowering through C++.
Unexpectedly, this improves codesize for microMIPS as previous floating
point setcc expansions would materialize 0 and 1 into GPRs before using
the relevant mov[tf].[sd] instruction. Now $zero is used directly.
Reviewers: dsanders, vkalintiris, zoran.jovanovic
Differential Review: https://reviews.llvm.org/D23118
llvm-svn: 281022
Chris Dewhurst [Fri, 9 Sep 2016 09:08:13 +0000 (09:08 +0000)]
[Sparc][LEON] Unit test for CASA instruction supported by some LEON processors added.
llvm-svn: 281021
Gor Nishanov [Fri, 9 Sep 2016 05:39:00 +0000 (05:39 +0000)]
[Coroutines] Part13: Handle single edge PHINodes across suspends
Summary:
If one of the uses of the value is a single edge PHINode, handle it.
Original:
%val = something
<suspend>
%p = PHINode [%val]
After Spill + Part13:
%val = something
%slot = gep val.spill.slot
store %val, %slot
<suspend>
%p = load %slot
Plus tiny fixes/changes:
* use correct index for coro.free in CoroCleanup
* fixup id parameter in coro.free to allow authoring coroutine in plain C with __builtins
Reviewers: majnemer
Subscribers: mehdi_amini, llvm-commits
Differential Revision: https://reviews.llvm.org/D24242
llvm-svn: 281020
Amaury Sechet [Fri, 9 Sep 2016 04:50:38 +0000 (04:50 +0000)]
Rationalise the attribute getter/setter methods on Function and CallSite.
Summary:
While woring on mapping attributes in the C API, it clearly appeared that the recent changes in the API on the C++ side left Function and Call/Invoke with an attribute API that grew in an ad hoc manner. This makes it difficult to work with it, because one doesn't know which overloads exists and which do not.
Make sure that getter/setter function exists for both enum and string version. Remove inconsistent getter/setter, unless they have many callsites.
This should make it easier to work with attributes in the future.
This doesn't change how attribute works.
Reviewers: bkramer, whitequark, mehdi_amini, void
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D21514
llvm-svn: 281019
Amaury Sechet [Fri, 9 Sep 2016 04:42:49 +0000 (04:42 +0000)]
Update clang for D21514. NFC
Summary: As per title.
Reviewers: ahatanak, bkramer, whitequark, mehdi_amini, void
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D21515
llvm-svn: 281018
George Burgess IV [Fri, 9 Sep 2016 02:45:48 +0000 (02:45 +0000)]
[Docs] Fix typos, remove trailing whitespace.
Avoided wrapping NullabilityDocs at 80cols, since that would've made
this diff much bigger, and never-ending lines seems to be the style for
many of the null-related docs.
llvm-svn: 281017
Kostya Serebryany [Fri, 9 Sep 2016 02:38:28 +0000 (02:38 +0000)]
[libFuzzer] improve -print_pcs to not print new PCs coming from libFuzzer itself
llvm-svn: 281016
Kostya Serebryany [Fri, 9 Sep 2016 02:13:27 +0000 (02:13 +0000)]
[sanitizer] fix a potential buffer overflow due to __sanitizer_symbolize_pc (need to put a zero after strncmp). LOL
llvm-svn: 281015
Kostya Serebryany [Fri, 9 Sep 2016 01:57:38 +0000 (01:57 +0000)]
[libFuzzer] remove unneeded call
llvm-svn: 281014
Craig Topper [Fri, 9 Sep 2016 01:36:10 +0000 (01:36 +0000)]
[AVX-512] Add VPCMP instructions to the load folding tables and make them commutable.
llvm-svn: 281013
Craig Topper [Fri, 9 Sep 2016 01:36:04 +0000 (01:36 +0000)]
[AVX-512] Add more integer vector comparison tests with loads. Some of these show opportunities where we can commute to fold loads.
Commutes will be added in a followup commit.
llvm-svn: 281012
Vedant Kumar [Fri, 9 Sep 2016 01:32:55 +0000 (01:32 +0000)]
[llvm-cov] Emit a summary in the report directory's index
llvm-cov writes out an index file in '-output-dir' mode, albeit not a
very informative one. Try to fix that by using the CoverageReport API to
include some basic summary information in the index file.
llvm-svn: 281011
Vedant Kumar [Fri, 9 Sep 2016 01:32:51 +0000 (01:32 +0000)]
[llvm-cov] Constify some methods (NFC)
llvm-svn: 281010
Vedant Kumar [Fri, 9 Sep 2016 01:32:49 +0000 (01:32 +0000)]
[llvm-cov] Add an API to prepare file reports (NFC)
It would be nice to prepare file reports (using the CoverageReport API)
without actually rendering them to the console. I plan on using this to
flesh out the 'index' files in the coverage views.
llvm-svn: 281009
Vedant Kumar [Fri, 9 Sep 2016 01:32:47 +0000 (01:32 +0000)]
[llvm-cov] Speculate fix for a Windows-only test (NFC)
This test should have broken after r280896. Fix up the test case
speculatively, since I don't have a way to test it.
I wonder why I didn't get any angry bot emails about this. Maybe none of
the win32 bots test llvm-cov? That could explain it, since the test says
it 'REQUIRES: system-windows', which is restricted to win32 hosts.
Also: why is 'system-windows' not defined for non-win32 Windows bots?
llvm-svn: 281008
Kostya Serebryany [Fri, 9 Sep 2016 01:17:03 +0000 (01:17 +0000)]
[libFuzzer] remove use_traces=1 since use_value_profile seems to be strictly better
llvm-svn: 281007
David Majnemer [Fri, 9 Sep 2016 01:07:01 +0000 (01:07 +0000)]
[X86] Tighten up a comment which confused x64 ABI terminology.
The x64 ABI has two major function types:
- frame functions
- leaf functions
A frame function is one which requires a stack frame. A leaf function
is one which does not. A frame function may or may not have a frame
pointer.
A leaf function does not require a stack frame and may never modify SP
except via a return (RET, tail call via JMP).
A frame function which has a frame pointer is permitted to use the LEA
instruction in the epilogue, a frame function without which doesn't
establish a frame pointer must use ADD to adjust the stack pointer epilogue.
Fun fact: Leaf functions don't require a function table entry
(associated PDATA/XDATA).
llvm-svn: 281006
Michael Kuperstein [Fri, 9 Sep 2016 00:49:29 +0000 (00:49 +0000)]
[X86] Add more baseline tests for "irregular" shuffles. NFC.
This adds more tests for shuffles where the output width does not match
the input width and/or the output is generated from more than two inputs.
llvm-svn: 281005
Rui Ueyama [Fri, 9 Sep 2016 00:25:56 +0000 (00:25 +0000)]
Fix -output option.
Usually, options that are longer than one character can be preceded
either by "-" or "--", but options starting with "o" are exceptions
because they conflict with "-o" option. They have to be preceded by
"--".
llvm-svn: 281004
Hans Wennborg [Thu, 8 Sep 2016 23:35:10 +0000 (23:35 +0000)]
Win64: Don't use REX prefix for direct tail calls
The REX prefix should be used on indirect jmps, but not direct ones.
For direct jumps, the unwinder looks at the offset to determine if
it's inside the current function.
Differential Revision: https://reviews.llvm.org/D24359
llvm-svn: 281003
Mike Aizatsky [Thu, 8 Sep 2016 23:29:37 +0000 (23:29 +0000)]
.clang-tidy: parameters and members should be CamelCased.
Summary: Not sure why is it missing.
Differential Revision: https://reviews.llvm.org/D24364
llvm-svn: 281002
Rui Ueyama [Thu, 8 Sep 2016 23:26:54 +0000 (23:26 +0000)]
Fix bug in -nostdlib.
We still have to skip a token inside SEARCH_DIR() when -nostdlib is
specified. Previuosly, we didn't skip it, so it caused a parse error.
llvm-svn: 281001
Saleem Abdulrasool [Thu, 8 Sep 2016 23:17:34 +0000 (23:17 +0000)]
llvm-objdump: add missing ) in help output, NFC
Add a missing ')' in the help output. NFC.
llvm-svn: 281000
Richard Smith [Thu, 8 Sep 2016 23:14:54 +0000 (23:14 +0000)]
C++ Modules TS: Add parsing and some semantic analysis support for
export-declarations. These don't yet have an effect on name visibility;
we still export everything by default.
llvm-svn: 280999
Vedant Kumar [Thu, 8 Sep 2016 22:53:19 +0000 (22:53 +0000)]
[Driver] Report invalid -mtune/-mcpu parameters when -arch=arm64
Differential Revision: https://reviews.llvm.org/D23643
llvm-svn: 280998
Albert Gutowski [Thu, 8 Sep 2016 22:32:19 +0000 (22:32 +0000)]
Implement MS _rot intrinsics
Reviewers: thakis, Prazek, compnerd, rnk
Subscribers: majnemer, cfe-commits
Differential Revision: https://reviews.llvm.org/D24311
llvm-svn: 280997
Kostya Serebryany [Thu, 8 Sep 2016 22:21:13 +0000 (22:21 +0000)]
[libFuzzer] add missing docs
llvm-svn: 280996
Dehao Chen [Thu, 8 Sep 2016 21:53:33 +0000 (21:53 +0000)]
Remove debug info when hoisting instruction from then/else branch.
Summary: The hoisted instruction is executed speculatively. It could affect the debugging experience as user would see gdb go into code that may not be expected to execute. It will also affect sample profile accuracy by assigning incorrect frequency to source within then/else branch.
Reviewers: davidxl, dblaikie, chandlerc, kcc, echristo
Subscribers: mehdi_amini, probinson, eric_niebler, andreadb, llvm-commits
Differential Revision: https://reviews.llvm.org/D24164
llvm-svn: 280995
Rui Ueyama [Thu, 8 Sep 2016 21:46:21 +0000 (21:46 +0000)]
Define writeUint and use it. NFC.
llvm-svn: 280994
Sanjay Patel [Thu, 8 Sep 2016 21:40:21 +0000 (21:40 +0000)]
[InstCombine] regenerate checks
llvm-svn: 280993
Matthew Simpson [Thu, 8 Sep 2016 21:38:26 +0000 (21:38 +0000)]
[LV] Ensure proper handling of multi-use case when collecting uniforms
The test case included in r280979 wasn't checking what it was supposed to be
checking for the predicated store case. Fixing the test revealed that the
multi-use case (when a pointer is used by both vectorized and scalarized memory
accesses) wasn't being handled properly. We can't skip over
non-consecutive-like pointers since they may have looked consecutive-like with
a different memory access.
llvm-svn: 280992