Alexander Potapenko [Tue, 6 Aug 2013 14:07:46 +0000 (14:07 +0000)]
[libsanitizer] Drive-by fix for -Wempty-body in sanitizer_common_syscalls.inc
This makes the file consistently use { } around the if statements containing the PRE_/POST_ macros.
llvm-svn: 187797
John Thompson [Tue, 6 Aug 2013 14:05:43 +0000 (14:05 +0000)]
Fix comment typo.
llvm-svn: 187796
Tim Northover [Tue, 6 Aug 2013 13:58:03 +0000 (13:58 +0000)]
ARM: implement allowTruncateForTailCall
Now that it's in place, it seems silly not to let ARM make use of the extra
tail call opportunities.
llvm-svn: 187795
Rafael Espindola [Tue, 6 Aug 2013 13:16:28 +0000 (13:16 +0000)]
Add a release not about llvm-ar.
Thanks to Bill Wendling for the reminder.
llvm-svn: 187794
Tim Northover [Tue, 6 Aug 2013 12:50:45 +0000 (12:50 +0000)]
Remove oddly named libraries with "make uninstall-local"
Patch by Edward-san.
llvm-svn: 187793
Alexey Samsonov [Tue, 6 Aug 2013 10:49:15 +0000 (10:49 +0000)]
Store compile unit corresponding to each chain of inlined debug info entries. No functionality change.
llvm-svn: 187792
Elena Demikhovsky [Tue, 6 Aug 2013 10:40:45 +0000 (10:40 +0000)]
LLVM Interpreter: fixed bug 16694
fix for: Bug 16694 - ExecutionEngine/test-interp-vec-loadstore.ll failing on powerpc-darwin8 (http://llvm.org/bugs/show_bug.cgi?id=16694)
The ExecutionEngine/test-interp-vec-loadstore.ll test has been failing on powerpc-darwin8 (on other platforms it passed)
the reason of fail was wrong output by printf. this output is checked by FileCheck, but on little-endian powerpc the output numeric data were printed inside out and FileCheck reported fail.
the printfs have been replaced by checking data inside test and numeric output has been replaced by the text output like : "int test passed, float test passed". The text output is checked by FileCheck.
the dependency on data layout has been removed.
done by Yuri Veselov (Intel)
llvm-svn: 187791
Alexey Samsonov [Tue, 6 Aug 2013 10:32:39 +0000 (10:32 +0000)]
Add LLVM-style RTTI to DIContext/DWARFContext classes
llvm-svn: 187790
Evgeniy Stepanov [Tue, 6 Aug 2013 09:54:33 +0000 (09:54 +0000)]
[sanitizer] Fix NULL arg handling in setitimer/getitimer interceptors.
llvm-svn: 187789
Evgeniy Stepanov [Tue, 6 Aug 2013 09:29:01 +0000 (09:29 +0000)]
[sanitizer] Intercept sched_getaffinity.
llvm-svn: 187788
Tim Northover [Tue, 6 Aug 2013 09:12:35 +0000 (09:12 +0000)]
Refactor isInTailCallPosition handling
This change came about primarily because of two issues in the existing code.
Niether of:
define i64 @test1(i64 %val) {
%in = trunc i64 %val to i32
tail call i32 @ret32(i32 returned %in)
ret i64 %val
}
define i64 @test2(i64 %val) {
tail call i32 @ret32(i32 returned undef)
ret i32 42
}
should be tail calls, and the function sameNoopInput is responsible. The main
problem is that it is completely symmetric in the "tail call" and "ret" value,
but in reality different things are allowed on each side.
For these cases:
1. Any truncation should lead to a larger value being generated by "tail call"
than needed by "ret".
2. Undef should only be allowed as a source for ret, not as a result of the
call.
Along the way I noticed that a mismatch between what this function treats as a
valid truncation and what the backends see can lead to invalid calls as well
(see x86-32 test case).
This patch refactors the code so that instead of being based primarily on
values which it recurses into when necessary, it starts by inspecting the type
and considers each fundamental slot that the backend will see in turn. For
example, given a pathological function that returned {{}, {{}, i32, {}}, i32}
we would consider each "real" i32 in turn, and ask if it passes through
unchanged. This is much closer to what the backend sees as a result of
ComputeValueVTs.
Aside from the bug fixes, this eliminates the recursion that's going on and, I
believe, makes the bulk of the code significantly easier to understand. The
trade-off is the nasty iterators needed to find the real types inside a
returned value.
llvm-svn: 187787
Serge Pavlov [Tue, 6 Aug 2013 08:44:18 +0000 (08:44 +0000)]
Unbreak Debug build on Windows
llvm-svn: 187786
Larisse Voufo [Tue, 6 Aug 2013 07:37:09 +0000 (07:37 +0000)]
Update status of support for variable templates on website.
llvm-svn: 187785
Larisse Voufo [Tue, 6 Aug 2013 07:33:00 +0000 (07:33 +0000)]
A bit of clean up based on peer's feedback...
llvm-svn: 187784
Craig Topper [Tue, 6 Aug 2013 07:23:12 +0000 (07:23 +0000)]
Simplify vector lane handling math a bit. No functional change intended.
llvm-svn: 187783
Richard Smith [Tue, 6 Aug 2013 07:09:20 +0000 (07:09 +0000)]
PR16755: When initializing or modifying a bitfield member in a constant
expression, truncate the stored value to the size of the bitfield.
llvm-svn: 187782
Craig Topper [Tue, 6 Aug 2013 06:54:25 +0000 (06:54 +0000)]
Simplify math a little bit.
llvm-svn: 187781
NAKAMURA Takumi [Tue, 6 Aug 2013 06:38:37 +0000 (06:38 +0000)]
Target/*/CMakeLists.txt: Add the dependency to CommonTableGen explicitly for each corresponding CodeGen.
Without explicit dependencies, both per-file action and in-CommonTableGen action could run in parallel.
It races to emit *.inc files simultaneously.
llvm-svn: 187780
Craig Topper [Tue, 6 Aug 2013 06:05:05 +0000 (06:05 +0000)]
Replace EVT with MVT in isHorizontalBinOp as it is only called with legal types.
llvm-svn: 187779
NAKAMURA Takumi [Tue, 6 Aug 2013 05:56:39 +0000 (05:56 +0000)]
AsmPrinter/CMakeLists.txt: Add explicit dependency to intrinsics_gen here.
llvm-svn: 187778
NAKAMURA Takumi [Tue, 6 Aug 2013 05:56:32 +0000 (05:56 +0000)]
Option/CMakeLists.txt: Don't use target_link_libraries. LLVMBuild knows dependencies.
llvm-svn: 187777
Larisse Voufo [Tue, 6 Aug 2013 05:49:26 +0000 (05:49 +0000)]
Removed hack that was used to properly restore the nested name specifier of qualified variable template ids. It turns out that the current implementation was just not logical setup for it. This commit has made it so.
llvm-svn: 187776
Craig Topper [Tue, 6 Aug 2013 05:41:22 +0000 (05:41 +0000)]
Put an llvm_unreachable at the end of getSplatIndex as its loop should never find all undef elements.
llvm-svn: 187775
Craig Topper [Tue, 6 Aug 2013 05:07:37 +0000 (05:07 +0000)]
Check against >= 0 instead of != -1 in getSplatIndex because it generally compiles to better code and is equivalent for shuffle indices.
llvm-svn: 187774
Craig Topper [Tue, 6 Aug 2013 05:01:21 +0000 (05:01 +0000)]
Remove trailing whitespace and fix an 80-column violation. No functional change.
llvm-svn: 187773
Tobias Grosser [Tue, 6 Aug 2013 04:36:45 +0000 (04:36 +0000)]
ScopInfo: add a testcase that share parameters within nested start.
Contributed-by: Star Tan <tanmx_star@yeah.net>
llvm-svn: 187772
Craig Topper [Tue, 6 Aug 2013 04:12:40 +0000 (04:12 +0000)]
Simplify code slightly. No functional change.
llvm-svn: 187771
Larisse Voufo [Tue, 6 Aug 2013 03:57:41 +0000 (03:57 +0000)]
Fixing commit r187768: Moved diagnosis of forward declarations of variable templates from Parser to Sema.
llvm-svn: 187770
Richard Trieu [Tue, 6 Aug 2013 03:44:10 +0000 (03:44 +0000)]
Fix for PR16570: when comparing two function pointers, discard qualifiers when
comparing non-reference function parameters. The qualifiers don't matter for
comparisons.
llvm-svn: 187769
Larisse Voufo [Tue, 6 Aug 2013 03:43:07 +0000 (03:43 +0000)]
Moved diagnosis of forward declarations of variable templates from Parser to Sema.
llvm-svn: 187768
NAKAMURA Takumi [Tue, 6 Aug 2013 03:31:55 +0000 (03:31 +0000)]
clang/test/Driver/crash-report.c: Mark it as XFAIL:mingw32, for now. Investigating.
I met a message, "fatal error: file 'nul' modified since it was first processed".
llvm-svn: 187767
NAKAMURA Takumi [Tue, 6 Aug 2013 03:29:01 +0000 (03:29 +0000)]
AST/DeclTemplate.h: Possibly fix a warning. [-Wuninitialized]
FIXME: Would '0' be an appropriate value in SequenceNumber?
llvm-svn: 187766
NAKAMURA Takumi [Tue, 6 Aug 2013 03:28:54 +0000 (03:28 +0000)]
SemaTemplateInstantiateDecl.cpp: Suppress a warning. [-Wunused-variable]
llvm-svn: 187765
Tom Stellard [Tue, 6 Aug 2013 02:43:45 +0000 (02:43 +0000)]
Factor FlattenCFG out from SimplifyCFG
Patch by: Mei Ye
llvm-svn: 187764
Eric Christopher [Tue, 6 Aug 2013 01:38:27 +0000 (01:38 +0000)]
Allow 4 as a valid debug info version.
llvm-svn: 187763
Larisse Voufo [Tue, 6 Aug 2013 01:03:05 +0000 (01:03 +0000)]
Started implementing variable templates. Top level declarations should be fully supported, up to some limitations documented as FIXMEs or TODO. Static data member templates work very partially. Static data member templates of class templates need particular attention...
llvm-svn: 187762
Shuxin Yang [Tue, 6 Aug 2013 00:45:32 +0000 (00:45 +0000)]
Add some comment to LTOCodeGenerator class
llvm-svn: 187761
Hans Wennborg [Tue, 6 Aug 2013 00:20:31 +0000 (00:20 +0000)]
clang-cl: Implement the /Tc, /TC, /Tp and /TP options.
These are used to specify source files, and whether to treat source
files as C or C++.
Differential Revision: http://llvm-reviews.chandlerc.com/D1290
llvm-svn: 187760
John Thompson [Tue, 6 Aug 2013 00:17:40 +0000 (00:17 +0000)]
Add test for header guard work-around.
llvm-svn: 187759
Matt Arsenault [Tue, 6 Aug 2013 00:16:21 +0000 (00:16 +0000)]
Fix missing -*- C++ -*-s
llvm-svn: 187758
John Thompson [Mon, 5 Aug 2013 23:55:14 +0000 (23:55 +0000)]
Fixed uninitialized member, plus recommit of 187736.
llvm-svn: 187757
Bill Wendling [Mon, 5 Aug 2013 23:43:18 +0000 (23:43 +0000)]
FileCheckize some of the testcases.
llvm-svn: 187756
Bill Wendling [Mon, 5 Aug 2013 23:29:16 +0000 (23:29 +0000)]
Fix grammar.
llvm-svn: 187755
Tom Stellard [Mon, 5 Aug 2013 22:45:56 +0000 (22:45 +0000)]
R600/SI: Add missing test for r187749
llvm-svn: 187754
David Majnemer [Mon, 5 Aug 2013 22:43:06 +0000 (22:43 +0000)]
[ms-cxxabi] Properly mangle member pointers
There were three things missing from the original implementation:
- We would omit the 'E' qualifier for members int 64-bit mode.
- We would not exmaine the qualifiers in 'IsMember' mode.
- We didn't generate the correct backref to the base class.
llvm-svn: 187753
Eric Christopher [Mon, 5 Aug 2013 22:32:28 +0000 (22:32 +0000)]
Recommit previous cleanup with a fix for c++98 ambiguity.
llvm-svn: 187752
David Majnemer [Mon, 5 Aug 2013 22:26:46 +0000 (22:26 +0000)]
[ms-cxxabi] Handle template-template arguments
Template-template arguments appear to be a rather simple encoding of the
template's templated tag type.
llvm-svn: 187751
Tom Stellard [Mon, 5 Aug 2013 22:22:07 +0000 (22:22 +0000)]
R600: Implement TargetLowering::getVectorIdxTy()
We use MVT::i32 for the vector index type, because we use 32-bit
operations to caculate offsets when dynamically indexing vectors.
llvm-svn: 187749
Tom Stellard [Mon, 5 Aug 2013 22:22:01 +0000 (22:22 +0000)]
TargetLowering: Add getVectorIdxTy() function v2
This virtual function can be implemented by targets to specify the type
to use for the index operand of INSERT_VECTOR_ELT, EXTRACT_VECTOR_ELT,
INSERT_SUBVECTOR, EXTRACT_SUBVECTOR. The default implementation returns
the result from TargetLowering::getPointerTy()
The previous code was using TargetLowering::getPointerTy() for vector
indices, because this is guaranteed to be legal on all targets. However,
using TargetLowering::getPointerTy() can be a problem for targets with
pointer sizes that differ across address spaces. On such targets,
when vectors need to be loaded or stored to an address space other than the
default 'zero' address space (which is the address space assumed by
TargetLowering::getPointerTy()), having an index that
is a different size than the pointer can lead to inefficient
pointer calculations, (e.g. 64-bit adds for a 32-bit address space).
There is no intended functionality change with this patch.
llvm-svn: 187748
Eric Christopher [Mon, 5 Aug 2013 22:07:30 +0000 (22:07 +0000)]
Revert "Use existing builtin hashing functions to make this routine more"
This reverts commit r187745.
llvm-svn: 187747
Arnold Schwaighofer [Mon, 5 Aug 2013 22:03:41 +0000 (22:03 +0000)]
Reverting r187736. It broke some public bots:
http://lab.llvm.org:8011/builders/clang-native-arm-cortex-a9/builds/10248
http://bb.pgr.jp/builders/cmake-clang-i686-mingw32/builds/3327
llvm-svn: 187746
Eric Christopher [Mon, 5 Aug 2013 22:00:50 +0000 (22:00 +0000)]
Use existing builtin hashing functions to make this routine more
simple.
llvm-svn: 187745
Eric Christopher [Mon, 5 Aug 2013 21:40:57 +0000 (21:40 +0000)]
Change parent hashing algorithm to be non-recursive and elaborate
greatly on many comments in the code.
llvm-svn: 187742
David Majnemer [Mon, 5 Aug 2013 21:33:59 +0000 (21:33 +0000)]
[ms-cxxabi] Mangle nullptr template arguments
MSVC mangles nullptr template arguments identically to zero literals.
llvm-svn: 187741
Marshall Clow [Mon, 5 Aug 2013 21:23:28 +0000 (21:23 +0000)]
Implement NULL iterators for <list> re: N3644
llvm-svn: 187740
Michael Gottesman [Mon, 5 Aug 2013 21:07:07 +0000 (21:07 +0000)]
[bugpoint] Allow the user to specify the path to opt on the commandline.
llvm-svn: 187739
Daniel Jasper [Mon, 5 Aug 2013 20:26:17 +0000 (20:26 +0000)]
Add option to disable module loading.
This patch was created by Lawrence Crowl and reviewed in:
http://llvm-reviews.chandlerc.com/D963
llvm-svn: 187738
Hans Wennborg [Mon, 5 Aug 2013 20:14:43 +0000 (20:14 +0000)]
lit.cfg: better check for MSYS
When running the tests under Cygwin using non-Cygwin python,
the platform would be Windows and there would be bash on the path,
so this check for MSYS would not work correctly.
llvm-svn: 187737
John Thompson [Mon, 5 Aug 2013 19:15:50 +0000 (19:15 +0000)]
Fixed incorrect include file exit detection. Added work-around to avoid error on header guard in nested include. Fixed a couple of coding standard issues on variable names.
llvm-svn: 187736
Richard Smith [Mon, 5 Aug 2013 18:49:43 +0000 (18:49 +0000)]
Implement C++'s restrictions on the type of an expression passed to a vararg
function: it can't be 'void' and it can't be an initializer list. We give a
hard error for these rather than treating them as undefined behavior (we can
and probably should do the same for non-POD types in C++11, but as of this
change we don't).
Slightly rework the checking of variadic arguments in a function with a format
attribute to ensure that certain kinds of format string problem (non-literal
string, too many/too few arguments, ...) don't suppress this error.
llvm-svn: 187735
Serge Pavlov [Mon, 5 Aug 2013 18:38:16 +0000 (18:38 +0000)]
Regression test for PR12699
llvm-svn: 187734
Peter Collingbourne [Mon, 5 Aug 2013 17:48:04 +0000 (17:48 +0000)]
Introduce an optimisation for special case lists with large numbers of literal entries.
Our internal regex implementation does not cope with large numbers
of anchors very efficiently. Given a ~3600-entry special case list,
regex compilation can take on the order of seconds. This patch solves
the problem for the special case of patterns matching literal global
names (i.e. patterns with no regex metacharacters). Rather than
forming regexes from literal global name patterns, add them to
a StringSet which is checked before matching against the regex.
This reduces regex compilation time by an order of roughly thousands
when reading the aforementioned special case list, according to a
completely unscientific study.
No test cases. I figure that any new tests for this code should
check that regex metacharacters are properly recognised. However,
I could not find any documentation which documents the fact that the
syntax of global names in special case lists is based on regexes.
The extent to which regex syntax is supported in special case lists
should probably be decided on/documented before writing tests.
Differential Revision: http://llvm-reviews.chandlerc.com/D1150
llvm-svn: 187732
Peter Collingbourne [Mon, 5 Aug 2013 17:47:59 +0000 (17:47 +0000)]
Introduce Regex::isLiteralERE function.
This will be used to implement an optimisation for literal entries
in special case lists.
Differential Revision: http://llvm-reviews.chandlerc.com/D1278
llvm-svn: 187731
Andrew Kaylor [Mon, 5 Aug 2013 17:12:35 +0000 (17:12 +0000)]
Additional fixes/xfails for icc tests
llvm-svn: 187730
Jordan Rose [Mon, 5 Aug 2013 16:02:02 +0000 (16:02 +0000)]
[analyzer] Clarify that r187624 is a hack and should be fixed better later.
Tracked by <rdar://problem/
14648821>.
llvm-svn: 187729
Tobias Grosser [Mon, 5 Aug 2013 15:14:15 +0000 (15:14 +0000)]
ScopInfo: Split start value from SCEVAddRecExpr to enable parameter sharing.
SCoP invariant parameters with the different start value would deter parameter
sharing. For example, when compiling the following C code:
void foo(float *input) {
for (long j = 0; j < 8; j++) {
// SCoP begin
for (long i = 0; i < 8; i++) {
float x = input[j * 64 + i + 1];
input[j * 64 + i] = x * x;
}
}
}
Polly would creat two parameters for these memory accesses:
p_0: {0,+,256}
p_2: {4,+,256}
[j * 64 + i + 1] => MemRef_input[o0] : 4o0 = p_1 + 4i0
[j * 64 + i] => MemRef_input[o0] : 4o0 = p_0 + 4i0
These parameters only differ from start value. To enable parameter sharing,
we split the start value from SCEVAddRecExpr, so they would share a single
parameter that always has zero start value:
p0: {0,+,256}<%for.cond1.preheader>
[j * 64 + i + 1] => MemRef_input[o0] : 4o0 = 4 + p_1 + 4i0
[j * 64 + i] => MemRef_input[o0] : 4o0 = p_0 + 4i0
Such translation can make the polly-dependence much faster.
Contributed-by: Star Tan <tanmx_star@yeah.net>
llvm-svn: 187728
Aaron Ballman [Mon, 5 Aug 2013 13:47:03 +0000 (13:47 +0000)]
Silencing an MSVC11 type conversion warning.
llvm-svn: 187727
Alexey Samsonov [Mon, 5 Aug 2013 13:20:39 +0000 (13:20 +0000)]
Lint fixes
llvm-svn: 187726
Alexey Samsonov [Mon, 5 Aug 2013 13:19:49 +0000 (13:19 +0000)]
80-cols
llvm-svn: 187725
Elena Demikhovsky [Mon, 5 Aug 2013 12:17:06 +0000 (12:17 +0000)]
LLVM Interpreter: This patch implements vector support for cast operations (zext, sext, uitofp, sitofp, trunc, fpext, fptosi, fptrunc, bitcast) and shift operations (shl, ashr, lshr) for integer and floating point data types.
Added tests.
Done by Yuri Veselov (mailto:Yuri.Veselov@intel.com).
llvm-svn: 187724
Richard Sandiford [Mon, 5 Aug 2013 11:23:46 +0000 (11:23 +0000)]
[SystemZ] Use BRCT and BRCTG to eliminate add-&-compare sequences
This patch just uses a peephole test for "add; compare; branch" sequences
within a single block. The IR optimizers already convert loops to
decrement-and-branch-on-nonzero form in some cases, so even this
simplistic test triggers many times during a clang bootstrap and
projects/test-suite run. It looks like there are still cases where we
need to more strongly prefer branches on nonzero though. E.g. I saw a
case where a loop that started out with a check for 0 ended up with a
check for -1. I'll try to look at that sometime.
I ended up adding the Reference class because MachineInstr::readsRegister()
doesn't check for subregisters (by design, as far as I could tell).
llvm-svn: 187723
Benjamin Kramer [Mon, 5 Aug 2013 11:11:11 +0000 (11:11 +0000)]
Don't leak passes if added outside of the area determined by Started/Stopped flags.
llvm-svn: 187722
Richard Sandiford [Mon, 5 Aug 2013 11:07:38 +0000 (11:07 +0000)]
[SystemZ] Add definitions for BRCT and BRCTG
llvm-svn: 187721
Richard Sandiford [Mon, 5 Aug 2013 11:03:20 +0000 (11:03 +0000)]
[SystemZ] Use LOAD AND TEST to eliminate comparisons against zero
llvm-svn: 187720
Richard Sandiford [Mon, 5 Aug 2013 11:00:53 +0000 (11:00 +0000)]
[SystemZ] Add LOAD AND TEST instructions
Just the definitions and MC support. The next patch uses them for codegen.
llvm-svn: 187719
Richard Sandiford [Mon, 5 Aug 2013 10:58:53 +0000 (10:58 +0000)]
[SystemZ] Split out comparison elimination into a separate pass
Perhaps predictably, doing comparison elimination on the fly during
SystemZLongBranch turned out to be a bad idea. The next patches make
use of LOAD AND TEST and BRANCH ON COUNT, both of which require
changes to earlier instructions.
No functionality change intended.
llvm-svn: 187718
Elena Demikhovsky [Mon, 5 Aug 2013 08:52:21 +0000 (08:52 +0000)]
AVX-512 set: added mask operations, lowering BUILD_VECTOR for i1 vector types.
Added intrinsics and tests.
llvm-svn: 187717
Craig Topper [Mon, 5 Aug 2013 06:17:21 +0000 (06:17 +0000)]
Use a shuffle with undef elements instead of inserting 0s in the 128-bit to 256-bit casting intrinsics to improve performance. Thanks to Katya Romanova for identifying this issue.
llvm-svn: 187716
David Majnemer [Mon, 5 Aug 2013 04:53:41 +0000 (04:53 +0000)]
Sema: Don't assume a nested name specifier holds a type
Sema::PerformObjectMemberConversion assumed that the Qualifier it was
given holds a type. However, the specifier could hold just a namespace.
In this case, we should ignore the qualifier and not attempt to cast to
it.
llvm-svn: 187715
Nadav Rotem [Mon, 5 Aug 2013 04:31:05 +0000 (04:31 +0000)]
Update the release notes about the status of the vectorizers.
llvm-svn: 187714
Nadav Rotem [Mon, 5 Aug 2013 04:27:34 +0000 (04:27 +0000)]
Update the docs.
llvm-svn: 187713
Reed Kotler [Sun, 4 Aug 2013 23:56:53 +0000 (23:56 +0000)]
Add the saving of S2. This is needed for some of the floating point
helper functions. This can be optimized out later when the remaining
parts of the helper function work is moved into the Mips16HardFloat pass.
For now it forces us to use the 32 bit save/restore instructions instead
of the 16 bit ones.
llvm-svn: 187712
Bob Wilson [Sun, 4 Aug 2013 23:55:24 +0000 (23:55 +0000)]
Remove "lto_on_osx" xfails, now that -rdynamic works on Darwin.
Note that this will require a recent version of the linker for Darwin
builds with LTO to pass these tests.
llvm-svn: 187711
Bob Wilson [Sun, 4 Aug 2013 22:06:11 +0000 (22:06 +0000)]
Build with the $RDYNAMIC flag on Darwin as well as other platforms.
Part of <rdar://problem/
14620988>
llvm-svn: 187710
Timur Iskhodzhanov [Sun, 4 Aug 2013 17:30:04 +0000 (17:30 +0000)]
Emit the constructor for abstract classes when using -cxx-abi microsoft, fixes PR16735
llvm-svn: 187709
Benjamin Kramer [Sun, 4 Aug 2013 16:06:43 +0000 (16:06 +0000)]
clang-tidy's modules depend on it.
Fixes shared cmake build.
llvm-svn: 187708
Daniel Jasper [Sun, 4 Aug 2013 15:56:30 +0000 (15:56 +0000)]
Fix clang-tidy dependencies and bad file comment.
This addresses comments in post-commit review of r187345.
llvm-svn: 187707
Benjamin Kramer [Sun, 4 Aug 2013 12:05:16 +0000 (12:05 +0000)]
X86: Turn fp selects into mask operations.
double test(double a, double b, double c, double d) { return a<b ? c : d; }
before:
_test:
ucomisd %xmm0, %xmm1
ja LBB0_2
movaps %xmm3, %xmm2
LBB0_2:
movaps %xmm2, %xmm0
after:
_test:
cmpltsd %xmm1, %xmm0
andpd %xmm0, %xmm2
andnpd %xmm3, %xmm0
orpd %xmm2, %xmm0
Small speedup on Benchmarks/SmallPT
llvm-svn: 187706
Elena Demikhovsky [Sun, 4 Aug 2013 10:46:07 +0000 (10:46 +0000)]
AVX-512 set: added VEXTRACTPS instruction
llvm-svn: 187705
Tim Northover [Sun, 4 Aug 2013 10:00:45 +0000 (10:00 +0000)]
X86: specify CPU on new test to fix atom buildbot
Apparently Atoms use lea for stack adjustment, which we weren't
looking for.
llvm-svn: 187704
Tim Northover [Sun, 4 Aug 2013 09:35:57 +0000 (09:35 +0000)]
X86: correct tail return address calculation
Due to the weird and wondeful usual arithmetic conversions, some
calculations involving negative values were getting performed in
uint32_t and then promoted to int64_t, which is really not a good
idea.
Patch by Katsuhiro Ueno.
llvm-svn: 187703
Benjamin Kramer [Sun, 4 Aug 2013 09:06:29 +0000 (09:06 +0000)]
AsmParser: Store MacroLikeBodies on the side so they don't get leaked.
llvm-svn: 187702
Reed Kotler [Sun, 4 Aug 2013 01:13:25 +0000 (01:13 +0000)]
Clean up code for Mips16 large frame handling.
llvm-svn: 187701
Benjamin Kramer [Sat, 3 Aug 2013 22:43:29 +0000 (22:43 +0000)]
PPCAsmParser: Stop leaking names.
Store them in a place that gets cleaned up properly.
llvm-svn: 187700
Benjamin Kramer [Sat, 3 Aug 2013 22:18:45 +0000 (22:18 +0000)]
Unbreak llvm-rtdyld build.
llvm-svn: 187699
Benjamin Kramer [Sat, 3 Aug 2013 22:16:37 +0000 (22:16 +0000)]
MachObjectFile: Don't leak on error.
llvm-svn: 187698
Benjamin Kramer [Sat, 3 Aug 2013 22:16:31 +0000 (22:16 +0000)]
llvm-rtdyld: Don't leak memory managers.
Dyld never outlives MemMgr, just put both on the stack.
llvm-svn: 187697
Benjamin Kramer [Sat, 3 Aug 2013 22:16:24 +0000 (22:16 +0000)]
ARMAsmParser: Plug a leak.
Using an object to do the cleanup may look like overkill, but it's safer and nicer than putting deletes everywhere.
llvm-svn: 187696
Benjamin Kramer [Sat, 3 Aug 2013 22:16:16 +0000 (22:16 +0000)]
Stop leaking register infos in the disassemblers.
llvm-svn: 187695
Craig Topper [Sat, 3 Aug 2013 17:40:38 +0000 (17:40 +0000)]
Add support for passing -1 to __builtin_shufflevector to signify an undefined element value to match IR capabilities.
llvm-svn: 187694