A couple improvements:
1) Add ability to limit fullMask size when KMP_HW_SUBSET limits resources.
2) Make KMP_HW_SUBSET work for affinity_none, and only limit fullMask in this case.
Patch by Andrey Churbanov.
Differential Revision: http://reviews.llvm.org/D21528
llvm-svn: 273278
Summary:
Code generation for Cortex-A72/Cortex-A73 was accidentally changed
by r271555, which was a NFCI. The isCortexA57() predicate was not true
for Cortex-A72/Cortex-A73 before r271555 (since it was checking the CPU
string). Because Cortex-A72/Cortex-A73 inherit all features from Cortex-A57,
all decisions previously guarded by isCortexA57() are now taken.
This change restores the behaviour before r271555 by adding separate
ProcA72/ProcA73, which have the required features to preserve code
generation.
Reviewers: kristof.beyls, aadg, mcrosier, rengolin
Subscribers: mcrosier, llvm-commits, aemerson, t.p.northover, MatzeB, rengolin
Differential Revision: http://reviews.llvm.org/D21182
llvm-svn: 273277
There was a segfault in the stubs library in posix_memalign because
of a bad parameter. The fix is to send address of the pointer as a
parameter. Also added check of result of posix_memalign.
Patch by Andrey Churbanov.
Differential Revision: http://reviews.llvm.org/D21529
llvm-svn: 273276
This change appends the process id to the KMP_STATS_FILE (if specified) which
enables MPI processes to output their stats to separate files.
Differential Revision: http://reviews.llvm.org/D21386
llvm-svn: 273273
Summary:
We have switched to using features for all heuristics, but
the tests for these are still using -mcpu, which means we
are not directly testing the features.
This converts at least some of the existing regression tests
to use the new features.
This still leaves the following features untested:
merge-narrow-ld
predictable-select-expensive
alternate-sextload-cvt-f32-pattern
disable-latency-sched-heuristic
Reviewers: mcrosier, t.p.northover, rengolin
Subscribers: MatzeB, aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D21288
llvm-svn: 273271
This is part of the effort for asan to support Windows 64 bit.
Patch by Wei Wang
Differential Revision: http://reviews.llvm.org/D21525
llvm-svn: 273270
The large offset is being tested on Windows 10 (which has larger usable
virtual address space than Windows 8 or earlier)
Patch by: Wei Wang
Differential Revision: http://reviews.llvm.org/D21523
llvm-svn: 273269
When you have a map holding a unique_ptr, hold a reference to the raw
pointer instead of the unique pointer. The unique_ptr will be moved on
rehash.
llvm-svn: 273268
Summary:
The MSVC compiler complains about implicit conversion of 32-bits constant to
64-bit when using this shiting pattern 1 << (<64-bit expr>).
Reviewers: rnk
Subscribers: kcc, llvm-commits, wang0109, kubabrecka, chrisha
Differential Revision: http://reviews.llvm.org/D21524
llvm-svn: 273267
Summary:
The MSVC compiler is complaining about invalid flags. The CFLAGS are passed
to the linker (i.e. /Gy-, ...).
Reviewers: rnk
Subscribers: llvm-commits, kubabrecka, chrisha
Differential Revision: http://reviews.llvm.org/D21554
llvm-svn: 273265
This patch adds the __sanitizer_start_switch_fiber and
__sanitizer_finish_switch_fiber methods inspired from what can be found here
2ea64dd249 .
These methods are needed when the compiled software needs to implement
coroutines, fibers or the like. Without a way to annotate them, when the program
jumps to a stack that is not the thread stack, __asan_handle_no_return shows a
warning about that, and the fake stack mechanism may free fake frames that are
still in use.
Author: blastrock (Philippe Daouadi)
Reviewed in http://reviews.llvm.org/D20913
llvm-svn: 273260
Summary:
canCombineSinCosLibcall() would previously combine sin+cos into sincos for
GNUX32/GNUEABI/GNUEABIHF regardless of whether UnsafeFPMath were set or not.
However, GNU would only combine them for UnsafeFPMath because sincos does not
set errno like sin and cos do. It seems likely that this is an oversight.
Reviewers: t.p.northover
Subscribers: t.p.northover, aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D21431
llvm-svn: 273259
It did not handle correctly cases without GEP.
The following loop wasn't vectorized:
for (int i=0; i<len; i++)
*to++ = *from++;
I use getPtrStride() to find Stride for memory access and return 0 is the Stride is not 1 or -1.
Differential revision: http://reviews.llvm.org/D20789
llvm-svn: 273257
This is a follow-up patch to r273218. GNU ld accepts both "--" and "-"
for all multi-letter options except "-o". This patch makes lld compatible
with that behavior.
llvm-svn: 273256
Summary:
Using isOutOfOrder makes the code more clear.
Reviewers: rengolin, atrick, hfinkel.
Differential Revision: http://reviews.llvm.org/D21548
llvm-svn: 273255
On PowerPC, if binutils and glibc are new enough, the linker uses
an optimized code sequence to implement __tls_get_addr call stub,
which will end up calling __tls_get_addr_opt instead of __tls_get_addr.
Thus, we need to intercept it in addition to __tls_get_addr.
This symbol is actually an alias of __tls_get_addr - its only purpose
is that its presence in glibc triggers the optimization in linker.
This means we can make our own intercepting symbol an alias as well.
This patch will make the linker attempt optimization even on older
glibc's (since it sees a defined __tls_get_addr_opt symbol in msan)
- however, this is only a very minor performance problem (the linker
generated code will never recognize a filled static TLS descriptor,
always burning a few cycles), not a correctness problem.
This fixes MSan's dtls_test.c, allowing us to finally enable MSan
on PowerPC64.
llvm-svn: 273250
`Inst` and `Op` variables are removed since they are not always
point to an instruction nor an operand. For 5-byte MOV instruction,
Op points to an instruction, which is confusing.
llvm-svn: 273246
This reverts commit r273019.
From email I sent to list:
> I don't think this makes sense. Either the linker you're using supports
> this feature, or it doesn't. Having it enabled for llc if your linker
> doesn't support it is not fun.
>
> Further note that this also affects basically all other code using llvm
> libraries -- other than Clang, which explicitly sets it back to false by
> default, unless you set the ENABLE_X86_RELAX_RELOCATIONS cmake flag to
> true.
>
> If you want to enable the relax mode across all llvm tools in some
> circumstances, I think it should be via moving the cmake flag from clang
> down into llvm.
>
> I'm going to revert this commit, since I both think it intrinsically
> doesn't make sense to do this, and because it's breaking some of our
> tools.
llvm-svn: 273245
This patch is to rewrite the function with a table-lookup-ish approach
so that it can read as a series of "convert this pattern to this" pattern.
llvm-svn: 273238
The export side is responsible for running any initializers, they are
run when the module is first loaded. Attempting to run an initializer
for the import side is not possible.
This fixes PR28216.
llvm-svn: 273237
which doesn't like against all the extra UI frameworks on ios)
so it now generates a binary called "debugserver-nonui" and puts
it in /usr/local/bin instead of /Developer/usr/bin.
Add some cruft to RNBDefs.h to get the version number (provided
by Xcode at build time) with either the name "debugserver" or
"debugserver_nonui" as appropriate.
Add the "debugserver-mini" target to the top level "ios" target
in lldb xcode project file, so this nonui debugserver will be
built along with the normal lldb / debugserver.
<rdar://problem/24730789>
llvm-svn: 273236
A number of warnings still remain, but these were the last of the
"unlexable code"-related ones (AFAICT).
I changed a few examples in docs/UsersManual.rst to showcase
-Wextra-tokens because it's already documented (-Wmultichar isn't), and
the sphinx C lexer apparently can't handle char literals like 'ab'. It
seemed like a better overall approach than just marking the code blocks
as none or console.
llvm-svn: 273232
Evandro Menezes