Author: ctopper
Date: Mon Oct 31 01:24:00 2016
New Revision: 285547
URL: http://llvm.org/viewvc/llvm-project?rev=285547&view=rev
Log:
[AVX-512] Add a regular expression to a test that was missed in r285540.
Modified:
cfe/trunk/test/CodeGen/avx512dq-builtins.c
Modified: cfe/trunk/test/CodeGen
Author: ctopper
Date: Mon Oct 31 00:49:11 2016
New Revision: 285545
URL: http://llvm.org/viewvc/llvm-project?rev=285545&view=rev
Log:
[AVX-512] Use selectd instead of selectps for _mm256_mask_extracti32x4_epi32.
Modified:
cfe/trunk/lib/Headers/avx512vlintrin.h
cfe/trunk/test/CodeGen/avx51
Author: majnemer
Date: Mon Oct 31 00:37:48 2016
New Revision: 285544
URL: http://llvm.org/viewvc/llvm-project?rev=285544&view=rev
Log:
Add support for __builtin_alloca_with_align
__builtin_alloca always uses __BIGGEST_ALIGNMENT__ for the alignment of
the allocation. __builtin_alloca_with_align a
Author: sepavloff
Date: Mon Oct 31 00:11:12 2016
New Revision: 285543
URL: http://llvm.org/viewvc/llvm-project?rev=285543&view=rev
Log:
Make output of ast-print closer to C++ code
Put semicolon after non-defining method declaration and a class
specialization body.
Modified:
cfe/trunk/lib/AST
Author: ctopper
Date: Sun Oct 30 23:30:51 2016
New Revision: 285539
URL: http://llvm.org/viewvc/llvm-project?rev=285539&view=rev
Log:
[AVX-512] Remove many of the masked 128/256-bit shift builtins and replace them
with unmasked builtins and selects.
Modified:
cfe/trunk/include/clang/Basic/Bu
Author: ctopper
Date: Sun Oct 30 23:30:56 2016
New Revision: 285540
URL: http://llvm.org/viewvc/llvm-project?rev=285540&view=rev
Log:
[AVX-512] Remove masked vector extract builtins and replace with native
shufflevectors and selects.
Unfortunately, the backend currently doesn't fold masks into t
Author: ericwf
Date: Sun Oct 30 22:42:50 2016
New Revision: 285538
URL: http://llvm.org/viewvc/llvm-project?rev=285538&view=rev
Log:
Add 'inline' but not 'always_inline' to std::strings destructor.
Adding both 'inline' and 'always_inline' to the destructor has been contentious.
However most of th
Author: ericwf
Date: Sun Oct 30 22:40:29 2016
New Revision: 285537
URL: http://llvm.org/viewvc/llvm-project?rev=285537&view=rev
Log:
Remove additional function template definitions from the dylib
Modified:
libcxx/trunk/include/string
libcxx/trunk/lib/abi/CHANGELOG.TXT
libcxx/trunk/lib
mehdi_amini added a comment.
I talked with Eric on IRC, he mentioned some benchmarks were ran. I'd like to
understand what was the baseline?
Here we add *both* the inline keyword and the always_inline attribute. I'd like
to know if there is a benchmarks that shows that always_inline is beneficia
loladiro added a comment.
I'm really out of my depth in this code, but it looks like that test case is
triggering the one code path in that function that is actually correct. Could
you adjust it to trigger the code path behind the first if statement?
Repository:
rL LLVM
https://reviews.llvm
Author: dergachev
Date: Sun Oct 30 22:08:48 2016
New Revision: 285533
URL: http://llvm.org/viewvc/llvm-project?rev=285533&view=rev
Log:
[analyzer] NumberObjectConversion: support more types, misc updates.
Support CFNumberRef and OSNumber objects, which may also be accidentally
converted to plain
This revision was automatically updated to reflect the committed changes.
Closed by commit rL285533: [analyzer] NumberObjectConversion: support more
types, misc updates. (authored by dergachev).
Changed prior to commit:
https://reviews.llvm.org/D25731?vs=76043&id=76349#toc
Repository:
rL LLV
mehdi_amini added a comment.
Before this gets re-committed, I'd like to understand why only the destructor?
https://reviews.llvm.org/D25624
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-co
Author: ericwf
Date: Sun Oct 30 21:46:25 2016
New Revision: 285532
URL: http://llvm.org/viewvc/llvm-project?rev=285532&view=rev
Log:
Optimize filesystem::path by providing weaker exception guarantees.
path uses string::append to construct, append, and concatenate paths.
Unfortunatly
string::appe
Hi All,
This problem was caused by the definition of
_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY.
The issue was fixed in r285531, which also adds linkage tests for
std::strings destructor.
@Adtiya feel free to recommit this change.
/Eric
On Fri, Oct 28, 2016 at 7:00 PM, Mehdi Amini via cfe-commit
Author: ericwf
Date: Sun Oct 30 21:07:23 2016
New Revision: 285531
URL: http://llvm.org/viewvc/llvm-project?rev=285531&view=rev
Log:
Fix _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to always have default visibility.
This prevent the symbols from being both externally available and hidden, which
cau
dvyukov added inline comments.
Comment at: lib/CodeGen/CodeGenFunction.cpp:717
+ // Ignore TSan memory acesses from within dealloc and all of its calees at
+ // run time.
This is unpleasant. We had recursive ignores in the old Valgrind-based tsan,
but manage
sammccall updated this revision to Diff 76348.
sammccall added a comment.
Fixing style.
https://reviews.llvm.org/D26132
Files:
lib/Format/TokenAnnotator.cpp
unittests/Format/FormatTest.cpp
Index: unittests/Format/FormatTest.cpp
=
sammccall created this revision.
sammccall added a reviewer: djasper.
sammccall added a subscriber: cfe-commits.
Herald added a subscriber: klimek.
Skip over AnnotatedLines with >50 levels of nesting; don't format them.
Reasoning:
- ExpressionParser uses a lot of stack for these, bad in some env
Author: ericwf
Date: Sun Oct 30 18:53:50 2016
New Revision: 285530
URL: http://llvm.org/viewvc/llvm-project?rev=285530&view=rev
Log:
Improve performance of constructing filesystem::path from strings.
This patch fixes a performance bug when constructing or appending to a path
from a string or c-st
Author: ericwf
Date: Sun Oct 30 18:30:38 2016
New Revision: 285526
URL: http://llvm.org/viewvc/llvm-project?rev=285526&view=rev
Log:
Rewrite std::filesystem::path iterators and parser
This patch entirely rewrites the parsing logic for paths. Unlike the previous
implementation this one stores info
Author: mehdi_amini
Date: Sun Oct 30 18:26:13 2016
New Revision: 285525
URL: http://llvm.org/viewvc/llvm-project?rev=285525&view=rev
Log:
Fix clang installed path to handle case where clang is invoked through a symlink
This code path is used when generating the path to libLTO.dylib, which
is pass
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM as well.
https://reviews.llvm.org/D26075
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listi
Author: ericwf
Date: Sun Oct 30 17:53:00 2016
New Revision: 285524
URL: http://llvm.org/viewvc/llvm-project?rev=285524&view=rev
Log:
Add start of filesystem benchmarks
Added:
libcxx/trunk/benchmarks/filesystem.bench.cpp
Modified:
libcxx/trunk/CMakeLists.txt
libcxx/trunk/benchmarks/CMa
jbcoe removed rL LLVM as the repository for this revision.
jbcoe updated this revision to Diff 76344.
jbcoe added a comment.
Remove mistakenly committed debugging output.
https://reviews.llvm.org/D26082
Files:
bindings/python/clang/cindex.py
bindings/python/tests/cindex/test_translation_uni
rmaprath added a comment.
This looks sensible to me. I'll let @EricWF approve.
/ Asiri
https://reviews.llvm.org/D26075
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
> As I said, this is not a decision based on the libc++ version.
It is at least partially based on the libc++ version. We obviously can't
enable the feature if libc++ does not yet support them.
/Eric
On Sun, Oct 30, 2016 at 2:00 PM, Joerg Sonnenberger via cfe-commits <
cfe-commits@lists.llvm.org
daphnediane added a comment.
ping.
FYI I have found some additional similar cases when
AlignConsecutiveDeclarations is also true, that I'll see if I can document/test
case etc. at some point.
https://reviews.llvm.org/D24703
___
cfe-commits mailin
Author: ericwf
Date: Sun Oct 30 15:05:52 2016
New Revision: 285523
URL: http://llvm.org/viewvc/llvm-project?rev=285523&view=rev
Log:
Mark thread exit test as unsupported w/o threads
Modified:
libcxxabi/trunk/test/cxa_thread_atexit_test.pass.cpp
Modified: libcxxabi/trunk/test/cxa_thread_atexi
idlecode added a comment.
@mgehre: Yes it does - in your case AST looks like `ifStmt(b) - CompoundStmt -
ifStmt(c)`, so ifStmt's don't have direct parent-child relations.
But there is another problem:
while(a)
if (b)
return 1;
else
doSomething(2);
In such case, generated f
On Sun, Oct 30, 2016 at 01:49:53PM -0600, Eric Fiselier via cfe-commits wrote:
> > E.g. presence of libc++ won't tell you if you can use sized deallocation
> > as that's a ABI library issue.
>
> The value of __libcpp_version could easily be updated by vendors to store
> the version of the system
> E.g. presence of libc++ won't tell you if you can use sized deallocation
as that's a ABI library issue.
The value of __libcpp_version could easily be updated by vendors to store
the version of the system dylib,
so I don't see why this wouldn't work.
> So, why again should clang care about the
On Sun, Oct 30, 2016 at 12:54:28PM -0600, Eric Fiselier wrote:
> Richard requested this change. Take a look at
> https://reviews.llvm.org/D26044 for more rational.
I don't see much rational and in fact, I disagree with some of the
mentioned items. E.g. presence of libc++ won't tell you if you can
mgehre added a comment.
With this fix, is there still a warning on the following code?
if(b) {
...
if(c)
return;
else
doSomething()
...
}
I would expect that the check still warns on it.
https://reviews.llvm.org/D26125
Richard requested this change. Take a look at
https://reviews.llvm.org/D26044 for more rational.
On Sat, Oct 29, 2016 at 2:28 PM, Joerg Sonnenberger via cfe-commits <
cfe-commits@lists.llvm.org> wrote:
> On Fri, Oct 28, 2016 at 06:06:50AM -, Eric Fiselier via cfe-commits
> wrote:
> > Author:
jbcoe retitled this revision from "Incomplete support for Python 3 in libclang
python bindings" to "Support for Python 3 in libclang python bindings".
jbcoe updated the summary for this revision.
jbcoe updated this revision to Diff 76338.
jbcoe added a comment.
Python bindings tests now pass in P
djasper added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
idlecode wrote:
> djasper wrote:
> >
idlecode planned changes to this revision.
idlecode added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
--
djasper added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
idlecode wrote:
> djasper wrote:
> >
idlecode marked an inline comment as done.
idlecode added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
--
djasper added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
idlecode wrote:
> djasper wrote:
> >
idlecode marked an inline comment as done.
idlecode added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
--
Author: mzuckerm
Date: Sun Oct 30 09:54:05 2016
New Revision: 285519
URL: http://llvm.org/viewvc/llvm-project?rev=285519&view=rev
Log:
Fixing "type" issue for (epi32)
and replaceing hardcoded inf with clang builtin inf "__builtin_inff()" for
float ({max|min}_{pd|ps})
Modified:
cfe/trunk/l
djasper added inline comments.
Comment at: clang-tidy/readability/ElseAfterReturnCheck.cpp:28
stmt(forEach(
- ifStmt(hasThen(stmt(
+ ifStmt(unless(hasParent(ifStmt())),
+ hasThen(stmt(
I think this now effectively does:
idlecode created this revision.
idlecode added reviewers: djasper, malcolm.parsons, omtcyfz.
idlecode added a subscriber: cfe-commits.
Fix applied to else in a cascade if statements changed
program semantics in cases when not all previous branches
resulted in flow change.
Fixes PR30652
https://
Author: ctopper
Date: Sun Oct 30 02:11:34 2016
New Revision: 285516
URL: http://llvm.org/viewvc/llvm-project?rev=285516&view=rev
Log:
[AVX-512] Remove masked 128/256-bit builtins for vpmaddwd and vpmaddubsw.
Replace with unmasked builtins and select.
Modified:
cfe/trunk/include/clang/Basic/B
46 matches
Mail list logo
