, Eric Fiselier via cfe-commits
> <cfe-commits@lists.llvm.org> wrote:
> > Cleanup node-type handling in the unordered containers
>
> This seems to have broken __gnu_cxx::hash_map (used by hash.cpp and
> hash2.cpp in the test-suite). A smaller reproducer is:
>
> #include
Author: ericwf
Date: Mon Feb 8 17:47:13 2016
New Revision: 260172
URL: http://llvm.org/viewvc/llvm-project?rev=260172=rev
Log:
Revert r260012 due to __gnu_cxx::hash_map breakage
Removed:
libcxx/trunk/test/libcxx/containers/unord/key_value_traits.pass.cpp
>> template class _LIBCPP_TYPE_VIS_ONLY
>> __hash_map_iterator;
>> template class _LIBCPP_TYPE_VIS_ONLY
>> __hash_map_const_iterator;
>>
>> should not they be prefixed with __gnu_cxx:: ?
>>
>> Clang says that std::__1::__hash_const_iterator and
>
::__1::__hash_const_iterator and
> __gnu_cxx::__hash_map_const_iterator are not friends and
> __non_const_iterator at ext/hash_map:432 is private.
>
>
> On Wed, Feb 10, 2016 at 12:46 PM, Eric Fiselier via cfe-commits
> <cfe-commits@lists.llvm.org> wrote:
> > Author: ericwf
> > Date
EricWF added inline comments.
Comment at: include/experimental/iterator:85
@@ +84,3 @@
+template
+ostream_joiner& operator=(const _Tp& __v)
+{
EricWF wrote:
> People are going to try and use this as an assignment operator. We either
> need to static
EricWF added a comment.
- Please add the header to `test/libcxx/double_include.sh.cpp`
- Please add a `_LIBCPP_VERSION` test in `test/libcxx/experimental/iterator`
- The `ostream_joiner` type and it's members should be given visibility
attributes.
Comment at:
EricWF added a comment.
One last thing. The spec implicitly defines the copy/move/assignment operators
for "ostream_joiner" by making "_Delim" an exposition only member. Essentially
"ostream_joiner" should only copy/moveable if "_Delim" is. I think we should
add tests that ensure we actually
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM.
http://reviews.llvm.org/D16605
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
EricWF added a comment.
I'm OK signing off on this myself.
http://reviews.llvm.org/D16584
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
EricWF added a comment.
ping.
http://reviews.llvm.org/D13420
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
EricWF added a comment.
> This code triggers CFI warnings.
What are CFI warnings? Sorry if that's a dumb question, I just don't recognize
"CFI".
Repository:
rL LLVM
http://reviews.llvm.org/D16738
___
cfe-commits mailing list
EricWF added a comment.
> This also could be fixed in a different way by replacing C-style
> casts with reinterpret_cast<>, which, from my reading of the
> standard, is allowed in this context.
I agree that using `void*` to represent raw memory is the better approach than
Author: ericwf
Date: Tue Jan 19 15:58:49 2016
New Revision: 258198
URL: http://llvm.org/viewvc/llvm-project?rev=258198=rev
Log:
Add more missing license headers
Modified:
libcxx/trunk/src/include/atomic_support.h
libcxx/trunk/utils/gen_link_script/gen_link_script.py
;> >> -I
> >>> >>
> >>> >>
> /Users/thakis/src/chrome/src/third_party/llvm/projects/libcxxabi/include
> >>> >> -stdlib=libc++ -std=c++1z -fdeprecated-macro -fdebug-compilation-dir
> >>> >>
> >>> >>
> /Users/thakis/s
Author: ericwf
Date: Tue Jan 19 16:07:10 2016
New Revision: 258201
URL: http://llvm.org/viewvc/llvm-project?rev=258201=rev
Log:
Add missing license headers
Modified:
libcxxabi/trunk/test/libcxxabi/test/config.py
libcxxabi/trunk/test/support/timer.hpp
Modified:
Author: ericwf
Date: Tue Jan 19 15:51:07 2016
New Revision: 258195
URL: http://llvm.org/viewvc/llvm-project?rev=258195=rev
Log:
Mark slow ASAN/MSAN tests as XFAIL for now.
Modified:
Author: ericwf
Date: Tue Jan 19 15:52:04 2016
New Revision: 258196
URL: http://llvm.org/viewvc/llvm-project?rev=258196=rev
Log:
Add missing license headers
Modified:
libcxx/trunk/test/libcxx/compiler.py
libcxx/trunk/test/libcxx/test/config.py
libcxx/trunk/test/libcxx/test/executor.py
I'll merge all my commits tonight. Thanks Hans.
/Eric
On Wed, Jan 20, 2016 at 10:38 AM, Hans Wennborg wrote:
> No problem. Please go ahead and merge with utils/release/merge.sh, or
> let me know if you'd prefer me to do it.
>
> On Tue, Jan 19, 2016 at 7:09 PM, Eric Fiselier
EricWF added a comment.
ping.
http://reviews.llvm.org/D13420
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: ericwf
Date: Thu Jan 21 19:12:23 2016
New Revision: 258469
URL: http://llvm.org/viewvc/llvm-project?rev=258469=rev
Log:
Merging r258249:
r258249 | ericwf | 2016-01-19 16:42:10 -0700 (Tue, 19 Jan 2016) | 7 lines
Author: ericwf
Date: Thu Jan 21 19:13:27 2016
New Revision: 258470
URL: http://llvm.org/viewvc/llvm-project?rev=258470=rev
Log:
Merging r258277:
r258277 | ericwf | 2016-01-19 20:04:50 -0700 (Tue, 19 Jan 2016) | 1 line
Fix
Hi Hans,
CC Marshall
FYI All my pending libc++ and libc++abi commits have been pushed to the
release_38 branch.
/Eric
On Thu, Jan 21, 2016 at 1:51 PM, Eric Fiselier wrote:
> I'll merge all my commits tonight. Thanks Hans.
>
> /Eric
>
> On Wed, Jan 20, 2016 at 10:38 AM, Hans
Author: ericwf
Date: Thu Jan 21 19:10:25 2016
New Revision: 258467
URL: http://llvm.org/viewvc/llvm-project?rev=258467=rev
Log:
Merging r258198:
r258198 | ericwf | 2016-01-19 14:58:49 -0700 (Tue, 19 Jan 2016) | 1 line
Add
Author: ericwf
Date: Thu Jan 21 19:11:34 2016
New Revision: 258468
URL: http://llvm.org/viewvc/llvm-project?rev=258468=rev
Log:
Merging r258201:
r258201 | ericwf | 2016-01-19 15:07:10 -0700 (Tue, 19 Jan 2016) | 1 line
Add
EricWF added a comment.
In http://reviews.llvm.org/D12354#331776, @dexonsmith wrote:
> This patch looks correct to me. Is there any reason it wasn't committed?
I was concerned about using a function-local static in the library headers, I
don't think libc++ does that anywhere else and I
Author: ericwf
Date: Thu Jan 21 19:09:23 2016
New Revision: 258464
URL: http://llvm.org/viewvc/llvm-project?rev=258464=rev
Log:
Merging r258196:
r258196 | ericwf | 2016-01-19 14:52:04 -0700 (Tue, 19 Jan 2016) | 1 line
Add
Author: ericwf
Date: Thu Jan 21 19:08:10 2016
New Revision: 258463
URL: http://llvm.org/viewvc/llvm-project?rev=258463=rev
Log:
Merging r258217:
r258217 | ericwf | 2016-01-19 16:06:29 -0700 (Tue, 19 Jan 2016) | 1 line
Fix
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM after addressing the `IncompleteClass` comment.
Thanks for the patch!
Comment at: test/std/utilities/meta/meta.rel/is_convertible.pass.cpp:217
@@ -209,1 +216,3 @@
+
EricWF added a reviewer: mclow.lists.
EricWF added a subscriber: mclow.lists.
EricWF added a comment.
Adding @mclow.lists as a reviewer.
http://reviews.llvm.org/D16360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
Author: ericwf
Date: Fri Jan 22 00:25:47 2016
New Revision: 258491
URL: http://llvm.org/viewvc/llvm-project?rev=258491=rev
Log:
Add __uncvref type for use in later patches
Modified:
libcxx/trunk/include/type_traits
Modified: libcxx/trunk/include/type_traits
URL:
EricWF added a comment.
Overall the patch looks good but I have a few concerns.
> - If argument.first can be trivially converted to key_type, don't alloc.
I'm concerned with this part of the change because:
- The `is_trivially_*` traits are often not available and can sometimes blow up.
- It
EricWF added a comment.
> - Did I successfully match the coding style? (I'm kind of lost without
> clang-format TBH.)
The style looks pretty good. I'll comment on any nits I have.
> - Should I separate the change to __construct_node_hash() into a separate
> prep commit? (I would if this were
On Thu, Jan 21, 2016 at 10:35 PM, Duncan P. N. Exon Smith <
dexonsm...@apple.com> wrote:
>
> > On 2016-Jan-21, at 17:59, Eric Fiselier wrote:
> >
> > EricWF added a comment.
> >
> > In http://reviews.llvm.org/D12354#331776, @dexonsmith wrote:
> >
> >> This patch looks correct to
Hey Duncan,
I know this isn't required, but would it be possible to put this on
phabricator?
/Eric
On Mon, Jan 18, 2016 at 4:34 PM, Duncan P. N. Exon Smith <
dexonsm...@apple.com> wrote:
> ping
>
> > On 2016-Jan-11, at 16:23, Duncan P. N. Exon Smith
> wrote:
> >
> > ping
Hi Hans,
Please merge this commit into 3.8. I need it in order to un-revert the
libc++abi commit r256323.
Marshall can you OK this?
/Eric
On Tue, Jan 19, 2016 at 4:06 PM, Eric Fiselier via cfe-commits <
cfe-commits@lists.llvm.org> wrote:
> Author: ericwf
> Date: Tue Jan 19 17:06:2
Author: ericwf
Date: Tue Jan 19 17:37:51 2016
New Revision: 258222
URL: http://llvm.org/viewvc/llvm-project?rev=258222=rev
Log:
Recommit r256322: Fix PR25898 - Check for incomplete pointers types in
can_catch(...)
This patch re-commits r256322 and r256323. They were reverted due to a OS X
test
Hi Hans,
I fixed the problem that caused Nico to revert this. I would like to
re-land this in 3.8.
This depends on libc++ commit r258217 also landing in 3.8.
@Marshall can you OK this?
/Eric
On Tue, Jan 19, 2016 at 4:42 PM, Eric Fiselier via cfe-commits <
cfe-commits@lists.llvm.org>
Author: ericwf
Date: Tue Jan 19 19:26:30 2016
New Revision: 258270
URL: http://llvm.org/viewvc/llvm-project?rev=258270=rev
Log:
Add link to 3rd party GDB pretty-printers
Modified:
libcxx/trunk/docs/UsingLibcxx.rst
Modified: libcxx/trunk/docs/UsingLibcxx.rst
URL:
Hi Hans,
Sorry to make this more complicated, but can you also include the follow-up
libc++abi commit r258277.
/Eric
On Tue, Jan 19, 2016 at 7:08 PM, Marshall Clow
wrote:
>
>
> On Tue, Jan 19, 2016 at 3:48 PM, Eric Fiselier wrote:
>
>> Hi Hans,
>>
>> I
Author: ericwf
Date: Tue Jan 19 17:42:10 2016
New Revision: 258249
URL: http://llvm.org/viewvc/llvm-project?rev=258249=rev
Log:
Recommit r256322: Fix PR25898 - Check for incomplete pointers types in
can_catch(...)
This patch re-commits r256322 and r256323. They were reverted due to a OS X
test
Author: ericwf
Date: Tue Jan 19 17:06:29 2016
New Revision: 258217
URL: http://llvm.org/viewvc/llvm-project?rev=258217=rev
Log:
Fix enviroment variables when running shell scripts
Modified:
libcxx/trunk/test/libcxx/test/config.py
Modified: libcxx/trunk/test/libcxx/test/config.py
URL:
Hi Hans,
I would like to merge the following commits into 3.8.
1. r258196 (libc++)
2. r258198 (libc++)
3. r258201 (libc++abi)
These commits simply add missing license headers.
@Marshall Can you please OK this?
/Eric
___
cfe-commits mailing list
Author: ericwf
Date: Tue Jan 19 21:04:50 2016
New Revision: 258277
URL: http://llvm.org/viewvc/llvm-project?rev=258277=rev
Log:
Fix link flags order in RUN command.
Modified:
libcxxabi/trunk/test/incomplete_type.sh.cpp
Modified: libcxxabi/trunk/test/incomplete_type.sh.cpp
URL:
EricWF created this revision.
EricWF added a reviewer: mclow.lists.
EricWF added a subscriber: cfe-commits.
This bug was originally fixed in http://reviews.llvm.org/D7201.
However it was broken again by the fix to
https://llvm.org/bugs/show_bug.cgi?id=22605.
This patch re-fixes __wrap_iter
Author: ericwf
Date: Tue Jan 19 23:44:11 2016
New Revision: 258292
URL: http://llvm.org/viewvc/llvm-project?rev=258292=rev
Log:
Use TEST_STD_VER instead of __has_feature to detect noexcept. This fixes the
test with GCC.
Modified:
Author: ericwf
Date: Tue Jan 19 22:59:57 2016
New Revision: 258287
URL: http://llvm.org/viewvc/llvm-project?rev=258287=rev
Log:
Mark some test XFAIL for GCC 4.9 due to missing is_trivial* traits
Modified:
libcxx/trunk/test/std/iterators/stream.iterators/istream.iterator/types.pass.cpp
EricWF created this revision.
EricWF added reviewers: mclow.lists, rsmith.
EricWF added a subscriber: cfe-commits.
According to the C++ standard isn't allowed to define `true`
`false` or `bool`. However these macros are sometimes defined by the compilers
`stdbool.h`.
Clang defines the macros
EricWF added a comment.
Alright, so this is looking really good. Once we land this would you be
interested in applying the same optimization to `emplace` calls?
Comment at: include/__hash_table:863
@@ -862,2 +862,3 @@
_LIBCPP_INLINE_VISIBILITY
-pair
EricWF added a comment.
In http://reviews.llvm.org/D16360#334781, @dexonsmith wrote:
> Great, I should have time to clean this up tomorrow afternoon.
>
> Regarding emplace, this patch as is has tests for emplace, but
> they depend on r258575 being in tree. You asked me to revert
> that...
EricWF added a subscriber: EricWF.
EricWF accepted this revision.
EricWF added a reviewer: EricWF.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM.
@dsanders: What platform are you on that doesn't have en_US.UTF-8? If your
serious about testing libc++ you should
Author: ericwf
Date: Tue Jan 26 14:24:30 2016
New Revision: 258852
URL: http://llvm.org/viewvc/llvm-project?rev=258852=rev
Log:
Fix PR26103 - Error calling is_convertible with incomplete type. Patch from
Michael Daniels.
Modified:
libcxx/trunk/include/type_traits
EricWF closed this revision.
EricWF added a comment.
Committed as r258852. Thanks for the patch.
http://reviews.llvm.org/D16344
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: ericwf
Date: Tue Jan 26 14:31:01 2016
New Revision: 258855
URL: http://llvm.org/viewvc/llvm-project?rev=258855=rev
Log:
Remove dead code missed in r258852.
Modified:
libcxx/trunk/include/type_traits
Modified: libcxx/trunk/include/type_traits
URL:
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
Obviously LGTM. @hans I'm OK with this going into 3.8.
http://reviews.llvm.org/D16584
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
somehow in the
> include
> >>>> path for a C89 compilation?
> >>>
> >>>
> >>> Eric, Marshall: Ping ^
> >>>
> >>>>
> >>>>
> >>>>>
> >>>>> On Tue, Oct 13, 2015 at 7:3
EricWF updated this revision to Diff 46071.
EricWF added a comment.
Address @mclow.lists comment about `__get_node_unchecked()`'s name.
http://reviews.llvm.org/D15836
Files:
include/__config
include/forward_list
test/std/containers/sequences/forwardlist/incomplete.pass.cpp
Index:
EricWF accepted this revision.
EricWF added a reviewer: EricWF.
EricWF added a comment.
This revision is now accepted and ready to land.
Accepting on phab to reflect @mclow.lists LGTM.
Comment at: include/forward_list:369
@@ +368,3 @@
+}
+__node_pointer
This revision was automatically updated to reflect the committed changes.
Closed by commit rL25: [libcxx] Fix undefined behavior in forward_list
(authored by EricWF).
Changed prior to commit:
http://reviews.llvm.org/D15836?vs=46071=46073#toc
Repository:
rL LLVM
Author: ericwf
Date: Tue Jan 26 18:11:54 2016
New Revision: 25
URL: http://llvm.org/viewvc/llvm-project?rev=25=rev
Log:
[libcxx] Fix undefined behavior in forward_list
Summary:
This patch is similar to the fix but it has a few differences. This
patch doesn't use a `__link_pointer`
Author: ericwf
Date: Tue Jan 26 18:49:20 2016
New Revision: 258893
URL: http://llvm.org/viewvc/llvm-project?rev=258893=rev
Log:
Fix broken commit r25. I missed adding two pointer conversions
Modified:
libcxx/trunk/include/forward_list
Modified: libcxx/trunk/include/forward_list
URL:
The direction and change LGTM. Feel free to commit once the change has some
tests.
On Fri, Jan 8, 2016 at 5:45 PM, Duncan P. N. Exon Smith <
dexonsm...@apple.com> wrote:
> No one is waiting for me on this, are they? FWIW, this seems like
> a good solution to me. I'll see what I can do
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM.
I don't see anything in the Itanium ABI spec that would prefer one demangling
over another.
I also double checked that `scanf("%a", ...)` generates the expected
bit-pattern for each
gt;>
> >>
> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include
> >>
> >>
> /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.11.sdk/usr/include
> >>
> >>
> /Ap
On Fri, Jan 22, 2016 at 10:44 AM, Duncan P. N. Exon Smith <
dexonsm...@apple.com> wrote:
>
> > On 2016-Jan-21, at 22:22, Eric Fiselier wrote:
> >
> >
> >
> > On Thu, Jan 21, 2016 at 10:35 PM, Duncan P. N. Exon Smith <
> dexonsm...@apple.com> wrote:
> >
> > > On 2016-Jan-21, at
EricWF added a subscriber: EricWF.
EricWF added a comment.
How are you getting this warning? I can't generate it with Clang 3.8 on 64bit
linux with -m32. In that configuration int and long have the same size.
http://reviews.llvm.org/D16480
___
Author: ericwf
Date: Wed Feb 17 18:20:34 2016
New Revision: 261180
URL: http://llvm.org/viewvc/llvm-project?rev=261180=rev
Log:
Get working again
Modified:
libcxx/trunk/include/__hash_table
Modified: libcxx/trunk/include/__hash_table
URL:
Author: ericwf
Date: Wed Feb 17 18:21:09 2016
New Revision: 261181
URL: http://llvm.org/viewvc/llvm-project?rev=261181=rev
Log:
Commit tests missing from r261180.
Added:
libcxx/trunk/test/libcxx/containers/gnu_cxx/
libcxx/trunk/test/libcxx/containers/gnu_cxx/hash_map.pass.cpp
gt; On 2016-Feb-11, at 04:25, Eric Fiselier via cfe-commits <
> cfe-commits@lists.llvm.org> wrote:
> >
> > The following changes are planed for future revisions:
> >
> > * Fix LWG issue #2469 by delegating 'unordered_map::operator[]' to use
> >'__e
...@gmail.com> wrote:
> >>>
> >>> Hi,
> >>>
> >>> hash_map still looks broken to me.
> >>> I don't have a simple reproducer, but these declarations in
> __hash_table:
> >>>
> >>> template class _LIBCPP_TYPE_VIS_
> How big a maintenance burden is it? Is there a way to reduce the burden?
It's actually not *that* big of a maintenance burden because it's mostly
just a shallow wrapper around <__hash_table>.
What's been frustrating me is that there were *NO TESTS*. So modifying
<__hash_table> is a stab in
EricWF accepted this revision.
EricWF added a reviewer: EricWF.
EricWF added a comment.
This revision is now accepted and ready to land.
Got approval from @mclow.lists offline.
http://reviews.llvm.org/D16346
___
cfe-commits mailing list
Author: ericwf
Date: Fri Feb 19 18:16:41 2016
New Revision: 261381
URL: http://llvm.org/viewvc/llvm-project?rev=261381=rev
Log:
Add stdbool.h wrapper for libc++
Summary:
According to the C++ standard isn't allowed to define `true`
`false` or `bool`. However these macros are sometimes defined
EricWF accepted this revision.
EricWF added a reviewer: EricWF.
EricWF added a comment.
This revision is now accepted and ready to land.
@mclow.lists said this was OK to commit if he couldn't think of a better idea.
It's unfortunate to use `` to forward declare containers but ``
already
Author: ericwf
Date: Fri Feb 19 18:19:45 2016
New Revision: 261382
URL: http://llvm.org/viewvc/llvm-project?rev=261382=rev
Log:
Make __wrap_iter work with GCC again
Summary:
This bug was originally fixed in http://reviews.llvm.org/D7201.
However it was broken again by the fix to
Author: ericwf
Date: Fri Feb 19 18:24:43 2016
New Revision: 261383
URL: http://llvm.org/viewvc/llvm-project?rev=261383=rev
Log:
Fix PR26622 - Make CheckLibcxxAtomic.cmake use the libc++ headers.
Modified:
libcxx/trunk/cmake/Modules/CheckLibcxxAtomic.cmake
Modified:
Author: ericwf
Date: Fri Feb 19 19:11:55 2016
New Revision: 261391
URL: http://llvm.org/viewvc/llvm-project?rev=261391=rev
Log:
Work around GCC bug in .fail.cpp tests
Modified:
libcxx/trunk/test/libcxx/test/format.py
Modified: libcxx/trunk/test/libcxx/test/format.py
URL:
EricWF added a comment.
Overall this patch is *almost* there. My only objection is that this
optimization neglects "unordered_set". Optimally we would also catch
"unordered_set.emplace(42)"
Comment at: include/__hash_table:103
@@ -102,1 +102,3 @@
+template ::type>
EricWF added a comment.
@dexonsmith Actually is it OK if I contribute the tests for this patch? Your's
are in no way bad, but I want to test this in a similar manner to
`unord.map.modifiers/insert_allocator_requirments.pass.cpp` and it's not fair
to ask you to do that.
However if your willing
Author: ericwf
Date: Fri Feb 19 23:28:30 2016
New Revision: 261416
URL: http://llvm.org/viewvc/llvm-project?rev=261416=rev
Log:
Cleanup node-type handling in the associative containers.
This patch is very similar to r260431.
This patch is the first in a series of patches that's meant to better
Author: ericwf
Date: Sat Feb 20 01:12:17 2016
New Revision: 261419
URL: http://llvm.org/viewvc/llvm-project?rev=261419=rev
Log:
Remove all usages of "const" node pointer typedefs in the assoc containers.
The "const" pointer typedefs such as "__node_const_pointer" and
"__node_base_const_pointer"
Author: ericwf
Date: Sat Feb 20 01:59:16 2016
New Revision: 261421
URL: http://llvm.org/viewvc/llvm-project?rev=261421=rev
Log:
Rename <__hash_table> traits so they don't conflict with similar traits.
In particular <__tree> defines many of the same types of traits classes.
Modified:
EricWF added a comment.
Did you consider simply wrapping the library in "-Wl,-Bstatic" and
"-Wl,-Bdynamic" linker flags? It seems like that would work almost as well as
explicitly naming the target.
Comment at: test/libcxx/test/config.py:463
@@ -462,2 +462,3 @@
else:
EricWF added inline comments.
Comment at: test/CMakeLists.txt:10
@@ -9,2 +9,3 @@
pythonize_bool(LLVM_BUILD_32_BITS)
+pythonize_bool(LIBCXX_ENABLE_SHARED)
pythonize_bool(LIBCXXABI_ENABLE_SHARED)
This seems wrong for out of tree builds because
Author: ericwf
Date: Thu Feb 11 12:21:18 2016
New Revision: 260556
URL: http://llvm.org/viewvc/llvm-project?rev=260556=rev
Log:
Revert r260514 because it has a bogus commit message.
Modified:
libcxx/trunk/include/unordered_map
EricWF added a comment.
The method @bcraig is describing is documented here:
http://libcxx.llvm.org/docs/BuildingLibcxx.html#id4
http://reviews.llvm.org/D18174
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
EricWF added a subscriber: EricWF.
EricWF added a comment.
@raj.khem Can you file a bug against libc++abi please?
http://reviews.llvm.org/D18174
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
EricWF added a comment.
That patch looks hugely dangerous because you've now changed when your static
mutex's get initialized. Please file a bug against libc++.
http://reviews.llvm.org/D18174
___
cfe-commits mailing list
EricWF added a comment.
Yes exactly. `// REQUIRES: ` tells LIT to skip the test (and report it
as UNSUPPORTED) whenever the feature is not present.
http://reviews.llvm.org/D14731
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
EricWF added inline comments.
Comment at: include/atomic:859
@@ +858,3 @@
+template <> _LIBCPP_CONSTEXPR bool __libcpp_always_lock_free = 2 ==
ATOMIC_CHAR32_T_LOCK_FREE;
+template <> _LIBCPP_CONSTEXPR bool __libcpp_always_lock_free = 2 ==
ATOMIC_WCHAR_T_LOCK_FREE;
+template <>
EricWF added a comment.
So I fixed up LIT so that it also adds "-Werror=thread-safety" for both
".pass.cpp" and ".fail.cpp" tests. Could you apply it to your patch?
https://gist.github.com/EricWF/8a0bfb6ff02f8c9f9940
Comment at: include/__mutex_base:37
@@ -30,1 +36,3 @@
+
EricWF added a comment.
So this LGTM except for one last change (I'm sorry). LIT now knows when
"-Werror=thread-safety" is being passed. Could you change the tests guarded by
"#if !defined(__clang__) || !__has_attribute(aquire_capability)` to say "//
REQUIRES: thread-safety" instead? I prefer
EricWF added inline comments.
Comment at: include/__mutex_base:37
@@ -30,1 +36,3 @@
+
+class _LIBCPP_TYPE_VIS _LIBCPP_THREAD_SAFETY_ANNOTATION(capability("mutex"))
mutex
{
I appreciate the super thorough answer!
http://reviews.llvm.org/D14731
EricWF added a comment.
By including `fallback_malloc.ipp` in a new C++ file we now have 2 different
OOM emergancy buffers. I think it would be preferable to use only one. This
would require a fairly large restructuring of `fallback_malloc.ipp` though.
http://reviews.llvm.org/D17815
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM.
Repository:
rL LLVM
http://reviews.llvm.org/D18107
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
EricWF added a comment.
No worries, I'll try and take a look this weekend. Thanks.
http://reviews.llvm.org/D16360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
It seems like we are not losing test coverage with these removals. And they
have caused us a ton of pain.
I agree lets just get rid of them!
LGTM.
http://reviews.llvm.org/D18187
I agree with Nico. Adding _LIBCPP_INLINE_VISIBILITY should fix the problem,
but it's a bit of a hack. Also we should only *add* additional signatures
if possible since existing signatures already exist in the dylib so we
can't modify them.
On Fri, Mar 11, 2016 at 8:32 AM, Nico Weber via
On Fri, Mar 11, 2016 at 8:52 AM, Duncan Exon Smith via cfe-commits <
cfe-commits@lists.llvm.org> wrote:
> There's a longer term fix involving availability attributes. My patch to
> the test system for deployment targets is holding up a series of patches
> that tell the compiler when API was
EricWF accepted this revision.
EricWF added a comment.
This revision is now accepted and ready to land.
LGTM after the inline question is addressed.
Comment at: test/libcxxabi/test/config.py:38
@@ +37,3 @@
+super(Configuration, self).configure_features()
+if
Author: ericwf
Date: Wed Mar 16 22:30:56 2016
New Revision: 263688
URL: http://llvm.org/viewvc/llvm-project?rev=263688=rev
Log:
Make std::addressof constexpr in C++17 (Clang only).
Added:
601 - 700 of 2481 matches
Mail list logo