Commit Graph

4739 Commits

Author SHA1 Message Date
Martin Storsjo 11b02759a3 [CMake] Convert paths to the right form in standalone builds on Windows
The paths output from llvm-config --cmakedir and from clang
--print-libgcc-file-name can contain backslashes, while CMake
can't handle the paths in this form.

This matches what compiler-rt already does (since SVN r203789
and r293195).

Differential Revision: https://reviews.llvm.org/D48356

llvm-svn: 335172
2018-06-20 21:03:34 +00:00
Richard Smith 7ebd641fcf Fix libcxx tests after clang r334677.
Feature test macro versions may have a trailing L.

llvm-svn: 334917
2018-06-17 19:58:45 +00:00
Marshall Clow b6b7a3bb78 Remove P0771, which was not passed in Rapperswil
llvm-svn: 334894
2018-06-16 18:03:29 +00:00
Stephan T. Lavavej cb0d20519e [libcxx] [test] Strip trailing whitespace. NFC.
llvm-svn: 334676
2018-06-14 00:12:20 +00:00
Stephan T. Lavavej 896e499e38 [libcxx] [test] Update msvc_stdlib_force_include.hpp.
MSVC's STL removed _SCL_SECURE_NO_WARNINGS.

MSVC's STL implemented feature-test macros.

llvm-svn: 334675
2018-06-14 00:12:14 +00:00
Petr Hosek a9e7908e30 [CMake] Use common variable for all header targets NFC
This simplifies the handling of header targets.

llvm-svn: 334477
2018-06-12 06:58:06 +00:00
Petr Hosek 0aa70faf9b [CMake] Add a missing target dependency on C++ ABI headers
This resolves the breakage introduced in r334468 which results in
build error when using CMake Makefile generator.

llvm-svn: 334470
2018-06-12 03:31:03 +00:00
Petr Hosek c20c182df3 Reland "Use custom command and target to install libc++ headers"
Using file(COPY FILE...) has several downsides. Since the file command
is only executed at configuration time, any changes to headers made
after the initial CMake execution are ignored. This can lead to subtle
errors since the just built Clang will be using stale libc++ headers.
Furthermore, since the headers are copied prior to executing the build
system, this may hide missing dependencies on libc++ from other LLVM
components.

This changes replaces the use of file(COPY FILE...) command with a
custom command and target which addresses all aforementioned issues and
matches the implementation already used by other LLVM components that
also install headers like Clang builtin headers.

Differential Revision: https://reviews.llvm.org/D44773

llvm-svn: 334468
2018-06-12 03:10:02 +00:00
Marshall Clow e044d1e504 Update the to-do list with motions from Rapperswil.
llvm-svn: 334467
2018-06-12 02:45:30 +00:00
Volodymyr Sapsai d1df7b7ee7 Mark the test using <experimental/memory_resource> to require c++experimental.
When built against the old libc++ version the test was causing linker error

Undefined symbols for architecture x86_64:
  "std::experimental::fundamentals_v1::pmr::new_delete_resource()", referenced from:
      void test_evil<WidgetV0, WidgetV0>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV0, WidgetV1>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV0, WidgetV2>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV0, WidgetV3>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV1, WidgetV0>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV1, WidgetV1>() in construct_piecewise_pair_evil.pass.cpp.o
      void test_evil<WidgetV1, WidgetV2>() in construct_piecewise_pair_evil.pass.cpp.o
      ...

llvm-svn: 334431
2018-06-11 19:42:27 +00:00
Eric Fiselier cb5b004a9b Remove unused code from __functional_base. NFC.
Patch from Arthur O'Dwyer.

`__user_alloc_construct_impl` is used by <experimental/memory_resource>, but
this `__user_alloc_construct` is never used.

Also, `<experimental/memory_resource>` doesn't need a full definition of
`std::tuple`; just the forward declaration in `<__tuple>` will suffice.

Reviewed as https://reviews.llvm.org/D46806

llvm-svn: 334069
2018-06-06 06:42:27 +00:00
Eric Fiselier c797bb23b8 Fix test failures after r334053.
llvm-svn: 334056
2018-06-06 00:13:49 +00:00
Eric Fiselier 8cef7fd75a Fix PR37694 - std::vector doesn't correctly move construct allocators.
C++2a[container.requirements.general]p8 states that when move constructing
a container, the allocator is move constructed. Vector previously copy
constructed these allocators. This patch fixes that bug.

Additionally it cleans up some unnecessary allocator conversions
when copy constructing containers. Libc++ uses
__internal_allocator_traits::select_on_copy_construction to select
the correct allocator during copy construction, but it unnecessarily
converted the resulting allocator to the user specified allocator
type and back. After this patch list and forward_list no longer
do that.

Technically we're supposed to be using allocator_traits<allocator_type>::select_on_copy_construction,
but that should seemingly be addressed as a separate patch, if at all.

llvm-svn: 334053
2018-06-05 22:32:52 +00:00
Erik Pilkington f52318b47b Fix a strict aliasing violation in map and unordered_map.
These containers type-punned between pair<K, V> and pair<const K, V> as an
optimization. This commit instead provides access to the pair via a pair of
references that assign through to the underlying object. It's still undefined to
mutate a const object, but clang doesn't optimize on this for data members, so
this should be safe.

Differential revision: https://reviews.llvm.org/D47607

llvm-svn: 333948
2018-06-04 20:38:23 +00:00
JF Bastien 84d7a4a344 Mark __c11_atomic_load as const
Summary:
C++11 onwards specs the non-member functions atomic_load and atomic_load_explicit as taking the atomic<T> by const (potentially volatile) pointer. C11, in its infinite wisdom, decided to drop the const, and C17 will fix this with DR459 (the current draft forgot to fix B.16, but that’s not the normative part).

This patch fixes the libc++ version of the __c11_atomic_load builtins defined for GCC's compatibility sake.

D47618 takes care of the clang side.

Discussion: http://lists.llvm.org/pipermail/cfe-dev/2018-May/058129.html

<rdar://problem/27426936>

Reviewers: EricWF, mclow.lists

Subscribers: christof, cfe-commits

Differential Revision: https://reviews.llvm.org/D47613

llvm-svn: 333776
2018-06-01 18:02:53 +00:00
JF Bastien 766128e697 Filesystem tests: un-confuse write time
Summary:
The filesystem test was confused about access versus write / modification time. The spec says:

  file_time_type last_write_time(const path& p, error_code& ec) noexcept;
  Returns: The time of last data modification of p, determined as if by the value of the POSIX stat structure member st_mtime obtained as if by POSIX stat(). The signature with argument ec returns file_time_type::min() if an error occurs.

The test was looking at st_atime, not st_mtime, when comparing the result from last_write_time. That was probably due to using a pair instead of naming things nicely or using types. I opted to rename things so it's clearer.

This used to cause test bot failures.

<rdar://problem/40648859>

Reviewers: EricWF, mclow.lists, aemerson

Subscribers: christof, cfe-commits

Differential Revision: https://reviews.llvm.org/D47557

llvm-svn: 333723
2018-06-01 04:59:48 +00:00
Eric Fiselier 18567eacc5 Update ABI lists after change in r333467.
r333467 updated the symbols exported by libc++.so/dylib by changing
the ODR usage of __uncaught_exception/__uncaught_exceptions. This
should not be a breaking change.

llvm-svn: 333481
2018-05-29 23:52:14 +00:00
JF Bastien 2fec6dc56e Mark deduction guide tests as failing on apple-clang-9
As discussed here: http://lists.llvm.org/pipermail/cfe-dev/2018-May/058116.html
The tests fail on clang-5, as well as apple-clang-9. Mark them as such.

llvm-svn: 333479
2018-05-29 23:28:04 +00:00
Marshall Clow 3a92ecc8e6 Fix embarrasing typo in uncaught_exceptions. Update tests to really test this. Thanks to Peter Klotz for calling my attention to this.
llvm-svn: 333467
2018-05-29 22:25:42 +00:00
Marshall Clow 5a823d8d63 Mark __clear_and_shrink() as noexcept. This prevents the generation of a catch block and call to terminate in string's move assignment. Thanks to Howard for the 'catch'.
llvm-svn: 333435
2018-05-29 17:04:37 +00:00
Eric Fiselier bd2e949869 LWG 2969 "polymorphic_allocator::construct() shouldn't pass resource()"
Patch from Arthur O'Dwyer.

In the TS, `uses_allocator` construction for `pair` tried to use an allocator
type of `memory_resource*`, which is incorrect because `memory_resource*` is
not an allocator type. LWG 2969 fixed it to use `polymorphic_allocator` as the
allocator type instead.

https://wg21.link/lwg2969

(D47090 included this in `<memory_resource>`; at Eric's request, I've split
this out into its own patch applied to the existing
`<experimental/memory_resource>` instead.)

Reviewed as https://reviews.llvm.org/D47109

llvm-svn: 333384
2018-05-29 00:08:47 +00:00
Marshall Clow c58f9b787f Fix up the final bits of breakage due to clang v5 generating bad implicit template deduction guides - specifically for copy-ctors
llvm-svn: 333381
2018-05-28 19:20:21 +00:00
Marshall Clow aefe07a4c1 Mark the template deduction tests as UNSUPPORTED on clang 5, because it deduces the wrong type.
llvm-svn: 333376
2018-05-28 15:42:47 +00:00
JF Bastien f764dbd023 Revert "Add nonnull; use it for atomics"
That's r333325, as well as follow-up "Fix GCC handling of ATOMIC_VAR_INIT"
r333327.

Marshall asked to revert:

Let's have a discussion about how to implement this so that it is more friendly
to people with installed code bases. We've had *extremely* loud responses to
unilaterally adding warnings - especially ones that can't be easily disabled -
to the libc++ code base in the past.

llvm-svn: 333351
2018-05-26 19:44:45 +00:00
JF Bastien 925faf274c Fix GCC handling of ATOMIC_VAR_INIT
r333325 from D47225 added warning checks, and the test was written to be C++11 correct by using ATOMIC_VAR_INIT (note that the committee fixed that recently...). It seems like GCC can't handle ATOMIC_VAR_INIT well because it generates 'type 'std::atomic<int>' cannot be initialized with an initializer list' on bot libcxx-libcxxabi-x86_64-linux-ubuntu-cxx03. Drop the ATOMIC_VAR_INITs since they weren't required to test the diagnostics.

llvm-svn: 333327
2018-05-26 00:13:53 +00:00
JF Bastien 1be7517aa9 Add nonnull; use it for atomics
Summary:
The atomic non-member functions accept pointers to std::atomic / std::atomic_flag as well as to the non-atomic value. These are all dereferenced unconditionally when lowered, and therefore will fault if null. It's a tiny gotcha for new users, especially when they pass in NULL as expected value (instead of passing a pointer to a NULL value). We can therefore use the nonnull attribute to denote that:

  - A warning should be generated if the argument is null
  - It is undefined behavior if the argument is null (because a dereference will segfault)

This patch adds support for this attribute for clang and GCC, and sticks to the subset of the syntax both supports. In particular, work around this GCC oddity:
  https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60625

The attributes are documented:

  - https://gcc.gnu.org/onlinedocs/gcc-4.0.0/gcc/Function-Attributes.html
  - https://clang.llvm.org/docs/AttributeReference.html#nullability-attributes

I'm authoring a companion clang patch for the __c11_* and __atomic_* builtins, which currently only warn on a subset of the pointer parameters.

In all cases the check needs to be explicit and not use the empty nonnull list, because some of the overloads are for atomic<T*> and the values themselves are allowed to be null.

<rdar://problem/18473124>

Reviewers: arphaman, EricWF

Subscribers: aheejin, christof, cfe-commits

Differential Revision: https://reviews.llvm.org/D47225

llvm-svn: 333325
2018-05-25 23:43:53 +00:00
JF Bastien 41b4ba1f1e Fix optional<char> test breakage
It seems GCC and clang disagree. Talked to mclow on IRC, disabling for now.

llvm-svn: 333317
2018-05-25 21:32:27 +00:00
JF Bastien 39281931ec Fix array deduction guide test breakage
No matching constructor

llvm-svn: 333315
2018-05-25 21:17:43 +00:00
JF Bastien fc53b60a14 Fix optional deduction guide test breakage
llvm-svn: 333308
2018-05-25 20:43:57 +00:00
Marshall Clow 3854964531 Add one more test for optional
llvm-svn: 333252
2018-05-25 02:22:54 +00:00
Marshall Clow f35b4bc395 Add deduction guides for optional
llvm-svn: 333251
2018-05-25 02:08:49 +00:00
Akira Hatanaka 7b9bee064a Do not define template specialization __libcpp_is_floating_point<__fp16>
if the compiler is not clang.

gcc doesn't allow using __fp16 on non-ARM targets.

llvm-svn: 333108
2018-05-23 17:50:41 +00:00
Akira Hatanaka ad3be5c454 Teach __libcpp_is_floating_point that __fp16 and _Float16 are
floating-point types.

rdar://problem/40377353

llvm-svn: 333103
2018-05-23 17:31:09 +00:00
Marshall Clow 50eae5a3e5 Mark more bits of P0433 as complete.
llvm-svn: 333058
2018-05-23 03:22:59 +00:00
Marshall Clow edd5e29cfe Implement deduction guides for basic_regex
llvm-svn: 333050
2018-05-23 01:57:02 +00:00
Volodymyr Sapsai b9c0b637cc [libcxx] [test] Mark the test as unsupported by apple-clang-8.1.
llvm-svn: 333011
2018-05-22 18:46:16 +00:00
Marshall Clow 85d26075f9 Change the names of two private methods: allocate -> __vallocate and deallocate -> __vdeallocate. NFC. This change triggered by https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61806, which shows up when we implement deduction guides for the container adaptors.The names have a 'v' in them because WIN32 has a macro named __deallocate. (sigh).
llvm-svn: 332996
2018-05-22 16:20:28 +00:00
Marshall Clow acd629cd3c Missed the tests for the deduction guides for prority_queue
llvm-svn: 332931
2018-05-22 02:19:38 +00:00
Marshall Clow 5b8b8b5dce Deduction guides for the container adaptors - queue, stack, and priority_queue
llvm-svn: 332927
2018-05-22 01:57:53 +00:00
Marshall Clow df8f754792 Implement deduction guides for vector
llvm-svn: 332901
2018-05-21 21:30:12 +00:00
Marshall Clow 4a227e582b Deduction guides for list
llvm-svn: 332818
2018-05-20 14:05:31 +00:00
Marshall Clow e076700b77 Implement deduction guides for forward_list
llvm-svn: 332811
2018-05-19 16:02:05 +00:00
Marshall Clow 926b0b2bf5 Remove expression '1L + INT_MAX', because it overflows on machines where int/long are the same size
llvm-svn: 332797
2018-05-19 03:09:05 +00:00
Marshall Clow dbb6f8a817 Implement deduction guides for <deque>
llvm-svn: 332785
2018-05-18 23:44:13 +00:00
Marshall Clow 929b4ce7aa Disable 'missing-braces' warning
llvm-svn: 332779
2018-05-18 22:21:33 +00:00
Marshall Clow 0ca8c0895c Implement deduction guides for <array>; Reviewed as https://reviews.llvm.org/D46964
llvm-svn: 332768
2018-05-18 21:01:04 +00:00
Billy Robert O'Neal III 8b3e138aad [libcxx] [test] Remove unused local typedef in test/std/numerics/rand/rand.eng/rand.eng.lcong/default.pass.cpp
llvm-svn: 332571
2018-05-17 04:59:34 +00:00
Billy Robert O'Neal III 610e0fd5d5 Add void casts to suppress nodiscard on linear_congruential_engine.
llvm-svn: 332567
2018-05-17 02:58:26 +00:00
Peter Collingbourne acd1f4e738 Condition usage of locale stdlib functions on Android API version
Some *_l functions were not available in some versions of Bionic. This CL
checks that the NDK version supports the functions, and if not, falls back
on the corresponding functions that don't take a locale.

Patch by Tom Anderson!

Differential Revision: https://reviews.llvm.org/D46558

llvm-svn: 332543
2018-05-16 22:40:12 +00:00
Volodymyr Sapsai 85d3a702da Emit an error when include <atomic> after <stdatomic.h>
C11 defines `kill_dependency` as a macro in <stdatomic.h>. When you
include <atomic> after <stdatomic.h>, the macro clashes with
`std::kill_dependency` and causes multiple errors. Explicit error should
help in diagnosing those errors.

No change for working code that includes <atomic> before <stdatomic.h>.

rdar://problem/27435938

Reviewers: rsmith, EricWF, mclow.lists, jfb

Reviewed By: jfb

Subscribers: jfb, jkorous-apple, christof, bumblebritches57, JonChesterfield, smeenai, cfe-commits

Differential Revision: https://reviews.llvm.org/D45470

llvm-svn: 332413
2018-05-15 22:38:31 +00:00