Libc++ was incorrectly reporting an error when the target of create_directory
already exists, but was not a directory. This behavior is not specified
in the most recent standard, which says no error should be reported.
Additionally, libc++ failed to report an error when the attribute directory
path didn't exist or didn't name a directory. This has been fixed as well.
Although it's not clear if we should call status or symlink_status on the
attribute directory. This patch chooses to still call status.
llvm-svn: 337888
Previously the <experimental/filesystem> didn't guard its
contents in any dialect. However, the implementation implicitly
requires at least C++11, and the tests have always been marked
unsupported in C++03. This patch puts a header guard around the
contents to avoid exposing them before C++11.
Additionally, it replaces all of the usages of _NOEXCEPT or
_LIBCPP_CONSTEXPR with the keyword directly, since we can
expect the compiler to implement those by now.
llvm-svn: 337884
Summary:
This is not guaranteed to work since the characters after '__has_include('
have special lexing rules that can't possibly be applied when
__has_include is generated by a macro. It also breaks the crash reproducers
generated by -frewrite-includes (see https://llvm.org/pr37990).
Reviewers: EricWF, rsmith, mclow.lists
Reviewed By: mclow.lists
Differential Revision: https://reviews.llvm.org/D49067
llvm-svn: 337824
When adding the new tests for the filesystem_error::what method,
I incorrectly removed a test case and replaced it with something else.
This patch restores that test case
llvm-svn: 337764
This patch implements the `what()` for filesystem errors. The message
includes the 'what_arg', any paths that were specified, and the
error code message.
Additionally this patch refactors how errors are created, making it easier
to report them correctly.
llvm-svn: 337664
This patch improves both the performance, and the safety of the
copy_file implementation.
The performance improvements are achieved by using sendfile on
Linux and copyfile on OS X when available.
The TOCTOU hardening is achieved by opening the source and
destination files and then using fstat to check their attributes to
see if we can copy them.
Unfortunately for the destination file, there is no way to open
it without accidentally creating it, so we first have to use
stat to determine if it exists, and if we should copy to it.
Then, once we're sure we should try to copy, we open the dest
file and ensure it names the same entity we previously stat'ed.
llvm-svn: 337649
First, <experimental/filesystem> didn't correctly guard
against min/max macros. This adds the proper push/pop macro guards.
Second, an internal time helper had been renamed but the test for
it hadn't been updated. This patch updates those tests.
llvm-svn: 337520
Summary:
This patch implements directory_entry caching *almost* as specified in P0317r1. However, I explicitly chose to deviate from the standard as I'll explain below.
The approach I decided to take is a fully caching one. When `refresh()` is called, the cache is populated by calls to `stat` and `lstat` as needed.
During directory iteration the cache is only populated with the `file_type` as reported by `readdir`.
The cache can be in the following states:
* `_Empty`: There is nothing in the cache (likely due to an error)
* `_IterSymlink`: Created by directory iteration when we walk onto a symlink only the symlink file type is known.
* `_IterNonSymlink`: Created by directory iteration when we walk onto a non-symlink. Both the regular file type and symlink file type are known.
* `_RefreshSymlink` and `_RefreshNonSymlink`: A full cache created by `refresh()`. This case includes dead symlinks.
* `_RefreshSymlinkUnresolved`: A partial cache created by refresh when we fail to resolve the file pointed to by a symlink (likely due to permissions). Symlink attributes are cached, but attributes about the linked entity are not.
As mentioned, this implementation purposefully deviates from the standard. According to some readings of the specification, and the Windows filesystem implementation, the constructors and modifiers which don't pass an `error_code` must throw when the `directory_entry` points to a entity which doesn't exist. or when attribute resolution fails for another reason.
@BillyONeal has proposed a more reasonable set of requirements, where modifiers other than refresh ignore errors. This is the behavior libc++ currently implements, with the expectation some form of the new language will be accepted into the standard.
Some additional semantics which differ from the Windows implementation:
1. `refresh` will not throw when the entry doesn't exist. In this case we can still meet the functions specification, so we don't treat it as an error.
2. We don't clear the path name when a constructor fails via refresh (this will hopefully be changed in the standard as well).
It should be noted that libstdc++'s current implementation has the same behavior as libc++, except for point (2).
If the changes to the specification don't get accepted, we'll be able to make the changes later.
[1] http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2016/p0317r1.html
Reviewers: mclow.lists, gromer, ldionne, aaron.ballman
Subscribers: BillyONeal, christof, cfe-commits
Differential Revision: https://reviews.llvm.org/D49530
llvm-svn: 337516
* Remove unused type from is_assignable.pass.cpp
* Don't specialize `common_type<::X<float>>` in common_type.pass.cpp, which violates the requirements of [meta.trans.other]/5
llvm-svn: 336618
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
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
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
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
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
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
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
