Introduce a `_LIBCPP_HAS_BITSCAN64` macro to specify if the 64-bit
variant of the bitscan family of APIs is available.  This avoids
duplicating the check in the support header.

llvm-svn: 290924

When a parameter pack has multiple corresponding arguments, and some subset of
them are overloaded functions, it's possible that some subset of the parameters
are non-deduced contexts. In such a case, keep deducing from the remainder of
the arguments, and resolve the incomplete pack against whatever other
deductions we've performed for the pack.

GCC, MSVC, and ICC give three different bad behaviors for this case; what we do
now (and what we did before) don't exactly match any of them, sadly :( I'm
getting a core issue opened to specify more precisely how this should be
handled.

llvm-svn: 290923

Fixes D27786.

llvm-svn: 290922

These tests were using malloc()'s return value without checking for null,
which MSVC's /analyze rightly warns about. Asserting that the pointer is
non-null both expresses the test's intention and silences the warning.

Fixes D27785.

llvm-svn: 290921

Previously, -Wtautological-overlap-compare did not warn on cases where the
boolean expression was in an assignment or return statement.  This patch
should cause all boolean statements to be passed to the tautological compare
checks in the CFG analysis.

This is one of the issues from PR13101

llvm-svn: 290920

Summary: ThinLTO needs to pass down the sample profile file path to linker.

Reviewers: tejohnson, davidxl, mehdi_amini

Subscribers: bruno, cfe-commits

Differential Revision: https://reviews.llvm.org/D27832

llvm-svn: 290919

This test was testing that we could correctly find the parent of a DIE, but it was actually just testing the special case where a DIE's depth was 1. This corrects that error by adding an extra level into the the DWARF to ensure that we correctly get the parent by looking for the parent with a depth that is 1 less than the current depth.

Differential Revision: https://reviews.llvm.org/D28261

llvm-svn: 290918

lldb-gtest binary that xcode builds for -scheme lldb-gtest; these tests
won't run on macosx systems.  Fixes testsuite failures we started
seeing after 290819.

<rdar://problem/29853778> 

llvm-svn: 290917

I accidentally omitted these when refactoring this code. This caused
problems when building parts of the test-suite on MacOS.

llvm-svn: 290916

As per post-commit review for r289993 (D27775), we can only safely
import a type as a decl if it has an Identifier, as the Name alone
is not enough to be unique across modules.

llvm-svn: 290915

I'm not sure what determines the minor version, but it appears
that it's possible for a fully updated, release version of
VS2015 with Update 3 can go (at least) as low as 19.00.24213.1.
Updating the compiler version check to account for this so we
don't generate superfluous warnings.

llvm-svn: 290914

llvm-svn: 290913

I wrote this patch before seeing the comment in:
https://reviews.llvm.org/D27114
...that suggests we should actually be canonicalizing the other way.

So just in case we decide this is the right way, we might as well
have a cleaner implementation.

llvm-svn: 290912

Replace the use of _WIN32 in libc++. Replace most use with a C runtime
check _LIBCPP_MSVCRT or the new _LIBCPP_WIN32 to indicate that we are
using the Win32 API. Use a new _LIBCPP_WCHAR_IS_UCS2 to indicate that we
are on an environment that has a short wchar_t.

llvm-svn: 290910

As Pete points out in r290905, CallSite lets us avoid duplicating this!

llvm-svn: 290909

Allows LLVM to build with LLVM_USE_OPROFILE=True.

Patch by Mark Dewing. Thanks Mark!

llvm-svn: 290908

llvm-svn: 290907

The special case to widen the integer literal zero when passed to
variadic function calls should only apply to variadic functions, not
unprototyped functions. This is consistent with what MSVC does. In this
test case, MSVC uses a 4-byte store to pass the 5th argument to 'kr' and
an 8-byte store to pass the zero to 'v':

  void v(int, ...);
  void kr();
  void f(void) {
    v(1, 2, 3, 4, 0);
    kr(1, 2, 3, 4, 0);
  }

Aaron Ballman discovered this issue in https://reviews.llvm.org/D28166

llvm-svn: 290906

Use getReturnedArgOperand() instead of rolling our own.  Note that it's
equivalent because there can only be one 'returned' operand.

The existing code was also incorrect: there already was awkward logic to
ignore callee/EH blocks, but operands can now also be operand bundles,
in which case we'll look for non-existent parameter attributes.

Unfortunately, this isn't observable in-tree, as it only crashes when
exercising the regular call lowering logic with operand bundles.
Still, this is a nice small cleanup anyway.

llvm-svn: 290905

This patch cleans up private methods for NVPTX OpenMP codegen. It converts private
members to static functions to follow the coding style of CGOpenMPRuntime.cpp and
declutter the header file.

Reviewers: ABataev
Differential Revision: https://reviews.llvm.org/D28124

llvm-svn: 290904

llvm-svn: 290903

llvm-svn: 290902

Also sort the existing list by value.

llvm-svn: 290901

Summary:
Enable disabled loopidiom test. Apparently we handle it now.
Maybe due to improvements to AA.

Reviewers: atrick, danielcdh, hfinkel

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28171

llvm-svn: 290900

llvm-svn: 290899

https://reviews.llvm.org/D17840

This patch enables private, firstprivate, and lastprivate clauses for the OpenMP distribute directive.
Regression tests differ from the similar case of the same clauses on the for directive, by removing a reference to two global variables g and g1. This is necessary because: 1. a distribute pragma is only allowed inside a target region; 2. referring a global variable (e.g. g and g1) in a target region requires the program to enclose the variable in a "declare target" region; 3. declare target pragmas, which are used to define a declare target region, are currently unavailable in clang (patch being prepared).
For this reason, I moved the global declarations into local variables.

llvm-svn: 290898

Provide a distinct contents for semBogus and semPPCDoubleDouble in order
to prevent compilers from collapsing them to a single memory address,
while we heavily rely on every semantic having distinct address.

This happens if insecure optimization collapsing identical values is
enabled. As a result, APFloats of semBogus are indistinguishable from
semPPCDoubleDouble -- and whenever the move constructor is used, the old
value beings being incorrectly recognized as a semPPCDoubleDouble.

Since the values in semPPCDoubleDouble are not used anywhere,
we can easily solve this issue via altering the value of one of the
fields and therefore ensuring that the collapse can not occur.

Differential Revision: https://reviews.llvm.org/D28112

llvm-svn: 290896

Previously it parsed /proc/<pid>/maps for every module separately
resulting in a very slow response time. This CL add some caching and
optimizes the implementation to improve the code from O(n*m) to O(n+m)
where n is the number of modules requested and m is the number of
files mapped into memory.

Differential revision: https://reviews.llvm.org/D28233

llvm-svn: 290895

Add a field indicating the associated check for every replacement to the YAML
report generated with the '-export-fixes' option.  Update
clang-apply-replacements to handle the new format.

Patch by Alpha Abdoulaye!

Differential revision: https://reviews.llvm.org/D26137

llvm-svn: 290893

Add a field indicating the associated check for every replacement to the YAML
report generated with the '-export-fixes' option.  Update
clang-apply-replacements to handle the new format.

Patch by Alpha Abdoulaye!

Differential revision: https://reviews.llvm.org/D26137

llvm-svn: 290892

Reviewers: ioeric

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D28228

llvm-svn: 290891

r290874 enabled the s390x test, which caused the rest of the tests to start
misbehaving. This is because this test switches the selected platform and the
change persists.

This fixes it by explicitly resetting the platform in a similar way to the gcore
tests do. Potentially we should consider re-setting the platform globally
between each test run to better protect tests from each other.

llvm-svn: 290890

Differential revision: https://reviews.llvm.org/D27576

Reviewers: EricWF
llvm-svn: 290889

r281179 Introduced an externally threaded variant of the libc++ library. This
patch adds support for a similar library variant for libc++abi.

Differential revision: https://reviews.llvm.org/D27575

Reviewers: EricWF
llvm-svn: 290888

This commit fixes a crash that occurs when -print-decl-contexts AST consumer
tries to print an unhandled declaration.

rdar://19467234

Differential Revision: https://reviews.llvm.org/D26964

llvm-svn: 290887

This commit fixes a crash that occurs when -print-decl-contexts AST consumer
tries to print an unhandled declaration.

rdar://19467234

Differential Revision: https://reviews.llvm.org/D26964

llvm-svn: 290886

This commit fixes a crash that occurs when -print-decl-contexts AST consumer
tries to print an unhandled declaration.

rdar://19467234

Differential Revision: https://reviews.llvm.org/D26964

llvm-svn: 290885

This commit fixes a crash that occurs when -print-decl-contexts AST consumer
tries to print an unhandled declaration.

rdar://19467234

Differential Revision: https://reviews.llvm.org/D26964

llvm-svn: 290884

Summary:
modernize-pass-by-value doesn't warn about value parameters that
cannot be moved, so performance-unnecessary-value-param should.

Reviewers: aaron.ballman, flx, alexfh

Subscribers: JDevlieghere, cfe-commits

Differential Revision: https://reviews.llvm.org/D28022

llvm-svn: 290883

This commit fixes a crash that occurs when -print-decl-contexts AST consumer
tries to print an unhandled declaration.

rdar://19467234

Differential Revision: https://reviews.llvm.org/D26964

llvm-svn: 290882