llvm-svn: 170133

Patch by Anastasi Voitova with with small fixes by me.

llvm-svn: 170132

llvm-svn: 170131

llvm-svn: 170130

llvm-svn: 170129

This assumes (1 << n) is always not zero. Consider n is greater than word size.
Although I know it is undefined, this transforms undefined behavior hidden.

This led clang unexpected behavior with some failures. I will investigate to fix undefined shl in clang.

llvm-svn: 170128

Patch contributed by Philip Craig!

llvm-svn: 170127

to package reST-formatted documentation.

llvm-svn: 170126

Remove little endian specification from SPIR data layout - SPIR doesn't define endiannes in the data layout.

llvm-svn: 170125

llvm-svn: 170124

llvm-svn: 170123

[asan] extend overflow-in-qsort.cc to check both unwind kinds. One more test for SlowUnwind. Move both to Linux dir (no slow unwind on Mac)

llvm-svn: 170122

llvm-svn: 170121

llvm-svn: 170120

llvm-svn: 170119

llvm-svn: 170118

[asan] add two asan flags: fast_unwind_on_fatal and fast_unwind_on_malloc to allow using the slow CFI-based unwinder

llvm-svn: 170117

seems to cause some weird stack overflow

llvm-svn: 170116

[Sanitizer] Add new header - sanitizer_platform_interceptors.h, which should define whether or not given function should be intercepted on a given platform.

llvm-svn: 170115

llvm-svn: 170114

llvm-svn: 170113

llvm-svn: 170112

llvm-svn: 170111

llvm-svn: 170110

llvm-svn: 170109

bit-patterns which are not valid values for enumerated or boolean types.
These checks are the ubsan analogue of !range metadata.

llvm-svn: 170108

bit-patterns which are not valid values for enumerated or boolean types.
These checks are the ubsan analogue of !range metadata.

llvm-svn: 170107

it seems to be breaking self-host for a few people and is PR14592.

This reverts commit r170024.

llvm-svn: 170106

This reverts commit r170095 since it appears to be breaking the bots.

llvm-svn: 170105

Accordingly, add helper funtions getSimpleValueType (in parallel to
getValueType) in SDValue, SDNode, and TargetLowering.

This is the first, in a series of patches.

This is the second attempt. In the first attempt (r169837), a few
getSimpleVT() were hoisted too far, detected by bootstrap failures.

llvm-svn: 170104

[sanitizer] replace size_t with SIZE_T in interceptors (same for ssize_t, off_t and off64_t). This is done to avoid confusion and to make the code work with compilers that define size_t w/o any includes (MSVC)

llvm-svn: 170103

Based on code written by Riyadh Baghdadi.

llvm-svn: 170102

llvm-svn: 170101

llvm-svn: 170100

llvm-svn: 170099

llvm-svn: 170098

[sanitizer] change the way SizeClassAllocator64 allocated memory from the system: instead of one huge mmap(NORESERVE) it does one huge mprotect and then does small on-demand mmaps. This allows us to call OnMap callbacks which are required to poison newly allocated memory in asan

llvm-svn: 170097

Patch by Andrew Turner.

llvm-svn: 170096

On MachO, sections also have segment names. When a tool looking at a .o file
prints a segment name, this is what they mean. In reality, a .o has only one,
anonymous, segment.

This patch adds a MachO only function to fetch that segment name. I named it
getSectionFinalSegmentName since the main use for the name seems to be informing
the linker with segment this section should go to.

The patch also changes MachOObjectFile::getSectionName to return just the
section name instead of computing SegmentName,SectionName.

llvm-svn: 170095

llvm-svn: 170094