Justin Lebar 2e4ecfdebe [CUDA] Implement __ldg using intrinsics. ... Summary: Previously it was implemented as inline asm in the CUDA headers. This change allows us to use the [addr+imm] addressing mode when executing ld.global.nc instructions. This translates into a 1.3x speedup on some benchmarks that call this instruction from within an unrolled loop. Reviewers: tra, rsmith Subscribers: jhen, cfe-commits, jholewinski Differential Revision: http://reviews.llvm.org/D19990 llvm-svn: 270150		2016-05-19 22:49:13 +00:00
clang	[CUDA] Implement __ldg using intrinsics.	2016-05-19 22:49:13 +00:00
clang-tools-extra	[include-fixer] Fix unused variable warning in Release builds.	2016-05-19 16:57:57 +00:00
compiler-rt	[profile] entry eviction support in value profiler	2016-05-19 21:35:34 +00:00
debuginfo-tests	New round of fixes for "Always compile debuginfo-tests for the host triple"	2014-10-18 23:47:59 +00:00
libclc	math: Use single precision fmax in sp path	2016-05-17 19:44:01 +00:00
libcxx	Cleanup superfluous std:: qualifiers in <type_traits>	2016-05-18 23:09:24 +00:00
libcxxabi	libc++abi: make __cxa_call_unexpected visible	2016-05-11 23:56:37 +00:00
libunwind	unwind: remove last instance of -Wexpansion-to-defined	2016-04-26 01:11:29 +00:00
lld	Fix the function to set the section VMA/LMA fields in case of using	2016-05-19 18:15:54 +00:00
lldb	Remove a should have been deleted extra assignment to a variable.	2016-05-19 22:22:57 +00:00
llgo	[llgo] llgoi: separate evaluation from printing	2016-04-25 01:18:20 +00:00
llvm	Reapply r263460: [SpillPlacement] Fix a quadratic behavior in spill placement.	2016-05-19 22:40:37 +00:00
openmp	Remove unnecessary unistd.h header from tests.	2016-05-18 21:36:34 +00:00
polly	Revert "Optimistic assume required invariant loads to be invariant"	2016-05-19 13:47:34 +00:00