Dmitry Vyukov afdcc96d9f tsan: optimize memory access functions ... The optimization is two-fold: First, the algorithm now uses SSE instructions to handle all 4 shadow slots at once. This makes processing faster. Second, if shadow contains the same access, we do not store the event into trace. This increases effective trace size, that is, tsan can remember up to 10x more previous memory accesses. Perofrmance impact: Before: [ OK ] DISABLED_BENCH.Mop8Read (2461 ms) [ OK ] DISABLED_BENCH.Mop8Write (1836 ms) After: [ OK ] DISABLED_BENCH.Mop8Read (1204 ms) [ OK ] DISABLED_BENCH.Mop8Write (976 ms) But this measures only fast-path. On large real applications the speedup is ~20%. Trace size impact: On app1: Memory accesses : 1163265870 Including same : 791312905 (68%) on app2: Memory accesses : 166875345 Including same : 150449689 (90%) 90% of filtered events means that trace size is effectively 10x larger. llvm-svn: 209897		2014-05-30 13:36:29 +00:00
clang	Remove unnecessary test.	2014-05-30 13:19:28 +00:00
clang-tools-extra	Revert "Remove redundant check discovered in post-commit review of r209505."	2014-05-30 13:31:14 +00:00
compiler-rt	tsan: optimize memory access functions	2014-05-30 13:36:29 +00:00
debuginfo-tests	…
libclc	Add more log related float constants	2014-05-29 21:30:28 +00:00
libcxx	Add a _LIBCPP_CONSTEXPR that was missed in r170026.	2014-05-30 12:09:47 +00:00
libcxxabi	Fix typo in comment.	2014-05-14 17:16:57 +00:00
lld	[ELF] Fix incorrect sorting of .init_array / .fini_array sections.	2014-05-30 05:29:46 +00:00
lldb	iOS simulator cleanup to make sure we use "*-apple-ios" for iOS simulator apps and binaries.	2014-05-29 21:33:45 +00:00
llvm	ARM: use AAPCS-style prologues for embedded MachO.	2014-05-30 13:23:06 +00:00
openmp	Add testsuite from OpenUH	2014-05-10 17:02:09 +00:00
polly	Added option for n-dimensional rectangular tiling	2014-05-28 17:21:02 +00:00