- 860a319 Fix Sema and IRGen for atomic compound assignment so it has the right semantics when promotions are involved. by Eli Friedman · 12 years ago
- cda8df0 Add XOP frcz instrinsics. by Craig Topper · 12 years ago
- 5daa4de Add XOP permute intrinsics. by Craig Topper · 12 years ago
- ecbc009 Add XOP shift and compare intrinsics. by Craig Topper · 12 years ago
- 3335eb7 Add XOP vprot* instruction intrinsics by Craig Topper · 12 years ago
- 18b73ea More XOP intrinsics by Craig Topper · 12 years ago
- 239c030 Begin adding XOP intrinsics by Craig Topper · 12 years ago
- 1c6e95f Rename a test case to a more generic name. This is a completely useless by Chandler Carruth · 12 years ago
- ec22f56 User better API for vla in compund literals. // rdar://11485774 by Fariborz Jahanian · 12 years ago
- 2ccc0f9 When emitting compund literal of vla pointer elements, make by Fariborz Jahanian · 12 years ago
- e14da79 reverse r158117. by Fariborz Jahanian · 12 years ago
- 3077577 When doing arithmatic on vla pointer, make sure by Fariborz Jahanian · 12 years ago
- f4bd4d8 Fix a bug with va_arg and vectors on Darwin x86-32. <rdar://problem/11592208>. by Eli Friedman · 12 years ago
- 97075fb Add fma3 intrinsic header file. by Craig Topper · 12 years ago
- 2a6e784 Upgrade test for the fma4->fma intrinsic change. by Benjamin Kramer · 12 years ago
- 3c0bc15 Add builtin for pclmulqdq instruction. by Craig Topper · 12 years ago
- 6b97877 Add an ammintrin.h header for SSE4a intrinsics. by Benjamin Kramer · 12 years ago
- 2813420 add CodeGen support for the alloc_size attribute by Nuno Lopes · 12 years ago
- c2c1e4f Un-XFAIL CodeGen/builtins-nvptx.c now that the proper changes have by Justin Holewinski · 12 years ago
- a1bfcbd XFAIL this test, which does not pass on trunk since the grand by John McCall · 12 years ago
- 2c585b9 Replace PTX back-end with NVPTX back-end in all places where Clang cares by Justin Holewinski · 12 years ago
- c679485 If the first argument of __builtin_object_size can be folded to a constant by Richard Smith · 12 years ago
- def1849 wire -fbounds-checking to the new LLVM bounds checking pass by Nuno Lopes · 12 years ago
- 3e86a04 revert the usage of the objectsize intrinsic with 3 parameters (to match LLVM r157255) by Nuno Lopes · 12 years ago
- ba16ac3 XFAIL this test on MIPS. by Akira Hatanaka · 12 years ago
- 6565b8a We were already passing the case in pr9356. Add it to attributes.c to make by Rafael Espindola · 12 years ago
- ee409a9 Add support for __attribute__((hot)) and __attribute__((cold)). by Benjamin Kramer · 12 years ago
- 91338cf Coerce byval aggregate arguments to integers whose size matches the integer by Akira Hatanaka · 12 years ago
- 8aeb147 Fix handling of vector return types. by Akira Hatanaka · 12 years ago
- 060ddb0 hopefully fix a bunch of ARM buildbot failures by Nuno Lopes · 12 years ago
- 08a9ae9 update calls to objectsize intrinsic to match LLVM r156473 by Nuno Lopes · 12 years ago
- a135f2c The address of a vla is actually complex and requires a dereference. by Eric Christopher · 12 years ago
- 4ccf004 hopefully unbreak some buildbots by Nuno Lopes · 12 years ago
- 9eecef1 make this test a little more useful. I'll add more tests for bound checking once -fbounds-checking patch gets in by Nuno Lopes · 12 years ago
- fd00eec This patch adds a new Clang compiler flag "-gline-tables-only". by Alexey Samsonov · 12 years ago
- 6c803f7 During block layout, after padding up to the max field alignment, by John McCall · 12 years ago
- c5713cf Get rid of storelv4si builtin as it can be expressed directly. This is general by Chad Rosier · 12 years ago
- e593921 Currently __builtin_annotation() only annotates an i32. by Julien Lerouge · 12 years ago
- 6ea4841 Fix a bug with block layout when the block contains something by John McCall · 12 years ago
- 454393e Add atan, atan2, exp, and log to the builtin math library functions. by Chad Rosier · 12 years ago
- c9fe605 enable TBAA when -fthread-sanitizer is given, even with -O0 or -relaxed-aliasing by Kostya Serebryany · 12 years ago
- 87380aa Forward declarations should take a context. This helps the debugger by Eric Christopher · 12 years ago
- a8063c3 Do not use stdint.h, driver might provide invalid location for it. Instead, provide the types directly. by Anton Korobeynikov · 12 years ago
- 49a110d Convert vperm2f128 and vperm2i128 intrinsics back to using llvm intrinsics. Unfortunately, these instructions have behavior that can't be modeled with shuffle vector. by Craig Topper · 12 years ago
- 7d6c8a1 Attempt to fix test. by Eli Friedman · 12 years ago
- 7fea7c8 Make sure EmitMoveFromReturnSlot is passing the correct alignment to by Chad Rosier · 12 years ago
- 3c6ed1b Remove support for -fast-math metadata for the moment. by Duncan Sands · 12 years ago
- 10e675d Generate fpmath metadata when -ffast-math. Note that no optimizations are hooked by Duncan Sands · 12 years ago
- 377ecc7 Propagate alignment on lvalues through EmitLValueForField. PR12395. by Eli Friedman · 12 years ago
- b5491f3 Change _mm256_permute4x64_epi64 and _mm256_permute4x64_pd to use builtin_shufflevector instead of specific builtins. Old builtins will be removed from llvm now that vpermq/vpermpd are supported by shuffle lowering code. by Craig Topper · 12 years ago
- 2d7cb06 Use MDBuilder to help with metadata creation. by Duncan Sands · 12 years ago
- 00000b0 First set of tests for ARM homogenous aggregates. C only. C++ will follow. by Anton Korobeynikov · 12 years ago
- 51b9240 Implement __atomic_fetch_nand and __atomic_nand_fetch to complete our set of by Richard Smith · 12 years ago
- 2c39d71 Implement the missing pieces needed to support libstdc++4.7's <atomic>: by Richard Smith · 12 years ago
- ff34d40 Implement support for 18 of the GNU-compatible __atomic builtins. by Richard Smith · 12 years ago
- fafbf06 Provide, and document, a set of __c11_atomic_* intrinsics to implement C11's by Richard Smith · 12 years ago
- e1b2abc AtomicExpr: make ASTStmtReader a friend and remove setters. Also fix saving by Richard Smith · 12 years ago
- b2561c7 This test actually needs label names. by Eric Christopher · 12 years ago
- c078337 For debug and coverage analysis if we're not optimizing go ahead by Eric Christopher · 12 years ago
- 87b9c03 Make the variant of __builtin_shufflevector that takes the shuffle indexes as a vector actually usable. Patch by David Neto. PR12465. by Eli Friedman · 12 years ago
- 0ea6164 Add more constraint registers for mips. by Eric Christopher · 12 years ago
- 5aeaca3 Correctly check argument types for some vector macros in smmintrin.h. Put parentheses around uses of vector macro arguments. by Craig Topper · 12 years ago
- 5228982 Add _mm_minpos_epu16 to smmintrin.h. Fixes PR12399. by Craig Topper · 12 years ago
- 5629646 Fix shuffle vector calculation for mm_permute_ps. Fixes PR 12401. by Craig Topper · 12 years ago
- 84392d0 The UTF16 string referenced by a CFString should go into the __TEXT,__ustring by Bill Wendling · 12 years ago
- 882784d Relax the test slightly so that it should pass on other people's computers. by David Chisnall · 12 years ago
- 3a7d69b Call out to GCC-compatible runtime functions for atomic ops that we can't use by David Chisnall · 12 years ago
- d1f853d Add better support for $fp and $sp for mips inline asm support. by Eric Christopher · 12 years ago
- 5432ff2 [tests] Fix test failure in release mode. by Daniel Dunbar · 12 years ago
- 8cb4a07 add tbaa metadata to vtable pointer loads/stores by Kostya Serebryany · 12 years ago
- c3f8955 Add back r153360 with a fix for enums that cover all the 32 bit values. by Rafael Espindola · 12 years ago
- 88a569a Revert r153360 (and r153380), "Second part of PR12251. Produce the range metadata in clang for booleans and". by NAKAMURA Takumi · 12 years ago
- e27af29 Second part of PR12251. Produce the range metadata in clang for booleans and by Rafael Espindola · 12 years ago
- 263e0a6 Small cleanup. by Bill Wendling · 12 years ago
- 8ff6059 Use correct FileCheck regexp. by Bill Wendling · 12 years ago
- acb5310 When an MMX output variable is tied to the input variable, we have to implicitly by Bill Wendling · 12 years ago
- bce35a2 line endings by Matt Beaumont-Gay · 12 years ago
- e5a8aeb Make sure we correctly set the alignment for vector loads and stores associated with vector element lvalues. Patch by Kevin Schoedel (with some minor modifications by me). by Eli Friedman · 12 years ago
- ff99320 Don't try to create "store atomic" instructions of non-integer types; they aren't supported at the moment. PR12040. by Eli Friedman · 12 years ago
- a43ef3e When a variable has a specified asm name, but isn't using the register storage class, the asm name doesn't specify a register. PR12244. by Eli Friedman · 12 years ago
- fc055f9 [frontend] Fix how the frontend handles -fno-inline. AFAIK clang now matches by Chad Rosier · 12 years ago
- bd86ad5 Fix codegen for vld{3,4}_dup intrinsics. by James Molloy · 12 years ago
- 8fbe385 [frontend] Add support for disabling the "inline" keyword using by Chad Rosier · 12 years ago
- efe9c0d Fix the long double to be of width/align 64. Rename va_list_test to by Roman Divacky · 12 years ago
- e45b9b7 Use ZeroLengthBitfieldAlignment for AAPCS, as well as APCS-GNU. by James Molloy · 12 years ago
- edfac03 IRgen/ABI/x86_64: Avoid passing small structs using byval sometimes. by Daniel Dunbar · 12 years ago
- 634a4b1 [driver] What was implemented in r152130 was actually -fno-inline-functions, not by Chad Rosier · 12 years ago
- 9f86af8 Whitespace. by Chad Rosier · 12 years ago
- 250008b [driver] Add support for -fno-inline. rdar://10972766 by Chad Rosier · 12 years ago
- 5bd20ff From his comment in PR12168, Eli seems confused about the alignment assumptions by Chris Lattner · 12 years ago
- d6e7356 add a testcase for PR12094 and fix a crash on pointer to incomplete type, by Chris Lattner · 12 years ago
- 75a96d0 I referenced the incorrect rdar in my previous commit (r151745). Add the by Chad Rosier · 12 years ago
- 3f3335d Allocate TargetLibraryInfo for the CodeGen passes. Otherwise, it's instantiated by Chad Rosier · 12 years ago
- 16f1f71 Reapply r151638 and r151641. by James Molloy · 12 years ago
- 32a9a75 Revert r151638 because it causes assertion hit on PCH creation for Cocoa.h by Argyrios Kyrtzidis · 12 years ago
- fbcf040 Correctly track tags and enum members defined in the prototype of a function, and ensure they are properly scoped. by James Molloy · 12 years ago
- fce4715 Add missing code for compound literals of complex type. <rdar://problem/10938628> by Eli Friedman · 12 years ago
- 07d1642 Test case for r151429/r151430, which ensures llvm.lifetime intrinsics are not by Chad Rosier · 12 years ago
- 5a13d4d Fix a stupid mistake in r151133. Reported to me by Joerg Sonnenberger. by Eli Friedman · 12 years ago
- 9428772 Turned on support for __declspecs: noreturn, noinline, nothrow and naked in MS compatibility mode. by Aaron Ballman · 12 years ago