- a124f94 Make LowerSIGN_EXTEND_INREG split 256-bit vectors when AVX1 is enabled and use AVX2 shifts when AVX2 is enabled. by Craig Topper · 13 years ago
- 7912ef9 Less template, more virtual! Refactoring suggested by Chris in code review. by Nick Lewycky · 13 years ago
- 173862e Refactor code to use new attribute getters on CallSite for NoCapture and ByVal. by Nick Lewycky · 13 years ago
- 742e5cf test/CodeGen/X86/block-placement.ll: Relax expressions for Win32. by NAKAMURA Takumi · 13 years ago
- b0dadb9 The logic for breaking the CFG in the presence of hot successors didn't by Chandler Carruth · 13 years ago
- a91a3e0 Make an obviously const interface actually be marked as const. by Chandler Carruth · 13 years ago
- b09a3aa XFAIL this test until I figure out what indvars is doing here (or find someone who does) by Benjamin Kramer · 13 years ago
- 13c3c75 SCEV: Actually set overflow flags on add expressions. by Benjamin Kramer · 13 years ago
- 2901243 Add some comments to the latest test case I added here to document what by Chandler Carruth · 13 years ago
- 0d86d46 Add code for lowering v32i8 shifts by a splat to AVX2 immediate shift instructions. Remove 256-bit splat handling from LowerShift as it was already handled by PerformShiftCombine. by Craig Topper · 13 years ago
- 745a86b Use 256-bit vcmpeqd for creating an all ones vector when AVX2 is enabled. by Craig Topper · 13 years ago
- ba798c5 Remove some of the special classes that worked around an old tablegen limitation of not being able to remove redundant bitconverts from patterns. by Craig Topper · 13 years ago
- 98fc729 Custom lower AVX2 variable shift intrinsics to shl/srl/sra nodes and remove the intrinsic patterns. by Craig Topper · 13 years ago
- 03300ec Move the handling of unanalyzable branches out of the loop-driven chain by Chandler Carruth · 13 years ago
- 6bf57b0 Test cases for SSSE3/AVX integer horizontal add/sub. by Craig Topper · 13 years ago
- 54f952a Synthesize SSSE3/AVX 128-bit horizontal integer add/sub instructions from add/sub of appropriate shuffle vectors. by Craig Topper · 13 years ago
- 3113384 Collapse X86 PSIGNB/PSIGNW/PSIGND node types. by Craig Topper · 13 years ago
- 1666cb6 Extend VPBLENDVB and VPSIGN lowering to work for AVX2. by Craig Topper · 13 years ago
- 787a88f Remove some unnecessary filtering checks from X86 disassembler table build. by Craig Topper · 13 years ago
- 60d9a92 Remove unused parameters from the AVX maskmov classes. by Craig Topper · 13 years ago
- 5865a8d Fix a corner case in updating LoopInfo after fully unrolling an outer loop. by Andrew Trick · 13 years ago
- cbbe33f Add AVX2 vpbroadcast support by Nadav Rotem · 13 years ago
- 5a3a9c9 [asan] workaround for reg alloc bug 11395: don't instrument functions with large chunks of inline assembler by Kostya Serebryany · 13 years ago
- 424fe0e Guard call to getRegForValue with isTypeLegal check to avoid unnecessary work/dead code. by Chad Rosier · 13 years ago
- ce35d8b DISubrange supports unsigned lower/upper array bounds, so let's not fake it in the end while emitting DWARF. If a FE needs to encode signed lower/upper array bounds then we need to extend DISubrange or ad DISignedSubrange. by Devang Patel · 13 years ago
- 2e7fb2f quick fix: remove GlobalVariable::GlobalVariable mistakenly commited at r144933. For some reason this compiles on linux by Kostya Serebryany · 13 years ago
- 4f30524 Fix an overly general check in SimplifyIndvar to handle useless phi cycles. by Andrew Trick · 13 years ago
- 7cf2a04 fall back to explicit list of allowed linkages when instrumenting globals in asan; add a test check that asan does not touch linkonce_odr by Kostya Serebryany · 13 years ago
- f249449 Fix bug in RefCountedBase/RefCountedBaseVPTR where the reference count was accidentally copied as part of the copy constructor. This could result in objects getting leaked because there reference count was too high. by Ted Kremenek · 13 years ago
- 944d82b Add TODO comment. by Chad Rosier · 13 years ago
- d90a191 Fix SSE/AVX integer comparison patterns to understand that all integer vector loads are promoted to i64 vector loads so patterns need a bitconvert. Also slightly simplify the AVX2 variable shift patterns by using the predefined bitconvert pattern fragments. by Craig Topper · 13 years ago
- 2fb82ce Dead code. by Chad Rosier · 13 years ago
- 478b06c When fast iseling a GEP, accumulate the offset rather than emitting a series of by Chad Rosier · 13 years ago
- ec43d1f Remove seemingly unnecessary duplicate VROUND definitions. by Craig Topper · 13 years ago
- c78daaf x86/windows issues fixed. by Chris Lattner · 13 years ago
- 9d434db Add support for custom names for library functions in TargetLibraryInfo. Add a custom name for fwrite and fputs on x86-32 OSX. Make SimplifyLibCalls honor the custom by Eli Friedman · 13 years ago
- d5889d8 llvm-build: Attempt to work around a CMake Makefile generator bug that doesn't by Daniel Dunbar · 13 years ago
- 3bdb3c9 Don't unconditionally set the kill flag. rdar://10456186 by Chad Rosier · 13 years ago
- d224c78 Turn on vzeroupper insertion on call boundaries for AVX; it works as far as I know, and I'd like to see wider testing. by Eli Friedman · 13 years ago
- 708276d build/make/test: Get rid of unused BUGPOINT_TOPTS variable. by Daniel Dunbar · 13 years ago
- 4db4add Make sure to replace the chain properly when DAGCombining a LOAD+EXTRACT_VECTOR_ELT into a single LOAD. Fixes PR10747/PR11393. by Eli Friedman · 13 years ago
- 11ba26d Object/COFF: Support common symbols. by Michael J. Spencer · 13 years ago
- 9b1671b Remove obsolete test. by Jim Grosbach · 13 years ago
- 2abba84 Generalize the fixup info for ARM mode. by Jim Grosbach · 13 years ago
- 47a906a Update test for r144842. by Jim Grosbach · 13 years ago
- 620db89 Lower 64-bit constant pool node. by Akira Hatanaka · 13 years ago
- 9b944a8 Lower 64-bit block address. by Akira Hatanaka · 13 years ago
- b84acd2 Fix encoding of NOP used for padding in ARM mode .align. by Jim Grosbach · 13 years ago
- 74c7634 Add patterns for 64-bit tglobaladdr, tblockaddress, tjumptable and tconstpool by Akira Hatanaka · 13 years ago
- 4fd40b3 64-bit jump register instruction. by Akira Hatanaka · 13 years ago
- 2b89498 Another missing X86ISD::MOVLPD pattern. rdar://10450317 by Evan Cheng · 13 years ago
- 40a86ee ARM assembly parsing for shifted register operands for MOV instruction. by Jim Grosbach · 13 years ago
- efed3d1 Clean up debug printing of ARM shifted operands. by Jim Grosbach · 13 years ago
- 053e69a Add fast-isel stats to determine who's doing all the work, the by Chad Rosier · 13 years ago
- f91488c Fix the stats collection for fast-isel. The failed count was only accounting by Chad Rosier · 13 years ago
- 2929de4 There are already problems with building LLVM under VS2005, and it's by Chandler Carruth · 13 years ago
- b598b04 ARM assmebly two operand forms for LSR, ASR, LSL, ROR register. by Jim Grosbach · 13 years ago
- 48b368b ARM assembly parsing for RRX mnemonic. by Jim Grosbach · 13 years ago
- cd75e44 Added missing comment about new custom lowering of DEC64 by Pete Cooper · 13 years ago
- c3aa7c5 Disable expensive two-address optimizations at -O0. rdar://10453055 by Evan Cheng · 13 years ago
- 508a1f4 Check to make sure we can select the instruction before trying to put the by Chad Rosier · 13 years ago
- 14117c4 Disable the assertion again. Looks like fastisel is still generating bad kill markers. by Evan Cheng · 13 years ago
- 23f2207 ARM mode aliases for bitwise instructions w/ register operands. by Jim Grosbach · 13 years ago
- d0405aa Fix tablegen warning: hasSideEffects is inferred for eh_sjlj_dispatchsetup. by Bob Wilson · 13 years ago
- 5c283e9 lib/Target/ARM/CMakeLists.txt: Disable optimization in ARMISelLowering.cpp also on MSC15(aka VS9). Seems miscompiled. by NAKAMURA Takumi · 13 years ago
- b95fc31 Sink codegen optimization level into MCCodeGenInfo along side relocation model by Evan Cheng · 13 years ago
- f1b41dd Record landing pads with a SmallSetVector to avoid multiple entries. by Bob Wilson · 13 years ago
- 12755b0 Fix the execution domain on a bunch of SSE/AVX instructions. by Craig Topper · 13 years ago
- 20c918d Update the SP in the SjLj jmpbuf whenever it changes. <rdar://problem/10444602> by Bob Wilson · 13 years ago
- eaab6ef Fix ARM SjLj-EH dispatch setup code. <rdar://problem/10444602> by Bob Wilson · 13 years ago
- 2713d04 Remove code to enable execution dependency fix pass on VR256. VR128 is sufficient after r144636. by Craig Topper · 13 years ago
- 0a405ae Revert r144568 now that r144730 has fixed the fast-isel kill marker bug. by Evan Cheng · 13 years ago
- f8f558d Fix typo in test. by Nick Lewycky · 13 years ago
- ae10dd2 Merge isObjectPointerWithTrustworthySize with getPointerSize. Use it when by Nick Lewycky · 13 years ago
- 9bad88a If the 2addr instruction has other kills, don't move it below any other uses since we don't want to extend other live ranges. by Evan Cheng · 13 years ago
- 2bee6a8 RescheduleKillAboveMI() must backtrack to before the rescheduled DBG_VALUE instructions. rdar://10451185 by Evan Cheng · 13 years ago
- ae7db7a Process all uses first before defs to accurately capture register liveness. rdar://10449480 by Evan Cheng · 13 years ago
- ee94dc2 Fix testcase. by Eli Friedman · 13 years ago
- d577df8 CONCAT_VECTORS can have more than two operands. PR11389. by Eli Friedman · 13 years ago
- b91b600 Add a couple asserts so it will be easier to debug if we accidentally pass indexed loads/stores to the legalizer. by Eli Friedman · 13 years ago
- a9a4f5e Remove extra ,. by Michael J. Spencer · 13 years ago
- 800e03f AddressSanitizer, first commit (compiler module only) by Kostya Serebryany · 13 years ago
- 8a2549f Object/Archive: Give Child a operator < for map. by Michael J. Spencer · 13 years ago
- dc29663 Support/COFF: Add structs and enums from the standard for image files. by Michael J. Spencer · 13 years ago
- f81285c llvm-objdump: Ignore non-objects in archives. by Michael J. Spencer · 13 years ago
- a2a2d1f test commit to verify that commit access works (added blank line) by Kostya Serebryany · 13 years ago
- 99aa14f Rename MVT::untyped to MVT::Untyped to match similar nomenclature. by Owen Anderson · 13 years ago
- 79f0bfc Fix SCEV overly optimistic back edge taken count for multi-exit loops. by Andrew Trick · 13 years ago
- f56c60b Add FIXME comment. by Chad Rosier · 13 years ago
- 3805d85 Enable -widen-vmovs by default. by Jakob Stoklund Olesen · 13 years ago
- 8368f74 Stabilize the output of the dwarf accelerator tables. Fixes a comparison by Eric Christopher · 13 years ago
- 22b34cc GEPs with all zero indices are trivially coalesced by fast-isel. For example, by Chad Rosier · 13 years ago
- e43862b ARM assembly parsing for register range syntax for VLD/VST register lists. by Jim Grosbach · 13 years ago
- 6ac5b16 Merge ObjCPropertyDebugInfo.html into SourceLevelDebugging.html by Devang Patel · 13 years ago
- 5b2fb20 ARM assembly parsing for data type suffices on NEON VMOV aliases. by Jim Grosbach · 13 years ago
- de63112 Fix MSVC warnings by adding a cast. by Nadav Rotem · 13 years ago
- f8c10e5 AVX: Add support for vbroadcast from BUILD_VECTOR and refactor some of the vbroadcast code. by Nadav Rotem · 13 years ago
- 62f009a jakob fixed X87 inline asm! by Chris Lattner · 13 years ago
- 2fdd005 add ImmutableSet/Map dox, patch by Caitlin Sadowski! by Chris Lattner · 13 years ago
- ec0af2f test/CodeGen/X86/dec-eflags-lower.ll: Relax expression for win32 x64. by NAKAMURA Takumi · 13 years ago