Programming Languages Research Group: Git

2015-11-17	Ahmed Bougacha	[AArch64] Promote f16 SELECT_CC CC operands when op...	commit \| commitdiff \| tree
2015-11-11	Ahmed Bougacha	[MC] Use LShr for constant evaluation of ">>" on non...	commit \| commitdiff \| tree
2015-11-06	Ahmed Bougacha	[AArch64][FastISel] Don't even try to select vector...	commit \| commitdiff \| tree
2015-11-06	Ahmed Bougacha	[X86] Fold (trunc (i32 (zextload i16))) into vbroadcast.	commit \| commitdiff \| tree
2015-11-06	Ahmed Bougacha	[X86] SRL non-LSB extracts when folding to truncating...	commit \| commitdiff \| tree
2015-11-06	Ahmed Bougacha	[X86] Don't fold non-LSB extracts into truncating broadcasts.	commit \| commitdiff \| tree
2015-10-01	Ahmed Bougacha	[X86] Don't custom-lower vNi32 uint_to_fp when unsafe...	commit \| commitdiff \| tree
2015-09-26	Ahmed Bougacha	[ARM] Don't generate clrex for pre-v7 targets.	commit \| commitdiff \| tree
2015-09-22	Ahmed Bougacha	[ARM] Emit clrex in the expanded cmpxchg fail block.	commit \| commitdiff \| tree
2015-09-22	Ahmed Bougacha	[AArch64] Emit clrex in the expanded cmpxchg fail block.	commit \| commitdiff \| tree
2015-09-12	Ahmed Bougacha	[CodeGen] Fix AtomicExpand invalidation issue caused...	commit \| commitdiff \| tree
2015-09-11	Ahmed Bougacha	[CodeGen] Refactor TLI/AtomicExpand interface to make...	commit \| commitdiff \| tree
2015-09-11	Ahmed Bougacha	[CodeGen] Rename AtomicRMWExpansionKind to AtomicExpansionKind.	commit \| commitdiff \| tree
2015-09-10	Ahmed Bougacha	[AArch64] Match FI+offset in STNP addressing mode.	commit \| commitdiff \| tree
2015-09-10	Ahmed Bougacha	[AArch64] Match base+offset in STNP addressing mode.	commit \| commitdiff \| tree
2015-09-10	Ahmed Bougacha	[AArch64] Support selecting STNP.	commit \| commitdiff \| tree
2015-09-10	Ahmed Bougacha	[CodeGen] Make x86 nontemporal store patfrags generic...	commit \| commitdiff \| tree
2015-09-02	Ahmed Bougacha	[X86] Require 32-byte alignment for 32-byte VMOVNTs.	commit \| commitdiff \| tree
2015-09-02	Ahmed Bougacha	[X86] Cleanup nontemporal tests a little. NFC.	commit \| commitdiff \| tree
2015-09-02	Ahmed Bougacha	[X86] Cleanup nontemporal fragments. NFCI.	commit \| commitdiff \| tree
2015-09-02	Ahmed Bougacha	[AArch64] More consistently separate asm opc and operands...	commit \| commitdiff \| tree
2015-09-02	Ahmed Bougacha	[AArch64] Consistently separate asm opc and operands...	commit \| commitdiff \| tree
2015-09-01	Ahmed Bougacha	[ARM] Don't abort on variable-idx extractelt in ReconstructS...	commit \| commitdiff \| tree
2015-09-01	Ahmed Bougacha	[AArch64] Lower READCYCLECOUNTER using MRS PMCCTNR_EL0.	commit \| commitdiff \| tree
2015-08-28	Ahmed Bougacha	[CodeGen] Support (and default to) expanding READCYCLECOUNTE...	commit \| commitdiff \| tree
2015-08-27	Ahmed Bougacha	[CodeGen] Check FoldConstantArithmetic result before...	commit \| commitdiff \| tree
2015-08-27	Ahmed Bougacha	[TableGen][DAGISel] Dedup predicates with same code...	commit \| commitdiff \| tree
2015-08-27	Ahmed Bougacha	[TableGen] Remove dead code. NFC.	commit \| commitdiff \| tree
2015-08-20	Ahmed Bougacha	[X86] Look for scalar through one bitcast when lowering...	commit \| commitdiff \| tree
2015-08-20	Ahmed Bougacha	[X86] Add some broadcast-from-memory tests.	commit \| commitdiff \| tree
2015-08-20	Ahmed Bougacha	[X86] Replace avx2 broadcast intrinsics with native IR.	commit \| commitdiff \| tree
2015-08-19	Ahmed Bougacha	[AArch64] Improve short-form diags on long-form Match_Invali...	commit \| commitdiff \| tree
2015-08-14	Ahmed Bougacha	[AArch64] Fix FMLS scalar-indexed-from-2s-after-neg...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[AArch64] Provide "too few operands" diags on short...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[CodeGen] Mark the promoted FCOPYSIGN result FP_ROUND...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[AArch64] Cleanup vector-fcopysign.ll test. NFC.	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[AArch64] Also custom-lowering mismatched vector/f16...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[CodeGen] Assert on getNode(FP_EXTEND) with a smaller...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[CodeGen] When Promoting, don't extend the 2nd FCOPYSIGN...	commit \| commitdiff \| tree
2015-08-13	Ahmed Bougacha	[CodeGen] Simplify getNode(*EXT/TRUNC) type size assert...	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64] Rename FP formats to be more consistent....	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64] Add isel support for f16 indexed LD/ST.	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64][v8.1a] The "pan" sysreg isn't MSR-specific...	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64] Remove unnecessary "break". NFC.	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64] Use SDValue bool operator. NFC.	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[AArch64] Vector FCOPYSIGN supports Custom-lowering...	commit \| commitdiff \| tree
2015-08-04	Ahmed Bougacha	[CodeGen] Fix FCOPYSIGN legalization to account for...	commit \| commitdiff \| tree
2015-06-25	Ahmed Bougacha	[X86] Accept hasAVX512() as well as hasFMA() when generating...	commit \| commitdiff \| tree
2015-06-25	Ahmed Bougacha	[X86] Cleanup fma tests a little bit. NFC.	commit \| commitdiff \| tree
2015-06-24	Ahmed Bougacha	[X86] Don't generate vbroadcasti128 for v4i64 splats...	commit \| commitdiff \| tree
2015-06-24	Ahmed Bougacha	[X86] update_llc_test_checks vector-shuffle-*. NFC.	commit \| commitdiff \| tree
2015-06-22	Ahmed Bougacha	[X86] Teach load folding to accept scalar _Int users...	commit \| commitdiff \| tree
2015-06-19	Ahmed Bougacha	[ARM] Look through concat when lowering in-place shuffles...	commit \| commitdiff \| tree
2015-06-19	Ahmed Bougacha	[ARM] Factor out two-result shuffle matching. NFCI.	commit \| commitdiff \| tree
2015-06-19	Ahmed Bougacha	[ARM] Add D-sized vtrn/vuzp/vzip tests, and cleanup...	commit \| commitdiff \| tree
2015-06-17	Ahmed Bougacha	[CodeGenPrepare] Generalize inserted set from truncs...	commit \| commitdiff \| tree
2015-06-16	Ahmed Bougacha	[AArch64] Generalize extract-high DUP extension to...	commit \| commitdiff \| tree
2015-06-16	Ahmed Bougacha	[AArch64] Robustize neon-2velem-high test. NFC.	commit \| commitdiff \| tree
2015-06-11	Ahmed Bougacha	[CodeGen] ArrayRef'ize cond/pred in various TII APIs...	commit \| commitdiff \| tree
2015-06-04	Ahmed Bougacha	[GlobalMerge] Take into account minsize on Global users...	commit \| commitdiff \| tree
2015-05-29	Ahmed Bougacha	[TableGen][AsmMatcherEmitter] Only parse isolated tokens...	commit \| commitdiff \| tree
2015-05-29	Ahmed Bougacha	[TableGen][AsmMatcherEmitter] Factor out AsmOperand...	commit \| commitdiff \| tree
2015-05-29	Ahmed Bougacha	[IR] fptrunc-of-fptrunc isn't an EliminableCastPair.	commit \| commitdiff \| tree
2015-05-22	Ahmed Bougacha	[AArch64][CGP] Sink zext feeding stxr/stlxr into the...	commit \| commitdiff \| tree
2015-05-22	Ahmed Bougacha	[AArch64] Robustize atomic cmpxchg test a little more...	commit \| commitdiff \| tree
2015-05-22	Ahmed Bougacha	[AArch64] Robustize atomic cmpxchg test. NFC.	commit \| commitdiff \| tree
2015-05-21	Ahmed Bougacha	[MemCpyOpt] Do move the memset, but look at its dest...	commit \| commitdiff \| tree
2015-05-21	Ahmed Bougacha	[MemCpyOpt] Pass Instruction to IRBuilder, no need...	commit \| commitdiff \| tree
2015-05-20	Ahmed Bougacha	[MemCpyOpt] Don't move the memset when optimizing memset...	commit \| commitdiff \| tree
2015-05-16	Ahmed Bougacha	[MemCpyOpt] Turn memcpy from just-memset'd source into...	commit \| commitdiff \| tree
2015-05-16	Ahmed Bougacha	[MemCpyOpt] Remove dead argument. NFC.	commit \| commitdiff \| tree
2015-05-16	Ahmed Bougacha	Remove dead code in testcase. NFC.	commit \| commitdiff \| tree
2015-05-14	Ahmed Bougacha	[CodeGen] Use standard -not gnueabi- naming for f16...	commit \| commitdiff \| tree
2015-05-11	Ahmed Bougacha	[MemCpyOpt] Look at any dependency -not just source...	commit \| commitdiff \| tree
2015-05-06	Ahmed Bougacha	[ARM][FastISel] Use TST #1 instead of CMP #0 for select.	commit \| commitdiff \| tree
2015-05-06	Ahmed Bougacha	[WinEH] Reset WinEHPrepare::SEHExceptionCodeSlot when...	commit \| commitdiff \| tree
2015-04-30	Ahmed Bougacha	Flip r236172 testcase RUN option ordering for BSD sed...	commit \| commitdiff \| tree
2015-04-28	Ahmed Bougacha	[MC] Use LShr for constant evaluation of ">>" on ELF...	commit \| commitdiff \| tree
2015-04-28	Ahmed Bougacha	[MC] Split MCBinaryExpr::Shr into LShr and AShr.	commit \| commitdiff \| tree
2015-04-28	Ahmed Bougacha	[MC] Move getBinOpPrecedence into AsmParser. NFC.	commit \| commitdiff \| tree
2015-04-27	Ahmed Bougacha	[AArch64] Also combine vector selects fed by non-i1...	commit \| commitdiff \| tree
2015-04-27	Ahmed Bougacha	[AArch64] Don't assert when combining (v3f32 select...	commit \| commitdiff \| tree
2015-04-21	Ahmed Bougacha	[MemCpyOpt] Use the raw i8* dest when optimizing memset...	commit \| commitdiff \| tree
2015-04-18	Ahmed Bougacha	[MemCpyOpt] Don't force i64 when promoting memset/memcpy...	commit \| commitdiff \| tree
2015-04-18	Ahmed Bougacha	[MemCpyOpt] Promote both memset/memcpy sizes if differently...	commit \| commitdiff \| tree
2015-04-18	Ahmed Bougacha	[GlobalMerge] Look at uses to create smaller global...	commit \| commitdiff \| tree
2015-04-17	Ahmed Bougacha	[AArch64] Don't force MVT::Untyped when selecting LD1LANEpost.	commit \| commitdiff \| tree
2015-04-17	Ahmed Bougacha	Fix another typo in r235224 testcase. NFC.	commit \| commitdiff \| tree
2015-04-17	Ahmed Bougacha	[MemCpyOpt] Optimize double-storing by memset+memcpy.	commit \| commitdiff \| tree
2015-04-17	Ahmed Bougacha	Fix typo in r235224 testcase. NFC.	commit \| commitdiff \| tree
2015-04-17	Ahmed Bougacha	[AArch64] Avoid vector->load dependency cycles when...	commit \| commitdiff \| tree
2015-04-16	Ahmed Bougacha	[AArch64] Don't assert on f16 in DUP PerfectShuffle...	commit \| commitdiff \| tree
2015-04-16	Ahmed Bougacha	[CodeGen] Re-apply r234809 (concat of scalars), with...	commit \| commitdiff \| tree
2015-04-13	Ahmed Bougacha	[CodeGen] Combine concat_vectors of scalars into build_vector.	commit \| commitdiff \| tree
2015-04-11	Ahmed Bougacha	[CodeGen] Split -enable-global-merge into ARM and AArch64...	commit \| commitdiff \| tree
2015-04-10	Ahmed Bougacha	[AArch64] Promote f16 operations to f32.	commit \| commitdiff \| tree
2015-04-09	Ahmed Bougacha	[CodeGen] Combine concat_vector of trunc'd scalar to...	commit \| commitdiff \| tree
2015-04-07	Ahmed Bougacha	[ARM] Mark a bunch of .td Operands with type _MEMORY.	commit \| commitdiff \| tree
2015-04-01	Ahmed Bougacha	[SimplifyLibCalls] Ignore nobuiltin/unavailable fortified...	commit \| commitdiff \| tree
2015-03-31	Ahmed Bougacha	[X86] Generate MOVNT for all vector types.	commit \| commitdiff \| tree
next

C/C++ LLVM-based compilers that forbids OOTA behaviors

RSS Atom