Add X86 MMX type to bitcode and Type.
[oota-llvm.git] / lib / Target /
2010-09-04 Bruno Cardoso LopesEmit target specific nodes to handle palignr. Do not...
2010-09-04 Bruno Cardoso LopesEmit target specific nodes to handle splats starting...
2010-09-04 Bruno Cardoso LopesEmit target specific nodes for isPSHUFHWMask and isPSHU...
2010-09-04 Bruno Cardoso LopesEmit target specific nodes for isSHUFPMask
2010-09-04 Bruno Cardoso LopesPrevious isMOVLMask matching already emits targets...
2010-09-04 Bruno Cardoso LopesOne more check from the original isShuffleMaskLegal...
2010-09-04 Bruno Cardoso LopesRemove a duplicated but useless check that i've inserte...
2010-09-04 Bruno Cardoso LopesRefactor some code and remove the extra checks for...
2010-09-03 Bruno Cardoso LopesRemove check for unpckh mask
2010-09-03 Bruno Cardoso LopesRemove check for unpckl mask
2010-09-03 Bruno Cardoso LopesInline isShuffleMaskLegal into LowerVECTOR_SHUFFLE...
2010-09-03 Bruno Cardoso LopesReapply considered harmfull part of rr112934 and r112942.
2010-09-03 Dale JohannesenRemove the rest of the nonexistent 64-bit AVX instructions.
2010-09-03 Bruno Cardoso LopesReapply last harmless part of r112934, the pattern...
2010-09-03 Bruno Cardoso LopesReintroduce a simple function refactoring done in r1129...
2010-09-03 Bruno Cardoso LopesReapply piecies of r112942 and r112934 which don't do
2010-09-03 Bruno Cardoso LopesReapply Fix comment
2010-09-03 Daniel DunbarRevert r112934, "- Use specific nodes to match unpckl...
2010-09-03 Daniel DunbarRevert r112938 "Fix comment", which depends on r112934...
2010-09-03 Daniel DunbarRevert r112942, "Use punpckh and unpckh family of nodes...
2010-09-03 Jim GrosbachRe-apply r112883:
2010-09-03 Jim GrosbachCheck the local frame alignment for determining whether...
2010-09-03 Bob WilsonFinish converting the rest of the NEON VLD instructions...
2010-09-03 Daniel DunbarRevert "For ARM stack frames that utilize variable...
2010-09-03 Benjamin KramerZap dead code.
2010-09-03 Bruno Cardoso LopesAVX doesn't support mm operations neither its instrinsics.
2010-09-03 Bruno Cardoso LopesUse punpckh and unpckh family of nodes instead of using...
2010-09-03 Bob WilsonReplace NEON vabdl, vaba, and vabal intrinsics with...
2010-09-03 Bruno Cardoso LopesFix comment
2010-09-03 Bruno Cardoso Lopes- Use specific nodes to match unpckl masks.
2010-09-03 Eric ChristopherSimple branch instruction support.
2010-09-03 Jakob Stoklund OlesenDon't call Predicate_* from X86 target.
2010-09-03 Jakob Stoklund OlesenRemove Predicate_* calls from MBlaze and XCore
2010-09-03 Jakob Stoklund OlesenRemove Predicate_* calls from Mips
2010-09-02 Eric ChristopherAdd basic support for materializing constants (includin...
2010-09-02 Anton KorobeynikovProperly emit __chkstk call instead of __alloca on...
2010-09-02 Bruno Cardoso LopesMove insertps mask decoding to header file
2010-09-02 Anton KorobeynikovRevert win64 changes. They seem to be incomplete
2010-09-02 Jim GrosbachFor ARM stack frames that utilize variable sized object...
2010-09-02 Anton KorobeynikovProperly allocate win64 shadow reg area.
2010-09-02 Bruno Cardoso LopesMove decoding of insertps back to avoid unused warnings...
2010-09-02 Dan GohmanDon't narrow the load and store in a load+twiddle+store...
2010-09-02 Jim Grosbachtrailing whitespace
2010-09-02 Jim Grosbachremove trailing whitespace
2010-09-02 Bruno Cardoso LopesMove x86 specific shuffle mask decoding to its own...
2010-09-02 Jim Grosbachhandle case where a register class is specified
2010-09-02 Jim GrosbachNow that register allocation properly considers reserve...
2010-09-02 Jim GrosbachMask out reserved registers when constructing the set...
2010-09-02 Bob WilsonFill in a missing comment.
2010-09-02 Bob WilsonConvert VLD1 and VLD2 instructions to use pseudo-instru...
2010-09-02 Bruno Cardoso LopesReplace unpckl_undef and unpckh_undef matching with...
2010-09-02 Bruno Cardoso LopesMove condition out to prepare for more matching
2010-09-02 Bruno Cardoso LopesRemove checking for isUNPCKL_v_undef_Mask, the specific...
2010-09-02 Bruno Cardoso Lopesbecome more strict about when it's safe to use X86ISD...
2010-09-02 Eric ChristopherClang's -ccc-host-triple was ignoring the arch specifie...
2010-09-02 Eric ChristopherThis should be TargetMaterializeConstant instead.
2010-09-02 Eric ChristopherOne definition of isThumb is plenty, thanks.
2010-09-02 Jim GrosbachRemove trailing whitespace
2010-09-02 Eric ChristopherRework arm fast-isel load and store handling. Move...
2010-09-02 Jim Grosbachtrivial cleanup
2010-09-01 Jim GrosbachSimplify the tGPR register class now that the register...
2010-09-01 Bob WilsonRemove NEON vmull, vmlal, and vmlsl intrinsics, replaci...
2010-09-01 Bruno Cardoso LopesRevert r112689, avoid those kind of checks cause they...
2010-09-01 Bruno Cardoso LopesUsing target specific nodes for shuffle nodes makes...
2010-09-01 Eric ChristopherSome basic store support.
2010-09-01 Eric ChristopherAdd some more load types in.
2010-09-01 Chris Lattnerzap dead code.
2010-09-01 Chris Lattnertemporarily revert r112664, it is causing a decoding...
2010-09-01 Bruno Cardoso LopesUse movlps, movlpd, movss and movsd specific nodes...
2010-09-01 Bruno Cardoso Lopesminor change, simplify some logic
2010-09-01 Bruno Cardoso LopesMove some functions around so they can be used for...
2010-08-31 Bill WendlingWe have a chance for an optimization. Consider this...
2010-08-31 Bruno Cardoso LopesUse x86 specific MOVSLDUP node, add more patterns to...
2010-08-31 Bruno Cardoso LopesUse x86 specific MOVSHDUP node and add more patterns...
2010-08-31 Bill WendlingAnd ANDS pattern to match the t2ANDS pattern.
2010-08-31 Jakob Stoklund OlesenMake %EFLAGS unallocatable.
2010-08-31 Bruno Cardoso LopesUse MOVHLPS node instead of matching using movhlps...
2010-08-31 Bruno Cardoso LopesUse MOVLHPS and MOVHLPS x86 nodes whenever possible...
2010-08-31 Jim GrosbachSP relative offsets need to be adjusted by the local...
2010-08-31 Jim Grosbachthis assert should just be a condition, since this...
2010-08-31 Bill Wendling- Cleanup some whitespaces.
2010-08-31 Bruno Cardoso LopesUse X86ISD::MOVSS and MOVSD to represent the movl mask...
2010-08-31 Eric ChristopherRewrite slightly so we can expand for floating point...
2010-08-30 Eric ChristopherIf we have an unhandled type then assert, we shouldn...
2010-08-30 Anton KorobeynikovExpand MOVi32imm in ARM mode after regalloc. This provides
2010-08-30 Bill WendlingUse the existing T2I_bin_s_irs pattern instead of creat...
2010-08-30 Jakob Stoklund OlesenRemember to clear the shadow kill flag at the same...
2010-08-30 Bob WilsonRemove NEON vmovn intrinsic, replacing it with vector...
2010-08-30 Jim GrosbachMake ARM add rN, sp, #imm instructions rematerializable...
2010-08-30 Bob WilsonWhen expanding NEON VST pseudo instructions, if the...
2010-08-30 Bill WendlingCreate Thumb2sI_cpsr and T2sI_cpsr. These new classes...
2010-08-29 Kalle RaiskilaFix lowering of INSERT_VECTOR_ELT in SPU.
2010-08-29 Bill WendlingFix whitespaces. No functionality changes.
2010-08-29 Bob WilsonRemove NEON vaddl, vaddw, vsubl, and vsubw intrinsics...
2010-08-29 Eli FriedmanA couple of small missed optimizations.
2010-08-29 Bill Wendling- Add a parameter to T2I_bin_irs for those patterns...
2010-08-29 Chris Lattneradd a bunch more common shuffles to the instprinter.
2010-08-29 Bill WendlingName ANDflag to ANDS, which is less stupid.
2010-08-29 Bill WendlingFile missing from last commit.
2010-08-29 Bill WendlingCreate an ARMISD::AND node. This node is exactly like...
next