move target-independent opcodes out of TargetInstrInfo
[oota-llvm.git] / lib / Target / ARM / ARMISelDAGToDAG.cpp
2010-02-09 Chris Lattnermove target-independent opcodes out of TargetInstrInfo
2010-01-19 Evan ChengFix r93758. Use isel patterns instead of c++ selection...
2010-01-18 Jim GrosbachPatch by David Conrad:
2010-01-17 Bob WilsonFix an off-by-one error that caused the chain operand...
2010-01-05 Dan GohmanChange SelectCode's argument from SDValue to SDNode...
2009-11-24 Anton KorobeynikovMaterialize global addresses via movt/movw pair, this...
2009-11-21 Evan ChengAdd predicate operand to NEON instructions. Fix lots...
2009-11-20 Evan ChengFix codegen of conditional move of immediates. We were...
2009-11-19 Evan ChengRefactor cmov selection code out to a separate function...
2009-11-19 Evan Cheng80 col violation.
2009-11-09 Jim GrosbachUse Unified Assembly Syntax for the ARM backend.
2009-11-07 Jim GrosbachSupport alignment specifier for NEON vld/vst instructions
2009-11-05 Dan GohmanRemove uninteresting and confusing debug output.
2009-11-02 Bob WilsonPrune unnecessary include.
2009-10-27 Johnny ChenTest commit. Added '.' to the comment line.
2009-10-22 Evan ChengDon't generate sbfx / ubfx with negative lsb field...
2009-10-21 Evan ChengMatch more patterns to movt.
2009-10-14 Bob WilsonRemove unused variables to fix build warning.
2009-10-14 Bob WilsonRefactor code to select NEON VST intrinsics.
2009-10-14 Bob WilsonRefactor code to select NEON VLD intrinsics.
2009-10-14 Bob WilsonMore refactoring. NEON vst lane intrinsics can share...
2009-10-14 Bob WilsonRefactor code for selecting NEON load lane intrinsics.
2009-10-13 Bob WilsonMore Neon clean-up: avoid the need for custom-lowering...
2009-10-13 Bob WilsonRevise ARM inline assembly memory operands to require...
2009-10-13 Sandeep PatelFix method name in comment, per Bob Wilson.
2009-10-13 Sandeep PatelAdd ARMv6T2 SBFX/UBFX instructions. Approved by Anton...
2009-10-09 Bob WilsonAdd codegen support for NEON vst4lane intrinsics with...
2009-10-08 Bob WilsonAdd codegen support for NEON vst3lane intrinsics with...
2009-10-08 Bob WilsonAdd codegen support for NEON vst2lane intrinsics with...
2009-10-08 Bob WilsonAdd codegen support for NEON vld4lane intrinsics with...
2009-10-08 Bob WilsonAdd codegen support for NEON vld3lane intrinsics with...
2009-10-08 Bob WilsonAdd codegen support for NEON vld2lane intrinsics with...
2009-10-08 Bob WilsonClean up some unnecessary initializations.
2009-10-08 Bob WilsonClean up a comment (indentation was wrong).
2009-10-08 Bob WilsonAdd codegen support for NEON vst4 intrinsics with ...
2009-10-08 Bob WilsonAdd codegen support for NEON vst3 intrinsics with ...
2009-10-08 Bob WilsonAdd codegen support for NEON vst2 intrinsics with ...
2009-10-07 Bob WilsonAdd codegen support for NEON vld4 intrinsics with ...
2009-10-07 Bob WilsonAdd codegen support for NEON vld3 intrinsics with ...
2009-10-07 Bob WilsonAdd codegen support for NEON vld2 intrinsics with ...
2009-10-07 Bob WilsonAdd codegen support for NEON vst4 intrinsics with 128...
2009-10-07 Bob WilsonAdd codegen support for NEON vst3 intrinsics with 128...
2009-10-07 Bob WilsonAdd codegen support for NEON vst2 intrinsics with 128...
2009-10-07 Bob WilsonAdd codegen support for NEON vld4 intrinsics with 128...
2009-10-07 Bob WilsonAdd codegen support for NEON vld3 intrinsics with 128...
2009-10-07 Bob WilsonRearrange code for selecting vld2 intrinsics. No funct...
2009-10-06 Bob WilsonAdd codegen support for NEON vld2 operations on quad...
2009-09-28 Bob WilsonPass the optimization level when constructing the ARM...
2009-09-27 Anton KorobeynikovUse movt/movw pair to materialize 32 bit constants...
2009-09-25 Dan GohmanRename getTargetNode to getMachineNode, for consistency...
2009-09-01 Bob WilsonAdd support for generating code for vst{234}lane intrin...
2009-09-01 Bob WilsonGenerate code for vld{234}_lane intrinsics.
2009-08-26 Bob WilsonRemove unneeded ARM-specific DAG nodes for VLD* and...
2009-08-22 Devang PatelRecord variable debug info at ISel time directly.
2009-08-21 Anton KorobeynikovFix some typos and use type-based isel for VZIP/VUZP...
2009-08-21 Anton KorobeynikovAdd nodes & dummy matchers for some v{zip,uzp,trn}...
2009-08-21 Bob WilsonRemove Neon intrinsics for VZIP, VUZP, and VTRN. We...
2009-08-14 Evan ChengIndentation.
2009-08-14 Bob WilsonDuring legalization, change Neon vdup_lane operations...
2009-08-13 Owen AndersonPush LLVMContexts through the IntegerType APIs.
2009-08-12 Evan ChengShrink Thumb2 movcc instructions.
2009-08-12 Bob WilsonAdd missing chain operands for VLD* and VST* instructions.
2009-08-11 Evan ChengShrinkify Thumb2 r = add sp, imm.
2009-08-11 Owen AndersonSplit EVT into MVT and EVT, the former representing...
2009-08-11 Jim GrosbachWhitespace cleanup. Remove trailing whitespace.
2009-08-11 Evan ChengFix Thumb2 load / store addressing mode matching code...
2009-08-11 Bob WilsonUse vAny type to get rid of Neon intrinsics that differ...
2009-08-10 Dan GohmanFix a bug where DAGCombine was producing an illegal...
2009-08-10 Owen AndersonRename MVT to EVT, in preparation for splitting SimpleV...
2009-08-10 Evan ChengHandle the constantfp created during post-legalization...
2009-08-08 Anton KorobeynikovUse VLDM / VSTM to spill/reload 128-bit Neon registers
2009-08-08 Bob WilsonImplement Neon VZIP and VUZP instructions. These are...
2009-08-08 Bob WilsonImplement Neon VTRN instructions. For now, anyway...
2009-08-07 Evan ChengIt turns out most of the thumb2 instructions are not...
2009-08-06 Bob WilsonImplement Neon VST[234] operations.
2009-08-06 Bob WilsonNeon does not actually have VLD{234}.64 instructions.
2009-08-05 Bob WilsonChange DAG nodes for Neon VLD2/3/4 operations to return...
2009-08-03 Bob WilsonLower CONCAT_VECTOR during legalization instead of...
2009-08-01 Evan ChengSplit t2MOVCCs since some assemblers do not recognize...
2009-07-30 David GoodwinRemove redundant match for frame index from imm8 addrmo...
2009-07-30 David GoodwinCleanup and include code selection for some frame index...
2009-07-26 Evan ChengRename tMOVhi2lor to tMOVgpr2tgpr. It's not moving...
2009-07-24 Owen AndersonRevert the ConstantInt constructors back to their 2...
2009-07-24 David GoodwinCorrectly handle the Thumb-2 imm8 addrmode. Specialize...
2009-07-22 Evan ChengUse getTargetConstant instead of getConstant since...
2009-07-22 Evan ChengEliminate a redudant check Eli pointed out.
2009-07-21 Evan ChengFix ARM isle code that optimize multiply by constants...
2009-07-20 David GoodwinUse t2LDRri12 for frame index loads.
2009-07-15 David GoodwinThumb-2 only support [base_reg + offset_reg] addressing...
2009-07-14 Owen AndersonMove EVER MORE stuff over to LLVMContext.
2009-07-14 David GoodwinCheck for PRE_INC and POST_INC.
2009-07-14 David GoodwinhasThumb2() does not mean we are compiling for thumb...
2009-07-11 Evan ChengSmarter isel of ldrsb / ldrsh. Only make use of these...
2009-07-11 Evan ChengMajor changes to Thumb (not Thumb2). Many 16-bit instru...
2009-07-09 Evan ChengFix ldrd / strd address mode matching code. It allows...
2009-07-08 David GoodwinUse common code for both ARM and Thumb-2 instruction...
2009-07-08 Evan ChengChange how so_imm and t2_so_imm are handled. At instruc...
2009-07-08 Torok EdwinImplement changes from Chris's feedback.
2009-07-07 Evan ChengAdd Thumb2 movcc instructions.
2009-07-07 Evan ChengAdd some more Thumb2 multiplication instructions.
next