projects / oota-llvm.git / history
? search:
summary | shortlog | log | commit | commitdiff | tree
first ⋅ prev ⋅ next
note to self: *save* file, then check it in
[oota-llvm.git] / lib / Target / X86 /
2006-03-31 Evan ChengExpand all INSERT_VECTOR_ELT (obviously bad) for now. tree | commitdiff
2006-03-31 Evan ChengTypo tree | commitdiff
2006-03-31 Evan ChengOk for vector_shuffle mask to contain undef elements. tree | commitdiff
2006-03-30 Evan ChengMake sure all possible shuffles are matched. tree | commitdiff
2006-03-30 Evan ChengMore logical ops patterns tree | commitdiff
2006-03-30 Evan ChengAdd support for _mm_cmp{cc}_ss and _mm_cmp{cc}_ps intri... tree | commitdiff
2006-03-30 Evan ChengAdd 128-bit pmovmskb intrinsic support. tree | commitdiff
2006-03-29 Evan ChengChange SSE pack operation definitions to fit what the... tree | commitdiff
2006-03-29 Evan Cheng- Added some SSE2 128-bit packed integer ops. tree | commitdiff
2006-03-29 Evan ChengNeed to special case splat after all. Make the second... tree | commitdiff
2006-03-29 Evan ChengFloating point logical operation patterns should match... tree | commitdiff
2006-03-29 Evan Cheng- More shuffle related bug fixes. tree | commitdiff
2006-03-29 Evan ChengAnother entry about shuffles. tree | commitdiff
2006-03-29 Evan Cheng- Only use pshufd for v4i32 vector shuffles. tree | commitdiff
2006-03-28 Evan ChengAdded aliases to scalar SSE instructions, e.g. addss... tree | commitdiff
2006-03-28 Evan ChengFixing buggy code. tree | commitdiff
2006-03-28 Jim LaskeyExpose base register for DwarfWriter. Refactor code... tree | commitdiff
2006-03-28 Jim LaskeyAdded missing paren on behalf of Ramana Radhakrishnan. tree | commitdiff
2006-03-28 Evan ChengMissed X86::isUNPCKHMask tree | commitdiff
2006-03-28 Evan Chengmovlps and movlpd should be modeled as two address... tree | commitdiff
2006-03-28 Evan ChengUpdate tree | commitdiff
2006-03-28 Evan ChengTypo tree | commitdiff
2006-03-28 Evan Cheng* Prefer using operation of matching types. e.g unpcklp... tree | commitdiff
2006-03-28 Evan ChengAdded a couple of entries about movhps and movlhps. tree | commitdiff
2006-03-28 Evan ChengAll unpack cases are now being handled. tree | commitdiff
2006-03-28 Evan Cheng- Clean up / consoladate various shuffle masks. tree | commitdiff
2006-03-28 Evan ChengModel unpack lower and interleave as vector_shuffle... tree | commitdiff
2006-03-27 Jim LaskeyTranslate llvm target registers to dwarf register numbe... tree | commitdiff
2006-03-27 Chris Lattnerunbreak the build tree | commitdiff
2006-03-27 Evan ChengUse pcmpeq to generate vector of all ones. tree | commitdiff
2006-03-27 Nate BegemanSelectionDAGISel can now natively handle Switch instruc... tree | commitdiff
2006-03-26 Nate BegemanReadme note tree | commitdiff
2006-03-26 Evan ChengRemove X86:isZeroVector, use ISD::isBuildVectorAllZeros... tree | commitdiff
2006-03-25 Evan ChengAdded missing (any_extend (load ...)) patterns. tree | commitdiff
2006-03-25 Evan ChengBuild arbitrary vector with more than 2 distinct scalar... tree | commitdiff
2006-03-25 Chris Lattner#include Intrinsics.h into all dag isels tree | commitdiff
2006-03-25 Evan ChengAdded SSE cachebility ops tree | commitdiff
2006-03-25 Evan ChengInstruction encoding bug tree | commitdiff
2006-03-25 Evan ChengAdded 128-bit packed integer subtraction. tree | commitdiff
2006-03-25 Evan ChengAdded CVTTPS2PI. tree | commitdiff
2006-03-25 Evan ChengAdded CVTSS2SI. tree | commitdiff
2006-03-24 Evan ChengSupport for scalar to vector with zero extension. tree | commitdiff
2006-03-24 Evan ChengAdded LDMXCSR tree | commitdiff
2006-03-24 Chris Lattnerplug the intrinsics into the patterns for movmsk* tree | commitdiff
2006-03-24 Jim LaskeyAdd dwarf register numbering to register data. tree | commitdiff
2006-03-24 Evan ChengHandle BUILD_VECTOR with all zero elements. tree | commitdiff
2006-03-24 Chris LattnerGabor points out that we can't spell. :) tree | commitdiff
2006-03-24 Evan ChengAll v2f64 shuffle cases can be handled. tree | commitdiff
2006-03-24 Evan ChengMore efficient v2f64 shuffle using movlhps, movhlps... tree | commitdiff
2006-03-24 Evan ChengA new entry tree | commitdiff
2006-03-24 Evan ChengHandle more shuffle cases with SHUFP* instructions. tree | commitdiff
2006-03-23 Evan ChengTypo tree | commitdiff
2006-03-23 Jim LaskeyAdd support to locate local variables in frames (early... tree | commitdiff
2006-03-23 Jim LaskeyChange interface to DwarfWriter. tree | commitdiff
2006-03-23 Chris LattnerFix the encodings of these new instructions, hopefully... tree | commitdiff
2006-03-23 Evan ChengFollowing icc's lead: use movdqa to load / store 128... tree | commitdiff
2006-03-23 Chris LattnerEliminate IntrinsicLowering from TargetMachine. tree | commitdiff
2006-03-23 Evan ChengAdd v4i32 <-> v4f32 bitconvert patterns. tree | commitdiff
2006-03-23 Evan ChengAdd 128-bit integer vector load and add (for testing). tree | commitdiff
2006-03-23 Nate BegemanAdd support for 8 bit immediates with 16/32 bit cmp... tree | commitdiff
2006-03-22 Evan ChengAdded a ValueType operand to isShuffleMaskLegal().... tree | commitdiff
2006-03-22 Evan ChengSHUFP* are two address code. tree | commitdiff
2006-03-22 Evan ChengSome clean up. tree | commitdiff
2006-03-22 Evan Cheng- Supposely movlhps is faster / better than unpcklpd. tree | commitdiff
2006-03-22 Evan Cheng- Implement X86ISelLowering::isShuffleMaskLegal().... tree | commitdiff
2006-03-22 Evan Cheng- VECTOR_SHUFFLE of v4i32 / v4f32 with undef second... tree | commitdiff
2006-03-22 Evan ChengFix PSHUF* and SHUF* jit code emission problems tree | commitdiff
2006-03-22 Chris Lattnerfix a warning tree | commitdiff
2006-03-22 Evan ChengSome splat and shuffle support. tree | commitdiff
2006-03-22 Evan ChengAdd a couple more pseudo instructions. tree | commitdiff
2006-03-21 Evan ChengDidn't mean to check this in. No MMX support yet. tree | commitdiff
2006-03-21 Evan Cheng- Use movaps to store 128-bit vector integers. tree | commitdiff
2006-03-21 Chris LattnerThese targets don't support EXTRACT_VECTOR_ELT, though... tree | commitdiff
2006-03-21 Evan ChengCombine 2 entries tree | commitdiff
2006-03-21 Evan ChengAdd a note about x86 register coallescing tree | commitdiff
2006-03-21 Evan Cheng- Remove scalar to vector pseudo ops. They are just... tree | commitdiff
2006-03-21 Evan Chengx86 ISD::SCALAR_TO_VECTOR support. tree | commitdiff
2006-03-21 Evan ChengJunk unused vector register classes. tree | commitdiff
2006-03-20 Chris LattnerAdd a build_vector node tree | commitdiff
2006-03-20 Evan ChengMove a few things around. tree | commitdiff
2006-03-19 Chris Lattneradd a note with a testcase tree | commitdiff
2006-03-19 Evan ChengVector undef's tree | commitdiff
2006-03-19 Evan ChengTurning on LSR by default tree | commitdiff
2006-03-19 Evan ChengRemember which tests are hurt by LSR. tree | commitdiff
2006-03-19 Chris Lattnerrename these nodes tree | commitdiff
2006-03-18 Evan ChengUse the generic vector register classes VR64 / VR128... tree | commitdiff
2006-03-17 Evan ChengDarwin should use _setjmp/_longjmp instead of setjmp... tree | commitdiff
2006-03-17 Evan ChengMove some pattern fragments to the right files. tree | commitdiff
2006-03-17 Chris LattnerDisable x86 fastcc from passing args in registers tree | commitdiff
2006-03-17 Chris LattnerParameterize the number of integer arguments to pass... tree | commitdiff
2006-03-17 Evan ChengAlso fold MOV8r0, MOV16r0, MOV32r0 + store to MOV8mi... tree | commitdiff
2006-03-17 Evan ChengAdd some missing entries to X86RegisterInfo::foldMemory... tree | commitdiff
2006-03-17 Evan Cheng- Nuke 16-bit SBB instructions. We'll never use them. tree | commitdiff
2006-03-17 Nate BegemanRemove BRTWOWAY* tree | commitdiff
2006-03-16 Evan ChengA new entry. tree | commitdiff
2006-03-16 Evan ChengBug fix: condition inverted. tree | commitdiff
2006-03-16 Evan ChengAdded a way for TargetLowering to specify what values... tree | commitdiff
2006-03-13 Evan ChengAdded getTargetLowering() to TargetMachine. Refactored... tree | commitdiff
2006-03-13 Evan ChengAdd LSR hooks. tree | commitdiff
2006-03-09 Evan ChengAdd option -enable-x86-lsr to enable x86 loop strength... tree | commitdiff
next
C/C++ LLVM-based compilers that forbids OOTA behaviors