2012-01-12 |
Nadav Rotem | Fix a bug in the AVX 256-bit shuffle code in cases... |
blob | commitdiff | raw |
2012-01-11 |
Rafael Espindola | Explicitly set the scale to 1 on some segstack prologue... |
blob | commitdiff | raw | diff to current |
2012-01-11 |
Nadav Rotem | Fix a bug in the lowering of BUILD_VECTOR for AVX.... |
blob | commitdiff | raw | diff to current |
2012-01-10 |
Lang Hames | Fixed order of operands in comment to match code. |
blob | commitdiff | raw | diff to current |
2012-01-10 |
Bill Wendling | For i386, don't use the generic code. |
blob | commitdiff | raw | diff to current |
2012-01-10 |
Craig Topper | Fix a crash in AVX2 when trying to broadcast a double... |
blob | commitdiff | raw | diff to current |
2012-01-10 |
Craig Topper | Remove hasXMM/hasXMMInt functions. Move callers to... |
blob | commitdiff | raw | diff to current |
2012-01-10 |
Craig Topper | Remove hasSSE*orAVX functions and change all callers... |
blob | commitdiff | raw | diff to current |
2012-01-09 |
Craig Topper | Change some places that were checking for AVX OR SSE1... |
blob | commitdiff | raw | diff to current |
2012-01-08 |
Victor Umansky | Reverted commit #147601 upon Evan's request. |
blob | commitdiff | raw | diff to current |
2012-01-07 |
Benjamin Kramer | Remove VectorExtras. This unused helper was written... |
blob | commitdiff | raw | diff to current |
2012-01-07 |
Craig Topper | Remove unnecessary check of hasAVX(). It's already... |
blob | commitdiff | raw | diff to current |
2012-01-07 |
Eric Christopher | Make the 'x' constraint work for AVX registers as well. |
blob | commitdiff | raw | diff to current |
2012-01-05 |
Victor Umansky | Peephole optimization of ptest-conditioned branch in... |
blob | commitdiff | raw | diff to current |
2012-01-05 |
Bill Wendling | Replace the uint64_t -> double convertion algorithm... |
blob | commitdiff | raw | diff to current |
2012-01-04 |
Evan Cheng | For x86, canonicalize max |
blob | commitdiff | raw | diff to current |
2012-01-03 |
Chad Rosier | Fix 80-column violations. |
blob | commitdiff | raw | diff to current |
2012-01-03 |
Nadav Rotem | Revert 147426 because it caused pr11696. |
blob | commitdiff | raw | diff to current |
2012-01-03 |
Chad Rosier | Enhance DAGCombine for transforming 128->256 casts... |
blob | commitdiff | raw | diff to current |
2012-01-02 |
Craig Topper | Miscellaneous shuffle lowering cleanup. No functional... |
blob | commitdiff | raw | diff to current |
2012-01-02 |
Craig Topper | Make CanXFormVExtractWithShuffleIntoLoad reject loads... |
blob | commitdiff | raw | diff to current |
2012-01-02 |
Nadav Rotem | Optimize the sequence blend(sign_extend(x)) to blend... |
blob | commitdiff | raw | diff to current |
2011-12-31 |
Craig Topper | Merge X86 SHUFPS and SHUFPD node types. |
blob | commitdiff | raw | diff to current |
2011-12-29 |
Craig Topper | Make LowerBUILD_VECTOR keep node vector types consisten... |
blob | commitdiff | raw | diff to current |
2011-12-29 |
Craig Topper | Remove some elses after returns. |
blob | commitdiff | raw | diff to current |
2011-12-29 |
Craig Topper | Remove trailing spaces. Fix an assert to use && instead... |
blob | commitdiff | raw | diff to current |
2011-12-28 |
Eli Friedman | Fix type-checking for load transformation which is... |
blob | commitdiff | raw | diff to current |
2011-12-28 |
Elena Demikhovsky | Fixed a bug in LowerVECTOR_SHUFFLE and LowerBUILD_VECTOR. |
blob | commitdiff | raw | diff to current |
2011-12-27 |
Craig Topper | Add handling of x86_avx2_pmovmskb to computeMaskedBitsF... |
blob | commitdiff | raw | diff to current |
2011-12-24 |
Chandler Carruth | Use standard promotion for i8 CTTZ nodes and i8 CTLZ... |
blob | commitdiff | raw | diff to current |
2011-12-24 |
Chandler Carruth | Add systematic testing for cttz as well, and fix the... |
blob | commitdiff | raw | diff to current |
2011-12-24 |
Chandler Carruth | Expand more when we have a nice 'tzcnt' instruction... |
blob | commitdiff | raw | diff to current |
2011-12-24 |
Chandler Carruth | Switch the lowering of CTLZ_ZERO_UNDEF from a .td patte... |
blob | commitdiff | raw | diff to current |
2011-12-22 |
Chad Rosier | Fix 80-column violations. |
blob | commitdiff | raw | diff to current |
2011-12-21 |
Chad Rosier | No case stmt for BUILD_VECTOR in PerformDAGCombine... |
blob | commitdiff | raw | diff to current |
2011-12-20 |
Chandler Carruth | Begin teaching the X86 target how to efficiently codege... |
blob | commitdiff | raw | diff to current |
2011-12-18 |
Benjamin Kramer | Another variadics tweak. |
blob | commitdiff | raw | diff to current |
2011-12-18 |
Benjamin Kramer | Use the fancy new VariadicFunction template instead... |
blob | commitdiff | raw | diff to current |
2011-12-17 |
Craig Topper | Remove an unused X86ISD node type. |
blob | commitdiff | raw | diff to current |
2011-12-17 |
Benjamin Kramer | X86: Factor the bswap asm matching to be slightly less... |
blob | commitdiff | raw | diff to current |
2011-12-17 |
Lang Hames | Make sure that the lower bits on the VSELECT condition... |
blob | commitdiff | raw | diff to current |
2011-12-16 |
Craig Topper | Don't try to match 'unpackl/h v, v' for 32xi8 and 16xi1... |
blob | commitdiff | raw | diff to current |
2011-12-15 |
Chad Rosier | Fix assert in LowerBUILD_VECTOR for v16i16 type on... |
blob | commitdiff | raw | diff to current |
2011-12-15 |
Lang Hames | Fix VSELECT operand order. Was previously backwards... |
blob | commitdiff | raw | diff to current |
2011-12-15 |
Chad Rosier | Use SmallVector/assign(), rather than std::vector/push_... |
blob | commitdiff | raw | diff to current |
2011-12-15 |
Chad Rosier | Add support for lowering fneg when AVX is enabled. |
blob | commitdiff | raw | diff to current |
2011-12-13 |
Chandler Carruth | Initial CodeGen support for CTTZ/CTLZ where a zero... |
blob | commitdiff | raw | diff to current |
2011-12-11 |
Craig Topper | Remove some remants of the old palign pattern fragment... |
blob | commitdiff | raw | diff to current |
2011-12-09 |
Benjamin Kramer | X86: Add patterns for the various rounding ops for... |
blob | commitdiff | raw | diff to current |
2011-12-08 |
Owen Anderson | Don't explicitly marked libm rounding ops as legal... |
blob | commitdiff | raw | diff to current |
2011-12-08 |
Owen Anderson | Teach SelectionDAG to match more calls to libm function... |
blob | commitdiff | raw | diff to current |
2011-12-06 |
Craig Topper | Add X86ISD::HADD/HSUB to getTargetNodeName |
blob | commitdiff | raw | diff to current |
2011-12-06 |
Craig Topper | Merge floating point and integer UNPCK X86ISD node... |
blob | commitdiff | raw | diff to current |
2011-12-06 |
Craig Topper | Clean up some of the shuffle decoding code for UNPCK... |
blob | commitdiff | raw | diff to current |
2011-12-06 |
Craig Topper | Merge isSHUFPMask and isCommutedSHUFPMask into single... |
blob | commitdiff | raw | diff to current |
2011-12-06 |
Jakob Stoklund Olesen | Use logarithmic units for basic block alignment. |
blob | commitdiff | raw | diff to current |
2011-12-05 |
Craig Topper | Remove some leftover remnants that once tried to create... |
blob | commitdiff | raw | diff to current |
2011-12-05 |
Craig Topper | Clean up and optimizations to the X86 shuffle lowering... |
blob | commitdiff | raw | diff to current |
2011-12-02 |
Nick Lewycky | Move global variables in TargetMachine into new TargetO... |
blob | commitdiff | raw | diff to current |
2011-12-02 |
Craig Topper | Reduce duplicate code in isHorizontalBinOp and add... |
blob | commitdiff | raw | diff to current |
2011-12-02 |
Craig Topper | Add instruction selection support for horizontal add... |
blob | commitdiff | raw | diff to current |
2011-11-30 |
Nadav Rotem | X86: PerformOrCombine introduced a vselect node with... |
blob | commitdiff | raw | diff to current |
2011-11-30 |
Craig Topper | Add instruction selection support for AVX2 horizontal... |
blob | commitdiff | raw | diff to current |
2011-11-30 |
Craig Topper | Merge VPERM2F128/VPERM2I128 ISD node types. |
blob | commitdiff | raw | diff to current |
2011-11-30 |
Craig Topper | Merge decoding of VPERMILPD and VPERMILPS shuffle masks... |
blob | commitdiff | raw | diff to current |
2011-11-29 |
Craig Topper | Fix issues in shuffle decoding around VPERM* instructio... |
blob | commitdiff | raw | diff to current |
2011-11-28 |
Craig Topper | Add X86 instruction selection for VPERM2I128 when AVX2... |
blob | commitdiff | raw | diff to current |
2011-11-28 |
Craig Topper | Make isCommutedVSHUFP more like the way isCommutedSHUFP... |
blob | commitdiff | raw | diff to current |
2011-11-27 |
Craig Topper | Merge detecting and handling for VSHUFPSY and VSHUFPDY... |
blob | commitdiff | raw | diff to current |
2011-11-26 |
Craig Topper | Merge 128-bit and 256-bit X86ISD node types for VPERMIL... |
blob | commitdiff | raw | diff to current |
2011-11-26 |
Craig Topper | Collapse X86ISD node types for PUNPCKH*, PUNPCKL*,... |
blob | commitdiff | raw | diff to current |
2011-11-24 |
Craig Topper | Remove 256-bit specific node types for UNPCKHPS/D and... |
blob | commitdiff | raw | diff to current |
2011-11-24 |
Craig Topper | Remove AVX2 specific X86ISD node types for PUNPCKH... |
blob | commitdiff | raw | diff to current |
2011-11-23 |
Benjamin Kramer | X86: Use btq for bit tests if the immediate can't be... |
blob | commitdiff | raw | diff to current |
2011-11-23 |
Elena Demikhovsky | I added several lines in X86 code generator that allow... |
blob | commitdiff | raw | diff to current |
2011-11-22 |
Craig Topper | Fix shuffle decoding logic to handle UNPCKLPS/UNPCKLPD... |
blob | commitdiff | raw | diff to current |
2011-11-22 |
Craig Topper | Add methods for querying minimum SSE version along... |
blob | commitdiff | raw | diff to current |
2011-11-21 |
Craig Topper | Lowering for v32i8 to VPUNPCKLBW/VPUNPCKHBW when AVX2... |
blob | commitdiff | raw | diff to current |
2011-11-21 |
Craig Topper | Add support for lowering 256-bit shuffles to VPUNPCKL... |
blob | commitdiff | raw | diff to current |
2011-11-21 |
Craig Topper | Make LowerSIGN_EXTEND_INREG split 256-bit vectors when... |
blob | commitdiff | raw | diff to current |
2011-11-20 |
Craig Topper | Add code for lowering v32i8 shifts by a splat to AVX2... |
blob | commitdiff | raw | diff to current |
2011-11-19 |
Craig Topper | Use 256-bit vcmpeqd for creating an all ones vector... |
blob | commitdiff | raw | diff to current |
2011-11-19 |
Craig Topper | Custom lower AVX2 variable shift intrinsics to shl... |
blob | commitdiff | raw | diff to current |
2011-11-19 |
Craig Topper | Synthesize SSSE3/AVX 128-bit horizontal integer add... |
blob | commitdiff | raw | diff to current |
2011-11-19 |
Craig Topper | Collapse X86 PSIGNB/PSIGNW/PSIGND node types. |
blob | commitdiff | raw | diff to current |
2011-11-19 |
Craig Topper | Extend VPBLENDVB and VPSIGN lowering to work for AVX2. |
blob | commitdiff | raw | diff to current |
2011-11-18 |
Nadav Rotem | Add AVX2 vpbroadcast support |
blob | commitdiff | raw | diff to current |
2011-11-15 |
Nadav Rotem | AVX: Add support for vbroadcast from BUILD_VECTOR and... |
blob | commitdiff | raw | diff to current |
2011-11-15 |
Pete Cooper | Added custom lowering for load->dec->store sequence... |
blob | commitdiff | raw | diff to current |
2011-11-15 |
Jay Foad | Remove some unnecessary includes of PseudoSourceValue.h. |
blob | commitdiff | raw | diff to current |
2011-11-14 |
Pete Cooper | Changed SSE4/AVX <2 x i64> extract and insert ops to... |
blob | commitdiff | raw | diff to current |
2011-11-14 |
Craig Topper | Add neverHasSideEffects, mayLoad, and mayStore to many... |
blob | commitdiff | raw | diff to current |
2011-11-13 |
Craig Topper | Add BLSI, BLSMSK, and BLSR to getTargetNodeName. |
blob | commitdiff | raw | diff to current |
2011-11-12 |
Craig Topper | Add more AVX2 shift lowering support. Move AVX2 variabl... |
blob | commitdiff | raw | diff to current |
2011-11-11 |
Craig Topper | Add lowering for AVX2 shift instructions. |
blob | commitdiff | raw | diff to current |
2011-11-09 |
Nadav Rotem | AVX2: Add patterns for variable shift operations |
blob | commitdiff | raw | diff to current |
2011-11-09 |
Nadav Rotem | Add AVX2 support for vselect of v32i8 |
blob | commitdiff | raw | diff to current |
2011-11-09 |
Craig Topper | Add instruction selection for AVX2 integer comparisons. |
blob | commitdiff | raw | diff to current |
2011-11-09 |
Craig Topper | Add AVX2 instruction lowering for add, sub, and mul. |
blob | commitdiff | raw | diff to current |
2011-11-08 |
Pete Cooper | Added invariant field to the DAG.getLoad method and... |
blob | commitdiff | raw | diff to current |
next |