Commits · 4d22f15a2fc195dd7589bb3fd887f95a2720812f · Roger Ferrer / llvm-epi · GitLab

Jul 13, 2018
- [cfi-verify] Support AArch64. · 06e7e579
  Joel Galenson authored Jul 13, 2018
  06e7e579
- Add parens to silence Wparentheses warning, introduced by 336990 · ac8cb22a
  Erich Keane authored Jul 13, 2018
  ac8cb22a
- [SLH] Introduce a new pass to do Speculative Load Hardening to mitigate · 90358e1e
  Chandler Carruth authored Jul 13, 2018
  90358e1e
- [x86] Teach the EFLAGS copy lowering to handle much more complex control · caa7b03a
  Chandler Carruth authored Jul 13, 2018
  caa7b03a
- [X86] Prefer MOVSS/SD over BLEND under optsize in isel. · a8efe59a
  Craig Topper authored Jul 13, 2018
  a8efe59a
- [X86] Remove isel patterns that turns packed add/sub/mul/div+movss/sd into... · 2ab325ba
  Craig Topper authored Jul 13, 2018
  2ab325ba
- [X86] Add AVX512 equivalents of some isel patterns so we get EVEX instructions. · 3a134772
  Craig Topper authored Jul 12, 2018
  3a134772
Jul 12, 2018
- Revert r336950 and r336951 "[X86] Add AVX512 equivalents of some isel patterns... · b0053b79
  Craig Topper authored Jul 12, 2018
  b0053b79
- [X86] Add AVX512 equivalents of some isel patterns so we get EVEX instructions. · 3b837b6b
  Craig Topper authored Jul 12, 2018
  3b837b6b
- foo · b01a3553
  Craig Topper authored Jul 12, 2018
  b01a3553
- [X86][FastISel] Support EVEX version of sqrt. · 57c4585b
  Craig Topper authored Jul 12, 2018
  57c4585b
- [X86] Connect the flags user from PCMPISTR instructions to the correct node from the instruction. · abc307e6
  Craig Topper authored Jul 12, 2018
  abc307e6
- [X86][FastISel] Choose EVEX instructions when possible when lowering... · d43f5823
  Craig Topper authored Jul 12, 2018
  d43f5823
- [X86][SSE] Utilize ZeroableElements for canWidenShuffleElements · 44b89fa9
  Simon Pilgrim authored Jul 12, 2018
  44b89fa9
- [X86][AVX] Use Zeroable mask to improve shuffle mask widening · 8a463897
  Simon Pilgrim authored Jul 12, 2018
  8a463897
- [X86] Remove i128 type from FR128 regclass. · 73c7dcef
  Craig Topper authored Jul 12, 2018
  73c7dcef
- [X86] Remove patterns and ISD nodes for the old scalar FMA intrinsic lowering. · 73347ec0
  Craig Topper authored Jul 12, 2018
  73347ec0
- [x86] Fix another trivial bug in x86 flags copy lowering that has been · b4faf4ce
  Chandler Carruth authored Jul 12, 2018
  b4faf4ce
- [X86] Add patterns to use VMOVSS/SD zero masking for scalar f32/f64 select with zero. · be996bd2
  Craig Topper authored Jul 12, 2018
  be996bd2
- [x86] Fix EFLAGS copy lowering to correctly handle walking past uses in · 1c8234f6
  Chandler Carruth authored Jul 12, 2018
  1c8234f6
- [X86] Remove and autoupgrade the scalar fma intrinsics with masking. · 034adf26
  Craig Topper authored Jul 12, 2018
  034adf26
Jul 11, 2018
- [X86] Remove patterns for inserting a load into a zero vector. · 38b290f7
  Craig Topper authored Jul 11, 2018
  38b290f7
- [X86] Fix MayLoad/HasSideEffect flag for (V)MOVLPSrm instructions. · 483db141
  Andrea Di Biagio authored Jul 11, 2018
  483db141
- [X86] The TEST instruction is eliminated when BSF/TZCNT is used · 02867f0f
  Craig Topper authored Jul 11, 2018
  02867f0f
- [X86] Remove some composite MOVSS/MOVSD isel patterns. · 1d6a80cd
  Craig Topper authored Jul 11, 2018
  1d6a80cd
- [X86] Remove AddedComplexity from all patterns that use X86vzmovl as their root. · 27c77fe4
  Craig Topper authored Jul 10, 2018
  27c77fe4
- [X86] Teach X86InstrInfo::commuteInstructionImpl to use MOVSD/MOVSS for BLEND... · 860ab496
  Craig Topper authored Jul 10, 2018
  860ab496
Jul 10, 2018
- [X86] Remove X86ISD::MOVLPS and X86ISD::MOVLPD. NFCI · dea0b88b
  Craig Topper authored Jul 10, 2018
  dea0b88b
- [X86] Remove dead SDNode object from X86InstrFragmentsSIMD.td. NFC · fb302d01
  Craig Topper authored Jul 10, 2018
  fb302d01
- [X86] Remove AddedComplexity from register form of NOT. NFCI · 04ded1ac
  Craig Topper authored Jul 10, 2018
  04ded1ac
- [X86] Remove AddedComplexity from MMX_X86movw2d patterns. · 0f6275ab
  Craig Topper authored Jul 10, 2018
  0f6275ab
- [GlobalISel][X86_64] Support for G_SITOFP · 48ca0550
  Alexander Ivchenko authored Jul 10, 2018
  48ca0550
- [X86][SSE] Prefer BLEND(SHL(v,c1),SHL(v,c2)) over MUL(v, c3) · d32ca2c0
  Simon Pilgrim authored Jul 10, 2018
  d32ca2c0
- [X86] Use IsProfitableToFold to block vinsertf128rm in favor of insert_subreg... · 08b81a55
  Craig Topper authored Jul 10, 2018
  08b81a55
- [X86] Remove some seemingly unnecessary patterns. · db73f564
  Craig Topper authored Jul 10, 2018
  db73f564
- [X86] Correct vfixupimm load patterns to look for an integer load, not a... · 866a377e
  Craig Topper authored Jul 10, 2018
  866a377e
- [X86] Remove FloatVT from X86VectorVTInfo in X86InstrAVX512.td · e4f46e4f
  Craig Topper authored Jul 10, 2018
  e4f46e4f
Jul 09, 2018
- [X86][TLI] DAGCombine: Unfold variable bit-clearing mask to two shifts. · 5ccae175
  Roman Lebedev authored Jul 09, 2018
  5ccae175
- [X86] In combineFMA, make sure we bitcast the result of isFNEG back the... · 47170b31
  Craig Topper authored Jul 09, 2018
  47170b31
- [X86] Remove some patterns that include a bitcast of a floating point load to an integer type. · e9cff7d4
  Craig Topper authored Jul 09, 2018
  e9cff7d4