Commits · bbb3dffd0acac9097917a1e3a3e3f81e8c5d3433 · Roger Ferrer / llvm-epi-0.8 · GitLab

This is an archived project. Repository and other project resources are read-only.

Aug 14, 2016
- [InstCombine] add test for missing vector icmp fold · bbb3dffd
  Sanjay Patel authored Aug 14, 2016
  bbb3dffd
- [InstCombine] add test for missing vector icmp fold · 66a3457a
  Sanjay Patel authored Aug 14, 2016
  66a3457a
- [AVX-512] Mark VPMADDWD as commutable to match SSE/AVX version. · 37e8c544
  Craig Topper authored Aug 14, 2016
  37e8c544
- [AVX-512] Add masked commutable floating point max/min instructions to folding tables. · c677e97d
  Craig Topper authored Aug 14, 2016
  c677e97d
- [AVX-512] Add masked logical operations to memory folding tables. · 29fbdc30
  Craig Topper authored Aug 14, 2016
  29fbdc30
- [AVX512] Fix VFPCLASSSD/VFPCLASSSS intrinsic lowering. The i1 result should be... · 505f2cc4
  Igor Breger authored Aug 14, 2016
  505f2cc4
- elf-core: set powerpc vector byte order correctly · 0767b40e
  Ed Maste authored Aug 14, 2016
  0767b40e
- autogenerate checks · 6fc00b0a
  Igor Breger authored Aug 14, 2016
  6fc00b0a
- [AVX512] Fix insertelement i1 lowering. · 8672408d
  Igor Breger authored Aug 14, 2016
  8672408d
- Revert "gold: add a cast to appease std::max NFC" · 98541b09
  Saleem Abdulrasool authored Aug 14, 2016
  98541b09
- Revert "CodeGen: If Convert blocks that would form a diamond when tail-merged." · 68be1eb8
  Diana Picus authored Aug 14, 2016
  68be1eb8
- Revert "Codegen: Don't tail-duplicate blocks with un-analyzable fallthrough." · 35ccf53e
  Diana Picus authored Aug 14, 2016
  35ccf53e
- [IRCE] Change variable grouping; NFC · 35459f0e
  Sanjoy Das authored Aug 14, 2016
  35459f0e
- [IRCE] Create llvm::Loop instances for cloned out loops · 2143447c
  Sanjoy Das authored Aug 14, 2016
  2143447c
- [IRCE] Don't iterate on loops that were cloned out · 7a18a238
  Sanjoy Das authored Aug 14, 2016
  7a18a238
- [IRCE] Add better DEBUG diagnostic; NFC · 43fdc543
  Sanjoy Das authored Aug 14, 2016
  43fdc543
- Fix bitcode auto-upgrade when using bitcode lazy loading · a71002e7
  Mehdi Amini authored Aug 14, 2016
  a71002e7
- Fix unsupported relocation type R_HEX_6_X' for symbol .rodata · 822ee88a
  Ron Lieberman authored Aug 13, 2016
  822ee88a
- Revert "Fix bitcode auto-upgrade when using bitcode lazy loading" · 466a64e2
  Mehdi Amini authored Aug 13, 2016
  466a64e2
- [IRCE] Fix test case; NFC · 1b1272f5
  Sanjoy Das authored Aug 13, 2016
  1b1272f5
- [IRCE] Be resilient in the face of non-simplified loops · 2a2f14d7
  Sanjoy Das authored Aug 13, 2016
  2a2f14d7
- Fix bitcode auto-upgrade when using bitcode lazy loading · e62aaf23
  Mehdi Amini authored Aug 13, 2016
  e62aaf23
- Revert "Revert "Invariant start/end intrinsics overloaded for address space"" · 8c629ecf
  Mehdi Amini authored Aug 13, 2016
  8c629ecf
- Revert "Invariant start/end intrinsics overloaded for address space" · 164ac651
  Mehdi Amini authored Aug 13, 2016
  164ac651
- [IRCE] Use dyn_cast instead of explicit isa/cast; NFC · f2b7bafa
  Sanjoy Das authored Aug 13, 2016
  f2b7bafa
- [IRCE] Use range-for; NFC · d1d62a13
  Sanjoy Das authored Aug 13, 2016
  d1d62a13
Aug 13, 2016
- [ADT] Add a reserve() method to DenseSet as well as an insert() for R-value · fa0f96b0
  Mehdi Amini authored Aug 13, 2016
  fa0f96b0
- Revert "[ADT] Add a reserve method to DenseSet as well as an insert() for R-value" · bf001093
  Mehdi Amini authored Aug 13, 2016
  bf001093
- Limit DenseMap::setNumEntries input to 1<<31, in accordance with the 31 bits... · 782788b7
  Yaron Keren authored Aug 13, 2016
  782788b7
- Add missing REQUIRES in sancov/print_coverage_pcs.test: it requires aarch64 as well now · 7079efe1
  Mehdi Amini authored Aug 13, 2016
  7079efe1
- [ADT] Add a reserve method to DenseSet as well as an insert() for R-value · d866d8a0
  Mehdi Amini authored Aug 13, 2016
  d866d8a0
- [x86] add tests to show missed 64-bit immediate merging · 08c87667
  Sanjay Patel authored Aug 13, 2016
  08c87667
- Test commit · f24939b1
  Aditya Kumar authored Aug 13, 2016
  f24939b1
- [X86] Add a check of isCommutable at the top of... · 8c372a31
  Craig Topper authored Aug 13, 2016
  8c372a31
- [AVX-512] Add isCommutable to scalar FMA3 instructions. · eafdbecc
  Craig Topper authored Aug 13, 2016
  eafdbecc
- [AVX-512] Add commutable flags to 132 form FMA3 instructions. · 5f2441d8
  Craig Topper authored Aug 13, 2016
  5f2441d8
- [X86] Remove patterns for (vzmovl (insert_subvector undef,... · e5115aa4
  Craig Topper authored Aug 13, 2016
  e5115aa4
- [AVX-512] Remove an AddedComplexity that was prioritizing basic vzmovl... · 3f8126e6
  Craig Topper authored Aug 13, 2016
  3f8126e6
- [AVX-512] Add patterns to support VZEXT_MOVL from 512-bit vectors with 64-bit and 32-bit elements. · 600685d5
  Craig Topper authored Aug 13, 2016
  600685d5
- [PM] Port LoopDataPrefetch to new pass manager · 1eca6bc6
  Teresa Johnson authored Aug 13, 2016
  1eca6bc6