Started 1 mo 4 days ago
Took 3 hr 15 min on green-dragon-18

Success Build rL:372070 - C:372063 - #630 (Sep 16, 2019 10:42:04 PM)

Revisions
  • http://llvm.org/svn/llvm-project/llvm/trunk : 372070
  • http://llvm.org/svn/llvm-project/cfe/trunk : 372063
  • http://llvm.org/svn/llvm-project/compiler-rt/trunk : 372056
  • http://llvm.org/svn/llvm-project/debuginfo-tests/trunk : 364589
  • http://llvm.org/svn/llvm-project/zorg/trunk : 371835
  • http://llvm.org/svn/llvm-project/libcxx/trunk : 372034
  • http://llvm.org/svn/llvm-project/clang-tools-extra/trunk : 372037
Changes
  1. [git-llvm] Do not reinvent `@{upstream}` (take 2)

    This makes git-llvm more of a thin wrapper around git while temporarily
    maintaining backwards compatibility with past git-llvm behavior.

    Using @{upstream} makes git-llvm more robust when used with a nontrivial
    local repository.

    https://reviews.llvm.org/D67389 (detail/ViewSVN)
    by davezarzycki
  2. [X86] Split oversized vXi1 vector arguments and return values into scalars on avx512 targets.

    Previously we tried to split them into narrower v64i1 or v16i1
    pieces that each got promoted to vXi8 and then passed in a zmm
    or xmm register. But this crashes when you need to pass more
    pieces than available registers reserved for argument passing.

    The scalarizing done here generates much longer and slower code,
    but is consistent with the behavior of avx2 and earlier targets
    for these types.

    Fixes PR43323. (detail/ViewSVN)
    by ctopper
  3. [X86] Allow masked VBROADCAST instructions to be turned into BLENDM with a broadcast load to avoid a copy.

    The BLENDM instructions allow an 2 sources and an independent
    destination while masked VBROADCAST has the destination tied
    to the source. (detail/ViewSVN)
    by ctopper
  4. [X86] Add support for commuting EVEX VCMP instructons with any immediate value.

    Previously we limited to the EQ/NE/TRUE/FALSE/ORD/UNORD immediates. (detail/ViewSVN)
    by ctopper
  5. [X86] Add test case for missed opportunity to commute a VCMP instruction after unfolding one load in order to fold another load. (detail/ViewSVN)
    by ctopper
  6. [X86] Enable commuting of EVEX VCMP for all immediate values during isel. (detail/ViewSVN)
    by ctopper

Started by an SCM change (2 times)

This run spent:

  • 1 hr 0 min waiting;
  • 3 hr 15 min build duration;
  • 4 hr 15 min total from scheduled to completion.
LLVM/Clang Warnings: 1 warning.
    Test Result (no failures)