FailedChanges

Summary

  1. [libc++] Add a benchmark for std::map operations (details)
  2. [MachineBasicBlock] Fix a typo in function copySuccessor (details)
  3. NFC: Add whitespace change to .git-blame-ignore-revs (details)
  4. [X86][AVX] lowerShuffleWithSHUFPS - handle missed canonicalization cases. (details)
  5. [lldb] Add -l/--language option to script command (details)
  6. [VectorCombine] Don't vectorize scalar load under asan/hwasan/memtag/tsan (details)
Commit 74a9c6d7e1c49cd0e3a8e8072b8aa03f7a84caff by Louis Dionne
[libc++] Add a benchmark for std::map operations

Before tackling http://llvm.org/PR38722, make sure there is a baseline
benchmark.

Differential Revision: https://reviews.llvm.org/D62778
The file was addedlibcxx/benchmarks/map.bench.cpp
Commit 243ffd0cade71ddca9b0dffec1c8e9084b0f7745 by carrot
[MachineBasicBlock] Fix a typo in function copySuccessor

The condition used to decide if need to copy probability should be reversed.

Differential Revision: https://reviews.llvm.org/D87417
The file was modifiedllvm/lib/CodeGen/MachineBasicBlock.cpp (diff)
Commit 4ddd985ca941e48a016e8d7270921b4aa76afbe1 by Louis Dionne
NFC: Add whitespace change to .git-blame-ignore-revs
The file was modified.git-blame-ignore-revs (diff)
Commit a43e68b58b085797e2f1435765255ebd431db297 by llvm-dev
[X86][AVX] lowerShuffleWithSHUFPS - handle missed canonicalization cases.

PR47534 exposes a case where calling lowerShuffleWithSHUFPS directly from a derived repeated mask (found by is128BitLaneRepeatedShuffleMask) results in us using an non-canonicalized mask.

The missed canonicalization in this case is trivial - just commute the mask so we have more (swapped) LHS than RHS references so lowerShuffleWithSHUFPS can handle it.
The file was modifiedllvm/test/CodeGen/X86/vector-shuffle-avx512.ll (diff)
The file was modifiedllvm/lib/Target/X86/X86ISelLowering.cpp (diff)
Commit 127faae7529aee7e8508abebbc19212ce30bbf27 by Jonas Devlieghere
[lldb] Add -l/--language option to script command

Make it possible to run the script command with a different language
than currently selected.

  $ ./bin/lldb -l python
  (lldb) script -l lua
  >>> io.stdout:write("Hello, World!\n")
  Hello, World!

When passing the language option and a raw command, you need to separate
the flag from the script code with --.

  $ ./bin/lldb -l python
  (lldb) script -l lua -- io.stdout:write("Hello, World!\n")
  Hello, World!

Differential revision: https://reviews.llvm.org/D86996
The file was addedlldb/test/Shell/ScriptInterpreter/Lua/lua-python.test
The file was modifiedllvm/lib/Support/MemoryBuffer.cpp (diff)
The file was modifiedlldb/source/Commands/CommandObjectScript.h (diff)
The file was modifiedlldb/source/Commands/Options.td (diff)
The file was modifiedlldb/test/Shell/ScriptInterpreter/Lua/lua.test (diff)
The file was modifiedlldb/source/Commands/CommandObjectScript.cpp (diff)
The file was addedlldb/test/Shell/ScriptInterpreter/Python/python.test
Commit 4452cc4086aca1a424b2cd40da9fa120add522e7 by i
[VectorCombine] Don't vectorize scalar load under asan/hwasan/memtag/tsan

Similar to the tsan suppression in
`Utils/VNCoercion.cpp:getLoadLoadClobberFullWidthSize` (rL175034; load widening used by GVN),
the D81766 optimization should be suppressed under tsan due to potential
spurious data race reports:

  struct A {
    int i;
    const short s; // the load cannot be vectorized because
    int modify;    // it overlaps with bytes being concurrently modified
    long pad1, pad2;
  };
  // __tsan_read16 does not know that some bytes are undef and accessing is safe

Similarly, under asan, users can mark memory regions with
`__asan_poison_memory_region`. A widened load can lead to a spurious
use-after-poison error. hwasan/memtag should be similarly suppressed.

`mustSuppressSpeculation` suppresses asan/hwasan/tsan but not memtag, so
we need to exclude memtag in `vectorizeLoadInsert`.

Note, memtag suppression can be relaxed if the load is aligned to the
its granule (usually 16), but that is out of scope of this patch.

Reviewed By: spatel, vitalybuka

Differential Revision: https://reviews.llvm.org/D87538
The file was modifiedllvm/test/Transforms/VectorCombine/X86/load.ll (diff)
The file was modifiedllvm/lib/Transforms/Vectorize/VectorCombine.cpp (diff)