This is an automated email from the git hooks/post-receive script.

unknown user pushed a change to branch master in repository llvm.

from 225f28a6dba [Hexagon] Use MachineInstrBuilder instead of changing instr [...] new 998914d3010 Enable vectorizer-maximize-bandwidth by default. new 52c64527391 TableGen.cmake: Use DEPFILE for Ninja Generator with CMake>=3.7. new 1a1f5442634 [AMDGPU] simplify add x, *ext (setcc) => addc|subb x, 0, setcc

The 3 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "adds" were already present in the repository and have only been added to this reference.

Summary of changes: cmake/modules/TableGen.cmake | 29 +++++++++-- lib/Target/AMDGPU/AMDGPUInstrInfo.td | 10 ++++ lib/Target/AMDGPU/SIISelLowering.cpp | 39 ++++++++++++++ lib/Target/AMDGPU/SIISelLowering.h | 1 + lib/Target/AMDGPU/VOP2Instructions.td | 9 ++++ lib/Transforms/Vectorize/LoopVectorize.cpp | 2 +- test/CodeGen/AMDGPU/combine-cond-add-sub.ll | 43 ++++++++++++++++ .../AArch64/loop-vectorization-factors.ll | 32 ++++++------ .../LoopVectorize/AArch64/reduction-small-size.ll | 16 +++--- test/Transforms/LoopVectorize/ARM/gcc-examples.ll | 6 +-- .../LoopVectorize/X86/fp64_to_uint32-cost-model.ll | 4 +- test/Transforms/LoopVectorize/X86/gcc-examples.ll | 15 +++--- .../LoopVectorize/X86/masked_load_store.ll | 60 ++++++++++++---------- test/Transforms/LoopVectorize/X86/no_fpmath.ll | 2 +- .../LoopVectorize/X86/no_fpmath_with_hotness.ll | 2 +- .../LoopVectorize/X86/reduction-crash.ll | 2 +- .../X86/vectorization-remarks-loopid-dbg.ll | 2 +- .../LoopVectorize/X86/vectorization-remarks.ll | 2 +- 18 files changed, 205 insertions(+), 71 deletions(-) create mode 100644 test/CodeGen/AMDGPU/combine-cond-add-sub.ll