This is an automated email from the git hooks/post-receive script.
unknown user pushed a change to branch master in repository llvm.
from 1639229 AMDGPU: Constant fold when immediate is materialized new eb75eea [AArch64] Consider all vector types for FeatureSlowMisaligned [...] new 9975cec [NVPTX] Add CHECK-LABEL where appropriate to fast-math.ll test. new e9bf848 [TM] Restore default TargetOptions in TargetMachine::resetTar [...] new 108d72a [loop-unroll] Properly populate LoopInfo for loops cloned in [...] new e9fdb8b [InstCombine] add a wrapper for a common pair of transforms; NFCI
The 5 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: include/llvm/Target/TargetMachine.h | 1 + lib/Target/AArch64/AArch64TargetTransformInfo.cpp | 23 ++++---- lib/Target/TargetMachine.cpp | 9 ++- lib/Transforms/InstCombine/InstCombineAddSub.cpp | 12 +--- lib/Transforms/InstCombine/InstCombineAndOrXor.cpp | 28 ++------- lib/Transforms/InstCombine/InstCombineInternal.h | 19 ++++-- .../InstCombine/InstCombineMulDivRem.cpp | 31 +++------- lib/Transforms/InstCombine/InstCombineShifts.cpp | 9 +-- .../InstCombine/InstructionCombining.cpp | 20 ++++--- lib/Transforms/Utils/LoopUnrollRuntime.cpp | 8 ++- test/Analysis/CostModel/AArch64/store.ll | 58 ++++++++++++++++--- test/CodeGen/NVPTX/fast-math.ll | 17 ++---- test/CodeGen/PowerPC/change-no-infs.ll | 67 ++++++++++++++++++++++ test/CodeGen/X86/change-unsafe-fp-math.ll | 56 ++++++++++++++++++ 14 files changed, 246 insertions(+), 112 deletions(-) create mode 100644 test/CodeGen/PowerPC/change-no-infs.ll create mode 100644 test/CodeGen/X86/change-unsafe-fp-math.ll