diff options
-rw-r--r-- | llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp | 13 | ||||
-rw-r--r-- | llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp | 6 |
2 files changed, 19 insertions, 0 deletions
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp index d49b7b8ea9d..c07ca85d3a4 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp @@ -1428,6 +1428,19 @@ Instruction *InstCombiner::visitCallInst(CallInst &CI) { } break; + case Intrinsic::x86_avx_storeu_ps_256: + case Intrinsic::x86_avx_storeu_pd_256: + case Intrinsic::x86_avx_storeu_dq_256: + // Turn X86 storeu -> store if the pointer is known aligned. + if (getOrEnforceKnownAlignment(II->getArgOperand(0), 32, DL, II, AC, DT) >= + 32) { + Type *OpPtrTy = + PointerType::getUnqual(II->getArgOperand(1)->getType()); + Value *Ptr = Builder->CreateBitCast(II->getArgOperand(0), OpPtrTy); + return new StoreInst(II->getArgOperand(1), Ptr); + } + break; + case Intrinsic::x86_vcvtph2ps_128: case Intrinsic::x86_vcvtph2ps_256: { auto Arg = II->getArgOperand(0); diff --git a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp index b5afe0a62ce..ad70a70a0d6 100644 --- a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp +++ b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp @@ -687,6 +687,9 @@ static bool isAddressUse(Instruction *Inst, Value *OperandVal) { case Intrinsic::x86_sse_storeu_ps: case Intrinsic::x86_sse2_storeu_pd: case Intrinsic::x86_sse2_storeu_dq: + case Intrinsic::x86_avx_storeu_ps_256: + case Intrinsic::x86_avx_storeu_pd_256: + case Intrinsic::x86_avx_storeu_dq_256: if (II->getArgOperand(0) == OperandVal) isAddress = true; break; @@ -711,6 +714,9 @@ static MemAccessTy getAccessType(const Instruction *Inst) { case Intrinsic::x86_sse_storeu_ps: case Intrinsic::x86_sse2_storeu_pd: case Intrinsic::x86_sse2_storeu_dq: + case Intrinsic::x86_avx_storeu_ps_256: + case Intrinsic::x86_avx_storeu_pd_256: + case Intrinsic::x86_avx_storeu_dq_256: AccessTy.MemTy = II->getArgOperand(0)->getType(); break; } |