diff options
Diffstat (limited to 'llvm/lib/Analysis/ConstantFolding.cpp')
| -rw-r--r-- | llvm/lib/Analysis/ConstantFolding.cpp | 15 |
1 files changed, 8 insertions, 7 deletions
diff --git a/llvm/lib/Analysis/ConstantFolding.cpp b/llvm/lib/Analysis/ConstantFolding.cpp index 7341dad..dd98b62 100644 --- a/llvm/lib/Analysis/ConstantFolding.cpp +++ b/llvm/lib/Analysis/ConstantFolding.cpp @@ -2631,14 +2631,14 @@ static Constant *ConstantFoldScalarCall1(StringRef Name, case Intrinsic::nvvm_ceil_d: return ConstantFoldFP( ceil, APF, Ty, - nvvm::GetNVVMDenromMode( + nvvm::GetNVVMDenormMode( nvvm::UnaryMathIntrinsicShouldFTZ(IntrinsicID))); case Intrinsic::nvvm_fabs_ftz: case Intrinsic::nvvm_fabs: return ConstantFoldFP( fabs, APF, Ty, - nvvm::GetNVVMDenromMode( + nvvm::GetNVVMDenormMode( nvvm::UnaryMathIntrinsicShouldFTZ(IntrinsicID))); case Intrinsic::nvvm_floor_ftz_f: @@ -2646,7 +2646,7 @@ static Constant *ConstantFoldScalarCall1(StringRef Name, case Intrinsic::nvvm_floor_d: return ConstantFoldFP( floor, APF, Ty, - nvvm::GetNVVMDenromMode( + nvvm::GetNVVMDenormMode( nvvm::UnaryMathIntrinsicShouldFTZ(IntrinsicID))); case Intrinsic::nvvm_rcp_rm_ftz_f: @@ -2679,11 +2679,12 @@ static Constant *ConstantFoldScalarCall1(StringRef Name, case Intrinsic::nvvm_round_ftz_f: case Intrinsic::nvvm_round_f: case Intrinsic::nvvm_round_d: { - // Use APFloat implementation instead of native libm call, as some - // implementations (e.g. on PPC) do not preserve the sign of negative 0. + // nvvm_round is lowered to PTX cvt.rni, which will round to nearest + // integer, choosing even integer if source is equidistant between two + // integers, so the semantics are closer to "rint" rather than "round". bool IsFTZ = nvvm::UnaryMathIntrinsicShouldFTZ(IntrinsicID); auto V = IsFTZ ? FTZPreserveSign(APF) : APF; - V.roundToIntegral(APFloat::rmNearestTiesToAway); + V.roundToIntegral(APFloat::rmNearestTiesToEven); return ConstantFP::get(Ty->getContext(), V); } @@ -2708,7 +2709,7 @@ static Constant *ConstantFoldScalarCall1(StringRef Name, return nullptr; return ConstantFoldFP( sqrt, APF, Ty, - nvvm::GetNVVMDenromMode( + nvvm::GetNVVMDenormMode( nvvm::UnaryMathIntrinsicShouldFTZ(IntrinsicID))); // AMDGCN Intrinsics: |