Index: llvm/lib/Target/PowerPC/PPCISelLowering.cpp =================================================================== --- llvm/lib/Target/PowerPC/PPCISelLowering.cpp +++ llvm/lib/Target/PowerPC/PPCISelLowering.cpp @@ -1317,6 +1317,8 @@ MaxLoadsPerMemcmpOptSize = 4; } + IsStrictFPEnabled = true; + // Let the subtarget (CPU) decide if a predictable select is more expensive // than the corresponding branch. This information is used in CGP to decide // when to convert selects into branches. Index: llvm/lib/Target/PowerPC/PPCInstrInfo.td =================================================================== --- llvm/lib/Target/PowerPC/PPCInstrInfo.td +++ llvm/lib/Target/PowerPC/PPCInstrInfo.td @@ -3477,7 +3477,7 @@ def : Pat<(f64 (extloadf32 xaddr:$src)), (COPY_TO_REGCLASS (LFSX xaddr:$src), F8RC)>; -def : Pat<(f64 (fpextend f32:$src)), +def : Pat<(f64 (any_fpextend f32:$src)), (COPY_TO_REGCLASS $src, F8RC)>; } Index: llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll =================================================================== --- llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll +++ llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll @@ -7168,19 +7168,19 @@ ; PC64LE-NEXT: addis 3, 2, .LCPI133_0@toc@ha ; PC64LE-NEXT: addis 4, 2, .LCPI133_1@toc@ha ; PC64LE-NEXT: addis 5, 2, .LCPI133_2@toc@ha -; PC64LE-NEXT: lfs 1, .LCPI133_0@toc@l(3) +; PC64LE-NEXT: lfs 3, .LCPI133_0@toc@l(3) ; PC64LE-NEXT: lfs 2, .LCPI133_1@toc@l(4) -; PC64LE-NEXT: lfs 3, .LCPI133_2@toc@l(5) +; PC64LE-NEXT: lfs 1, .LCPI133_2@toc@l(5) ; PC64LE-NEXT: blr ; ; PC64LE9-LABEL: constrained_vector_fpext_v3f32: ; PC64LE9: # %bb.0: # %entry ; PC64LE9-NEXT: addis 3, 2, .LCPI133_0@toc@ha -; PC64LE9-NEXT: lfs 1, .LCPI133_0@toc@l(3) +; PC64LE9-NEXT: lfs 3, .LCPI133_0@toc@l(3) ; PC64LE9-NEXT: addis 3, 2, .LCPI133_1@toc@ha ; PC64LE9-NEXT: lfs 2, .LCPI133_1@toc@l(3) ; PC64LE9-NEXT: addis 3, 2, .LCPI133_2@toc@ha -; PC64LE9-NEXT: lfs 3, .LCPI133_2@toc@l(3) +; PC64LE9-NEXT: lfs 1, .LCPI133_2@toc@l(3) ; PC64LE9-NEXT: blr entry: %result = call <3 x double> @llvm.experimental.constrained.fpext.v3f64.v3f32(