[PATCH] AArch64 Neon Scalar Floating-point Compare Instructions
Tim Northover
t.p.northover at gmail.com
Tue Oct 8 08:30:52 PDT 2013
+def int_aarch64_neon_vceqs :
+ Intrinsic<[llvm_v1i32_ty], [llvm_v1f32_ty, llvm_v1f32_ty], [IntrNoMem]>;
+def int_aarch64_neon_vceqd :
+ Intrinsic<[llvm_v1i64_ty], [llvm_v1f64_ty, llvm_v1f64_ty], [IntrNoMem]>;
I'm not sure about these names. It's not at all obvious to me that
int_aarch64_neon_vceq acts on integer types but these versions with
's' and 'd' act on floats. It might be better to use just a single
overloaded version.
Cheers.
Tim.
More information about the llvm-commits
mailing list