[Mlir-commits] [compiler-rt] [llvm] [libc] [flang] [clang] [clang-tools-extra] [lldb] [libcxx] [mlir] [AArch64] add intrinsic to generate a bfi instruction (PR #79672)

[Rama Malladi]

RamaMalladiAWS wrote:

> OK. We would not usually add intrinsics like this without a strong motivating case, that could not be optimized in some other way. It is better to use target independent options when available, and inline assembly is available as a fallback if it is really needed. But I would recommend that they use normal and/or/shift operations and let us know about places the compiler isn't optimizing them as well as it could be.

I completely agree with the approach @davemgreen. In this case, the IR sequence wasn't optimized to a `bfi`. I can try to get a test-case and create an issue for better sequence if feasible. Thanks again.

https://github.com/llvm/llvm-project/pull/79672