[llvm] [WIP][AMDGPU] Support the RF inline asm constraint (PR #195123)

[Jun Wang via llvm-commits]

================
@@ -0,0 +1,478 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
+
+; RUN: llc -mtriple=amdgcn -mcpu=tonga < %s | FileCheck --check-prefix=VI %s
----------------
jwanggit86 wrote:

fixed.

https://github.com/llvm/llvm-project/pull/195123