[clang] [llvm] [AMDGPU] Implement 'llvm.get.fpenv' and 'llvm.set.fpenv' (PR #83906)

Joseph Huber via cfe-commits cfe-commits at lists.llvm.org
Mon Mar 4 13:35:12 PST 2024


================
@@ -1122,7 +1122,7 @@ class S_SETREG_B32_Pseudo <list<dag> pattern=[]> : SOPK_Pseudo <
   pattern>;
 
 def S_SETREG_B32 : S_SETREG_B32_Pseudo <
-  [(int_amdgcn_s_setreg (i32 SIMM16bit:$simm16), i32:$sdst)]> {
+  [(int_amdgcn_s_setreg (i32 timm:$simm16), i32:$sdst)]> {
----------------
jhuber6 wrote:

Okay, I think I got it to work by casting it to `int16_t` only on the store instruction. My previous mistake seems to have been changing the constant for both `set` and `get. It passes all the AMDGPU codegen tests now.

https://github.com/llvm/llvm-project/pull/83906


More information about the cfe-commits mailing list