[llvm] [AMDGPU][True16][MC] true16 for v_alignbit_b32 (PR #119409)
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Tue Dec 17 17:16:42 PST 2024
================
@@ -3003,6 +3060,40 @@ def : GCNPat <
sub1)
>;
+let True16Predicate = UseFakeTrue16Insts in
+def : GCNPat <
+ (i64 (bswap i64:$a)),
+ (REG_SEQUENCE VReg_64,
+ (V_BFI_B32_e64 (S_MOV_B32 (i32 0x00ff00ff)),
----------------
arsenm wrote:
Can you conditionalize the use of the source modifier pattern in the input? something like `!if(SupportsSrcMods, (VOP3Mods $src0), $src0)`
https://github.com/llvm/llvm-project/pull/119409
More information about the llvm-commits
mailing list