[llvm-branch-commits] [llvm] [AtomicExpand] Add bitcasts when expanding load atomic vector (PR #148900)

[Simon Pilgrim via llvm-branch-commits]

================
@@ -1220,6 +1220,11 @@ def : Pat<(v2i64 (scalar_to_vector (i64 (atomic_load_64 addr:$src)))),
 def : Pat<(v2i64 (scalar_to_vector (i64 (atomic_load_64 addr:$src)))),
   (VMOV64toPQIZrm addr:$src)>, Requires<[HasAVX512]>;
 
+def : Pat<(v2i64 (atomic_load_128_v2i64 addr:$src)),
+           (VMOVAPDrm addr:$src)>; // load atomic <2 x i64>
+def : Pat<(v4i32 (atomic_load_128_v4i32 addr:$src)),
+           (VMOVAPDrm addr:$src)>; // load atomic <4 x i32>
----------------
RKSimon wrote:

These need to be tagged with Requires<[UseAVX] and the SSE2/AVX512 variants adding as well (MOVAPDrm / VMOVAPDZ128rm)

https://github.com/llvm/llvm-project/pull/148900