[llvm-branch-commits] [llvm] [AtomicExpand] Add bitcasts when expanding load atomic vector (PR #120716)

[Simon Pilgrim via llvm-branch-commits]

================
@@ -1211,6 +1211,11 @@ def : Pat<(v4i32 (scalar_to_vector (i32 (atomic_load_32 addr:$src)))),
 def : Pat<(v2i64 (scalar_to_vector (i64 (atomic_load_64 addr:$src)))),
            (MOV64toPQIrm  addr:$src)>; // load atomic <2 x i32,float>
 
+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 require SSE/AVX/AVX512 variants (see below) - but x86 doesn't guarantee atomics for anything above 8 bytes (and those must be aligned to avoid cacheline crossing).

https://github.com/llvm/llvm-project/pull/120716