[llvm] [LSV][AMDGPU] Vectorize unordered and monotonic atomic loads (PR #190152)

[Harrison Hao via llvm-commits]

================
@@ -130,17 +130,24 @@ namespace {
 // (We could in theory remove element-size from the this tuple.  We'd just need
 // to fix up the vector packing/unpacking code.)
 using EqClassKey =
-    std::tuple<const Value * /* result of getUnderlyingObject() */,
-               unsigned /* AddrSpace */,
-               unsigned /* Load/Store element size bits */,
-               char /* IsLoad; char b/c bool can't be a DenseMap key */
+    std::tuple<const Value *, // result of getUnderlyingObject()
+               unsigned,      // AddrSpace
+               unsigned,      // Load/Store element size bits
+               char,          // IsLoad; char b/c bool can't be a DenseMap key
+               unsigned,      // AtomicOrdering
----------------
harrisonGPU wrote:

Okay! I got it! Thanks! I forgot it just now. I have updated it. :-)

https://github.com/llvm/llvm-project/pull/190152