[llvm] [NVPTX] Support for memory orderings for cmpxchg (PR #126159)

[Artem Belevich via llvm-commits]

================
@@ -177,6 +177,7 @@ define float @atomicrmw_add_f32_generic(ptr %addr, float %val) {
 
 ; CHECK-LABEL: atomicrmw_add_f16_generic
 define half @atomicrmw_add_f16_generic(ptr %addr, half %val) {
+; CHECK: membar.sys
----------------
Artem-B wrote:

Let's convert this test to autogenerated checks, too.

https://github.com/llvm/llvm-project/pull/126159