[llvm] [Matrix] Optimize static extracts with ShapeInfo (PR #141815)

[Jon Roelofs via llvm-commits]

================
@@ -55,11 +55,11 @@ define void @multiply_ntt(ptr %A, ptr %B, ptr %C, ptr %R) {
 ; REMARK-NEXT: Function:        multiply_ntt
 ; REMARK-NEXT: Args:
 ; REMARK-NEXT:   - String:          'Lowered with '
-; REMARK-NEXT:   - NumStores:       '4'
+; REMARK-NEXT:   - NumStores:       '0'
 ; REMARK-NEXT:   - String:          ' stores, '
-; REMARK-NEXT:   - NumLoads:        '10'
+; REMARK-NEXT:   - NumLoads:        '3'
 ; REMARK-NEXT:   - String:          ' loads, '
-; REMARK-NEXT:   - NumComputeOps:   '38'
+; REMARK-NEXT:   - NumComputeOps:   '0'
----------------
jroelofs wrote:

FIXME: not sure what's going on here. I didn't expect these to change.

https://github.com/llvm/llvm-project/pull/141815