[llvm] [SROA] Unfold gep of index select (PR #80983)

[Artem Belevich via llvm-commits]

================
@@ -158,11 +158,20 @@ bb:
 
 define i32 @test_select_idx_memcpy(i1 %c, ptr %p) {
----------------
Artem-B wrote:

This test looks rather odd right now. It's not at all obvious that the change is beneficial. We're still copying the same amount of data, but doing it in multiple calls of smaller chunks, and we have more GEPs.

Also we're testing both memcpy handling and gep(select(const)) now and it may be nice to separate the two.


https://github.com/llvm/llvm-project/pull/80983