[llvm] [AArch64][GISel] Always fold G_SHL into addressing mode where possible, unless the subtarget has addr-lsl-slow-14 (PR #96603)

[via llvm-commits]

================
@@ -12,17 +12,16 @@ declare void @foo()
 define i16 @halfword(ptr %ctx, i32 %xor72) nounwind {
----------------
Him188 wrote:

This was actually coped from `AArch64/aarch64-fold-lslfast.ll`. Let's add GISEL RUN lines to this file and delete the new file.


https://github.com/llvm/llvm-project/pull/96603