[llvm] AMDGPU: Report unaligned scratch access as fast if supported by tgt (PR #158036)

[via llvm-commits]

================
@@ -2090,10 +2090,16 @@ bool SITargetLowering::allowsMisalignedMemoryAccessesImpl(
   if (AddrSpace == AMDGPUAS::PRIVATE_ADDRESS ||
       AddrSpace == AMDGPUAS::FLAT_ADDRESS) {
     bool AlignedBy4 = Alignment >= Align(4);
+    if (Subtarget->hasUnalignedScratchAccessEnabled()) {
----------------
macurtis-amd wrote:

Updated title and description.

https://github.com/llvm/llvm-project/pull/158036