[clang] [llvm] [AMDGPU] Add alignment attr & propagate alignment through make.buffer.rsrc inst (PR #145278)

[via llvm-commits]

================
@@ -85,7 +85,7 @@ __amdgpu_buffer_rsrc_t test_amdgcn_make_buffer_p0_nullptr(short stride, int num,
 
 // CHECK-LABEL: @test_amdgcn_make_buffer_p1_nullptr(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP0:%.*]] = tail call ptr addrspace(8) @llvm.amdgcn.make.buffer.rsrc.p8.p1(ptr addrspace(1) null, i16 [[STRIDE:%.*]], i32 [[NUM:%.*]], i32 [[FLAGS:%.*]])
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call align 4294967296 ptr addrspace(8) @llvm.amdgcn.make.buffer.rsrc.p8.p1(ptr addrspace(1) null, i16 [[STRIDE:%.*]], i32 [[NUM:%.*]], i32 [[FLAGS:%.*]])
----------------
Shoreshen wrote:

Hi @shiltian , the first operand of this is 0 pointer `__builtin_amdgcn_make_buffer_rsrc((global void *)0LL, stride, num, flags);`, so the attribute regard it as the maximum alignment...

https://github.com/llvm/llvm-project/pull/145278