[clang] [AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

[Shilei Tian via cfe-commits]

================
@@ -0,0 +1,194 @@
+// RUN: %clang_cc1 -triple amdgcn-- -target-cpu gfx1100 -S -verify=expected -o - %s
+// REQUIRES: amdgpu-registered-target
----------------
shiltian wrote:

IIRC, we usually use `-err.cl` suffix for error check tests.

https://github.com/llvm/llvm-project/pull/140210