[clang] cuda clang: Add support for CUDA surfaces (PR #132883)

[Artem Belevich via cfe-commits]

================
@@ -0,0 +1,3329 @@
+// RUN: %clang_cc1 -triple nvptx-unknown-unknown -fcuda-is-device -O3 -o - %s -emit-llvm | FileCheck %s
+// RUN: %clang_cc1 -triple nvptx64-unknown-unknown -fcuda-is-device -O3 -o - %s -emit-llvm | FileCheck %s
+#include "../Headers/Inputs/include/cuda.h"
----------------
Artem-B wrote:

Relative include paths are a problem for tests -- some run remotely with only the test directory staged, so there may not always be ../Headers directory where the tests actually run. Replicating test input changes is OK.

https://github.com/llvm/llvm-project/pull/132883