[SYCL][CUDA] Enable group_ballot tests (#581)

This was added in intel/llvm#5020

This patch should only be merged after intel/llvm#5020 is merged.

Author: npmiller

SYCL/SubGroupMask/Basic.cpp

@@ -1,6 +1,6 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
 // REQUIRES: gpu
-// UNSUPPORTED: cuda, hip
+// UNSUPPORTED: hip
 // GroupNonUniformBallot capability is supported on Intel GPU only
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // UNSUPPORTED: ze_debug-1,ze_debug4

SYCL/SubGroupMask/GroupSize.cpp

@@ -1,6 +1,6 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
 // REQUIRES: gpu
-// UNSUPPORTED: cuda, hip
+// UNSUPPORTED: hip
 // GroupNonUniformBallot capability is supported on Intel GPU only
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // UNSUPPORTED: ze_debug-1,ze_debug4