[clang] [Clang] Default OpenMP and C/C++ compilations to uniform blocks (PR #221279)

[Nick Sarnie via cfe-commits]

================
@@ -140,6 +140,10 @@ class SPIRVTargetCodeGenInfo : public CommonSPIRTargetCodeGenInfo {
     return getABIInfo().getTarget().getTriple().getVendor() !=
            llvm::Triple::AMD;
   }
+  bool hasUniformWorkGroupLaunch() const override {
+    return getABIInfo().getTarget().getTriple().getVendor() ==
----------------
sarnex wrote:

Can @KseniyaTikhomirova or @tahonermann double check the expected behavior for SYCL/SPIR-V, I'm not really familiar with kernel launch semantics, thanks!

https://github.com/llvm/llvm-project/pull/221279