[mlir][gpu] Increase default SM version from 35 to 50

[guraypp]

Summary

Current SM version is 35 but it is deprecated long time ago. D155563 introduced ptxas compilations, using sm_35 causes failures in builtbot. This change increase default SM version to 50.

[guraypp]

builtbot failures due to sm_35 is below

lib/libmlir_runner_utils.so --entry-point-result=void
<unknown>:0: error: Invoking ptxas is failed, see the file: ptxas fatal   : Value 'sm_35' is not defined for option 'gpu-name'

mlir/lib/Dialect/GPU/Transforms/SerializeToCubin.cpp

 class SerializeToCubinPass
     : public PassWrapper<SerializeToCubinPass, gpu::SerializeToBlobPass> {
   static llvm::once_flag initializeBackendOnce;
 
 public:
   MLIR_DEFINE_EXPLICIT_INTERNAL_INLINE_TYPE_ID(SerializeToCubinPass)
 
   SerializeToCubinPass(StringRef triple = "nvptx64-nvidia-cuda",
-                       StringRef chip = "sm_35", StringRef features = "+ptx60",
+                       StringRef chip = "sm_50", StringRef features = "+ptx60",
                        int optLevel = 2, bool dumpPtx = false,
                        bool usePtxas = true, StringRef ptxasParams = {});
 
   StringRef getArgument() const override { return "gpu-to-cubin"; }
   StringRef getDescription() const override {
     return "Lower GPU kernel function to CUBIN binary "
            "annotations";
   }

mlir/test/lib/Dialect/GPU/TestConvertGPUKernelToCubin.cpp

private:
   // Serializes PTX to CUBIN.
   std::unique_ptr<std::vector<char>>
   serializeISA(const std::string &isa) override;
 };
 } // namespace
 
 TestSerializeToCubinPass::TestSerializeToCubinPass() {
   this->triple = "nvptx64-nvidia-cuda";
-  this->chip = "sm_35";
+  this->chip = "sm_50";
   this->features = "+ptx60";
 }
 
 void TestSerializeToCubinPass::getDependentDialects(
     DialectRegistry &registry) const {
   registerNVVMDialectTranslation(registry);
   gpu::SerializeToBlobPass::getDependentDialects(registry);
 }