Adjust clang format to the new versions, revert broken macro layout (#714)

projects/hip-tests/catch/performance/kernelLaunch/hipExtLaunchKernel.cc

@@ -31,8 +31,7 @@ THE SOFTWARE.
  * Contains performance tests for kernel launch overhead benchmarking.
  */
 
-template <KernelType kernel_type, bool timer_type>
-class ExtLaunchKernelBenchmark
+template <KernelType kernel_type, bool timer_type> class ExtLaunchKernelBenchmark
     : public KernelLaunchBenchmark<ExtLaunchKernelBenchmark<kernel_type, timer_type>, timer_type> {
  public:
   constexpr void LaunchKernel() {

projects/hip-tests/catch/performance/kernelLaunch/hipLaunchCooperativeKernel.cc

@@ -32,8 +32,7 @@ THE SOFTWARE.
  * Contains performance tests for kernel launch overhead benchmarking.
  */
 
-template <KernelType kernel_type, bool timer_type>
-class LaunchCooperativeKernelBenchmark
+template <KernelType kernel_type, bool timer_type> class LaunchCooperativeKernelBenchmark
     : public KernelLaunchBenchmark<LaunchCooperativeKernelBenchmark<kernel_type, timer_type>,
                                    timer_type> {
  public:

projects/hip-tests/catch/performance/kernelLaunch/hipLaunchKernel.cc

@@ -31,8 +31,7 @@ THE SOFTWARE.
  * Contains performance tests for kernel launch overhead benchmarking.
  */
 
-template <KernelType kernel_type, bool timer_type>
-class LaunchKernelBenchmark
+template <KernelType kernel_type, bool timer_type> class LaunchKernelBenchmark
     : public KernelLaunchBenchmark<LaunchKernelBenchmark<kernel_type, timer_type>, timer_type> {
  public:
   constexpr void LaunchKernel() {

projects/hip-tests/catch/performance/kernelLaunch/triple_chevron.cc

@@ -31,8 +31,7 @@ THE SOFTWARE.
  * Contains performance tests for kernel launch overhead benchmarking.
  */
 
-template <KernelType kernel_type, bool timer_type>
-class TripleChevronBenchmark
+template <KernelType kernel_type, bool timer_type> class TripleChevronBenchmark
     : public KernelLaunchBenchmark<TripleChevronBenchmark<kernel_type, timer_type>, timer_type> {
  public:
   constexpr void LaunchKernel() {

projects/hip-tests/catch/performance/memcpy/hipMemcpyWithKernel.cc

@@ -68,9 +68,9 @@ class MemcpyHtoDKernelDtoHv2AsyncBenchmark
     REQUIRE(sum == count * (count - 1) / 2);
   }
 };
-template <typename BenchmarkType>
-static void RunBenchmark(LinearAllocs host_allocation_type, LinearAllocs device_allocation_type,
-                         size_t size) {
+template <typename BenchmarkType> static void RunBenchmark(LinearAllocs host_allocation_type,
+                                                           LinearAllocs device_allocation_type,
+                                                           size_t size) {
   BenchmarkType benchmark;
   if (size < 1_KB) {
     benchmark.AddSectionName(std::to_string(size));

projects/hip-tests/catch/performance/warpSync/warpSync.cc

@@ -124,8 +124,8 @@ __global__ void reduceOpSync(T* __restrict__ output, const T* __restrict__ input
   }
 }
 
-template <class T, template <typename> class Op>
-class AtomicBenchmark : public Benchmark<AtomicBenchmark<T, Op>> {
+template <class T, template <typename> class Op> class AtomicBenchmark
+    : public Benchmark<AtomicBenchmark<T, Op>> {
  public:
   void operator()(T* output, const T* input, int numItems, unsigned long long mask) {
     dim3 blockDim = {kBlockDim};
@@ -157,8 +157,8 @@ class AtomicBenchmark : public Benchmark<AtomicBenchmark<T, Op>> {
   }
 };
 
-template <class T, template <typename> class Op>
-class ReduceSyncBenchmark : public Benchmark<ReduceSyncBenchmark<T, Op>> {
+template <class T, template <typename> class Op> class ReduceSyncBenchmark
+    : public Benchmark<ReduceSyncBenchmark<T, Op>> {
  public:
   void operator()(T* output, T* input, int numItems, unsigned long long mask) {
     dim3 blockDim = {kBlockDim};