[rocm7.0_internal_testing] Prevent static initialization of at::cuda::warp_size() (#2293)

Fixes SWDEV-540240, SWDEV-540309, SWDEV-539989

### Error
```
#24 437.7   what():  HIP error: no ROCm-capable device is detected
#24 437.7 HIP kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
#24 437.7 For debugging consider passing AMD_SERIALIZE_KERNEL=3
#24 437.7 Device-side assertions were explicitly omitted for this error check; the error probably arose while initializing the DSA handlers.
#24 437.7 Exception raised from c10_hip_check_implementation at /pytorch/c10/hip/HIPException.cpp:44 (most recent call first):
#24 437.7 frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x88 (0x7f272de18738 in /usr/local/lib/python3.10/dist-packages/torch/lib/libc10.so)
#24 437.7 frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0x55 (0x7f272ddb42ed in /usr/local/lib/python3.10/dist-packages/torch/lib/libc10.so)
...
#24 437.7 frame #7: at::cuda::getCurrentDeviceProperties() + 0x9 (0x7f270b5874e9 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_hip.so)
#24 437.7 frame #8: at::cuda::warp_size() + 0x9 (0x7f270b587509 in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_hip.so)
#24 437.7 frame #9: <unknown function> + 0x81ac8b (0x7f2709c27c8b in /usr/local/lib/python3.10/dist-packages/torch/lib/libtorch_hip.so)
```

### Explanation

80cca70
created a static global variable that used `at::cuda::warp_size()` to
initialize its value, which needs GPUs to be visible to query device
properties. However, GPUs are not present on CPU-only build systems.

### Solution
Convert static variable into a static function, thus preventing static
initialization.

### Validation

http://rocm-ci.amd.com/job/pyt_whl_docker_mainline/1461/artifact/build_artifacts.txt/*view*/

Ran microbenchmark to confirm basic functionality:
```
root@ubb4-rack-22:/var/lib/jenkins/pytorch-micro-benchmarking# python3 micro_benchmarking_pytorch.py --network resnet50
INFO: running forward and backward for warmup.
INFO: running the benchmark..
OK: finished running benchmark..
--------------------SUMMARY--------------------------
Microbenchmark for network : resnet50
Num devices: 1
Dtype: FP32
Mini batch size [img] : 64
Time per mini-batch : 0.10158218145370483
Throughput [img/sec] : 630.0317544289736=
```

Author: ethanwee1

aten/src/ATen/native/cuda/Embedding.cu

@@ -369,7 +369,7 @@ Tensor & embedding_renorm_cuda_(Tensor & self, const Tensor & indices,
 
     int warp_size = at::cuda::warp_size();
     TORCH_INTERNAL_ASSERT(num_threads() % warp_size == 0 &&
-                  num_threads() <= cuda_utils::kCUDABlockReduceMaxThreads,
+                  num_threads() <= cuda_utils::kCUDABlockReduceMaxThreads(),
                   "BlockReduceSum requires all warps be active");
     const int64_t *num_unique_indices_ptr = num_unique_indices.const_data_ptr<int64_t>();
     dim3 grid = unique_indices.numel();

aten/src/ATen/native/cuda/MultinomialKernel.cu

@@ -86,7 +86,7 @@ void renormRows(Tensor& t) {
   TORCH_CHECK(props != nullptr);
   int numSM = props->multiProcessorCount;
   const int64_t maxThreads = std::min(
-      props->maxThreadsPerBlock, cuda_utils::kCUDABlockReduceMaxThreads);
+      props->maxThreadsPerBlock, cuda_utils::kCUDABlockReduceMaxThreads());
 
   int warp_size = at::cuda::warp_size();
   dim3 grid(rows < numSM * 4 ? rows : numSM * 4);

aten/src/ATen/native/cuda/TensorModeKernel.cu

@@ -207,7 +207,7 @@ void handle_fused_mode(
   constexpr int num_threads = size / 2;
   int warp_size = at::cuda::warp_size();
   TORCH_INTERNAL_ASSERT(num_threads % warp_size == 0 &&
-                num_threads <= cuda_utils::kCUDABlockReduceMaxThreads, "");
+                num_threads <= cuda_utils::kCUDABlockReduceMaxThreads(), "");
   const auto memsize =
       (sizeof(scalar_t) * size) + (2 * size * sizeof(unsigned int));
   compute_mode<scalar_t, size>

aten/src/ATen/native/cuda/block_reduce.cuh

@@ -15,9 +15,13 @@ constexpr int kCUDABlockReduceNumThreads = 512;
 // ROCm NOTE: C10_WARP_SIZE should only be used inside device functions,
 // and kCUDABlockReduceMaxThreads is a host-side variable.
 #ifdef USE_ROCM
-static const int kCUDABlockReduceMaxThreads = at::cuda::warp_size() * at::cuda::warp_size();
+static int kCUDABlockReduceMaxThreads() {
+    return at::cuda::warp_size() * at::cuda::warp_size();
+}
 #else
-constexpr int kCUDABlockReduceMaxThreads = C10_WARP_SIZE * C10_WARP_SIZE;
+constexpr int kCUDABlockReduceMaxThreads() {
+    return C10_WARP_SIZE * C10_WARP_SIZE;
+}
 #endif
 
 // Sums `val` across all threads in a warp.