bugfix: fix shared memory alignment conflict in sampling.cuh (#1402)

<!-- .github/pull_request_template.md -->

## 📌 Description

Fix compilation issue on gb200 due to shared mem alignment conflict.

## 🔍 Related Issues

#1400

## 🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull
request, please make sure the following items are complete.

### ✅ Pre-commit Checks

- [ ] I have installed `pre-commit` by running `pip install pre-commit`
(or used your preferred method).
- [ ] I have installed the hooks with `pre-commit install`.
- [ ] I have run the hooks manually with `pre-commit run --all-files`
and fixed any reported issues.

> If you are unsure about how to set up `pre-commit`, see [the
pre-commit documentation](https://pre-commit.com/).

## 🧪 Tests

- [ ] Tests have been added or updated as needed.
- [ ] All tests are passing (`unittest`, etc.).

## Reviewer Notes

<!-- Optional: anything you'd like reviewers to focus on, concerns, etc.
-->

Author: 842974287

include/flashinfer/sampling.cuh

@@ -739,8 +739,8 @@ __global__ void SamplingFromLogitsKernel(DType* logits, IdType* output, IdType*
   const uint32_t row_idx = indices == nullptr ? bx : indices[bx];
   using SharedMem = typename BlockReduce<DataAndIndex<DType, IdType>, BLOCK_THREADS,
                                          REDUCE_ALGORITHM>::TempStorage;
-  extern __shared__ __align__(alignof(SharedMem)) uint8_t smem_sampling[];
-  auto& temp_storage = reinterpret_cast<SharedMem&>(smem_sampling);
+  extern __shared__ __align__(alignof(SharedMem)) uint8_t smem_sampling_logit[];
+  auto& temp_storage = reinterpret_cast<SharedMem&>(smem_sampling_logit);
 
   vec_t<DType, VEC_SIZE> logits_vec;
   DataAndIndex<DType, IdType> max_data = {-cuda::std::numeric_limits<DType>::infinity(), 0};