Fix stream ordering for brute-force L2 sqrt post-processing

[dantegd]

Run the fused brute-force KNN distance post-processing on the same selected CUDA stream as fusedL2Knn.

Problem

In brute_force_knn_impl(), a stream is selected using:

auto stream = raft::resource::get_next_usable_stream(handle, i);

fusedL2Knn is launched on this selected stream. For L2SqrtExpanded and L2SqrtUnexpanded, the subsequent square-root map is instead launched using the original handle, whose main stream may differ from the selected stream:

fusedL2Knn(..., stream, ...);

raft::linalg::map(
  handle,
  ...);

There is no dependency making the main stream wait for the selected stream before the map reads the KNN output. The later stream-pool synchronization does not repair this read-before-write race.

Under concurrent cuML HDBSCAN workloads using managed memory on GB300, the corrupted CUDA context was subsequently reported by RAFT MST as:

copy_if failed on 2nd step: cudaErrorIllegalAddress

Related cuML issue: NVIDIA/cuml#8510
Companion RAFT PR: NVIDIA/raft#3125

[betatim]

Do we need more review on this or can we merge it?

Getting this merged would (hopefully) make the cuml nightly CI happy

[achirkin]

/merge