Claude review: Re: [PATCH 1/2] drm/amdgpu: protect waitq access with userq_mutex in wait IOCTL

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

**Already fixed upstream via refcounting.** The current tree uses:
```c
waitq = amdgpu_userq_get(userq_mgr, wait_info->waitq_id);
```
(line 866) instead of the raw `xa_load()` this patch targets. The `amdgpu_userq_get()` function takes a kref under xa_lock, and `amdgpu_userq_put()` is called at the end (line 951), providing proper lifetime protection without holding the mutex across the entire loop.

**Even evaluated against the older base, there is a concern:**

- **Sleeping under mutex**: The patch holds `userq_mutex` across `dma_fence_wait(fences[i], true)` (line 909 in the diff context). `dma_fence_wait` with `intr=true` can sleep for an extended/unbounded period waiting for GPU work to complete. Holding `userq_mutex` during this time blocks all other userqueue operations (create, destroy, signal) for this file descriptor. This is a significant liveness concern — a slow or stalled fence would effectively deadlock the entire userqueue subsystem for the process. The refcounting approach in drm-next avoids this problem entirely.

---
Generated by Claude Code Patch Reviewer