Claude review: drm/i915/display: Add drm helper to check pr optimization support

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

Patch Review

**Documentation typo:**
```
 * Returns %true if the PR BW optimiation is supported for @tunnel.
```
"optimiation" → "optimization"

**Redundancy with existing tree code:** The drm-next tree already reads `DP_PANEL_REPLAY_OPTIMIZATION_SUPPORT` from `DP_TUNNELING_CAPABILITIES` in the `drm_dp_tunnel_detect()` debug print (line ~793 of `drm_dp_tunnel.c`), but doesn't store it. This patch correctly adds the `pr_optimization_support` bitfield and caches the value at tunnel creation time in `create_tunnel()`. However, note that `create_tunnel()` is called from `drm_dp_tunnel_detect()`, and the capability is already read from the same `regs` struct there. The approach is consistent with how `bw_alloc_supported` is handled — good.

**Missing update in `update_tunnel_regs()`:** The existing code checks whether `bw_alloc_supported` has changed across hotplug/resume in `update_tunnel_regs()` (around line 628). The new `pr_optimization_support` field is only set in `create_tunnel()` but is never updated or checked for changes in `update_tunnel_regs()`. If the tunnel bridge firmware were updated or the bridge changed, the cached value could become stale. Consider whether this needs a similar change-detection path. This is a minor concern since capability changes between detect calls are unlikely, but it's an inconsistency with the `bw_alloc_supported` pattern.

**No `EXPORT_SYMBOL_GPL`:** The existing exports in `drm_dp_tunnel.c` use `EXPORT_SYMBOL`. This patch follows the same convention, which is consistent. Just noting this is fine.

Overall this patch is clean and straightforward.

---
Generated by Claude Code Patch Reviewer