Claude review: drm/bridge: pass down IRQ_HPD to the drivers

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

Patch Review

This is the bulk of the mechanical plumbing — extending `hpd_cb`, `hpd_notify` (bridge funcs callback), `oob_hotplug_event` (connector funcs callback), and all their callers/implementers to take the extra `enum drm_connector_status_extra` parameter.

**Issue: Spurious whitespace change in ti-tfp410.c**

```c
-
 static inline struct tfp410 *
 drm_connector_to_tfp410(struct drm_connector *connector)
```

A blank line between two `static inline` function definitions is removed. This is unrelated to the patch purpose and will create unnecessary merge noise.

**Issue: `drm_bridge_hpd_notify()` hardcodes `DRM_CONNECTOR_NO_EXTRA_STATUS`**

In `drm_bridge.c`:
```c
 	if (bridge->hpd_cb)
-		bridge->hpd_cb(bridge->hpd_data, status);
+		bridge->hpd_cb(bridge->hpd_data, status, DRM_CONNECTOR_NO_EXTRA_STATUS);
```

This is correct for this intermediate step since patch 3 renames this to `drm_bridge_hpd_notify_extra()` and makes it pass through. But it does mean patch 2 is not fully bisectable for any driver that wants to use `drm_bridge_hpd_notify()` with extra status between patches 2 and 3. This is fine since no such caller exists at that point.

**Observation: `drm_bridge_connector_handle_hpd` unconditionally sets `connector->status`**

```c
	mutex_lock(&dev->mode_config.mutex);
	connector->status = status;
	mutex_unlock(&dev->mode_config.mutex);

	drm_bridge_connector_hpd_notify(connector, status, extra_status);

	drm_kms_helper_connector_hotplug_event(connector);
```

When an IRQ_HPD event arrives with `status == connector_status_connected` (which is what DP altmode sends), this will still set `connector->status` and fire `drm_kms_helper_connector_hotplug_event()`. This means every IRQ_HPD also triggers a full connector hotplug re-detection cycle. For MST sideband messages that can be frequent, this could cause unnecessary work. Is this intentional? The MSM driver's internal HPD path (via `DP_DP_IRQ_HPD_INT_MASK`) only queues `EV_IRQ_HPD_INT` without re-triggering the full hotplug cycle. Consider whether `drm_bridge_connector_handle_hpd` should skip the status update and hotplug event when only `extra_status` is set (i.e., the status hasn't actually changed).

**Intel i915: extra_status is ignored**

The `intel_dp_oob_hotplug_event()` function gains the `extra_status` parameter but doesn't use it. Looking at the current implementation, i915 only tracks connect/disconnect state transitions via `oob_hotplug_last_state`. If an IRQ_HPD arrives with the same connection status, the `hpd_high != test_bit(...)` check will be false and the event will be silently dropped. This means DP MST events via USB-C will be lost on i915. This may be intentional for now (the cover letter says only MSM is wired up), but it's worth noting as a gap — i915 USB-C DP MST will still be broken.

---
Generated by Claude Code Patch Reviewer