Claude review: drm/bridge: display-connector: trigger initial HPD event for DP

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

Patch Review

**Purpose:** After enabling HPD, schedule a work item to send the current cable status for DP connectors, so that cables plugged before boot are detected.

**Analysis:**

The work_struct approach is the correct design. Calling `drm_bridge_hpd_notify()` directly from `hpd_enable` would deadlock on `bridge->hpd_mutex`. The deferred work runs in system workqueue context where the mutex is not held:

```c
static void display_connector_hpd_work(struct work_struct *work)
{
	struct display_connector *conn = container_of(work, struct display_connector, hpd_work);
	struct drm_bridge *bridge = &conn->bridge;
	drm_bridge_hpd_notify(bridge, display_connector_detect(bridge));
}
```

`display_connector_detect()` calls `gpiod_get_value_cansleep()`, which is fine from workqueue context (can sleep).

The ordering in `hpd_disable` is correct — cancel the work before disabling the IRQ:

```c
if (conn->bridge.type == DRM_MODE_CONNECTOR_DisplayPort)
	cancel_work_sync(&conn->hpd_work);
disable_irq(conn->hpd_irq);
```

`cancel_work_sync` ensures the work has either completed or been cancelled before proceeding, preventing a use-after-disable scenario.

The conditional `INIT_WORK` in probe:

```c
if (conn->hpd_irq >= 0 && type == DRM_MODE_CONNECTOR_DisplayPort)
	INIT_WORK(&conn->hpd_work, display_connector_hpd_work);
```

is consistent with the conditional usage in `hpd_enable`/`hpd_disable` — both check `conn->bridge.type == DRM_MODE_CONNECTOR_DisplayPort`.

**Minor concern — teardown safety:** `display_connector_remove()` does not call `cancel_work_sync()`. This relies on the DRM framework calling `drm_bridge_hpd_disable()` (which calls `hpd_disable` → `cancel_work_sync`) before the bridge is removed. This is the expected teardown order and is how other bridge drivers work, so it's acceptable. Adding a defensive cancel in `remove()` would be difficult anyway since the work is only initialized for DP connectors.

**Commit message typos:**

```
Note, thia issue only affects drivers which set OP_HPD but not OP_DETECT
```
"thia" → "this"

```
The HPD pin state is equalc to the display plugged state.
```
"equalc" → "equal"

**No functional issues found. Two trivial typos in the commit message.**

---
Generated by Claude Code Patch Reviewer