Hi,
On Mon, Jun 16, 2025 at 9:24 AM Doug Anderson <dianders@xxxxxxxxxxxx> wrote:
Hi,
On Mon, Jun 16, 2025 at 2:32 AM Jayesh Choudhary <j-choudhary@xxxxxx> wrote:
@@ -1220,6 +1231,27 @@ static void ti_sn65dsi86_debugfs_init(struct drm_bridge *bridge, struct dentry *
debugfs_create_file("status", 0600, debugfs, pdata, &status_fops);
}
+static void ti_sn_bridge_hpd_enable(struct drm_bridge *bridge)
+{
+ struct ti_sn65dsi86 *pdata = bridge_to_ti_sn65dsi86(bridge);
+
+ /*
+ * Device needs to be powered on before reading the HPD state
+ * for reliable hpd detection in ti_sn_bridge_detect() due to
+ * the high debounce time.
+ */
+
+ pm_runtime_get_sync(pdata->dev);
+}
+
+static void ti_sn_bridge_hpd_disable(struct drm_bridge *bridge)
+{
+ struct ti_sn65dsi86 *pdata = bridge_to_ti_sn65dsi86(bridge);
+
+ pm_runtime_mark_last_busy(pdata->dev);
+ pm_runtime_put_autosuspend(pdata->dev);
nit: you don't need the pm_runtime_mark_last_busy() here, do you? Just
call pm_runtime_put_autosuspend().
Aside from the nit, this looks reasonable to me now.
Reviewed-by: Douglas Anderson <dianders@xxxxxxxxxxxx>
What's the plan here? I can just remove the
`pm_runtime_mark_last_busy()` and land it if people are on board with
that (and if it works fine for Jayesh). If Jayesh wants to post a v6
to make it more legit, I can land that. I probably won't land anything
myself past Wednesday (California time) since I'm about to go offline
for 2 weeks and wouldn't want to land and bolt.
-Doug