Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit baa5a334 authored by Sandeep Panda's avatar Sandeep Panda
Browse files

drm/msm/sde: fix unwanted panic in case of ESD failure 



Currently in driver if ESD thread has detected panel failure,
then from commit thread the panel status check API returns early.
This unnecessarily causes panic, if ping pong transfer timeout has
happened because of panel failure. Fix this by adding check
for any pending ESD recovery sequence in panel status check API.

Change-Id: Ia1e9cbeef3c94eb2a0d258f53fce8b848179e10a
Signed-off-by: default avatarSandeep Panda <spanda@codeaurora.org>
parent 2b01ab76

drivers/gpu/drm/msm/sde/sde_connector.c

+10 −2
Original line number Diff line number Diff line
@@ -1848,6 +1848,7 @@ static void _sde_connector_report_panel_dead(struct sde_connector *conn) 
int sde_connector_esd_status(struct drm_connector *conn)

{

	struct sde_connector *sde_conn = NULL;

	struct dsi_display *display;

	int ret = 0;



	if (!conn)
@@ -1857,10 +1858,17 @@ int sde_connector_esd_status(struct drm_connector *conn) 
	if (!sde_conn || !sde_conn->ops.check_status)

		return ret;



	display = sde_conn->display;



	/* protect this call with ESD status check call */

	mutex_lock(&sde_conn->lock);

	ret = sde_conn->ops.check_status(&sde_conn->base, sde_conn->display,

								true);

	if (atomic_read(&(display->panel->esd_recovery_pending))) {

		SDE_ERROR("ESD recovery already pending\n");

		mutex_unlock(&sde_conn->lock);

		return -ETIMEDOUT;

	}

	ret = sde_conn->ops.check_status(&sde_conn->base,

					 sde_conn->display, true);

	mutex_unlock(&sde_conn->lock);



	if (ret <= 0) {