linux-kernel - Re: [PATCH 3/5] drm: dw-hdmi: update CEC phys addr and EDID on HPD event

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Mon, 2 Sep 2019 10:49:34 +0000
From:   Jonas Karlman <jonas@...boo.se>
To:     Neil Armstrong <narmstrong@...libre.com>,
        Andrzej Hajda <a.hajda@...sung.com>
CC:     David Airlie <airlied@...ux.ie>, Daniel Vetter <daniel@...ll.ch>,
        Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
        Maxime Ripard <mripard@...nel.org>,
        Sean Paul <sean@...rly.run>,
        Laurent Pinchart <Laurent.pinchart@...asonboard.com>,
        Jernej Skrabec <jernej.skrabec@...l.net>,
        Hans Verkuil <hverkuil@...all.nl>,
        Jerome Brunet <jbrunet@...libre.com>,
        "dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 3/5] drm: dw-hdmi: update CEC phys addr and EDID on HPD
 event

On 2019-09-02 10:10, Neil Armstrong wrote:
> On 01/09/2019 18:14, Jonas Karlman wrote:
>> Update CEC phys addr and EDID on HPD event, fixes lost CEC phys addr and
>> stale EDID when HDMI cable is unplugged/replugged or AVR is powered on/off.
>>
>> Signed-off-by: Jonas Karlman <jonas@...boo.se>
>> ---
>>  drivers/gpu/drm/bridge/synopsys/dw-hdmi.c | 16 +++++++++-------
>>  1 file changed, 9 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> index 91d86ddd6624..0f07be1b5914 100644
>> --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> @@ -2189,6 +2189,7 @@ static int dw_hdmi_connector_update_edid(struct drm_connector *connector,
>>  static enum drm_connector_status
>>  dw_hdmi_connector_detect(struct drm_connector *connector, bool force)
>>  {
>> +	enum drm_connector_status status;
>>  	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi,
>>  					     connector);
>>  
>> @@ -2198,7 +2199,14 @@ dw_hdmi_connector_detect(struct drm_connector *connector, bool force)
>>  	dw_hdmi_update_phy_mask(hdmi);
>>  	mutex_unlock(&hdmi->mutex);
>>  
>> -	return hdmi->phy.ops->read_hpd(hdmi, hdmi->phy.data);
>> +	status = hdmi->phy.ops->read_hpd(hdmi, hdmi->phy.data);
>> +
>> +	if (status == connector_status_connected)
>> +		dw_hdmi_connector_update_edid(connector, false);
>> +	else
>> +		cec_notifier_phys_addr_invalidate(hdmi->cec_notifier);
>> +
>> +	return status;
>>  }
>>  
>>  static int dw_hdmi_connector_get_modes(struct drm_connector *connector)
>> @@ -2438,12 +2446,6 @@ static irqreturn_t dw_hdmi_irq(int irq, void *dev_id)
>>  		dw_hdmi_setup_rx_sense(hdmi,
>>  				       phy_stat & HDMI_PHY_HPD,
>>  				       phy_stat & HDMI_PHY_RX_SENSE);
>> -
>> -		if ((phy_stat & (HDMI_PHY_RX_SENSE | HDMI_PHY_HPD)) == 0) {
>> -			mutex_lock(&hdmi->cec_notifier_mutex);
>> -			cec_notifier_phys_addr_invalidate(hdmi->cec_notifier);
>> -			mutex_unlock(&hdmi->cec_notifier_mutex);
>> -		}
>>  	}
>>  
>>  	if (intr_stat & HDMI_IH_PHY_STAT0_HPD) {
>>
> Won't it possibly trigger twice edid readouts on each HPD event,
> one for CEC and one for modes ? It seems sane but not very efficient to me...

Yes there may be EDID readout twice with this change in case both detect and get_modes gets called.
Guess we could save the EDID readout in detect and reuse it in get_modes or similar?

The issue observed is that CEC phys addr never gets updated after an invalidation (HPD)
until get_modes is called, when TV/AVR triggers an EDID refresh using a 100ms HPD pulse
or an AVR is powered on/off the CEC phys addr, EDID and ELD never gets refreshed as
only detect is called and not get_modes.

Regards,
Jonas

>
> Anyway:
> Reviewed-by: Neil Armstrong <narmstrong@...libre.com>