diff mbox series

drm/sun4i: hdmi: Improve compatibility with non-hotplug capable connectors

Message ID 20181116171830.23465-1-plaes@plaes.org (mailing list archive)
State New, archived
Headers show
Series drm/sun4i: hdmi: Improve compatibility with non-hotplug capable connectors | expand

Commit Message

Priit Laes Nov. 16, 2018, 5:18 p.m. UTC 
From: Priit Laes <priit.laes@paf.com>

Even though HDMI connector features hotplug detect pin (HPD), there are
devices that which do not support it. For these devices fall back to
additional check on I2C bus. Of course, there might be also devices
that do not wire DDC pins too, so we don't really know whether cable
has been connected.

Signed-off-by: Priit Laes <plaes@plaes.org>
Signed-off-by: Priit Laes <priit.laes@paf.com>
---
 drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

Comments

Maxime Ripard Nov. 19, 2018, 8:19 a.m. UTC | #1 
Hi,

On Fri, Nov 16, 2018 at 07:18:29PM +0200, Priit Laes wrote:
> From: Priit Laes <priit.laes@paf.com>
> 
> Even though HDMI connector features hotplug detect pin (HPD), there are
> devices that which do not support it.

Which devices?

> For these devices fall back to additional check on I2C bus. Of
> course, there might be also devices that do not wire DDC pins too,
> so we don't really know whether cable has been connected.

Again, which devices?

> 
> Signed-off-by: Priit Laes <plaes@plaes.org>
> Signed-off-by: Priit Laes <priit.laes@paf.com>

You only need one :)

> ---
>  drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 12 ++++++++----
>  1 file changed, 8 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> index 061d2e0d9011..bded09af1340 100644
> --- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> @@ -238,14 +238,18 @@ sun4i_hdmi_connector_detect(struct drm_connector *connector, bool force)
>  	struct sun4i_hdmi *hdmi = drm_connector_to_sun4i_hdmi(connector);
>  	unsigned long reg;
>  
> -	if (readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> +	if (!readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
>  			       reg & SUN4I_HDMI_HPD_HIGH,
>  			       0, 500000)) {
> -		cec_phys_addr_invalidate(hdmi->cec_adap);
> -		return connector_status_disconnected;
> +		return connector_status_connected;
>  	}
>  
> -	return connector_status_connected;
> +	if (!IS_ERR(hdmi->i2c) && drm_probe_ddc(hdmi->i2c))
> +		return connector_status_connected;
> +
> +	cec_phys_addr_invalidate(hdmi->cec_adap);
> +
> +	return connector_status_unknown;

You're doing basically two things in that patch, first adding the
fallback to the DDC probe if the hotplug mechanism couldn't detect the
display, and then returning a status unknown if both fail.

While I don't really have an opinion on the first one, it's mandatory
for every HDMI device to be able to retrieve the EDID through the
DDC. If a device was to disallow that, it would violate the HDMI, and
I'm not sure we want to start supporting those devices.

Maxime
Priit Laes Nov. 19, 2018, 8:50 a.m. UTC | #2 
On Mon, Nov 19, 2018 at 09:19:34AM +0100, Maxime Ripard wrote:
> Hi,
> 
> On Fri, Nov 16, 2018 at 07:18:29PM +0200, Priit Laes wrote:
> > From: Priit Laes <priit.laes@paf.com>
> > 
> > Even though HDMI connector features hotplug detect pin (HPD), there are
> > devices that which do not support it.
> 
> Which devices?

Device I have here is labelled "AMATIC INDUSTRIES PT-MULTI-1" and
based on the TFP401APZP chip.

> 
> > For these devices fall back to additional check on I2C bus. Of
> > course, there might be also devices that do not wire DDC pins too,
> > so we don't really know whether cable has been connected.
> 
> Again, which devices?

OK, let's skip the part without DDC. I was probably thinking about
VGA cables when I was writing that..

> > 
> > Signed-off-by: Priit Laes <plaes@plaes.org>
> > Signed-off-by: Priit Laes <priit.laes@paf.com>
> 
> You only need one :)
> 
> > ---
> >  drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 12 ++++++++----
> >  1 file changed, 8 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > index 061d2e0d9011..bded09af1340 100644
> > --- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > @@ -238,14 +238,18 @@ sun4i_hdmi_connector_detect(struct drm_connector *connector, bool force)
> >  	struct sun4i_hdmi *hdmi = drm_connector_to_sun4i_hdmi(connector);
> >  	unsigned long reg;
> >  
> > -	if (readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> > +	if (!readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> >  			       reg & SUN4I_HDMI_HPD_HIGH,
> >  			       0, 500000)) {
> > -		cec_phys_addr_invalidate(hdmi->cec_adap);
> > -		return connector_status_disconnected;
> > +		return connector_status_connected;
> >  	}
> >  
> > -	return connector_status_connected;
> > +	if (!IS_ERR(hdmi->i2c) && drm_probe_ddc(hdmi->i2c))
> > +		return connector_status_connected;
> > +
> > +	cec_phys_addr_invalidate(hdmi->cec_adap);
> > +
> > +	return connector_status_unknown;
> 
> You're doing basically two things in that patch, first adding the
> fallback to the DDC probe if the hotplug mechanism couldn't detect the
> display, and then returning a status unknown if both fail.

Agreed. 'connector_status_disconnected' is the way to go.

> While I don't really have an opinion on the first one, it's mandatory
> for every HDMI device to be able to retrieve the EDID through the
> DDC. If a device was to disallow that, it would violate the HDMI, and
> I'm not sure we want to start supporting those devices.

Yes, Even if someone runs into those non-spec devices, then there's
also possibility to use the force argument.

Thanks for review!

> 
> Maxime
> 
> -- 
> Maxime Ripard, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com
Russell King (Oracle) Nov. 19, 2018, 10:26 a.m. UTC | #3 
On Mon, Nov 19, 2018 at 09:19:34AM +0100, Maxime Ripard wrote:
> Hi,
> 
> On Fri, Nov 16, 2018 at 07:18:29PM +0200, Priit Laes wrote:
> > From: Priit Laes <priit.laes@paf.com>
> > 
> > Even though HDMI connector features hotplug detect pin (HPD), there are
> > devices that which do not support it.
> 
> Which devices?
> 
> > For these devices fall back to additional check on I2C bus. Of
> > course, there might be also devices that do not wire DDC pins too,
> > so we don't really know whether cable has been connected.
> 
> Again, which devices?
> 
> > 
> > Signed-off-by: Priit Laes <plaes@plaes.org>
> > Signed-off-by: Priit Laes <priit.laes@paf.com>
> 
> You only need one :)
> 
> > ---
> >  drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 12 ++++++++----
> >  1 file changed, 8 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > index 061d2e0d9011..bded09af1340 100644
> > --- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > @@ -238,14 +238,18 @@ sun4i_hdmi_connector_detect(struct drm_connector *connector, bool force)
> >  	struct sun4i_hdmi *hdmi = drm_connector_to_sun4i_hdmi(connector);
> >  	unsigned long reg;
> >  
> > -	if (readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> > +	if (!readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> >  			       reg & SUN4I_HDMI_HPD_HIGH,
> >  			       0, 500000)) {
> > -		cec_phys_addr_invalidate(hdmi->cec_adap);
> > -		return connector_status_disconnected;
> > +		return connector_status_connected;
> >  	}
> >  
> > -	return connector_status_connected;
> > +	if (!IS_ERR(hdmi->i2c) && drm_probe_ddc(hdmi->i2c))
> > +		return connector_status_connected;
> > +
> > +	cec_phys_addr_invalidate(hdmi->cec_adap);
> > +
> > +	return connector_status_unknown;
> 
> You're doing basically two things in that patch, first adding the
> fallback to the DDC probe if the hotplug mechanism couldn't detect the
> display, and then returning a status unknown if both fail.
> 
> While I don't really have an opinion on the first one, it's mandatory
> for every HDMI device to be able to retrieve the EDID through the
> DDC. If a device was to disallow that, it would violate the HDMI, and
> I'm not sure we want to start supporting those devices.

There is also the problem that HDMI uses the HPD signal to indicate
that the source should re-read the EDID due to the EDID changing.
In HDMI, you don't necessarily have a fixed-for-all-time EDID, but
one which can change depending on what devices are in the HDMI path.

Consider, for example, an AV amplifier which needs to subsitute the
audio capabilities when it is turned on, but when in standby needs
to pass through the TVs audio capabilities.  It informs the source
by momentarily deasserting the HDMI HPD signal, which is the HDMI
way to inform the source that the EDID should be re-read.

If you're going to use "read EDID" as the hotplug method, I think
you need to keep track of when it changes so that EDID updates are
correctly handled.
Maxime Ripard Nov. 20, 2018, 8:58 a.m. UTC | #4 
On Mon, Nov 19, 2018 at 10:26:38AM +0000, Russell King - ARM Linux wrote:
> On Mon, Nov 19, 2018 at 09:19:34AM +0100, Maxime Ripard wrote:
> > Hi,
> > 
> > On Fri, Nov 16, 2018 at 07:18:29PM +0200, Priit Laes wrote:
> > > From: Priit Laes <priit.laes@paf.com>
> > > 
> > > Even though HDMI connector features hotplug detect pin (HPD), there are
> > > devices that which do not support it.
> > 
> > Which devices?
> > 
> > > For these devices fall back to additional check on I2C bus. Of
> > > course, there might be also devices that do not wire DDC pins too,
> > > so we don't really know whether cable has been connected.
> > 
> > Again, which devices?
> > 
> > > 
> > > Signed-off-by: Priit Laes <plaes@plaes.org>
> > > Signed-off-by: Priit Laes <priit.laes@paf.com>
> > 
> > You only need one :)
> > 
> > > ---
> > >  drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 12 ++++++++----
> > >  1 file changed, 8 insertions(+), 4 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > > index 061d2e0d9011..bded09af1340 100644
> > > --- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > > +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
> > > @@ -238,14 +238,18 @@ sun4i_hdmi_connector_detect(struct drm_connector *connector, bool force)
> > >  	struct sun4i_hdmi *hdmi = drm_connector_to_sun4i_hdmi(connector);
> > >  	unsigned long reg;
> > >  
> > > -	if (readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> > > +	if (!readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
> > >  			       reg & SUN4I_HDMI_HPD_HIGH,
> > >  			       0, 500000)) {
> > > -		cec_phys_addr_invalidate(hdmi->cec_adap);
> > > -		return connector_status_disconnected;
> > > +		return connector_status_connected;
> > >  	}
> > >  
> > > -	return connector_status_connected;
> > > +	if (!IS_ERR(hdmi->i2c) && drm_probe_ddc(hdmi->i2c))
> > > +		return connector_status_connected;
> > > +
> > > +	cec_phys_addr_invalidate(hdmi->cec_adap);
> > > +
> > > +	return connector_status_unknown;
> > 
> > You're doing basically two things in that patch, first adding the
> > fallback to the DDC probe if the hotplug mechanism couldn't detect the
> > display, and then returning a status unknown if both fail.
> > 
> > While I don't really have an opinion on the first one, it's mandatory
> > for every HDMI device to be able to retrieve the EDID through the
> > DDC. If a device was to disallow that, it would violate the HDMI, and
> > I'm not sure we want to start supporting those devices.
> 
> There is also the problem that HDMI uses the HPD signal to indicate
> that the source should re-read the EDID due to the EDID changing.
> In HDMI, you don't necessarily have a fixed-for-all-time EDID, but
> one which can change depending on what devices are in the HDMI path.
> 
> Consider, for example, an AV amplifier which needs to subsitute the
> audio capabilities when it is turned on, but when in standby needs
> to pass through the TVs audio capabilities.  It informs the source
> by momentarily deasserting the HDMI HPD signal, which is the HDMI
> way to inform the source that the EDID should be re-read.
> 
> If you're going to use "read EDID" as the hotplug method, I think
> you need to keep track of when it changes so that EDID updates are
> correctly handled.

I didn't think about that, thanks for bringing it up!

Maxime
Priit Laes Dec. 5, 2018, 1:18 p.m. UTC | #5 
On Tue, Nov 20, 2018 at 09:58:41AM +0100, Maxime Ripard wrote:
> On Mon, Nov 19, 2018 at 10:26:38AM +0000, Russell King - ARM Linux wrote:
> > On Mon, Nov 19, 2018 at 09:19:34AM +0100, Maxime Ripard wrote:
> > > Hi,
> > > 
> > > On Fri, Nov 16, 2018 at 07:18:29PM +0200, Priit Laes wrote:
> > > > From: Priit Laes <priit.laes@paf.com>
> > > > 
> > > > Even though HDMI connector features hotplug detect pin (HPD), there are
> > > > devices that which do not support it.
> > > 
> > > Which devices?
> > > 
> > > > For these devices fall back to additional check on I2C bus. Of
> > > > course, there might be also devices that do not wire DDC pins too,
> > > > so we don't really know whether cable has been connected.
> > > 
> > > Again, which devices?
> > > 
> > > > 
> > > > Signed-off-by: Priit Laes <plaes@plaes.org>
> > > > Signed-off-by: Priit Laes <priit.laes@paf.com>
> > > 
> > > You only need one :)
> > > 
> > > > ---
> > > 
> > > You're doing basically two things in that patch, first adding the
> > > fallback to the DDC probe if the hotplug mechanism couldn't detect the
> > > display, and then returning a status unknown if both fail.
> > > 
> > > While I don't really have an opinion on the first one, it's mandatory
> > > for every HDMI device to be able to retrieve the EDID through the
> > > DDC. If a device was to disallow that, it would violate the HDMI, and
> > > I'm not sure we want to start supporting those devices.
> > 
> > There is also the problem that HDMI uses the HPD signal to indicate
> > that the source should re-read the EDID due to the EDID changing.
> > In HDMI, you don't necessarily have a fixed-for-all-time EDID, but
> > one which can change depending on what devices are in the HDMI path.
> > 
> > Consider, for example, an AV amplifier which needs to subsitute the
> > audio capabilities when it is turned on, but when in standby needs
> > to pass through the TVs audio capabilities.  It informs the source
> > by momentarily deasserting the HDMI HPD signal, which is the HDMI
> > way to inform the source that the EDID should be re-read.
> > 
> > If you're going to use "read EDID" as the hotplug method, I think
> > you need to keep track of when it changes so that EDID updates are
> > correctly handled.
> 
> I didn't think about that, thanks for bringing it up!

Well, currently this is broken anyway, becuse we are doing the polling
due to missing HPD interrupt. You can test by quickly switching between
two different monitors that both have proper HPD pin.

> 
> Maxime
> 
> -- 
> Maxime Ripard, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com
diff mbox series

Patch

diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
index 061d2e0d9011..bded09af1340 100644
--- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
+++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
@@ -238,14 +238,18 @@  sun4i_hdmi_connector_detect(struct drm_connector *connector, bool force)
 	struct sun4i_hdmi *hdmi = drm_connector_to_sun4i_hdmi(connector);
 	unsigned long reg;
 
-	if (readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
+	if (!readl_poll_timeout(hdmi->base + SUN4I_HDMI_HPD_REG, reg,
 			       reg & SUN4I_HDMI_HPD_HIGH,
 			       0, 500000)) {
-		cec_phys_addr_invalidate(hdmi->cec_adap);
-		return connector_status_disconnected;
+		return connector_status_connected;
 	}
 
-	return connector_status_connected;
+	if (!IS_ERR(hdmi->i2c) && drm_probe_ddc(hdmi->i2c))
+		return connector_status_connected;
+
+	cec_phys_addr_invalidate(hdmi->cec_adap);
+
+	return connector_status_unknown;
 }
 
 static const struct drm_connector_funcs sun4i_hdmi_connector_funcs = {