[4/4] drm/i915: force full detect on sink count change
diff mbox

Message ID 1440665312-26698-5-git-send-email-sivakumar.thulasimani@intel.com
State New
Headers show

Commit Message

Sivakumar Thulasimani Aug. 27, 2015, 8:48 a.m. UTC
From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>

This patch checks for changes in sink count between short pulse
hpds and forces full detect when there is a change.

This will allow both detection of hotplug and unplug of panels
through dongles that give only short pulse for such events.

v2: changed variable type from u8 to bool (Jani)
    return immediately if perform_full_detect is set(Siva)

Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
---
 drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
 1 file changed, 22 insertions(+), 5 deletions(-)

Comments

Jani Nikula Sept. 1, 2015, 10:42 a.m. UTC | #1
On Thu, 27 Aug 2015, Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> wrote:
> From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>
>
> This patch checks for changes in sink count between short pulse
> hpds and forces full detect when there is a change.
>
> This will allow both detection of hotplug and unplug of panels
> through dongles that give only short pulse for such events.
>
> v2: changed variable type from u8 to bool (Jani)
>     return immediately if perform_full_detect is set(Siva)

The general idea LGTM, however see below.

>
> Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
>  1 file changed, 22 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index 834f513..279e52c 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -4370,26 +4370,39 @@ go_again:
>   *  4. Check link status on receipt of hot-plug interrupt
>   */
>  static void
> -intel_dp_check_link_status(struct intel_dp *intel_dp)
> +intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
>  {
>  	struct drm_device *dev = intel_dp_to_dev(intel_dp);
>  	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
>  	struct intel_crtc *crtc =
>  		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
>  	u8 sink_irq_vector;
> +	u8 old_sink_count = intel_dp->sink_count;
> +	bool ret;
>  	u8 link_status[DP_LINK_STATUS_SIZE];
>  
>  	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
>  
> +	*perform_full_detect = false;
> +
>  	/* Try to read receiver status if the link appears to be up */
>  	if (!intel_dp_get_link_status(intel_dp, link_status)) {
>  		return;
>  	}
>  
> -	/* Now read the DPCD to see if it's actually running */
> -	if (!intel_dp_get_dpcd(intel_dp)) {
> +	/* Now read the DPCD to see if it's actually running
> +	 * Don't return immediately if dpcd read failed,
> +	 * if sink count was 1 and dpcd read failed we need
> +	 * to do full detection
> +	 */
> +	ret = intel_dp_get_dpcd(intel_dp);
> +
> +	if (old_sink_count != intel_dp->sink_count)
> +		*perform_full_detect = true;

The point I was trying to make earlier is that how can you trust
intel_dp->sink_count if intel_dp_get_dpcd failed? You don't know where
it failed, and ->sink_count may or may not have changed. Maybe it has
changed for real but you didn't have a chance to read it just yet.

> +
> +	/* No need to proceed if we are going to do full detect */
> +	if (!ret || *perform_full_detect)
>  		return;
> -	}

So if you're returning because of !ret, *perform_full_detect may be true
or false, you don't know. I'd like the code to be explicit about what
you want the caller to do when intel_dp_get_dpcd returns false, full
detect or not.

BR,
Jani.

>  
>  	if (!intel_encoder->base.crtc)
>  		return;
> @@ -5031,13 +5044,17 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
>  		}
>  
>  		if (!intel_dp->is_mst) {
> +			bool full_detect;
>  			/*
>  			 * we'll check the link status via the normal hot plug path later -
>  			 * but for short hpds we should check it now
>  			 */
>  			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
> -			intel_dp_check_link_status(intel_dp);
> +			intel_dp_check_link_status(intel_dp, &full_detect);
>  			drm_modeset_unlock(&dev->mode_config.connection_mutex);
> +
> +			if (full_detect)
> +				goto put_power;
>  		}
>  	}
>  
> -- 
> 1.7.9.5
>
Sivakumar Thulasimani Sept. 1, 2015, 11:39 a.m. UTC | #2
On 9/1/2015 4:12 PM, Jani Nikula wrote:
> On Thu, 27 Aug 2015, Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> wrote:
>> From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>
>>
>> This patch checks for changes in sink count between short pulse
>> hpds and forces full detect when there is a change.
>>
>> This will allow both detection of hotplug and unplug of panels
>> through dongles that give only short pulse for such events.
>>
>> v2: changed variable type from u8 to bool (Jani)
>>      return immediately if perform_full_detect is set(Siva)
> The general idea LGTM, however see below.
>
>> Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
>> ---
>>   drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
>>   1 file changed, 22 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
>> index 834f513..279e52c 100644
>> --- a/drivers/gpu/drm/i915/intel_dp.c
>> +++ b/drivers/gpu/drm/i915/intel_dp.c
>> @@ -4370,26 +4370,39 @@ go_again:
>>    *  4. Check link status on receipt of hot-plug interrupt
>>    */
>>   static void
>> -intel_dp_check_link_status(struct intel_dp *intel_dp)
>> +intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
>>   {
>>   	struct drm_device *dev = intel_dp_to_dev(intel_dp);
>>   	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
>>   	struct intel_crtc *crtc =
>>   		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
>>   	u8 sink_irq_vector;
>> +	u8 old_sink_count = intel_dp->sink_count;
>> +	bool ret;
>>   	u8 link_status[DP_LINK_STATUS_SIZE];
>>   
>>   	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
>>   
>> +	*perform_full_detect = false;
>> +
>>   	/* Try to read receiver status if the link appears to be up */
>>   	if (!intel_dp_get_link_status(intel_dp, link_status)) {
>>   		return;
>>   	}
>>   
>> -	/* Now read the DPCD to see if it's actually running */
>> -	if (!intel_dp_get_dpcd(intel_dp)) {
>> +	/* Now read the DPCD to see if it's actually running
>> +	 * Don't return immediately if dpcd read failed,
>> +	 * if sink count was 1 and dpcd read failed we need
>> +	 * to do full detection
>> +	 */
>> +	ret = intel_dp_get_dpcd(intel_dp);
>> +
>> +	if (old_sink_count != intel_dp->sink_count)
>> +		*perform_full_detect = true;
> The point I was trying to make earlier is that how can you trust
> intel_dp->sink_count if intel_dp_get_dpcd failed? You don't know where
> it failed, and ->sink_count may or may not have changed. Maybe it has
> changed for real but you didn't have a chance to read it just yet.
>
>> +
>> +	/* No need to proceed if we are going to do full detect */
>> +	if (!ret || *perform_full_detect)
>>   		return;
>> -	}
> So if you're returning because of !ret, *perform_full_detect may be true
> or false, you don't know. I'd like the code to be explicit about what
> you want the caller to do when intel_dp_get_dpcd returns false, full
> detect or not.
>
> BR,
> Jani.
the idea i had in mind was if intel_dp_get_dpcd failed we should do a 
full detect.

my assumption here was that sink_count would be zero so it would take care
of this, but now that it seems to be a wrong assumption how about the 
following code ?
this makes it explicit that if dpcd read failed we will do full detect.

+	if ((old_sink_count != intel_dp->sink_count) || !ret)
+		*perform_full_detect = true;

+
+	/* No need to proceed if we are going to do full detect */
+	if (*perform_full_detect)
  		return;

>
>>   
>>   	if (!intel_encoder->base.crtc)
>>   		return;
>> @@ -5031,13 +5044,17 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
>>   		}
>>   
>>   		if (!intel_dp->is_mst) {
>> +			bool full_detect;
>>   			/*
>>   			 * we'll check the link status via the normal hot plug path later -
>>   			 * but for short hpds we should check it now
>>   			 */
>>   			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
>> -			intel_dp_check_link_status(intel_dp);
>> +			intel_dp_check_link_status(intel_dp, &full_detect);
>>   			drm_modeset_unlock(&dev->mode_config.connection_mutex);
>> +
>> +			if (full_detect)
>> +				goto put_power;
>>   		}
>>   	}
>>   
>> -- 
>> 1.7.9.5
>>
Jani Nikula Sept. 1, 2015, 1:04 p.m. UTC | #3
On Tue, 01 Sep 2015, Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> wrote:
> On 9/1/2015 4:12 PM, Jani Nikula wrote:
>> On Thu, 27 Aug 2015, Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> wrote:
>>> From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>
>>>
>>> This patch checks for changes in sink count between short pulse
>>> hpds and forces full detect when there is a change.
>>>
>>> This will allow both detection of hotplug and unplug of panels
>>> through dongles that give only short pulse for such events.
>>>
>>> v2: changed variable type from u8 to bool (Jani)
>>>      return immediately if perform_full_detect is set(Siva)
>> The general idea LGTM, however see below.
>>
>>> Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
>>> ---
>>>   drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
>>>   1 file changed, 22 insertions(+), 5 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
>>> index 834f513..279e52c 100644
>>> --- a/drivers/gpu/drm/i915/intel_dp.c
>>> +++ b/drivers/gpu/drm/i915/intel_dp.c
>>> @@ -4370,26 +4370,39 @@ go_again:
>>>    *  4. Check link status on receipt of hot-plug interrupt
>>>    */
>>>   static void
>>> -intel_dp_check_link_status(struct intel_dp *intel_dp)
>>> +intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
>>>   {
>>>   	struct drm_device *dev = intel_dp_to_dev(intel_dp);
>>>   	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
>>>   	struct intel_crtc *crtc =
>>>   		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
>>>   	u8 sink_irq_vector;
>>> +	u8 old_sink_count = intel_dp->sink_count;
>>> +	bool ret;
>>>   	u8 link_status[DP_LINK_STATUS_SIZE];
>>>   
>>>   	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
>>>   
>>> +	*perform_full_detect = false;
>>> +
>>>   	/* Try to read receiver status if the link appears to be up */
>>>   	if (!intel_dp_get_link_status(intel_dp, link_status)) {
>>>   		return;
>>>   	}
>>>   
>>> -	/* Now read the DPCD to see if it's actually running */
>>> -	if (!intel_dp_get_dpcd(intel_dp)) {
>>> +	/* Now read the DPCD to see if it's actually running
>>> +	 * Don't return immediately if dpcd read failed,
>>> +	 * if sink count was 1 and dpcd read failed we need
>>> +	 * to do full detection
>>> +	 */
>>> +	ret = intel_dp_get_dpcd(intel_dp);
>>> +
>>> +	if (old_sink_count != intel_dp->sink_count)
>>> +		*perform_full_detect = true;
>> The point I was trying to make earlier is that how can you trust
>> intel_dp->sink_count if intel_dp_get_dpcd failed? You don't know where
>> it failed, and ->sink_count may or may not have changed. Maybe it has
>> changed for real but you didn't have a chance to read it just yet.
>>
>>> +
>>> +	/* No need to proceed if we are going to do full detect */
>>> +	if (!ret || *perform_full_detect)
>>>   		return;
>>> -	}
>> So if you're returning because of !ret, *perform_full_detect may be true
>> or false, you don't know. I'd like the code to be explicit about what
>> you want the caller to do when intel_dp_get_dpcd returns false, full
>> detect or not.
>>
>> BR,
>> Jani.
> the idea i had in mind was if intel_dp_get_dpcd failed we should do a 
> full detect.
>
> my assumption here was that sink_count would be zero so it would take care
> of this, but now that it seems to be a wrong assumption how about the 
> following code ?
> this makes it explicit that if dpcd read failed we will do full detect.
>
> +	if ((old_sink_count != intel_dp->sink_count) || !ret)
> +		*perform_full_detect = true;
>
> +
> +	/* No need to proceed if we are going to do full detect */
> +	if (*perform_full_detect)
>   		return;

Ok by me.

Jani.



>
>>
>>>   
>>>   	if (!intel_encoder->base.crtc)
>>>   		return;
>>> @@ -5031,13 +5044,17 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
>>>   		}
>>>   
>>>   		if (!intel_dp->is_mst) {
>>> +			bool full_detect;
>>>   			/*
>>>   			 * we'll check the link status via the normal hot plug path later -
>>>   			 * but for short hpds we should check it now
>>>   			 */
>>>   			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
>>> -			intel_dp_check_link_status(intel_dp);
>>> +			intel_dp_check_link_status(intel_dp, &full_detect);
>>>   			drm_modeset_unlock(&dev->mode_config.connection_mutex);
>>> +
>>> +			if (full_detect)
>>> +				goto put_power;
>>>   		}
>>>   	}
>>>   
>>> -- 
>>> 1.7.9.5
>>>
>
> -- 
> regards,
> Sivakumar
>
Daniel Vetter Sept. 2, 2015, 9:13 a.m. UTC | #4
On Thu, Aug 27, 2015 at 02:18:32PM +0530, Sivakumar Thulasimani wrote:
> From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>
> 
> This patch checks for changes in sink count between short pulse
> hpds and forces full detect when there is a change.
> 
> This will allow both detection of hotplug and unplug of panels
> through dongles that give only short pulse for such events.
> 
> v2: changed variable type from u8 to bool (Jani)
>     return immediately if perform_full_detect is set(Siva)
> 
> Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>

That's still incomplete since in the hotplug work function we check
whether ->detect status has changed. If that doesn't happen (e.g. sink
count changes from 1->2 or whatever) then we'll fail to generate the
required uevent.

I suspect the only clean way to fix this is to merge all the dp hpd
handling together into one place and stop using ->detect for some parts of
it. Then we would have one place only that decides whether anything
changed, and whether we need to send anything to userspace or not.
-Daniel

> ---
>  drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
>  1 file changed, 22 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index 834f513..279e52c 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -4370,26 +4370,39 @@ go_again:
>   *  4. Check link status on receipt of hot-plug interrupt
>   */
>  static void
> -intel_dp_check_link_status(struct intel_dp *intel_dp)
> +intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
>  {
>  	struct drm_device *dev = intel_dp_to_dev(intel_dp);
>  	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
>  	struct intel_crtc *crtc =
>  		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
>  	u8 sink_irq_vector;
> +	u8 old_sink_count = intel_dp->sink_count;
> +	bool ret;
>  	u8 link_status[DP_LINK_STATUS_SIZE];
>  
>  	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
>  
> +	*perform_full_detect = false;
> +
>  	/* Try to read receiver status if the link appears to be up */
>  	if (!intel_dp_get_link_status(intel_dp, link_status)) {
>  		return;
>  	}
>  
> -	/* Now read the DPCD to see if it's actually running */
> -	if (!intel_dp_get_dpcd(intel_dp)) {
> +	/* Now read the DPCD to see if it's actually running
> +	 * Don't return immediately if dpcd read failed,
> +	 * if sink count was 1 and dpcd read failed we need
> +	 * to do full detection
> +	 */
> +	ret = intel_dp_get_dpcd(intel_dp);
> +
> +	if (old_sink_count != intel_dp->sink_count)
> +		*perform_full_detect = true;
> +
> +	/* No need to proceed if we are going to do full detect */
> +	if (!ret || *perform_full_detect)
>  		return;
> -	}
>  
>  	if (!intel_encoder->base.crtc)
>  		return;
> @@ -5031,13 +5044,17 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
>  		}
>  
>  		if (!intel_dp->is_mst) {
> +			bool full_detect;
>  			/*
>  			 * we'll check the link status via the normal hot plug path later -
>  			 * but for short hpds we should check it now
>  			 */
>  			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
> -			intel_dp_check_link_status(intel_dp);
> +			intel_dp_check_link_status(intel_dp, &full_detect);
>  			drm_modeset_unlock(&dev->mode_config.connection_mutex);
> +
> +			if (full_detect)
> +				goto put_power;
>  		}
>  	}
>  
> -- 
> 1.7.9.5
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Sivakumar Thulasimani Sept. 3, 2015, 8:24 a.m. UTC | #5
On 9/2/2015 2:43 PM, Daniel Vetter wrote:
> On Thu, Aug 27, 2015 at 02:18:32PM +0530, Sivakumar Thulasimani wrote:
>> From: "Thulasimani,Sivakumar" <sivakumar.thulasimani@intel.com>
>>
>> This patch checks for changes in sink count between short pulse
>> hpds and forces full detect when there is a change.
>>
>> This will allow both detection of hotplug and unplug of panels
>> through dongles that give only short pulse for such events.
>>
>> v2: changed variable type from u8 to bool (Jani)
>>      return immediately if perform_full_detect is set(Siva)
>>
>> Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
> That's still incomplete since in the hotplug work function we check
> whether ->detect status has changed. If that doesn't happen (e.g. sink
> count changes from 1->2 or whatever) then we'll fail to generate the
> required uevent.
>
> I suspect the only clean way to fix this is to merge all the dp hpd
> handling together into one place and stop using ->detect for some parts of
> it. Then we would have one place only that decides whether anything
> changed, and whether we need to send anything to userspace or not.
> -Daniel
i doubt sink count change from 1->2 is valid.
1)  we support only edp and SST displays in intel_dp_detect.
we should handle all mst related code in intel_dp_hpd_pulse.
2) check_link_status is called from withing !mst check so i dont see it 
change from 1 to 2.
3) even if it changes from 1 to 2, for the given port without mst there 
is no way to
switch between first sink to second one outside mst logic.

correct me if i am wrong, the only reason to call intel_dp_detect is to 
see if
"one display is connected or disconnected" which i feel is fine with 
current code.

But i do agree that we need to clean up intel_dp_hpd_pulse since it does 
a lot of things
then repeats those same operations in intel_dp_detect. for example we 
read dpcd 0-14,
link status and probe oui only to repeat the same steps in the 
intel_dp_detect. considering
that our dpcd read has 1ms delay this can add up pretty quick.

if the intention of intel_dp_hpd_pulse is to handle mst displays, that 
should be checked first
and continue only if display is mst otherwise we should just jump to 
intel_dp_detect.
>> ---
>>   drivers/gpu/drm/i915/intel_dp.c |   27 ++++++++++++++++++++++-----
>>   1 file changed, 22 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
>> index 834f513..279e52c 100644
>> --- a/drivers/gpu/drm/i915/intel_dp.c
>> +++ b/drivers/gpu/drm/i915/intel_dp.c
>> @@ -4370,26 +4370,39 @@ go_again:
>>    *  4. Check link status on receipt of hot-plug interrupt
>>    */
>>   static void
>> -intel_dp_check_link_status(struct intel_dp *intel_dp)
>> +intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
>>   {
>>   	struct drm_device *dev = intel_dp_to_dev(intel_dp);
>>   	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
>>   	struct intel_crtc *crtc =
>>   		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
>>   	u8 sink_irq_vector;
>> +	u8 old_sink_count = intel_dp->sink_count;
>> +	bool ret;
>>   	u8 link_status[DP_LINK_STATUS_SIZE];
>>   
>>   	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
>>   
>> +	*perform_full_detect = false;
>> +
>>   	/* Try to read receiver status if the link appears to be up */
>>   	if (!intel_dp_get_link_status(intel_dp, link_status)) {
>>   		return;
>>   	}
>>   
>> -	/* Now read the DPCD to see if it's actually running */
>> -	if (!intel_dp_get_dpcd(intel_dp)) {
>> +	/* Now read the DPCD to see if it's actually running
>> +	 * Don't return immediately if dpcd read failed,
>> +	 * if sink count was 1 and dpcd read failed we need
>> +	 * to do full detection
>> +	 */
>> +	ret = intel_dp_get_dpcd(intel_dp);
>> +
>> +	if (old_sink_count != intel_dp->sink_count)
>> +		*perform_full_detect = true;
>> +
>> +	/* No need to proceed if we are going to do full detect */
>> +	if (!ret || *perform_full_detect)
>>   		return;
>> -	}
>>   
>>   	if (!intel_encoder->base.crtc)
>>   		return;
>> @@ -5031,13 +5044,17 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
>>   		}
>>   
>>   		if (!intel_dp->is_mst) {
>> +			bool full_detect;
>>   			/*
>>   			 * we'll check the link status via the normal hot plug path later -
>>   			 * but for short hpds we should check it now
>>   			 */
>>   			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
>> -			intel_dp_check_link_status(intel_dp);
>> +			intel_dp_check_link_status(intel_dp, &full_detect);
>>   			drm_modeset_unlock(&dev->mode_config.connection_mutex);
>> +
>> +			if (full_detect)
>> +				goto put_power;
>>   		}
>>   	}
>>   
>> -- 
>> 1.7.9.5
>>
>> _______________________________________________
>> Intel-gfx mailing list
>> Intel-gfx@lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Patch
diff mbox

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 834f513..279e52c 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -4370,26 +4370,39 @@  go_again:
  *  4. Check link status on receipt of hot-plug interrupt
  */
 static void
-intel_dp_check_link_status(struct intel_dp *intel_dp)
+intel_dp_check_link_status(struct intel_dp *intel_dp, bool *perform_full_detect)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
 	struct intel_crtc *crtc =
 		to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
 	u8 sink_irq_vector;
+	u8 old_sink_count = intel_dp->sink_count;
+	bool ret;
 	u8 link_status[DP_LINK_STATUS_SIZE];
 
 	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 
+	*perform_full_detect = false;
+
 	/* Try to read receiver status if the link appears to be up */
 	if (!intel_dp_get_link_status(intel_dp, link_status)) {
 		return;
 	}
 
-	/* Now read the DPCD to see if it's actually running */
-	if (!intel_dp_get_dpcd(intel_dp)) {
+	/* Now read the DPCD to see if it's actually running
+	 * Don't return immediately if dpcd read failed,
+	 * if sink count was 1 and dpcd read failed we need
+	 * to do full detection
+	 */
+	ret = intel_dp_get_dpcd(intel_dp);
+
+	if (old_sink_count != intel_dp->sink_count)
+		*perform_full_detect = true;
+
+	/* No need to proceed if we are going to do full detect */
+	if (!ret || *perform_full_detect)
 		return;
-	}
 
 	if (!intel_encoder->base.crtc)
 		return;
@@ -5031,13 +5044,17 @@  intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
 		}
 
 		if (!intel_dp->is_mst) {
+			bool full_detect;
 			/*
 			 * we'll check the link status via the normal hot plug path later -
 			 * but for short hpds we should check it now
 			 */
 			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-			intel_dp_check_link_status(intel_dp);
+			intel_dp_check_link_status(intel_dp, &full_detect);
 			drm_modeset_unlock(&dev->mode_config.connection_mutex);
+
+			if (full_detect)
+				goto put_power;
 		}
 	}