Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-07-10 Thread Manasi Navare 
Matt, is the new revision already submitted to the M-L?

Manasi

On Tue, Jun 26, 2018 at 07:54:19AM -0700, Atwood, Matthew S wrote:
> ill uprev by EOD
> 
> From: Navare, Manasi D
> Sent: Monday, June 25, 2018 5:39 PM
> To: Atwood, Matthew S
> Cc: intel-gfx@lists.freedesktop.org; Vivi, Rodrigo; ble...@chromium.org; 
> Zanoni, Paulo R
> Subject: Re: [Intel-gfx] [PATCH] drm/dp: implement 
> EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT
> 
> What is the status of this patch?
> I am hitting this issue with one of the DP 1.4 sinks where the correct
> values of DPCD are present at this new offset of 0x2200 else I see
> bad values for REV and Max_link_rate etc..
> So we absolutely need this for DP 1.4 sinks.
> 
> Manasi
> 
> On Wed, May 16, 2018 at 09:33:42AM -0700, matthew.s.atw...@intel.com wrote:
> > From: Matt Atwood 
> >
> > According to DP spec (2.9.3.1 of DP 1.4) if
> > EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in DPCD
> > 02200h through 0220Fh shall contain the DPRX's true capability. These
> > values will match 0h through Fh, except for DPCD_REV,
> > MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
> >
> > Read from DPCD once for all 3 values as this is an expensive operation.
> > Spec mentions that all of address space 02200h through 0220Fh should
> > contain the right information however currently only 3 values can
> > differ.
> >
> > There is no address space in the intel_dp->dpcd struct for addresses
> > 02200h through 0220Fh, and since so much of the data is a identical,
> > simply overwrite the values stored in 0h through Fh with the
> > values that can be overwritten from addresses 02200h through 0220Fh
> >
> > Signed-off-by: Matt Atwood 
> > ---
> >  drivers/gpu/drm/i915/intel_dp.c | 14 ++
> >  include/drm/drm_dp_helper.h |  5 +++--
> >  2 files changed, 17 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/intel_dp.c 
> > b/drivers/gpu/drm/i915/intel_dp.c
> > index dde92e4af5d3..899ebc5cece6 100644
> > --- a/drivers/gpu/drm/i915/intel_dp.c
> > +++ b/drivers/gpu/drm/i915/intel_dp.c
> > @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
> >sizeof(intel_dp->dpcd)) < 0)
> >   return false; /* aux transfer failed */
> >
> > + if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> > + DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> > + intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
> > + uint8_t dpcd_ext[16];
> > +
> > + if (drm_dp_dpcd_read(_dp->aux, DP_DP13_DPCD_REV,
> > + _ext, sizeof(dpcd_ext)) < 0)
> > + return false; /* aux transfer failed */
> > +
> > + intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
> > + intel_dp->dpcd[DP_MAX_LINK_RATE] = dpcd_ext[DP_MAX_LINK_RATE];
> > + intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> > + dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> > + }
> >   DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd), 
> > intel_dp->dpcd);
> >
> >   return intel_dp->dpcd[DP_DPCD_REV] != 0;
> > diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h
> > index c01564991a9f..757bd5913f3d 100644
> > --- a/include/drm/drm_dp_helper.h
> > +++ b/include/drm/drm_dp_helper.h
> > @@ -123,8 +123,9 @@
> >  # define DP_FRAMING_CHANGE_CAP   (1 << 1)
> >  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or 
> > higher */
> >
> > -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> > -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> > +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> > +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> > +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2? */
> >
> >  #define DP_ADAPTER_CAP   0x00f   /* 1.2 */
> >  # define DP_FORCE_LOAD_SENSE_CAP (1 << 0)
> > --
> > 2.17.0
> >
> > ___
> > Intel-gfx mailing list
> > Intel-gfx@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> 
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-06-26 Thread Atwood, Matthew S 
ill uprev by EOD

From: Navare, Manasi D
Sent: Monday, June 25, 2018 5:39 PM
To: Atwood, Matthew S
Cc: intel-gfx@lists.freedesktop.org; Vivi, Rodrigo; ble...@chromium.org; 
Zanoni, Paulo R
Subject: Re: [Intel-gfx] [PATCH] drm/dp: implement 
EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

What is the status of this patch?
I am hitting this issue with one of the DP 1.4 sinks where the correct
values of DPCD are present at this new offset of 0x2200 else I see
bad values for REV and Max_link_rate etc..
So we absolutely need this for DP 1.4 sinks.

Manasi

On Wed, May 16, 2018 at 09:33:42AM -0700, matthew.s.atw...@intel.com wrote:
> From: Matt Atwood 
>
> According to DP spec (2.9.3.1 of DP 1.4) if
> EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in DPCD
> 02200h through 0220Fh shall contain the DPRX's true capability. These
> values will match 0h through Fh, except for DPCD_REV,
> MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
>
> Read from DPCD once for all 3 values as this is an expensive operation.
> Spec mentions that all of address space 02200h through 0220Fh should
> contain the right information however currently only 3 values can
> differ.
>
> There is no address space in the intel_dp->dpcd struct for addresses
> 02200h through 0220Fh, and since so much of the data is a identical,
> simply overwrite the values stored in 0h through Fh with the
> values that can be overwritten from addresses 02200h through 0220Fh
>
> Signed-off-by: Matt Atwood 
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 14 ++
>  include/drm/drm_dp_helper.h |  5 +++--
>  2 files changed, 17 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index dde92e4af5d3..899ebc5cece6 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
>sizeof(intel_dp->dpcd)) < 0)
>   return false; /* aux transfer failed */
>
> + if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> + DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> + intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
> + uint8_t dpcd_ext[16];
> +
> + if (drm_dp_dpcd_read(_dp->aux, DP_DP13_DPCD_REV,
> + _ext, sizeof(dpcd_ext)) < 0)
> + return false; /* aux transfer failed */
> +
> + intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
> + intel_dp->dpcd[DP_MAX_LINK_RATE] = dpcd_ext[DP_MAX_LINK_RATE];
> + intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> + dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> + }
>   DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd), 
> intel_dp->dpcd);
>
>   return intel_dp->dpcd[DP_DPCD_REV] != 0;
> diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h
> index c01564991a9f..757bd5913f3d 100644
> --- a/include/drm/drm_dp_helper.h
> +++ b/include/drm/drm_dp_helper.h
> @@ -123,8 +123,9 @@
>  # define DP_FRAMING_CHANGE_CAP   (1 << 1)
>  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or higher 
> */
>
> -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2? */
>
>  #define DP_ADAPTER_CAP   0x00f   /* 1.2 */
>  # define DP_FORCE_LOAD_SENSE_CAP (1 << 0)
> --
> 2.17.0
>
> ___
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-06-25 Thread Manasi Navare 
What is the status of this patch?
I am hitting this issue with one of the DP 1.4 sinks where the correct
values of DPCD are present at this new offset of 0x2200 else I see
bad values for REV and Max_link_rate etc..
So we absolutely need this for DP 1.4 sinks.

Manasi

On Wed, May 16, 2018 at 09:33:42AM -0700, matthew.s.atw...@intel.com wrote:
> From: Matt Atwood 
> 
> According to DP spec (2.9.3.1 of DP 1.4) if
> EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in DPCD
> 02200h through 0220Fh shall contain the DPRX's true capability. These
> values will match 0h through Fh, except for DPCD_REV,
> MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
> 
> Read from DPCD once for all 3 values as this is an expensive operation.
> Spec mentions that all of address space 02200h through 0220Fh should
> contain the right information however currently only 3 values can
> differ.
> 
> There is no address space in the intel_dp->dpcd struct for addresses
> 02200h through 0220Fh, and since so much of the data is a identical,
> simply overwrite the values stored in 0h through Fh with the
> values that can be overwritten from addresses 02200h through 0220Fh
> 
> Signed-off-by: Matt Atwood 
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 14 ++
>  include/drm/drm_dp_helper.h |  5 +++--
>  2 files changed, 17 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index dde92e4af5d3..899ebc5cece6 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
>sizeof(intel_dp->dpcd)) < 0)
>   return false; /* aux transfer failed */
>  
> + if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> + DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> + intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
> + uint8_t dpcd_ext[16];
> +
> + if (drm_dp_dpcd_read(_dp->aux, DP_DP13_DPCD_REV,
> + _ext, sizeof(dpcd_ext)) < 0)
> + return false; /* aux transfer failed */
> +
> + intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
> + intel_dp->dpcd[DP_MAX_LINK_RATE] = dpcd_ext[DP_MAX_LINK_RATE];
> + intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> + dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> + }
>   DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd), 
> intel_dp->dpcd);
>  
>   return intel_dp->dpcd[DP_DPCD_REV] != 0;
> diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h
> index c01564991a9f..757bd5913f3d 100644
> --- a/include/drm/drm_dp_helper.h
> +++ b/include/drm/drm_dp_helper.h
> @@ -123,8 +123,9 @@
>  # define DP_FRAMING_CHANGE_CAP   (1 << 1)
>  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or higher 
> */
>  
> -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2? */
>  
>  #define DP_ADAPTER_CAP   0x00f   /* 1.2 */
>  # define DP_FORCE_LOAD_SENSE_CAP (1 << 0)
> -- 
> 2.17.0
> 
> ___
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-05-17 Thread Jani Nikula 
On Thu, 17 May 2018, "Atwood, Matthew S"  wrote:
> On Thu, 2018-05-17 at 12:50 +0300, Jani Nikula wrote:
>> On Wed, 16 May 2018, Dhinakaran Pandiyan > om> wrote:
>> > Why overwrite all values if this is an expensive operation? From
>> > what I
>> > can see, you'll need to read only h - 5h
> was mostly future proofing, we can get away with only reading 6 values.
> the expense is to read 1, any number after that doesnt cost alot. That
> being said sure thing.

With the dpcd read, memcmp, memcpy, and debug logging written based on
sizeof(dpcd_ext), it'll be trivial to just adjust the size of the local
array if needed.

>> Surely this is not XXX 1.2? ;)
> I found it in a dp1.2 spec that Rodrigo had, I had originally found it
> as a change with dp1.3. Earlier versions of the patch that added
> DP_TRAINING_AUX_RD_MASK has had dp1.3 until he showed me that. If you'd
> like Ill change it.

I'm not looking it up now, but please just update the XXX as best you
can.

BR,
Jani.


-- 
Jani Nikula, Intel Open Source Graphics Center
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-05-17 Thread Atwood, Matthew S 
On Thu, 2018-05-17 at 12:50 +0300, Jani Nikula wrote:
> On Wed, 16 May 2018, Dhinakaran Pandiyan  om> wrote:
> > On Wed, 2018-05-16 at 09:33 -0700, matthew.s.atw...@intel.com
> > wrote:
> > > From: Matt Atwood 
> > > 
> > > According to DP spec (2.9.3.1 of DP 1.4) if
> > > EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses
> > > in
> > > DPCD
> > > 02200h through 0220Fh shall contain the DPRX's true capability.
> > > These
> > > values will match 0h through Fh, except for DPCD_REV,
> > > MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
> > > 
> > > Read from DPCD once for all 3 values as this is an expensive
> > > operation.
> > > Spec mentions that all of address space 02200h through 0220Fh
> > > should
> > > contain the right information however currently only 3 values can
> > > differ.
> > > 
> > > There is no address space in the intel_dp->dpcd struct for
> > > addresses
> > > 02200h through 0220Fh, and since so much of the data is a
> > > identical,
> > > simply overwrite the values stored in 0h through Fh with
> > > the
> > > values that can be overwritten from addresses 02200h through
> > > 0220Fh
> 
> Without reading the spec, this commit message makes one think there's
> no
> point in any of this. Please mention this is for backward
> compatibility
> with older source devices that trip over because of newer DPCD.
sure thing
> 
> > Why overwrite all values if this is an expensive operation? From
> > what I
> > can see, you'll need to read only h - 5h
was mostly future proofing, we can get away with only reading 6 values.
the expense is to read 1, any number after that doesnt cost alot. That
being said sure thing.
> > 
> > > 
> > > Signed-off-by: Matt Atwood 
> > > ---
> > >  drivers/gpu/drm/i915/intel_dp.c | 14 ++
> > >  include/drm/drm_dp_helper.h |  5 +++--
> > >  2 files changed, 17 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/intel_dp.c
> > > b/drivers/gpu/drm/i915/intel_dp.c
> > > index dde92e4af5d3..899ebc5cece6 100644
> > > --- a/drivers/gpu/drm/i915/intel_dp.c
> > > +++ b/drivers/gpu/drm/i915/intel_dp.c
> > > @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp
> > > *intel_dp)
> > >sizeof(intel_dp->dpcd)) < 0)
> > >   return false; /* aux transfer failed */
> > >  
> > > + if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> > > + DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> > > + intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
> > 
> > The same section in the spec also says - "The Extended Receiver
> > Capability registers at DPCD Addresses 02200h through 0220Fh shall
> > contain the DPRX’s true capability, while the original Base
> > Receiver
> > Capability registers at DPCD Addresses 0h through Fh might
> > indicate DPCD r1.1, a MAX_LINK_RATE of 2.7Gbps/lane, and no DFP to
> > avoid interoperability issues ..."
> > 
> > Which means, intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14 is
> > probably
> > not going to be true for panels you want to read the extended
> > capabilities for. 
> 
> Agreed. Only check DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT.
sure
> 
> > 
> > > + uint8_t dpcd_ext[16];
> 
> u8.
> 
> > > +
> > > + if (drm_dp_dpcd_read(_dp->aux,
> > > DP_DP13_DPCD_REV,
> > > + _ext, sizeof(dpcd_ext)) < 0)
> > > + return false; /* aux transfer failed */
> 
> Please don't return false here. Just waltz on.
sure
> 
> Like DK said, please read the minimal amount. Please memcmp those six
> bytes against the already read DPCD, and debug log the *old* bytes if
> there's a diff (new bytes debug logged below), and memcpy the new
> parts
> in place.
I really like this idea. thanks!
> 
> > > +
> > > + intel_dp->dpcd[DP_DPCD_REV] =
> > > dpcd_ext[DP_DPCD_REV];
> > > + intel_dp->dpcd[DP_MAX_LINK_RATE] =
> > > dpcd_ext[DP_MAX_LINK_RATE];
> > > + intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> > > + dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> > > + }
> > >   DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp-
> > > >dpcd),
> > > intel_dp->dpcd);
> > >  
> > >   return intel_dp->dpcd[DP_DPCD_REV] != 0;
> > > diff --git a/include/drm/drm_dp_helper.h
> > > b/include/drm/drm_dp_helper.h
> > > index c01564991a9f..757bd5913f3d 100644
> > > --- a/include/drm/drm_dp_helper.h
> > > +++ b/include/drm/drm_dp_helper.h
> > 
> > This should be a separate patch as it's outside i915.
no problem
> 
> Yes.
> 
> > 
> > > @@ -123,8 +123,9 @@
> > >  # define DP_FRAMING_CHANGE_CAP   (1 << 1)
> > >  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp
> > > v1.2 or
> > > higher */
> > >  
> > > -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2?
> > > */
> > > -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2?
> > > */
> > > +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX
> >

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-05-17 Thread Jani Nikula 
On Wed, 16 May 2018, Dhinakaran Pandiyan  wrote:
> On Wed, 2018-05-16 at 09:33 -0700, matthew.s.atw...@intel.com wrote:
>> From: Matt Atwood 
>> 
>> According to DP spec (2.9.3.1 of DP 1.4) if
>> EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in
>> DPCD
>> 02200h through 0220Fh shall contain the DPRX's true capability. These
>> values will match 0h through Fh, except for DPCD_REV,
>> MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
>> 
>> Read from DPCD once for all 3 values as this is an expensive
>> operation.
>> Spec mentions that all of address space 02200h through 0220Fh should
>> contain the right information however currently only 3 values can
>> differ.
>> 
>> There is no address space in the intel_dp->dpcd struct for addresses
>> 02200h through 0220Fh, and since so much of the data is a identical,
>> simply overwrite the values stored in 0h through Fh with the
>> values that can be overwritten from addresses 02200h through 0220Fh

Without reading the spec, this commit message makes one think there's no
point in any of this. Please mention this is for backward compatibility
with older source devices that trip over because of newer DPCD.

> Why overwrite all values if this is an expensive operation? From what I
> can see, you'll need to read only h - 5h
>
>> 
>> Signed-off-by: Matt Atwood 
>> ---
>>  drivers/gpu/drm/i915/intel_dp.c | 14 ++
>>  include/drm/drm_dp_helper.h |  5 +++--
>>  2 files changed, 17 insertions(+), 2 deletions(-)
>> 
>> diff --git a/drivers/gpu/drm/i915/intel_dp.c
>> b/drivers/gpu/drm/i915/intel_dp.c
>> index dde92e4af5d3..899ebc5cece6 100644
>> --- a/drivers/gpu/drm/i915/intel_dp.c
>> +++ b/drivers/gpu/drm/i915/intel_dp.c
>> @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
>>   sizeof(intel_dp->dpcd)) < 0)
>>  return false; /* aux transfer failed */
>>  
>> +if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
>> +DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
>> +intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
>
> The same section in the spec also says - "The Extended Receiver
> Capability registers at DPCD Addresses 02200h through 0220Fh shall
> contain the DPRX’s true capability, while the original Base Receiver
> Capability registers at DPCD Addresses 0h through Fh might
> indicate DPCD r1.1, a MAX_LINK_RATE of 2.7Gbps/lane, and no DFP to
> avoid interoperability issues ..."
>
> Which means, intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14 is probably
> not going to be true for panels you want to read the extended
> capabilities for. 

Agreed. Only check DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT.

>
>> +uint8_t dpcd_ext[16];

u8.

>> +
>> +if (drm_dp_dpcd_read(_dp->aux,
>> DP_DP13_DPCD_REV,
>> +_ext, sizeof(dpcd_ext)) < 0)
>> +return false; /* aux transfer failed */

Please don't return false here. Just waltz on.

Like DK said, please read the minimal amount. Please memcmp those six
bytes against the already read DPCD, and debug log the *old* bytes if
there's a diff (new bytes debug logged below), and memcpy the new parts
in place.

>> +
>> +intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
>> +intel_dp->dpcd[DP_MAX_LINK_RATE] =
>> dpcd_ext[DP_MAX_LINK_RATE];
>> +intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
>> +dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
>> +}
>>  DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd),
>> intel_dp->dpcd);
>>  
>>  return intel_dp->dpcd[DP_DPCD_REV] != 0;
>> diff --git a/include/drm/drm_dp_helper.h
>> b/include/drm/drm_dp_helper.h
>> index c01564991a9f..757bd5913f3d 100644
>> --- a/include/drm/drm_dp_helper.h
>> +++ b/include/drm/drm_dp_helper.h
>
> This should be a separate patch as it's outside i915.

Yes.

>
>> @@ -123,8 +123,9 @@
>>  # define DP_FRAMING_CHANGE_CAP  (1 << 1)
>>  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or
>> higher */
>>  
>> -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
>> -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
>> +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2?
>> */
>> +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2?
>> */
>> +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2?

Surely this is not XXX 1.2? ;)

BR,
Jani.

>> */
>>  
>>  #define DP_ADAPTER_CAP  0x00f   /* 1.2 */
>>  # define DP_FORCE_LOAD_SENSE_CAP(1 << 0)
> ___
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Jani Nikula, Intel Open Source Technology Center
___
Intel-gfx mailing list

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-05-16 Thread Manasi Navare 
On Wed, May 16, 2018 at 03:31:52PM -0700, Dhinakaran Pandiyan wrote:
> On Wed, 2018-05-16 at 09:33 -0700, matthew.s.atw...@intel.com wrote:
> > From: Matt Atwood 
> > 
> > According to DP spec (2.9.3.1 of DP 1.4) if
> > EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in
> > DPCD
> > 02200h through 0220Fh shall contain the DPRX's true capability. These
> > values will match 0h through Fh, except for DPCD_REV,
> > MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
> > 
> > Read from DPCD once for all 3 values as this is an expensive
> > operation.
> > Spec mentions that all of address space 02200h through 0220Fh should
> > contain the right information however currently only 3 values can
> > differ.
> > 
> > There is no address space in the intel_dp->dpcd struct for addresses
> > 02200h through 0220Fh, and since so much of the data is a identical,
> > simply overwrite the values stored in 0h through Fh with the
> > values that can be overwritten from addresses 02200h through 0220Fh
> 
> Why overwrite all values if this is an expensive operation? From what I
> can see, you'll need to read only h - 5h
> 
> > 
> > Signed-off-by: Matt Atwood 
> > ---
> >  drivers/gpu/drm/i915/intel_dp.c | 14 ++
> >  include/drm/drm_dp_helper.h |  5 +++--
> >  2 files changed, 17 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/intel_dp.c
> > b/drivers/gpu/drm/i915/intel_dp.c
> > index dde92e4af5d3..899ebc5cece6 100644
> > --- a/drivers/gpu/drm/i915/intel_dp.c
> > +++ b/drivers/gpu/drm/i915/intel_dp.c
> > @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
> >      sizeof(intel_dp->dpcd)) < 0)
> >     return false; /* aux transfer failed */
> >  
> > +   if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> > +   DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> > +   intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {
> 
> The same section in the spec also says - "The Extended Receiver
> Capability registers at DPCD Addresses 02200h through 0220Fh shall
> contain the DPRX’s true capability, while the original Base Receiver
> Capability registers at DPCD Addresses 0h through Fh might
> indicate DPCD r1.1, a MAX_LINK_RATE of 2.7Gbps/lane, and no DFP to
> avoid interoperability issues ..."
> 
> Which means, intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14 is probably
> not going to be true for panels you want to read the extended
> capabilities for.

Yes thats a good point. So just check on the 
EXTENDED_EXTENDED_RECEIVER_CAP_FIELD_PRESENT
should suffice.
 
> 
> > +   uint8_t dpcd_ext[16];
> > +
> > +   if (drm_dp_dpcd_read(_dp->aux,
> > DP_DP13_DPCD_REV,
> > +   _ext, sizeof(dpcd_ext)) < 0)
> > +   return false; /* aux transfer failed */
> > +
> > +   intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
> > +   intel_dp->dpcd[DP_MAX_LINK_RATE] =
> > dpcd_ext[DP_MAX_LINK_RATE];
> > +   intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> > +   dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> > +   }
> >     DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd),
> > intel_dp->dpcd);
> >  
> >     return intel_dp->dpcd[DP_DPCD_REV] != 0;
> > diff --git a/include/drm/drm_dp_helper.h
> > b/include/drm/drm_dp_helper.h
> > index c01564991a9f..757bd5913f3d 100644
> > --- a/include/drm/drm_dp_helper.h
> > +++ b/include/drm/drm_dp_helper.h
> 
> This should be a separate patch as it's outside i915.

And also send this patch to dri-devel M-L

Manasi

> 
> > @@ -123,8 +123,9 @@
> >  # define DP_FRAMING_CHANGE_CAP (1 << 1)
> >  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or
> > higher */
> >  
> > -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> > -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> > +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2?
> > */
> > +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2?
> > */
> > +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2?
> > */
> >  
> >  #define DP_ADAPTER_CAP 0x00f   /* 1.2 */
> >  # define DP_FORCE_LOAD_SENSE_CAP   (1 << 0)
> ___
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH] drm/dp: implement EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT

2018-05-16 Thread Dhinakaran Pandiyan 
On Wed, 2018-05-16 at 09:33 -0700, matthew.s.atw...@intel.com wrote:
> From: Matt Atwood 
> 
> According to DP spec (2.9.3.1 of DP 1.4) if
> EXTENDED_RECEIVER_CAPABILITY_FIELD_PRESENT is set the addresses in
> DPCD
> 02200h through 0220Fh shall contain the DPRX's true capability. These
> values will match 0h through Fh, except for DPCD_REV,
> MAX_LINK_RATE, DOWN_STREAM_PORT_PRESENT.
> 
> Read from DPCD once for all 3 values as this is an expensive
> operation.
> Spec mentions that all of address space 02200h through 0220Fh should
> contain the right information however currently only 3 values can
> differ.
> 
> There is no address space in the intel_dp->dpcd struct for addresses
> 02200h through 0220Fh, and since so much of the data is a identical,
> simply overwrite the values stored in 0h through Fh with the
> values that can be overwritten from addresses 02200h through 0220Fh

Why overwrite all values if this is an expensive operation? From what I
can see, you'll need to read only h - 5h

> 
> Signed-off-by: Matt Atwood 
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 14 ++
>  include/drm/drm_dp_helper.h |  5 +++--
>  2 files changed, 17 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_dp.c
> b/drivers/gpu/drm/i915/intel_dp.c
> index dde92e4af5d3..899ebc5cece6 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -3738,6 +3738,20 @@ intel_dp_read_dpcd(struct intel_dp *intel_dp)
>    sizeof(intel_dp->dpcd)) < 0)
>   return false; /* aux transfer failed */
>  
> + if (intel_dp->dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
> + DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT &&
> + intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14) {

The same section in the spec also says - "The Extended Receiver
Capability registers at DPCD Addresses 02200h through 0220Fh shall
contain the DPRX’s true capability, while the original Base Receiver
Capability registers at DPCD Addresses 0h through Fh might
indicate DPCD r1.1, a MAX_LINK_RATE of 2.7Gbps/lane, and no DFP to
avoid interoperability issues ..."

Which means, intel_dp->dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14 is probably
not going to be true for panels you want to read the extended
capabilities for. 

> + uint8_t dpcd_ext[16];
> +
> + if (drm_dp_dpcd_read(_dp->aux,
> DP_DP13_DPCD_REV,
> + _ext, sizeof(dpcd_ext)) < 0)
> + return false; /* aux transfer failed */
> +
> + intel_dp->dpcd[DP_DPCD_REV] = dpcd_ext[DP_DPCD_REV];
> + intel_dp->dpcd[DP_MAX_LINK_RATE] =
> dpcd_ext[DP_MAX_LINK_RATE];
> + intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] =
> + dpcd_ext[DP_DOWNSTREAMPORT_PRESENT];
> + }
>   DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd),
> intel_dp->dpcd);
>  
>   return intel_dp->dpcd[DP_DPCD_REV] != 0;
> diff --git a/include/drm/drm_dp_helper.h
> b/include/drm/drm_dp_helper.h
> index c01564991a9f..757bd5913f3d 100644
> --- a/include/drm/drm_dp_helper.h
> +++ b/include/drm/drm_dp_helper.h

This should be a separate patch as it's outside i915.

> @@ -123,8 +123,9 @@
>  # define DP_FRAMING_CHANGE_CAP   (1 << 1)
>  # define DP_DPCD_DISPLAY_CONTROL_CAPABLE (1 << 3) /* edp v1.2 or
> higher */
>  
> -#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2? */
> -# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2? */
> +#define DP_TRAINING_AUX_RD_INTERVAL 0x00e   /* XXX 1.2?
> */
> +# define DP_TRAINING_AUX_RD_MASK0x7F/* XXX 1.2?
> */
> +# define DP_EXTENDED_RECEIVER_CAP_FIELD_PRESENT (1 << 7)/* XXX 1.2?
> */
>  
>  #define DP_ADAPTER_CAP   0x00f   /* 1.2 */
>  # define DP_FORCE_LOAD_SENSE_CAP (1 << 0)
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx