| Message ID | 1456243798-12453-1-git-send-email-javier@osg.samsung.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Javier, On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: > The chip internal signal generator was modelled as an input connector > and represented as a media entity but isn't really a connector so the > driver was changed to use the V4L2_CID_TEST_PATTERN control instead. > > Remove the signal generator input from the list of connectors in the > tvp5150 DT binding document as well since isn't a connector anymore. > > Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> > > --- > Hello, > > I think is OK to change this DT binding because is only in the media tree > for now and not in mainline yet and also is expected to change more since > there are still discussions about how input connectors will be supported > by the Media Controller framework in the media subsystem. I think that's fine, yes Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> I haven't noticed the patch that introduced this early enough I'm afraid, and I think we still have issues with those bindings. The tvp5150 node should *not* contain connector subnodes, the connectors nodes should use the bindings defined in Documentation/devicetree/bindings/display/connector/ and be linked to the tvp5150 node using the OF graph bindings (ports and endpoints). Do you think you could fix that ? > Documentation/devicetree/bindings/media/i2c/tvp5150.txt | 3 --- > 1 file changed, 3 deletions(-) > > diff --git a/Documentation/devicetree/bindings/media/i2c/tvp5150.txt > b/Documentation/devicetree/bindings/media/i2c/tvp5150.txt index > daa20e43a8e3..d13f8b8e235d 100644 > --- a/Documentation/devicetree/bindings/media/i2c/tvp5150.txt > +++ b/Documentation/devicetree/bindings/media/i2c/tvp5150.txt > @@ -30,9 +30,6 @@ The possible values for the "input" property are: > 1: Composite1 > 2: S-Video > > -and on a tvp5150am1 and tvp5151 there is another: > - 4: Signal generator > - > The list of valid input connectors are defined in > dt-bindings/media/tvp5150.h header file and can be included by device tree > source files.
Hello Laurent, On 02/23/2016 01:16 PM, Laurent Pinchart wrote: > Hi Javier, > > On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: >> The chip internal signal generator was modelled as an input connector >> and represented as a media entity but isn't really a connector so the >> driver was changed to use the V4L2_CID_TEST_PATTERN control instead. >> >> Remove the signal generator input from the list of connectors in the >> tvp5150 DT binding document as well since isn't a connector anymore. >> >> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> >> >> --- >> Hello, >> >> I think is OK to change this DT binding because is only in the media tree >> for now and not in mainline yet and also is expected to change more since >> there are still discussions about how input connectors will be supported >> by the Media Controller framework in the media subsystem. > > I think that's fine, yes > > Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > Thanks. > I haven't noticed the patch that introduced this early enough I'm afraid, and > I think we still have issues with those bindings. > Yes, I posted those patches and got merged before we had the discussion about input connectors over IRC so I didn't know what was the correct way to do it. > The tvp5150 node should *not* contain connector subnodes, the connectors nodes > should use the bindings defined in > Documentation/devicetree/bindings/display/connector/ and be linked to the > tvp5150 node using the OF graph bindings (ports and endpoints). > Agreed. > Do you think you could fix that ? > Yes I will, I'm waiting for the input connectors discussions to settle so I can post a final version of the DT bindings following what is agreed by all. Best regards,
Hi Javier, On Tuesday 23 February 2016 13:27:51 Javier Martinez Canillas wrote: > On 02/23/2016 01:16 PM, Laurent Pinchart wrote: > > On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: > >> The chip internal signal generator was modelled as an input connector > >> and represented as a media entity but isn't really a connector so the > >> driver was changed to use the V4L2_CID_TEST_PATTERN control instead. > >> > >> Remove the signal generator input from the list of connectors in the > >> tvp5150 DT binding document as well since isn't a connector anymore. > >> > >> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> > >> > >> --- > >> Hello, > >> > >> I think is OK to change this DT binding because is only in the media tree > >> for now and not in mainline yet and also is expected to change more since > >> there are still discussions about how input connectors will be supported > >> by the Media Controller framework in the media subsystem. > > > > I think that's fine, yes > > > > Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > > Thanks. > > > I haven't noticed the patch that introduced this early enough I'm afraid, > > and I think we still have issues with those bindings. > > Yes, I posted those patches and got merged before we had the discussion > about input connectors over IRC so I didn't know what was the correct way > to do it. > > > The tvp5150 node should *not* contain connector subnodes, the connectors > > nodes should use the bindings defined in > > Documentation/devicetree/bindings/display/connector/ and be linked to the > > tvp5150 node using the OF graph bindings (ports and endpoints). > > Agreed. > > > Do you think you could fix that ? > > Yes I will, I'm waiting for the input connectors discussions to settle so I > can post a final version of the DT bindings following what is agreed by all. > Shouldn't we revert the patch that introduced connectors support in the DT bindings in the meantime then, to avoid known to be broken bindings from hitting mainline in case we can't fix them in time for v4.6 ?
Hi Laurent, On 02/23/2016 03:02 PM, Laurent Pinchart wrote: > Hi Javier, > > On Tuesday 23 February 2016 13:27:51 Javier Martinez Canillas wrote: >> On 02/23/2016 01:16 PM, Laurent Pinchart wrote: >>> On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: >>>> The chip internal signal generator was modelled as an input connector >>>> and represented as a media entity but isn't really a connector so the >>>> driver was changed to use the V4L2_CID_TEST_PATTERN control instead. >>>> >>>> Remove the signal generator input from the list of connectors in the >>>> tvp5150 DT binding document as well since isn't a connector anymore. >>>> >>>> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> >>>> >>>> --- >>>> Hello, >>>> >>>> I think is OK to change this DT binding because is only in the media tree >>>> for now and not in mainline yet and also is expected to change more since >>>> there are still discussions about how input connectors will be supported >>>> by the Media Controller framework in the media subsystem. >>> >>> I think that's fine, yes >>> >>> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> >> >> Thanks. >> >>> I haven't noticed the patch that introduced this early enough I'm afraid, >>> and I think we still have issues with those bindings. >> >> Yes, I posted those patches and got merged before we had the discussion >> about input connectors over IRC so I didn't know what was the correct way >> to do it. >> >>> The tvp5150 node should *not* contain connector subnodes, the connectors >>> nodes should use the bindings defined in >>> Documentation/devicetree/bindings/display/connector/ and be linked to the >>> tvp5150 node using the OF graph bindings (ports and endpoints). >> >> Agreed. >> >>> Do you think you could fix that ? >> >> Yes I will, I'm waiting for the input connectors discussions to settle so I >> can post a final version of the DT bindings following what is agreed by all. >> > > Shouldn't we revert the patch that introduced connectors support in the DT > bindings in the meantime then, to avoid known to be broken bindings from > hitting mainline in case we can't fix them in time for v4.6 ? > Yes, that would be a good idea. I've seen recently though a DT binding doc that was marked as unstable / work in progress and I wonder if that's a new accepted convention for DT binding docs or is just something that slipped through review. The commit I'm talking about is f07b4e49d27e ("Documentation: bindings: berlin: consider our dt bindings as unstable") but I don't see anything documented in Documentation/devicetree/bindings/ABI.txt. In any case, I'm fine with either marking the DT binding doc as unstable or to revert the patch that added the connectors portion to the tvp5150 DT binding. Best regards,
Hi Javier, On Tuesday 23 February 2016 15:23:48 Javier Martinez Canillas wrote: > On 02/23/2016 03:02 PM, Laurent Pinchart wrote: > > On Tuesday 23 February 2016 13:27:51 Javier Martinez Canillas wrote: > >> On 02/23/2016 01:16 PM, Laurent Pinchart wrote: > >>> On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: > >>>> The chip internal signal generator was modelled as an input connector > >>>> and represented as a media entity but isn't really a connector so the > >>>> driver was changed to use the V4L2_CID_TEST_PATTERN control instead. > >>>> > >>>> Remove the signal generator input from the list of connectors in the > >>>> tvp5150 DT binding document as well since isn't a connector anymore. > >>>> > >>>> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> > >>>> > >>>> --- > >>>> Hello, > >>>> > >>>> I think is OK to change this DT binding because is only in the media > >>>> tree for now and not in mainline yet and also is expected to change > >>>> more since there are still discussions about how input connectors will > >>>> be supported by the Media Controller framework in the media subsystem. > >>> > >>> I think that's fine, yes > >>> > >>> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > >> > >> Thanks. > >> > >>> I haven't noticed the patch that introduced this early enough I'm > >>> afraid, and I think we still have issues with those bindings. > >> > >> Yes, I posted those patches and got merged before we had the discussion > >> about input connectors over IRC so I didn't know what was the correct way > >> to do it. > >> > >>> The tvp5150 node should *not* contain connector subnodes, the connectors > >>> nodes should use the bindings defined in > >>> Documentation/devicetree/bindings/display/connector/ and be linked to > >>> the tvp5150 node using the OF graph bindings (ports and endpoints). > >> > >> Agreed. > >> > >>> Do you think you could fix that ? > >> > >> Yes I will, I'm waiting for the input connectors discussions to settle so > >> I can post a final version of the DT bindings following what is agreed by > >> all. > > > > Shouldn't we revert the patch that introduced connectors support in the DT > > bindings in the meantime then, to avoid known to be broken bindings from > > hitting mainline in case we can't fix them in time for v4.6 ? > > Yes, that would be a good idea. I've seen recently though a DT binding doc > that was marked as unstable / work in progress and I wonder if that's a new > accepted convention for DT binding docs or is just something that slipped > through review. I'm not sure if it's an established practice but I certainly like it. However, in this specific case, we know that the bindings are broken, so I think a revert would be better. > The commit I'm talking about is f07b4e49d27e ("Documentation: bindings: > berlin: consider our dt bindings as unstable") but I don't see anything > documented in Documentation/devicetree/bindings/ABI.txt. > > In any case, I'm fine with either marking the DT binding doc as unstable or > to revert the patch that added the connectors portion to the tvp5150 DT > binding.
Hi Laurent, On 02/23/2016 03:28 PM, Laurent Pinchart wrote: > Hi Javier, [snip] >>> >>> Shouldn't we revert the patch that introduced connectors support in the DT >>> bindings in the meantime then, to avoid known to be broken bindings from >>> hitting mainline in case we can't fix them in time for v4.6 ? >> >> Yes, that would be a good idea. I've seen recently though a DT binding doc >> that was marked as unstable / work in progress and I wonder if that's a new >> accepted convention for DT binding docs or is just something that slipped >> through review. > > I'm not sure if it's an established practice but I certainly like it. However, > in this specific case, we know that the bindings are broken, so I think a > revert would be better. > Ok, I'll post a revert then that supersedes this patch. Best regards,
On Tue, Feb 23, 2016 at 08:28:14PM +0200, Laurent Pinchart wrote: > Hi Javier, > > On Tuesday 23 February 2016 15:23:48 Javier Martinez Canillas wrote: > > On 02/23/2016 03:02 PM, Laurent Pinchart wrote: > > > On Tuesday 23 February 2016 13:27:51 Javier Martinez Canillas wrote: > > >> On 02/23/2016 01:16 PM, Laurent Pinchart wrote: > > >>> On Tuesday 23 February 2016 13:09:58 Javier Martinez Canillas wrote: > > >>>> The chip internal signal generator was modelled as an input connector > > >>>> and represented as a media entity but isn't really a connector so the > > >>>> driver was changed to use the V4L2_CID_TEST_PATTERN control instead. > > >>>> > > >>>> Remove the signal generator input from the list of connectors in the > > >>>> tvp5150 DT binding document as well since isn't a connector anymore. > > >>>> > > >>>> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> > > >>>> > > >>>> --- > > >>>> Hello, > > >>>> > > >>>> I think is OK to change this DT binding because is only in the media > > >>>> tree for now and not in mainline yet and also is expected to change > > >>>> more since there are still discussions about how input connectors will > > >>>> be supported by the Media Controller framework in the media subsystem. > > >>> > > >>> I think that's fine, yes > > >>> > > >>> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > > >> > > >> Thanks. > > >> > > >>> I haven't noticed the patch that introduced this early enough I'm > > >>> afraid, and I think we still have issues with those bindings. > > >> > > >> Yes, I posted those patches and got merged before we had the discussion > > >> about input connectors over IRC so I didn't know what was the correct way > > >> to do it. > > >> > > >>> The tvp5150 node should *not* contain connector subnodes, the connectors > > >>> nodes should use the bindings defined in > > >>> Documentation/devicetree/bindings/display/connector/ and be linked to > > >>> the tvp5150 node using the OF graph bindings (ports and endpoints). > > >> > > >> Agreed. > > >> > > >>> Do you think you could fix that ? > > >> > > >> Yes I will, I'm waiting for the input connectors discussions to settle so > > >> I can post a final version of the DT bindings following what is agreed by > > >> all. > > > > > > Shouldn't we revert the patch that introduced connectors support in the DT > > > bindings in the meantime then, to avoid known to be broken bindings from > > > hitting mainline in case we can't fix them in time for v4.6 ? > > > > Yes, that would be a good idea. I've seen recently though a DT binding doc > > that was marked as unstable / work in progress and I wonder if that's a new > > accepted convention for DT binding docs or is just something that slipped > > through review. > > I'm not sure if it's an established practice but I certainly like it. However, > in this specific case, we know that the bindings are broken, so I think a > revert would be better. It is not normal practice, though there are some exceptions. > > The commit I'm talking about is f07b4e49d27e ("Documentation: bindings: > > berlin: consider our dt bindings as unstable") but I don't see anything > > documented in Documentation/devicetree/bindings/ABI.txt. That one was simply not reviewed by anyone that would object. It's been almost a year, so they should be stable now... Rob -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/Documentation/devicetree/bindings/media/i2c/tvp5150.txt b/Documentation/devicetree/bindings/media/i2c/tvp5150.txt index daa20e43a8e3..d13f8b8e235d 100644 --- a/Documentation/devicetree/bindings/media/i2c/tvp5150.txt +++ b/Documentation/devicetree/bindings/media/i2c/tvp5150.txt @@ -30,9 +30,6 @@ The possible values for the "input" property are: 1: Composite1 2: S-Video -and on a tvp5150am1 and tvp5151 there is another: - 4: Signal generator - The list of valid input connectors are defined in dt-bindings/media/tvp5150.h header file and can be included by device tree source files.
The chip internal signal generator was modelled as an input connector and represented as a media entity but isn't really a connector so the driver was changed to use the V4L2_CID_TEST_PATTERN control instead. Remove the signal generator input from the list of connectors in the tvp5150 DT binding document as well since isn't a connector anymore. Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com> --- Hello, I think is OK to change this DT binding because is only in the media tree for now and not in mainline yet and also is expected to change more since there are still discussions about how input connectors will be supported by the Media Controller framework in the media subsystem. Best regards, Javier Documentation/devicetree/bindings/media/i2c/tvp5150.txt | 3 --- 1 file changed, 3 deletions(-)