microchip-csi2dc: Remove VC support for now
Subsystems (2)
PATCH 1/1 microchip-csi2dc: Remove VC support for now
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)
drivers/media/platform/atmel/microchip-csi2dc.c +2 -16
@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &
@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Signature · 1 line
2.30.2
Mauro Carvalho Chehab
maintainerEm Wed, 2 Feb 2022 17:36:09 +0200
Sakari Ailus <[email protected]> escreveu:
Quoted 5 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Cc Eugen Hristev <[email protected]>
Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Quoted 48 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Quoted 7 lines · Mauro Carvalho Chehab
Em Wed, 2 Feb 2022 17:36:09 +0200
Sakari Ailus <[email protected]> escreveu:Quoted 4 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Quoted 6 lines
Quoted 1 line
Cc Eugen Hristev <[email protected]>Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Hi ,
I am not happy with this change. It looks like I wasn't even CC-ed on the original patch e-mail.
The effect on the driver will be that everything will be treated as virtual channel=0 .
I do not yet understand why we are about to remove V4L2_MBUS_CSI2_CHANNEL_* as I remember this was just introduced. Is there any alternative in place ?
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
In the end you guys are the maintainers for the subsystem and can have this change if you like, I am more unhappy about the fact that changes happen suddenly and without notice.
Eugen
Quoted 49 lines
Quoted 49 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Quoted 9 lines · On 2/8/22 10:00 AM, Eugen Hristev - M18282 wrote:
Quoted 8 lines · Mauro Carvalho Chehab
Em Wed, 2 Feb 2022 17:36:09 +0200
Sakari Ailus <[email protected]> escreveu:Quoted 5 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Cc Eugen Hristev <[email protected]>
Having a closer look, I am added here to the Cc: list but your git send-email did not send it to me. Perhaps you have a CC-suppress activated.
Quoted 81 lines
Quoted 5 lines
Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Hi ,
I am not happy with this change. It looks like I wasn't even CC-ed on the original patch e-mail.
The effect on the driver will be that everything will be treated as virtual channel=0 .
I do not yet understand why we are about to remove V4L2_MBUS_CSI2_CHANNEL_* as I remember this was just introduced. Is there any alternative in place ?My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
In the end you guys are the maintainers for the subsystem and can have this change if you like, I am more unhappy about the fact that changes happen suddenly and without notice.
Eugen
Quoted 49 lines
Quoted 49 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Hi Eugen,
Quoted 14 lines · On Tue, Feb 08, 2022 at 08:00:19AM +0000, [email protected] wrote:
Quoted 6 lines · Mauro Carvalho Chehab
Em Wed, 2 Feb 2022 17:36:09 +0200 Sakari Ailus escreveu:
Quoted 4 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Quoted 6 lines
Quoted 1 line
Cc Eugen Hristev <[email protected]>Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Quoted 7 lines
Hi ,
I am not happy with this change. It looks like I wasn't even CC-ed on the original patch e-mail.
The effect on the driver will be that everything will be treated as virtual channel=0 .
I don't think there's any risk of regression, as we have no driver setting any of the V4L2_MBUS_CSI2_CHANNEL_[123] flags in the kernel.
Quoted 2 lines
I do not yet understand why we are about to remove V4L2_MBUS_CSI2_CHANNEL_* as I remember this was just introduced.
Those flags were added in 2011. If you think of that as "just introduced" then I understand why you would be unhappy about "sudden changes" mentioned below ;-)
Quoted 1 line
Is there any alternative in place ?
Virtual channels have never been properly supported in V4L2. This is going to change with "[PATCH v10 00/38] v4l: subdev internal routing and streams" ([1]).
[1] https://lore.kernel.org/all/[email protected]
Quoted 7 lines
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
When regressions are introduced this makes sense, but here we're dropping a feature that isn't used as no kernel driver selects a VC different than 0.
Quoted 52 lines
In the end you guys are the maintainers for the subsystem and can have this change if you like, I am more unhappy about the fact that changes happen suddenly and without notice.
Quoted 48 lines
Quoted 48 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Signature · 3 lines
Regards, Laurent Pinchart
Hello everyone,
Quoted 11 lines · Laurent Pinchart
Quoted 7 lines
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
When regressions are introduced this makes sense, but here we're dropping a feature that isn't used as no kernel driver selects a VC different than 0.
I'd like to add that the purpose of virtual channels in CSI-2 is to be able to transport multiple logically separate streams over a single link. That has never been supported in V4L2 so these flags have been unusable to begin with.
The streams set a lot of people have been working on on their turn is the way to support virtual channels (as well as other features such as data types), and it does not involve using mbus flags (as the virtual channel isn't bound to the bus).
The intent was to cc the patch to Eugen but it seems my gitconfig prevented that.
Signature · 3 lines
Kind regards, Sakari Ailus
Quoted 20 lines · Sakari Ailus
Hello everyone,
Quoted 11 lines · Laurent Pinchart
Quoted 7 lines
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
When regressions are introduced this makes sense, but here we're dropping a feature that isn't used as no kernel driver selects a VC different than 0.
I'd like to add that the purpose of virtual channels in CSI-2 is to be able to transport multiple logically separate streams over a single link. That has never been supported in V4L2 so these flags have been unusable to begin with.
Quoted 7 lines
The streams set a lot of people have been working on on their turn is the way to support virtual channels (as well as other features such as data types), and it does not involve using mbus flags (as the virtual channel isn't bound to the bus).
The intent was to cc the patch to Eugen but it seems my gitconfig prevented that.
Hello Sakari,
I appreciate you taking time to explain this and the intent to cc me on the patch.
Thanks,
Eugen
Quoted 6 lines
--
Kind regards,Sakari Ailus
Quoted 28 lines · Laurent Pinchart
Hi Eugen,
Quoted 14 lines · On Tue, Feb 08, 2022 at 08:00:19AM +0000, [email protected] wrote:
Quoted 6 lines · Mauro Carvalho Chehab
Em Wed, 2 Feb 2022 17:36:09 +0200 Sakari Ailus escreveu:
Quoted 4 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Quoted 6 lines
Quoted 1 line
Cc Eugen Hristev <[email protected]>Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Quoted 7 lines
Hi ,
I am not happy with this change. It looks like I wasn't even CC-ed on the original patch e-mail.
The effect on the driver will be that everything will be treated as virtual channel=0 .
I don't think there's any risk of regression, as we have no driver setting any of the V4L2_MBUS_CSI2_CHANNEL_[123] flags in the kernel.
Quoted 6 lines
Quoted 2 lines
I do not yet understand why we are about to remove V4L2_MBUS_CSI2_CHANNEL_* as I remember this was just introduced.
Those flags were added in 2011. If you think of that as "just introduced" then I understand why you would be unhappy about "sudden changes" mentioned below ;-)
Hello Laurent,
I am sorry, I was referring to the get_mbus_config which was recently added, which in my case uses V4L2_MBUS_CSI2_CHANNEL_* . Without the get_mbus_config, I wasn't supposed to use the V4L2_MBUS_CSI2_CHANNEL_* at all .
Hence my confusion in adding them into the same bucket.
My driver uses these flags because my initial implementation to get the VC from the devicetree was rejected and this was suggested instead.
Quoted 8 lines
Quoted 2 lines
Is there any alternative in place ?
Virtual channels have never been properly supported in V4L2. This is going to change with "[PATCH v10 00/38] v4l: subdev internal routing and streams" ([1]).
In that case then I suppose the support in the csi2dc driver has to be rewritten
Quoted 9 lines
Quoted 8 lines
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
Quoted 3 lines
When regressions are introduced this makes sense, but here we're dropping a feature that isn't used as no kernel driver selects a VC different than 0.
I still disagree that these should be removed before the patch you mentioned earlier about adding 'subdev internal routing and streams' is applied. I see things in the way that after the new support is available, old flags , even if unused, can be removed.
You can disagree, of course.
Thank you for sharing your thoughts,
Eugen
Quoted 59 lines
Quoted 52 lines
In the end you guys are the maintainers for the subsystem and can have this change if you like, I am more unhappy about the fact that changes happen suddenly and without notice.
Quoted 47 lines
Quoted 47 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }Quoted 1 line
Quoted 1 line
drivers/media/platform/atmel/microchip-csi2dc.c static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) · line 371
Quoted 1 line
drivers/media/platform/atmel/microchip-csi2dc.c +0
-0static void csi2dc_vp_update (struct csi2dc_device *csi2dc)--
Regards,Laurent Pinchart
Hi Eugen,
Quoted 47 lines · On Tue, Feb 08, 2022 at 11:36:16AM +0000, [email protected] wrote:
Quoted 26 lines · Laurent Pinchart
Quoted 14 lines · On Tue, Feb 08, 2022 at 08:00:19AM +0000, [email protected] wrote:
Quoted 6 lines · Mauro Carvalho Chehab
Em Wed, 2 Feb 2022 17:36:09 +0200 Sakari Ailus escreveu:
Quoted 4 lines
As part of removing mbus config flags, remove VC flag use in the microchip-csi2dc driver. The support can be reintroduced later on as part of the streams patches.
Quoted 6 lines
Quoted 1 line
Cc Eugen Hristev <[email protected]>Hmm... that sounds a regression to me. What effects this will cause at the driver?
Eugen, any comments?
Quoted 7 lines
Hi ,
I am not happy with this change. It looks like I wasn't even CC-ed on the original patch e-mail.
The effect on the driver will be that everything will be treated as virtual channel=0 .
I don't think there's any risk of regression, as we have no driver setting any of the V4L2_MBUS_CSI2_CHANNEL_[123] flags in the kernel.
Quoted 6 lines
Quoted 2 lines
I do not yet understand why we are about to remove V4L2_MBUS_CSI2_CHANNEL_* as I remember this was just introduced.
Those flags were added in 2011. If you think of that as "just introduced" then I understand why you would be unhappy about "sudden changes" mentioned below ;-)
Hello Laurent,
I am sorry, I was referring to the
get_mbus_configwhich was recently added, which in my case uses V4L2_MBUS_CSI2_CHANNEL_* . Without theget_mbus_config, I wasn't supposed to use the V4L2_MBUS_CSI2_CHANNEL_* at all .
Hence my confusion in adding them into the same bucket.My driver uses these flags because my initial implementation to get the VC from the devicetree was rejected and this was suggested instead.
Quoted 2 lines
Quoted 1 line
Is there any alternative in place ?
Quoted 8 lines
Quoted 5 lines
Virtual channels have never been properly supported in V4L2. This is going to change with "[PATCH v10 00/38] v4l: subdev internal routing and streams" ([1]).
In that case then I suppose the support in the csi2dc driver has to be rewritten
Only when you'll need to support virtual channels other than VC 0. It's a non-negligible effort if you need to capture multiple streams from different VCs simultaneously, but if it's only about selecting the correct VC number for one stream, it's much easier. As far as I understand, your hardware supports the latter, not the former, right ?
Do you have any particular CSI-2 source in mind that produces different VCs by the way ?
Quoted 12 lines
Quoted 11 lines
Quoted 7 lines
My opinion is that if we want to replace something existing with a new API or something else, we should first add the new support, block any new adopters for the old API such that everyone uses the new API, and only after that convert the old API clients to the new API. So 'can be reintroduced later on' is not okay. We can't remove things in the hope that it would be reintroduced later. Just my personal take on this, feel free to have a different opinion.
When regressions are introduced this makes sense, but here we're dropping a feature that isn't used as no kernel driver selects a VC different than 0.
Quoted 5 lines
I still disagree that these should be removed before the patch you mentioned earlier about adding 'subdev internal routing and streams' is applied. I see things in the way that after the new support is available, old flags , even if unused, can be removed.
You can disagree, of course.
I agree it would be nice, but the multiplexed streams series will still need time to land (it's been out there for ages, we've resumed work on it about a year ago, and Tomi has done a really amazing job moving this forward). I'd like to avoid blocking the V4L2_MBUS_CSI2_CHANNEL_* removal until multiplexed streams get merged if possible. If there was an urgent need to support different VCs in your driver the situation would of course be different, but if that's not foreseen until we get proper support for multiplexed streams, I think moving forward with all the rework in parallel is best. V4L2 has few active contributors for core code, it's really hard modernizing the internals of the subsystem with the resource shortage, I'd like to avoid making it even harder :-)
Quoted 55 lines
Thank you for sharing your thoughts,
EugenQuoted 52 lines
Quoted 52 lines
In the end you guys are the maintainers for the subsystem and can have this change if you like, I am more unhappy about the fact that changes happen suddenly and without notice.
Quoted 48 lines
Quoted 48 lines
Signed-off-by Sakari Ailus <[email protected]>--- .../media/platform/atmel/microchip-csi2dc.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-)drivers/media/platform/atmel/microchip-csi2dc.c +2
-16@@ -348,24 +348,15 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)if (ret == -ENOIOCTLCMD ) {dev_dbg (csi2dc->dev,"no remote mbus configuration available\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; }if (ret) {dev_err (csi2dc->dev,"failed to get remote mbus configuration\n" ); - goto csi2dc_get_mbus_config_defaults; +return 0 ; } -if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_0 ) - csi2dc->vc =0 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_1 ) - csi2dc->vc =1 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_2 ) - csi2dc->vc =2 ; -else if (mbus_config.flags &V4L2_MBUS_CSI2_CHANNEL_3 ) - csi2dc->vc =3 ; -dev_dbg (csi2dc->dev,"subdev sending on channel %d\n" , csi2dc->vc); csi2dc->clk_gated = mbus_config.flags &@@ -375,11 +366,6 @@ static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc) csi2dc->clk_gated ? "gated" : "free running");return 0 ; - -csi2dc_get_mbus_config_defaults: - csi2dc->vc =0 ;/* Virtual ID 0 by default */ - -return 0 ; }static void csi2dc_vp_update (struct csi2dc_device *csi2dc)
Signature · 3 lines
Regards, Laurent Pinchart
Hi Eugen,
Quoted 8 lines · On Tue, Feb 08, 2022 at 11:36:16AM +0000, [email protected] wrote:
I am sorry, I was referring to the
get_mbus_configwhich was recently added, which in my case uses V4L2_MBUS_CSI2_CHANNEL_* . Without theget_mbus_config, I wasn't supposed to use the V4L2_MBUS_CSI2_CHANNEL_* at all .
Hence my confusion in adding them into the same bucket.My driver uses these flags because my initial implementation to get the VC from the devicetree was rejected and this was suggested instead.
As noted by Laurent, this feature is not supported in mainline currently. Beyond that, no transmitter driver uses virtual channel different from zero, so the case for setting non-zero virtual channel in absence of multiple streams does not exist.
Also the interfaces required to make use of multiple streams do not exist in mainline at the moment. They will be added by the streams patchset eventually, and adding support for them requires extra driver work.
Thus, we're not losing any functionality by dropping the defines and potentially associated driver code. Therefore I see no reason to postpone these patches.
Signature · 3 lines
Kind regards, Sakari Ailus
Quoted 17 lines · Sakari Ailus
Hi Eugen,
Quoted 8 lines · On Tue, Feb 08, 2022 at 11:36:16AM +0000, [email protected] wrote:
I am sorry, I was referring to the
get_mbus_configwhich was recently added, which in my case uses V4L2_MBUS_CSI2_CHANNEL_* . Without theget_mbus_config, I wasn't supposed to use the V4L2_MBUS_CSI2_CHANNEL_* at all .
Hence my confusion in adding them into the same bucket.My driver uses these flags because my initial implementation to get the VC from the devicetree was rejected and this was suggested instead.
As noted by Laurent, this feature is not supported in mainline currently. Beyond that, no transmitter driver uses virtual channel different from zero, so the case for setting non-zero virtual channel in absence of multiple streams does not exist.
Quoted 7 lines
Also the interfaces required to make use of multiple streams do not exist in mainline at the moment. They will be added by the streams patchset eventually, and adding support for them requires extra driver work.
Thus, we're not losing any functionality by dropping the defines and potentially associated driver code. Therefore I see no reason to postpone these patches.
Hi Sakari,
I understand. I am fine with the patch.
Eugen
Quoted 6 lines
--
Kind regards,Sakari Ailus
Quoted 22 lines · On Tue, Feb 15, 2022 at 08:10:18PM +0000, [email protected] wrote:
Quoted 21 lines · Sakari Ailus
Hi Eugen,
Quoted 8 lines · On Tue, Feb 08, 2022 at 11:36:16AM +0000, [email protected] wrote:
I am sorry, I was referring to the
get_mbus_configwhich was recently added, which in my case uses V4L2_MBUS_CSI2_CHANNEL_* . Without theget_mbus_config, I wasn't supposed to use the V4L2_MBUS_CSI2_CHANNEL_* at all .
Hence my confusion in adding them into the same bucket.My driver uses these flags because my initial implementation to get the VC from the devicetree was rejected and this was suggested instead.
As noted by Laurent, this feature is not supported in mainline currently. Beyond that, no transmitter driver uses virtual channel different from zero, so the case for setting non-zero virtual channel in absence of multiple streams does not exist.
Also the interfaces required to make use of multiple streams do not exist in mainline at the moment. They will be added by the streams patchset eventually, and adding support for them requires extra driver work.
Quoted 7 lines
Quoted 3 lines
Thus, we're not losing any functionality by dropping the defines and potentially associated driver code. Therefore I see no reason to postpone these patches.
Hi Sakari,
I understand. I am fine with the patch.
Thank you, Eugen.
Signature · 1 line
Sakari Ailus