mchehab: I'm going through patchwork and delegating patches to me where appropriate. With a bit of luck I'll have time to review patches and make one or more pull requests today or Friday.
mchehab: hverkuil: Hi I pushed a V4L bus formats patch + documentation to describe the HDMI bus formats used for the Synopsys HDMI TX IP
It seems coherent, but I'm not sure about the 4:2:0 bus formats
Jose Abreu from Synopsys confirmed me the format, but I'm not sure about the define naming...
It's on my todo list to review. Ping me if you haven't heard from me by Tuesday.
narmstrong: ^^^^
hverkuil: thanks, no hurry !
mchehab: sailus: I see patches being posted to linux-media for drivers/staging/media/atomisp, but that's not in mainline, nor has anything been posted to linux-media.
Do either of you know anything about this? Is it from linux-next?
'918 files changed, 204643 insertions(+)' not bad
mchehab: Bom dia!
hverkuil: I've no idea
maybe it was added at via staging tree
sailus: moikka!
mchehab: I need to leave in a minute.
yeah, staging. I've complained to Greg about us not being informed.
Nothing urgent, let's talk when we're both around. :-)
atomisp?
y
Yeah, I noticed that, too.
same happened with bcm2835
I don't dare to look at the code.
But. I'll be back later. Have a nice evening!
bcm2835 is actually an special case...
as, on this chipset, the SoC boots from GPU
and the linux driver needs to talk to the GPU to access the V4L2 device...
via an interface that it is only at staging, and being modified
anyway, I agreed with Greg that it will only be moved out of staging after our review
but I've no idea about what's there at the atomisp driver
Re bcm2835: that's still no excuse not to CC us. How hard is that?
it was c/c to our ML: [PATCH 0/6] staging: BCM2835 MMAL V4L2 camera driver
I'm not finding any posts regarding atomisp either at linux-media or at lkml (except for fixup patches)
Greg claims he informed you.
didn't find any e-mail about it on my inboxes
hverkuil: found it. the e-mail was c/c linux-media, but it was big
likely, vger rejected it
I'll forward it to you
it was sent to an usual e-mail
ok
about 1MB
just sent
just what I needed :-)
:-)
I wonder why they merged it to staging instead of cleaning it up beforehand
It looks like a direct dump from Android drivers into staging.
that's not very polite
OMG, it uses videobuf, not vb2 :-(
it will never move out of staging
nobody has the resources to clean up 200k lines
unless someone is really committed to make it happen, no
right.
Intel has
it depends on whether they want to invest in that
resources and priorities
one more reason to remove vb1...
pinchartl: well now you can't, unless you cleanup that driver
larsc: I have no issue with breaking a staging driver
they're broken anyway :-)
I'd happily throw it away just to be able to get rid of vb1. But there are still too many drivers that use it (bttv).
how can we get bttv converted to vb2 ? it's the biggest blocker
there are others too. But nobody has time to work on it.
I've been sitting on a 80% finished vb2 conversion of cx231xx for ages (22 months) without being able to finish it.
does atomisp use soc-camera ?
no, that's one sin that they didn't commit.
ok
but it includes like 20 sensor drivers
are you surprised ? :-)
using subdev API though, so that's a good thing
but all accessing state from the videodev
    struct atomisp_input_stream_info *stream_info =
        (struct atomisp_input_stream_info *)fmt->reserved;
well done
in a sensor driver
hm regarding the section "it worked before"... what is wrong here? http://pastebin.com/Snref1R2 its on debian 3.16.0-4-amd64
Bus 004 Device 006: ID 046d:08ce Logitech, Inc. QuickCam Pro 5000
hverkuil: you've got a patch in your mailbox that removes vb1 support from soc-camera