[07:58] <fling> kernfs: can not remove 'max_openers', no directory -> https://bpaste.net/show/XL3TC
[07:58] <fling> modprobe: ERROR: could not insert 'v4l2loopback': Bad address
[07:59] <fling> how to fix? ^
[07:59] <fling> Is it something new in 5.4 preventing me from modprobing it?
[08:01] <hverkuil> fling: doesn't ring a bell for me.
[08:02] <fling> works with 5.2 with the same config and all the older kernels hmmm
[08:03] <fling> hverkuil: am I doing something wrong?
[08:08] <fling> Should I bugreport?
[08:26] <hverkuil> fling: v4l2loopback is an out-of-tree driver, we don't support this. If you get the same error with an in-kernel driver such as vivid, then let us know.
[08:30] <fling> hverkuil: ok, thank. Can I replace it with something in-kernel?
[08:38] <hverkuil> vivid can do something similar, but it is a bit painful to set up (you need to enable video loopback and switch the input and output to 'HDMI' and set the right timings for input and output)
[08:52] <ezequielg> fling: you could try to put loopback behavior in vimc, iirc.
[08:56] <LazyGrizzly> ping pinchartl
[09:27] *** henriknj has quit IRC (Remote host closed the connection)
[09:27] *** z3ntu_ has quit IRC (Write error: Connection reset by peer)
[09:27] *** johanhelsing has quit IRC (Write error: Connection reset by peer)
[09:27] *** nldvos[m] has quit IRC (Write error: Connection reset by peer)
[09:27] *** dafna21 has quit IRC (Remote host closed the connection)
[09:27] *** wonchul[m] has quit IRC (Read error: Connection reset by peer)
[09:27] *** kbingham[m] has quit IRC (Remote host closed the connection)
[10:43] <fling> ezequielg: what is vimc?
[11:13] *** jherland has quit IRC (Quit: leaving)
[11:43] <ndufresne> fling: it's another type of virtual driver, this one with a media controller (multiple configurable nodes)
[12:52] <mfelsch> sailus: ping
[13:33] <bparrot_> hverkuil: ping
[13:33] <hverkuil> bparrot_: pong
[13:34] <bparrot_> hverkuil: under what circumstance does the 'slave-mode' DT property would need to be used?
[13:36] <hverkuil> I don't really know much more than this: Documentation/devicetree/bindings/media/video-interfaces.txt
[13:37] <hverkuil> sailus might have done more with this.
[13:38] <bparrot_> yeah I only see it used here: arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi
[13:39] <hverkuil> I'm not sure if it is ever actually used. I suspect it might actually be a remnant of soc-camera.
[13:39] <bparrot_> internally it is specified in our dts but i am wondering if that's correct.
[13:39] <bparrot_> probably based on the older commit.
[13:39] <bparrot_> hopefully sailus can confirm...
[13:40] <hverkuil> I can't really think of a use case for slave mode.
[13:42] <bparrot_> alright thanks
[15:22] <sailus> bparrot_: My recollection is that this was some rather curious hardware-specific configuration flag.
[15:22] <sailus> And AFAIR hverkuil once told me about it. :-)
[15:22] <sailus> In Prague actually.
[15:23] <sailus> Perhaps the direction of the parallel clock was from the bridge to the sensor?
[15:23] <bparrot_> sailus: right that is what i gather so it would only make sense on the sesnor side and not on the bridge side
[15:24] <bparrot_> if that was the case
[15:30] <sailus> I guess the bridge would also need to set it, to provide the clock.
[15:31] <sailus> Ah, this is about sync signals, not the clock.
[15:31] <sailus> bparrot_: Don't tell me you're contemplating using this? :-)
[15:31] <hverkuil> that it is just about sync signals is weird. It makes sense for the clock signal, but not for the syncs.
[15:32] <hverkuil> It == slave mode
[15:34] <bparrot_> sailus: no i am not. I just see it used internally and was wondering if that was a mistake and can remove it...
[16:40] <fling> ndufresne: do you have a link?
[16:44] *** benjiG has left 
[17:28] <mfelsch> sailus: Can we discuss the connector/endpoint parsing tomorrow here?
[17:57] *** snawrocki has quit IRC (Ping timeout: 250 seconds)
[17:57] *** padovan has quit IRC (Ping timeout: 250 seconds)
[18:21] *** kbingham has quit IRC (Ping timeout: 240 seconds)
[18:23] *** kos_tom has quit IRC (Ping timeout: 250 seconds)
[21:48] <ndufresne> fling: you could start from this one I guess, https://www.kernel.org/doc/html/latest/media/v4l-drivers/vimc.html
[21:50] <ndufresne> koike: ^ can you attach an OUTPUT entity to any of the virtual blocks ?
[21:51] <ndufresne> ah nevermind, just noticed the RGB/YUV Input entity, and the inactive link in the graph
[21:52] <ndufresne> fling: so basically, you'll have to activate the link between "RGB/YUV Input" entity, this entity is backed by a /dev/videoN device, and is where you send your stream to
[21:52] <ndufresne> then activate the link from "Scaler" to "RGB/YUV Capture"
[21:54] <ndufresne> when that is setup (you can do that in script at boot reall), you can stream to the Input entity (it's input from the driver perspective), and let you app use the Capture entity, which will work similarly to a v4l2loopback