tzanger: I think more information is needed , like the kernel version,  ,and the outpu of v4l2-ctl -d0 --all
tzanger: what i.MX8 ? different SoCs use different drivers. are you using the mainline kernel or an NXP BSP kernel ?
dafna2[m]: ping
pong
there's something that has been bothering me for a loooong time with vimc, and I wanted to have you opinion on it
it's about the scaler
it uses a fixed upscaling factor, which can be set through a module parameter, and defaults to 3
first of all, 3 is horrible. most standard resolutions are not multiples of 3 in both directions
but more importantly, this doesn't match what real devices typically do
which is scaling with a variable factor
and mostly downscaling. lots of ISPs can't upscale
it reduces vimc's usefulness as a test driver
I'd like to fix that
but it would likely result in a different configuration of the vimc pipeline
and thus a different behaviour towards applications
what do you think ?
there were patches sent to fix it long time ago ,
what do you mean by 'different configuration of the pipeline'?
for instance, if we switch the driver to do downscaling instead of upscaling, it won't be able to produce the same output resolutions
if we want to make this look more like a real device, the sensor subdev should output a fixed resolution
https://lkml.org/lkml/2020/1/21/505
ok, so that could be revived
it would still use up-scaling though
but I guess it could also support downscaling
was there any particular blocker ? it seems that the issues mentioned in that e-mail thread shouldn't be too hard to address
(as for the sensor subdev configuration, I suppose we could address that later)
for some reason he didn't sent v5 and that was forgotten
he write clearly in the commit log `Now any resolution can be set at src (even smaller ones)`
ok
I'll have a look
thanks
so the sensor should change to have only a set of supprted resolutions instead of a range?
I think that would make sense, but on the other hand, now that I think about it, making the sensor configurable has the advantage that it could emulate different sensor resolution
s
which could be done through a module parameter, but that would be less flexible
dwlsalmeida: hey there, I've allowed myself to let a certain jagan_ know that you can help reply requestions around running our V4L2 work on Cedrus
he got a broken sample there he'd like to carry on for the debugging, while I'm away, could be nice to help him on how to dump v4l2 controls, ffmpeg from LibreELEC is working
hello there linux-media people :) I've gone through the process of putting together my first kernel patch, and in trying to send it to the majordomo address I get "Undeliverable" bounce back.. I'm trying to send from my hotmail account.. was wondering if this was the SpamBot in action, and if so, what can I do to be able to submit a patch?
hi HemPeter 
and welcome :-)
first of all, have you read the patch submission documentation (Documentation/process/submitting-patches.rst) ?
the "No MIME, no links, no compression, no attachments.  Just plain text" section is of particular interest
it mentions git send-email, which I highly recommend using
sending a patch from the hotmail web UI is probably not possible
Hi pinchartl =) Thanks for your welcome.
I had been going off this page, which I believe has the same info as the documentation: https://www.linuxtv.org/wiki/index.php/Development:_Submitting_Patches#Developer.27s_Certificate_of_Origin_1.1
I had submitted as plain text.. could paste it here, as it's not a big one ;)   wasn't from the WebUI, was from Thunderbird client
Even an e-mail with "subscribe linux-media" in the body bounced.. so it looks like it's a M$ problem :'(
https://docs.microsoft.com/en-gb/exchange/mail-flow-best-practices/non-delivery-reports-in-exchange-online/fix-error-code-550-5-7-1-in-exchange-online
what address did you send the patch to ?
majordomo@vger.kernel.org
the mailing list address is linux-media@vger.kernel.org
had been following the instructions here: https://www.linuxtv.org/wiki/index.php/Development:_How_to_submit_patches#Patch_Preparation , the link is under "Post your patches to the Linux-Media Mailing List"
cool, thanks, I'll give that a shot now
the link is clearly incorrect :-S
mchehab: ^^
meow, it bounced again
can try from yahoo or gmail..
when sending to linux-media@vger.kernel.org ?
yeah
Delivery has failed to these recipients or groups:  linux-media@vger.kernel.org Your message wasn't delivered because the destination email system rejected your message for security or policy reasons. For example, the email address might only accept messages from certain senders, or it might not accept certain types of messages, like those larger than a specific size.  Contact the recipient (by phone, for example) and work with them and their email a
e what policy or setting blocked your message and what you should do to make sure that future messages from you won't be rejected.  For more information, see Status code 5.7.1.
can you send the patch to laurent.pinchart@ideasonboard.com ?
the exact same way you tried to send it to linux-media@vger.kernel.org
I'll check if I see anything wrong
sent
not bounced =)
it's an HTML e-mail
also, the body of the e-mail contains
From 6e89b8808bcb5b90de80146968694794e53907e1 Mon Sep 17 00:00:00 2001
From: Pete Hemery <petehemery@hotmail.com>
Date: Thu, 8 Jul 2021 19:02:40 +0100
Subject: [PATCH] media: gspca/sn9c20x: Add ability to control built-in webcam LEDs
that shouldn't be in the body
D: ok =/
please use git send-email instead of thunderbird, you'll have many issues
will try, cheers
FYI, another place the e-mail I tried is linked: http://vger.kernel.org/vger-lists.html#linux-media
majordomo@vger.kernel.org is the right address to send a subscription request, but it's not the list address
and I suspect you subscription request was an HTML e-mail too
View-->Message Body As-->Plain Text
may help
(see Documentation/process/email-clients.rst)
ok, I've been semi drunkenly struggling with git send-email.. do you think removing the header you mentioned and sending as plain text is worth a shot?
if not I can try again when I have less tired eyes on
meh, seems to be Hotmail being the problem =/ it's become a tomorrow problem. Thanks for your help pinchartl =)