ChanServ changed the topic of #cedrus to: Cedrus development channel (http://linux-sunxi.org/Cedrus) | *only registered users can talk* | Logs at https://oftc.irclog.whitequark.org/cedrus/
Daanct12 has joined #cedrus
Daanct12 has quit [Read error: Connection reset by peer]
Daanct12 has joined #cedrus
Daanct12 has quit [Quit: Quit]
<linkmauve> Hi, I’ve written a JPEG decoder, do you know of a userland program I could test it with?
<gamiee> Hi linkmauve. Where you integrated it ?
<linkmauve> I added a cedrus_jpeg.c file, and did pretty much the same changes as for VP8 support.
<linkmauve> I’m looking for a V4L2 userland tool which would take a JPEG file and output the raw file, in NV12 format from what I understand of the driver.
<gamiee> This is good question, I didn't even know that V4L2 req api supports JPEG decoders
<linkmauve> Ah hmm, perhaps I’ll have to add that too.
<linkmauve> I’ve found multiple other JPEG decoders in the tree (I’m working on top of Linus’s master), but didn’t check whether they use the requests API.
jernej has quit [Remote host closed the connection]
jernej has joined #cedrus
<linkmauve> gamiee, how do you determine whether a driver is stateless?
<linkmauve> Ah, MEDIA_CONTROLLER_REQUEST_API.
<linkmauve> gamiee, anything I should keep in mind while adding JPEG decoding support?
<linkmauve> I don’t think we need any control or anything, the output should exactly match other decoders AIUI.
<gamiee> linkmauve: tbh, no idea, I don't that much about V4L2 Request API internals.
<gamiee> I don't know *
<linkmauve> For encoders it will be something else, but for now I’m focusing on just decoding, as I want pictures to appear faster on my phone. :)
<linkmauve> Alright, thanks anyway!
<gamiee> np, good luck :)
<linkmauve> Thanks. ^^
<linkmauve> Would it even make sense to still do the JPEG header parsing in the kernel, when we could require the request client to provide it as metadata alongside?