ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development - Did you try looking at our wiki? https://linux-sunxi.org - Don't ask to ask. Just ask and wait for an answer! - This channel is logged at https://oftc.irclog.whitequark.org/linux-sunxi
aggi_ has joined #linux-sunxi
vagrantc has joined #linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
bauen1 has quit [Ping timeout: 480 seconds]
jernej has quit [Quit: Free ZNC ~ Powered by LunarBNC: https://LunarBNC.net]
jernej has joined #linux-sunxi
jernej has quit [Remote host closed the connection]
cnxsoft has joined #linux-sunxi
jernej has joined #linux-sunxi
mripard_ has quit [Remote host closed the connection]
mripard has joined #linux-sunxi
Daanct12 has joined #linux-sunxi
Danct12 has quit [Ping timeout: 480 seconds]
vagrantc has quit [Quit: leaving]
Daanct12 has quit [Remote host closed the connection]
JohnDoe_71Rus has joined #linux-sunxi
Danct12 has joined #linux-sunxi
apritzel has joined #linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
cnxsoft has quit []
cnxsoft has joined #linux-sunxi
apritzel has joined #linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
bauen1 has joined #linux-sunxi
apritzel has joined #linux-sunxi
veremitz has quit [Read error: No route to host]
veremitz has joined #linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
apritzel has joined #linux-sunxi
rajkosto has joined #linux-sunxi
Benjojo__ has quit []
Benjojo has joined #linux-sunxi
cnxsoft has quit []
ftg has quit [Ping timeout: 480 seconds]
pcBob has joined #linux-sunxi
<pcBob> I am struggling to get USB working on the F1C100s, according to the support matrix in sunxi wiki it is "Work in Progress" - what is the current status on this?
<pcBob> But which kernel version are they targeting?
tuxd3v has joined #linux-sunxi
<tuxd3v> hello guys, can anyone help me to understand the diferences between Hantro and Cedrus drivers?
<tuxd3v> are they for the same thing?
<gamiee> hi tuxd3v, no they are not
<tuxd3v> or Hantro uses Cedrus beneath
<tuxd3v> or underneath
<tuxd3v> :)
<gamiee> Hantro driver is for SoCs which are using "Hantro" IP Cores for video encoding decoding (for example Rockchips uses them, but also some others), and Cedrus is driver for Allwinner SoCs IP core for video decoding / encoding
<tuxd3v> gamiee, many thanks
<gamiee> you're welcome tuxd3v :)
<tuxd3v> but as I read, Hantro is now adding support for Allwiner H6 :/
<tuxd3v> :)
<gamiee> tuxd3v: what? Where ????
rajkosto has quit [Read error: Connection reset by peer]
<gamiee> Oh wait, you're right
<tuxd3v> I saw it on phoronix
<tuxd3v> but the commit doesn't talk about allwinner H6, only the article..
<gamiee> tuxd3v: okay, it looks like following: Allwinner H6 have H.264, H.265 and VP9 decoder (and some others). H.264 and H.265 is decoded by custom Allwinner IP Core (mostly mentioned as Cedar), but Allwinner doesn't have IP Core for VP9, so they purchased Hantro VP9 Decoding core, which is present in H6.
<tuxd3v> ho... many thanks for the explanation :)
<tuxd3v> I was confuse :)
<gamiee> I was firstly confused too haha, but now it giving a sense. But still, let's ping @jernej if I explained it correctlx
<tuxd3v> :)
tuxd3v has quit [Read error: Connection reset by peer]
tuxd3v has joined #linux-sunxi
vagrantc has joined #linux-sunxi
megi has joined #linux-sunxi
<jernej> gamiee, tuxd3v: that is more or less correct. H6 has two video decoding cores, Cedrus and Hantro G2. Well, Cedrus can also encode.
<gamiee> @jernej: hi and thank you for clarification. Does this means that H6 have also H.264 and H.265 decoder from Hantro core ?
<jernej> In my opinion, G2 was added because Cedrus has not finished VP9 support. For example, H616 has only Cedrus, which also supports VP9.
<jernej> G2 does support VP9 and maybe HEVC too, but nothing else.
<tuxd3v> jernej, many thanks,then for what I understand H6 has Cedrus encode/decode for h26{4,5}, and Hantro for vp9 decode?
<gamiee> Ahh, interesting, and yes, most likely it wasn't finished
<jernej> I haven't tested HEVC yet, current driver would need some adjustments, because it's a bit older variant
<jernej> at one point Google gave G2 cores for free to SoC designers to speed up VP9 adoption
<jernej> and it's now mandatory for Android certification
<gamiee> Jernej, please, do you know how was this VP9 integration was done in BSP SDK? It was implemented into CedarC library ?
<jernej> yes, there is vp9hw plugin which uses it
<jernej> (cedar library was restructured a long time ago to be more modular, all codecs are implemented as plugins)
<jernej> before, there was only vp9soft plugin
<gamiee> Cool, thanks for the info! Btw fun fact, until V831 SDK, CedarC had all encoders inside libvencoder.so , but now it's same separated as decoding plugins, and the encoding "layer" was open-sourced
<jernej> I forgot already how the plugin is called on H616
<jernej> tuxd3v: yes, at least it's used in such way in BSP
<jernej> note, there are two Hantro cores, G1 for a bit older codecs and G2 for newer
<jernej> I think there are even newer variants now, but I'm not sure
aggi_ has left #linux-sunxi [#linux-sunxi]
<jernej> gamiee: this common layer is most probably not very interesting from RE standpoint, at least that is the case for decoders
aggi has joined #linux-sunxi
<tuxd3v> jernej, many thanks :)
apritzel has quit [Ping timeout: 480 seconds]
<gamiee> @jernej: yeah, but in case of encoding where everything is in one library, I reversed the common layer, although source is available from the source released by AW from 2015
apritzel has joined #linux-sunxi
JohnDoe_71Rus has quit []
apritzel has quit [Ping timeout: 480 seconds]
vagrantc has quit [Quit: leaving]
gamelaster has joined #linux-sunxi
gamelaster has quit [Remote host closed the connection]
gamiee2 has joined #linux-sunxi
<libv> gamiee: the reason why that was open source was because allwinner was caught redhanded
<libv> using LGPL code from ffmpeg and something they stole from a company later bought by google
<gamiee> @libv : ahh yes, of course, they doing it still
<libv> that shitstorm happened just before NextThingCo did their 9usd board kickstarter, and since they were an american company, they could actually be held liable for spreading that code, and it was really bad publicity on kickstarter
<libv> so that then triggered nexthingco to pay free-electrons, now bootlin, for working on whatever allwinner A13 the variant was called with the ram included in the same package
<libv> </history lesson ;)>
<gamiee> LOL, good to know, I didn't absolutely know about this. Only thing I remember is exploit in their kernel to get root access
<libv> that was nothing, just some stupidity
apritzel has joined #linux-sunxi
<libv> the cedarx and other GPL violations were quite more severe and some actions were very deliberate
jernej has quit [Remote host closed the connection]
<gamiee> hmm it stills quite surprises me how NTC get Allwinner to make their own spin of SoC
<buZz> GR8 you mean? afaik that was just 'order >x , we'll laser another letter on it for ya'
<gamiee> buZz: yeah
<buZz> there wasnt anything different from the R8 that i'm aware of, are you suggesting there was?
<gamiee> afaik it's same
evadot has quit [Quit: leaving]
evadot has joined #linux-sunxi
<libv> gamiee: NTC was started with some former allwinner employees
<libv> i think allwinner wanted an rpi story for itself
<buZz> libv: wasnt that tom cubie ? :)
<buZz> the 'allwinner pi' dreamboy
<libv> the fact that it had that with the cubietruck, and then totally stonewalled the community that came to exist around it, they overlooked that
<buZz> hehe
jernej has joined #linux-sunxi
<buZz> ah yeah , cubieboard 3.x
<gamiee> libv: what happened with cubietruck ?
<buZz> not much
<buZz> basically only support for cubie* was this channel
<libv> buZz: were you involved then?
<libv> gamiee: tom cubie started with a crowdsource campaign that was a resounding success
<libv> he ended up starting a business around it, but then ended up falling out with his business partners (no details given)
<libv> then went on to do raxda rock
<buZz> libv: i had a cubieboard 1 and 2 sent to me from tomcubie himself yeah
<libv> we all here owe tom cubie a lot
<buZz> gifted em away, ppl still use em to this day!
<libv> without tom cubie, no mele a1000 would've been sold to the western world over ebay
<buZz> :)
<libv> tom also provided some other support that was somewhat grey
<buZz> preinstalled kodi with warezstuff?
<libv> that's not what i meant
<libv> and i have no idea about kodi
<libv> but without tom cubie, no A1000, no linux-sunxi; and then no cubieboard, or it's partial rip-offs banana pi, other $shitpis
<buZz> ah well, many companies reselling those android tv machines get sued by gov/copyright offices here
<libv> its even
<buZz> for them selling them with kodi + DRM breaking plugins preinstalled , sometimes even tor/vpn stuff for international streams
<gamiee> noooo wayyy :D
<gamiee> I guess you all know what Reuuimlla Technology is ?
<libv> buZz: tom cubie was an allwinner employee, who noticed the potential for the allwinner A10 and the mele A1000 with a uart port already soldered onto it
<buZz> never heard of
<libv> buZz: allwinner in reverse
<buZz> haha, ok
<buZz> Reuuimlla Technology Co., Ltd is an active company located in Sheffield, South Yorkshire.
<buZz> Registered Address 12 Allwinner Drive
<buZz> that seems fake :P
<gamiee> That was the first thing I wanted to say, someone estabishled a company with this name in UK :D :D
<buZz> a sir 'Johnny James Bravo
<buZz> lol
<buZz> johnny bravo , riiiiight
<gamiee> Yeah just reading it, lmaooo :D how is this even possible?
<gamiee> But even more funnier, I thought this is only in old BSP kernels... But I found two source codes with this even in mainline
<buZz> with 'this' ?
<buZz> ah, that name
<libv> David LanzenDoerfer was claiming to work for allwinner around the time the cedarx lgpl shitstorm hit
<gamiee> also, owner of reuuimllatech.com is some chinese guy :D
<gamiee> Like a year ago, in this domain, there was same allwinner page, but horizontally mirrored :D
<libv> he was not happy with the shitstorm that was mainly driven by myself, but he ended up "seeing the light" a year or so later, when whatever he was trying to achieve on his own also hit a wall
<libv> he's still active in other projects, he has a talk for the upcoming fosdem
<libv> not sure what about, but i have seen his name pop up in the mls
<libv> so that reuuimlla mmc code was probably a rework of existing allwinner code
<libv> probably in an attempt of allwinner to try to gain control of the upstream story?
<libv> anyway, this was a time when allwinner lines of communication were opened, then closed, then opened then closed, and nothing useful ever happened
<gamiee> Still I wonder why not anyone changed the texts to normal form.
<gamiee> it never worked
<gamiee> And probably newer work
<libv> which is why the emergency brake was pulled with the lgpl violations
<buZz> www.reuuimllatech.com was already inside those files in 2017
<libv> that did not happen in a vacume, other routes were tried before
<libv> somewhere after the 2015 shitstorm allwinner actually paid into linaros set top box or whatever project
<gamiee> buZz: it's here since 2014 when David Lanzendorfer commited it
<libv> which i was then critical of, as it seemed as if allwinner then considered itself a fully paid up member of the open source community
<buZz> :)
<buZz> good
<libv> and then everything died properly with allwinner
<buZz> i was looking in the sun8i one ;)
gamiee2 has quit [Ping timeout: 480 seconds]
<apritzel> I heard that Allwinner's Linaro involvement "coincided" with some Chromebook "call for bids"?
<apritzel> but they never showed up at any Connects, AFAIK, even though some engineers were officially registered
<libv> iirc, rockchip was already a staple for chromebooks back then
<libv> or was that just after that?
<libv> but interesting, i had not heard that
<libv> heh, march '15 is when the rk3288 invasion came
<libv> arnd explained a few months ago that allwinner just paid in, and then nothing, and then the membership lapsed after 2ys
<libv> but interesting, the misguided attempt at being open in 2014/2015 could've completely been triggered by the call for bids of google
<libv> ah, and the joining of linaro was march 2014
<libv> so yes, sounds very plausible
megi has quit [Quit: WeeChat 3.3]
megi has joined #linux-sunxi
megi1 has joined #linux-sunxi
megi has quit [Ping timeout: 480 seconds]
<libv> lemaker.org is a lapsed domain now
<libv> interesting
<libv> filled with what seems to be some manga stuff now
ftg has joined #linux-sunxi
<libv> "establishing the open source innovation ecosystem", right.
<juri_> i smell bullshit.
<libv> ?
<libv> oh, from bpi, yes, this was clear from the start
<libv> it was a matter of time before this $shitpi vanished
<libv> i'm surprised orangepi is still alive
pcBob has quit [Remote host closed the connection]