mcbrick has quit [Remote host closed the connection]
SamSagaZ has left #linux-rockchip ["Once you know what it is you want to be true, instinct is a very useful device for enabling you to know that it is"]
Tsvetan2 has joined #linux-rockchip
Tsvetan3 has quit [Ping timeout: 268 seconds]
sjoerd` is now known as sjoerd
sjoerd is now known as sjoerd`
menomc has joined #linux-rockchip
mnemoc has quit [*.net *.split]
menomc is now known as mnemoc
Astralix1 has joined #linux-rockchip
Astralix has quit [Ping timeout: 264 seconds]
eebrah has joined #linux-rockchip
sjoerd` is now known as sjoerd
sjoerd is now known as sjoerd`
Tsvetan3 has joined #linux-rockchip
Tsvetan2 has quit [Ping timeout: 240 seconds]
atiti has joined #linux-rockchip
Tsvetan3 is now known as Tsvetan
deviker has joined #linux-rockchip
jeremb_ has quit [Ping timeout: 257 seconds]
_brain has joined #linux-rockchip
<_brain>
yo
rellla has joined #linux-rockchip
AstralixNB has joined #linux-rockchip
AstralixNB has left #linux-rockchip [#linux-rockchip]
AstralixNB has joined #linux-rockchip
massi_ has joined #linux-rockchip
utente has joined #linux-rockchip
<utente>
hello, im waitig to receive radxa board based on RK3188. I see there is linux 3.0.66 avaiable, but i would like to know if a more recent (i.e.3,10) is near to be released with rk3188 patchs
<AstralixNB>
There are mainly two development activities on the RK front. One is kernel 3.0.xx from RK and current Android / Linux developers. The other is a mainline integration of RK chips, where only 1.5 people work on (I count myself only half, as I am currently busy with different things.)
<AstralixNB>
The mainline integration started around 3.10 and is currently based on 3.11 if I am correct.
<_brain>
I'm also working on my own kernel where I have essentially taken what I need from the Android kernel and put it in a 3.0.65 kernel
<_brain>
It's working, and I'm getting rid of lots of android stuff
<AstralixNB>
does it make sense to fork another kernel?
<_brain>
how do you mean?
jeremb_ has joined #linux-rockchip
<AstralixNB>
We already have several repos online... so why doing the job again instead of supporting current work and bring it to another step forward
<AstralixNB>
the 3.0.36+ kernel from RK is based on mainline 3.0.66 as I was told by RK engineers
<utente>
if 3.0 is stable and working with all hardwaree, there is no many need to move to 3.10 or later. but if there are some lack of functionality, to mobe haead is a need. of course in function of manpower avaiable. im not a kernel haker, just a end user. my final goal is to have rk on radxa board fully functional. if the kernel is 3.0 or 3.10 became not so important, the focus is the 100% functionality.
<AstralixNB>
Yes and no
<_brain>
That makes sense, as thats why I went on using 3.0.65. Problem is I have only found repos with lots of messy code. The RK engineers haven't done this well at all.
<AstralixNB>
There are interesting new features in 3.10+ if you only think about kernel flash filesystem.
<_brain>
Not to mention there is no git history on whatever work they have done
<AstralixNB>
So 3.10 brings you an option to get rid of rknand drivers
<AstralixNB>
It is chinese way of coding
<_brain>
The reason why I did this and forked yet again is so that I have a somewhat clean kernel with commit history, making it easier to merge fixes from mainline
<AstralixNB>
There is no diffference comapred to Allwinner, ESS, Mediatek
<utente>
AstralixNB, "chicense way of coding" maybe means "crap soft to support nice hardware"?
<AstralixNB>
Yes
<AstralixNB>
ugly dirty code
<utente>
i got the same with cubieboard: a nice boarda d poor software support.
<AstralixNB>
I guess they are just payed in lines / minute
Theueip has quit [Quit: Leaving...]
<utente>
the best way to produce poor qualito code :(
<AstralixNB>
It has not changed since I worked on ESS DVD code back in 2006
<_brain>
It might be better to patch vanilla with whatever RK-code is needed, and not use their released code
<utente>
i think _brain said the true
<_brain>
So for me I only have the options; help with 3.10, or use my fork of 3.0.65 that adds RK stuff
<AstralixNB>
_brain, mmind00 and I do exactly this. For getting RK mainlined we write code from scratch
<AstralixNB>
But in 3.10 things are pretty different. There is no arch/cpu thing anymore as, like Linus Torvalds said, that ARM thing ist a pain in the ass.
<AstralixNB>
In RK Cores is nothing new compared to other cores. So many IPs are already existent and you need to book them via dts system.
<_brain>
the best would obviously be if RK would release most, if not all, technical docs for their platforms
<_brain>
:>
<_brain>
in the end it would be easier for them, having the community write the drivers and get it into mainline
<AstralixNB>
I am not sure if it really brings you an advantage
<_brain>
any 3.10 code I can look at, by the way?
<AstralixNB>
It would be a good thing to leave documentation to people who know to make it as well as you should leave programming to guys who know about that. But they only have technical people with some documentation skills and some programming skills...
<mnemoc>
_brain: in the case of allwinner they only started to provide *some* documentation after mach-sunxi hitted mainline
<AstralixNB>
go to my github?
<AstralixNB>
github.com/Astralix and check the link to mmind00 too#
_whitelogger has joined #linux-rockchip
<_brain>
great stuff
<_brain>
what's the state of the code btw? bootable?
<utente>
_brain, S.o.C. producer are gelous of their docs, and dotr realize thart not share them mens less selling of thei oroiducts. i rememnber te fanny story where husband, cut avay his "man detail" to avoid hes wife enjoy it.
tonikasch has joined #linux-rockchip
<mnemoc>
utente: they also fear to be sued by some patent trolls if they give too much info
<utente>
mnemoc, i know that "enemy" can get profit from your info, but if you dont share, also "friend" get damaged by lackness of info and your soc will be seen as not good enough simply because coders cannot write dedicated code.. at my eyes, they are not able to balance "friend" and "foe"
<mnemoc>
utente: layers can do much more harm than the help friends can give
<mnemoc>
lawyers*
<utente>
no problem, apparently freescale are more "haker friendly", so people will move on those SoCs.
<utente>
at the momen they are still at 40nm, but i home next generation will move to 28ns as RK is. so bye bye RK and welmome i.MX6. simple.
<mnemoc>
utente: there is a channel for imx6 work too
<utente>
nice to know. as soon as i will have that soc, i will join the channel.
<AstralixNB>
_brain, the wip branch did not boot as we had some issues to implement the 64bit counters. Then we encountered that there is already a 64 bit counter devel ongoing and we waited for it to be inserted in 3.11.
<AstralixNB>
but RK3066 boots with that code
<AstralixNB>
I'll catch up with 3.11 as soon as my radxa board arrives.
<tonikasch>
3.11?
<AstralixNB>
3.12, sorry
<tonikasch>
mmm, anyway, android branch, I guess
<AstralixNB>
ah, you missed the discussion... we where talking about mainlining RK
<AstralixNB>
I hope Android and mainline kernel will merge some day...
<tonikasch>
me too
<tonikasch>
is easier to mainline RK now that radxa rock board is being sent to developers and some more specs are available?
<AstralixNB>
nope, there are no more specs than before
<tonikasch>
aha
<tonikasch>
:(
<AstralixNB>
you mean specs of the chip, not the board, I guess
<_brain>
AstralixNB: cool, I'll check out the code, maybe there is something I can even contribute with
<AstralixNB>
_brain, welcomed
morfoh has joined #linux-rockchip
<morfoh>
moin
<tonikasch>
so how would you clone only kernel source for 3.12? I'm not used to Android
<AstralixNB>
moin :)
<AstralixNB>
tonikasch, github.com/Astralix or mmind00, just hook up on that, we did that for you already ;)
<tonikasch>
wow!
<AstralixNB>
The problem isn't to get a fork out of kernel.org
<AstralixNB>
The problem is to get the new code back into kernel.org
<tonikasch>
I see...
<AstralixNB>
we should work on that silent and then leave it to some people to post the patches on the mailing list
<AstralixNB>
None of ou will get a line of code into the kernel on the first run
<AstralixNB>
You should have some experience of how to pass code to the responsible maintainers
<AstralixNB>
I am in a delay as I am pretty busy in my official job
<AstralixNB>
But I'll start over these days again
<tonikasch>
oh, ok! didn't know if I was looking at the wrong repo
<tonikasch>
:)
<AstralixNB>
no, you're on the right repo.
<AstralixNB>
I have to pull the latest changes and move up to 3.12-rcX
<AstralixNB>
Hope that I can do that till end of the week
<tonikasch>
aha
<tonikasch>
mmm, I see, a lot of coding needed but for the moment I haven't got the needed skills
<AstralixNB>
kernel ist probably bare metal coding. That is a different thing than to start on top of a system
<AstralixNB>
I don't know if the one or the other is more difficult, but it is different
<tonikasch>
I'll see if I can do something but I'm still studying the computer degree...
<tonikasch>
well, have to go, bye
tonikasch has quit [Quit: Bye!]
eebrah has quit [Ping timeout: 265 seconds]
utente has quit [Read error: Operation timed out]
utente has joined #linux-rockchip
deviker has quit [Remote host closed the connection]
Galland has joined #linux-rockchip
lunra has quit [Remote host closed the connection]
eebrah|noPC has joined #linux-rockchip
eebrah|noPC has quit [Ping timeout: 252 seconds]
Tsvetan2 has joined #linux-rockchip
Tsvetan has quit [Ping timeout: 268 seconds]
ganbold_ has joined #linux-rockchip
tonikasch has joined #linux-rockchip
<tonikasch>
Galland: I made a new commit with recent rtl8188etl (rtl8188eu) drivers in my github, should I cancel (if possible) the previous pull request and make a new one with latest commit?
leowt has joined #linux-rockchip
<Galland>
don't worry, I'll take care of that
<Galland>
where did u get those new drivers and do they work?