vinifr has quit [Remote host closed the connection]
Seppoz has quit [Ping timeout: 264 seconds]
ojn has quit [Ping timeout: 264 seconds]
awafaa has quit [Ping timeout: 240 seconds]
jeremb has quit [Ping timeout: 256 seconds]
steev has quit [Ping timeout: 256 seconds]
jeremb has joined #linux-sunxi
steev has joined #linux-sunxi
arete74 has quit [Ping timeout: 264 seconds]
arete74 has joined #linux-sunxi
ZetaNeta has quit [Ping timeout: 260 seconds]
arete74 has quit [Ping timeout: 248 seconds]
_hipboi_ has joined #linux-sunxi
hipboi_ has quit [Ping timeout: 260 seconds]
Tsvetan has joined #linux-sunxi
arete74 has joined #linux-sunxi
arete74 has quit [Ping timeout: 245 seconds]
boycottg00gle has joined #linux-sunxi
ojn has joined #linux-sunxi
awafaa has joined #linux-sunxi
<boycottg00gle>
i managed to kill hdmi output using a10disp (https://github.com/hglm/a10disp) on a olimex a20 (a10disp uses /dev/disp ioctls) - now the strange part: the problem is persistent / long power off didn't help - any suggestions?
\\Mr_C\\ has quit []
Colani1200|work has joined #linux-sunxi
deasy has joined #linux-sunxi
bsdfox\ has quit [Read error: Operation timed out]
arete74 has joined #linux-sunxi
n01 has joined #linux-sunxi
qlex has joined #linux-sunxi
* qlex
says hi
<qlex>
can any1 point to linaro/ubuntu/android images for marsboard a20 ?
<qlex>
i cant seem to find them on linux-sunxi.org
<qlex>
and on marsboard.com i think they only mention a10 board, and not a20
<oliv3r>
mornin' all
<oliv3r>
back @$work
rellla has joined #linux-sunxi
rellla has quit [Client Quit]
rellla has joined #linux-sunxi
arete74 has quit [Ping timeout: 256 seconds]
arete74 has joined #linux-sunxi
panda84kde has joined #linux-sunxi
rz2k has joined #linux-sunxi
atiti has joined #linux-sunxi
_hipboi_ has quit [Ping timeout: 276 seconds]
<qlex>
nobody to help with my a20 marsboard - when i "burn" the image on sd card (linaro from marsboard), board doesnt boot
eagles0513875_ has joined #linux-sunxi
<eagles0513875_>
hey guys im trying to use build root on my cubieboard 2 a20 board it seems like there is only an a10 port what do i need to do to get this woorking on my a20
<steev>
eagles0513875_: well it would help to explain what "this" is
qlex1 has joined #linux-sunxi
<eagles0513875_>
steev: i mean the buildroot images that i compiled and put on the microsd card for my cubieboard
qlex has quit [Ping timeout: 264 seconds]
<steev>
eagles0513875_: well, i can't really help, i don't have a cubieboard yet - just bought mine last night, but you should definitely list what all you've tried to do that has failed if you want assistance
_hipboi_ has joined #linux-sunxi
<eagles0513875_>
steev: buildroot from git has support for the cubieboard but the a10 it seems like i managed to compile and copy everything to the micros yet there is no support for the a20
<boycottg00gle>
managed to build booting u-boot and kernel for olimex a20 following http://linux-sunxi.org/Building_on_Debian using u-boot current branch and stage/sunxi-3.4 kernel branch - unfortunately somehow killed hdmi now
<boycottg00gle>
i really don't understand how using ioctl's could kill hdmi persistently
<boycottg00gle>
being a hardware noob: could i have killed the hw?
<eagles0513875_>
steev: there is no mention of the a20 on that first link
<eagles0513875_>
steev: i love the board but android which comes default doesnt suite what im trying to accomplish
<eagles0513875_>
ill eventually need to get the cubie trucker
<mnemoc>
cb2 -> cb + A20
<steev>
eagles0513875_: uh, the repository is from cubieboard2...
<steev>
mnemoc: hola!
<steev>
ltns
<steev>
(mostly on my end)
<eagles0513875_>
ahh ok
<eagles0513875_>
steev: would you recommmend trying out that build script you linked me above?
<steev>
maybe try the stuff from this repo eagles0513875_ ? like i said though, i don't have a board, and have no experience, so definitely wait to see what others say
<mnemoc>
steev: yeah.... very long. busy at work or playing with other toys?
<steev>
mnemoc: both?
<mnemoc>
:)
<eagles0513875_>
thanks steev :)
<steev>
eagles0513875_: i'd try it myself, but the most i could do was prove that it built something
<panda84kde>
Good morning everybody! Is the Mali400 on the A10 overclockable? If it is, anybody tested?
<eagles0513875_>
mnemoc: ahh ok :) thats much clearer now hehe
<mnemoc>
eagles0513875_: still, I expect a large migration OUT of the clouds
<eagles0513875_>
ya im heading that way as well :)
<steev>
mnemoc: i don't
<eagles0513875_>
mnemoc: migration to where though if out of the clouds
<eagles0513875_>
technically the article you linked makes sense in the sense that you are giving up privacy for your email to get from destination a to b as the headers of the email have to be read
<steev>
i've been against the cloud for ages (pretty much since people started calling it "the cloud" - though really i just *really* hate that buzzword (and i give my buddies that work at Rackspace shit about it constantly))
<mnemoc>
internet connections are faster and cheaper.... it's not that hard for mid businesses to host their own private "cloud" anymore
<steev>
which has to happen for any email, not just gmail
<steev>
kinda how email was designed
<eagles0513875_>
im not hosting my own cloud i use a vps provider but still i have control over my server
<eagles0513875_>
which you guys are reminding me i need to fix my postfix setup to work with my ssl cert
<steev>
eagles0513875_: the point is, if you use email, the headers HAVE to be disclosed so that the mail can be routed at all, for anyone to expect that to be private is... looney
<eagles0513875_>
steev: in the article mnemoc mentioned the previous statement was rather mis leading in the sense it wasnt clear that only the headers are being read by 3rd parties
<steev>
eagles0513875_: unless the email is encrypted, the entire mail is sent plaintext so... anyone sniffing along the line could, at least in theory, read it
<mnemoc>
well.... for spam filtering content IS read
<eagles0513875_>
steev: agreed. thats another pain in the ass how gmail doesnt support people using pgp keys
<steev>
mnemoc: for certain definitions of read
<mnemoc>
:)
<steev>
it's not like google pays people to physically read each email (sign me up, if they do)
<mnemoc>
of course it's not people, but software does analyse and collect (parts) of the content of your mails before it arrives to your mailbox
<mnemoc>
which at least for me it's fine
<eagles0513875_>
mnemoc: so they can tailor make the advertisements
<steev>
idgaf myself, anything that needs to be private isn't sent over the internet at all
<mnemoc>
eagles0513875_: the ads part is supposed to be done at UI leve
<mnemoc>
level*
FR^2 has joined #linux-sunxi
<eagles0513875_>
mnemoc: has to be as well at email level they need to know whats in there to tailor make the ads to the content of your email
<mnemoc>
eagles0513875_: yes, but iirc mail is read while you see it on your browser, and sent back to the ads server, where the tailored ads are rendered for you
<mnemoc>
not internally from your mailbox
notmart has joined #linux-sunxi
notmart has joined #linux-sunxi
<Colani1200|work>
Hi all
<steev>
mnemoc: so how's mainline support going?
<eagles0513875_>
mnemoc: do you have any other suggestion in terms of getting a different system on the a20
<Colani1200|work>
Does the A10 have a temperature sensor that can be read e.g. in Debian?
kaspter has quit [Ping timeout: 248 seconds]
geecko has joined #linux-sunxi
<mnemoc>
pretty well afaik, but I've been very away for the last 3-4 months
fluo75|2 has joined #linux-sunxi
fluo75|2 has quit [Client Quit]
fluo75 has joined #linux-sunxi
* qlex1
sees more activity on chat - so might as well ask now
<qlex1>
could any1 point me to a working images (for SD) for marsboard a20 ?
<steev>
mnemoc: ah, work or other toys? :P
BluesBoy has joined #linux-sunxi
<qlex1>
i just put linaro on sd and my marsboard doesnt even start (getting "no signal" on hdmi monitor)
<eagles0513875_>
qlex1: im trying to get a bare bones root setup on mine sadly buildroot doesnt support the a20 for my board yet :(
<qlex1>
once i put it out, marsboard a20 starts with android on nand
<mnemoc>
steev: work and settling down in berlin
<eagles0513875_>
qlex1: sounds like me are you on a cubieboard 2
<qlex1>
my cubie2 is already out on tests in production environment using stock android
<steev>
mnemoc: whoa, berlin eh?
<steev>
weren't you in HK?
<qlex1>
but i have marsboard a20 here in office
<mnemoc>
steev: spain :)
<steev>
oh
<qlex1>
chipset is the same, but i guess images from cubie2 wont work on mars2 ?
<eagles0513875_>
qlex1: :-/ was hoping you could potentially help me get rid of android on mine and put a bare bones root system
<mripard>
eagles0513875_: I don't remember very well how it the support is made in buildroot, but I guess it's just a matter of changing the kernel branch/tree/defconfig, and the fex file to use, isn't it?
<mnemoc>
i've never gone to asia
shineworld has joined #linux-sunxi
<eagles0513875_>
mripard: i followed the simple readme it has for the board let me paste it for you hold on
<steev>
qlex1: i did a quick google search and saw something about the cb2 images working for the marsboard, but, i have neither so i can't say (and thus why i didn't answer when you asked)
<mripard>
qlex1: did you plug a serial on your board to see if it's not booting at all or just failing to bring up the HDMI?
<eagles0513875_>
maybe thats what i should do too lol
<eagles0513875_>
on the cubie 2 how should i plugin the serial its serial to usb but what device would i plug it into as my monitor doesnt have any usb ports if thats necessary
<mripard>
eagles0513875_: ok. so, change the buildroot configuration once you have loaded the cubieboard_defconfig, and change it to use a kernel for the A20, and the fex file for the cubie2
<mripard>
I don't know what's the kernel tree to use for now for the A20
<mripard>
hans one?
<qlex1>
mripard: i got a serial cable (UART) that i used for APC board - but i dont know if it'd work for mars ?
<n01>
eagles0513875_: back ... yeah that's me ... as mripard said should be enough modify defconfig and such
<qlex1>
it has RST,3v3,5v,txd and gnd bolts
<eagles0513875_>
n01: noobie question how as this is all new to me re embedded systems
<eagles0513875_>
n01: should i hook up my serial cable and see why nothing is loading?
<n01>
eagles0513875_: that is the first step :)
<eagles0513875_>
n01: do you mind if i pm you please
<mripard>
qlex1: well, it depends if it's a TTL or a RS232 one
<eagles0513875_>
as i have a hunch im going to drive everyone crazy with my noobie questions
<n01>
eagles0513875_: better if we talk here. this is totally IT
<eagles0513875_>
mripard: im having 2nd thoughts about that
<eagles0513875_>
im wondering if open embedded would get me up and running quicker
<eagles0513875_>
i think im going to try it
<mripard>
I find OE way more complicated, but it's your call :)
<eagles0513875_>
how so?
<n01>
mripard: could you ping my patch? I'm getting tired of being ignored :(
<mripard>
n01: yeah, I'm just back from vacation
Tiny-noob has joined #linux-sunxi
<mripard>
I saw your message from yesterday, that was on my today-todo-list :)
<n01>
thx :)
<eagles0513875_>
ok so i guess time to read the manual
<eagles0513875_>
just really want to get this system up and running asap
<mripard>
eagles0513875_: buildroot uses basically the UI/configuration system used in a lot of other places, like the kernel, uclibc, crosstool-ng, etc.
<n01>
eagles0513875_: if you need help in modifying my recipe just ask
<mripard>
OE has gazillions features, and configurations system of its own
<eagles0513875_>
ahh ok :)
<eagles0513875_>
n01: what should the first step for it be?
<mnemoc>
and dominated by yocto anyway
<n01>
connect a serial cable :)
<eagles0513875_>
ok
<mripard>
qlex1: it should be fine then yes
<eagles0513875_>
can tell you for sure steev that the default android on the cubie 2 isnt optimized enough for the hardware
<qlex1>
mripard: are there any tutorial which ones and where i should plug (meaning which cables in uart cable
Soru has quit [Read error: Connection reset by peer]
<eagles0513875_>
next nubbish question what colour goes onto what pin on the board my cable has no labels
<mripard>
qlex1: on the marsboard? none that I'm aware of
Soru has joined #linux-sunxi
<qlex1>
mripard: what do you mean by "none". its marsboard a20, i think there is UART on it
<mripard>
I'd look at either the PCB or the schematics, and plug TX to TX, RX to RX, and GND to GND.
<qlex1>
ok
<qlex1>
and 3v3 or 5v ?
<mnemoc>
3v3
<mnemoc>
just skip connecting the vcc
<qlex1>
vcc ?
Seppoz has joined #linux-sunxi
<eagles0513875_>
n01: next nubbish question im on an ubuntu vm on imac it picked up the serial to usb adapater. i just installed putty how do i go about connecting to it
<mnemoc>
qlex1: the 4th pin, which tends to bring 5v and when 3v3 can make your board not power down
<eagles0513875_>
even mnemoc ^ included in the above question
<n01>
eagles0513875_: in vm you should have something like /dev/ttyUSB0 ... use that dev to connect using minicom/putty
Colani1200|work has quit [Ping timeout: 264 seconds]
<mripard>
eagles0513875_: yes
<eagles0513875_>
ok that is probably why things arent working
<mripard>
to change the kernel and fex file you'll be using
<eagles0513875_>
mripard: does that mean i dont need to run make def conf or what ever that command is?
<n01>
eagles0513875_: first you run defconfig and then menuconfig to change stuff for A20
<mripard>
you need to run it, because the configuration only changes slightly
Seppoz has quit [Remote host closed the connection]
Seppoz has joined #linux-sunxi
<mnemoc>
there is a sun7i_defconfig
<eagles0513875_>
n01: ok so if i have already run it i just need to run make menu config at this point correct
<eagles0513875_>
and for some reason i get this rwxrwxrwx 1 root root 13 Aug 19 11:16 pci-0000:00:1d.0-usb-0:2.1:1.0-port0 -> ../../ttyUSB0 and using /dev/ttyUSB0 in putty it doesnt seem to work
Seppoz has quit [Remote host closed the connection]
<eagles0513875_>
n01: seeing as i ran defconfig already can the make menuconfig be run at this point
<n01>
eagles0513875_: yes you can run make menuconfig ... what does "it doesnt seem to work" mean?
<n01>
did u correctly setup baudrate & co?
<steev>
eagles0513875_: i don't have mine yet! soon
<eagles0513875_>
n01: i used default 96000 baud rate
<eagles0513875_>
steev: im just giving you some feed back from mine
<steev>
eagles0513875_: i have no interest in running android on mine :)
<n01>
AFAIK should be 115200
<eagles0513875_>
me neither steev
<eagles0513875_>
im workin on a project to develop a low cost point of sales system
<eagles0513875_>
so atm runningand testing stuff on cubieboard at this point
<eagles0513875_>
found it it compiled the kernel for the a8
vicenteH has joined #linux-sunxi
<eagles0513875_>
this should do the trick now :)
<mripard>
a8?
<eagles0513875_>
a20 is a dual cor a7
<eagles0513875_>
when i ran make menuconfig the default seems to be set there to an a8
<mripard>
what default?
<eagles0513875_>
the cpu default
<mripard>
ah
<eagles0513875_>
that is probably why it didnt work :)
<eagles0513875_>
that readme is a bit mis leading
<mripard>
so it compiles the userspace and toolchain with the cortex-a8 instruction set
<mripard>
it's not kernel related at all
<eagles0513875_>
as there is no mention of running make menu config after running the default config
<n01>
eagles0513875_: nope ... you still have to modify other stuff to make it working on A20
<eagles0513875_>
n01: :-/ like what
<boycottg00gle>
i think this should be "make menuconfig" not "make menu config" ?
<mripard>
eagles0513875_: of course. because you're reading the *cubieboard* readme
<mripard>
not the cubieboard2 one.
<n01>
eagles0513875_: I wrote that readme :)
<eagles0513875_>
n01: oh
<n01>
what are you complainig for? lol
<eagles0513875_>
lol nothing but dont you think there should be a mention of running make menuconfig before running make
<n01>
let me check
<eagles0513875_>
n01: im on latest git version available from master
<n01>
eagles0513875_: ???
<n01>
it is correct
<n01>
eagles0513875_: I think you shold really read the manual, no offense
<eagles0513875_>
none taken
Colani1200|work has joined #linux-sunxi
<n01>
eagles0513875_: are you the guy who wrote me an email?
<eagles0513875_>
n01: nope
<n01>
uh ok
qlex1 has left #linux-sunxi [#linux-sunxi]
mcbrick has quit [Quit: Leaving]
boycottg00gle has quit [Remote host closed the connection]
<eagles0513875_>
i must say i am learning alot here :)
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
eagles0513875_ has quit [Ping timeout: 250 seconds]
_hipboi_ has quit [Quit: Leaving]
<oliv3r>
time to open my mailling list mailbox
<oliv3r>
800 mails :(
<panda84kde>
oliv3r: I hope for you they're not 800 digests...
<oliv3r>
haha, i don't do digests :)
<oliv3r>
i find them highly annoying and un-replyable
<mripard>
oliv3r: back from holidays ? :)
<oliv3r>
was home for the past week, but yeah back at work now
<oliv3r>
so today i planned for backlog
<oliv3r>
tomorrow i'll do my a10 and a20 tests and then finally resend my patch :)
Tiny-noob has quit []
<jinzo>
I had to unsubscribe from linux-sunxi list, too much patches (which is very good) and not enough time to keep up
<oliv3r>
i did fix my PHY during vacation though :)
<oliv3r>
linux-sunxi is really low volume list though :p
<oliv3r>
linux-media is reasonably low but still causes some significant mails
shineworld has quit [Remote host closed the connection]
Super-noob has joined #linux-sunxi
Super-noob has left #linux-sunxi [#linux-sunxi]
Olaff has joined #linux-sunxi
Super-noob has joined #linux-sunxi
Seppoz has joined #linux-sunxi
<oliv3r>
mnemoc: around?
<mnemoc>
oliv3r: sort of
<oliv3r>
mnemoc: you've worked with lxc right
<mnemoc>
oliv3r: sort of :)
<oliv3r>
i'm playing with it now, and using the libvirt frontend, there it asks if i want to create a 'virtual operating system continer' or an 'application container'
<mnemoc>
i have two machines using it. but still voodoo
<oliv3r>
any clue on what the difference is?
<mnemoc>
i totally failed with virt-lxc
<oliv3r>
hehe, it seems to work so far :p
<mnemoc>
ended up doing everything manually, and with lxc-foo
<oliv3r>
i run libvirt in a qemu xubuntu guest :)
<oliv3r>
yeah I read abou that too
<oliv3r>
but wanted to manage the whole thing with libvirt; creating a container isn't possible though
<mnemoc>
and you need some fancy xml templates
<mnemoc>
yuck
<oliv3r>
yeah i kinda skipped that :p
<oliv3r>
i did create a debian base using qemu even i think ;p[
<oliv3r>
but any clue as to what difference an 'os container' vs an 'application container' is?
<mnemoc>
the second may not care abotu the OS and so you get whatever thing installed....
<mnemoc>
but no idea
<mnemoc>
test and see :)
<oliv3r>
yeah i tried googling, but wasn't quite obvious either
<oliv3r>
from what i understand, a lxc is a 'chrooted' /, so that's a full OS i'd say
<oliv3r>
if you'd only want to run some application from your host in an LXC, you'd have application virtualization, but then you still either need some shared access (shared libraries etc?) or god knows what
<mnemoc>
oliv3r: the only diff is that you don't run a separated kernel/hw emulation
<mnemoc>
but you can have a gentoo guest of a debian host in chroot
<oliv3r>
well lxc doesn't run any hardware emulation anyway
<oliv3r>
and gentoo in debian (or vice versa) sounds like OS application
Colani1200|work has quit [Ping timeout: 256 seconds]
<mnemoc>
with vserver/lxc you share kernel and hw, but userpsace can be different
<mnemoc>
app container might be more limited in the list of packages installed
<oliv3r>
i suppose, but how that differentiates itself ..
<mnemoc>
while virtual os impliees full userspace
<mripard>
or maybe it just uses the same rootfs, instead of chrooting in the application case ?
<mnemoc>
that makes sense. no chroot at all, only isoletaed app process
<mnemoc>
with less capabilities etc
<mripard>
(I don't know, but such a setup would make sense I guess)
shineworld has joined #linux-sunxi
<oliv3r>
i can see it as a sensible thing
<oliv3r>
i just trying to grasp how it would end up being installed
<oliv3r>
you start with a 'new' lxc, so you have no storage etc?
<mnemoc>
try and see
<oliv3r>
maybe an application container shares the rootfs like mripard said
<oliv3r>
mnemoc: yeah when i have time
<oliv3r>
700 mails to do :)
Colani1200|work has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
woprr has quit [Quit: Ex-Chat]
boycottg00gle has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
<oliv3r>
anybody here thinking of getting an ubuntu-edge?
shineworld has quit [Remote host closed the connection]
woprr has joined #linux-sunxi
<n01>
it will depend on iOS 7 :)
<mnemoc>
too expensve for a toy
<n01>
we should organize a linux-sunxi dev meetup some day
<mnemoc>
Tsvetan invited to sofia
hansg has joined #linux-sunxi
<mnemoc>
all inclusive except the transport to there
<n01>
oh O_O I missed it
<oliv3r>
how does it depend on iOS7?
<oliv3r>
mnemoc: really? I was in austria! which is so close!
<mnemoc>
ubuntu-edge =)
<oliv3r>
how does ubunutu-edge depend on ios7? ios7 will never be open, so that's a non-issue :p
<oliv3r>
i find it very expensive too, but still
<n01>
uhm, I could think about it
<mnemoc>
[linux-sunxi] [OT] Linux-Sunxi developers meeting at Openfest 2-3 November Sofia, Bulgaria
<mnemoc>
2013-08-12
<n01>
anybody already planned to go?
<n01>
mnemoc: yep, I have to catch up with my emails after vacation
<oliv3r>
oh i haven't read that yet :p
<mnemoc>
i'm also very late on my mails.... found that one by accident
<mripard>
n01: I'm thinking about going there, but I already have quite a lot of travels and conferences until november :S
<n01>
mripard: lucky you :D
<n01>
uhm I need to ping my SO then :D
<mripard>
n01: :)
Colani1200|work has quit [Ping timeout: 264 seconds]
vinifr has joined #linux-sunxi
<oliv3r>
yeah I'm still trying to go to fosdem; i doubt she'll let me go to sofia in september
<oliv3r>
erm november
<oliv3r>
unless i win the lottery and get 2 tickets
<oliv3r>
200E per person, 400 for us together, ouch :(
vicenteH has quit [Ping timeout: 240 seconds]
<n01>
oliv3r: you can stay at fosdem while she drains your credit card in shoes and dresses :)
<mripard>
who need shoes and dresses in brussels? beer and chocolate! :)
<oliv3r>
haha, i don't have a CC and am poor :p
<oliv3r>
shoes and chocolate and i'll be a deal :)
Soru has quit [Quit: No Ping reply in 180 seconds.]
Soru has joined #linux-sunxi
<eagles0513875_>
hno`: ping i have a question regarding uboot and cubieboard2 when you get a chance
boycottg00gle has quit [Remote host closed the connection]
atiti has quit [Ping timeout: 246 seconds]
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
jlj_ has quit [Ping timeout: 256 seconds]
jlj has quit [Ping timeout: 245 seconds]
\\Mr_C\\ has quit []
geecko has quit [Ping timeout: 256 seconds]
jlj has joined #linux-sunxi
jlj_ has joined #linux-sunxi
<hno`>
eagles0513875_, sure. just ask. See topic.
<eagles0513875_>
hno`: does boot support the a20 (a7) cpu? it seems like the version that buildroot uses seems to just hang
jlj has quit [Read error: Connection reset by peer]
<hno`>
eagles0513875_, I don't know what buildroot uses. but out sunxi-current branch do support cubieboard2.
<eagles0513875_>
hno`: by default buildroot when you run defconf for cubiboard it has this as the address for the custom git repo https://github.com/linux-sunxi/u-boot-sunxi.git and revision 1a8ac55429f7f8cc9f100c1cf2dc0195cf81e76f are those correct
<hno`>
you can't use a u-boot built for cubieboard on cubieboard2. It's different hardware.
<n01>
eagles0513875_: use the sunxi-current branch and the latest revision
<eagles0513875_>
n01: what branch am i currently using
<Turl>
sigh, their wiki got flooded by *a lot* of spam :(
<n01>
hi Turl
<Turl>
n01: did you get any news on the watchdog driver?
sanka has joined #linux-sunxi
<n01>
Turl: still waiting for approval, gh
<soletti>
thx Turl
<soletti>
thought there was a respository or something
<Turl>
n01: do you want me to send a ping mail?
<n01>
mripard ping-ed already today
<Turl>
n01: ok then
sanka has quit [Quit: Leaving]
<wingrime>
Turl: Ir clocks are settled in mainline?
<Turl>
wingrime: not yet
<Turl>
hno: your mail about uboot branches managed to confuse me even more :P
deasy has quit [Quit: Quitte]
n01 has quit [Remote host closed the connection]
<hno>
Turl, sorry about that. Short version, use sunxi branch.
<Turl>
hno: the user part was ok, but I got lost on the paragraph for developers.
<Turl>
you say they need to get merged to u-boot sunxi and then the sunxi branch
<hno>
"For those working on u-boot this means that staging of changes should be done outside of the main sunxi-uboot repository and merged into sunxi u-boot after suitable review process, and then to the sunxi branch. The sunxi-current branch is no more."
<hno>
Is it the "and then to the sunxi branch" part that is confusing?
<hno>
what I mean is that changes should be meged into the sunxi branch.
<hno>
Turl^
<Turl>
hno: yes, kind of
<Turl>
hno: aren't "sunxi u-boot" and "the sunxi branch" the same?
<hno>
"sunxi u-boot" is meant as a collective reference to the software as a whole, not a specific branch.
<Turl>
but we're going to have just 1 branch so it's going to be the same right?
<hno>
there will likely be wip/ branches from time to time when there is burts of shared development, and also many release stags.
atiti has quit [Ping timeout: 245 seconds]
<hno>
s/stags/tags/
<Turl>
ok, I think I understood now
<hno>
but only one main branch.
<Turl>
mripard: ping?
deasy has joined #linux-sunxi
<Turl>
sigh
<Turl>
now google rejects my email because I don't have a PTR record on v6 :(
<hno>
Quite understandable.
<hno>
mail servers should have PTR..
rz2k has quit []
<Turl>
hno: it worked good for ages though
<Turl>
and my IP is a good citizen, and complies with SPF and DKIM
* hno
really do not like SPF.
<Turl>
why?
<hno>
To error prone, and makes a major headache for simple mail lists.
<Turl>
really?
<Turl>
it's just a list of the IPs that can send email from your domain
<Turl>
and it doesn't even need to be exhaustive
jemk has quit [Quit: bye]
<oliv3r>
hno: so new stuff in review mode ona branch, then tested in u-boot repo and finally pushed to sunxi repo when its 'accepted'
<hno>
oliv3r, review primarily in external branches (your, mine, hans, whoevers repo) and discussed on mailinglist prefarbly.
<hno>
then merged to sunxi branch in main repo when "done".
<hno>
or "accepted".
<hno>
if ther is shared development then wip/.... branches in main repo is appropriate.
<hno>
to ease syncronisation between developers working on the same thing.
<oliv3r>
ok fair nuff
<oliv3r>
wait,t here is no sunxi-u-boot repo seperate from the sunxi repo, brainfart!
<hno>
?
<hno>
you have one, I have one. hand have one. Many others have one..
<oliv3r>
yeah but not an official sunxi-uboot repo
<oliv3r>
we have linux-sunxi/u-boot
<oliv3r>
for some reason i was thinking we had sunxi-uboot/u-boot
<geecko>
need someone who has some experience with the kernel
<hno>
We have linux-sunxi/u-boot-sunxi only
<oliv3r>
yep, i got confused somehow, sorry
<hno>
geecko, just ask what you need to ask. See topic.
bsdfox has quit [Ping timeout: 240 seconds]
<geecko>
hno, but it's not related to the allwinner kernel.
<geecko>
it's a general thing though
<hno>
geecko, just ask anyway
<geecko>
ok well
<hno>
if not appropriate then someone will tell you where to ask.
<geecko>
i got mali kernel drivers from odroid source, put them in the cyanogenmod kernel for common exynos4 devices (not related to allwinner, allright ;) )
<oliv3r>
hopefully it's not about the allwinner kernel ;)
<geecko>
they should compile
<geecko>
-but
<geecko>
when i build the kernel, they are excluded from the build
<geecko>
i checked the defconfig, and took the values from odroid, so it _should_ work
<hno>
mali isn't kernel.. that's ARM attempt at working around the kernel. But..
<hno>
that sounds like a mali Makefile integration issue.
<geecko>
hno, i'm talking specifically about the kernel-side drivers
<hno>
I know. So am I.
<oliv3r>
oh those are messy iirc
<geecko>
hno, i triple checked everything and didn't see anything suspicious
<geecko>
will quadruple-check then
<Turl>
geecko: if you push your tree somewhere I can have a look
<geecko>
it's embarrassing: i spent two days on this issue, and i can't fix it
<geecko>
Turl, you are awesome.
<hno>
everytime I look at ARM mali driver Makefiles I decide I do not want to look there again.. certainly do not envy the ones who have attempted to clean them up for sunxi kernel.
<Turl>
hno: disabled v6 and it goes through just fine, so I hope it's just the PTR
<Turl>
hno: yeah, they're ugly and often broken
<Turl>
hno: it's been improving though, r2p* were a pain, but r3p* needed just some little tweaks
<hno>
don't rememer which version I looked too hard at, but probably some releases back.
<geecko>
Turl, btw do you know about r3p2 blobs for android?
<geecko>
i have libEGL, libGLES*, and libMali.so but no libUMP.so
<hno>
libUMP.so is hardware integration specific iirc.
<Turl>
libUMP is theorically buildable from source I think
<hno>
yes.
<Turl>
I never got it to build though, but I saw there's a repo for it somewhere for sunxi
<geecko>
oh the headache
<hno>
geecko, welcome to blob land. headache guaranteed.
<geecko>
would libUMP.so from r3p1 work with other binaries from r3p2 ? that would simplify EVERYTHING
<geecko>
hno, i'd like to thanks Samsung, Insignal and ARM here.
<geecko>
Turl, what if i send arch/ and drivers/ only?
<hno>
there is many many more to thank for blob land.. those are fairly small players in that field.
<Turl>
geecko: I thought you'd have a git tree somewhere like github