hno changed the topic of #linux-sunxi to: /Allwinner/sunxi development discussion - Don't ask to ask. Just ask and wait! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi
<granite_crusher> pendrive is not bad, if to plug out one and replug other it will get mounted
<arokux2> granite_crusher: hm.. let me clarify. you can mount each of the pendrives if only one of them is plugged in. however if two are plugged in you cannot mount the second one?
<granite_crusher> arokux2: no, not one but only two
<granite_crusher> and 3rd can't be mounted
<granite_crusher> even dmesg reports it got pluged, and says it is sdc sdc1
<granite_crusher> but same do <blkid> it shows just two from 3
<granite_crusher> arokux2: I should test it in kernel 3.4, it is possible that problem can be on usb onboard hub
<arokux2> granite_crusher: hm.. possible.
<arokux2> granite_crusher: if you connect that third hub, everything works?
<arokux2> if you connect it alone?
<granite_crusher> arokux2: yeah first two works
<granite_crusher> arokux2: oh you mean that way, yeah
<arokux2> granite_crusher: ok. do you have external usb hub?
<granite_crusher> arokux2: no I dont
<arokux2> granite_crusher: ok, otherwise if you had a powered usb hub you could have tried to connect three pendrives to it, but ok.
<granite_crusher> arokux2: I have SD with debian and kernel 3.0 I just will test there
<arokux2> granite_crusher: I'll ask Tsvetan from Olimex if he thinks the power won't be enough for 3 pendrives.
<arokux2> granite_crusher: oh, perfect.
<granite_crusher> arokux2: I also thought it could be power
<arokux2> granite_crusher: hm.. wait. do you actually have a partition on that third pendrive?
<granite_crusher> but, have schematics there it is shown that it is 1000mA, (the host is intended on one of 4 chanels have wifi so it is powerfull enought)
<granite_crusher> arokux2: 3td pendrive is ok, if I mount it first or second it gets mounted
<arokux2> granite_crusher: ah, you have. i see.
<granite_crusher> the rule is simple as jungle rule - who's first!
<arokux2> :)
<arokux2> granite_crusher: ok, 3 pendrives.. weird. please tell me you findings once you tried legacy kernel
Gerwin_J has joined #linux-sunxi
<arokux2> granite_crusher: alsamixer and jack need to be installed, they are not in that rootfs I supplied to you. but you better test them after you have a distro on a usb stick mounted as rootfs
<granite_crusher> arokux2: blkid gets all 3
<arokux2> granite_crusher: on legacy kernel?
<granite_crusher> arokux2: but most funny thing is that it is not sda1 sdb1 sdc1 but... uba1 ubb1 ubc1
<granite_crusher> arokux2: yeah on kernel... how to check exactlly?
<arokux2> uname -a
<granite_crusher> 3.0.52+
<arokux2> granite_crusher: you cannot mount usb stick as rootfs because the driver wasn't loaded yet.. so the kernel didn't know about /dev/sda1 yet. we need to find a way to force it to enumerate all usb devices first.
<granite_crusher> arokux2: probally initrd...
<arokux2> granite_crusher: should be possible to deal without it.
<arokux2> granite_crusher: and usb support is built-in any way
<arokux2> granite_crusher: initrd is used to add flexibility -- so you do not build stuff into the kernel
<granite_crusher> arokux2: hope these findings will help, I think it is time to sleep
<arokux2> granite_crusher: yes of course, very helpful indeed. good night!
granite_crusher has quit [Quit: Page closed]
\\Mr_C\\ has joined #linux-sunxi
ykchavan has quit [Ping timeout: 252 seconds]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
arokux2 has quit [Remote host closed the connection]
<akaizen> is anyone good with Allwinner android here?
Gerwin_J has quit [Quit: Gerwin_J]
awafaa has quit [Ping timeout: 260 seconds]
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
TheSeven has quit [Read error: Operation timed out]
TheSeven has joined #linux-sunxi
\\Mr_C\\ has quit []
awafaa has joined #linux-sunxi
hramrach has joined #linux-sunxi
rellla has joined #linux-sunxi
rellla has quit [Client Quit]
forcev is now known as FunkyPenguin
FunkyPenguin has quit [Changing host]
FunkyPenguin has joined #linux-sunxi
wingrime has joined #linux-sunxi
rz2k has joined #linux-sunxi
<oliv3r> next set of pinctrl patches work
Black_Horseman has quit [Ping timeout: 260 seconds]
<wingrime> oliv3r: thats can be our cedar
<oliv3r> registers match?
<oliv3r> serial is designware
<oliv3r> so i wouldn't be supprised
<oliv3r> but this page can be very usefull to test/expect limitations features
<wingrime> oliv3r: codecs match
<oliv3r> dts/ace3 too?
<oliv3r> still it's very intersting bit to know
<oliv3r> as you know the capabilities and limitaions now better
<oliv3r> also means the IP doesn't have complete shit :p
<wingrime> oliv3r: do you know witch kind of encode here?
<oliv3r> no clue, what's a uu file?
<oliv3r> could be some unicode encoding
<oliv3r> like all codes -> ascii
<wingrime> oliv3r: I find non blob
wingrime_ has joined #linux-sunxi
rz2k has quit []
<oliv3r> wingrime: they have full source for their VPU? i didn't know that existed
Cubear has joined #linux-sunxi
<Cubear> lo
<wingrime_> oliv3r: blobs too
<oliv3r> wingrime_: i don't udnerstand :(
<oliv3r> is rk blob-free?
<wingrime_> oliv3r: no
<oliv3r> wingrime_: ok thought so :)
<oliv3r> wingrime_: did you get your CT yet?
<wingrime_> oliv3r: no
<wingrime_> oliv3r: you get?
<oliv3r> nah
<oliv3r> i did get track & confirm for rdxa
<oliv3r> its mailed from sweden strangly enough
<wingrime_> oliv3r: post pics
<wingrime_> oliv3r: crap
<wingrime_> oliv3r: looked too rk sources and blobs
<wingrime_> oliv3r: regs complitily diffrent
Super-noob has quit [Ping timeout: 245 seconds]
hipboi has joined #linux-sunxi
hipboi has quit [Quit: Leaving]
hipboi has joined #linux-sunxi
hipboi has quit [Ping timeout: 252 seconds]
akaizen_ has joined #linux-sunxi
akaizen has quit [Ping timeout: 248 seconds]
hipboi has joined #linux-sunxi
<wingrime_> hipboi: are you worked on melis?
<hipboi> wingrime, no
<wingrime_> hipboi: just seeking old/new blobs for cedarx...
<wingrime_> hipboi: old blobs can be very usefull for RE
<hipboi> how old it should be
massi has joined #linux-sunxi
<wingrime_> hipboi: old enought cedarx be in hw
<wingrime_> hipboi: cedarx codecs-set looks very alike with rk's but rk registers complitily different
<wingrime_> hipboi: rk use design ware ace video decoder, also I have no idea where cedarx can be also, (not in aw's hw)
<hipboi> wingrime, :O
<wingrime_> hipboi: cedarx aw's or 3rd IP ?
<hipboi> wingrime_, what do you mean by not in aw's hw
<wingrime_> hipboi: other fab-less vendors
<hipboi> yes
<hipboi> cedarx is only on allwinner platform
<wingrime_> hipboi: but why cedarx have so many unused tings
<hipboi> wingrime_, haha
<hipboi> because it's original from others
<wingrime_> hipboi: for exapmle AVS jizun codec have no any code
<hipboi> do you thing that a three year founded company can make a complete arm soc?
<hipboi> with everything self developed?
<wingrime_> hipboi: indeed no, but question, where they get verilog code...
<wingrime_> hipboi: opencores ?
<hipboi> wingrime_, no
<wingrime_> hipboi: they just cut-and-paste IPs ....
<wingrime_> hipboi: I wonder that worked
<hipboi> wingrime_, well ... i can not tell
<hipboi> wingrime_, to find the original cedarx, i think we had better look at the early chips like F18, F20...
<hipboi> wingrime_, but that's why the Chinese soc are so high integrated and also low cost...
<wingrime_> hipboi: good point, at least honest
<hipboi> wingrime_, i realized this long before ...
<wingrime_> hipboi: As I know there is melis for a10
<wingrime_> hipboi: do you know where I can get it?
<hipboi> write an email to benn@cubietech
<hipboi> see if he can get from allwinner internal people
<hipboi> i am at shenzhen now
<hipboi> for him, it's just a walk after dinner to allwinner office
<hipboi> with a u disk and copy it back
<wingrime_> hipboi: cool, for me even binaries useful
<wingrime_> hipboi: runable binaries more useful
<wingrime_> hipboi: but there some HW without docs/blobs thats totaly hell
<wingrime_> hipboi: as I know we have Codec Engine that can do zlib's deflate on hw , but no code/no docs /only some reference....
<wingrime_> hipboi: do you remeber aw chip history, witch chip was first, and that comes then?
<hipboi> wingrime_, the first allwinner chips are not branded by allwinner
arokux2 has joined #linux-sunxi
<hipboi> there are names such as boxchip
<hipboi> sochip
<hipboi> it's said that they used arm core but they did not pay license to arm
<hipboi> i need to check out at the famous Chinese forum
<hipboi> imp3.net
<hipboi> they begin the discussion of Chinese chip solution since mp3 time
Black_Horseman has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
<wingrime_> hipboi: witch chip was firstly in mass production ?
<wingrime_> oliv3r: ping
<wingrime_> hipboi: thats understandable , any start-up need money for first. legality, license can wait
<arokux2> hi ppl
<hipboi> wingrime_, SC8600 was in 2009
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
Black_Horseman has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
<hipboi> wingrime_, it was F10, F15 i think
<wingrime_> hipboi: wow, I find SDK for SC8600
<wingrime_> hipboi: sunii thats means sun2i
<wingrime_> hipboi: but was sun-1 ?
<hipboi> wingrime_, i found the so called sdk too...
<hipboi> i think sun-i was never massive produced...
<wingrime_> hipboi: thants you, I find ms-pro regs in sdk header
n01 has joined #linux-sunxi
<wingrime_> hipboi: SoChip SC9800 == BoxChip F10 ,, SoChip SC8600 == BoxChip F15
<hipboi> wingrime_, it seems
wingrime has quit [Ping timeout: 252 seconds]
wingrime_ has quit [Ping timeout: 240 seconds]
wingrime has joined #linux-sunxi
ykchavan has joined #linux-sunxi
wingrime_ has joined #linux-sunxi
<arokux2> wingrime: maybe you can document on our wiki, like history or smth, otherwise it'll get lost
Black_Horseman has quit [Read error: Connection reset by peer]
<n01> 'moin
geecko has joined #linux-sunxi
granite_crusher has joined #linux-sunxi
massi has quit [Remote host closed the connection]
<granite_crusher> arokux1 or arokux2: I looked more into issue with usb in u-boot
massi has joined #linux-sunxi
<arokux2> granite_crusher: arokux2 for now :)
<arokux2> granite_crusher: http://linux-sunxi.org/User:Arokux
<arokux2> granite_crusher: I know what you should do to wait for usb stick popup with your /dev/sda1
<arokux2> granite_crusher: Add root=/dev/sda1 rootwait to your cmdline.
<granite_crusher> yeah I saw in your wiki section, nice
<granite_crusher> well, if to continue with this usb error in u-boot
<granite_crusher> I get it just with two devices - one is semi pro grade (M-audio Fast track pro) usb soundcard, other is also kind of audio equipment - usb MIDI keyboard (CME M-key v2)
<wingrime> oliv3r: ping
<arokux2> granite_crusher: I see. are they both full speed?
<granite_crusher> arokux2: low-speed mouse, keyboard or high speed webcam or eth-usb dongle gets ok and visible with <usb tree>
<arokux2> granite_crusher: I see. (might be a bug in U-Boot usb stack too.)
<granite_crusher> arokux2: I just dmsg'ed it on pc, yep bowth is full-speed
diego71_ is now known as diego71
<arokux2> granite_crusher: where do you know mouse is low-speed? checked it?
<granite_crusher> arokux2: in addition, u-boot can read from 3 pendrives at once
<arokux2> granite_crusher: other thing.. blkid may output wrong things. to verify 3 pendrives, try to mount all of them.
<granite_crusher> arokux2: I checked it with dmesg
<granite_crusher> on pc
<granite_crusher> arokux2: I tried and it failed
<arokux2> granite_crusher: ah, nice, so only full speed devices won't work.
<granite_crusher> arokux2: identical behaviour as blkid
<wingrime> mnemoc: when you last time saw Ithamar ?
<arokux2> granite_crusher: what was the error when you tried to mount it?
<arokux2> granite_crusher: is it in that dmesg you sent me yesterday?
<granite_crusher> arokux2: no it is not in that log
<granite_crusher> arokux2: one liner
<granite_crusher> mounting /dev/sdc1 on /mnt/apacer failed: No such file or directory
<n01> tm512: nice post on reddit ;)
<arokux2> granite_crusher: hm.. was there /dev/sdc1, have you checked?
hramrach has quit [Ping timeout: 240 seconds]
<granite_crusher> arokux2: dmesg showed it got plugged and that it is sdc
<granite_crusher> but neither blkid can see it neither can it be mounted
<arokux2> granite_crusher: yeah, and then there was also line sdc: sdc1, have you noticed?
<granite_crusher> yeah
<arokux2> granite_crusher: so I wonder if there was /dev/sdc1 device file.
<granite_crusher> arokux2: I am sending you log excerpt
<granite_crusher> arokux2: http://sprunge.us/FcjJ
choko has joined #linux-sunxi
<arokux2> granite_crusher: please take a look if there is /dev/sdc1
<granite_crusher> arokux2: ok I will boot and check
<arokux2> granite_crusher: you can do it later some time.
hipboi has quit [Ping timeout: 260 seconds]
<arokux2> granite_crusher: I also wanted to give you a hint. you can save several u-boot commands in one env variable and the do: run <var_name>
<arokux2> granite_crusher: take a look here.
<granite_crusher> arokux2: can I have few such files?
<granite_crusher> and choose one of it at boot?
<arokux2> granite_crusher: you can then save the environment and use it next time.
akaizen_ has quit [Remote host closed the connection]
<arokux2> granite_crusher: you can define several variables
<arokux2> granite_crusher: and then do: run var1 or run var2
<arokux2> granite_crusher: take a look here: http://linux-sunxi.org/Possible_setups_for_hacking_on_mainline
<choko> Hi, arokux
<choko> i checked my mini usb port is only for power
hramrach has joined #linux-sunxi
<granite_crusher> arokux2: where do it save it?
<granite_crusher> arokux2: SD card, SD fat partition, or USB partition?
<arokux2> granite_crusher: sd card
<granite_crusher> cool
<arokux2> granite_crusher: there is a special place reserved for environment
<arokux2> choko: so you cannot use it to connect a usb device?
<choko> no i can't do that
<arokux2> choko: i see.. :(
<arokux2> choko: btw, you compiled your kernel by yourself?
<choko> any other options ?
<choko> yes compiled by myself
<choko> but i also used Fedor wich downloaded
<choko> Fedora
<granite_crusher> arokux2: I think after usb start I should force u-boot wait a bit, how to do that?
<choko> If you have any other compiled version for a10s give a link ;)
<arokux2> granite_crusher: next command won't start till the last finishes, so no need to do smth
<arokux2> choko: yep, you said u used fedora and 3.4.61.. and I was surprised later because that fedora image is rather old. but now interesting question comes. does it work the same under fedora kernel (which version is it, btw?) and 3.4.61 kernel?
<arokux2> choko: ok.. but what kernel version is inside? just run: uname -a
<choko> but i can't login right now, somthing fuckup with password
<arokux2> choko: oh shit
<choko> i mounted on my linux and changed /etc/shadow but still don't work
<choko> ah btw this ofcourse booting from my sdcard not nand partition
<granite_crusher> arokux2: You can delete that problem 2 of 3 usb, there is no /dev/sdc* files in this buildroot initramfs
<choko> maybe this is a problem
<arokux2> granite_crusher: sorry?
<arokux2> granite_crusher: why should I delete if there is no /dev/sdc1? (you can edit it yourself!)
<arokux2> choko: why have you changed /etc/shadow?! it is where the passwords are stored.
<granite_crusher> arokux2: I found that in /dev/ there is plenty of sda* (1 2 3 ... 9) sdb*(1 2 ... 9) but no sdc
<choko> yes i changed to password which i knew
<granite_crusher> arokux2: I mean in wiki, that ppl wouldnt think there is something wrong
<arokux2> granite_crusher: why so many sda*? this is not good
<granite_crusher> I think it is because of buildroot
<arokux2> granite_crusher: .. no idea.
<arokux2> granite_crusher: you can get rid of this buildroot and try to put you rootfs on usb stick
<arokux2> granite_crusher: do you have debian rootfs?
<granite_crusher> arokux2: that initramfs have lots of devices... If they would be real ... http://sprunge.us/iJIR
<arokux2> granite_crusher: some video* too!
<arokux2> granite_crusher: no idea. i think it is best to switch to some distro. you'd nee a sane package management too.
<granite_crusher> arokux2: is there difference where will be my uImage and dtb, can it be just in / or should I copy it to /boot/?
<granite_crusher> I have deboostraped debian, which worked with 3.4 kernel
rz2k has joined #linux-sunxi
<n01> granite_crusher: are you using devtmpfs?
<granite_crusher> devtmpfs?
<n01> 15:01 < granite_crusher> arokux2: I found that in /dev/ there is plenty of sda* (1 2 3 ... 9) sdb*(1 2 ... 9) but no sdc
<arokux2> granite_crusher: nope, you can put your kernel where ever you want. however..... i think it is better to put it at some place where you have a shared access to it.
<n01> it seems like you are not using devtmpfs
<arokux2> granite_crusher: if you later would like to change your kernel and try things out, you wouldn't like to mess up with usb sticks
<arokux2> granite_crusher: I recommend you get your usb2ethernet working and set up tftpserver (very easy)
<granite_crusher> n01: Ichecked my kernel config it is enabled
<n01> granite_crusher: are you using udev/mdev?
<n01> or a static dev allocation map
<arokux2> n01: whatever is in buildroot...
<granite_crusher> n01: where to check it?
<n01> arokux2: in buildroot it is configurable obviously
<n01> granite_crusher: buildroot menuconfig
<arokux2> wait
<n01> but are you using my config for A10?
<n01> R2_ROOTFS_DEVICE_TABLE="system/device_table.txt"
<n01> BR2_ROOTFS_STATIC_DEVICE_TABLE="system/device_table_dev.txt"
leowt has joined #linux-sunxi
ykchavan has quit [Ping timeout: 245 seconds]
leowt has quit [Remote host closed the connection]
leowt has joined #linux-sunxi
<arokux2> do you know cheap-shipping online shop for connectors & stuff?
<granite_crusher> arokux2: me?
<arokux2> anybody
<Sonicadvance1> digikey
<arokux2> Sonicadvance1: +18 EUR
<Sonicadvance1> oh
<jelly-home> arokux2: cheap-shipping to which country
<arokux2> jelly-home: Germany
tzafrir has quit [Ping timeout: 264 seconds]
<arokux2> granite_crusher: ah! still there?
<arokux2> granite_crusher: just realized: the file /dev/sdc1 is not created because it's not in that static list of device files in buildroot
<arokux2> granite_crusher: so it should work, I think --> removing it from wiki
<n01> Turl: ping
<granite_crusher> arokux2: good
<granite_crusher> arokux2: I changed in my config of kernel rootdelay=10 into rootwait as you suggested in wiki... no luck, i get kernel panic in same place... maybe usb should be /dev/uba1 not sda1?
<arokux2> granite_crusher: oh wait.. so you had rootdelay=10 already? rootwait just means to wait indefinitely
<arokux2> granite_crusher: so as it hasn't waited i assume you change hasn't had any influence..
bamvor has quit [Ping timeout: 245 seconds]
bamvor has joined #linux-sunxi
<granite_crusher> arokux2: maybe I should force it to use these in kernel by setting <Kernel command line type> into "always use default..."?
<arokux2> granite_crusher: do not know anything about that. I normally set bootargs in u-boot
<arokux2> granite_crusher: try this
<arokux2> granite_crusher: setenv bootargs console=ttyS0 root=/dev/sda1 rootwait
<granite_crusher> arokux2: I am stupid... to use bootargs in uboot it is much easier than to recompile kernel....
<arokux2> granite_crusher: now you know it... :)
<arokux2> granite_crusher: so if you want to continue to hack you'd need to setup things so everything is automatic, only minimal interaction. I can give you tips later. I can boot different kernel each 10 sec, i think :)
<granite_crusher> arokux2: cool
<aep> anyone knows how to rotate the lcd output?
<granite_crusher> arokux2: I think after console=ttyS0 I should add ",115200" because now I get just
<granite_crusher> Starting kernel ...
<granite_crusher> arokux2: and USB led is flashing (means it is used
<arokux2> granite_crusher: wait a bit
<arokux2> granite_crusher: we should have added earlyprinkt probably.. (not sure)
<arokux2> granite_crusher: but you will get a shell later if it works
<n01> you have to enable earlyprintk on the right tty
<n01> mripard: ping
<arokux2> Sonicadvance1: shipping - 0.0EUR
<granite_crusher> arokux2: I am not shure, but problem could have been in putty (it stops working time to time, and this was one of it)
<granite_crusher> arokux2: I got to it :)
<arokux2> granite_crusher: :)
leowt has quit [Quit: leowt]
hipboi has joined #linux-sunxi
<naobsd> arokux: I have USB wifi dongle
<arokux2> naobsd: can you try if it works with your cubieboard and netbsd?
<arokux2> granite_crusher: so next step is to get usb2ethernet working, right?
<granite_crusher> arokux2: I just tested 3usb is working
<arokux2> granite_crusher: +1
<granite_crusher> but /dev/ is overcrowded with some junk
<granite_crusher> sd* apears dynamically with every insert of us
<granite_crusher> *usb
<naobsd> well, I don't have NetBSD userland(rootfs) for now :(
<arokux2> granite_crusher: yeah, you cannot really rely on /de/sd* names, you should use UUIDs...
<naobsd> arokux: can you tell me summary of devices which are working, or not working?
<arokux2> naobsd: I see. are u interested in testing a mainline a bit? I have a problem with on-board wifi, I wonder if it is because its on-board.
<granite_crusher> arokux2: yeah, especially if some other pendrive would be insert, when enumeration gets different
<arokux2> granite_crusher: maybe there is something like UUID to /dev/sd* mapping you can set, then you can use /dev/sd*
<granite_crusher> arokux2: I counted about 200 tty* in /dev/.......... what a hell?
<naobsd> arokux: is on board USB wifi powered properly? I guess it's controlled by GPIO
<arokux2> naobsd: do not know, but i do same gpio trick for both controllers..
skoti has joined #linux-sunxi
<granite_crusher> arokux2: I remember somewhere in the internets were instructions how to use /dev/disk/by-uuid/<UUID> instead of /dev/sd*
<naobsd> arokux: what I said is not about host controller, I guess there is board specific GPIO pin to give power for on-board USB wifi... is it detected?
<arokux2> granite_crusher: take a look, it is distro agnostic: https://wiki.archlinux.org/index.php/Persistent_block_device_naming
<arokux2> naobsd: hm... how do I know? :(
<arokux2> naobsd: here is fex for my board, it should be there, right? https://github.com/linux-sunxi/sunxi-boards/blob/master/sys_config/a10/mele_a1000.fex#L762
<mouchon> hello
<naobsd> arokux: well... it seems there is no information :( so it should work w/o any special handling for power. if on-board wifi can be found, it should be powered
<mouchon> i try to refactor my i2c patch to avoid ugly code. But still have a choice problem. should i use inline function or macro the kernel doc is not clear for me
jacq has joined #linux-sunxi
<arokux2> naobsd: can I measure with multimeter? I've tried each pair of pins and for some there were 3.3V, so I assume it is powered, right?
<arokux2> hi mouchon, nice to see you here.
<arokux2> mouchon: to be honest I do not know... but there was some discussion here about inline vs define
<arokux2> mouchon: take a look at this func here too: http://lxr.free-electrons.com/source/include/linux/device.h#L915
<arokux2> mouchon: maybe you can find how other ppl do it in the kernel.
leowt has joined #linux-sunxi
<mouchon> also where to put the macro ?: top of file or near the usage of it ?
<naobsd> arokux: sorry, not sure
<n01> mouchon: top is better
<arokux2> mouchon: they do not cast even. you should do the same.
<naobsd> arokux: is there any a1000 internal photo?
<mouchon> ok will look at links
<naobsd> I should learn how to build/use mainline kernel...
<n01> (void *) is never casted
<naobsd> arokux: left bottom is wifi?
<arokux2> naobsd: yep
<wingrime> oliv3r: ping
<naobsd> arokux: no idea which pin is VBUS...
<naobsd> arokux: is on-board wifi detected? is kernel driver attached and error/panic at some point? or it's totally not found?
<arokux2> naobsd: it is found, everything seems to work, but once I try "ip link set wlan0 up" I get hardware erorrs.
<arokux2> naobsd: http://sprunge.us/Qihe
<naobsd> arokux: then it should not be power issue, sorry
<naobsd> arokux: did you tried USB device which has EP more than 1?
<arokux2> naobsd: usb2ethernet for example?
<naobsd> arokux: probably yes
<arokux2> naobsd: yep, it works.
<arokux2> naobsd: if it is power - than it's good. you know it works in sunx-3.4. it is not board specific kernel. so the power enablement should come from fex, right?
<arokux2> hipboi: am I right?
<arokux2> hipboi: we're talking about the power supply to the on-board wifi module
<naobsd> arokux: if it works with 3.4, board specific power thing should not be needed because it seems there is no such information in fex
vinifr has joined #linux-sunxi
<naobsd> arokux: well, ehci1? do you have any physical USB port connected to ehci1?
<arokux2> naobsd: wifi is connected to ehci2.
<arokux2> Tsvetan: you there?
<wingrime> hipboi: thats strange ... melis not use import in modules, there is no module name in file, so, ONLY call by OFFSET ??? wtf
<naobsd> oh, dmesg said ehci1... I guess ehci2 was 2nd instance
<wingrime> hipboi: I mean no function names
<hipboi> wingrime, because that sdk was also REed
<hipboi> if we are talk about the same one by Russian
<n01> wingrime: do you know if we have in mainline something to differentiate among A10/20/etc.. ? otherwise I'll use fdt
<wingrime> hipboi: no, I unpacked some fw
massi has quit [Quit: Sto andando via]
<wingrime> hipboi: drivers, modules, are elf, but I can't find any function names
<arokux2> naobsd: do you know what those gpio section could mean?
<wingrime> hipboi: so I wonder how it do export
<wingrime> hipboi: thats SDK not looks REed
<arokux2> hipboi: do you have idea what pin can be VBUS at the wifi module, bottom left? http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/A10_device_a1000_inside.jpg
<naobsd> gpio section should be for gpio driver
jacq is now known as fluo75
fredy has quit [Excess Flood]
balage has left #linux-sunxi [#linux-sunxi]
fredy has joined #linux-sunxi
<mouchon> hi i modified the ic2 patch using platform_get_drvdata function. Should i submit the patch as new subject or add it on current one ?
<rz2k> wingrime: wingrime: wingrime:
Soru has quit [Read error: No route to host]
<arokux2> mouchon: what ever you like. have you configured git send-email? :p
<mouchon> yes i did
<arokux2> mouchon: perfect!
<arokux2> rz2k: can I have your attention for a sec.. do you know which pin is VBUS for wifi module (bottom left)? http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/A10_device_a1000_inside.jpg
<n01> mouchon: you have to submit a v2
<mouchon> ok will do that later
<n01> so your subject will be [PATCH v2]
Soru has joined #linux-sunxi
<arokux2> thanks moofree
<arokux2> err
<arokux2> thanks mouchon
arokux2 has quit [Remote host closed the connection]
arokux2 has joined #linux-sunxi
<rz2k> arokux2: oh
<rz2k> dont remember, do you have a multimeter?
<wingrime> who is Дмитрий Снежинский?
<arokux2> rz2k: yes, I have one!
<rz2k> then boot device to any kernel where you have everything working and find +5V
jemk has joined #linux-sunxi
<arokux2> rz2k: I can find only 3.3V. is it bad?
<rz2k> probably this module needs 3.3v regulator
<rz2k> and has supply of that sort
<rz2k> honestly, I have no experience with these micro wifi modules
<arokux2> :$
<rz2k> I mean in my projects, so I could tell you exactly whats happenning
<rz2k> you should trace the board and find where 3v3 from wifi module is connected. probably to transistor that enables this regulator, which is controlled by gpio
<rz2k> or read the 3.4 wifi code
<rz2k> I bet it has something like that somewhere
<arokux2> rz2k: mainline driver works with wifi...
_enrico_ has joined #linux-sunxi
<arokux2> rz2k: so there is nothing special that allwinner plugged in, afaik
<arokux2> rz2k: the only thing is fex.. or?
<rz2k> oh, mainline works? then whats the problem, why are you interested in how power supply is tied to this module?
<arokux2> rz2k: no.. not that. there are two drivers for this wifi module, one from realtek the other one in mainline
<arokux2> rz2k: so the mainline driver works.
<arokux2> rz2k: not mainline kernel, but the linux kernel driver from 3.4 works. (and backports from 3.10 works too)
leowt has quit [Quit: leowt]
mcbrick has joined #linux-sunxi
<arokux2> rz2k: hm.. I cannot find a datasheet for this module, do you think there could be one somewhere I do not know?
Soru has quit [Read error: Connection reset by peer]
choko has quit [Quit: Page closed]
vinifr has quit [Ping timeout: 240 seconds]
Soru_ has joined #linux-sunxi
<arokux2> rz2k: I have measured all the possible voltages in the working and non-working state, they are the same.
<arokux2> rz2k: does it suggest the power is correctly provided to the wifi module?
Soru_ has quit [Read error: Connection reset by peer]
tzafrir has joined #linux-sunxi
rellla has joined #linux-sunxi
<rz2k> arokux2: yes, it does
<arokux2> rz2k: then I cannot understand what is the problem...
<rz2k> describe it in detail @ ML and we can think it out together
<arokux2> rz2k: you can take a look at the thread here: http://thread.gmane.org/gmane.linux.usb.general/94872
<arokux2> rz2k: I was trying to aks a USB maintainer..
buZz has quit [Read error: Connection reset by peer]
buZz has joined #linux-sunxi
sdboot has joined #linux-sunxi
hipboi has quit [Quit: Leaving]
<sdboot> How do I boot from an SD card on a tablet that always boot from NAND first? Boot order is SD0, NAND, SD3, right?
<wingrime> sdboot: yes
<wingrime> sdboot: but be sure you use sd0 not sd3
<sdboot> I think that's the problem. If SD3 is connected, is there any way to boot a custom Linux image from SD?
<wingrime> sdboot: kill nand
<wingrime> sdboot: just short some data lines
<wingrime> sdboot: boot will be failback to sd3
<arokux2> wingrime: do you have Mele A1000?
<wingrime> arokux2: no
<sdboot> No A20 A70x
<wingrime> sdboot: you can kill nand boot signature , erase first sectors
<sdboot> I have a backup of the NAND partitions, so I should be able to recover it right?
djahma has joined #linux-sunxi
<wingrime> sdboot: you can simply short nand ic wires when boot
rellla has quit [Remote host closed the connection]
<djahma> hi there, has anyone managed to get A10 wifi working? I've got a tablet I just can't manage to connect to internet. currently compiling sunxi 3.4 branch for more modules but often get errors
Cubear has quit [Quit: leaving]
geecko has quit [Quit: Quitte]
<rz2k> arokux2: no idea really
<rz2k> seems like errors on usb bus
<rz2k> hardware ones
Vespine has joined #linux-sunxi
Vespine is now known as Cubear
<Turl> n01: pong
_enrico_ has quit [Read error: Connection reset by peer]
jemk has quit [Remote host closed the connection]
<arokux2> rz2k: yeah.. do you have *any* idea how can I track the difference between two behaviors?
<arokux2> Turl: do you plan to push PLL6 soon?
sdboot has quit [Ping timeout: 250 seconds]
_enrico_ has joined #linux-sunxi
<n01> Turl: rtc driver is almost the same for a10 and a20 but for some constants. To distinguish between the two I'm using two different of_device_id
<n01> there is any other way in mainline to do that?
<n01> in sunxi-3.4 there is sunxi_is_sun7i()
<arokux2> n01: do you have a wifi dongle? :p (don't remember if I asked you..)
<n01> arokux2: yeah
<arokux2> n01: could you test my tree with usb support please? you might have heard there are some problems with on-board wlan
<n01> "on-board wlan" ??
<arokux2> n01: yes, why are you surprised?
<n01> are you talking about cb2+
<arokux2> n01: no, I have another board: Mele A1000
<n01> hoooo, now it makes sense
<arokux2> n01: but I would like you to test wifi dongle, so that I know it is not special to wifi..
<arokux2> n01: I suppose it should work.
<n01> arokux2: I'll put it in TODO list
<arokux2> n01: I can test your rtc patch on Mele, it is sun4i
Gerwin_J has joined #linux-sunxi
<n01> yeah the problem is that one :) I want to finish with this driver asap
<arokux2> n01: no prob
<arokux2> n01: same here :D
<Turl> n01: just use two diff compatibles, and pass the constants with that
<n01> ok, i'm doing so now
<n01> good
<Turl> n01: you can pass a real pointer instead of the integer stuffing oliv3r does and use that
<n01> Turl: yeah, I'm using of_find_matching_node + of_match_node actually
<Turl> n01: why not of_match_device?
<n01> didn't know of of_match_device
<Turl> n01: it's used on sid :p
<n01> well, of_find_matching_node + of_match_node is used in sunxi-board.c ;)
djahma has quit [Quit: Leaving]
<arokux2> Turl: does it work as described here in mainline? usb is mentioned here.
<Turl> arokux2: eh?
<arokux2> Turl: what?
<Turl> arokux2: mainline has no cpufreq?
<arokux2> Turl: I don't know :(
<arokux2> what is that?
<Turl> what is what? cpufreq?
<Turl> DVFS
<arokux2> Turl: I'm just asking if the mainline works the way it is described in comment to that function.
<arokux2> and the limitation of apb is 150Mhz. for usb connecting,
<arokux2> * the frequency of ahb must not lower than 60Mhz.
<granite_crusher> arokux2 and Turl: cpufreq shouldn't work in mainline because there is no AXP drivers
aexl has joined #linux-sunxi
<aexl> moin.
<aexl> does anybody have a mele m5. and knows if it got a uart interface?
<arokux2> aexl: oh, we'd love to have some mele m5 pictures....
<Turl> arokux2: cat clk_summary and see for yourself? :) dunno
<Turl> granite_crusher: you can get along with no axp and scale just the freq and not the voltage
<Turl> but it's half the fun :)
<arokux2> Turl: ah, right.. thanks
<aexl> arokux: hm. this looks comprehensive: androidpc.es/blog/2013/05/10/mele-m5-con-soc-a20-dualcore-en-oferta-en-su-tienda-oficial
<arokux2> aexl: the pictures of the board inside..
Gerwin_J has quit [Quit: Gerwin_J]
<arokux2> Turl: "the limitation of apb is 150Mhz. for usb connecting" do you think its the upper or lower bound?
<Turl> arokux2: you're reading that wrong
<Turl> " for usb connecting,
<Turl> * the frequency of ahb must not lower than 60Mhz."
<arokux2> *sigh* it is 168Mhz now..
<brain_> ok this is just annoying, I changed my bootargs in u-boot to reduce the bootdelay and now the device just loads boot0 after attempting to start the kernel...i don't get it.
<brain_> i'm sure its just not finding anything in nand but why would a simple change like that cause this? should I try to upgrade u-boot?
<brain_> from my understanding their is no u-boot source with nand support tho?
<brain_> I'll be back, hopefully someone can help out here. gotta boot into crap windows to livesuit it.
<arokux2> brain_: yes, our u-boot does not support nand
<brain_> :( sucks.
<granite_crusher> arokux2: I am using your tree, so I should ask you, why rtc doesnt work? it is written that clock drivers was merged into 3.10 and I2C (rtc is atached throught it) at 3.11.
<arokux2> n01: ^ <--- this guy is working on this driver afaik, my tree is based on the last stuff, dunno.
<Turl> granite_crusher: what device?
<arokux2> Turl: A13
<Turl> granite_crusher: most devices don't have an external i2c rct though
<granite_crusher> Turl: board or clock?
<Turl> granite_crusher: hm?
<Turl> granite_crusher: if your device has you probably just need to add the info to the device tree
<granite_crusher> Turl: it is soldered on A13-olinuxino... device tree hmm...
brain_ has quit [Remote host closed the connection]
<granite_crusher> is there somethere info how to make stuff in these device trees?
<arokux2> granite_crusher: yeah, there is a wiki
<arokux2> granite_crusher: but maybe take a look at other sunxi dts..
wingrime has quit [Read error: Operation timed out]
<Turl> granite_crusher: let me look at the O-A13 schematics
<granite_crusher> arokux2: good idea, copy paste and modify :)
wingrime_ has quit [Ping timeout: 248 seconds]
<granite_crusher> arokux2: by the way I got booting with UUID, just not exactly as links you sent, and its kind tricky it needs hybrid partition table MBR and GPT in one (GPT for kernel, MBR for u-boot)
wingrime has joined #linux-sunxi
<arokux2> granite_crusher: ok, good that you figured it out. have you configured network?
wingrime_ has joined #linux-sunxi
<granite_crusher> arokux2: still not, I am trying to do it now
<arokux2> granite_crusher: ok let us know in case of problems..
<Turl> granite_crusher: I'll make a patch for the rtc if you want, looks like it's already supported kernel wise
deasy has joined #linux-sunxi
<granite_crusher> Turl: you mean that rtc on chip is supproted in kernel?
<granite_crusher> at least I thought so
ZetaNeta has quit [Remote host closed the connection]
<granite_crusher> I enabled it while in config
ZetaNeta has joined #linux-sunxi
<Turl> granite_crusher: yes, you just need to declare it on the dt
<granite_crusher> Turl: cool, should I try to do it? (patch, I never did it and I am not good in git, but would be great to learn it)
<arokux2> granite_crusher: yes, this is a really good task to start how to do a patch
<arokux2> granite_crusher: if you like to learn it - go for it
<arokux2> granite_crusher: will you?
<Turl> granite_crusher: sure, go ahead :)
<arokux2> Turl: .. I've noticed strange behavior..
<Turl> mripard: ping, found a bug in our dts :)
<granite_crusher> arokux2 and Turl: I would like to try, hopefully I will not fail
<arokux2> granite_crusher: for each your feature you want to work on create a branch in git, and work there.
<mripard> Turl: hmmm, ok, what is it ?
<Turl> mripard: i2c nodes lack addr/size-cells
<Turl> mripard: except the a10s ones :)
Gerwin_J has joined #linux-sunxi
<mripard> Turl: hmmm, ok. and how is it a bug ?
<mripard> it's something missing, sure, but only until actually someone's use it
<mripard> and that person will probably send a patch for it, right ? ;)
<Turl> mripard: well it's not a bug, it's a "missing feature" :)
fluo75 has quit [Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/]
<arokux2> Turl: you've told me about busybox devmem, is there some standalone utility like this? that one seems to be present in busybox only
<Turl> arokux2: devmem2
<Turl> the license is bigger than the code :P
<granite_crusher> arokux2: I cant get that net I get this:
<granite_crusher> CONFIG_PACKET (Packet socket) and CONFIG_FILTER (Socket Filtering) sould be enabled in kernel
<arokux2> granite_crusher: well perfect then. enable them in the kernel
<arokux2> :)
<arokux2> granite_crusher: what was the program u used?
<granite_crusher> I just tried ifup eth0 and it spited out I need these
<arokux2> granite_crusher: ok. just enable them in kernel and everything should work
<granite_crusher> hmm if I search for them there is no such...
<arokux2> CONFIG_PACKET is there
<arokux2> granite_crusher: or you mean CONFIG_FILTER?
<granite_crusher> and if I search outside menuconfig there is hundred of files with that
<granite_crusher> both
<arokux2> granite_crusher: you should press "/" in menu config
<mnemoc> wingrime_: I haven't seen Ithamar seens he announced his packing/unpacking tools
<arokux2> then enter CONFIG_PACKET and take a look where it resides.. take a look at full path to it and its prompt
<mnemoc> wingrime_: can you download the pdf of that datasheet somehow?
Black_Horseman has joined #linux-sunxi
<granite_crusher> arokux2: probably I disabled something and can't get it with /
<arokux2> granite_crusher: you should get it..
<arokux2> mnemoc: add my key plz
<mnemoc> arokux2: I'm on that. will give you access to /SDK too, so you can maintain that :p
<hno> granite_crusher, / searches everything, even those disabled by other options.
<arokux2> mnemoc: yeah, but limit my access as much as possible
<arokux2> hno: hi.. can I ask you something?
<hno> arokux2, you can always ask.
<arokux2> hno: gpio_set_one_pin_io_status, gpio_set_one_pin_pull, gpio_write_one_pin_value
<arokux2> hno: do you know what these funcs do? ( I know the meaning of the first one)
<hno> arokux2, where are those? kernel or u-boot?
<arokux2> hno: kernel, 3.4
<hno> gpio_set_one_pin_pull sets I/O pull level, pull up/down strength.
<hno> gpio_set_one_pin_driver_level sets the I/O drive strength. Don't know what levels are available.
<arokux2> hno: I wasn't asking about that one
<mnemoc> arokux2: ssh arokux@sunxi.org ...... SDK/ is dl..../SDK and public_html os dl..../users/arokux
<arokux2> mnemoc: thanks
<hno> gpio_write_one_pin_value set the GPIO output value for a pin.
<hno> arokux2, I know you didn't ask for the _drive one, but it's very closely related to the _pull one.
<wingrime_> mnemoc: tryed , but ....
<arokux2> hno: ah, ok, I've thought you have mixed it with the other one I've asked about, sorry.
<hno> arokux2, see the PIO chapter in the user manaual. These all map directly to the corresponding fields for the pins in PIO.
<arokux2> hno: another question. how should I go about different gpio pins needed for different boards to enable usb?
<arokux2> hno: should there be defines per board, or how is it done?
<hno> arokux2, it's just different GPIO pins for VBUS power control. Noting strange. What do you mean exactly? For mainline it goes into DT. For 3.4 it's in script.bin.
<arokux2> hno: yep, my question is where it goes in u-boot.
<arokux2> hno: sorry, wasn't clear i'm talking about u-boot now..
<hno> Ah, there... a define with the GPIO number that needs to be enabled when enabling u-boot. can easily be set in boards.cfg .
focus_it has joined #linux-sunxi
<arokux2> alright. thanks hno
drachensun has quit [Read error: Operation timed out]
<arokux2> hno: how are you doing?
<n01> granite_crusher: what about rct
<n01> *rtc
<arokux2> hno: btw, mainline rtl driver works just fine with wlan
akaizen has joined #linux-sunxi
<tgaz> yay. musb gadget serial console working after FEL boot.
<arokux2> tgaz: you've added it?
<arokux2> tgaz: just curios, i'm not in context
<tgaz> arokux2, no, debugged it. jukivili added it
<tgaz> i added a musb_writeb(mbase, MUSB_FADDR, 0); in musb_g_reset() to get it to work
<tgaz> the musb driver assumes it starts up with FADDR=0 it seems
drachensun has joined #linux-sunxi
<hno> arokux2, I am a bit stressed at the moment, preparing last pieces for product launch.
<arokux2> hno: yeah, I know that is why i'm asking. hope everything goes as planned.
<hno> arokux2, will know in about 24 hours..
<arokux2> hno: :)
wingrime_ has quit [Read error: Connection reset by peer]
akaizen has quit [Remote host closed the connection]
rzk has joined #linux-sunxi
rz2k has quit [Ping timeout: 248 seconds]
skoti has quit [Quit: leaving]
massi has joined #linux-sunxi
<granite_crusher> n01: well I got error while booting mainline kernel about rtc, but looks it is problem what there is no rtc defined in device tree, and I am going to try add it
rzk has quit []
<Turl> granite_crusher: yeah but it's a different rtc :) yours is an external one
<Turl> n01 is working on the SoC-internal one
<n01> granite_crusher: which kernel on what boaord
<Turl> n01: A13 olinuxino iirc, has an i2c rtc from nxp
<n01> hooo, ok mine is internal for sun4i/7i
<n01> granite_crusher: good luck with that :)
<n01> anyway driver is done. I need to test it again
<arokux2> hno: I cannot understand something. if i won't set gpio to output 1 in 3.4, usb hub won't enumerate wlan adapter, however if you won't set gpio to 1 in mainline, wlan adapter will be enumerated regardless
<arokux2> hno: oh, forget it, it is enabled by somebody already
<granite_crusher> arokux2: ethernet is working :) (I succesfully apt-get update)
<arokux2> granite_crusher: cool. now set up tftp and forget about plugging and unplugging of a usb stick
focus_it has quit [Quit: Leaving]
<granite_crusher> arokux2: I have bad feeling about usb on that board... it have led which in both legacy and mainline kernel keeps lighting while it doesn't do so on laptop, (even if used as live usb), can it be result of that USB_EN open?
<granite_crusher> arokux2: I am talking about usb storage
<arokux2> granite_crusher: no idea, sorry
<arokux2> granite_crusher: led will never be on on laptop?
<granite_crusher> arokux2: no, just at moment it is read
<granite_crusher> or write
<arokux2> granite_crusher: hm.. I see, you'd need to ask Tsvetan
<arokux2> mripard: how do you think, who can set GPIO output 1 if not your code?
<arokux2> mripard: ignore.
<n01> who is Carlos Quijano?
<arokux2> hm.. do you guys know where gpios are touched in 3.4?
<arokux2> except of some obvious places..
aexl has quit [Quit: Leaving.]
<hno> arokux2, 3.4 kernel is quite dirty.. PIO is touched directly in too many places. But most are using the sys_config.c functions discussed earlier, or the implicit one requesting gpios which sets them to what is in script.bin.
<arokux2> hno: I've found it. they clear it in *_request func
<arokux2> hno: I do not know what to do.. I have 200 lines of code which works in 3.4 but fails in mainline
massi has quit [Quit: Leaving]
<arokux2> hno: maybe this is some regression in mainline in ehci-hcd.. I have no idea how I should track it down now
tinti has joined #linux-sunxi
_enrico_ has quit [Quit: Bye]
wingrime has quit [Ping timeout: 240 seconds]
akaizen has joined #linux-sunxi
<Turl> granite_crusher: did you get rtc working? :)
<Turl> n01: dunno :)
akaizen has quit [Remote host closed the connection]
<n01> I'm writing a new defconfig and scripts for mainline kernel in buildroot
<n01> what time is there Turl
<Turl> n01: 8PM
<Turl> n01: maybe in #buildroot someone recognises him
akaizen has joined #linux-sunxi
<Turl> n01: the guy is apparently spanish though, so 1AM for him :)
<n01> yeah, same here :)
<Turl> n01: btw, what device specific stuff does buildroot need?
<Turl> I've been using it for quite a while now
<Turl> and don't recall needing any black magic :)
<n01> Turl: just a new defconfig and a couple of script for making the sd
<arokux2> is there something like hrtimers in mainline for sunxi?
<Turl> n01: ah, so it's uboot integration, nice
<Turl> arokux2: yeah, maxime sent hrtimer patches like this week
<Turl> arokux2: I've picked them in sunxi-devel if you want to test drive them :)
<arokux2> Turl: I really have no idea what I can test :$
<Turl> arokux2: I asked the same to maxime, he said that if it boots it's working :)
<arokux2> Turl: I mean about usb.......
<Turl> arokux2: ah
<Turl> you think it's latency sensitive?
<Turl> arokux2: well, it's worth a shot I suppose
<arokux2> Turl: well... you know sometimes I see such errors in 3.4 too, but then it restores and works.
<arokux2> Turl: also, take a look here: drivers/usb/host/ehci-timer.c
<arokux2> Turl: it uses hrtimers...
<arokux2> Turl: Alan told it could be timing issue.
<arokux2> Turl: i'm not sure these timers are used, though.
<Turl> arokux2: go grab the hrtimer patches from lakml and give it a go then
<Turl> the old timer is just 24Mhz, might not have enough resolution
<arokux2> Turl: he has added them to everything except sun4i. is it a matter of adding support to dt?
<Turl> arokux2: no, sun4i doesn't have hrtimers
<Turl> the hw block is only present on sun5i+
<arokux2> Turl: ok...
<Turl> arokux2: I could give it a try on my sun7i though
<Turl> is your tree up to date?
<arokux2> Turl: yep
akaizen has quit [Remote host closed the connection]
<arokux2> Turl: ehci-timer.c wasn't there at 3.4 times.
<arokux2> Turl: but wait, what are you going to try? you do not have wifi module anyway
akaizen has joined #linux-sunxi
akaizen has quit [Remote host closed the connection]
<Turl> hm good point
<Turl> :P
<arokux2> nobody wants to try out a wifi dongle on mainline, what a shame :(
<arokux2> Turl: so sun4i timer has also resolution of 24Mhz?
<Turl> arokux2: it runs with osc24M so it should
<Turl> arokux2: I just remembered I have two ancient bluetooth USB thingies, they do the same as my soundcard
<Turl> (ie, just raise 2 interrupts and do nothing, and render the port unusable until next reboot)
<arokux2> Turl: you mean prevent usb from working?
<Turl> yeah
<Turl> I really think all of these things would work if we had OHCI support :)
<Turl> my pendrive still works
<arokux2> Turl: ... I can add OHCI, no problem, but why should I do it now?
<Turl> arokux2: so you can distinguish between "doesn't work because there's no OHCI support" vs "doesn't work because EHCI is broken"? :)
<Turl> arokux2: also, maybe wifi works in ohci only mode
<arokux2> Turl: you can try to put these sticks with legacy kernel
<arokux2> Turl: no wifi works on EHCI..
<Turl> USB2 devices theorically work just fine on USB1 hosts
<Turl> just like USB3 devices work on USB2 hosts
<Turl> it'll just be slow :)
<arokux2> Turl: maybe. you can easily comment out OHCI/EHCI (config won't help) in legacy 3.4 and see what happens.
<arokux2> granite_crusher: still working?
<granite_crusher> arokux2: trying to understand that device tree.... fex looked much more simple...