≡
[11:55:09] <cahbtexhuk_> question (maybe provocating) - what is the cheapest SBC?
[11:55:45] <cahbtexhuk_> all options are welcome, apart from "strip down tplink, it's awsum"
[11:56:12] <cahbtexhuk_> I need a linux host, usb, few gpios
[11:57:16] <Tsvetan> imx233-nano
[11:57:45] <Tsvetan> only eur 17.60 for 50+ pcs
[11:58:11] <cahbtexhuk_> I wish I needed 50+, but thanks
[11:59:56] <Tsvetan> eur 22 for 1 pce :)
[12:33:39] <tkoskine> +VAT if buying as a private person inside Europe.
[12:34:44] <Tsvetan> sure as for every other purchase
[12:45:34] <weigon> what is the latest kernel revid that is know to work on A20?
[12:45:49] <Tsvetan> 3.4
[12:46:10] <weigon> the one from http://olimex.wordpress.com/2013/11/05/building-the-ultimate-debian-sd-card-for-linux-with-kernel-3-4-for-a20-olinuxino-micro/ works, but linux-sunxi/stage-3.4 latest doesn't
[12:46:39] <Tsvetan> there is tag you can use for the git
[12:48:02] <tkoskine> sunxi-3.4 branch should work.
[12:49:06] <Tsvetan> weigon why do you make your own image and not use the one from the wiki?
[12:50:14] <weigon> Tsvetan: 1) 4G image on 16G sd-card is a bit of waste, 2) I would prefer the ubuntu rootfs instead of the debian one
[12:50:51] <Tsvetan> there are many things to consider, recently we see people who claim their board crash or do not work properly and they usually all have built their own images, when we ask them to use the image from the wiki they report board is working
[12:51:31] <Tsvetan> ok
[12:51:59] <weigon> the oopses are around do_lookup_rcu
[12:53:17] <weigon> ACTION was using 3.4.79+ (9c77e95e1e60d13699a67e440327b0b664eb9ff6)
[12:54:09] <Tsvetan> the wiki image is build with 3.4.67+ so you do not follow the instructions ;)
[12:54:54] <weigon> Tsvetan: quoting tkoskine: (11:48:02) tkoskine: sunxi-3.4 branch should work. :)
[12:55:34] <Tsvetan> linux-sunxi development is a bit chaotic in the morning PWM for LCD backlight is positive, in the afternoon it's negative as somebody decided so, next day vice versa again
[12:55:57] <Tsvetan> to stop this madness we just select one kernel release with tag and modify all around it the rest
[12:56:23] <Tsvetan> not saying that .79 will not work but you have to go through long list of features and check if your image match the kernel settings
[12:56:34] <tkoskine> I am having this kernel on my a20: http://files.tkoskine.me/linux-sunxi/olinuxino-a20/kernel-3.4.79-2014-02-13/ (sorry, no .config available, /proc/config.gz might give you one)
[12:57:18] <Tsvetan> it's not problem to build the image with any kernel 3.4.x but just quite lot of time is wasted on testing everything each time you compile it
[12:57:26] <weigon> tkoskine: I'll give it a spin
[12:58:44] <weigon> Tsvetan: good to know. Now I know what I can expect.
[12:59:12] <weigon> my A20 is just a headless build-box.
[13:00:27] <weigon> the debootstrap script gives my a very tiny footprint for this setup.
[14:45:41] <ssvb> weigon: do you have complete logs of these oopses?
[14:47:38] <ssvb> weigon: there are way too many different allwinner based devices
[14:48:30] <ssvb> weigon: and if olimex boards users do not cooperate, then it's difficult to ensure that they are fully supported :)
[14:49:47] <weigon> I'll do something like "git bisect" to find out when it broke
[14:50:13] <weigon> ACTION running a build on the board right now and doesn't want to interrupt it
[14:50:35] <ssvb> weigon: not a build of a kernel I hope?
[14:51:39] <weigon> just the MySQL Server 5.7.4 release ... something small
[14:51:51] <ssvb> ok :)
[14:52:26] <ssvb> thanks, it you can bisect these oopses, that would be very helpful
[14:53:12] <ssvb> but the first things to check would be the fex file, u-boot and kernel config
[14:55:31] <ssvb> gcc version used for building the kernel *might* be also the culprit in theory, things like https://github.com/linux-sunxi/linux-sunxi/commit/e6fffec6636f do happen sometimes
[14:57:00] <ssvb> but as Tsvetan said, use their images to verify that at least the hardware works correctly, and then move on to whatever works best for you :)
[14:58:09] <ssvb> weigon: is this your board? http://linux-sunxi.org/A20-olinuxino-micro
[15:08:02] <Aristide> Hi !
[15:08:18] <Aristide> My External HDD is not power on A20 USB ports :o
[15:08:49] <Aristide> Its an small Hard Disk Drive, and power is enought for use it on my laptop
[15:09:06] <weigon> ssvb: yep
[15:10:32] <ssvb> weigon: you can try to check the fex file referenced from there and also the build instructions
[15:11:40] <ssvb> weigon: things like the fex file there actually may be not up to date, so you can compare it with what olimex offers to use for their images
[16:19:54] <weigon> http://olimex.wordpress.com/2013/09/18/7795/ says 8388cf06936aaf05342fb912ed6fc1a3b04e5afb was used. But I can't find that rev in the stage/sunxi-3.4 ... my git foo is weak ;(
[16:22:38] <ssvb> weigon: the stage/sunxi-3.4 branch is always rebased after promiting/merging it to sunxi-3.4
[16:23:00] <ssvb> weigon: the chances are that this commit id is not alive anymore
[16:23:07] <weigon> yeah, guessed so
[16:23:37] <weigon> how would I bisect this then?
[16:23:50] <ssvb> oh, actually it is here - https://github.com/linux-sunxi/linux-sunxi/commit/8388cf06936aaf05342fb912ed6fc1a3b04e5afb
[16:24:12] <ssvb> so it's indeed all about your git skills
[16:24:27] <ssvb> try "git checkout 8388cf06936aaf05342fb912ed6fc1a3b04e5afb"
[16:27:01] <weigon> that's my problem: fatal: reference is not a tree: 8388cf06936aaf05342fb912ed6fc1a3b04e5afb ...
[16:27:14] <weigon> $ git show <revid> doesn't show it either
[16:27:33] <weigon> $ git branch
[16:27:33] <weigon> * stage/sunxi-3.4
[16:27:47] <weigon> that's the only branch I have
[16:29:26] <ssvb> try git fetch
[16:30:04] <weigon> git blame shows, that the change came in with: 24b8158a
[16:30:09] <ssvb> "git fetch" will sync your local copy with the repository on github
[16:30:22] <ssvb> what kind of change?
[16:30:57] <weigon> ssvb: https://github.com/linux-sunxi/linux-sunxi/commit/8388cf06936aaf05342fb912ed6fc1a3b04e5afb is https://github.com/linux-sunxi/linux-sunxi/commit/24b8158a for me
[16:31:55] <weigon> ssvb: the 1st link as no "branch"-info in github, the 2nd has "branch: sunxi-3.4"
[16:32:46] <ssvb> yes, 24b8158a is likely how it got merged into sunxi-3.4
[16:32:48] <weigon> ssvb: as you said, it is rebased, but not in the sunxi-3.4 branch anymore
[16:33:05] <weigon> ok, I'll bisect against that
[16:33:22] <ssvb> this commit id does not necessarily represent the same state as 8388cf06936aaf05342fb912ed6fc1a3b04e5afb
[16:33:55] <weigon> good point
[16:34:08] <weigon> let's see if the older build is still in the branch
[16:34:11] <ssvb> bad boys from olimex should not have used a commit from the stage branch in their instructions :)
[16:35:00] <ssvb> but if you just need a working kernel, then you can use some of the linux-sunxi tags
[16:35:32] <ssvb> actually I'm not even sure why the current sunxi-3.4 (3.4.79) does not work for you
[16:39:42] <ssvb> weigon: how did you end up in your broken state by the way? did you follow the http://linux-sunxi.org/Manual_build_howto ?
[16:40:32] <weigon> ssvb: I followed http://olimex.wordpress.com/2013/09/18/7795/ but used the latest rev from linux-sunxi
[16:41:39] <ssvb> do you have no access to 8388cf06936aaf05342fb912ed6fc1a3b04e5afb ?
[16:45:00] <weigon> ssvb: http://pastebin.com/raw.php?i=vYzpvEuQ is the oops I get btw
[16:46:04] <ssvb> very interesting
[16:46:29] <ssvb> do you have a complete dmesg log from the start of bootup?
[16:47:49] <weigon> rcu is read-copy-update ... that is quite likely very sensitive to wrong optimizations
[16:47:59] <ssvb> you can probably indeed use 24b8158a to see if it works better, but if not, I have pushed it here - https://github.com/ssvb/linux-sunxi/tree/8388cf06936a
[16:48:49] <ssvb> still it looks like 3.4.61 with minimal changes
[16:48:50] <weigon> gcc 4.8.1 is being used here
[16:52:16] <weigon> ssvb: the complete dump I have only split into /var/log/messages and /var/log/debug
[17:01:22] <ssvb> weigon: anyway, right now you are trying to reproduce the build instructions from the olimex blog, right?
[17:01:40] <weigon> ssvb: yep
[17:01:53] <ssvb> the only thing that is missing is the 8388cf06936aaf05342fb912ed6fc1a3b04e5afb commit id
[17:02:01] <weigon> ssvb: yep
[17:02:24] <weigon> ... and for some reason after a simple reboot it doesn't crash right away anymore
[17:03:18] <ssvb> hmm, 3.4.79 does not crash anymore? or the older one like 24b8158a?
[17:03:34] <weigon> 3.4.79
[17:03:45] <ssvb> can you get the dmesg log?
[17:04:01] <ssvb> and your kernel config from /proc/config.gz
[17:05:02] <weigon> let me reboot to get a full log
[17:05:34] <ssvb> it would be interesting to check if there is something obviously wrong or suspicious there
[17:06:35] <ssvb> btw, is your power supply powerful enough? do you connect anything power hungry such as sata hard drive?
[17:07:58] <weigon> ssvb: I do have a sata drive attached, yes. 12V 0.5A it can provide.
[17:08:24] <weigon> I also have a fully charged 6600mAh LiPo
[17:09:18] <weigon> ... attached
[17:09:21] <ssvb> 3.5" sata?
[17:09:36] <weigon> 2.5"
[17:09:52] <ssvb> it should not need 12v then
[17:10:27] <weigon> ACTION phrased it awkwardly
[17:10:59] <ssvb> Tsvetan: can LiPo battery provide enough power for a sata hard drive?
[17:11:00] <weigon> ssvb: I do have a 2.5" sata drive attached, yes. The power supply provides 12V 0.5A (6W)
[17:13:20] <NeuhNeuh> Hi !
[17:13:33] <ssvb> weigon: I think that 6W is too weak, that would be equivalent to only a little bit more than 1A at 5V
[17:13:49] <weigon> https://www.olimex.com/wiki/A20-OLinuXino-MICRO#Power_supply_and_consumption says the A20-board needs between 2-3W
[17:14:22] <ssvb> you can barely only power the board alone without a hard drive with that
[17:14:28] <NeuhNeuh> nand don't appear in /dev/ :(
[17:14:55] <weigon> ssvb: can the axp209 report how much power is drawn in?
[17:15:06] <ssvb> weigon: yes
[17:15:34] <weigon> is that exposed in /sys or /proc somewhere?
[17:16:23] <NeuhNeuh> I get this list of dev entry http://wstaw.org/m/2014/04/07/plasma-desktopzJ9178.png but none about nand flash
[17:17:23] <ssvb> weigon: incidentally that is handled by https://www.mail-archive.com/linux-sunxi@googlegroups.com/msg03487.html (which is not pushed to linux-sunxi git yet but lives in my private branch) :)
[17:18:46] <ssvb> weigon: oh, but I'm not sure if sata power consumption is going to be included there on your board
[17:19:25] <ssvb> weigon: but you can always try with and without a sata hdd drive attached and compare the reported numbers
[17:20:42] <weigon> luckily I have a rootfs on both, the sd-card and the hdd ;)
[17:21:15] <ssvb> weigon: the https://www.olimex.com/wiki/A20-OLinuXino-MICRO#Power_supply_and_consumption wiki is likely not including the power consumption of a hard drive
[17:22:01] <ssvb> weigon: this looks more like the board without any power hungry peripherals
[17:22:36] <ssvb> weigon: and when the hard drive is spinning up (if it is a mechanical drive), the power usage spikes quite high
[17:24:43] <weigon> .oO( debootstrap is a safe bet to get a oops )
[17:25:56] <weigon> this time from sys_rename(), but still the same do_path_lookup -> path_lookupat -> ... do_lookup -> __do_lookup_rcu
[17:26:51] <ssvb> just put your sata hard drive aside and run some testing without it, if the oopses disappear, then you simply don't have enough power
[17:26:59] <weigon> *nod*
[17:28:31] <NeuhNeuh> Its strange
[17:28:58] <NeuhNeuh> Without microSD card, Olinuxino boot on Nand flash ... But with SD Card, I don't have access to nand flash (not list in fdisk -l or on /disk) :/
[17:29:10] <NeuhNeuh> *on /dev/disk
[17:32:10] <NeuhNeuh> Because speed of my microSD is very ... Not good x) (/dev/mmcblk0: Timing cached reads: 702 MB in 2.00 seconds = 350.95 MB/sec - Timing buffered disk reads: 50 MB in 3.03 seconds = 16.51 MB/sec)
[17:33:58] <zwenni> i wasn't here from the beginning so which os are you talking about ?
[17:34:33] <NeuhNeuh> zwenni: My Olinuxino can boot on Nand flash without problem :) But I don't have access to Nandflash from Debian (running on this microSD)
[17:34:56] <NeuhNeuh> Is not listed in fdisk -l, or /proc/partitions or /dev/disk/by-path
[17:35:02] <zwenni> yeah i think debian isn't able at all to access nand flash
[17:35:21] <NeuhNeuh> zwenni: oO'
[17:35:41] <zwenni> i had the same problem but didnt found any solution
[17:36:28] <NeuhNeuh> zwenni: So i have an other question
[17:36:29] <zwenni> i didn't know excatly anymore but i think there aren't the drivers todo this
[17:36:34] <NeuhNeuh> Hard disk drive is not power on on USB
[17:36:49] <NeuhNeuh> Its normal ?
[17:37:14] <zwenni> how is the harddisk drive power ?
[17:37:18] <zwenni> just over usb ?
[17:37:37] <NeuhNeuh> Yes its a small disk
[17:37:46] <NeuhNeuh> And Computer's USB Port is enought for power
[17:37:52] <NeuhNeuh> (One cable)
[17:37:57] <weigon> ssvb: oops without SATA :)
[17:38:41] <zwenni> if i remember right the maxium power on usb is about 500mA
[17:38:47] <zwenni> or 350mA
[17:39:11] <zwenni> refer to the docs from the olimex wiki
[17:39:45] <NeuhNeuh> Ok, my HDD need 5V and 0.1A x)
[17:39:51] <NeuhNeuh> 1.0A*
[17:39:58] <zwenni> same here
[17:40:04] <NeuhNeuh> Ok
[17:40:08] <zwenni> so you may need a powerdes usb hub
[17:40:15] <zwenni> powered*
[17:40:26] <ssvb> weigon: oh, do they look similar to what you had before?
[17:41:14] <weigon> yep, it always ends up in do_lookup_rcu after doing some file-operation
[17:41:41] <ssvb> weigon: in any case, it makes sense to try the olimex instructions for the older kernel, because they are supposed to be verified to work on this hardware
[17:41:48] <weigon> path_openat() -> ... -> __d_lookup_rcu()
[17:42:39] <weigon> ssvb: yep, I'll build older kernels and see when they start to not crash anymore
[17:47:12] <NeuhNeuh> So, zwenni the problem is maybe an unloaded Kernel module
[17:47:21] <NeuhNeuh> I try to get list of modules but -l return an error :
[17:47:30] <NeuhNeuh> root@a20-OLinuXino:/# modprobe -l
[17:47:30] <NeuhNeuh> modprobe: invalid option -- 'l'
[17:47:31] <NeuhNeuh> :')
[18:19:42] <youmix> Hi
[18:20:35] <youmix> my olinuxino a13 doesn't detect my usb 3g modem by /dev/ttyUSB
[18:21:14] <Claude> youmix, what does dmesg say?
[18:21:25] <youmix> AND i HAVE CONFIG_USB_SERIAL and CONFIG_USB_SERIAL_FTDI_SIO built as modules
[18:21:38] <youmix> soory for the capital letters
[18:21:58] <Claude> hmm isn't a typical usb 3g modem not an cdc-acm device?
[18:22:44] <youmix> yes is the mod-usb3g by olimex
[18:23:05] <youmix> it says thar is compatible with debia
[18:23:09] <youmix> n
[18:24:09] <Claude> yeah , unplug the modem , plug it in again and type in dmesg. should tell you if the kernel can talk to your modem
[18:25:27] <Claude> best is copy your dmesg output (last couple , 30 or so , lines) here and post the url http://pastie.org/
[18:25:43] <szwenni> someone here ever tried to connect a usb touchscreen to android on a20 ?
[18:25:47] <youmix> it says usb 2-13: new full-speed USB device number 8 using sw-ehci
[18:26:13] <Claude> youmix, nothing more?
[18:26:20] <youmix> if I use lsusb it says that it is a huawei modem
[18:26:40] <Claude> yeah but dmesg tell if the kernel has a driver for your modem
[18:27:02] <youmix> no, nothing more Claude
[18:28:38] <youmix> the rest is for the mouse and the keyboard
[18:28:43] <Claude> okay i guess the usb cdc-acm driver is missing then
[18:29:35] <youmix> it says that us(2.9): stall on control pipe a lot
[18:29:56] <youmix> how can I install the cdc-acm driver Claude?
[18:30:09] <Claude> kernel config
[18:30:38] <youmix> Do I need to build the kernel?
[18:31:31] <Claude> yeah or a module
[18:32:58] <youmix> how can I build a module? I am a newbie I use the r18 image for the a13
[18:37:58] <szwenni> make M=/pazh/to/module/source
[18:38:35] <Claude> https://www.olimex.com/wiki/A13-OLinuXino
[18:39:27] <Claude> youmix, you might check the 3g modem first on windows or linux if it really works :)
[18:39:52] <youmix> yes it works in both Claude
[18:40:26] <Claude> ah you have a linux box? what does dmesg say on that box when you plug in the modem? should give you a hint on the driver then
[18:45:24] <youmix> i'm going to try
[18:52:43] <szwenni> get usb-utils package then with usb-devics command you can get everything you want to know
[18:55:12] <Claude> got to go..
[18:55:45] <youmix> claude this is my dmesg on ubuntu: CONFIG_USB_SERIAL and CONFIG_USB_SERIAL_FTDI_SIO built as modules
[18:55:48] <youmix> sorry
[18:56:10] <youmix> this one: www.pastie.org/9001019
[18:57:12] <youmix> as you can see it creates the ttyUSB
[18:58:06] <youmix> szwenni how can i install packages on the olinuxino via USB? I don't have the wifi version
[18:58:56] <szwenni> debian as system ?
[19:00:03] <youmix> I use Ubuntu on my laptop and Debian on the olinuxino szwenni
[19:01:07] <szwenni> install or transfer ?
[19:01:35] <szwenni> ahnow i understand you want to insatll the usb-utils package right ?
[19:02:14] <youmix> yes
[19:02:26] <youmix> if that is going to help me...
[19:02:54] <youmix> I would like to know how can I install the cdc-acm module in the olinuxino
[19:02:57] <szwenni> yeah don't but ithink you can see what driver is loaded to use the 3g module on ubuntu
[19:03:17] <youmix> aah ok I'll try
[19:03:39] <szwenni> so you have to install the usbutils on ubuntu
[19:03:56] <szwenni> apt-get install usbutils
[19:04:49] <szwenni> the connect the 3g modem to ubuntu an run 'usb-devices'
[19:06:03] <szwenni> maybe 'usb-devices| grep HSDPA
[19:06:07] <szwenni> '
[19:08:52] <youmix> thank you szwenni I'm trying that
[19:09:30] <szwenni> tell me whats the result
[19:11:21] <szwenni> brb
[19:13:18] <youmix> szwenni www.pastie.org/9001052
[19:18:50] <youmix> szwenni do you have any idea?
[19:28:09] <doubtman> How can I install the cdc-acm module in a olinuxino?
[19:52:28] <szwenni> so this looks like their is ni real driver
[19:53:10] <szwenni> the 3g modem is connect via usb right
[19:53:22] <szwenni> and just 1 usb port
[20:10:22] <a13oli> I'm trying to build my own kernel but when i wrote the instruction:
[20:10:23] <a13oli> make A13-OLinuXino CROSS_COMPILE=arm-linux-gnueabihf-
[20:10:28] <a13oli> it gives me an error
[20:10:42] <a13oli> I think it is because the instruction is not complete
[20:10:54] <a13oli> but it is in the manual this way
[20:11:13] <a13oli> http://olimex.wordpress.com/2014/01/15/building-debian-linux-bootable-sd-card-with-hardware-accelerated-video-decoding-and-kernel-3-4-for-a13-olinuxino/
[20:14:00] <a13oli> the error says system not configured
[20:26:59] <szwenni> you run make menuconfig or something ?
[20:28:23] <a13oli> i run make A13-OLinuXino CROSS_COMPILE=arm-linux-gnueabihf- which is what it says in the manual
[20:30:40] <szwenni> ehh in which directory you run ?
[20:32:51] <szwenni> and is their a .config in that directory
[20:51:54] <szwenni> a13oli you got sth ?
[21:35:25] <oliv3r> Tsvetan: I need a few words with you!
[21:35:29] <oliv3r> when wan we talk!
[21:35:47] <oliv3r> can
[21:54:21] <szwenni> someone here who can help me building android
[21:54:44] <szwenni> or who else had a fail a booting the kernel
[21:55:17] <szwenni> axp_mfd 0-0034: failes reading at 0x03
IRC #olimex 2014-04-07
[11:54:43] <cahbtexhuk_> hi all[11:55:09] <cahbtexhuk_> question (maybe provocating) - what is the cheapest SBC?
[11:55:45] <cahbtexhuk_> all options are welcome, apart from "strip down tplink, it's awsum"
[11:56:12] <cahbtexhuk_> I need a linux host, usb, few gpios
[11:57:16] <Tsvetan> imx233-nano
[11:57:45] <Tsvetan> only eur 17.60 for 50+ pcs
[11:58:11] <cahbtexhuk_> I wish I needed 50+, but thanks
[11:59:56] <Tsvetan> eur 22 for 1 pce :)
[12:33:39] <tkoskine> +VAT if buying as a private person inside Europe.
[12:34:44] <Tsvetan> sure as for every other purchase
[12:45:34] <weigon> what is the latest kernel revid that is know to work on A20?
[12:45:49] <Tsvetan> 3.4
[12:46:10] <weigon> the one from http://olimex.wordpress.com/2013/11/05/building-the-ultimate-debian-sd-card-for-linux-with-kernel-3-4-for-a20-olinuxino-micro/ works, but linux-sunxi/stage-3.4 latest doesn't
[12:46:39] <Tsvetan> there is tag you can use for the git
[12:48:02] <tkoskine> sunxi-3.4 branch should work.
[12:49:06] <Tsvetan> weigon why do you make your own image and not use the one from the wiki?
[12:50:14] <weigon> Tsvetan: 1) 4G image on 16G sd-card is a bit of waste, 2) I would prefer the ubuntu rootfs instead of the debian one
[12:50:51] <Tsvetan> there are many things to consider, recently we see people who claim their board crash or do not work properly and they usually all have built their own images, when we ask them to use the image from the wiki they report board is working
[12:51:31] <Tsvetan> ok
[12:51:59] <weigon> the oopses are around do_lookup_rcu
[12:53:17] <weigon> ACTION was using 3.4.79+ (9c77e95e1e60d13699a67e440327b0b664eb9ff6)
[12:54:09] <Tsvetan> the wiki image is build with 3.4.67+ so you do not follow the instructions ;)
[12:54:54] <weigon> Tsvetan: quoting tkoskine: (11:48:02) tkoskine: sunxi-3.4 branch should work. :)
[12:55:34] <Tsvetan> linux-sunxi development is a bit chaotic in the morning PWM for LCD backlight is positive, in the afternoon it's negative as somebody decided so, next day vice versa again
[12:55:57] <Tsvetan> to stop this madness we just select one kernel release with tag and modify all around it the rest
[12:56:23] <Tsvetan> not saying that .79 will not work but you have to go through long list of features and check if your image match the kernel settings
[12:56:34] <tkoskine> I am having this kernel on my a20: http://files.tkoskine.me/linux-sunxi/olinuxino-a20/kernel-3.4.79-2014-02-13/ (sorry, no .config available, /proc/config.gz might give you one)
[12:57:18] <Tsvetan> it's not problem to build the image with any kernel 3.4.x but just quite lot of time is wasted on testing everything each time you compile it
[12:57:26] <weigon> tkoskine: I'll give it a spin
[12:58:44] <weigon> Tsvetan: good to know. Now I know what I can expect.
[12:59:12] <weigon> my A20 is just a headless build-box.
[13:00:27] <weigon> the debootstrap script gives my a very tiny footprint for this setup.
[14:45:41] <ssvb> weigon: do you have complete logs of these oopses?
[14:47:38] <ssvb> weigon: there are way too many different allwinner based devices
[14:48:30] <ssvb> weigon: and if olimex boards users do not cooperate, then it's difficult to ensure that they are fully supported :)
[14:49:47] <weigon> I'll do something like "git bisect" to find out when it broke
[14:50:13] <weigon> ACTION running a build on the board right now and doesn't want to interrupt it
[14:50:35] <ssvb> weigon: not a build of a kernel I hope?
[14:51:39] <weigon> just the MySQL Server 5.7.4 release ... something small
[14:51:51] <ssvb> ok :)
[14:52:26] <ssvb> thanks, it you can bisect these oopses, that would be very helpful
[14:53:12] <ssvb> but the first things to check would be the fex file, u-boot and kernel config
[14:55:31] <ssvb> gcc version used for building the kernel *might* be also the culprit in theory, things like https://github.com/linux-sunxi/linux-sunxi/commit/e6fffec6636f do happen sometimes
[14:57:00] <ssvb> but as Tsvetan said, use their images to verify that at least the hardware works correctly, and then move on to whatever works best for you :)
[14:58:09] <ssvb> weigon: is this your board? http://linux-sunxi.org/A20-olinuxino-micro
[15:08:02] <Aristide> Hi !
[15:08:18] <Aristide> My External HDD is not power on A20 USB ports :o
[15:08:49] <Aristide> Its an small Hard Disk Drive, and power is enought for use it on my laptop
[15:09:06] <weigon> ssvb: yep
[15:10:32] <ssvb> weigon: you can try to check the fex file referenced from there and also the build instructions
[15:11:40] <ssvb> weigon: things like the fex file there actually may be not up to date, so you can compare it with what olimex offers to use for their images
[16:19:54] <weigon> http://olimex.wordpress.com/2013/09/18/7795/ says 8388cf06936aaf05342fb912ed6fc1a3b04e5afb was used. But I can't find that rev in the stage/sunxi-3.4 ... my git foo is weak ;(
[16:22:38] <ssvb> weigon: the stage/sunxi-3.4 branch is always rebased after promiting/merging it to sunxi-3.4
[16:23:00] <ssvb> weigon: the chances are that this commit id is not alive anymore
[16:23:07] <weigon> yeah, guessed so
[16:23:37] <weigon> how would I bisect this then?
[16:23:50] <ssvb> oh, actually it is here - https://github.com/linux-sunxi/linux-sunxi/commit/8388cf06936aaf05342fb912ed6fc1a3b04e5afb
[16:24:12] <ssvb> so it's indeed all about your git skills
[16:24:27] <ssvb> try "git checkout 8388cf06936aaf05342fb912ed6fc1a3b04e5afb"
[16:27:01] <weigon> that's my problem: fatal: reference is not a tree: 8388cf06936aaf05342fb912ed6fc1a3b04e5afb ...
[16:27:14] <weigon> $ git show <revid> doesn't show it either
[16:27:33] <weigon> $ git branch
[16:27:33] <weigon> * stage/sunxi-3.4
[16:27:47] <weigon> that's the only branch I have
[16:29:26] <ssvb> try git fetch
[16:30:04] <weigon> git blame shows, that the change came in with: 24b8158a
[16:30:09] <ssvb> "git fetch" will sync your local copy with the repository on github
[16:30:22] <ssvb> what kind of change?
[16:30:57] <weigon> ssvb: https://github.com/linux-sunxi/linux-sunxi/commit/8388cf06936aaf05342fb912ed6fc1a3b04e5afb is https://github.com/linux-sunxi/linux-sunxi/commit/24b8158a for me
[16:31:55] <weigon> ssvb: the 1st link as no "branch"-info in github, the 2nd has "branch: sunxi-3.4"
[16:32:46] <ssvb> yes, 24b8158a is likely how it got merged into sunxi-3.4
[16:32:48] <weigon> ssvb: as you said, it is rebased, but not in the sunxi-3.4 branch anymore
[16:33:05] <weigon> ok, I'll bisect against that
[16:33:22] <ssvb> this commit id does not necessarily represent the same state as 8388cf06936aaf05342fb912ed6fc1a3b04e5afb
[16:33:55] <weigon> good point
[16:34:08] <weigon> let's see if the older build is still in the branch
[16:34:11] <ssvb> bad boys from olimex should not have used a commit from the stage branch in their instructions :)
[16:35:00] <ssvb> but if you just need a working kernel, then you can use some of the linux-sunxi tags
[16:35:32] <ssvb> actually I'm not even sure why the current sunxi-3.4 (3.4.79) does not work for you
[16:39:42] <ssvb> weigon: how did you end up in your broken state by the way? did you follow the http://linux-sunxi.org/Manual_build_howto ?
[16:40:32] <weigon> ssvb: I followed http://olimex.wordpress.com/2013/09/18/7795/ but used the latest rev from linux-sunxi
[16:41:39] <ssvb> do you have no access to 8388cf06936aaf05342fb912ed6fc1a3b04e5afb ?
[16:45:00] <weigon> ssvb: http://pastebin.com/raw.php?i=vYzpvEuQ is the oops I get btw
[16:46:04] <ssvb> very interesting
[16:46:29] <ssvb> do you have a complete dmesg log from the start of bootup?
[16:47:49] <weigon> rcu is read-copy-update ... that is quite likely very sensitive to wrong optimizations
[16:47:59] <ssvb> you can probably indeed use 24b8158a to see if it works better, but if not, I have pushed it here - https://github.com/ssvb/linux-sunxi/tree/8388cf06936a
[16:48:49] <ssvb> still it looks like 3.4.61 with minimal changes
[16:48:50] <weigon> gcc 4.8.1 is being used here
[16:52:16] <weigon> ssvb: the complete dump I have only split into /var/log/messages and /var/log/debug
[17:01:22] <ssvb> weigon: anyway, right now you are trying to reproduce the build instructions from the olimex blog, right?
[17:01:40] <weigon> ssvb: yep
[17:01:53] <ssvb> the only thing that is missing is the 8388cf06936aaf05342fb912ed6fc1a3b04e5afb commit id
[17:02:01] <weigon> ssvb: yep
[17:02:24] <weigon> ... and for some reason after a simple reboot it doesn't crash right away anymore
[17:03:18] <ssvb> hmm, 3.4.79 does not crash anymore? or the older one like 24b8158a?
[17:03:34] <weigon> 3.4.79
[17:03:45] <ssvb> can you get the dmesg log?
[17:04:01] <ssvb> and your kernel config from /proc/config.gz
[17:05:02] <weigon> let me reboot to get a full log
[17:05:34] <ssvb> it would be interesting to check if there is something obviously wrong or suspicious there
[17:06:35] <ssvb> btw, is your power supply powerful enough? do you connect anything power hungry such as sata hard drive?
[17:07:58] <weigon> ssvb: I do have a sata drive attached, yes. 12V 0.5A it can provide.
[17:08:24] <weigon> I also have a fully charged 6600mAh LiPo
[17:09:18] <weigon> ... attached
[17:09:21] <ssvb> 3.5" sata?
[17:09:36] <weigon> 2.5"
[17:09:52] <ssvb> it should not need 12v then
[17:10:27] <weigon> ACTION phrased it awkwardly
[17:10:59] <ssvb> Tsvetan: can LiPo battery provide enough power for a sata hard drive?
[17:11:00] <weigon> ssvb: I do have a 2.5" sata drive attached, yes. The power supply provides 12V 0.5A (6W)
[17:13:20] <NeuhNeuh> Hi !
[17:13:33] <ssvb> weigon: I think that 6W is too weak, that would be equivalent to only a little bit more than 1A at 5V
[17:13:49] <weigon> https://www.olimex.com/wiki/A20-OLinuXino-MICRO#Power_supply_and_consumption says the A20-board needs between 2-3W
[17:14:22] <ssvb> you can barely only power the board alone without a hard drive with that
[17:14:28] <NeuhNeuh> nand don't appear in /dev/ :(
[17:14:55] <weigon> ssvb: can the axp209 report how much power is drawn in?
[17:15:06] <ssvb> weigon: yes
[17:15:34] <weigon> is that exposed in /sys or /proc somewhere?
[17:16:23] <NeuhNeuh> I get this list of dev entry http://wstaw.org/m/2014/04/07/plasma-desktopzJ9178.png but none about nand flash
[17:17:23] <ssvb> weigon: incidentally that is handled by https://www.mail-archive.com/linux-sunxi@googlegroups.com/msg03487.html (which is not pushed to linux-sunxi git yet but lives in my private branch) :)
[17:18:46] <ssvb> weigon: oh, but I'm not sure if sata power consumption is going to be included there on your board
[17:19:25] <ssvb> weigon: but you can always try with and without a sata hdd drive attached and compare the reported numbers
[17:20:42] <weigon> luckily I have a rootfs on both, the sd-card and the hdd ;)
[17:21:15] <ssvb> weigon: the https://www.olimex.com/wiki/A20-OLinuXino-MICRO#Power_supply_and_consumption wiki is likely not including the power consumption of a hard drive
[17:22:01] <ssvb> weigon: this looks more like the board without any power hungry peripherals
[17:22:36] <ssvb> weigon: and when the hard drive is spinning up (if it is a mechanical drive), the power usage spikes quite high
[17:24:43] <weigon> .oO( debootstrap is a safe bet to get a oops )
[17:25:56] <weigon> this time from sys_rename(), but still the same do_path_lookup -> path_lookupat -> ... do_lookup -> __do_lookup_rcu
[17:26:51] <ssvb> just put your sata hard drive aside and run some testing without it, if the oopses disappear, then you simply don't have enough power
[17:26:59] <weigon> *nod*
[17:28:31] <NeuhNeuh> Its strange
[17:28:58] <NeuhNeuh> Without microSD card, Olinuxino boot on Nand flash ... But with SD Card, I don't have access to nand flash (not list in fdisk -l or on /disk) :/
[17:29:10] <NeuhNeuh> *on /dev/disk
[17:32:10] <NeuhNeuh> Because speed of my microSD is very ... Not good x) (/dev/mmcblk0: Timing cached reads: 702 MB in 2.00 seconds = 350.95 MB/sec - Timing buffered disk reads: 50 MB in 3.03 seconds = 16.51 MB/sec)
[17:33:58] <zwenni> i wasn't here from the beginning so which os are you talking about ?
[17:34:33] <NeuhNeuh> zwenni: My Olinuxino can boot on Nand flash without problem :) But I don't have access to Nandflash from Debian (running on this microSD)
[17:34:56] <NeuhNeuh> Is not listed in fdisk -l, or /proc/partitions or /dev/disk/by-path
[17:35:02] <zwenni> yeah i think debian isn't able at all to access nand flash
[17:35:21] <NeuhNeuh> zwenni: oO'
[17:35:41] <zwenni> i had the same problem but didnt found any solution
[17:36:28] <NeuhNeuh> zwenni: So i have an other question
[17:36:29] <zwenni> i didn't know excatly anymore but i think there aren't the drivers todo this
[17:36:34] <NeuhNeuh> Hard disk drive is not power on on USB
[17:36:49] <NeuhNeuh> Its normal ?
[17:37:14] <zwenni> how is the harddisk drive power ?
[17:37:18] <zwenni> just over usb ?
[17:37:37] <NeuhNeuh> Yes its a small disk
[17:37:46] <NeuhNeuh> And Computer's USB Port is enought for power
[17:37:52] <NeuhNeuh> (One cable)
[17:37:57] <weigon> ssvb: oops without SATA :)
[17:38:41] <zwenni> if i remember right the maxium power on usb is about 500mA
[17:38:47] <zwenni> or 350mA
[17:39:11] <zwenni> refer to the docs from the olimex wiki
[17:39:45] <NeuhNeuh> Ok, my HDD need 5V and 0.1A x)
[17:39:51] <NeuhNeuh> 1.0A*
[17:39:58] <zwenni> same here
[17:40:04] <NeuhNeuh> Ok
[17:40:08] <zwenni> so you may need a powerdes usb hub
[17:40:15] <zwenni> powered*
[17:40:26] <ssvb> weigon: oh, do they look similar to what you had before?
[17:41:14] <weigon> yep, it always ends up in do_lookup_rcu after doing some file-operation
[17:41:41] <ssvb> weigon: in any case, it makes sense to try the olimex instructions for the older kernel, because they are supposed to be verified to work on this hardware
[17:41:48] <weigon> path_openat() -> ... -> __d_lookup_rcu()
[17:42:39] <weigon> ssvb: yep, I'll build older kernels and see when they start to not crash anymore
[17:47:12] <NeuhNeuh> So, zwenni the problem is maybe an unloaded Kernel module
[17:47:21] <NeuhNeuh> I try to get list of modules but -l return an error :
[17:47:30] <NeuhNeuh> root@a20-OLinuXino:/# modprobe -l
[17:47:30] <NeuhNeuh> modprobe: invalid option -- 'l'
[17:47:31] <NeuhNeuh> :')
[18:19:42] <youmix> Hi
[18:20:35] <youmix> my olinuxino a13 doesn't detect my usb 3g modem by /dev/ttyUSB
[18:21:14] <Claude> youmix, what does dmesg say?
[18:21:25] <youmix> AND i HAVE CONFIG_USB_SERIAL and CONFIG_USB_SERIAL_FTDI_SIO built as modules
[18:21:38] <youmix> soory for the capital letters
[18:21:58] <Claude> hmm isn't a typical usb 3g modem not an cdc-acm device?
[18:22:44] <youmix> yes is the mod-usb3g by olimex
[18:23:05] <youmix> it says thar is compatible with debia
[18:23:09] <youmix> n
[18:24:09] <Claude> yeah , unplug the modem , plug it in again and type in dmesg. should tell you if the kernel can talk to your modem
[18:25:27] <Claude> best is copy your dmesg output (last couple , 30 or so , lines) here and post the url http://pastie.org/
[18:25:43] <szwenni> someone here ever tried to connect a usb touchscreen to android on a20 ?
[18:25:47] <youmix> it says usb 2-13: new full-speed USB device number 8 using sw-ehci
[18:26:13] <Claude> youmix, nothing more?
[18:26:20] <youmix> if I use lsusb it says that it is a huawei modem
[18:26:40] <Claude> yeah but dmesg tell if the kernel has a driver for your modem
[18:27:02] <youmix> no, nothing more Claude
[18:28:38] <youmix> the rest is for the mouse and the keyboard
[18:28:43] <Claude> okay i guess the usb cdc-acm driver is missing then
[18:29:35] <youmix> it says that us(2.9): stall on control pipe a lot
[18:29:56] <youmix> how can I install the cdc-acm driver Claude?
[18:30:09] <Claude> kernel config
[18:30:38] <youmix> Do I need to build the kernel?
[18:31:31] <Claude> yeah or a module
[18:32:58] <youmix> how can I build a module? I am a newbie I use the r18 image for the a13
[18:37:58] <szwenni> make M=/pazh/to/module/source
[18:38:35] <Claude> https://www.olimex.com/wiki/A13-OLinuXino
[18:39:27] <Claude> youmix, you might check the 3g modem first on windows or linux if it really works :)
[18:39:52] <youmix> yes it works in both Claude
[18:40:26] <Claude> ah you have a linux box? what does dmesg say on that box when you plug in the modem? should give you a hint on the driver then
[18:45:24] <youmix> i'm going to try
[18:52:43] <szwenni> get usb-utils package then with usb-devics command you can get everything you want to know
[18:55:12] <Claude> got to go..
[18:55:45] <youmix> claude this is my dmesg on ubuntu: CONFIG_USB_SERIAL and CONFIG_USB_SERIAL_FTDI_SIO built as modules
[18:55:48] <youmix> sorry
[18:56:10] <youmix> this one: www.pastie.org/9001019
[18:57:12] <youmix> as you can see it creates the ttyUSB
[18:58:06] <youmix> szwenni how can i install packages on the olinuxino via USB? I don't have the wifi version
[18:58:56] <szwenni> debian as system ?
[19:00:03] <youmix> I use Ubuntu on my laptop and Debian on the olinuxino szwenni
[19:01:07] <szwenni> install or transfer ?
[19:01:35] <szwenni> ahnow i understand you want to insatll the usb-utils package right ?
[19:02:14] <youmix> yes
[19:02:26] <youmix> if that is going to help me...
[19:02:54] <youmix> I would like to know how can I install the cdc-acm module in the olinuxino
[19:02:57] <szwenni> yeah don't but ithink you can see what driver is loaded to use the 3g module on ubuntu
[19:03:17] <youmix> aah ok I'll try
[19:03:39] <szwenni> so you have to install the usbutils on ubuntu
[19:03:56] <szwenni> apt-get install usbutils
[19:04:49] <szwenni> the connect the 3g modem to ubuntu an run 'usb-devices'
[19:06:03] <szwenni> maybe 'usb-devices| grep HSDPA
[19:06:07] <szwenni> '
[19:08:52] <youmix> thank you szwenni I'm trying that
[19:09:30] <szwenni> tell me whats the result
[19:11:21] <szwenni> brb
[19:13:18] <youmix> szwenni www.pastie.org/9001052
[19:18:50] <youmix> szwenni do you have any idea?
[19:28:09] <doubtman> How can I install the cdc-acm module in a olinuxino?
[19:52:28] <szwenni> so this looks like their is ni real driver
[19:53:10] <szwenni> the 3g modem is connect via usb right
[19:53:22] <szwenni> and just 1 usb port
[20:10:22] <a13oli> I'm trying to build my own kernel but when i wrote the instruction:
[20:10:23] <a13oli> make A13-OLinuXino CROSS_COMPILE=arm-linux-gnueabihf-
[20:10:28] <a13oli> it gives me an error
[20:10:42] <a13oli> I think it is because the instruction is not complete
[20:10:54] <a13oli> but it is in the manual this way
[20:11:13] <a13oli> http://olimex.wordpress.com/2014/01/15/building-debian-linux-bootable-sd-card-with-hardware-accelerated-video-decoding-and-kernel-3-4-for-a13-olinuxino/
[20:14:00] <a13oli> the error says system not configured
[20:26:59] <szwenni> you run make menuconfig or something ?
[20:28:23] <a13oli> i run make A13-OLinuXino CROSS_COMPILE=arm-linux-gnueabihf- which is what it says in the manual
[20:30:40] <szwenni> ehh in which directory you run ?
[20:32:51] <szwenni> and is their a .config in that directory
[20:51:54] <szwenni> a13oli you got sth ?
[21:35:25] <oliv3r> Tsvetan: I need a few words with you!
[21:35:29] <oliv3r> when wan we talk!
[21:35:47] <oliv3r> can
[21:54:21] <szwenni> someone here who can help me building android
[21:54:44] <szwenni> or who else had a fail a booting the kernel
[21:55:17] <szwenni> axp_mfd 0-0034: failes reading at 0x03
