| 03:13 | mactimes | Hi. I'd like to install Debian on Cubox-i4pro (I4P-300-D) via netboot. I've tried following information from the forum and other tutorials I could find, but all of them seem too old to work properly. Would someone be willing to help me out on this, please? |
| 03:35 | mk01 | mactimes: if TFTP is an option for you (affaik PXE is not supported) |
| 03:40 | _unreal_ | mk01, hi |
| 03:40 | _unreal_ | I have an orignial cubox that was given to me, one of the 800mhz development one's? |
| 03:41 | _unreal_ | it has nothing on it and I dont know much about it. looking for help on getting it up and running. not sure what the best build for it would be |
| 03:44 | mactimes | mk01 Can't I just throw the files into mmcblk0p1? |
| 03:48 | mactimes | _unreal_ If you wanna keep it real 'simple', check this out: http://www.solid-run.com/wiki/Ignition |
| 03:49 | _unreal_ | all I know is I have the first edition I gues |
| 03:49 | _unreal_ | guess. it doesnt even have any markings on it. |
| 03:50 | _unreal_ | heh mactimes that wont work for me, I dont have the cuboxi |
| 03:50 | _unreal_ | cubox-i |
| 03:50 | mk01 | mactimes: so what you want ... autodeployement via network or you have physical access to the device and no problem writing SD card in another linux ? |
| 03:51 | mactimes | _unreal_ Sorry, I'm new to this too. Have made some tests, but nothing much that could be helpful in your case. |
| 03:51 | _unreal_ | http://www.cnx-software.com/2011/12/14/solid-run-cubox-open-source-platform-for-android-tv-media-center-and-nas-development/ |
| 03:51 | _unreal_ | this is the version I have |
| 03:51 | _unreal_ | there are no logos etc.. on it |
| 03:52 | mactimes | mk01 Either cases would work for me. |
| 03:53 | mactimes | mk01 What I'd like to have (which I don't know if it would be currently possible) is to start the installation via network, so that I can properly partition and encrypt the e-SATA disk connected to the Cubox during the installation. |
| 03:54 | mk01 | mactimes: the 'problem' is that CBi doesn't have permanent storage. so to boot (load netboot installer image) from network means - preparing SD with uboot to start uboot and load script you write (although oneliner) before to load .img via tftp and start it |
| 03:56 | mactimes | mk01 Could you help me around that? |
| 03:57 | mk01 | mactimes: sure, but that means that by the time we finish you can already boot into 2nd staged debootstrap (with 1st stage done on any linux you have up). |
| 03:58 | mactimes | mk01 I have console access and a SD card reader attached to my PC here. Currently running Debian Wheezy here. |
| 03:58 | mk01 | ok |
| 03:59 | mactimes | mk01 I'm ready when you are. Just tell me what to do. =-) |
| 03:59 | mk01 | and you want jessie or wheezy ? also headless or you want full "desktop" ? |
| 04:01 | mactimes | mk01 Well, I've tried wheezy from the ignition install and that was kind of 'broken'. lshw as root wouldn't work and as a regular user, not much information about it. |
| 04:01 | mactimes | mk01 So, I guess I'll go with Jessie. |
| 04:01 | mactimes | mk01 Full desktop. |
| 04:01 | _unreal_ | mk01, I'm also trying to figure out how to setup my cubox.... but I cant wait if your going to be around for a while |
| 04:01 | _unreal_ | CAN can WAIT |
| 04:01 | mactimes | mk01 I followed some guidelines for Wheezy, but in the end I didn't have Wireless working... |
| 04:02 | mk01 | _unreal_: i'm affraid I don't have any personal experience with the armada cubox... the one before CBi |
| 04:03 | _unreal_ | HUM |
| 04:03 | mactimes | mk01 I can have something similar to 'base install', and then I'll just 'apt-get install gnome'. Don't worry about the 'desktop' part. |
| 04:04 | _unreal_ | I believe http://download.solid-run.com/pub/solidrun/ the cubox link is the one for mine |
| 04:04 | _unreal_ | 2013 |
| 04:04 | mactimes | mk01 Just want to have the system to boot the netinstall properly, just like netinstall with SSH access or something like that. |
| 04:06 | mk01 | mactimes: yes I got the point what you want. there is one problem though - there is currently no distro providing FULL support for the hardware with it's packages... |
| 04:07 | mactimes | mk01 Just one question before you start: These files (http://ftp.nl.debian.org/debian/dists/jessie/main/installer-armhf/current/images/netboot/) are really useless for this purpose, right? |
| 04:07 | mactimes | mk01 Hummm. I see. |
| 04:09 | mactimes | mk01 No way I can start a Debian Installer, then, right? |
| 04:09 | mk01 | what is upstream (so part of kernels provided by distros like Debian) is probably bootable kernel on imx6 (CBi), but max what you get is txt(fb) console. |
| 04:11 | mactimes | mk01 Well, let's have it the _hard_ way, then. =) |
| 04:11 | mk01 | mactimes: the crucial point there is to have proper kernel to boot. ... |
| 04:11 | mactimes | mk01 Right. |
| 04:11 | mactimes | mk01 I have just followed that procedure to compile the latest kernel from that page. |
| 04:12 | mk01 | then the actual distro install is just few minutes on top. |
| 04:12 | mk01 | mactimes: ok ... |
| 04:12 | mk01 | mactime: and what was result ? |
| 04:14 | mactimes | mk01 I've decided to 'organize' the mess I had here with many trials. I'm just finishing compiling the modules again. |
| 04:15 | mactimes | mk01 The only think I had different from that is the ' export CROSS_COMPILE=/usr/bin/arm-linux-gnueabi-' which I used ' export CROSS_COMPILE=/usr/bin/arm-linux-gnueabihf-' instead. |
| 04:16 | mk01 | mactimes: I have not checked for a while but at what state are the custom installers scripts rabeeh was creating ? I have the feeling you was alrady refering to them few lines back (can't find it in history and don't remember the name) |
| 04:16 | mactimes | mk01 Yeah, I tried those scripts. |
| 04:17 | mk01 | and not working ? |
| 04:17 | mactimes | mk01 But the problem is that he 'personalized' so many things for his 'taste' that I ended up with a system with too many things to restore to the 'default' state. |
| 04:17 | mk01 | ah ok |
| 04:18 | mactimes | mk01 Since I'm doing this, not only for me, but also for a friend who is a bit 'paranoid' (who, in fact, bought 2 and gave me one), so, you see where it goes... |
| 04:20 | mk01 | mactimes: again , until you manage comfortably to build own kernel & vivante (gfx) support libs, you will need to rely on 'someones' kernel at least |
| 04:20 | mactimes | mk01 The latest one, I'm using to compile the new kernel is this http://www.solid-run.com/wiki/Building_the_kernel_and_u-boot_for_the_CuBox-i_and_the_HummingBoard with only that difference for the CROSS_COMPILE environment variable. |
| 04:21 | mk01 | mactimes: the difference you showed doesn't matter |
| 04:21 | mactimes | mk01 I see. |
| 04:22 | mk01 | maybe you can build your initial .img (with distro you want) with our build scripts we use to build XBian images |
| 04:23 | mk01 | it actually takes only XBian kernel & uboot (as one .deb) and all other from Distro you choose |
| 04:23 | mk01 | either Debian, or ubuntu or any other (.deb based). |
| 04:24 | mactimes | mk01 I have manually built the uboot too. It's already set up in the SD card. |
| 04:24 | mk01 | ok |
| 04:24 | mk01 | if you can boot your sdcard ... then just do debootstrapped two phase install |
| 04:25 | mk01 | that works as follows: |
| 04:25 | mk01 | on any linux with debootstrap tool available, you start it like this: |
| 04:26 | _unreal_ | mk01, I'm making a mirror of everything here http://download.solid-run.com/pub/solidrun/cubox/ |
| 04:27 | _unreal_ | I just dont know how to set it up |
| 04:27 | mk01 | debootstrap jessie --arch=armhf --foreign ./path-to-mounted-sd-partition-which-you-boot-on-cubox http://ftp.us.debian.org/debian |
| 04:29 | mactimes | mk01 Can I do that directly to the rootfs partition on the e-SATA which will be attached to the CBi? |
| 04:30 | mk01 | that will dl .debs needed for minimal root and unpack it to the ./path. then you boot the sd card in CBi and start /debootstrap/debootstrap --second-stage |
| 04:30 | mactimes | mk01 Hummmm. |
| 04:31 | mk01 | mactimes: yes, sure. until you have kernel on SD which is able to access eSata ... that's not technical problem of any kind |
| 04:31 | mactimes | mk01 BTW, do I *really* need to keep the '/boot' partition on the SD? Can't I just change the configuration from u-boot installed on the SD to boot from the e-SATA? |
| 04:31 | mk01 | mactimes: unfortunately you have to :-/ |
| 04:32 | mk01 | mactimes: as I told before, CBi doesn't have ANY storage itself, not even an eprom to hold uboot |
| 04:32 | mactimes | mk01 =( Ok. Hold on. I'll make the /boot partition the whole size of the SD then. I don't want to use it for anything else, it's too slow for regular use. |
| 04:33 | mk01 | mactimes: yes, I do it that way myself |
| 04:33 | mactimes | mk01 Even if the SPL and the u-boot.img have been set in the SD card as per that procedure? |
| 04:34 | mactimes | mk01 I'm sorry for the dumb questions, but this is really new to me. |
| 04:34 | mk01 | uboot+spl needs to be on the card. |
| 04:34 | mactimes | mk01 I'm trying to make sense from all of this. |
| 04:35 | mactimes | mk01 Yeah, that's what I mean. |
| 04:35 | mk01 | so that's clear |
| 04:35 | mk01 | now you need to load kernel & boot it, right ? |
| 04:36 | mactimes | mk01 The procedure tells me to do this: 'sudo dd if=SPL of=/dev/sdX bs=1K seek=1' and 'sudo dd if=u-boot.img of=/dev/sdX bs=1K seek=42'. Doesn't that already put SPL and uboot on the SD card? |
| 04:36 | mk01 | uboot can load kernel from SD card, from USB attached storage or from TFPT |
| 04:36 | mk01 | TFTP |
| 04:36 | mactimes | mk01 But not from the e-SATA device? |
| 04:36 | mk01 | (NOT SURE - DON'T KNOW) |
| 04:37 | mactimes | mk01 Ok. Hold on. Let me catch up here. |
| 04:37 | mk01 | ok |
| 04:37 | mk01 | no problem |
| 04:37 | mk01 | take your time |
| 04:37 | mk01 | imagine uboot+spl be a BIOS (as you know it) |
| 04:38 | _unreal_ | ouch not only do I have a cubox I have a cubox v1 |
| 04:45 | mk01 | _unreal_ http://archlinuxarm.org/platforms/armv7/marvell/cubox |
| 04:45 | mactimes | mk01 Hummm. I guess I'll be able to boot from e-SATA. Check this out: bmode - mmc0|normal|usb|sata|escpi1:0|escpi1:1|escpi1:2|escpi1:3|esdhc1|esdhc2|esdhc3|esdhc4 [noreset] |
| 04:45 | mactimes | mk01 That's from the console help of u-boot. |
| 04:45 | mactimes | mk01 Let me try something here... |
| 04:46 | mk01 | mactimes: could be that the raw storage (as needs to be used for USB works for sata too). ... |
| 04:48 | mactimes | mk01 Well, it can be. But I can have uEnv.txt configuration to load the boot portion with ext4load , just like it regularly does with mmc and usb devices. Or can't I? |
| 04:49 | mactimes | mk01 That way, *maybe*, I could have the /boot partition on the e-SATA device. Or not? |
| 04:50 | mk01 | mactimes: I don't think so ... but wait, will reboot with serial console |
| 04:50 | mactimes | mk01 Ok. |
| 04:55 | _unreal_ | ok following this http://archlinuxarm.org/platforms/armv7/marvell/cubox |
| 04:55 | _unreal_ | how would I extract http://archlinuxarm.org/os/ArchLinuxARM-cubox-latest.tar.gz |
| 04:56 | _unreal_ | root]$ tar ........ archlinux....tar.gz |
| 04:56 | mactimes | _unreal_ mkdir extracted_files |
| 04:57 | mactimes | _unreal_ tar -zxvf ArchLinuxARM-cubox-latest.tar.gz -C extracted_files |
| 04:58 | mk01 | mactimes: is the register (keeping info about boot device) value kept across cold reboots ? |
| 04:59 | mactimes | mk01 Well, I haven't tried yet, since I cannot identify the interface. But that would be as simples as 'setenv bmode sata' then 'saveenv', wouldn't it? |
| 05:00 | mk01 | mactimes: nope. bmode works differently. |
| 05:00 | mk01 | it writes the desired boot dev info into a register and reboots board |
| 05:00 | mactimes | mk01 You mean, via 'reset'? |
| 05:01 | mk01 | HummingBoard U-Boot > bmode mmc0 |
| 05:01 | mk01 | resetting ... |
| 05:01 | mk01 | U-Boot SPL 2013.10-rc4-ge4bc4c3-dirty (Dec 21 2014 - 05:46:01) |
| 05:01 | mk01 | Boot Device: SD1 |
| 05:01 | mk01 | spl: error reading image u-boot.img, err - -1 |
| 05:01 | mk01 | Load image from RAW... |
| 05:01 | mk01 | U-Boot 2013.10-rc4-ge4bc4c3-dirty (Dec 21 2014 - 05:46:01) |
| 05:01 | mk01 | CPU: Freescale i.MX6SOLO rev1.1 at 792 MHz |
| 05:01 | mk01 | Reset cause: WDOG |
| 05:01 | mk01 | Board: MX6-HummingBoard |
| 05:01 | mactimes | mk01 Is your box there a HummingBoard? |
| 05:02 | mk01 | this one you see yes |
| 05:02 | mk01 | HB1 |
| 05:02 | mactimes | mk01 Well, I try things manually here prior to putting all into a script. |
| 05:03 | mactimes | mk01 I usually stop autoboot, then do 'run autodetectfdt', and take it from there. |
| 05:04 | mactimes | mk01 I'm not sure if that's what you mean. =) |
| 05:06 | mk01 | no |
| 05:07 | mk01 | and just tested |
| 05:08 | mk01 | the boot dev in that register is kept only until you keep device running |
| 05:09 | mk01 | mactimes: also this only changes the location where SPL + uboot is loaded from |
| 05:09 | mactimes | mk01 Hummmm. |
| 05:09 | mactimes | mk01 Even if you run 'saveenv'? |
| 05:12 | mk01 | mactimes: env is stored as RAW onto the boot device |
| 05:12 | mactimes | mk01 The debootstrap target can be my rootfs on the e-SATA drive or *must* be the /boot partition on the SD card? |
| 05:13 | mk01 | so it's like board loads SPL, SPL loads env + uboot img |
| 05:13 | mactimes | mk01 Or can I just mount the e-sata under /mnt, then mount the SD into /mnt/boot and run debootstrap on /mnt? |
| 05:13 | mk01 | debootstrap rarget can be anything |
| 05:13 | mk01 | even local folder and you can COPY the files later |
| 05:13 | mk01 | (onto target storage) |
| 05:14 | mactimes | mk01 But copy where to? The rootfs or the boot partition? |
| 05:14 | mk01 | the folder you specify with debootstrap |
| 05:14 | mk01 | is considered "root" of future system |
| 05:14 | mk01 | that means if you specify /mnt |
| 05:15 | mk01 | it will create under /mnt actual mini rootfs ... so /mnt/etc, /mnt/var, /mnt/root .... |
| 05:15 | mk01 | so debootstrap target dir is FUTURE rootfs |
| 05:16 | mk01 | (I means from folder structure perspective ;) ) |
| 05:16 | mk01 | ((If I'm not making myself clear, feel free to tell me)) |
| 05:17 | mactimes | mk01 Perfectly clear. Your explanation has just clarified it. ;-) |
| 05:17 | mk01 | that means it is ROOTFS |
| 05:17 | mactimes | mk01 Hummm. Problem with debootstrap here. |
| 05:17 | mk01 | tell me |
| 05:18 | mactimes | mk01 debootstrap --arch=armhf --foreign jessie /mnt http://ftp.us.debian.org/debian failed with invalid signature. Then I tried debootstrap --arch=armhf --foreign jessie /mnt http://http.debian.net/debian which keeps failing retrieving packages. |
| 05:19 | mk01 | --no-check-gpg |
| 05:19 | mactimes | mk01 For the first? |
| 05:19 | mk01 | --no-check-certificate |
| 05:19 | mactimes | mk01 I'll give it a try. Hold on. |
| 05:22 | mk01 | the actual error means you don't have jessie gpg signatures in your dpkg key chain |
| 05:23 | mk01 | so --no-check-gpg should help you |
| 05:23 | mactimes | mk01 Keeps failing: W: Retrying failed download of http://ftp.us.debian.org/debian/dists/jessie/main/binary-armhf/Packages.gz |
| 05:24 | mk01 | then use your ftp.nl.debian.org (or what country are you in) |
| 05:26 | mactimes | mk01 Seems to work now. It's running yet. |
| 05:26 | mk01 | for cross check look here https://github.com/xbianonpi/xbian/blob/master/build.img#L147 and https://github.com/xbianonpi/xbian/blob/master/build.img#L156 |
| 05:27 | mk01 | mactimes: ok |
| 05:27 | mactimes | mk01 Nice! Retrieving and validating packages... |
| 05:28 | _unreal_ | mactimes, I'm getting operation not permitted, cant create sim links |
| 05:28 | _unreal_ | i'm using room |
| 05:29 | mactimes | _unreal_ Oh. |
| 05:29 | _unreal_ | root |
| 05:30 | _unreal_ | must be one of the switches or needing a switch |
| 05:32 | mactimes | mk01 Packages retrieved, validated and extracted. What now? |
| 05:32 | mactimes | _unreal_ Try this: touch test && ln -s test test_link |
| 05:32 | mactimes | _unreal_ What do you get? |
| 05:32 | mk01 | mactimes: now what you have is ROOTFS not finished - but possible to be started on target ARCH |
| 05:33 | mactimes | mk01 Remove the disk and the SD from this computer and connect them to the CBi? |
| 05:33 | _unreal_ | ? I touched my self |
| 05:33 | _unreal_ | ;) |
| 05:33 | mk01 | mactimes: that means with working kernel img you can start shell |
| 05:33 | mk01 | so |
| 05:33 | mk01 | as bootargs |
| 05:33 | mk01 | init=/bin/bash |
| 05:34 | _unreal_ | ? didnt do anyting what does that command do? |
| 05:34 | mk01 | mactimes: then you have to _finish_ the install with running the second stage |
| 05:35 | mk01 | https://github.com/xbianonpi/xbian/blob/master/build.img#L156 |
| 05:35 | mk01 | in your case |
| 05:35 | mk01 | it would be from the shell you get "/debootstrap/debootstrap --second-stage " |
| 05:36 | _unreal_ | sigh |
| 05:37 | mk01 | (this will actually properly install the unpacked packages ... so run "dpkg --configure" on them, build all the /var/lib stuff etc |
| 05:38 | mk01 | once it finishes you have jessie |
| 05:38 | mk01 | (not sure it will put any record into /etc/network/interfaces and /etc/apt/sources.list) so edit manually before continuing |
| 05:41 | mactimes | mk01 Hold on. I messed up a little bit here. Running debootstrap again. |
| 05:42 | mk01 | mactimes: if you install "qemu-arm-static" as part of debootstrap base install, you can run the second stage on your x86 machine too |
| 05:42 | mk01 | with simple chroot |
| 05:43 | mk01 | "chroot /your-target-dir debootstrap/debootstrap --second-stage" |
| 05:43 | _unreal_ | mactimes, I guess it is writing, but the error is that it cant change owner ship? |
| 05:43 | mactimes | mk01 I've done that before, but with Wheezy. Haven't made the bootstrap with jessie yet. Let's see how it works. I'll try to do it directly from the CBi. |
| 05:44 | mactimes | _unreal_ If you're currently logged in as root, you should have permission to do that. |
| 05:45 | mactimes | _unreal_ What are you trying to do now, exactly? |
| 05:46 | mactimes | mk01 Does it make any difference at all? |
| 05:46 | mk01 | mactimes: there is no change with jessie / wheezy / whatever ... you can even install ubuntu straight away on your wheezy. what changed probably (and is confusing you) is the two-step process now, but that's only because you are preparing rootfs for arch not compatible with your wheezy on pc you are currently using for install |
| 05:46 | _unreal_ | step 4 http://archlinuxarm.org/platforms/armv7/marvell/cubox |
| 05:46 | mactimes | mk01 I mean, using qemu-arm-static or doing it directly from the CBi? |
| 05:46 | mk01 | ach that |
| 05:47 | mk01 | no, no difference |
| 05:47 | _unreal_ | cubox should get an interface like the zipit |
| 05:47 | mk01 | the installed qemu only would allow to RUN the ROOTFS binaries straight away on your x86 box |
| 05:47 | mactimes | _unreal_ Have you already downloaded the files? |
| 05:47 | _unreal_ | its almost the same cpu. but it has a nice GUI boot loader custom made |
| 05:47 | _unreal_ | yes |
| 05:48 | mactimes | _unreal_ Have you 'cd' into the directory where the file is? |
| 05:48 | _unreal_ | yes |
| 05:48 | _unreal_ | and there is Content |
| 05:49 | mk01 | so you don't have to fiddle with try-test-see approach (when you are still learning about the new HW). so the finished target folder will be as any .img deployed (so simply ROOTFS in its proper form and shape) . |
| 05:49 | mk01 | (so just copy to sd/usb/sata and boot and done) |
| 05:52 | mactimes | mk01 Should 'debootstrap' have given be a kernel and ramdisk under /boot? |
| 05:52 | mactimes | s/be/me/g |
| 05:52 | mk01 | mactimes: no |
| 05:54 | mactimes | mk01 The 'zImage' I compiled from that guide should suffice then? |
| 05:54 | mk01 | mactimes: that question is not relevant to the rootfs you created itself. |
| 05:54 | mactimes | mk01 I mean, if I want to boot from the CBi to proceed with the second stage. |
| 05:55 | mk01 | mactimes: more it is relevant to the topic how you finally make your CBi to boot |
| 05:55 | mk01 | mactimes: ok you want to use SD for start ... and second-stage it on CBi ? |
| 05:56 | mactimes | mk01 I'd like to. I have done the second stage from the qemu thing. I'd like to learn a new approach, if you don't mind sparing the time. =) |
| 05:57 | mk01 | is you done second stage via qemu on your box |
| 05:57 | mk01 | then |
| 05:57 | mactimes | mk01 Nope. I've done that with Wheezy, a couple of days ago. |
| 05:57 | mk01 | ah ok |
| 05:57 | mk01 | so what, SD card for testing ? |
| 05:57 | mactimes | mk01 I have currently just done the 'first' portion. |
| 05:58 | mactimes | mk01 Yup! |
| 05:58 | mk01 | two partitions or just one as you told me at beginning ? |
| 05:58 | mactimes | mk01 SD card has single empty partition. |
| 05:58 | mk01 | ok |
| 05:58 | mk01 | SPL and uboot on it already ? |
| 05:58 | mactimes | mk01 bootstraping has been performed on the SATA disk. 1 root partition, a second one for SWAP |
| 05:59 | mactimes | mk01 YUP. |
| 05:59 | mactimes | mk01 bootstrapped into the root partition, of course. |
| 05:59 | mk01 | ok understand now your desired test setup |
| 05:59 | mactimes | |(*_*)| |
| 06:00 | mk01 | easiest is this : |
| 06:00 | mk01 | put zImage directly onto the SD card |
| 06:00 | mactimes | mk01 Hold on. Doing it. |
| 06:00 | mk01 | no folders are needed as DEFAULT uboot env is looking for "zImage" in 1st parition on SD in / or /boot |
| 06:02 | mactimes | mk01 Copied the compiled zImage and the imx6q-cubox-i.dtb to the SD partition |
| 06:02 | mk01 | ok |
| 06:02 | mk01 | because you want to boot from sata |
| 06:02 | mactimes | mk01 Should I connect the SD and the e-SATA to the CBi now? |
| 06:03 | mk01 | you would need to adapt imx6q-cubox-i.dts (before compile) or create uEnv.txt or boot script |
| 06:03 | mk01 | otherwise in .dtb file |
| 06:03 | mk01 | is hardcoded something like rootfs=/dev/mmcblk0p2 |
| 06:03 | mk01 | (if I rmember correctly) |
| 06:03 | mactimes | mk01 I prefer going the uEnv.txt rootfs=/dev/sda1 rootfstype=ext4 |
| 06:03 | mactimes | mk01 I think I've seen it while reading before. |
| 06:04 | mactimes | mk01 But no worries, I can do 'setenv bootargs' and set that up while testing before writing the uEnv.txt |
| 06:04 | mk01 | :)) ok |
| 06:04 | mactimes | mk01 When I get the correct parameters, then I write the uEnv.txt file. ;-) |
| 06:05 | mactimes | mk01 So, connect everything to the CBi now? |
| 06:05 | mk01 | you can write to ENV |
| 06:05 | mk01 | of uboot |
| 06:05 | mk01 | (also an option) |
| 06:05 | mactimes | mk01 That seems better than uEnv.txt |
| 06:05 | mk01 | one thing |
| 06:05 | mactimes | mk01 Change bootargs and saveenv to do that? |
| 06:05 | mk01 | remember to specify also the init= |
| 06:06 | mk01 | as we probably don't have /sbin/init yet and anyhow the rootfs is not yet finalised |
| 06:06 | mactimes | Hummm. Hold on. I'll connect everything and try to catch up. |
| 06:06 | mk01 | no problem |
| 06:07 | mactimes | mk01 At the console. |
| 06:08 | mactimes | CuBox-i U-Boot > run autodetectfdt |
| 06:08 | mk01 | nothing complicated. normally kernel at the end of kernel loading process calls the "init" binary (and uses path /sbin/ to locate it). specifyng init= as bootargs to kernel only changes that binary |
| 06:09 | mk01 | when you do manually |
| 06:09 | mactimes | mk01 init=/bin/bash? |
| 06:09 | mk01 | yep |
| 06:09 | mactimes | mk01 setenv bootargs 'console=ttymxc0,115200n8 rootfs=/dev/sda1 rootfstype=ext4 init=/bin/bash' |
| 06:10 | mactimes | mk01 Do I miss anything? |
| 06:10 | mk01 | no this should boot |
| 06:10 | mk01 | no. this should boot. |
| 06:10 | mactimes | CuBox-i U-Boot > boot ? |
| 06:10 | mk01 | toorfs |
| 06:10 | mk01 | ehm |
| 06:10 | mk01 | rootfs= is root= |
| 06:11 | mk01 | so 'console=ttymxc0,115200n8 root=/dev/sda1 rootfstype=ext4 init=/bin/bash' |
| 06:11 | mk01 | MAYBE |
| 06:11 | mactimes | mk01 fixed |
| 06:11 | mk01 | also |
| 06:11 | mk01 | rootwait |
| 06:11 | mk01 | and |
| 06:11 | mk01 | rw |
| 06:11 | mactimes | ok |
| 06:11 | mactimes | Any position? |
| 06:11 | mk01 | nope |
| 06:12 | mactimes | setenv bootargs 'console=ttymxc0,115200n8 root=/dev/sda1 rootwait rw rootfstype=ext4 init=/bin/bash' ? |
| 06:12 | mk01 | add |
| 06:12 | mk01 | console=tty1 |
| 06:12 | mk01 | and |
| 06:12 | mk01 | consoleblank=0 |
| 06:13 | mk01 | ach, if you really use serial |
| 06:13 | mk01 | then doesn;t matter |
| 06:13 | mactimes | mk01 So, the last one I wrote is enough? |
| 06:13 | mk01 | the other. but if yuo use monitor & keyboard add them |
| 06:13 | mk01 | YES |
| 06:14 | mk01 | although |
| 06:14 | mactimes | Hahahahah |
| 06:14 | mk01 | mactimes: we didn;t consider one thing |
| 06:14 | mactimes | Which is? |
| 06:14 | mactimes | display? |
| 06:15 | mk01 | I'm not sure if kernel after doing 'exec /bin/bash' keeps the ttymxc console |
| 06:15 | mk01 | you will tell me ;) |
| 06:16 | mactimes | mk01 I'll just do setenv bootargs 'console=ttymxc0,115200n8 root=/dev/sda1 rootwait rw rootfstype=ext4 init=/bin/bash video=mxcfb0:dev=hdmi,1920x1080M@60' |
| 06:16 | mk01 | HEH |
| 06:17 | mk01 | ok so |
| 06:17 | mk01 | what |
| 06:18 | mactimes | mk01 2.599339] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(179,2) |
| 06:18 | mk01 | is USB_STORAGE=y or =m ? |
| 06:18 | mk01 | in .config of kernel ? |
| 06:19 | mactimes | mk01 Not USB. e-SATA port |
| 06:19 | mk01 | ah |
| 06:19 | mk01 | sure |
| 06:19 | mk01 | sorry ;) |
| 06:20 | mk01 | ahci_imx.hotplug=1 |
| 06:21 | mk01 | as bootargs |
| 06:21 | mactimes | mk01 So weird! 2.572351] VFS: Cannot open root device "mmcblk0p2" or unknown-block(179,2): error -6 |
| 06:21 | mactimes | mk01 It's trying to use the SD partition 2 as root filesystem |
| 06:21 | mk01 | how you started the kernel ? |
| 06:21 | _unreal_ | is it mounted |
| 06:21 | _unreal_ | lsblk |
| 06:22 | mactimes | mk01 Just 'boot'. I'll try to load it and run bootm |
| 06:22 | mactimes | mk01 Hold on |
| 06:22 | mk01 | bootz |
| 06:22 | mk01 | wait I will check the default env |
| 06:25 | mactimes | CuBox-i U-Boot > setenv bootargs 'console=ttymxc0, root=/dev/sda1 rootwait rw rootfstype=ext4 init=/bin/bash video=mxcfb0:dev=hdmi,1920x1080M@60' |
| 06:25 | mactimes | CuBox-i U-Boot > ext4ls mmc 0:1 |
| 06:25 | mactimes | 4096 . |
| 06:25 | mactimes | 4096 .. |
| 06:25 | mactimes | 16384 lost+found |
| 06:25 | mactimes | 5993464 zImage |
| 06:25 | mactimes | 33568 imx6q-cubox-i.dtb |
| 06:25 | mactimes | CuBox-i U-Boot > ext4load mmc 0:1 10800000 /zImage |
| 06:25 | mactimes | 5993464 bytes read in 300 ms (19.1 MiB/s) |
| 06:25 | mactimes | CuBox-i U-Boot > bootz 10800000 |
| 06:25 | mactimes | Kernel image @ 0x10800000 [ 0x000000 - 0x5b73f8 ] |
| 06:25 | mactimes | Starting kernel ... |
| 06:25 | mactimes | mk01 And it hangs there... |
| 06:26 | mk01 | mactimes |
| 06:26 | mactimes | mk01 Yup? |
| 06:26 | mk01 | ok if you use boot |
| 06:26 | mk01 | it falls back to mmcboot |
| 06:26 | mk01 | which resets bootargs |
| 06:26 | mk01 | (if not in uEnv.txt) |
| 06:26 | mk01 | so |
| 06:26 | mactimes | mk01 But I've just used bootz now |
| 06:27 | mk01 | yes |
| 06:27 | mk01 | but you didn't load the .dtb |
| 06:27 | mk01 | look here |
| 06:27 | mactimes | Oh, yeah! |
| 06:27 | mk01 | on the script boot we use |
| 06:27 | mactimes | mk01 Your definitely right. But to provide a dtb, don't I have to provide a ramdisk also? |
| 06:28 | mk01 | (for reference ... https://github.com/xbianonpi/xbian-package-kernel/blob/master/build/imx6/content-tpl/boot/boot.scr.txt) |
| 06:28 | mk01 | no |
| 06:28 | mk01 | the syntax is |
| 06:28 | mactimes | mk01 Oh, yeah, I can use '-' to specify no ramdisk |
| 06:28 | mk01 | bootz ZIMAGEADDR INITRDADDR DTBADDR |
| 06:29 | mk01 | you can replace any with '-' |
| 06:29 | mk01 | yes |
| 06:30 | mactimes | mk01 Cross your fingers! =) |
| 06:30 | mactimes | CuBox-i U-Boot > setenv bootargs 'console=ttymxc0, root=/dev/sda1 rootwait rw rootfstype=ext4 init=/bin/bash video=mxcfb0:dev=hdmi,1920x1080M@60' |
| 06:30 | mactimes | CuBox-i U-Boot > ext4ls mmc 0:1 |
| 06:30 | mactimes | 4096 . |
| 06:30 | mactimes | 4096 .. |
| 06:30 | mactimes | 16384 lost+found |
| 06:30 | mactimes | 5993464 zImage |
| 06:30 | mactimes | 33568 imx6q-cubox-i.dtb |
| 06:30 | mactimes | CuBox-i U-Boot > run autodetectfdt |
| 06:30 | mactimes | CuBox-i U-Boot > ext4load mmc 0:1 10800000 /zImage |
| 06:30 | mactimes | 5993464 bytes read in 300 ms (19.1 MiB/s) |
| 06:30 | mactimes | CuBox-i U-Boot > ext4load mmc 0:1 40000000 /imx6q-cubox-i.dtb |
| 06:30 | mactimes | 33568 bytes read in 36 ms (910.2 KiB/s) |
| 06:30 | mactimes | CuBox-i U-Boot > bootz 10800000 - 40000000 |
| 06:30 | mactimes | mk01 Proceed? |
| 06:31 | mk01 | go for it |
| 06:31 | mk01 | and lets see |
| 06:31 | mk01 | what we forgot this time |
| 06:31 | mactimes | mk01 Weird characters. '???' |
| 06:31 | mk01 | wrong serial speed |
| 06:31 | mk01 | bps |
| 06:32 | mactimes | mk01 Although the HDMI 'fades' to black. |
| 06:32 | mk01 | you missed the 115200 (console=ttymxc0,) |
| 06:32 | mk01 | I don't know what speed is DEFAULT |
| 06:32 | mactimes | Damn |
| 06:32 | mactimes | 9600 is the default, but 115200 is the proper one. |
| 06:33 | mactimes | I'll do it again. |
| 06:33 | mk01 | no no |
| 06:33 | mk01 | wait |
| 06:33 | mk01 | change your speed to 9600 |
| 06:33 | mk01 | reset the line |
| 06:33 | mk01 | at least to see if all other is ok |
| 06:34 | mk01 | you know what ? |
| 06:34 | _unreal_ | sigh, I miss my old laptop it had a sd card port right in it. SO much faster reading/writing to cards this this stupid USB adapter |
| 06:35 | mk01 | adapt ENV and save |
| 06:35 | mk01 | that you don't type all the info over and over |
| 06:35 | mactimes | mk01 Just set the speed properly. |
| 06:36 | mactimes | mk01 It stopped on the console on '[ 2.503179] md: ... autorun DONE. |
| 06:36 | mactimes | [ 2.506596] Waiting for root device /dev/sda1...' |
| 06:36 | mactimes | mk01 MOOOOOOOVING ON! |
| 06:36 | mactimes | [ 45.639568] mxc_hdmi 20e0000.hdmi_video: mxc_hdmi_read_edid HDMI in HDMI mode |
| 06:36 | mactimes | [ 45.646771] mxc_hdmi 20e0000.hdmi_video: same edid |
| 06:36 | mactimes | [ 46.713601] mxc_hdmi 20e0000.hdmi_video: mxc_hdmi_read_edid HDMI in HDMI mode |
| 06:36 | mactimes | [ 46.720814] mxc_hdmi 20e0000.hdmi_video: same edid |
| 06:37 | mk01 | ok that's hotplug intr |
| 06:37 | mactimes | mk01 Hummmm.. Not really. |
| 06:37 | mk01 | ok tell me |
| 06:37 | mactimes | mk01 That's because I switched from D-SUB to HDMI, then back to D-SUB (I'm using the same monitor connected to both, my computer and the CBi). |
| 06:38 | mactimes | mk01 BUT.... It tells us that it has NOT hanged, right? |
| 06:38 | mk01 | ok and the serial ? |
| 06:38 | mactimes | [ 90.229156] imx-uart 2020000.serial: Rx FIFO overrun |
| 06:38 | mk01 | I mean |
| 06:39 | mk01 | what you have in the serial console connection |
| 06:39 | mactimes | It's still waiting for /dev/sda1... |
| 06:39 | mactimes | Ah. |
| 06:39 | mactimes | I have screen |
| 06:39 | mactimes | GNU/screen |
| 06:39 | mactimes | screen /dev/ttyUSB0 115200 |
| 06:40 | mactimes | What did you mean by 'adapt ENV and save' |
| 06:40 | mactimes | Missing '?" =) |
| 06:40 | mk01 | ok so it's like I thought ? after EXEC to init= the new process doesn't get the previous console on ttymxc0, right ? |
| 06:41 | mactimes | mk01 I think that it is, in fact, still waiting for /dev/sda1 |
| 06:41 | mk01 | ok |
| 06:41 | mactimes | mk01 Let me try changing the console to ttyS0 |
| 06:42 | mk01 | somehow in the process we forgot the 06:20 mk01: ahci_imx.hotplug=1 ? |
| 06:42 | mk01 | don't change the console to S0 |
| 06:42 | mactimes | Ok. |
| 06:42 | mk01 | as you really need ttymxc0 |
| 06:42 | mactimes | What should be the complete command line? |
| 06:43 | mk01 | only the other side needs to be redirected back to you ..... via getty for instance |
| 06:43 | mk01 | let's see that later |
| 06:43 | mk01 | first get the SDA1 |
| 06:44 | mk01 | and the save ENV means this - by default uboot starts $bootcmd |
| 06:44 | mk01 | string |
| 06:45 | mk01 | so you can do |
| 06:46 | mk01 | setenv bootcmd "setenv bootargs BLAH BLAH BLAH BLAH; load mmc 0:1 ADDR zImage; load mmc 0:1 ADDR2 .dtb; bootz ......;" |
| 06:46 | mk01 | saveenv |
| 06:46 | mactimes | mk01 I see. |
| 06:48 | mactimes | mk01 Hey! Shouldn't init be /debootstrap/debootstrap --second-stage by now? |
| 06:48 | _unreal_ | make the pain go away, its been writing to that SD card for an hour and still its only upto 94% |
| 06:51 | mk01 | mactimes: sure this should do the job |
| 06:51 | mk01 | but you wont see the stdout ... |
| 06:52 | mactimes | mk01 Hummm. |
| 06:52 | mk01 | you have copy of the debootstrap folder ? |
| 06:52 | mk01 | on your pc ? |
| 06:52 | mactimes | mk01 Nope. Make it directly to the sda1 |
| 06:52 | mactimes | mk01 Why? |
| 06:53 | mactimes | s/Make/Made/g |
| 06:53 | mk01 | I wanted to know if getty is already unpacked at /sbin |
| 06:56 | mactimes | mk01 Let me get the disk connected to the PC. Hold on. |
| 06:57 | mactimes | mk01 Connected and mounted. What do I do now? |
| 06:58 | mactimes | mk01 lrwxrwxrwx 1 root root 6 Dec 12 16:23 sbin/getty -> agetty |
| 07:05 | mactimes | mk01 Uhuuuullll! |
| 07:05 | mk01 | sorry was out |
| 07:05 | mactimes | mk01 I have no name!@(none):/# |
| 07:05 | mactimes | mk01 setenv bootargs 'console=ttymxc0,115200 root=/dev/sda1 rootwait rw rootfstype=ext4 init=/bin/bash video=mxcfb0:dev=hdmi,1024x768M@60' |
| 07:05 | mactimes | mk01 And that's the magic line! |
| 07:06 | mk01 | ok so you get the bash stdout/in redirected ? |
| 07:06 | mactimes | I have no name!@(none):/# /debootstrap/debootstrap --second-stage |
| 07:06 | mactimes | I: Installing core packages... |
| 07:06 | mk01 | ok |
| 07:06 | mactimes | mk01 I think it was the resolution of the screen |
| 07:06 | mk01 | fine then |
| 07:07 | mactimes | mk01 Dude, there are no words in the world which could express you my gratitude right now! |
| 07:08 | mactimes | mk01 So, here is a mere 'Thank you very, very much!' |
| 07:08 | mk01 | gladly ! |
| 07:09 | mactimes | mk01 It's installing! =) |
| 07:09 | mactimes | mk01 Just one more question... |
| 07:09 | mk01 | tell me |
| 07:10 | mactimes | mk01 Should I want the /home directory to be, in fact a /home partition encrypted, how should I have set that up? During the partitioning for the SATA disk, right? |
| 07:11 | mactimes | mk01 For cryptosetup/luksFS, how would I tell the system how to mount it? |
| 07:12 | mactimes | mk01 Or it should sort it out _automagically_ during the --second-stage thing? |
| 07:12 | mk01 | the systems (crypto , luks) have own init scripts |
| 07:12 | mk01 | they will add to the system |
| 07:13 | mactimes | mk01 All I have to do is create them before bootstraping on the sata disk? |
| 07:13 | mk01 | doesn't matter |
| 07:13 | mk01 | when you do it |
| 07:13 | mactimes | mk01 Could I do it after second-stage? |
| 07:13 | mk01 | even I would say, definitely not before |
| 07:13 | mk01 | yes |
| 07:14 | mactimes | mk01 I mean, could I create another partition on the disk and set the mount point for default options for that partition to /home and it would sort it out? |
| 07:14 | mactimes | s/for\ default/with\ default/g |
| 07:15 | mk01 | sure sure |
| 07:16 | mk01 | even the installer by default used during debian install process is doing most of that things AFTER all is installed |
| 07:16 | mk01 | before it is only gathering informations from you |
| 07:16 | mactimes | mk01 How do reboot the system now? |
| 07:16 | mk01 | 2nd stage finished ? |
| 07:17 | mactimes | mk01 reboot, shutdown, init not working. Already changed root password |
| 07:17 | mactimes | mk01 Yup. |
| 07:17 | mactimes | mk01 2nd stage finished. |
| 07:17 | mk01 | yes, there is no runlevel |
| 07:17 | mk01 | sync && reboot -fn |
| 07:17 | mactimes | mk01 Thanks! |
| 07:17 | mk01 | BUT |
| 07:18 | mk01 | now init will spawn VTs |
| 07:18 | mactimes | mk01 Yup. I still have to create the uEnv.txt file. =) |
| 07:18 | mactimes | mk01 And edit the inittab to respawn the ttymxc0 |
| 07:18 | mk01 | yes |
| 07:19 | mk01 | so reboot was quicker |
| 07:19 | mk01 | but as init spawns you will loose serial |
| 07:19 | mk01 | so you need to edit /etc/inittab before |
| 07:19 | mk01 | OR |
| 07:20 | mk01 | any other solution will work |
| 07:20 | mk01 | after |
| 07:20 | mk01 | but will be more costly |
| 07:20 | mk01 | like monitor + keyboard |
| 07:20 | mactimes | mk01 Sure |
| 07:21 | mk01 | with the encryption |
| 07:22 | mactimes | mk01 Dude, once again, thank you very much. |
| 07:22 | mk01 | I never used that on linux and macs have that directly at LVM level integrated, but what I remember from users cryptosetup installed a script into rcS |
| 07:22 | mk01 | to read config file and unlock all disks |
| 07:22 | mk01 | then loop them |
| 07:23 | mk01 | to /dev/X |
| 07:23 | mactimes | mk01 I'll try and rest a bit now. It's 4:22 AM here. |
| 07:23 | mk01 | so standard mountall (started at rc2) had "disks" like all other ... and mounted according to fstab |
| 07:23 | mactimes | mk01 I'll run the encryption tests later on. I'll let you know the results if you're around. |
| 07:23 | mk01 | it's 7:24 am HERE |
| 07:24 | mk01 | ::) |
| 07:24 | mactimes | mk01 I cannot thank you enough for all the help. |
| 07:24 | mactimes | mk01 So, once again, thank you! |
| 07:24 | mk01 | gladly |
| 07:24 | mk01 | nice day |
| 07:24 | mactimes | mk01 You too! |
| 14:49 | topi` | my Hummingboard has the BCM4330 wifi. Is it possible to make it go faster than 2.0MB/s? (mine is running in Master mode) |
| 14:49 | topi` | I believe the 802.11g can do 2.0MB/s |
| 14:50 | topi` | the wifi has a very small antenna, which seems to be glued into the Quad's heatsink. Maybe this antenna is sub-optimal? |
| 15:03 | sethstorm | topi`: probably, but it seems to be a best effort case |
| 15:06 | topi` | I want to figure out how to enable 802.11n for the bcm4330 |
| 15:06 | topi` | I'm using the brcmfmac kernel driver |
| 15:51 | topi` | hmm, it seems that gst-launch segfaults on given videos. I need to research what's so different with them |
| 15:51 | topi` | some videos (*.flv) seem to work just fine |
| 15:52 | topi` | and then I downloaded BigBuckBunny mp4 from blender's website, and the same playbin launch segfaults |
| 15:52 | topi` | hm, I'll try with a fakesink |
| 16:04 | sethstorm | topi`: encoding? |
| 16:07 | topi` | both streams use h.264 |
| 16:07 | topi` | and aac audio |
| 16:07 | topi` | I try to download several videos to see which ones segfault and which ones won't |
| 16:08 | topi` | mplayer plays everything just fine |
| 16:08 | topi` | (it's not very fast with the x11 video-output) |
| 16:14 | bencoh | (software upscaling ?) |
| 16:26 | mwcampbell | Can anyone here comment on the quality of the audio DAC on the HummingBoard i2eX (or Carrier Pro)? e.g. is there any click when the device is initialized, or any noticeable background noise? The RPi has both. |
| 16:27 | topi` | ok, it seems that the issue is with that single video (h.264, resolution 320x180), I downloaded a 720p version from blender's site and it works just fine on gstreamer |
| 16:28 | jnettlet | mwcampbell, I haven't tested in a while. There was definitely no background noise, but I can't remember if there was a click when starting and stopping the output. |
| 16:29 | jnettlet | I will try and make time to test later and record it for you to listen to |
| 16:29 | mwcampbell | jnettlet: thanks |
| 16:29 | jnettlet | we really haven't spent much time on it other to get it working because most everyone wants digital audio |
| 16:30 | mwcampbell | Presumably for home theater setups? |
| 16:32 | mwcampbell | For my current project, the intended audio outputs are either a car stereo or a stand-alone speaker (e.g. SoundCube) |
| 16:33 | mwcampbell | In that case, using a USB audio device seems a waste when the SoC has built-in audio support |
| 16:33 | mwcampbell | unless the analog output option on the Carrier Pro is itself just a USB device |
| 17:10 | topi` | I intend to run webkit on the HB, but it seems some sites just grab 100% cpu |
| 17:10 | topi` | when I try the same site under my i7 with virtualbox, no problems there whatsoever |
| 17:10 | topi` | I *think* this is because webkit has some uber-optimized js engine under x86_64, but nothing like that for armhf |
| 17:10 | topi` | any ideas or experiences? |
| 17:11 | topi` | the difference between 100% cpu and 0.1% cpu is just way too wide to be explained with intel being faster (it's not *THAT* much faster) |
| 17:11 | topi` | on most things, maybe around 10x faster |
| 17:45 | jnettlet | topi`, did you disable JIT for the javascript engine? |
| 17:45 | jnettlet | right now the jit engine is outputting junk on ARM |
| 20:24 | wbx | any kernel hackers online? |
| 20:24 | wbx | still having issues with DVI Mode disable interrupts and blank screen while watching a video via kodi |
| 21:00 | wbx | okay. consoleblank=0 solves it. |
| 21:03 | jnettlet | wbx, yes. |
| 21:03 | jnettlet | if you blank the screen in DVI mode interrupts have to be disabled because of a hardware bug. |
| 21:03 | jnettlet | but they will be re-enabled when the screen is unblanked |
| 21:04 | jnettlet | sounds like Kodi is not unblanking the screen properly |
| 21:21 | wbx | jnettlet: thanks for the explanation. consoleblank=0 is fine for me. |
| 21:21 | wbx | jnettlet: do you know if the vivante libraries are available in source? they are linked against glibc... |
| 21:40 | jnettlet | wbx, no the vivante libraries are only provided as binaries from freescale. |
| 21:40 | jnettlet | it is fine for binary libraries to link against GPL libs |
| 21:44 | wbx | LGPL libs. sure. but then you can't use another C library |
| 21:46 | wbx | are there any plans from freescale to provide the source? |
| 21:59 | jnettlet | they don't need to |
| 22:18 | Dom__ | can someone help me in using a kernel patch? |
| 22:25 | jnettlet | depends on the kernel and the patch |
| 22:25 | jnettlet | what are you trying to do? |
| 23:04 | Dom__ | jnettlet: i already handled it. thanks |
| 23:07 | Dom__ | some other question: how can i read the register 0x300a of i2c with the i2cget tool? it says invalid data-address... |
| 23:12 | mk01_ | Dom__: i2c tools are limited i2c conroller chip addresses |
| 23:12 | mk01_ | so no device addresses |
| 23:13 | Dom__ | hmm, is there an alternative to read a register value? |
| 23:14 | mk01_ | Dom__: i2c_master_send and i2c_master_receive |
| 23:15 | mk01_ | actually C code |
| 23:15 | mk01_ | don't know of a tool |
| 23:16 | mk01_ | do you have ov5647 cam ? |
| 23:16 | Dom__ | ok danks |
| 23:16 | Dom__ | thanks |
| 23:16 | Dom__ | yes |
| 23:16 | mk01_ | ok :) |
| 23:16 | mk01_ | 300a was the hint |
| 23:16 | Dom__ | ;) |
| 23:17 | mk01_ | how far are you with driver ? |
| 23:18 | Dom__ | i got a driver from cbxbiker. but doesn't work, or i'm stupid ^^ |
| 23:19 | Dom__ | thats why i want to check if the cam is installed correctly |
| 23:19 | mk01_ | and the driver is all OK working ? I mean offitially ? |
| 23:20 | Dom__ | no, we are testing it |
| 23:21 | mk01_ | nice |
| 23:22 | mk01_ | I was hacking mine own back in october and got all mipi communications with camera ok |
| 23:23 | mk01_ | but could not get actual DATA from ipu |
| 23:24 | mk01_ | and there was nobody to discuss further so I never returned to it |
| 23:24 | Dom__ | :) |
| 23:25 | Dom__ | maybe we can open a discuss group for us |
| 23:25 | Dom__ | it is better than irc, because of the different timezones |
| 23:25 | mk01_ | wait i will send you link for DL |
| 23:25 | mk01_ | you can cross check |
| 23:26 | mk01_ | I took old mipi driver (for v4l2-int) and rewrote for ov5647 ... so the control code can be of use for you |
| 23:27 | mk01_ | as v4l2-int was replaced with v4subsys |
| 23:27 | mk01_ | the ioctl part is not so relevant ... |
| 23:28 | Dom__ | ok |
| 23:29 | Dom__ | do you want to send me an email with link? |
| 23:29 | mk01_ | http://download.serveftp.net/other/ |
| 23:29 | Dom__ | ok |