| 23:59 | xraxor_ | what can i use to test my i4pro speed with geexbox? as in sd card, or even gpu |
| 23:59 | xraxor_ | is this possible? |
| 00:00 | cbxbiker61 | sd card speed? |
| 00:01 | cbxbiker61 | geexbox is pretty limited with respect to what is/can be installed, so it's not particularly good for running speed tests |
| 00:02 | _dab_ | jnettlet: do you have a test kernel yet? |
| 01:06 | dshankar | rabeeh: having an issue with the VPU on imx6 - after a few minutes of playing a video, the screen starts flickering and eventually stops displaying entirely. this occurs even with 480p files with <25% CPU usage, leading me to think it's no CPU-related or temperature-related, but something with the VPU or VPU memory |
| 01:07 | dshankar | have you encountered this before? MikeSeth said you might have |
| 01:07 | MikeSeth | it's 2 am in Israel, he's prolly asleep |
| 01:07 | MikeSeth | and you should too |
| 01:07 | MikeSeth | do bear in mind I am not entirely sure this was the case too |
| 01:07 | dshankar | ah! alright :-) |
| 01:08 | MikeSeth | I also remember someone bringing up buffer corruption here recently |
| 01:08 | MikeSeth | hmrhmrh |
| 01:08 | MikeSeth | sec I'll find it in the logs |
| 01:08 | MikeSeth | uhh |
| 01:08 | MikeSeth | the hack was |
| 01:09 | MikeSeth | 12:13 devmem 0x020e0018 32 0xffffffff |
| 01:09 | MikeSeth | 12:13 devmem2 0x020e0018 w 0xffffffff |
| 01:09 | MikeSeth | and it had to do with memory qos, the problem was characterized by IPU interrupt warning in dmesg |
| 01:09 | MikeSeth | that's all I got, no actual log :( |
| 01:13 | dshankar | yikes alright |
| 07:57 | jhonny | hi everyone, just got my Cubox-i :) really excited, hoping to install XBMC on it, any good guides out yet ? |
| 07:59 | tomlohave | http://cubox-i.com/getting-started/ |
| 08:03 | jhonny | Cool bananas ;) thanks very much |
| 08:03 | jhonny | Any particular build you'd recommend for running XBMC ? |
| 08:05 | durandle | I've been using GeexBox for a while and its pretty good. Dead easy to setup. |
| 08:06 | durandle | and thats on the old first gen cubox, I imagine it'll be a lot better on the new ones :) |
| 08:08 | jhonny | I'm hoping the performance is really good :) might install android on it later too and check out some emulators ;) it would be pretty awesome to run some N64 emu's on it ;) |
| 08:09 | jhonny | I think mine also has blue teeth .... might be able to get a PS3 controller working with it ? :) |
| 08:28 | durandle | Just a shame you can't use OpenELEC yet, I rather like it as a lightweight linux |
| 08:50 | jhonny | so i just tried dd-ing the following image to a 4gb class4 sdcard |
| 08:50 | jhonny | geexbox-devel-20131212-r16611.cuboxi-patched-cubox-i2ultra |
| 08:51 | jhonny | no dice, black screen, although my tv did switch to 1080p 60 hz mode |
| 08:51 | jhonny | the infra-red receiver on the front did light up red and so did the optical audio port on the side, so i don't know if the device is dead or not |
| 08:51 | jhonny | I'm going to try an android image now, see if that works |
| 09:08 | tomlohave | jhonny: try a more recent version. |
| 09:09 | jhonny | ok tomlohave will do :) |
| 09:09 | jhonny | i just tried android and it booted up, can't get a mac keyboard to work with it but at least it booted and i know the box is alive ;) |
| 09:10 | tomlohave | http://download.geexbox.org/snapshots/geexbox-xbmc-imx6-cuboxi/ |
| 09:10 | jhonny | I've got this image: geexbox-devel-20140114-r16709.cuboxi.img.xz |
| 09:10 | jhonny | only thing is its not specific to the i2 version ? |
| 09:10 | tomlohave | no |
| 09:10 | jhonny | will that mean the xbmc will work with only 1 core ? |
| 09:10 | tomlohave | it should work for all cuboxi |
| 09:10 | jhonny | ahh ok |
| 09:20 | jnettlet | jhonny, the Mac keyboards need a special kernel module that is probably not enabled in the kernel |
| 09:20 | jhonny | that's interesting, that android build doesn't have Playstore on it |
| 09:21 | jhonny | also another interesting this is i have the keyboard connected (which doesn't work) but the keyboard itself has a usb port so you can connect a mouse to it, so i connected my mouse and that works ;) lol |
| 09:25 | jnettlet | jhonny, we are still waiting for Google to certify the hardware and the playstore won't be available until that happens |
| 09:27 | jhonny | ahhh right, gotcha, didn't know about that |
| 09:58 | jnettlet | jhonny, if you want to play around I use this repository that distributes FOSS apps for android. https://f-droid.org |
| 09:58 | jhonny | nice, thank you ;) |
| 09:59 | jhonny | i just installed n64 emu from apkdad |
| 09:59 | jhonny | it runs Super Mario 64 at full speed ;) |
| 09:59 | jhonny | now ill just have to connect a Dual Shock controller and see if i can get a bluetooth connection :) that would be so sweet for retro gaming |
| 09:59 | jhonny | but now, ill try getting xbmc running again |
| 09:59 | jhonny | only problem is the files i downloaded are xz files, no idea how to open those up |
| 10:00 | jnettlet | what platform do you want to open them on? |
| 10:01 | jnettlet | you can use xz under linux or 7-zip under windows |
| 10:02 | jnettle | 10:02 * jnettlet has no idea about MacOS |
| 10:07 | jhonny | yep just downloaded xz and installed it on MacOSX |
| 10:54 | jhonny | thanks everyone for your help thus far |
| 10:55 | jhonny | I'm going to go play around with xbmc a bit |
| 10:55 | jhonny | thank you once again |
| 15:12 | jnettlet | MarcusVinter, when you get a chance can you send me the output of the mmcinfo command from uboot on your i2u? |
| 15:18 | MarcusVinter | Sure. Ill do it very shortly. |
| 15:21 | jnettlet | thanks |
| 15:29 | MarcusVinter | jnettlet > http://pastebin.com/52bgftCx |
| 15:30 | jnettlet | MarcusVinter, thanks a lot |
| 15:32 | MarcusVinter | No probs! |
| 15:34 | rabeeh | MarcusVinter: is this an UHS1 device? |
| 15:38 | MarcusVinter | Yeah. |
| 15:38 | MarcusVinter | It was really cheap, 16GB for less than �10. |
| 15:39 | MarcusVinter | Obviously my company paid for it anyway :P |
| 15:39 | jnettlet | and one more bit of info. can you type version at the u-boot prompt and post the info |
| 15:39 | MarcusVinter | Sure let me reboot. |
| 15:40 | MarcusVinter | http://pastebin.com/3btUm8RB |
| 15:40 | jnettlet | thanks. I will stop harassing you now |
| 15:41 | MarcusVinter | Haha its not a problem. I'm sure I haven't harassed you for the last time yet unfortunately. |
| 15:44 | xraxor_ | guys what can we do with this? http://sources.openelec.tv/tmp/image/misc/ |
| 15:44 | xraxor_ | was posted on OE IRC channel |
| 15:47 | jnettle | 15:47 * jnettlet stares at it blankly. |
| 15:48 | jnettlet | how is this different than geexbox or OE yocto? |
| 15:52 | xraxor_ | dnt know |
| 15:53 | xraxor_ | i just know Geexbox is playing that well with me, CEC keeps on breaking |
| 15:53 | xraxor_ | i didnt try the yocto image |
| 15:54 | _rmk_ | which kernel version are you using? |
| 15:54 | jnettlet | it depends if it is breaking in userspace or kernel |
| 15:54 | _rmk_ | jnettlet: its probably breaking in the kernel... from what I remember of the freescale driver |
| 15:55 | jnettle | 15:55 * jnettlet agrees |
| 15:55 | xraxor_ | im using geexbox latest snapshot, and same problem with all other GBox image |
| 15:56 | jnettlet | in which case a new image probably won't help unless the distro is holding back patches. Which I wouldn't discount |
| 15:56 | xraxor_ | i thought was geexbox problem |
| 15:57 | xraxor_ | it oly breaks when i turn off tv for a certain time, or am on the TV input when xbmc starts |
| 15:57 | xraxor_ | *only breaks |
| 15:57 | _rmk_ | jnettlet: I never committed my CEC driver because I couldn't test it on hummingboard |
| 15:58 | jnettlet | _rmk_, well now you have your shiny i4pro that should support it |
| 15:59 | _rmk_ | yea, except I don't want to wipe this SD card and I don't have another one available for it |
| 15:59 | jnettlet | _rmk_, just dd that sdcard to a backup image |
| 16:00 | _rmk_ | needs to be 8GB for my filesystem though |
| 16:01 | jnettlet | actually if we get you upgraded to the SPL uboot you can use the same SDHC card from both the HB and i4Pro |
| 16:01 | jnettlet | and boot them both from NFS if you would like |
| 16:02 | rabeeh | _rmk_: sudo dd if=SPL of=/dev/sdc bs=1K seek=1; sudo dd if=u-boot.img of=/dev/sdc bs=1K seek=42 |
| 16:02 | rabeeh | _rmk_: it autodetect if it's HB / CuBox-i and accordingly configures memory and machine ID |
| 16:02 | rabeeh | (although you don't need machine ID) |
| 16:03 | rabeeh | _rmk_: want ready to use images? |
| 16:16 | cardiel | xraxor_: did you try the openelec image? |
| 16:17 | _rmk_ | jnettlet: that'll probably work... sorry, was distracted by a phone call from the national health service made by a foreigner who couldn't understand basic English responses. |
| 16:17 | xraxor_ | no, im at work, and have no idea how to get around to installing it |
| 16:18 | jnettlet | _rmk_, are you booting with dtb's appended to the zImage or separate files? |
| 16:18 | _rmk_ | that's something I need to fix - I'd prefer the bootz method :) |
| 16:19 | jnettlet | my default loader detects if you pass it a zImage and automagically does a bootz for you. |
| 16:20 | jnettlet | the only difference you would need is to have your script detect the machine ID and change the dtb accordingly |
| 16:20 | _rmk_ | (when it takes five iterations to get the remote end to understand that you're about to pass the phone to the person who they want to speak to, you have to worry about patient confidentiality... and hope that they're not directly involved with patient care) |
| 16:21 | jnettle | 16:21 * jnettlet gets the added bonus of going through this in another language. |
| 16:21 | nikitis | Anyone know how to get a tracking number for your shipment? I've seen people on the forums with tracking numbers. |
| 16:24 | rabeeh | nikitis: you can contact [email protected] for that |
| 16:25 | _rmk_ | jnettlet: how are the ether macs handled on the cubox-i? is that still via an environment variable? |
| 16:26 | jnettlet | _rmk_, nope they are read from the fuses. |
| 16:26 | rabeeh | _rmk_: fuses |
| 16:26 | jnettlet | well with the HB1 they were read from the fuses, but the fuses weren't populate yet :-) |
| 16:26 | rabeeh | but if you do 'saveenv' and then move the micro sd to another box it will write some warning that the fuses and env mac addresses differ |
| 16:27 | _rmk_ | that gives me a way to select the appropriate dtb file |
| 16:27 | rabeeh | jnettlet: on HB1 those are not; but the production versions all have MAC addresses |
| 16:27 | rabeeh | (you can see them on the bottom of the box) |
| 16:27 | jnettlet | rabeeh, the production HB's right? |
| 16:27 | rabeeh | sorry; the production CuBox-i |
| 16:27 | rabeeh | HB is still not in production; it has all fixed MAC address |
| 16:27 | jnettlet | I assumed it was just the developer models that lacked the MAC |
| 16:28 | rabeeh | d0:63:00:00:00:00 |
| 16:28 | jnettle | 16:28 * jnettlet wonders if we should add an ARP to u-boot to see if there is MAC collision |
| 16:30 | nikitis | rabeeh: okay, thanks, I did so just now. last email I had said was expected to ship to me by Jan 15th. I didn't know if that meant it will leave them on the 15th or get to me by the 15th. |
| 16:30 | jnettle | 16:30 * jnettlet going to walk the dogs |
| 16:30 | MarcusVinter | Thats a good idea jnettlet |
| 16:31 | rabeeh | jnettlet: don't worry about that |
| 16:31 | rabeeh | i think no one got more than one HB sample; so they won't have any MAC collision :) |
| 17:06 | _rmk_ | well, the u-boot-imx6.git repository doesn't build - lots of undefined references at link time |
| 17:11 | xraxor_ | a new package for the vivante graphics was also release recently, gpu-viv-bin-mx6q ? |
| 17:22 | _rmk | 17:22 * _rmk_ assumes that http://imx.solid-run.com/wiki/index.php?title=Building_the_kernel_and_u-boot_for_the_CuBox-i_and_the_HummingBoard is outdated. |
| 17:23 | jhonny_ | hi everyone, is there a way to root android 4.3 on cubox-i2ultra ? |
| 17:28 | _5051 | jhonny_, not that I know of, but I have a few idea that I want to try when I receive the cubox |
| 17:28 | _5051 | basic thought process mentioned here: http://imx.solid-run.com/forums/viewtopic.php?f=2&t=227&start=10#p1972 |
| 17:29 | _505 | (Adium sucks when it comes to automatic nick changes if it signs you out) |
| 17:32 | jhonny_ | nice, ok, ill give that a quick go |
| 17:34 | _505 | I examined the install script, and they only thing you can't do is "/system/xbin/su --install" |
| 17:34 | _505 | but no idea what that actually does |
| 17:37 | rabeeh | _rmk_: it's updated not outdated |
| 17:37 | rabeeh | the build part |
| 17:37 | _rmk_ | well, uboot fails to build for me |
| 17:38 | _rmk_ | arch/arm/lib/libarm.o: In function `board_init_r': |
| 17:38 | _rmk_ | /home/rmk/git/u-boot-imx6/arch/arm/lib/board.c:593: undefined reference to `mmc_initialize' |
| 17:38 | _rmk_ | /home/rmk/git/u-boot-imx6/arch/arm/lib/board.c:655: undefined reference to `eth_initialize' |
| 17:38 | _rmk_ | arch/arm/lib/libarm.o:(.data.init_sequence+0x14): undefined reference to `env_init' |
| 17:38 | _rmk_ | arch/arm/cpu/armv7/start.o: In function `cpu_init_crit': |
| 17:38 | _rmk_ | (.text+0x10c): undefined reference to `lowlevel_init' |
| 17:38 | _rmk_ | ... and lots more ... |
| 17:38 | _rmk_ | I followed the instructions on that page |
| 17:39 | _rmk_ | $ make mx6_cubox-i_config |
| 17:39 | _rmk_ | Configuring for mx6_cubox-i - Board: mx6_cubox-i, Options: IMX_CONFIG=board/solidrun/mx6_cubox-i/imx6image.cfg,MX6QDL,SPL,FSL_ENV_IN_MMC |
| 17:40 | _rmk_ | then I just run make after that |
| 17:41 | _rmk_ | weird, that time it worked |
| 17:42 | rabeeh | which toolchain? |
| 17:42 | _rmk_ | same one that I've used for years for all my arm kernel builds |
| 17:42 | rabeeh | there were some issues with hardfp; but that's on division arthimetics |
| 17:42 | _rmk_ | and that I've used for other uboots |
| 17:42 | rabeeh | gcc 2.95? |
| 17:42 | _rmk_ | no, it's 4.3.something |
| 17:43 | _rmk_ | redoing the build a third time worked |
| 17:53 | iio7 | Yesterday my cubox pro booted fine from a USB harddrive. Today when I switched it on, with the same drive attached, it wont boot and it gives these errors: http://paste.debian.net/76431/ |
| 17:54 | MarcusVinter | you probs pulled out before she was done. |
| 17:54 | MarcusVinter | lol |
| 17:54 | MarcusVinter | shut it down while writing or something |
| 17:54 | MarcusVinter | I have no idea though, its just happened to me that way before. |
| 17:54 | MarcusVinter | you run a checkdisk on it? |
| 17:55 | iio7 | What are you talking about? Probs pulled out? |
| 17:55 | iio7 | The disk is fine. |
| 17:57 | MarcusVinter | It was a poor joke. I'm merely saying that the drive might have some disk errors if it wasnt shut down correctly last time. |
| 17:57 | iio7 | okay.. no the disk is fine I have checked it. |
| 17:58 | MarcusVinter | did you use fsck? |
| 17:58 | MarcusVinter | Its strange if it works working before though.. |
| 17:59 | MarcusVinter | was working* |
| 18:00 | iio7 | I think that my boot.scr maybe is wrong. I made it from this boot.txt: http://paste.debian.net/76436/ |
| 18:00 | jnettlet | _rmk_, "redoing the build a third time worked" always makes you feel good about things |
| 18:01 | _rmk_ | jnettlet: is there any dt file(s) for the cubox-i yet? |
| 18:02 | rabeeh | _rmk_: just tried; build every time |
| 18:04 | jnettlet | _rmk_, I have started one, but nothing proper. Basically it is an HB/cubox-i that includes the mx6q.dtsi instead of the mx6dl.dtsi |
| 18:05 | jnettlet | I have also tested adding uhs support, but the testing for that has been derailed a bit |
| 18:05 | _505 | iio7, can you make a serial connection and execute the boot commands manually? |
| 18:05 | _505 | that is easier for debugging |
| 18:05 | _rmk_ | jnettlet: there's nothing different about the MMC interface? |
| 18:05 | iio7 | _505, I'm doing that now. |
| 18:06 | jnettlet | _rmk_, not for basic support. for uhs you need to setup pin configs for 100mhz and 200mhz and add the 1.8 vselect |
| 18:06 | iio7 | _505, when I issue the "ext4load usb 0:1 0x02000000 /boot/uImage" part I get "Bad partition 1" |
| 18:07 | _505 | disk corrupt? or the partition table? |
| 18:08 | jnettlet | I think we can add that for the low end versions as well. The cubox-i devices all support the faster speeds, and theoretically the 1.8v switch should fail for the HB1 and make it work properly just clocking up to 50Mhz |
| 18:08 | iio7 | _505, no that's the strange part. I suspect that Cubox Pro isn't working well with this external USB disk, but I'm not sure. |
| 18:10 | _505 | which USB port are you using? Might be the case that only the top one works for booting |
| 18:11 | iio7 | Okay.. wait a second.. I just issued "usb start", then ran the command, now it works. |
| 18:11 | iio7 | Does "usb start" has to go into the boot.scr? |
| 18:11 | _rmk_ | jnettlet: when I get to hack the dt files... which is going to be very shortly... |
| 18:11 | jnettlet | _rmk_, I am using this right now http://fpaste.org/68668/98058491/ although that is configured for the framebuffer and freescale hdmi drivers. |
| 18:12 | jnettlet | although if you don't have a uhs1 card the usdhc2 changes shouldn't matter |
| 18:12 | jnettlet | everything else boots and works as intended with all 4 cores brought up. |
| 18:15 | _rmk_ | hmm. |
| 18:15 | _rmk_ | HummingBoard U-Boot > setenv bootcmdold ${bootcmd} |
| 18:15 | _rmk_ | setenv - set environment variables |
| 18:15 | _rmk_ | Usage: |
| 18:15 | _rmk_ | setenv [-f] name value ... |
| 18:15 | _rmk_ | ... |
| 18:17 | _rmk_ | HummingBoard U-Boot > echo "${bootcmd}" |
| 18:17 | _rmk_ | echo - echo args to console |
| 18:18 | _rmk_ | I think something's wrong with this uboot |
| 18:18 | jnettlet | well I think the three times you had to compile it for it to work might be a tip off :-) |
| 18:18 | Coolgeek | or ${bootcmd} is empty |
| 18:19 | _rmk_ | bootcmd=mmc dev ${mmcdev}; if mmc rescan; then if run ext2loadbootscript; then run bootscript; elif run fatloadbootscript; then run bootscript; else if run ext2loadbootenv || run fatloadbootenv; then run importbootenv; fi; if run ext2loadbootfile; then run ext2mmcboot; elif run fatloadbootfile; then run fatmmcboot; else run netboot; fi; fi; else run netboot; fi |
| 18:19 | _rmk_ | not empty. |
| 18:20 | _rmk_ | jnettlet: yea, maybe it needs to be rebuilt some more times :p |
| 18:20 | Coolgeek | try a clean build env |
| 18:24 | _rmk_ | well, a 4th build produces no apparant difference |
| 18:25 | iio7 | _505, when I issue these commands manually I get the device to boot http://paste.debian.net/76439/ |
| 18:26 | iio7 | _505, but when I make a boot.txt and insert the commands and then make a boot.scr, which I place in /boot/boot.scr then I get the "Bad partition" error and the device wont boot. |
| 18:26 | iio7 | What am I missing? |
| 18:27 | _505 | timing? not sure. Might take a few seconds for usb start to detect the device |
| 18:28 | _rmk_ | ok. setenv bootcmdold foobarbaz works |
| 18:28 | _rmk_ | but trying to use an environment variable on the command line results in the help being spat out for whatever the command was |
| 18:28 | iio7 | _505, is there some delay command? |
| 18:29 | Coolgeek | there is a command that wait for USB o be ready I think |
| 18:30 | iio7 | rootdelay, found it, gonna try that |
| 18:30 | _rmk_ | oops, I just crashed uboot's command interpreter |
| 18:33 | iio7 | Same problem. |
| 18:35 | iio7 | Ohh. I found something.. When the device boots it says "scanning bus for devices... 1 USB Device(s) found", but when I manually issue the "usb start" command it says "scanning bus for devices... 2 USB Device(s) found" |
| 18:36 | iio7 | So it doesn't even read the boot.scr file because it can't see the USB harddrive, it only sees the USB controller. |
| 18:37 | iio7 | But when I issue the usb start it sees both. |
| 18:37 | iio7 | How can I fix that? |
| 18:38 | Coolgeek | USB flash ? usb disk ? power internal via USB or external ? |
| 18:39 | iio7 | Its a USB external drive, power via USB. |
| 18:39 | Coolgeek | ok |
| 18:39 | iio7 | A small 2.5" seagate drive. |
| 18:39 | Coolgeek | I think he won't become ready fast enough |
| 18:39 | iio7 | Yes. |
| 18:40 | iio7 | So rootdelay isn't useful. |
| 18:41 | iio7 | Is that fixable? Mind you I'm a newbie cubox dude. |
| 18:41 | Coolgeek | http://plugcomputer.org/plugforum/index.php?topic=131.0;wap2 check here, the third |
| 18:43 | jnettlet | iio7, does it change if you swap the port that the usb drive is plugged into? |
| 18:43 | _rmk_ | jnettlet: and... this uboot stops booting my 3.13-rc7 kernel on the hb1 |
| 18:44 | iio7 | jnettlet, I have tried that as I understood cubox pro could only boot from the top usb plug. |
| 18:44 | iio7 | I have -> haven't |
| 18:44 | jnettlet | _rmk_, ??? are you using a custom uboot.scr, or uEnv.txt? |
| 18:44 | _rmk_ | hang on, it may be me messing up the dhcp server |
| 18:45 | _rmk_ | yea, it was. :) |
| 18:45 | iio7 | jnettle, no it's the same, it only detects 1 USB device. |
| 18:45 | jnettlet | phew, okay |
| 18:46 | jnettlet | iio7, does issuing usb reset at the uboot prompt list it? |
| 18:47 | iio7 | How do I get a prompt? Once it has failed loading its trying to boot from TFTP and is standing at "Loading: T T T T" |
| 18:47 | jnettlet | hopefully you can set your bootcmd to something like "usb start ; usb reset ; ext4load blah blah blah" |
| 18:48 | jnettlet | iio7, interrupt the boot process by hitting esc when uboot first loads |
| 18:48 | iio7 | jnettle, yes I think that is the solution, I just needs to figure out how that is done :) |
| 18:48 | jnettlet | you can also try cmd + c during the TFTP loading, but depending on how the uboot loading script is written that may give you a shell or nothing |
| 18:49 | iio7 | ctrl + c got me the shell |
| 18:49 | jnettlet | okay try usb-reset |
| 18:49 | jnettlet | sorry usb reset |
| 18:49 | iio7 | Still only 1 USB device(s) found |
| 18:50 | jnettlet | does it attempt to spin up your drive? can you hear it doing anything |
| 18:51 | iio7 | Oh, wait, sorry.. had to put the disk back in the upper usb. |
| 18:51 | iio7 | Now, usb reset works! |
| 18:52 | iio7 | Could someone give me a pointer as to how I insert a "usb reset" into uboot? |
| 18:53 | jnettlet | sure, do printenv bootcmd |
| 18:53 | iio7 | I guess this has something to do with SPI flash? |
| 18:53 | jnettlet | then paste what is prints out |
| 18:54 | iio7 | bootcmd=for device_name in ${trydev} ; do for partition in ${trypart} ; do for directory in ${trydir} ;do for fstype in ${tryfs}; do; |
| 18:56 | jnettlet | and you have your uboot.scr on the usb drive? |
| 18:57 | iio7 | in /boot/ on the usb drive I have boot.scr and uImage |
| 18:57 | jnettlet | regardless so you will just change things to set bootcmd 'usb reset; for device_name in ${trydev} ; do for partition in ${trypart} ; do for directory in ${trydir} ;do for fstype in ${tryfs}; do; |
| 18:58 | iio7 | Is that stored in flash? |
| 18:58 | jnettlet | then type boot and see if it works. If everything is okay, do the same thing again but instead of boot type save |
| 18:58 | jnettlet | don't store it in flash until you know it works ;-) |
| 18:59 | iio7 | When I type "boot" I get "syntax error" |
| 18:59 | jnettlet | printenv bootcmd again. |
| 19:00 | iio7 | Okay, it's missing the end. |
| 19:00 | _rmk_ | yay, that worked... on the hummingboard |
| 19:01 | jnettlet | \0/ |
| 19:01 | _rmk_ | that's now booting loading the initrd off the sd card, and pulling the kernel and dtb off the tftp server |
| 19:02 | _rmk_ | with a conditional (or two) on ${ethaddr} and ${cpu} to select an appropriate dtb |
| 19:02 | jnettlet | why do you have an initrd? |
| 19:02 | _rmk_ | to shutup ubuntu's upstart |
| 19:02 | jnettle | 19:02 * jnettlet shakes fist at upstart |
| 19:02 | _rmk_ | hmm, there's too many u's in that sentence |
| 19:03 | _rmk_ | if test ${ethaddr} = "d0:63:00:00:00:00"; then |
| 19:03 | _rmk_ | setenv fdtfile imx6dl-hummingboard.dtb |
| 19:03 | _rmk_ | else |
| 19:03 | _rmk_ | if test ${cpu} = "6SOLO"; then |
| 19:03 | jnettlet | :-D |
| 19:03 | _rmk_ | setenv fdtfile imx6dl-cubox-i.dtb |
| 19:03 | _rmk_ | else |
| 19:03 | _rmk_ | setenv fdtfile imx6q-cubox-i.dtb |
| 19:03 | _rmk_ | fi |
| 19:03 | _rmk_ | fi |
| 19:03 | iio7 | jnettle, I'm sorry but can I split the command up at the semicolons? |
| 19:03 | _rmk_ | now I just need to generate those dtbs :) |
| 19:03 | _rmk_ | ... which is a job for after food |
| 19:04 | jnettlet | iio7, they should all be a single line |
| 19:07 | iio7 | jnettlet, okay, got that.. "boot" gives me "syntax error".. heres printenv bootcmd: |
| 19:07 | iio7 | bootcmd=usb reset; for device_name in ${trydev}; do for partition in ${trypart}; do for directory in ${trydir}; do for fstype in ${tryfs}; do; |
| 19:08 | jnettlet | iio7, it seems like you are missing part after the last do; |
| 19:09 | jnettlet | iio7, I would reboot and then escape to the u-boot command prompt and look again |
| 19:09 | iio7 | Yes, I noticed from your paste that the final ' was missing. |
| 19:09 | iio7 | I'll do that.. thanks! |
| 19:11 | iio7 | I think its minicom, it cuts the end of. |
| 19:12 | iio7 | Yes, using screen instead gives a much longer command |
| 19:15 | iio7 | It freezes: http://paste.debian.net/76450/ |
| 19:16 | iio7 | the new bootcmd: http://paste.debian.net/76451/ |
| 19:23 | iio7 | It is like I can't get the entire command to paste, the end gets cut of. |
| 19:24 | jnettlet | iio7, well you can shorten that bootcmd if you are always going to boot from usb |
| 19:25 | jnettlet | if you want to keep it. To add "proper" usb support you should mimic something like if itest.s $device_name -eq ide; then if itest.s $ide_started -ne 1; then ide reset; but with ide instead of usb |
| 19:31 | iio7 | why ide instead of usb? |
| 19:32 | jnettlet | sorry other way around. |
| 19:32 | jnettlet | doing too much at once |
| 19:33 | iio7 | jnettlet, no problem, I'm grateful for your help and time. |
| 19:33 | iio7 | So I could just "remove" the ide part from the bootcmd since I don't need that. I only need to boot from usb, and if the usb should break down, then sd. |
| 19:33 | jnettlet | iio7, no problem good luck. you may also want to google u-boot bootcmd usb to get an idea of how to make a simple one that just loads your boot.scr direct form usb |
| 19:34 | jnettlet | really you don't need any of the complexity of that script if you only want to support the external usb device |
| 19:34 | iio7 | okay, thanks! |
| 19:35 | jnettlet | iio7, you can always reset back to the defaults so don't worry about breaking something. |
| 19:35 | iio7 | How can I reset it? |
| 19:38 | jnettlet | resetenv off the top of my head |
| 19:39 | iio7 | thanks a lot! |
| 19:46 | _rmk_ | bah, this is going to have to wait... cubox-i doesn't produce anything with that... I need to rescue the usb cable from powering the hummingboard |
| 19:46 | MarcusVinter | I thought the purpose of naming the C1 was to pick a name that expresses the modular side of things, the SOM. |
| 19:47 | _rmk_ | the carrier1 is the rpi-like board |
| 19:47 | _rmk_ | not the som, which is known as microsom |
| 19:47 | MarcusVinter | Oh.. |
| 19:47 | MarcusVinter | My bad. |
| 19:48 | MarcusVinter | the C1 is SOM too no? |
| 19:48 | MarcusVinter | why the other board mounted onto of the other? |
| 19:49 | _rmk_ | the cubox-i and the carrier1/hummingboard use the microsom, which has the iMX6 SoC, Ethernet phy and other bits on. |
| 19:50 | MarcusVinter | Oh its named in two parts? |
| 19:51 | iio7 | jnettlet, thanks a lot, the issue has been solved, the device is booting nicely and I don't need the boot.src now. |
| 19:52 | iio7 | Coolgeek, thanks for the link too! |
| 19:52 | MarcusVinter | Glad to hear that iio7 |
| 19:54 | MarcusVinter | Gotta run |
| 19:55 | _rmk_ | ok, well, the cubox-i dhcp'd but that's as far as it got |
| 19:59 | _rmk_ | ooh, that's better :) |
| 20:00 | _rmk_ | but I've got the dreaded magenta line down the left hand side... |
| 20:09 | _rmk_ | now booting from nfs :) |
| 20:10 | jmontleo | jnettlet, _rmk_ have you guys moved to using github yet? |
| 20:11 | _rmk_ | right, must leave (should've left 10mins ago) |
| 20:11 | rabeeh | somehow i broke the ethernet support in u-boot the last few weeks... i still can't figure out how |
| 21:07 | xraxor | loving the forum now :) |
| 21:09 | jnettlet | jmontleo, I have always used github |
| 21:10 | jnettlet | rabeeh, how so? the current uboot is working with ethernet on my image. |
| 21:11 | xraxor | guys maybe you should ask for a subforum for people to test your SW patches when needed |
| 21:11 | jmontleo | jnettlet, _rmk_ put a post on G+ that made it sound like he and/or others were ging to set up a github repo to carry future cubox-i/hummingboard work. Wasn't sure if that happened. |
| 21:11 | jmontleo | s/ging/going/ |
| 21:11 | jmontleo | I did find your 3.10.9 repo though :) |
| 21:12 | jnettlet | jmontleo, I think we are going to do that. Upstream is a bit slow. |
| 21:12 | jnettlet | that repo is working, but it is a bit behind now. I am fixing the 3.10.17 beta kernel and merging it with the linaro lts kernel for future releases |
| 21:13 | jnettle | 21:13 * jnettlet wants a kernel that doesn't need constant babysitting |
| 21:17 | rabeeh | jnettlet: use 3.0.35 :) |
| 21:18 | rabeeh | hi Ori_P |
| 21:18 | jnettlet | rabeeh, no babysitting not geriatric :-) |
| 21:18 | jnettlet | plus I want CMA and KMS and all the other goodies |
| 21:19 | Ori_P | hello rabeeh |
| 22:48 | hste_ | is sth wrong with the forum. I dont see the latest posts |
| 22:55 | hste_ | oh I found it was categorized now :) |
| 23:09 | _rmk | 23:09 * _rmk_ returns post-meeting. Now to sort out cubox-i stuff. :) |
| 23:12 | _dab_ | Is it possible to NFS boot from the i4pro? My uboot does not seem to see ethernet |
| 23:15 | _rmk_ | my i4pro is nfs booting at the moment, including fetching the kernel and dt over tftp |
| 23:18 | _dab_ | which uboot? I am using rabeehs SPL from http://imx.solid-run.com/forums/viewtopic.php?f=10&t=333&p=1888#p1885 |
| 23:18 | _rmk_ | I used the wiki page on building uboot/kernel |
| 23:20 | _dab_ | rmk can you tell me if uboot 'ping' works for you? |
| 23:20 | _rmk_ | dab: I haven't tried - I'm doing this: |
| 23:21 | _rmk_ | setenv autoload n; dhcp; tftp ${loadaddr} imx6.zImage; tftp ${fdt_addr} ${fdtfile} |
| 23:21 | _rmk_ | and that seems to be working |
| 23:23 | _dab_ | thanks, I will check the other uboot and see if I can figure out why ethernet fails on rabeehs binary |
| 23:23 | _rmk_ | rabeeh did mention that he'd broken ethernet and doesn't know why |
| 23:23 | _rmk_ | 19:11 < rabeeh> somehow i broke the ethernet support in u-boot the last few weeks... i still can't figure out how |
| 23:24 | _dab_ | aha |
| 23:24 | _dab_ | double thanks |
| 23:30 | _rmk | 23:30 * _rmk_ mails Fabio for comments on the thin magenta line. |
| 23:41 | _rmk_ | now... rabeeh's original carrier1 support had differences for the Ir receiver, I wonder if that's still true |
| 23:42 | _rmk_ | yep, it works on the hummingboard but not the cubox-i4 |
| 23:47 | _rmk_ | right, lets give that a go. |
| 23:55 | _rmk_ | that's better... that now works :) |