[00:07] Howdy === Jack87|Away is now known as Jack87 [00:58] Anyone active here? === asac_ is now known as asac [02:18] anyone know why this would be happening? http://yfrog.com/h6bbx1j === Jack87 is now known as Jack87|Away [04:17] Lopi: See bug 746133 [04:17] Launchpad bug 746133 in linux-ti-omap4 "Video loses sync on omap4" [High,Incomplete] https://launchpad.net/bugs/746133 === Jack87|Away is now known as Jack87 === Jack87 is now known as Jack87|Away === Jack87|Away is now known as Jack87 [06:31] * Jack87 wonders if folk here are considering the touchpad [06:36] of course, but hard to buy one === michaelh1 is now known as michaelh1|away === michaelh1|away is now known as michaelh1 [09:13] infinity, do you know when the .calc files in cdimage are needed? I see most but not all of the armel ones all set BOOT_SIZE_1=2 but I'd like to avoid it for mx5 if it is not strictly necessary [09:13] Building dependency tree... [09:13] E: Unable to locate package linux-linaro-lt-mx5 [09:13] P: Begin unmounting filesystems... [09:14] not sure what thre reserved space there is [09:14] janimo, hmm, that didnt go so well [09:14] ogra_, hmm, an -image- is missing [09:14] yep [09:14] well [09:14] the whole mucking around with cdimage is reminiscent of 'how many X are necessary to insert a lightbulb' jokes [09:15] we'll prevail eventually [09:15] you should just copy the omap script and start from there [09:15] ogra_, this is livecd right? I meant cdimage as the full build from livecd to publish [09:15] ogra_, yes I use the omap script [09:15] ah, good [09:16] right, the above is the live-build output [09:16] but here I deleted the explicit kernel name in the hope of it being picked up implicitly as per infinity's suggestion [09:17] welll, it still expects a toplevel metapackage ... [09:17] linux-image- is the linux-image metapackage only [09:17] ogra_, I think so too. So one still needs adding by jcrigby [09:17] yes, but not called linux-mx5 but linux-linaro-lt-mx5 [09:18] right [09:18] I thought the implicit rule (KERNEL_FLAVOUR) was about picking up the image not the meta [09:18] I was wrong [09:18] oh, you didnt use the other var ? [09:19] LB_LINUX_FLAVOURS="linaro-lt-mx5" [09:19] but then again we could have just left KERNEL_FLAVOUR untouched and created the linux-mx5 with the same result [09:19] ogra_, no, I used KERNEL_FLAVOURS only. Did I misread what infinity told me? Sigh [09:19] * janimo goes reread the logs [09:19] no, because other scriptds seem to use the snippet [09:19] and mx5 isnt the full uname string [09:19] right [09:19] ok [09:21] http://people.canonical.com/~ubuntu-archive/livefs-build-logs/oneiric/ btw [09:21] in case you want to check any live build logs yourself [09:22] ogra_, can you make the change in the script on antimony and see if that fixes (LB_LINUX_FLAVOURS instead of KERNEL_FLAVOURS) instead of me doind another shoot and miss upload? [09:23] no, live-build runs on the live builders ... not on antimony [09:23] i dotn have any access to fiddle on them, only http access to logs [09:23] so we need an upload [09:24] sigh [09:24] antimony is only dbeian-cd cdimage [09:24] *debian [09:24] so should I just change that varname? [09:25] did infinity not know that off the top of his head, with a medical examination pending and all? [09:26] well, i'm not sure that will do much else than KERNEL_FLAVOURS did [09:27] i think we're better off waiting for a meta upload and keep the rest as is [09:27] ogra_, agreed [09:27] so no upload for now === Jack87 is now known as Jack87|Away [09:27] unless infinity has any brilliant idea inbetween ... though i doubt he types in his sleep [09:28] (assuming thats what he does atm) [09:28] ogra_, so livecd.sh is no longer used right? [09:29] right [09:29] only BuildLiveCD and the stuff in the live-build subdir [09:29] I think we should put the 3 unused files there in the attic [09:29] anyone can find them in bzr history [09:29] might be there to makes sure we have a reference for possible regressions [09:30] cjwatson what do you think? ^ [09:30] but i agree we should drop them before final release [09:30] well it nevertheless confuses those looking at the code [09:30] seems to do much more than it actually does [09:30] livecd-rootfs was always brilliant in confusing people ... especially its massive amount of documentation :P [09:30] ogra_,unrelated: what about switching to ext4 before oneiric? [09:31] what could go wrong? [09:31] live-build could not support it [09:31] hmm, does it not support it on non-armel? [09:31] nothing beyond that, though i fear the u-boot stuff a bit [09:31] well linaro has switched to ext4 by default [09:31] and I test hwpacks like that [09:32] well, we can just try, but i suspect there are a good bunch of additional changes, it wont just be a switch we change [09:33] and the u-boot changes will bind our resources soon [09:36] any other ext3 assumptions in the code? [09:36] even the ext3 tools we call should work on ext4 [09:36] cdimage has some and debian-cd too [09:36] jasper indeed as well [09:37] for debian-cd changing CONF.sh might suffice [09:37] cdimage needs changes in bin/buildlive and bin/download-preinstalled-filesystems [09:38] we touch those these days anyway for imx53 [09:38] the linaro issue I ran into with ext4 is covered in our kernels [09:38] well, let me do a test run [09:38] CONFIG_LBDAF is set [09:39] * ogra_ changes buildlive ... that should suffice for seeing if live-build can cope on the buildd [09:44] ok, ac100 ubuntu-server build running with ext4 as default [09:44] lets see what comes out [09:45] argh ! [09:45] * ogra_ slaps forehead [09:45] silly me... i shouldnt test with a tar.gz image :P [09:46] luckily we have two buildds [09:46] * ogra_ fires off an omap server build on the other machine [09:49] I wish there was an easy way to install a local cdimage setup to test these before uploading things [09:50] well, even then you couldnt easily make sure that everything is always 100% in sync [09:50] it consists of to many parts to be really reliable [09:50] imho [09:50] working on these components is by far the most unpleasant for me, beats autoconf mucking in broken packages :) [09:51] you should join the cdimage team ... doing the changes on antimony makes working easier [09:51] ogra_, remember I woved not to join this team when I had to fight headless last cycle :) [09:51] and there are several levels of bzr trees, so its easy to roll back [09:51] isn't it the same craziness just via ssh? === michaelh1 is now known as on_reflection [09:51] or can you do exaclty what the build setup does in cron? [09:51] yes === on_reflection is now known as michaelh1 [09:52] i actually use the cron commands doing testbuilds [09:52] with additional env vars set to limit the build to one arch === doko_ is now known as doko [10:32] janimo, did you recently test any of our desktop images ? [10:32] on ac100 the greeter is so slow its close to unusable [10:35] janimo, hmm, so ext4 will need more changes in live-build it seems [10:35] i dont get an image file (everything else including logs , initrd and kernel files looks fine) [10:42] bah, so there is no ext4 support at all [10:45] I'm running custom kernel by Robert Nelson, which I compiled for beagle/ubuntu.. now when compiling lirc serial module for it, I run into a compiling issue: scripts/genksyms/genksyms: 1: Syntax error: "(" unexpected [10:45] ogra_, I tried testing panda a week ago, but did not come up :( [10:46] I checked in the linux source "../scripts/genksyms" with 'file genksyms' [10:46] genksyms: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, not stripped [10:46] scripts are not normally compiled in part of the kernel compilation ? [10:46] ogra_, I need to test desktop images for bootcharting them, but they keep being broken when I test :( [10:46] the code doesnt know it [10:49] yeah, currently the archive skew breaks them once again [10:49] anyway, no ext4 for us without extra effort, live-build doesnt support it at all [11:01] ogra_, oh well, at least you tested it. Shall we file a bug on it? [11:05] i think there ios a bug and an upstream patch ... one sec [11:05] bug 803547 [11:05] Launchpad bug 803547 in live-build "live-build lacks EXT4 support for binary image types" [Undecided,Invalid] https://launchpad.net/bugs/803547 [11:44] geez, the mkfs stuff in live-build is super inconsistent [11:45] sometimes it uses mkfs.${LB_CHROOT_FILESYSTEM}, sometimes mkfs.${MKFS} and often eough it is just hardcoded === michaelh1 is now known as michaelh1|away === mayday_jay is now known as mayday_jay-away [12:12] ok, i reverted the default to ext3 again (but left the other ext4 code in cdimage, so we dont need to touch that again) === mayday_jay-away is now known as mayday_jay === mayday_jay is now known as mayday_jay-away === mayday_jay-away is now known as mayday_jay [15:03] janimo, ogra_: I see from comments above that you need something changed in the mx5 meta pkg? Similar but different than what was discussed yesterday? [15:04] jcrigby, yes, similar issue, but different metapackage name [15:04] ogra_: janimo: also, we can already replace x-loader with u-boot-linaro spl for omap 4 [15:04] ogra_, has the exact error msg IIRC [15:04] well tested and no regression with 11.08 [15:04] rsalveti, I am all for it [15:04] but it is one of those cdimage jobs which only some folk can handle effectively [15:05] rsalveti, good to hear it works well [15:05] janimo: when can we push the new package? [15:05] or, who can we talk to to make it work? ogra_ ? [15:05] rsalveti, hmm, the new package switches to SPL with no backwards compat? [15:05] would be nice to push it before beta-1 [15:06] janimo, that is the one downside [15:06] janimo: nops, x-loader doesn't work anymore with it [15:06] that's why we need to be in sync [15:06] rsalveti, definitely before beta 1. We expect breakage these days. ogra would know better when to synchronize [15:06] rsalveti, ubuntu-arm meeting ATM [15:06] maybe something to bring up there ? [15:07] sure [15:07] jcrigby: #ubuntu-meeting [15:08] jcrigby, rsalveti there should be an images topic soon, and then it can be brought up [15:08] janimo: great [15:31] Hi there! Is there someone who can help me with rootstock? when trying to create a ubuntu rootfs, i get this: http://paste.ubuntu.com/674573/ [15:43] nicofs: try as root: lsof | grep /tmp/tmp.EvKThwZJAG/tmpmount/proc [15:43] This may show what process is still using that directory [15:44] dmart, what process is using the image is not so much my concern as the dpkg error... [15:44] Sometimes this is just caused by a race between some program terminating and unmounting that filesystem -- if you try again, you may find that it works [15:44] dmart, been trying for 2 days now... [15:45] dmart, --seed xubuntu-desktop works, so does ubuntu-minimal. [15:45] nicofs: oh, I see. [15:45] ubuntu-desktop fails at that stage. in natty, maverick and lucid [15:45] rootstock may spit out some logs you can look at ... I'm not sure offhand where to look, though [15:46] Have you got a more complete log? [15:46] jcrigby: ok, then it seems we only need a bug for that [15:46] as we need a FFe [15:46] dmart, it does normally - but not in that case... they should be in the same folder as the finished rootfs... [15:47] an explicit bug for the FFe or an existing bug to justify the FFe [15:47] nicofs: maybe the logs are being deleted on error, but I would hope not. Error conditions are when you _want_ the logs [15:47] dmart, what i pasted is what my console gave me... i can redirect the output of rootstock to a file and paste that... [15:47] jcrigby: bug for FFe [15:48] ok [15:48] nicofs: I suggest capturing the whole log ... there may be earlier errors [15:48] jcrigby: FFe: replace x-loader with new u-boot-linaro SPL [15:48] something like that [15:48] for omap 4 [15:48] rsalveti, ok got it [15:48] jcrigby: https://wiki.ubuntu.com/FreezeExceptionProcess [15:48] dmart, yes, i just need to run it again for that... can't scroll back so far in my console... [15:55] rsalveti, the uboot+SPL blueprint's last WI suggest UBOOT still boots with xloader. [15:56] janimo: it was tested, but the test result is not posted there [15:56] dmart, i could try --no-root - maybe that helps... [15:57] but is still works then with xloader? I thought xloader does not boot uboot, from what I understood [15:57] the new uboot that is [15:57] janimo: nops, just updated the bp [15:57] it doesn't work with x-loader anymore [15:57] ok [15:58] because to remove duplicated code, some part of u-boot was moved to spl [15:58] rsalveti, thanks for the clarification. I filed the FFE for uboot and linked the bp, so it helps if it is clear [15:58] so once you build with spl, it doesn't work with x-loader anymore [15:58] ok [15:58] janimo: great [16:01] nicofs: don't know -- I suggest you just repeat whatever you did last time, but capture the full log. The first line of your paste looks like the end of an error report, rather than the whole report [16:01] nicofs: The "script" command is useful for logging terminal sessions [16:02] dmart, i just started and appended " &> logfile" [16:03] rsalveti, I don't understand this line:Please note that we expect requesters to have an updated package already prepared and tested! You will need this anyway to provide proper build logs. [16:05] jcrigby: just to know that you already built the package locally and tested [16:05] and that we already did with 11.08 release [16:05] rsalveti, ok so the version in the ppa fills that part [16:05] yes [16:07] nicofs: that also should work [16:08] dmart, just takes quite a while... i guess about 15mins [16:25] dmart... ok different error... my bad... have to start again [16:45] just ran out of memory... [16:48] rsalveti, linaro-media-create also needs to be uploaded, as it now assumed uboot.bin right? [16:48] I tried 3 times today until it downed on me that what we were discussing today may be relevant to my error, not finding uboot.bin [16:49] I mean lmc needs uboot.img on panda now. Testing 11.08 daily hwpack [16:49] janimo: yeah, it's uploaded only at the PPA I believe [16:50] rsalveti, so I'll add a new FFE and we need to add that too. Is it also stable enough? [16:50] has the same cycle as the rest of linaro? [16:50] janimo: yes [16:50] ok [16:51] would be nice to have a FFe for it too [16:51] rsalveti, who is in charge of lmc. Can they upload or should I sponsor - asking them first if it is ok [16:51] guess james_w can upload it [16:52] rsalveti, actually not sure if it needs a FFe if it's a universe package and not part of default Ubuntu images [16:52] NCommander, does it? ^ [16:53] it is not technically a release team supervised feature [16:53] janimo: I thought all package updates that are not only bugfixing needs a FFe [16:54] so even universe-only changes [16:54] rsalveti, I may be wrong indeed. This technically fixes a bug - cannot make panda linaro images without [16:54] also adds features I guess [16:54] but is also a newer version [16:57] bbl, lunch [18:39] NCommander, https://bugs.launchpad.net/linaro-image-tools/+bug/834003 another FFE request [18:39] Ubuntu bug 834003 in linaro-image-tools "FFE: upload 11.08 to Oneiric" [Undecided,New] [18:39] closely related to the uboot one [19:06] Anyone know if the USB speed issue has been solved yet? [19:08] Nope. [19:08] such is life [19:14] Hey guys! [19:15] Can anyone help me make a chroot to boot onto my touchpad? [19:24] MrCurious: Keep an eye on bug 709245. Work is being done. [19:24] Launchpad bug 709245 in linux-ti-omap4 "ARM SMP scheduler performance bug" [High,Confirmed] https://launchpad.net/bugs/709245 [19:25] thanks gruemaster [19:25] PDunny: What kind of touchpad? [19:25] you weren't kidding.. longest bug page i have seen yet [19:26] You might be able to use the ubuntu-core image. [19:26] wow! almost an order of magnitude potential gains! that is going to be a game changer [19:27] GrueMaster - HP Touchpad [19:27] PDunny: Ok, so at least armv7. You should be able to use the ubuntu-core image as a base. [19:28] Gruemaster - Would the kubuntu-mobile project be suffice? It has most of the features I am looking for in it just not sure how to make the chroot [19:29] I don't know. I don't work with that image. It might though. [19:30] You could probably just loop mount the image and chroot into it that way. [19:31] ok .... hum .... not in my ubuntu partition right now:/ [19:32] Have yall had much luck with cellphones/tablets yet? [19:32] The touchpad community is essentially begging for something more than just webos so if you need testers for OMAP4 devices ..... [19:35] I thought the HP system was snapdragon. [19:35] that is what i thought as well [19:36] It is snapdragon but from the research I have done it looks like it is OMAP4 [19:37] * PDunny newb not sure [19:48] Both are dual core iirc. They may look the same for the most part. Need to check /proc/cpuinfo. [19:49] I had seen somewhere that it was on a compatable list as an omap4 processor so entirely possible I was wrong, of course I bookmarked it on my ubuntu partition === mayday_jay is now known as mayday_jay-away === michaelh1|away is now known as michaelh1 === michaelh1 is now known as michaelh1|away === michaelh1|away is now known as michaelh1 [21:54] hi [21:54] do you know the offcial forum for ubuntu arm support [21:55] Try #ubuntu-arm. Oh, wait... [21:55] :P [21:55] What can we help you with? === michaelh1 is now known as michaelh1|away [21:58] So kind of generic question.. would an ubuntu image just replace my android system partition? [21:58] or are we looking at modifying partitions for space? [21:59] What is the platform? We currently are only supporting TI dev platforms (beagle, beagleXM, panda), although there has been work in the community for other platforms. [21:59] I've got a Toshiba Thrive... it's been on my mind [22:00] and unfortunately not close enough to the other Toshiba Android device for me to say it'd be a simple push... [22:01] Unfortunately, it is not a platform that is currently being worked with, but someone in the community may have something based on ubuntu. I know there is a community behind the Toshiba AC100, which uses the same basic core hw. [22:02] I don't know if there is an active community yet for that system. [22:02] the AC100 doesn't seem close enough to the Thrive for me to think the Hw is compatible. [22:03] They are both nVidia Tegra2 dual core Arm Cortex A9. That in of itsself is a major step in the right direction. [22:04] I'm just trying to wrap my mind around how it would be booted, first of all... there is about 600mb of space in the System Partition to use [22:04] actually there may be more. [22:05] 688M according to df [22:05] What about other partitions? [22:05] We tend to repartition our ac100s a bit. [22:05] data - 2g, cache - 1007M [22:06] I believe the ac100 image uses a larger partition, like the user partition. [22:06] Though you can dual-boot android and Ubuntu if you feel the strange desire. It's just irksome. [22:06] Ah. [22:06] storage - 9G [22:06] (When we dual-boot, we just take over that storage partition) [22:07] When not, we reparition and turn system+data+cache+storage into one nice large one. [22:07] hmm [22:08] can't partition the thrive, we only have fastboot [22:08] We only have fastboot on the ac100 too. [22:08] Not sure how that relates. :P [22:09] thats when it gets scary for me I guess. lol [22:09] I'm willing to bet that the Thrive is pretty much just "an ac100 without a keyboard". [22:09] I'd put pretty good money on it, even. [22:09] But without touching one, it's hard to know for sure. :/ [22:09] I've looked at the hardware, too many other differences. [22:09] Care to come visit? :) [22:09] visit where? [22:09] Calgary, AB. [22:10] Too far [22:10] although I could probably give simple access to a shell, as close to touching one I can give. [22:14] unless you're willing to come to California ;-) [22:15] Well, I'll be in Santa Rosa in early September. :P [22:16] 7 hour drive to there [22:16] what are you doing up there? [22:17] Linux Plumbers. [22:23] I just wonder why it isn't as simple as pushing a new system partition to get a different linux distro booted on these arm devices.. or could it be? and it's just not ready yet? [22:23] I mean I understand the requirements that the kernel needs to know what it's on, etc. [22:25] Part of the problem is the vendor. Some (most) vendors don't want you mucking with their default OS (Windows 7, Android, OSX, etc). [22:25] So they don't make it easy for you to do so. [22:26] Tablets though have seemed to be fairly easy to root, and like the thrive, fastboot isn't locked. [22:26] Try to get factory warranty service on a system that originally sold with Windows 7 that you have installed Ubuntu on. [22:27] what fs type is http://cdimage.ubuntu.com/releases/11.04/release/ubuntu-11.04-preinstalled-headless-armel+omap.img.gz [22:27] That may be true, but have they posted specs on how to get the HW to work in the kernel? [22:27] no... lol... of course not [22:28] Lopi: The file is an image of an SD card with a fat32 partition and an ext3 partition. [22:28] GrueMaster: how can I mount it? === michaelh1|away is now known as michaelh1 [22:29] If you want to mount the fs with -o loop, there is a bit of a process to do it. First, use gunzip to uncompress it. [22:29] yeah I did that [22:29] Then type "file ubuntu-11.04-preinstalled-headless-armel+omap.img" to find out the partition locations. [22:30] okay [22:30] To mount a partition, type "sudo mount ubuntu-11.04-preinstalled-headless-armel+omap.img /mnt -o loop,offset=$((512*)). [22:30] Give me a sec and I can give you an exact example. [22:31] will this cause any issues if I'm using emulated filesystems? [22:31] Huh? [22:31] I'm assuming you are doing this on a Linux desktop system. [22:32] (i.e. Ubuntu Desktop). [22:32] yeah I am [22:32] what I'm asking is [22:32] can I get rid of these partitions and create a new image with this image [22:32] on my device, the ramdisk mounts the nand then mounts a loopback image to create an emulated fs [22:32] Yes, but it is not easy. [22:32] then pivot roots, etc. [22:34] So, back to the mount process, on my system "sudo mount ubuntu-11.04-preinstalled-headless-armel+omap.img mnt -o loop,offset=$((512*144585))" will mount the second partition. [22:34] I see [22:34] But I think what you would want to do is beyond the scope of the preinstalled images. [22:35] can I generate my own natty rootfs with rootstock? [22:35] Part of the boot process of these images is to resize the ext3 partition to fill the SD card being used, from 4G to 32G. [22:35] ah [22:35] It also runs oem-config to customize the image to the user. [22:35] yeah that would be a problem :/ [22:36] lilstevie was telling me about this [22:36] You might have a better time taking our new Ubuntu-Core image and expanding on it. It is smaller than the headless image by quite a bit, but is usefull to start with as a chroot environment. [22:37] hm okay [22:37] Someone needs to remind me to purge rootstock with fire for the next cycle. [22:37] where can I get ubuntu-core [22:37] You also can't boot directly to it as there is no kernel, but you can boot a kernel and have it mount a filesystem with ubuntu-core as / [22:37] I'll need a spec to make sure that whatever functionality it has lives sanely elsewhere. [22:38] http://cdimage.ubuntu.com/ubuntu-core/daily-preinstalled/ [22:38] (And don't download "current", it's not actually current... But the dated directories are) [22:38] btw, this is for Ubuntu on the iPad [22:38] (This will be fixed tomorrow :P) [22:38] you can see the linux kernel "booting" on it here: http://yfrog.com/hsu4bigj [22:38] Lopi: Then ubuntu-core is probably your best starting point. [22:39] GrueMaster: thanks for the help [22:39] Nice [22:41] it's going to be awhile before I have interaction with the image, but I figured I might as well get it booting now ;p === panda is now known as Guest45974 === michaelh1 is now known as michaelh1|away