**** BEGIN LOGGING AT Fri Aug 03 02:59:59 2012 Aug 03 03:28:49 frrgwregw, there are no ubuntu images for the beaglebone Aug 03 03:29:38 mturquette, i guess yuo have to wait for rsalveti to finish the port, pvr isnt ready for the new kernels atm Aug 03 03:30:26 mturquette: the leb itself should have the pvr driver already by default Aug 03 03:30:37 if used the lt-panda-x11-base Aug 03 03:30:38 ah Aug 03 03:31:04 ogra_: which timezone are you in this week? Aug 03 03:31:05 :-) Aug 03 03:31:33 boston Aug 03 03:31:44 tropical TZ as it seems :) Aug 03 03:32:12 got it :-) Aug 03 03:32:19 still like 30°C around 12pm and 90% humidity Aug 03 03:32:28 wow, nice :-) Aug 03 03:32:32 heh Aug 03 03:32:34 a bit too hot I'd say, but fine Aug 03 03:32:37 yeah, pretty unexpected Aug 03 03:33:22 * ogra_ is at the QA sprint getting automated image tests to run ... which work fine now thanks to you and john Aug 03 03:33:35 (which was pretty well noticed by mgmt) Aug 03 03:33:50 cool Aug 03 03:33:52 ;) Aug 03 03:33:55 ogra_: how things are going there in general? Aug 03 03:34:15 still missing anything at your initial plan for the sprint? Aug 03 03:35:18 no, i'm confident we'll be completely done tomorrow ... automated image booting is done, now only preseeding is missing Aug 03 03:36:03 cool Aug 03 03:36:13 would be nice to get a fully automated installer running :-) Aug 03 03:37:00 yup Aug 03 03:41:32 Fully automated installation is running all over the place. Aug 03 03:41:46 But automated image testing is indeed harder, so yay for efforts to make that go. Aug 03 03:43:12 :) Aug 03 03:51:07 rsalveti: the 12.07 leb? lsmod doesn't show it up Aug 03 03:53:14 mturquette: which image are you using? Aug 03 03:54:21 rsalveti: http://releases.linaro.org/12.07/ubuntu/leb-panda/lt-panda-x11-base_20120723-197-ubuntu-desktop.img.gz Aug 03 03:57:07 mturquette: hm, with that image it should work by default Aug 03 03:57:12 can you try reinstalling pvr-omap4? Aug 03 03:57:18 sudo apt-get install --reinstall pvr-omap4 Aug 03 03:57:57 * mturquette tries Aug 03 03:58:15 E: dpkg was interrupted, you must manually run 'dpkg --configure -a' to correct the problem. Aug 03 03:58:25 hmm, i may have deeper problems. this is a fresh imamge too. Aug 03 03:58:37 just pulled it and dd'd it to sd card today... Aug 03 03:59:05 hm, it shouldn't behave like that by default Aug 03 03:59:17 maybe it got corrupted while writing to the sd card Aug 03 04:00:36 rsalveti: yeah, i'll pull the image again tomorrow morning, check md5sums and try with multiple cards if needed Aug 03 04:00:52 rsalveti: thanks for the ideas. hopefully it will Just Work (tm) tomorrow. Aug 03 04:01:31 :-) Aug 03 04:05:56 rsalveti, iirc jimerickson filed a bug yesterday that showed some issue with dos2unix or some such ... Aug 03 04:05:56 on ubuntu though Aug 03 04:05:58 ib was wondering why pvr does need it ... but was to busy to actually take a look this week Aug 03 04:06:30 hm, will check that Aug 03 04:06:58 I remember it needed to convert a few files to unix so it could build successfully Aug 03 08:57:38 I have a vodafone webbook with ubuntu lucid on it that I would like to upgrade. /proc/cpuinfo says "Hardware: Freescale MX51 Babbage Board" Aug 03 09:03:44 I see "As of 10.10, Ubuntu has dropped support for the Freescale i.MX51 Babbage board." - should I look at debian instead ? Aug 03 09:05:11 wizzy: hm what arm revision is that anyways? Aug 03 09:05:42 Processor: ARMv7 Processor rev 5 (v7l) Aug 03 09:07:35 then chances are that it can be convinced to run ubuntu, no idea though how much work it would be. Aug 03 09:08:04 I have limited physical access - I don't seem to be able to open the case, but if I did would probably find a 4G flash chip soldered down. there is no CDROM. Aug 03 09:08:52 summary thread at https://lists.ubuntu.com/archives/ubuntu-za/2012-July/009425.html Aug 03 09:09:00 if you have access to the bootloader, via serial or such there's many things you can do. if you have not, chances to brick the device are considerably high. Aug 03 09:10:08 it has 3 partitions, not including beginning-of-disk, and a switch that puts me into a debian bootloader, that fails to find a cdrom Aug 03 09:10:44 no idea what you mean by "debian bootloader" Aug 03 09:11:12 wizzy: the thread you linked seems to recommend debian, though. Aug 03 09:11:13 https://lists.ubuntu.com/archives/ubuntu-za/2012-July/009438.html Aug 03 09:11:24 I *think* the debian bootloader is on the (unstructured) first partition (possibly as kernel+initrd) Aug 03 09:11:52 LetoThe2nd: yes - that is me :( Aug 03 09:12:14 AFAIK, there is no such thing as "the debian bootloader". hence i really have no idea what you are talking about. Aug 03 09:13:07 wizzy: We still support it in userspace just fine, but we don't ship babbage kernels anymore. Aug 03 09:13:18 ok - it puts me into the multi-step installer that was used (is used ?) on the alternate ubuntu disks and the debian installer Aug 03 09:13:52 how far can I push userspace and not bump the kernel ? Aug 03 09:14:19 Precise will mostly work on your current kernel, I suspect. Aug 03 09:14:29 Except for the possibility of things like binary video drivers and such. Aug 03 09:15:11 so if I pin the kernel and upgrade I might be OK ? Aug 03 09:15:30 It's worth an experiment. Aug 03 09:15:39 what other packages should I be careful of ? Aug 03 09:16:02 I dunno, havning never seen how the webbook was customised. Aug 03 09:16:09 having... Aug 03 09:16:24 if I brick it, how do I get a CD image that the installer will recognise ? Aug 03 09:16:47 Hrm? Aug 03 09:17:01 You shouldn't need a "CD" at all, didn't you just say there was a recovery partition? Aug 03 09:18:04 If that's not working right, I'm really the wrong person to ask. Aug 03 09:18:11 Like i said, I've never seen one of these devices. Aug 03 09:18:13 it has a little switch behind the battery that puts me into something else, that I am calling the debian installer. This fails at the find-cd step Aug 03 09:18:55 I have not bricked it yet - I am just terrified of doing so .. :) Aug 03 09:19:33 if that is what you call the recovery partition (I think it is) it is missing install media Aug 03 09:20:40 the whole flash is only $G Aug 03 09:20:45 err - 4G Aug 03 09:21:27 and it has no physical ethernet - only wireless Aug 03 09:21:56 Yeah, we've reached the point where I'd probably have to sit in front of one. Aug 03 09:22:05 (And also where it's 3:21am and I need to sleep) Aug 03 09:22:24 infinity: coffee? spice? coffee + spice? ;) Aug 03 09:22:44 11AM here - and we are beating the brits at cricket and mopping up gold medals :) Aug 03 09:23:45 infinity: thanks - I will consider my next steps Aug 03 09:27:15 marvin24, uploaded new package with your latest branch to quantal Aug 03 09:28:02 janimo: thanks! Aug 03 09:28:16 seems it's time to upgrade ;-) Aug 03 09:28:29 marvin24, the meta is not yet there, so it is not automatically taken Aug 03 09:28:40 but will be in tomorrow's image if it goes through today Aug 03 09:29:17 marvin24, I saw lockups a few times though with this kernel (maybe the older 3.1 too). Once on boot 'CPU0 stalled' Aug 03 09:29:40 but we'll fix them in time, so better have it uploaded for people to test Aug 03 09:30:27 yeah, I hope nvidia will make the finial r15 release before quantal Aug 03 09:30:52 janimo: ever tried to build mainline kernel for tegra? Aug 03 09:30:53 ah, looks the package is no longer in new, and nice guy infinity uploaded the meta too Aug 03 09:31:02 marvin24, had no time yet Aug 03 09:31:06 I plan to though Aug 03 09:31:16 it has the best chances to support multi-board Aug 03 09:31:23 will start with 3.5 then try out arm-soc if that does not cut it Aug 03 09:31:44 I am sure DT fixes multiboard but if graphics does not work yet it's not usable right? Aug 03 09:32:03 janimo: I beg to differ. Aug 03 09:32:08 but yes, I plan to check it out as soon as I do the next round of ac100 work Aug 03 09:32:09 janimo: display and fb is initialized Aug 03 09:32:22 so, xorg works kindof Aug 03 09:32:23 which is syncing up with ubuntu configs as much as possible Aug 03 09:32:24 janimo: (re: "the meta is not yet there") Aug 03 09:32:55 janimo: Oh, you mentioned that later. :P Aug 03 09:32:59 infinity, see my later statement in which not only I realize my mistale but I paise you for your niceness level Aug 03 09:33:15 IRC is not supposed to be read cronologically Aug 03 09:33:19 I'm not nice, just anal. Aug 03 09:33:28 hasthe modesetting driver arrived in quantal yet? Aug 03 09:33:43 anal guy infinity, uploaded meta. Aug 03 09:33:45 works for me too Aug 03 09:33:52 Kinky. Aug 03 09:33:58 sure Aug 03 09:34:40 seems so, https://launchpad.net/ubuntu/+source/xserver-xorg-video-modesetting/0.4.0-0ubuntu1 Aug 03 09:35:57 * janimo pulls out google to once and for all find out whether anal also has the meaning of thorough, picky, attentive to detail which geeky chats usually implies Aug 03 09:36:19 * janimo finds out if that is the case it is most certainly not the most popular meaning of the word by far Aug 03 09:38:32 marvin24, I do not know anything about the modesetting driver or what it is supposed to do exactly Aug 03 09:39:04 janimo: it is required for mainline kernel xorg Aug 03 09:39:18 marvin24, so then what does not work? DRM and GPU support? Power management Aug 03 09:39:19 ? Aug 03 09:39:29 the latter Aug 03 09:39:36 back in a few min Aug 03 09:40:00 modesetting driver is a simple framebuffer driver for xorg using kernel drm Aug 03 09:40:20 but the quantal package seems to require xorg-server 1.13 ... Aug 03 09:40:31 * marvin24 is confused Aug 03 09:42:14 janimo: It's shorthand for "anal-retentive", which has that meaning, yes. Aug 03 16:21:19 janimo: did you tried to install the new kernel? Aug 03 16:21:29 yes, yesterday Aug 03 16:21:30 I get: /tmp/fileIH3Shr: updated is too big for the Boot Image (8429568 vs 8388608 bytes) Aug 03 16:21:33 worked for me Aug 03 16:21:38 hmm Aug 03 16:21:40 but that's on precuse Aug 03 16:21:46 *precise ... Aug 03 16:21:47 I tried on precise too Aug 03 16:21:52 via dpkg -i Aug 03 16:21:58 same here Aug 03 16:22:06 what's you initrd size? Aug 03 16:22:17 hmm wait, I tried a .deb I build locally not the exact one from the archive Aug 03 16:22:23 I did a test it worked so I uploaded Aug 03 16:22:34 initrd size? I think it too is 8M Aug 03 16:22:51 no, 3446096 here Aug 03 16:23:25 was wrong, 4980696 Aug 03 16:23:30 kernel is 3446096 Aug 03 16:25:34 won't wake from suspend though Aug 03 16:34:54 huh Aug 03 16:44:25 janimo: for the initrd problem, the larger image includes a dm-mod.ko which isn't present with my self-compiled kernels Aug 03 16:44:42 I try to flash it with my old initrd Aug 03 16:45:34 it may turn out we need to disable some configs if we hit the initrd size limitation Aug 03 16:45:48 marvin24, is this too the >2M initrd issue? Aug 03 16:46:39 having DM is probably not that important but at one point it was requested by ubuntu/arm/server devs who did not yet have proper server hardware to test on Aug 03 16:47:20 janimo: no, flash-kernel doesn't like it because kernel+initrd becomes > 8 M Aug 03 16:47:35 ah there is an explicit check Aug 03 16:47:52 but now I wonder why the old initrd loads at all given the 2M limit Aug 03 16:47:55 but which of the two grew most? I did not change much of the kernel configs Aug 03 16:48:18 should the rebase on the new branch cause significant size increase? Aug 03 16:48:47 can't tell now, because kernel gave a nice oops on boot ;-) Aug 03 16:50:00 stack-protector ... Aug 03 16:50:12 during usb probe Aug 03 16:50:25 janimo: did you really tested it? Aug 03 16:52:25 I am running 3.1.10-4 now Aug 03 16:52:34 but as I said a locally built .deb Aug 03 16:52:39 not the exact one from the arch Aug 03 16:52:48 * janimo wgets that one and tries in a few minutes Aug 03 17:05:04 anyway, there should be no stack corruption Aug 03 17:17:14 kernel boots fine with 2M initrd Aug 03 17:18:47 janimo: sorry, false alarm Aug 03 17:19:03 I still had that 3M testfile with random data in my initrd Aug 03 17:19:16 marvin24, so all ok? Aug 03 17:19:36 wait, I'll retest again Aug 03 17:19:53 but flash-kernel should also check for the 2M initrd limit Aug 03 17:20:22 lot of users have problems with this Aug 03 17:22:59 and I think all tegra devices using fastboot have this problem Aug 03 17:29:45 janimo: ok, kernel boots, but resume fails Aug 03 17:30:27 ogra_, re flash-kernel ^ :) Aug 03 17:30:45 marvin24, whereas resume works in your own zImage? Aug 03 17:30:59 yes Aug 03 17:31:13 maybe I can get something from the serial console Aug 03 17:33:04 arrr, it works on my other ac100 (which has no wifi or 3g modem anymore) Aug 03 17:33:17 and that's the only one with a serial console Aug 03 17:47:31 janimo: can you give me a short course how to build my own kernel.deb? Aug 03 17:47:45 or is there some wiki page? Aug 03 17:49:13 marvin24, not usre if I put it in the wiki or just an email on the list Aug 03 17:49:16 a moment Aug 03 17:49:40 ah, I remember reading something from you on the ac100 list Aug 03 17:49:44 * marvin24 checks Aug 03 17:50:59 you need to checkout my packaging branch though which gives you two extra dirs, debian/ and debian.linaro (historical reason, I used linaro kernels when starting ac100 work) Aug 03 17:51:12 got it already .. Aug 03 17:51:45 and now? Aug 03 17:51:49 the packaging-3.1 branch Aug 03 17:52:18 do a git fetch + git reset --hard origin, I am just pushing out now the most recent changes Aug 03 17:52:23 for 3.1.10-4.5 Aug 03 17:52:43 and now, you need to have the arm cross gcc package installed Aug 03 17:53:21 and also sudo apt-get install libc6-dev-armhf-cross Aug 03 17:53:40 now in the kernel tree do Aug 03 17:53:47 I only see a Ubuntu 3.1.10-3.4 Aug 03 17:53:49 tag Aug 03 17:54:15 did try an update+reset as I said above? Aug 03 17:54:35 I just pushed the latest 3 min ago Aug 03 17:54:45 I hope I pushed to the right branch Aug 03 17:55:07 marvin24, hmm you're right it's not there Aug 03 17:55:22 not on the web interface Aug 03 17:55:52 but I forgot to fetch the tags Aug 03 17:55:53 (git fetch -t) Aug 03 17:56:09 just pushed now, forgot git push -u Aug 03 17:56:35 I though push -f should have been enough but no Aug 03 17:57:07 I have an alias fdr='fakeroot debian/rules' Aug 03 17:57:30 since that makefile's rules are commonly called from the cmdline when working with kernels Aug 03 17:57:34 so for example Aug 03 17:57:36 fdr clean Aug 03 17:57:38 fdr editconfigs Aug 03 17:57:41 fdr updateconfig Aug 03 17:57:53 ok Aug 03 17:58:10 do a git clean -xdf just to make sure there's no cruft there (just do not lose untracked files!) Aug 03 17:58:15 then fdr clean Aug 03 17:58:34 debuild -eDEB_BUILD_OPTIONS="parallel=3" -eCROSS_COMPILE=arm-linux-gnueabihf- -b -aarmhf -us -uc -nc Aug 03 17:58:55 change the parallel to suit your number of cores+1 (I think) Aug 03 18:02:05 seem I need to install some more package first ;-) Aug 03 18:02:55 ok, now it seems to do something Aug 03 18:04:42 you may be missing some other packages. If it fails to build the tools at the end you can turn that off by setting do_tools=false Aug 03 18:04:46 in debian.linaro/rules.d/armhf.mk Aug 03 18:08:37 poor athlon ... Aug 03 18:44:24 using the 3-Aug-2012 omap4+armhf image on a pandaboard ES. the installer asks me to designate a root and swap partition. when i do it fails with /dev/cdrom cannot be unmounted. why is the installer seeing a cdrom? Aug 03 19:22:00 jimerickson: What partitions are you setting up as root/swap? Aug 03 19:22:15 jimerickson: You're not trying to install to the SD card, are you (which it tells you not to do)? Aug 03 19:31:02 sadly it doesnt, it says you can install to a pre-made partition (which doesnt work either) Aug 03 19:32:23 ok so if i don't install to the sd where do i install to? usb? Aug 03 19:32:33 yes Aug 03 19:32:42 ok i will try it Aug 03 19:32:51 i will update the wiki instructions next week, once i'm home again Aug 03 19:33:06 just use guided partitioninig and let it do its job ;) Aug 03 19:33:11 ok Aug 03 19:33:20 it will default to the USB disk (if you attached one indeed) Aug 03 19:33:33 will do Aug 03 19:34:31 that should just finish without errors Aug 03 19:34:50 (and dont check the third party drivers option on the first page of the installer ;) ) Aug 03 19:35:38 ok Aug 03 19:37:55 can the usb drive be on a powered hub or does it have to be on the OTG port? Aug 03 19:38:35 no, it actually *needs* to be on the normal USB port (or a hub attached to this) Aug 03 19:38:41 OTG isnt running in host mode Aug 03 19:39:02 ok got it Aug 03 19:40:23 but last time i did this it didn't write a uboot partition. will be back in a bit. Aug 03 19:41:14 it will write u-boot to the SD card (as i told you last time we talked) Aug 03 19:41:32 the SD needs to stay in the slot even after install ... Aug 03 19:41:39 ok Aug 03 19:42:01 btw, many many thanks for all that testing, you rock ;) Aug 03 19:42:42 so now to boot panda i need the sdcard for booting and a usb drive for root? Aug 03 19:42:51 :) Aug 03 19:43:41 yep Aug 03 19:43:46 ok Aug 03 19:43:53 i am on it Aug 03 19:44:44 running root from SD card is very slow so we decided to switch the images to actually default to USB ... just lacing install documentation for this, i guess i should aslo blog about this change alongside creating new wikipages Aug 03 19:44:56 *lacking Aug 03 19:46:35 so now it will overwrite the installer portion at the end so it boots to usb root? Aug 03 19:56:04 i guess my class 10 32GB sd cards are kind of useless now. oh well. Aug 03 19:58:07 use them for something useful and get a 2G one ;) Aug 03 19:58:29 that i shall! Aug 03 20:00:54 marvin24, any luck with the build? Aug 03 20:01:16 marvin24, which is exactly your most mainline branch? for-next seems to be 3.2.0 based Aug 03 20:01:36 marvin24, nevermind, 3.5 Aug 03 20:01:49 I looked at video-next-test Aug 03 20:09:47 yay, srwarren is back in #ac100 ... Aug 03 20:10:03 * ogra_ has hope for a working tegra driver again :) Aug 03 20:32:13 install finished without errors as predicted by ogra_ but it booted to the installer again when done. what am i doing wrong? Aug 03 20:34:16 jimerickson, hmm, that might be my fault, look on the first partition of the SD and check if there are uEnv.txt and preEnv.txt files along with boot.scr Aug 03 20:34:25 if so, remove the two .txt files Aug 03 20:34:41 will do. Aug 03 20:34:53 flash-kernel-installer (teh bit making the images bootable) wasnt ported to deal with them yet Aug 03 20:35:05 on my list for next week Aug 03 20:36:01 ok so remove uEnv.txt and preEnv.txt and i should be good to go Aug 03 20:38:04 yes Aug 03 20:38:41 in future images these files will be used in the installed system too ... but currently it still needs boot.scr Aug 03 20:39:27 future = next week ...when i'm back in my normal timezone :) Aug 03 21:18:37 janimo: yes, finished, but I'll test it tomorrow Aug 03 21:18:48 problem may be usb related Aug 03 21:41:08 ogra_, removing both the .txt files makes it not boot. is there something i should edit in them? Aug 03 21:41:37 no, was the boot.scr there ? Aug 03 21:41:45 yes Aug 03 21:41:50 also how does it not boot :) Aug 03 21:41:58 (symptoms) Aug 03 21:42:17 just one led flashing blank screen Aug 03 21:43:00 give it some time, there is a bug with the splash that makes it just show a black screen until the login manager comes up Aug 03 21:43:14 ok will do Aug 03 21:43:24 keep an eye on the usb disk light ... you should see activity there Aug 03 21:59:42 my apologies ogra_ there is no boot.scr. i was wrong. Aug 03 22:01:11 oh Aug 03 22:01:33 do you still have the uEnv.txt and preEnv.txt files ? Aug 03 22:01:46 i can get them back Aug 03 22:02:02 if so, copy them back, then edit preEnv.txt: Aug 03 22:02:24 * ogra_ looks up the content you need Aug 03 22:03:40 bootargs=ro quiet splash Aug 03 22:03:44 that should do it Aug 03 22:03:55 * ogra_ just gets kicked out of the office here ... Aug 03 22:04:05 ok give me a minute Aug 03 22:04:12 and the hotel had a power outage the whole day, not sure i can come back online Aug 03 22:04:26 (though i expect that to be fixed now) Aug 03 22:04:37 ok i will be waiting Aug 03 22:19:26 ok that gets me to busybox with a (initramfs) prompt Aug 03 22:22:07 any error message above ? Aug 03 22:23:12 no Aug 03 22:25:52 shall i edit out "quiet" and try it? Aug 03 22:25:58 yep Aug 03 22:26:15 ok just a minute ;) Aug 03 22:29:43 well quite a bit of stuff but no error. last thing is starting udevd and detecting the hub. Aug 03 22:30:00 hmpf Aug 03 22:36:01 you can use the blkid command to find your UUID of the usb disk Aug 03 22:36:11 (blkid /dev/sda) Aug 03 22:36:22 ok Aug 03 22:36:45 then you can add root=UUID= and then the UUID the blkid command returned ... that should get you going Aug 03 22:36:52 not sure what happened there Aug 03 22:38:06 (That should probably be /dev/sda2 or something, not /dev/sda) Aug 03 22:38:59 ok i will try that infinity i am not getting a UUID with sda Aug 03 22:39:45 its the ext2 partion's UUID i want? Aug 03 22:40:17 what infinity said Aug 03 22:40:24 you want the ext4 Aug 03 22:40:43 ok no problem Aug 03 22:42:06 marvin24, built your for-next tree, does not boot though Aug 03 22:42:13 will try again next week :) Aug 03 22:50:47 ogra_, infinity , thank you very much guys it works!! Aug 03 22:55:50 infinity, has anything moved forward with Ti SGX armhf omap3 Aug 03 22:57:41 cvanvliet: I've not been keeping an eye on it personally, I'm assuming people who care will let me know if there's anything to be done. :P Aug 03 22:59:26 fair enough Aug 03 23:03:33 they are available, but I don’t know how official they are **** ENDING LOGGING AT Sat Aug 04 02:59:58 2012