**** BEGIN LOGGING AT Thu Jul 14 02:59:57 2011 Jul 14 04:59:56 04 or 11.10 work with the arm archictecture here's the link for the dreamplug: http://www.globalscaletechnologies.com/c-5-dreamplugs.aspx Jul 14 05:00:03 oops Jul 14 05:01:39 i just purchased dreamplug while it has ubuntu downloads my question is will 11.04 or 11.10 work with the arm archictecture here's the link for the dreamplug: http://www.globalscaletechnologies.com/c-5-dreamplugs.aspx Jul 14 05:04:56 my fault marvell kirkwood cpu Jul 14 05:09:52 anybody here? Jul 14 05:43:18 lintux, Hey. Ubuntu won't run on that device (although Marvell ships it with their own remix of Ubuntu) Jul 14 05:43:55 http://dreamplug.googlecode.com/files/Dreamplug%20-%20Change%20OS%20from%20%20Debian%20to%20Ubuntu-20110615.1.pdf has instructions Jul 14 05:44:25 Hrm. Timing :( Jul 14 05:48:39 persia: bad timing there Jul 14 05:49:06 Indeed. Jul 14 06:49:07 Hi Jul 14 06:52:44 Whether any netinstall option is avilable for ubuntu-arm Jul 14 06:53:19 I mean atleast for building RootFs Jul 14 06:58:23 janimo`, you there Jul 14 07:05:09 siji, yes Jul 14 07:05:43 hi Jul 14 07:06:02 I gone through your doc Jul 14 07:06:09 regd.devlopers meet Jul 14 07:06:40 janimo`, here am working with Beagleboard and ubuntu-arm Jul 14 07:07:36 And looking for a small ubuntu footprint version Jul 14 07:07:43 (with minimal things) Jul 14 07:08:47 So I think netinstall for arm will be good option Jul 14 07:09:10 siji, I have never used netinstall so cannot comment on this :) Jul 14 07:09:33 ok,now netinstall is not there for ARM Jul 14 07:10:22 So why cant you add it in the developers agenda ? Jul 14 07:32:37 siji, the session is about package build failures and aimed for those who want to help fix them. It is not about installing Ubuntu on ARM hw Jul 14 07:33:29 janimo`, ok Jul 14 07:34:09 siji, If you're looking for stable, small-footprint, check out the "headless" images. Jul 14 07:34:50 persia, ok Jul 14 07:35:09 http://cdimage.ubuntu.com/releases/natty/release/ Jul 14 07:35:17 For the beagleboard, you want the "OMAP3" image. Jul 14 07:35:35 persia, i have tried those Jul 14 07:35:53 The headless one? It wasn't small enough for you? Jul 14 07:38:09 Netinstall is only available for oneiric (in-development), at http://ports.ubuntu.com/dists/oneiric/main/installer-armel/current/images/omap/netboot/ Jul 14 07:38:30 But it may or may not work, depending on the current state of the archive. Jul 14 07:40:07 persia, ok Jul 14 07:40:31 let me try that Jul 14 07:44:08 Thanks for helping test :) Jul 14 07:45:51 persia, actually am in the process of deciding OS for one of our Device Jul 14 07:48:18 siji, What sort of device? Jul 14 07:49:35 An ubuntu based tablet Jul 14 07:50:19 *without android :) Jul 14 07:50:50 Oh, cool! That sounds kind of exciting. Jul 14 07:51:58 and i strongly prefer ubuntu Jul 14 07:52:35 Cool! Jul 14 10:56:41 persia: same pb, but it works with the one just before (2.21.51.20110421-6ubuntu1) Jul 14 11:22:29 persia: reported it at https://bugs.launchpad.net/ubuntu/+source/binutils/+bug/810402 Jul 14 11:22:30 Ubuntu bug 810402 in binutils "all native ocaml programs segfault on armel" [Undecided,New] Jul 14 14:33:47 persia, my protorype is running with 11.04 now Jul 14 14:34:04 Now try to enable touch screen Jul 14 14:34:09 8trying Jul 14 14:34:12 *trying Jul 14 15:07:32 hello, installing ubuntu-omap4-extras from trunk on ubuntu10.10 i get this error Error! Bad return status for module build on kernel: 2.6.35-903-omap4 (armel) Jul 14 15:07:32 Consult the make.log in the build directory Jul 14 15:07:33 /var/lib/dkms/pvr-omap4-kernel/1.7~git0f0b25f/build/ for more information. Jul 14 15:07:33 Traceback (most recent call last): Jul 14 15:07:33 File "/usr/share/apport/package-hooks/dkms.py", line 57, in Jul 14 15:07:34 report.write(open(apport.fileutils.make_report_path(report), 'w')) Jul 14 15:13:10 mahmoh: GrueMaster: next u-boot upload should have the unique mac address fixed on panda Jul 14 15:13:13 thanks to jcrigby Jul 14 15:13:19 still building Jul 14 15:13:32 Yes, I saw the email. Thanks. Jul 14 15:13:33 awesome! Jul 14 15:14:20 * mahmoh now has to convince jcrigby to make pxe a default (and /me needs to also investigate a boot failure by the likes of "test" with that setup) Jul 14 15:19:15 hey folks.. i just got wlan working on my pandaboard, got a quick q about wpa_supplicant binary vs functions.sh versions... Jul 14 15:19:35 basically /etc/wpa_supplicant/functions.sh includes a "-s" parameter when starting the wpa_supplicant daemon Jul 14 15:19:49 but this isn't supported by the wpa_supplicant that comes with ubuntu natty Jul 14 15:20:01 is this a bug? something wrong on my end? Jul 14 15:40:58 NCommander, oh, in case you didnt notice, persia provided a patch for netboot that now also publishes the MLO and u-boot.bin files ... so we have publically available unpacked binaries for both Jul 14 15:43:00 lool, so i'm just looking into adding the postinst/trigger bits to flash-kernel, whats the reason you focus on the ABI only in your original patch, shouldnt just the latest existing initrd suffice to get the version without splitting out the ABI Jul 14 15:43:12 Still need to post non-uImage versions of the kernel & initrd (or create an abootimg netboot.img) for usbboot. Jul 14 15:43:49 GrueMaster, meh, you should have said that when persia wrote the patch ... its just two more lines to the two he added :) Jul 14 15:44:11 I noted it back at the rally. Jul 14 15:44:20 bug ? Jul 14 15:44:26 NCommander was working on the d-i at the time. Jul 14 15:44:30 ah Jul 14 15:44:51 No one told me persia was working on it. Jul 14 15:45:06 he fixed the other issue a few days ago Jul 14 15:45:21 spontaneous as response to a user request Jul 14 15:46:06 but it was discussed largely in here, probably at a time where you were asleep though Jul 14 15:47:14 bug #808810 Jul 14 15:47:15 Launchpad bug 808810 in debian-installer "Please publish MLO and u-boot.bin on download page for OMAP netinst" [Wishlist,Fix released] https://launchpad.net/bugs/808810 Jul 14 15:47:20 I'm guessing that the u-boot that is there is already out of date, given the fix for die-id based mac in pxe. Jul 14 15:47:40 well, there was an upload today iirc Jul 14 15:48:29 - Generated unique usbethaddr (LP: #809015) Jul 14 15:48:48 i guess thats the one Jul 14 15:49:58 yep Jul 14 15:50:44 so next d-i rebuild then Jul 14 16:08:00 GrueMaster, any hint about this? http://dpaste.com/568564/ Jul 14 16:08:20 * GrueMaster looks Jul 14 16:09:25 I haven't seen that before, but I haven't looked at maverick in quite a while. Jul 14 16:09:52 http://dpaste.com/568555/ this is the make.log Jul 14 16:11:08 It looks almost like the maverick ppa got clobbered by the natty ppa release. Jul 14 16:11:09 robclarck (from ti) sent me here saying Jul 14 16:11:10 hmm.. you might want to ask on #ubuntu-arm ... I think there was some PM related API that changed in the kernel.. so when it upgrades the kernel, it tried to rebuild the old pvr kernel module, which failed.. Jul 14 16:11:10 I don't remember how to get around that.. I think perhaps remove the sgx packages, then update, then re-install them Jul 14 16:11:15 but makes no sense to me Jul 14 16:11:46 I'll see if I can reproduce. May take a while though. Jul 14 16:12:02 understandable Jul 14 16:17:55 Linux panda 2.6.35-903-omap4 #22-Ubuntu SMP PREEMPT Mon Mar 28 17:25:09 UTC 2011 armv7l GNU/Linux Jul 14 16:18:09 that is not the trunk update.. which should be 2.6.35-980 Jul 14 16:18:35 but the trunk driver ? Jul 14 16:18:52 mmmh Jul 14 16:19:01 so Jul 14 16:19:12 you are saying that i'm not using the up-to-date kernel Jul 14 16:19:23 you are using the release kernel Jul 14 16:19:39 that's strange Jul 14 16:19:40 yeah.. so I guess you end up w/ release kernel, but trunk userspace syslink libs and trunk firmware Jul 14 16:19:43 but not the release PPA for pvr Jul 14 16:19:57 ok, in which package i have te kernel? Jul 14 16:20:15 I seem to remember you need to uninstall pvr/sgx stuff to update.. otherwise the upgrade gets blocked by dkms compile error Jul 14 16:20:16 linx-omap4 should be the metapackage pulling in everything Jul 14 16:20:38 oh, that might be, could be that dkms is blocking Jul 14 16:20:48 uhm uhm Jul 14 16:21:01 dpkg drives me mad Jul 14 16:21:05 it is a bit ugly, but if you uninstall sgx/pvr stuff first, then upgrade, then re-install Jul 14 16:21:05 It may not have updated the boot partition. We have a bug filed about that. Jul 14 16:21:16 i've already tryied to uninstall sgx/pvr Jul 14 16:21:17 It isn't really dpkg's fault.. Jul 14 16:21:19 and upgrade Jul 14 16:21:23 Run "sudo flash-kernel" and it should update properly. Jul 14 16:21:24 as you said Jul 14 16:21:30 GrueMaster, indeed ! Jul 14 16:21:35 problem is kernel API that changed, so old pvr kernel module won't compile w/ new kernel headers and visa versa :-( Jul 14 16:21:38 why didnt i have that idea ! Jul 14 16:21:38 but no newer package were pulled Jul 14 16:21:44 also, i tryied to run flash-kernel Jul 14 16:21:48 * ogra_ is hacking on exactly that code atm :) Jul 14 16:22:59 phas@panda:~$ aptitude show linux-omap4 Jul 14 16:23:00 Package: linux-omap4 Jul 14 16:23:08 ... Jul 14 16:23:09 Version: 2.6.35.903.6 Jul 14 16:23:10 ugh, aptotude Jul 14 16:23:21 what's wrong with aptitude? Jul 14 16:23:24 thats just the meta Jul 14 16:23:34 check for the linux-image package Jul 14 16:24:06 Version: 2.6.35.903.6 Jul 14 16:24:20 (aptitude show linux-image-omap4 ...0 Jul 14 16:24:41 LPhas: Make sure you have maverick-updates in /etc/apt/sources.list Jul 14 16:25:02 iirc, that was also an issue with maverick. Jul 14 16:25:14 GrueMaster, yep, they seems to be there Jul 14 16:25:19 dpkg -l|grep linux-image Jul 14 16:25:22 try that Jul 14 16:25:34 phas@panda:~$ dpkg -l | grep linux-image Jul 14 16:25:34 ii linux-image-2.6.35-903-omap4 2.6.35-903.22 Linux kernel image for version 2.6.35 on TI OMAP4-based systems Jul 14 16:25:34 ii linux-image-omap4 2.6.35.903.6 Linux kernel image for the OMAP4 architecture. Jul 14 16:25:42 k, only 903 Jul 14 16:26:08 http://dpaste.com/568569/ sources.list Jul 14 16:26:50 hmm, how did you add the PPA Jul 14 16:27:06 that looks like manually hacked in, usually ppas end up in sources.list.d Jul 14 16:27:22 yep, i put them in manually Jul 14 16:27:23 (if you use the right tools) Jul 14 16:27:52 well, better follow the instructions next time :) Jul 14 16:28:03 wont do harm though Jul 14 16:28:32 its intresting that you dont get the new kernel from the trunkl ppa Jul 14 16:28:54 robclark, did you not provide a linux-meta package too in there ? Jul 14 16:29:17 couldn't i simply download the deb file and dpkg -i into it? Jul 14 16:29:33 well, there is a bug we'd like to identify :) Jul 14 16:30:31 i get your point Jul 14 16:30:54 so there is no -meta package Jul 14 16:31:00 thats the issue Jul 14 16:31:59 sudo apt-get install linux-image-2.6.35-980-omap4 linux-headers-2.6.35-980 Jul 14 16:32:02 try that one Jul 14 16:35:35 installing Jul 14 16:37:03 ogra_, no idea.. I wasn't really too involved in the deb packaging stuff Jul 14 16:37:25 well, thats definitely the blocker Jul 14 16:38:00 so i won "i found a bug" t-shirt? Jul 14 16:38:10 though its possible that one of the ubuntu-omap4-*-* packages would pull it in Jul 14 16:47:45 followup from #pandaboard.... Jul 14 16:48:08 LPhas: plz file a bug against the TI natty PPA Jul 14 16:48:28 will ping the folks who maintain it to see if they can address it Jul 14 16:49:14 uhm Jul 14 16:49:26 this is maverick not natty Jul 14 16:50:04 and i'm not 100% aware of what happened, not sure at least. Jul 14 16:50:33 basically the meta package of linux-omap4 was not updated to include newest linux kernel? Jul 14 16:51:26 no Jul 14 16:51:46 oops yeah maverick Jul 14 16:52:00 there was no metapackage created at all Jul 14 16:52:15 sob, now i've x not starting Jul 14 16:52:21 the existing one in the archive cant depend on a linux-image outside of the archive Jul 14 16:52:45 so someone would have had to create it inside the PPA with a higher version than the archive has Jul 14 16:52:54 which didnt happen Jul 14 16:53:23 FATAL: Module omap_gpu not found. Jul 14 16:53:28 that's strange Jul 14 16:53:42 oh, well, i probably got why Jul 14 17:16:39 mmh i keep getting omap_gpu not found on startx Jul 14 17:17:07 i removed and reinstalled ubuntu_omap4_extra thinking that previous installation compiled modules for the old kernel Jul 14 17:17:17 but nothing, it keeps not working Jul 14 17:25:16 LPhas: which ppa are you trying to use at maverick? Jul 14 17:25:22 release or trunk? Jul 14 17:25:25 trunk Jul 14 17:25:28 i have to use it Jul 14 17:25:41 so you should be using the linux-kernel from trunk Jul 14 17:25:42 realease worked fine Jul 14 17:25:46 i'am Jul 14 17:25:52 if you look at the history Jul 14 17:25:55 we found a bug Jul 14 17:26:15 that prevent me to automatically get the latest version of the kernel from trunk Jul 14 17:26:38 LetoThe2nd: 2.6.35-980.1release9? Jul 14 17:26:46 LetoThe2nd: sorry Jul 14 17:26:51 LPhas: 2.6.35-980.1release9? Jul 14 17:26:54 oh, ok Jul 14 17:26:57 yep Jul 14 17:27:08 sorry, huge backlog Jul 14 17:27:45 i noticed that linux-headers-2.7.35-980-omap4 was not installed Jul 14 17:27:45 Interesting. Fresh install of maverick, I am getting dependency failures trying to install ubuntu-omap4-extras. Jul 14 17:28:07 GrueMaster: ogra_: LPhas: http://www.omappedia.org/wiki/Ubuntu_OMAP_trunk Jul 14 17:28:11 GrueMaster, interesting Jul 14 17:28:15 that's why meta wasn't included I believe Jul 14 17:28:24 need to update x-loader/u-boot to make it work Jul 14 17:29:02 rsalveti, the "update the bootloader" part seems quite overcomplex Jul 14 17:29:17 shouldn't just sudo flash-kernel do the trick Jul 14 17:29:32 oh, it is intended Jul 14 17:29:34 LPhas: not for maverick Jul 14 17:29:36 oh my Jul 14 17:31:51 Ah, forgot that multiverse & universe were not added by default. yet another bug (and I now remember filing it back in the day). Jul 14 17:32:45 well, that is written on a lot of documents online Jul 14 17:53:17 * GrueMaster signs his life away to the TI liscense agreements in Maverick for the sake of testing. Jul 14 18:13:31 jeremiah, ping Jul 14 18:17:49 GrueMaster, you read them before clicking, right ?? especially the part talking about your first born i hope :) Jul 14 18:18:13 Bah. They can have him Jul 14 18:18:34 * GrueMaster is more concerned with the anti-cloning on Titan 6 clause. Jul 14 18:19:15 i dont think it says they want him :) Jul 14 18:19:45 i might misremember, but i think there was something abotu life long vacancy *g* Jul 14 18:44:17 So, about exposing netboot stuff: I tossed in those lines as a whim. The linked branch should make the process fairly obvious if anyone wants to include more files. Jul 14 18:44:54 I don't know what needs doing with abootimg to generate the right files for USB: if someone *really* wants to send me code to stick in the file, rather than writing a patch, that works. Jul 14 18:45:16 I'm also happy to review someone's patch adding abootimg code for syntax, integration, etc. if someone wants. Jul 14 18:45:30 (although I can't commit, so it would be a "looks sane" class of review) Jul 14 18:46:56 sgnb, Thanks for filing a bug about it. Looks like useful progress is happening there. Jul 14 18:48:49 ogra_, regarding the image list: it's not *my* list (if anyone's it's skaet's), and it has mx5 images for Server starting today and for Xubuntu Desktop starting on Tuesday, so not everything is supposed to wait until 0801 Jul 14 18:50:45 * micahg notices a mention of xubuntu ARM images... Jul 14 18:51:40 micahg, Yeah. When 0705 passed with no images in site, I bumped the date to 0719, because I figured two weeks was a nice round number, and someone else had set server/mx5 to 0715 Jul 14 18:52:20 From my memory of our last conversation about it, you didn't really care about the specific date, as long as there was something to test sufficiently before alpha-3 that you had a chance to fix obvious problems. Jul 14 18:53:11 persia: sounds good, now I just have to take care of my other stuff so I can actually play with this image :) Jul 14 18:53:54 sgnb, Apologies: I am having a bit of trouble following the discussion in bug #810402 : Does the 4.5 -> 4.6 change mean we need *another* ocaml transition for ABI shift? Is there something else happening? Jul 14 18:53:55 Launchpad bug 810402 in ocaml "all native ocaml programs segfault on armel" [High,Confirmed] https://launchpad.net/bugs/810402 Jul 14 18:55:03 note that there was an ocaml rebuild upload today Jul 14 18:55:21 which apparently ftbfs on all arches (i havent looked at the logs) Jul 14 18:55:56 Yeah, doko uploaded ocaml to build with a different GCC. Jul 14 18:56:05 It didn't work. Jul 14 18:56:58 * persia grumbles at the institutionalilsation of *yet another* resolution: "qHD", with a vertical component less than 600 pixels (and also less than the 576 pixels we chased last time someone had this sort of brilliant idea) Jul 14 18:58:31 well, but you should at least get back horizontally what you lose vertically :) Jul 14 18:58:41 But the discussion in 810402 seems to be missing some bits, or at least it doesn't make sense to me as a conversation. Jul 14 18:58:42 why else would they add a q in front :P Jul 14 18:58:52 'q' is for "Quarter". Jul 14 18:58:53 must be something special :) Jul 14 18:59:03 lol Jul 14 18:59:13 19020x1080, except only 1/4 of it, so 540 vertical. Jul 14 18:59:15 so not for quality ? Jul 14 18:59:22 960 horizontal is fine. Jul 14 18:59:27 sure Jul 14 18:59:35 960x540? better than iPhone4 resolution! :) Jul 14 18:59:38 well, thats close to PAL Jul 14 19:01:05 Still means that we either have to squash software or declare the resolution unsupported as a primary display Jul 14 19:03:43 i would go for the latter Jul 14 19:03:59 how many devices that fit a formfactor we target are there ? Jul 14 19:04:11 (with such resolution) Jul 14 19:07:55 janimo: I know where to find you when I do have porting questions, thanks :) Jul 14 19:08:41 micahg, ok ;) Jul 14 19:09:15 * janimo crosses fingers it is not chromium :) Jul 14 19:10:18 ogra_, An increasing number, since the resolution has a special name, and glass is available, rather than being considered weird. Jul 14 19:43:28 I'm trying to run pts/java-scimark2 and all the tests fail, exit non-zero, so it's not working for sure. Should I try the same tests with JemVM? (if that's correct?) Jul 14 19:44:50 mahmoh: It would be interesting to look at the failure log to see why it fails. I have openjdk-jre-headless running jenkins slave on panda. Jul 14 19:45:45 oh, and a crash in netboot too: http://paste.ubuntu.com/644324/ ( cmagina ) Jul 14 19:45:59 GrueMaster: good point, let me try that Jul 14 19:54:19 GrueMaster: it wasn't installed correctly b/c unzip wasn't installed! Is that a package bug I would guess? Jul 14 19:55:22 dependency issue. Should have resolved on it's own, but phoronix-test-suite relies on aptitude to install (which isn't part of the standard image). Jul 14 19:56:17 it's not? It's installed on my image (netboot w/ basic server), unzip was not Jul 14 19:56:25 I have seen a lot of dependency issues with phoronix. Jul 14 19:57:36 My netboot install with basic server & openssh-server didn't have aptitude. Not sure what else was missing. Had to install build-essential for the io tests as well. Jul 14 19:58:03 Possibly due to the debootstrap issues. Jul 14 19:58:32 GrueMaster: hm, build-essential was installed for me by pts, aptitude's on there, wonder why our installs are different, itneresting Jul 14 19:58:38 interesting Jul 14 19:59:13 I installed fresh Tuesday, right in the middle of the bootstrap issues. Jul 14 20:00:04 ok, I'm installing now on another board so I'll check Jul 14 20:00:24 actually, the install crashed (noted above) Jul 14 20:00:35 bbib Jul 14 20:07:40 and if anyone is nice enough to take a look at this and tell me why it's failing I'd appreciate it: http://paste.ubuntu.com/644336/ Jul 14 20:12:19 mahmoh: Install libaio-dev Jul 14 20:18:00 GrueMaster: awesome, thx Jul 14 20:19:05 oh, it's installed already Jul 14 20:39:54 Hy. I did this : http://seabright.co.nz/2011/03/29/building-the-ubuntu-pandaboard-kernel/ but it didn't seem to do anything. Where should the uImage be? Jul 14 20:40:13 arch/arm/boot Jul 14 20:43:33 it only compilse some object files but makes no linking... is this right? Jul 14 20:45:57 It might be I'm getting just to sleepy. I'll continue tomorrow. Jul 14 20:46:12 no that's not right Jul 14 20:46:51 thsi page assumes you've got an arm crosscompiler Jul 14 20:47:41 I was doing it on arm not on crosscompiler Jul 14 20:48:40 Okey... thanks. Jul 14 20:48:41 ok so you don't have CROSS_COMPILE= line ? Jul 14 20:48:57 in makefile? Jul 14 20:49:03 ... Jul 14 20:49:17 in the page you gave, it says to do export CROSS_COMPILE=blabla Jul 14 20:49:44 no Jul 14 20:49:52 i didn't need to.. Jul 14 20:50:02 i didn't crosscompile. Jul 14 20:50:14 I was running it on panda. Jul 14 20:50:59 so you didn't do what this page states like you first said ... Jul 14 20:51:52 eee... yes (angel).... i did allmost like it told Jul 14 20:52:35 anyway, what did the make say ? Jul 14 20:53:28 It just made lot of .o and finaly one ld. It didn't link. Jul 14 20:53:44 ld isn't a link tool ? Jul 14 20:53:45 interesting Jul 14 20:54:05 Only on ld. Jul 14 20:54:08 I did something wrong. I'll go to sleep... won't anoy you any more. Jul 14 20:54:13 Thnx. Jul 14 20:54:23 if you did something wrong there are errors in make log ... Jul 14 20:55:03 I'll come back with log tomorrow... there was no errors. Jul 14 20:55:19 Tnhx. Bb. Jul 14 23:48:36 Could anyone point me to nice documentation on boot.scr? I can't seem to find anything :( Jul 14 23:57:13 persia, just something generic or? (btw, if this is for the beagle, upstream has moved to uEnv.txt's instead, clear text..) Jul 14 23:57:46 Not for the beagle, and for a vendor uboot, so not even something where I can happily leverage upstream, sadly. Jul 14 23:58:33 But generic works. I basically want to understand conditionals, how program flow works, and how variables are used. Jul 14 23:59:08 ah, variables get fun.. they end up being 'true' if they are not defined.. Jul 14 23:59:25 In the end, I probably just want to setenv ramdisk/kernel/bootargs and bootm, but when digging through the pile of examples I have, it helps to have docs alongside to make sure I understand. Jul 14 23:59:30 Oh my :) Jul 14 23:59:38 kinda decent example here: https://github.com/RobertCNelson/flash-omap/blob/master/reset.cmd Jul 15 00:00:03 if i boot that on a pandaboard, where "beaglerev" is not defiend it'll drop into every if.. Jul 15 00:00:48 Wait, not only is the variable true if undefined, but the tests are automatically true if the variables are undefined? Jul 15 00:01:12 yeah, that's what i ran into, it was annoying.. Jul 15 00:01:44 so you have to be pretty careful with the end target.. Jul 15 00:01:46 In one of the examples I'm looking at, there's a construction that might work around that: "if test "_${foo}" = "_"; then:" Jul 15 00:02:24 (explicitly tests for null state, but presumably the same sort of thing could be extended to specifically testing a known state) Jul 15 00:02:26 actually, yeah that should work, didn't think of that.. Jul 15 00:03:01 Heh, and that's why sharing is good :) Jul 15 00:03:24 Mind you, having documentation that mentioned the flaw in the implementation of test and the workaround would be even better :) Jul 15 00:10:19 rcn-ee, Looking at your example, is that an entire menuing system for the flash on the beagles? Jul 15 00:11:02 the idea behind that was to stop people from accidently flashing nand the xM's.. Jul 15 00:11:33 there's no nand of course, but i got a couple bug reports with interesting errors... ;) Jul 15 00:11:51 Heh. Jul 15 00:12:30 I was wondering: would it be worth replacing the skeleton system we ship in Ubuntu with something like that (slightly more robust)? Jul 15 00:14:54 i've actually been working on something even crazier... autogenerate it at boot: https://github.com/RobertCNelson/netinstall/blob/master/mk_mmc.sh#L127 then with sed: https://github.com/RobertCNelson/netinstall/blob/master/mk_mmc.sh#L184 (so far the same script boots on imx53,beagle,panda,igevp2.. so not as many as you guys yet..) Jul 15 00:15:04 /boot/image flash time.. Jul 15 00:20:09 Wow! So this would be some generic framework that on first boot would (ideally) generate the perfect boot.scr for the target platform, and place that in the right place for the next boot? Jul 15 00:21:34 So, there's two points of wonder I have remaining: Jul 15 00:22:02 1) If we put several kernels in some directory available to u-boot, would this be able to select the right one? Jul 15 00:22:35 2) Would this allow us to construct a u-boot that was capable of booting on all those devices, or would there still need to be a separate compilation of u-boot for each image? Jul 15 00:22:58 i was actually hacking up rootstock to do the same thing.. install both the omap and imx image's then select the right one by which board was selected.. Jul 15 00:23:12 then one image for many boards.. (till device tree is ready) Jul 15 00:23:40 Could I convince you to do that hacking with debian-cd directly? Jul 15 00:24:38 That makes it trivial for us to use it to generate such generic images (as opposed to rootstock, where we'd have to port the code and we have to recommend nobody uses for production purposes) Jul 15 00:24:53 is that the same one ogra,lool.. keeps reminding me to use? ;) do you have the tree? i just finished adding debian armhf support to rootstock this last weekend.. ;) Jul 15 00:25:23 lp:~ubuntu-cdimage/debian-cd/ubuntu Jul 15 00:25:39 It is most likely precisely the same codebase everyone else keeps suggesting :) Jul 15 00:25:45 It's the one we use to build the images we ship. Jul 15 00:30:09 gosh Jul 15 00:30:35 ... ? Jul 15 00:31:34 My eyes are wide open at the shell pieces that rcn linked to, and also by the recommendation to hack it in debian-cd directly :-) Jul 15 00:31:42 in fact I'd recomment the opposite ;-) Jul 15 00:32:05 How do you mean? Jul 15 00:32:22 Have a simple interface to generate useful images in a script which consumes e.g. the name of the vmlinuz file to install and any other useful inputs and then use that from debian-cd Jul 15 00:32:23 I thought that the linked code was boot.scr code, rather than shell code. Jul 15 00:32:26 and other places Jul 15 00:32:39 And that one would just drop it in verbatim in debian-cd Jul 15 00:32:55 Ah, right. I should look at the top of the files. Jul 15 00:33:36 Yeah, having a utility that did the right thing, and having debian-cd call that utility is lots cleaner, if it's image-build-time, rather than first-boot-time that we calculate the boot script. Jul 15 00:34:00 I considered proposing linaro-image-tools to do this Jul 15 00:34:06 That said, I'd *much* rather see a first-boot-time solution. Jul 15 00:34:24 it has the advantage of being python and with decent test coverage, supporting a bunch of boards Jul 15 00:34:46 but I'm not entirely sure it's the right approach Jul 15 00:35:11 I think I should stop here, it's not reasonnable to start such a discussion in my local time Jul 15 00:35:23 Heh. Sleep well. Jul 15 00:35:44 I agree it would be nice to have a tool that did the right thing that we could use on the image builders *and* end user environments. Jul 15 00:36:12 I also feel fairly strongly that we should strive for a sufficiently generic model that we *don't* need to have N different images, all with the same rootfs. Jul 15 00:36:25 (or nearly the same: ignore things like /lib/modules for now) Jul 15 00:36:30 Yes, so there's actually a forum where we discuss exactly this Jul 15 00:36:50 exactly, one generic image is all we need, at some point the user will have to flash something, at that point is where the customization should be done. Jul 15 00:37:00 but to support current SoCs, we will need some middle layer adapting the boot of each SoC to some standard image format for a while, so the question remians Jul 15 00:37:34 rcn-ee, Users make mistakes. I want it to just work, without customisation. Jul 15 00:37:47 rcn-ee: Right now, there are different expectations between SoCs which don't allow this, but we could progressively transition in such a world Jul 15 00:38:11 lool, Indeed. It's messy now, until we have more mature bootloaders, and potentially a new generation of SoCs with ROM that integrates nicely with the richer bootloaders. Jul 15 00:38:21 however, we could have some generic image format right now, and provide mini-images for SoCs we care about which allow loading this image format Jul 15 00:39:03 lool, i was actually working on something silly. (till device tree's done..) have both and imx/omap image installed in the rootfs, then having the bootloader pick the correct one, based on what board the user selected... so you'd have one base iamge for imx and omap.. per say.. Jul 15 00:39:38 rcn-ee: so with Linaro "images" we split the soc specific bits out, and you combine them when writing the SD image Jul 15 00:39:55 This is the function of linaro-image-tools Jul 15 00:40:06 e.g. you download an Ubuntu rootfs, then imx specific bits then omap specific bits and you run linaro-image-tools to create omap or imx SD card images Jul 15 00:40:43 yeah, pretty much... Jul 15 00:41:09 while imx + OMAP might be combined in a single image, it's becoming exponentially complex when you factor more SoCs or things like aligning partitions, different partition types, and other constraints Jul 15 00:41:36 I like better the idea of merged images, letting the user select the right thing, but last I heard, one needed a different build of u-boot for each board (and sometimes each revision of a board), which makes this considerably harder to accomplish. Jul 15 00:41:42 like Samsung expects its bootloader near the imx one, so you wouldn't be able to do imx + samsung support Jul 15 00:43:06 yeah, i haven't expanded enough outside imx/omap yet to see the bigger picture yet.... ;) Jul 15 00:43:16 we didn't discuss detecting which SoC you're on; John Rigby had some idea that this could be done; it's also a fun research project, but it's pretty tricky Jul 15 00:43:51 and that's not needed in the case of interim middle layer images Jul 15 00:44:10 And it's very bootloader-specific. I'm not sure devicetree helps at all, because we still have to find a way to get to linux. Jul 15 00:44:19 just detecting variations of OMAP3 boards to infer their type of DDR is already quite hard to do reliably Jul 15 00:44:24 yeah the real issue here is you guys are *recompiling and reflashing u-boot* which is totally bad mojo and brick city awaits Jul 15 00:44:40 it's not really a problem of which bootloader Jul 15 00:44:48 the hardware doesn't provide the information to start with Jul 15 00:44:58 NekoXP: I don't think this relates at all Jul 15 00:45:13 but I can certainly say that we can't work on any such project if we aren't rebuilding u-boot ourselves Jul 15 00:45:14 NekoXP, We only do that for dev boards. We don't reflash u-boot for retail devices (or reflash u-boot past first install anyway) Jul 15 00:45:25 at least not on this target SoC, someone else will have to do it Jul 15 00:47:13 * lool 'night! & Jul 15 00:47:33 lool, Sleep well Jul 15 02:18:51 at least for omap it should be ok to have just one u-boot at least Jul 15 02:18:59 but still, we need something to provide the device id Jul 15 02:19:06 and that would be SPL or x-loader Jul 15 02:19:19 problem is that there's no pattern at all Jul 15 02:54:35 I have a question about network (Wired & Wireless), today I loaded this image on my board "http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/10.10/release/ubuntu-netbook-10.10-preinstalled-netbook-armel+omap4.img.gz". when it installed, I was at a desktop and cound not enable the wireless or wired network - does anyone know what I am missing? I am a newbe btw **** ENDING LOGGING AT Fri Jul 15 02:59:57 2011