[04:12] hi all, may i ask what is the status of GDB? i always run into SIGILL in dl_main... any idea? [04:13] Try asking in #linaro. They tend to work with those tools far more often, and I believe they may actually be supporting it on arm directly. [04:13] ok thanks! === Jack87|Away is now known as Jack87 [05:08] why cant i find the package that libXxf86misc.so is in?!?! [05:10] try "dpkg-query -S /libXxf86misc.so [05:11] nope [05:11] seems ubuntu chose not to include this library [05:11] Oh, are you looking for it to install? [05:13] X11_Xxf86misc_LIB [05:13] cant divine out the source package name [05:14] May not be in the images anymore. I am not seeing it on oneiric arm. [05:15] And it isn't on my natty amd64 netbook either. [05:15] seems to be depricated. [05:15] so much for libfreenect on ubuntu-arm [05:16] oddly enough on 11.04 ubuntu intel ITS THERE [05:16] stupid caps lock trigger finger [05:17] oh wait [05:17] intel debian had it [05:17] 11.04 was natty. [05:17] 10.04 maybe. [05:17] 11.04 arm ubuntu lacks it [05:17] now my facts feel a little straiter [05:18] there have been source packages for it in ubuntu until 10.10 as far as i can see [05:19] Well, the last build I have for it in armel is 2009-12-03 [05:20] i am on launchpad, and saus its not released for natty [05:20] Ok. I was looking direclty on my mirror, which has everything from 10.04 forward for arm. [05:21] But LP is more diffinative. [05:21] At any rate, I'm running on fumes. 13 hour workday. [05:22] night [05:23] Meh. That's hardly over half time. === Jack87 is now known as Jack87|Away [06:19] cvdemo has a decent frame rate on panda connected to a kinect [06:20] i am not entirely unpleased [07:22] Hi, does anyone know how to change the extraversion variable when creating kernel packages with dpkg-buildpackage? [07:22] I've tried EXTRAVERSION=-leon dpkg-buildpackage -B -uc -us but that didn't work :) [07:22] make-kpkg doesn't work. It complains about some missing commands from a x86 only package [08:47] x-loader-omap4 was deprecated in favour or x-loader - right? [08:49] ogra_: bug 806864 is for you [08:49] Launchpad bug 806864 in x-loader-omap4 "remove from repository" [Undecided,New] https://launchpad.net/bugs/806864 [08:55] hrw: You win a bug closure! ;) [09:05] fun is that I have this message on oneiric [09:06] once it will finish upgrade I will sent apport-collect [09:06] hrw: Which means you have/had the package installed at one point, and now dpkg has gotten grumpy about the stanza... [09:06] hrw: Perhaps worth filing a bug against dpkg for suddenly being irked by version numbers that it used to let slide. [09:07] un x-loader-omap4 (no description available) [09:07] ii x-loader-omap4-panda 1.5.0+git20110325+b6bbfe7-1ubuntu1 Board initialization helper for TI OMAP 4 Panda boards [09:07] hrw: Yes, note the "had"... The stanza is in there because you used to have it installed, I assume. [09:07] added dpkg to bug [09:07] hrw: Or you have maverick in your sources.list or something... [09:08] hrw: (apt-cache policy x-loader-omap4) [09:08] none [09:08] N: Unable to locate package x-loader-omap4 [09:10] Kay, then it's just stale info in the available DB from having it previously installed. [09:11] added data to bug [09:11] You can fix *your* issue with "dpkg --forget-old-unavail" probably. [09:11] But the part where dpkg suddenly got stricter is still a bug/misfeature. [09:13] will leave it like it is for now [09:13] infinity, hmm, do you knnow if GrueMaster re-tested omap3 images ? [09:13] ogra_: What's to re-test, we didn't change them, did we? [09:14] (the u-boort bug is supposed to be fixed, the iso tracker still lists it) [09:14] Oh. [09:14] we respun all images yesterday [09:14] I'm not sure what the status of omap3 stuff is, to be fair. You'll have to ask him in the morning. [09:14] i see he tested all omap4 and omap3 wont have a fixed kernel [09:14] i was just curious [09:15] Speaking of images, though. We still have resizing issues on desktop, apparently. [09:16] Fails to actually resize, explodes with ext3 errors. [09:16] Yet server works fine. [09:16] (Again, this is all second-hand from GrueMaster, so you might want to follow up with him) [09:17] yeah, we'll skip desktop for A2 i think [09:18] that will likely need some bigger changes, alonngside with the live-build ones [09:18] Fun, fun. [09:18] looks to me like the underlying filesystem isnt really resize friendly ... alongside with bugs in jasper [09:19] Well, I'll mangle live-build tomorrow (obviously not uploading until after A2), and pass it by you for a quick review. [09:20] I need to get involved in live-build upstream, I think. There are things I need to do to it that won't be pretty without upstream buy-in. [09:20] But first, I need sleep. [09:20] yeah, go to bed [09:21] Yes mom. [09:21] :) [09:23] hrw: Oh, wait. I just realised that issue you were seeing was a warning, not an error. [09:24] * Do not allow versions starting with non-digit when doing strict parsing, [09:24] warn otherwise. [09:24] hrw: Probably not really bug-worthy at all, just annoying that someone (glances at ogra_) used a policy-breaking version number in the past. :P [09:25] ?? [09:25] from where is that ? [09:26] ogra_: In maverick, you shipped an x-loader-omap4 that had an "L" as the first character in the version. [09:27] uh, yeah upstream ... [09:27] ogra_: New dpkg is stricter about policy compliance and now vomits warnings about that being in the status file. [09:27] hrm ... [09:27] (Traditionally, when upstreams have non-policy-compliant versions, we mangle them, like 0.L4-whatever, etc) [09:27] Anyhow, can't go back in time and fix it. [09:28] nope, sadly ... that stain stick on me now [09:28] And the number of users who will be upgrading from maverick on omap4 systems can probably be counted without removing my socks. [09:28] *sticks [09:28] oh, we might have more then ten that do it [09:28] not by much though i belive [09:28] Possibly, but I wouldn't say it's a large number. :P [09:29] I don't doubt there were a chunk of maverick omap4 users, but the tester/devel types are probably reinstalling every 3 minutes, and the "normal users" (all 4 of them) are more likely to just re-flash new images than upgrade, on these sorts of devices. [09:30] So. Meh. Whatever. [09:30] yep [09:31] I think I'll just close this dpkg bug then. [09:32] hrw: You can hit me later for having marked two tasks in a row as "invalid" on this bug. :P [09:32] hrw: But at least I left a friendly comment! [09:33] i wonder why lintian didnt complian, i never roll new packages without checking, i think i would have noticed such an error ... [09:34] Dunno. Seems odd that it wouldn't, since this has been in policy for over a decade. [09:34] Oh, but it's a "should", not a "must". [09:34] aww [09:34] Hence why dpkg is warning, not erroring, I guess. :P [09:34] Small mercies. [09:35] infinity: thx [09:36] hrw@panda:~$ LC_ALL=C dpkg --forget-old-unavail [09:36] dpkg: warning: obsolete '--forget-old-unavail' option, unavailable packages are automatically cleaned up. [09:36] ;D [09:36] I had to switch to English translation cause I was unable to understand Polish one ;D [09:38] Automatically cleaned up, eh? Seems not, in your case. :P [09:44] ;) [09:48] hrw: --clear-avail probably still works. :P [09:49] yes, it does [09:50] For entertainment value, there's a huge policy-versus-dpkg debate in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=620566 [09:50] Debian bug 620566 in dpkg "dpkg: "version number does not start with digit" is in" [Normal,Open] [09:56] hi all [09:57] iam running ubuntu 10.04 on my pandaboard [09:57] now i builded a kernel with xenomai [09:58] i replaced uImage of previously running kernel which i got from preinstalled images [09:58] but it giving me some errors [09:59] Is there any documentation to built the ubuntu filesystem for custom builded kernels [10:02] update-initramfs -c -kVER-OF-YOUR-KERNEL? [10:04] the error its giving me is unable to load modules [10:11] Did you install your modules after you compiled the kernel? [10:16] also copying the config to /boot/config-$version helps [10:16] iirc update-initramfs checks for that too [10:17] yes i installed my modules [10:18] update-initramfs -c -k 2.6.37.6 [10:19] you could add -v then you see what exactly it tries to do ... [10:19] might reveal you more info [10:35] ogra_: on the transformer I do not have the same issue with size [10:36] size ? [10:36] ogra_: my partition is 8MB flat, but can be extended if needed [10:36] ah, then you have a more friendly bootloader :) [10:36] the ac100 one can be changed as well [10:36] heh :) [10:36] in fact NCommander did that for many models areound we installed [10:37] well I am compressing the initramfs LZMA to fit [10:37] but its not really an enduser task to hack around in the bootloader [10:37] well resizing can be done with nvflash [10:37] my size probelm is a problem with this specific kernel i think [10:37] ah [10:38] ogra_: you seen it running on the transformer yet? [10:38] yep [10:38] pretty slow though [10:41] yeah [10:41] need to start on some optimizations [10:41] fb is the biggest issue [10:41] up to 70% usage on one core at a time [10:41] at any one time* [10:54] you could try the nvidia driver ;) [10:54] https://launchpad.net/~ac100/+archive/ppa/ [10:54] should work on the transformer too [10:55] GLES is still a problem though [10:58] I get a black screen [10:58] with that driver ? [10:58] and I still have no wifi :) [10:58] the nvidia driver [10:58] got an Xorg log ? [10:58] injected using their script [10:59] nothing on it [10:59] no idea what their script does or if it hooks into the right places in the system [11:01] well it is done through an xorg.conf in /etc/X11 [11:01] ?? [11:01] I still have everything in my fs, minus that config file [11:01] you mean the script puts that in place [11:01] yeah [11:01] right [11:02] well, it needs to mangle the ldconfig configuration, and should set the right alternative for the driver [11:02] ah [11:02] I will try through the PPA then [11:02] it should also put udev rules in place to get you the proper acls for the kernel devices the driver needs to access [11:02] still need to figure out this fing wifi though [11:03] well, there are several broadcom drivers you should be able to try [11:05] I don't know if that will help [11:08] my driver is identical to the xoom one apart from like 6 lines (wakelock defs) [11:09] the xoom one works, mine doesnt [11:13] ogra_, how is the new kernle [11:14] apw, all fine, thanks :) [11:15] update-initramfs -c -k [2.6.37.6] update-initramfs: No match. [11:15] update -initramfs -c -k -v [11:16] []2.6.37.6 ] [11:16] even that also gicing the same error [11:16] no match [11:17] i have moved config and system map files to boot [11:23] why did you put the version in brackets ? [11:28] when i run update-initramfs -c -k [11:29] it showing me options [11:29] test06:/media/disk-3/etc# update-initramfs -c -k /usr/sbin/update-initramfs: option requires an argument -- k Usage: /usr/sbin/update-initramfs [OPTION]... Options: -k [version] Specify kernel version or 'all' -c Create a new initramfs -u Update an existing initramfs -d Remove an existing initramfs -t Take over a custom initramfs with this one -b Set alternate bo [11:30] stm__: yeah replace [version] with the version not put it in the brackets :) [11:30] update-initramfs -c -k 2.6.37.6 [11:31] thanks [11:32] is that the only change i have do [11:32] inorder to boot withe the new uImage [11:37] ogra_: do you have sound working on the AC100 [11:38] no [11:38] not yet [11:38] there is active work going on though [11:38] ok :) [11:38] and i gave a device to one of the ubuntu alsa guys [11:38] but he is on paternity leave for a few weeks now [11:39] ah [11:40] others in #ac100 are hacking on it though [11:41] I can get aplay to show the device but not play anything [11:42] same here [11:42] what coded does the transformer use ? [11:42] *codec [11:42] WM8903 [11:43] hmm, no idea about that one [11:43] ac100 uses a relatek [11:43] *realtek [11:43] ah [11:44] alc53** [11:44] or so [12:20] hi all [12:24] the modules for the new kernel have only drivers directory [12:24] when i run update initramfs [12:24] it giving me the error [12:24] but you ran make mudules_install when you built your kernel ? [12:24] yes [12:24] i ran [12:25] it given me 2.6.37.6 directory [12:25] so /lib/modules/$(uname -r)/ should have modules if you selected any in the config [12:25] in lib [12:26] i mean it giving me the error cound not find crzpto directory [12:26] crypto [12:27] when you installed the crypto stuff should have been removed by the installer ... strange [12:27] unless you selected "encrypt my home dir" on the user creation page [12:28] you can manually uninstall cryptsetup though [12:28] if your kernel has no support for it [12:29] what is the relation with crypt stuff and initrd image [12:30] to unencrypt filesystems before they get mounted you need to get the password [12:30] that happens in initrd [12:31] for all filesystems that have to be in place before login [12:31] ok [12:33] iam sorry for these type questions . iam bit new to this . [12:34] then do i need to uninstall crzpto stuff [12:34] crypto stuff [12:34] to make update this initramfs [12:52] wm8xxx are nice supported in linux usually [13:00] lilstevie: broonie (Mark Brown) is maintainer of wm**** drivers === MrCurious_ is now known as MrCurious [13:24] @ogra : do i need to change boot args [13:24] for the root device [13:26] no, the installer should have set the UUID properly [13:30] Is it possible to install the prebuild arm Ubuntu rom on a random omap4 android phone without typing phone-specific commands? [13:31] strange, yesterday someone in #pandaboard asked nearly the same and couldn't believe that pandaboard images were not at all meant for his funky motorola thing. [13:43] what are "phone-specific" commands [13:43] whisteling the right tunes to imitate a keypad ? [13:43] heh [13:44] Mean you no need to modify command that you type according to the phone you have… [13:45] * LetoThe2nd guesses he refers to obscure rooting and flashing command from some homebrew-rom-installer board. [13:45] c3would you mind giving an example, link? [13:52] .. can't think of an example .. [14:02] then how did you get the idea it could be necessary? [14:06] As people prefer step by step instruction but to any things that need to test and correct themselves…… [14:09] *** statement parsing error: ENOTUNDERSTANDABLE - bailing out. [14:11] in another word people doesn't like thinking… === brendand_ is now known as brendand_n310 [14:52] hi [14:52] it i mean the booting prcess giving me the same result [14:53] uncompressing linux done [14:53] booting the kernel [15:29] ogra_: having the large partition is extremely nice, no cosntant out of disk spce warnings :-) [15:30] i never have that :) [15:30] but i also dont use much data on the local disk ... my payload lives externally [15:35] ogra_: Where is the delta review page? [15:35] ogra_: I don't see what delta I have [15:35] lag, no idea, i didnt even know about that spec :) [15:41] I need to figure out how to layout my emmc :p [15:41] I crave the speed of moving internal [15:42] emmc? Which platform? [15:42] transformer i guess [15:46] that one :) [15:46] not much different from ac100 in that regard [15:47] more ram, bigger panel and different wlan [15:47] gpt on the ac100? [15:47] ah, no [15:47] don't forget the different audio codec :p [15:47] i thought it used a fastboot based installer too [15:47] oh, yeah [15:47] well, they are both tegra2 :P [15:47] heh [15:48] is yours harmony or ventana [15:48] harmony based i think [15:48] * ogra_ isnt sure [15:48] ah ok, this one is ventana [15:48] bah, so they are totally different devices that pretend to be the same :P [15:48] heh [15:50] i have included earlyprintk /boot/boot.script stiliam seeing nothing after uncompressing linux [15:51] but it booting fine if i replaces the uimage [15:54] ogra_: wifi is not looking good on this device [15:55] there is a userland binary which runs after module is loaded and firmware has been uploaded === Jack87|Away is now known as Jack87 === Jack87 is now known as Jack87|Away === Jack87|Away is now known as Jack87 === Jack87 is now known as Jack87|Away [18:55] hi all [18:56] i have problem while booting with the custom uImage [18:56] its giving me uncompressing linux [18:57] if i change root to /dev/mmcblk0p2 [18:57] then the root device is notmounting [18:58] Do you have a rootfs on a device that your kernel detects as /dev/mmcblk0p2 ? [18:58] i have ubuntu rootfs on device [18:58] i have bulded a custom uImage [18:59] it is booting good if i used the preinstalled images from ubuntu [18:59] if i change the uImage [18:59] How did you construct the uImage? [18:59] then it just showing me uncompressing linux [19:00] i downloaded the kernel [19:00] then patched with adeos ipipe [19:00] and then compiled it [19:02] Which kernel did you download? [19:02] 2.6.37.6 [19:02] From where? [19:03] git://git.xenomai.org/ipipe-gch.git [19:04] Are you certain that this kernel includes all the patches from the Ubuntu kernel? My suspicion is that your new kernel isn't representing your storage in the same manner. [19:05] ok [19:05] then how can i boot this kernel with ubuntu [19:06] Firstly, you'd need to understand how that kernel represents the storage area into which you have placed the rootfs, and pass the correct argument. [19:06] Secondly, you'd need to be sure that kernel had the appropriate support for the Ubuntu booting procedures, etc. [19:09] Is there any link for the ubuntu booting procedures on pandaboard [19:11] No, but there's nothing remarkably special about the pandaboard in this case. [19:11] For the least-failure-prone path, I'd recommend starting with the Ubuntu omap4 kernel sources, and then trying to apply patches to that. [19:12] If that result doesn't work, you can be fairly sure that this is due to the patch. [19:12] Conversely, if you select an arbitrary kernel, the source of possible failures can be large. [19:17] yes but here the constraint is the ipipe support for pandaboard . They did not released it as a patch . thez released it as patched kernel [19:19] You referenced a git tree: you ought to be able to use git to determine which patches have been applied which are not applied in the Ubuntu tree. [19:24] ndec: i'm trying to get my new blaze board to boot ubuntu [19:24] ndec: but it's not even showing uboot [19:25] ndec: i'm trying to boot from SD === Jack87|Away is now known as Jack87 [19:31] demarchi: you need a special x-loader and u-boot for blaze [19:32] demarchi: for Maverick https://wiki.ubuntu.com/ARM/OMAPMaverickInstall === Jack87 is now known as Jack87|Away [19:32] then to support Natty I believe you can just use the same x-loader/u-boot combination [19:33] just need to check if the kernel supports blaze, and as I don't have one I never tested [19:33] GrueMaster: did you ever tested natty on a blaze? [19:33] demarchi: do you know which ES rev do you have at your blaze? [19:34] we only support >= ES2.x === Jack87|Away is now known as Jack87 [19:38] rsalveti: it's ES2.2 GP [19:39] rsalveti: what's the difference between EMU and GP? [19:40] demarchi: hm, quite new actually [19:40] demarchi: not sure, maybe you can find it at http://www.omappedia.org/wiki/Blaze [19:40] or maybe robclark can help us :-) [19:40] otherwise ndec is your man [19:42] demarchi, an EMU is sort of half way between a HS and GP device.. [19:42] HS has all the security and firewall stuff.. and is generally a pita to debug / use jtag, etc.. [19:42] GP is wide open.. [19:43] and EMU has all the security features but can be opened up.. basically intended for R&D.. [19:43] (maybe I paraphrase a bit.. but that is the general idea) [19:44] robclark: depending on my version, i need a different MLO, right? [19:44] yes.. or rather HS and (I think) EMU would need a signed bootloader [19:45] robclark: i have a GP one [19:45] ok, then don't worry about EMU/HS stuff ;-) [19:45] nice... [19:45] just use normal blaze/panda/whatever MLO.. depending on what board you are using [19:46] maybe the problem was because i didn't rename the gp_MLO to MLO when copying to the sd card? [19:46] yeah.. that would be a problem [19:55] robclark: it didn't work [19:55] robclark: no output on console :-/ [19:56] well.. what board, what bootloader? [19:57] fwiw, ubuntu itself isn't very chatty on the serial console.. but you should see *something* for the boot-loader.. [19:57] robclark: nothing... i don't see u-boot on this board [19:58] hmm.. I'm guessing the default ubuntu image has a panda boot-loader.. [19:59] so I'm assuming you replaced that already w/ something for blaze? [20:04] robclark: the guide for android has the uboot and mlo [20:04] robclark: i tried those [20:05] where did those come from? And are they actually for a GP device? [20:05] robclark: that should work for ubuntu as well [20:05] I think most of the android folks are using either EMU or HS [20:05] http://www.omappedia.org/wiki/Android:_Working_with_pre-built_binaries [20:05] there's a gp_MLO there [20:06] and looking at this thread, i think it should work: http://permalink.gmane.org/gmane.comp.hardware.microcontrollers.omap/69 [20:06] hmm.. [20:07] well, the non-gp MLO name seems to imply it is for an es2.1.. so it's possible you need something newer.. [20:10] robclark: humn.... true [20:10] well, MLO and u-boot are easy enough to build.. [20:10] robclark: i'll just try to compile a newer version [20:10] it might be easier to just clone the git trees and build your own.. [20:11] I think if you take whatever versions are used in 11.04 and just build em w/ sdp4430 config (or whatever it is called..) === Jack87 is now known as Jack87|Away [20:32] rsalveti: I think I tested Blaze back with Natty Beta 1 [20:33] Does it need anything than a different u-boot? Could support be added to our u-boot tree? [20:34] persia: needs support at the x-loader and u-boot [20:34] but TI didn't actually pushed the support upstream [20:34] they are maintaining it at their own tree at omapzoom [20:35] iirc, the mlo/u-boot for blaze from Maverick worked with natty as well. [20:35] Kernel is good. [20:35] just need to test if it works with es2.2 [20:35] If I get time, I'll check it out, but it is currently low priority for me. [20:35] rsalveti, Ah. Is there something deeply incompatible, or is it just a matter of not having gotten around to it yet? [20:35] I only have a 2.0 [20:36] persia: lack of hardware for the upstream maintainers and missing someone to actually push the patches upstream [20:36] so we can only wait TI to do it :-) [20:37] Well, or someone with compatible hardware and sufficient motivation, but yeah, that's most likely TI. [20:37] with pandas around, don't know why someone would use a blaze [20:37] I mean, from community [20:38] It has a nice plastic box, and nice screens, and bamboos are hard to find? [20:38] persia, usually bootloader needs to be updated for a new ES.. if nothing else, to add a new Si revision id.. but usually there are different settings for voltage rails, different memory settings, etc, etc.. [20:38] There's lots of us that don't like dev boards as a form factor. [20:38] persia: it's expensive, hard to get and so on :-) [20:38] robclark, Right. I thought that was just compile-time choices. [20:38] so it is safe to assume that bootloader needs to be rev'd whenever there is a new ES [20:38] no, it is even runtime.. [20:38]