[00:04] where would i be able to get an ARM devboard from? [00:06] can't seem to find these marvell dove boards on marvells website [00:08] Bre, the dove isn't very easy to find... Freescale's i.mx515 is sold by a company that's name escapes me for the moment... Other wise there are ton of omap35xx (cortex-a8) based boards... [00:10] is that the beagleboard? [00:10] beagleboard, overo, igep2, allwaysinnovating, etc.... yes.. [00:10] ah ok seen those for sale yeah [00:11] we use some omap flavour at work [07:59] After playing with moserial a bit, it seems *way* better than all the other tools we have to manage serial consoles. [10:00] lool, so i finally have a setup to strace apt ... it simply hangs in read() [10:01] looking at top apt-get only consumes a maximum of about 50% CPU [10:01] dpkg about 10% [10:01] nothing that looks serious at all [10:04] aha, now apt-get CPU usage rasies above 90% ... and i dont see any dpkg processes anymore (strace didnt change though) [10:09] ogra: Which kernel are you using? [10:09] the archive netinst kernel [10:09] no initramfs [10:09] It might be that you uncovered a race in apt or in qemu [10:10] i tried to stay as close to the rootstock setup as possible [10:10] You want to gdb apt I guess [10:11] well, that means i need to start over :( indeed i dont have a dbgsym package in the VM and next run of apt will work [10:11] * ogra curses how time consuming all this is [10:15] i wonder when apt was built last [10:16] hmm, feb 02 [10:16] ogra: You can dpkg install it [10:16] err 03 [10:16] You don't need to apt install -dbgsym [10:16] lool, but it wont replace the running process [10:16] ogra: It doesn't affect processes [10:16] ogra: gdb will pick up the symbols.... [10:16] oh, right [10:20] lool, no way ... db is locked [10:24] bah and scp hangs ... i cant copy the strace log :/ [10:24] ah, just took a bit [10:34] ogra: Just dpkg-deb unpack it [10:35] to late [10:36] * ogra killed the VM and already rebuilds a qemu minimal image [10:36] i'll just make sure to have all possible requirements installed this time in advance [10:37] http://launchpadlibrarian.net/40681851/apt-strace.log is the strace btw ... [10:38] i think the read is apt reading from dpkg here [10:38] so i think its more likely a dpkg bug than apt [12:16] ogra: did you ever get mxc_ts working? [12:16] touchscreen? [12:16] (EE) PreInit returned NULL for ""mxc_ts"" [12:18] babbage 3 has a touchscreen? [12:19] not sure ;) X complains about it [12:21] it has a touchscreen interface [12:22] asac, ask persia :) i think the netwalker has a touchscreen [12:22] It does [12:22] * persia is trying to fixure out which one [12:22] you might need a patches tslib for it [12:22] *patched [12:23] Yeah, it's mxc_ts [12:23] and probably a udev rule we dont have atm [12:23] * persia checks for tslib patches [12:23] ogra: Wouldn't I have most of it on the Netwalker? [12:24] persia, definately [12:24] * ogra checks the bsp [12:24] whee, there are a bunch of tslib patches in there [12:24] Doesn't need tslib patches. Netwalker is running 1.0-4ubuntu2 uploaded by NCommander. [12:25] And I don't have a udev rule for mxc_ts (although I do have ones for mxc_iim and mxc_vpu) [12:26] So I think it just works if the kernel does the right thing. [12:26] Unless there's somewhere else I should be looking. [12:26] geez there is a huge patch that enables input events in tslib [12:26] ah, most is autofoo [12:27] persia, seems most of the patches i see are changes to ts.conf [12:28] Where is that file? [12:28] and this one: http://paste.ubuntu.com/392488/ [12:28] /etc/ _ [12:28] the ts.conf should be in /etc [12:29] I'm running stock Jaunty /etc/ts.conf : no changes. [12:29] And I really don't have that udev rule. [12:29] -# module_raw input [12:29] +module_raw input [12:30] Hrm? I'm not showing any patches, but I have "module_raw input" rather than "# ..." [12:30] do you have module_raw commented ? [12:30] right, so that seems to be upstream [12:31] the rules should be in /etc/udev/ltib/10-imx.rules [12:31] I *don't* have module_raw commented. [12:31] yes [12:31] But I think we already uncomment in Ubuntu. [12:31] yes [12:32] `grep -r mxc /etc/udev/rules.d/*` doesn't show those rules. [12:32] I don't have /dev/input/keyboard or /dev/input/ts0, but I don't really care, because it works anyway. [12:32] heh [12:32] do you have an xorg.conf ? [12:32] * persia has a bundle of /dev/input/event* [12:33] I do, but it's 0 bytes long. [12:33] weird [12:35] Why? [12:36] ogra: did you have a working xorg.conf? [12:36] asac, for imx gpu stuff ? no [12:36] well, i did once [12:36] ok [12:36] but dont have it anymore [12:37] i am looking at the fbdev code and wonder how that does the probing [12:37] iirc you only need to add the right driver line [12:37] seems it only does that for PCI [12:37] right [12:37] how does it work for us then? [12:37] forget about autoprobing [12:37] well. i want to understand how the fbdev code does it ;) [12:37] NCommander did look into fb probing quite a while ago [12:37] Works for me. If someone identifies a package, I can hunt a patch. [12:37] UGH ! [12:37] seems that we probe for fbdev properly [12:37] but i dont see how that is happening [12:38] i wonder if the clock issues we have with resets are based on the fact that we dont have mxc_rtc enabled at all [12:38] ah so you get resets too now? [12:38] ;) [12:39] http://paste.ubuntu.com/392496/ [12:39] well, i have seen your issue when i removed power [12:39] # CONFIG_RTC_DRV_MXC_V2 is not set [12:39] # CONFIG_RTC_MXC is not set [12:40] persia, LoadModule: "evtouch" [12:40] heh [12:40] you dont use tslib at all [12:40] * ogra pats good old evtouch [12:40] Well, that explains it :) [12:40] to sad it wont work with the new xinput model [12:40] Why not? [12:41] nobody ported it [12:41] the desktop team took over touchscreens [12:41] but at "very low prio" [12:42] ogra: file a bug and get cooloney look [12:43] so anyone can explain why our X auto selects fb? [12:43] ;) [12:43] asac, well, i'D like to find our old conversation first, there must be a reason why we use mc13892 instead of mxc_rtc [12:43] asac, yes, NCommander should be able to [12:43] NCommander: there? [12:46] oh seems its doing pci access [12:48] ah ... ok [12:49] i think it iterates iver fb devices in a hardcoided manner if it doesnt find anything on pci ... but i'm not sure i remember correctly === ramana is now known as ramana-away [12:54] there is some stuff i cant even find a header for [12:54] fbdevHWInit [12:54] forinstance [12:55] oh got it === ramana-away is now known as ramana [13:26] ok seems fbdev is somewhat hardcoded ;) [13:27] or i am missing something ... if i ship the imx_drv.so as fbdev.so ... it will use that ;) [13:27] * asac goes #ubuntu-x [13:50] can someone send the contents of /proc/fb on dove? [14:09] plars, saeed: Hey, would you mind sharing the contents of /proc/fb on dove? it's for asac [14:09] asac: ^ [14:10] hi all [14:10] 0 GFX Layer 0 1 Video Layer 0 2 GFX Layer 1 3 Video Layer 1 [14:11] I have lcd and vga (two fb) enabled [14:12] lool, asac: xawtv fails on lucid [14:14] saeed: Do you know why it doens't display the kernel driver name? [14:15] saeed: asac is looking at extending the xorg-server logic to select a fb driver for ARM depending on the contents of /proc/fb; on x86, it nicely containts the kernel drivers [14:16] But on ARM, these are descriptive strings instead :-/ [14:17] lool: no idea, I'll forward your q to the lcd drivers maintainer [14:18] saeed: lool: thanks. much appreciated [14:37] (II) EXA(0): Offscreen pixmap area of 31981568 bytes [14:37] (II) EXA(0): Driver registered support for the following operations: [14:37] (II) Solid [14:37] (II) Copy [14:37] (II) Composite (RENDER acceleration) [14:37] (II) UploadToScreen [14:38] (II) DownloadFromScreen [14:38] (II) FBDEV(0): IMX EXA acceleration setup successful [14:38] ;) [14:53] That looks kinda familiar :) [14:55] asac, now use xranrd and see it explode [14:56] ogra: you ask for too much ;) ... i dont even have a monitor connected atm. fighting on other fronts first ;) [14:56] composite ? [14:56] wow the version i tested in karmic definately didnt have that === ramana is now known as ramana-away [15:32] ogra: [15:32] asac@babbage2:~$ xrandr [15:32] Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192 [15:32] VGA1 disconnected (normal left inverted right x axis y axis) [15:32] LVDS1 connected 1024x768+0+0 (normal left inverted right x axis y axis) 246mm x 185mm 1024x768 50.0*+ 85.0 75.0 70.1 60.0 40.0 832x624 74.6 800x600 85.1 72.2 75.0 60.3 56.2 640x480 85.0 72.8 75.0 60.0 59.9 720x400 85.0 640x400 85.1 640x350 85.1 [15:33] no crash ;) [15:33] well, but only 1024x768 [15:33] no monitor connected [15:33] let me do that ;) [15:33] is your monitor limited to that ? [15:33] ah [15:33] but no crash is a good thing [15:33] for me it did crash [15:34] but that was a pre-pre-pre-pre...-pre-release [15:35] though i could bump the resolution to 1440x900 with an xorg.conf [15:35] so yeah xrandr doesnt see that i have a great monitor [15:35] hmm [15:35] right, the EDID code still needs love [15:35] too bad that netbook-launcher doesnt install [15:36] would like to see if that works [15:36] does here [15:36] right. seems its a problem if you run apt-get instlal from minimal [15:36] * ogra is just running apt-get install ubuntu-netbook^ [15:36] for the third time today [15:36] on top of minimal [15:37] (note i'm installing a task, not a metapackage) [15:37] though the task shouldnt make a difference to the metapackage in that case [15:37] wow [15:38] it only worked when i did this: [15:38] sudo apt-get install netbook-launcher libbonoboui2-0 libgnome2-0 gvfs libgdu0 udisks libparted0 [15:38] had to specify those manually [15:38] hmm [15:38] odd [15:38] something is clearly screwed on your setup [15:38] heh [15:38] not sure how ;) [15:38] my setups usually work forever === ramana-away is now known as ramana [15:39] was your minimal install up to date ? [15:39] sure [15:39] oh [15:39] libparted1.8-12 [15:39] ah [15:39] needed to be removed [15:39] that was holding it [15:39] not sure how i got that [15:40] * asac hopes it was a intermediate packaging bug [15:40] nobody else sees it :) [15:40] nobody runs a minmal system and installs from there [15:40] but i think it was fixed ;) [15:41] i do every day :) [15:41] rootstock FTW :) its our best archive tester [15:41] but you rewinstall all the time [15:41] mine came from upgrading [15:41] ah [15:42] indeed i do a fresh minimal and then dump a netbook task on top === ramana is now known as ramana-away === ramana-away is now known as ramana [16:06] plars: any idea how to make paplay work? [16:06] aplay works fine to me on dove [16:07] saeed: in my tests, I was seeing it had some relation to sampling rate of the file I was trying to play [16:07] saeed: see bug #528524 [16:07] Launchpad bug 528524 in totem (Ubuntu Lucid) (and 5 other projects) "Sound not working in all apps on dove (affects: 2)" [High,New] https://launchpad.net/bugs/528524 [16:08] saeed: unfortunately I'm blocked on doing much further at the moment, my board isn't working well (possibly bad memory) [16:11] is dove x0 rev 1.4? [16:12] plars: i tried paplay /usr/share/sounds/alsa/Side_Left.wav, but I hear nothing [16:14] saeed: interesting, could you try paplay /usr/share/sounds/gnome/default/alerts/bark.ogg and paplay /usr/share/sounds/gnome/default/alerts/sonar.ogg [16:14] plars: silence [16:14] saeed: on both? [16:15] yes [16:15] plars: with aplay, I have to run with sudo [16:16] saeed: oh, did you adjust settings in alsamixer? [16:16] saeed: see: http://launchpadlibrarian.net/33792996/alsamixer.png for what it should look like [16:20] plars: yes, I've this settings [16:21] how can I check that pulseaudio uses the sound controls I've on the board? [16:22] Has anyone got experience building xbmc for ARM? I've got GLES working on my IGEP and I would like to try to get xbmc working... [16:23] saeed: odd... I did have a Y1 board for a while that I could not get *any* sound out of, no matter what. GrueMaster had much better success with his, even with the same settings. I never fully resolved before trading it out for the x0, but assumed it was possibly just a hardware problem on my board [16:25] NCommander: ping [16:37] saeed: how about calling aplay on those wav files mentioned in that bug? does that work on any of them? [16:38] plars: aplay works fine [16:38] saeed: or another thing to try might be: speaker-test -t wav -r 48000 and speaker-test -t wav -r 44100 [16:38] does one work, and the other does not? [16:39] * ogra doesnt get it ... [16:39] seems i cant reproduce the apt hang anymore [16:40] i reproduced it twice today ... very reliable ... then glib got out of sync ... now ubuntu-netbook just installs [16:40] * ogra scratches heads [16:40] -s [16:52] dmart: ping [16:52] Hi [16:58] plars: sudo speaker-test -t wav -r 48000 works [16:59] plars: speaker-test -t wav -r 44100 fails on bad sample rate [17:00] saeed: so it sounds like you are seeing basically what I was, but also paplay not playing *anything* [17:00] saeed: could you update the bug with your observations, and possibly collect the log files they were asking for? Would be nice to move the bug forward, but I can't in my current state until I get working board again [17:08] Has anyone compiled omapfbplay for Ubuntu? I had a bit of a go last night (cross compiling) but didn't get too far - I'd appreciated pointers! Or, anyone know a good way to get smooth video on OMAP3/Ubuntu? === ramana is now known as ramana-away === ramana-away is now known as ramana [17:53] hmm isnt there a trick to change SONAME for binary lib? [17:55] If the new name is the same length or shorter, you might be able to hack the string directly, but otherwise some tables would need to be rebuiilt --- I don't know of a tool which can do this. (Stripped) binaries might not contain enough info to do it safely. [17:57] yeah ok [17:57] so no chance i guess [17:58] i would like to append something ;) [18:00] Unfortunately you probably have to relink :/ [18:00] You could create an empty proxy library with the right soname which pulls the second one in, but that's a bit of a hack [18:02] nah ,) [18:02] we have to live with it unverseoined then ;) === ramana is now known as ramana-away [18:06] NCommander: ping === ramana-away is now known as ramana === ramana is now known as ramana-away === bjf is now known as bjf-afk === ramana-away is now known as ramana === ramana is now known as ramana-away === ramana-away is now known as ramana === ramana is now known as ramana-away === ramana-away is now known as ramana === ramana is now known as ramana-away === bjf-afk is now known as bjf === ramana-away is now known as ramana === ramana is now known as ramana-away [21:49] Hi all! I just came across this: http://news.softpedia.com/news/Acer-Prepares-Freescale-i-MX515-Powered-Smart-Monitor-131435.shtml [21:51] and it occured to me that the monitor could be ideal for running a mythfrontend and X server only, assuming it has the power... [21:52] I don't know if mythtv-frontend is packaged for ARM, but would it be realistic to do something like that with Ubuntu? [22:49] Hi All [22:49] Has anyone been able to get the ARM port working on Datawind Ubisurfer? [22:52] It is an ARM9, rather than ARM7 [22:55] You are mixing up arm7 and armv7 [22:55] (v7 is the generation of the processor architecture) [22:55] an ARM9 chip is actually a v5 processor, and thus not supported by ubuntu [22:56] ( arm versions are like intel numbering -- i386/486/686.. etc ) [22:57] Martyn: Thanks for clarifying. [22:57] ARM926EJ-S rev 5 (v5l) [22:58] (according to my quick bit of research) [22:58] This is therefore probably supported by upstream debian though [22:58] so you should, in theory, be able to get it running debian Lenny [22:58] Martyn: I don't know what distro the Ubisurfer currently runs, so someone must have ported to it. However, have not been able to find out any more. Datawind have not responded to my emial [22:58] it runs 8.04 [22:58] 9.04/9.10 dropped support for v5 [22:58] 10.4 drops support for v6 [23:00] I see that version now on this output: http://www.linuxquestions.org/questions/linux-laptop-and-netbook-25/datawind-ubisurfer-784464/ [23:02] Martyn: Do you know if anyone tested running 8.04 on Ubisurfer? I saw only some development boards by Marvel were listed as officially working. [23:06] no idea [23:06] sorry :( [23:15] no probs. [23:15] I was looknig for an ARM netbook, not found one yet.. well only this ubisurfer [23:19] now3d: there are these: http://www.pocketables.net/2009/09/first-impressions-of-the-sharp-netwalker-pcz1.html [23:20] now3d: from what I understand, pretty difficult to acquire if you are not in japan though [23:22] thanks for the tip! [23:23] ok, logging off now. thanks for help [23:23] now3d: ah, don't know how reputable the site is, but I did find one place that claims you can order [23:23] doh [23:24] http://www.geekstuff4u.com/sharp-netwalker-pc-z1.html for anyone else who might be interested