[00:22] persia: lol, now it doesnt even outputs sound... [00:52] persia here it is [00:52] http://dl.dropbox.com/u/1333955/soundtestarmpd.ogg [01:09] rlameiro, Cool. Please file a bug and attach it, so that folks who can fix it (not me) can have a look. Please use ubuntu-bug to file the bug *from the beagleboard* as this will collect more useful information. [01:11] well, i am installing apport right now [01:11] a bunch of libs [01:11] persia: what is the best way to see the cpu frq on the command line? [01:11] cpuinfo only shows MIPS [01:12] *BogoMIPS [01:13] I don't know if that's exposed to userspace except as strings: cpuinfo or dmesg output would be my expectations. [01:14] I could be mistaken though: I don't tend to do lots of low-level stuff. [01:14] Probably better to ask that sort of question generally: other folk might have better answers (and I think most ARM folk are familiar with more low-level bits than I) [01:16] well, i made cat /proc/cpuinfo [01:16] i am not eally sure if my processor is really runnig at 720Mhz [01:56] lp #642465 [01:56] Launchpad bug 642465 in puredata (Ubuntu) "Puredata outputs Very bad sound when using ALSA backend (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/642465 [01:56] persia: done, bug posted [01:57] persia: maybe you could triage it to the relevant teams/persons, i dont know who they are, i notified the ubuntuarmel porters team only === ian_brasil__ is now known as ian_brasil === ian_brasil_ is now known as ian_brasil [09:52] Anyone around? [09:53] nope [09:53] Hi Loic [09:53] Hey, how goes? [09:53] Meh [09:53] lag, Lots of folks are around: "/names" usually gives you a list. [09:53] Do you cross compile? [09:54] lag_: Right now I'm enjoying my Sunday, but I cross-compile from time to time, yes [09:54] :) [09:54] lag_: Anything in particular? userspace or kernel? [09:54] I usually build natively, but I want to do some work on the plane [09:54] Kernel [09:55] Ubuntu specific [09:55] I can cross compile [09:55] lag, You clearly need a Netwalker :) [09:55] But I would like to do it via fdr [09:55] lag_: Cross-compiling the kernel should be relatively easy; cross-compiling the Ubuntu kernel package is not trivial because they enable perf/ which means you need a couple of cross-build libs [09:55] lag_: fdr? [09:55] fakeroot debian/rules? [09:55] Yeah [09:55] I think the kernel team has a cross-compilation recipe to work around that. [09:56] But they're offline :) [09:56] And my plane leaves shortly [09:56] lag_: For kernel packages, just set CROSS_COMPILE=arm-linux-gnueabi- and it should work with fdr [09:56] yeah, but someone put it in the wiki... [09:56] That simple eh? [09:56] :) [09:57] lag_: Yup; even debuild -eCROSS_COMPILE=arm-linux-gnueabi- -aarmel should work [09:57] I'll give it a go [09:57] lag_: Problem is that you need to set some more vars if you want to replicate debuild or dpkg-buildpackage [09:57] I'm having issues installing hrw|gone's cross compiler at the moment [09:58] You're using the new shiny in-archive versions? [09:58] Of the cross-compiler? [09:58] I'm trying to install them [09:58] Yes, of the cross-compiler [09:59] The following packages have unmet dependencies: [09:59] libgomp1-armel-cross: Depends: gcc-4.5-arm-linux-gnueabi-base (= 4.5.1-7ubuntu1) but 4.5.1-6ubuntu1 is to be installed. [09:59] https://wiki.ubuntu.com/KernelTeam/KernelMaintenanceStarter#Cross-compiling%20for%20ARM is the kernel team's recommended way to do kernel cross-compilation on Ubuntu [09:59] lag_: ah that's temporary [09:59] As you know, native compilation should be done for validation, but that can happen when you land. [09:59] lag_: the packages are uninstallable right now [09:59] lag_: This should be fixed before final [09:59] lag, You can work around the temporary issue by working in an i386 chroot (mk-sbuild is a handy way to create one) [10:01] Boo [10:01] boo? [10:01] Downloading CodeSourcery [10:02] That's faster than `mk-sbuild --arch=i386 maverick`? [10:04] make: *** No rule to make target `build-omap4'. Stop. [10:05] Despite: flavours = omap4 [10:14] hello [10:15] hey dcordes [10:17] hi all, i have IGEP arm board .trying to boot it with ubuntu lucid. [10:17] but getting error as unsupported signal on HDMI monitor [10:18] ajay: Do you run a known working kernel ? [10:18] yes [10:18] dcordes, they have given kernel [10:18] ajay: What's the known working userspace that will give you working signal on HDMI monitor ? [10:19] dcordes, i am not getting.. [10:20] dcordes, i have followed all steps as http://free-electrons.com/blog/ubuntu-1004-igepv2/ [10:20] ok [10:20] but i dont have null modem cable so directly testing on monitor with hdmi cable [10:21] ajay: A very bad approach. You should setup the serial connection [10:21] Well, not necessarily. [10:22] Lots of folks have success with using monitors directly. The problem is that it's hard to troubleshoot low-level stuff (especially with rootstock) done that way. [10:22] ajay: Can you try flashing a known working kernel binary instead of building from a random revision ? [10:22] after doing google i got in beagle board we need to change bootargs of uboot as ompafb.mode=dvi:hd720 [10:23] The correct value for omapfb.mode is decided by the device to which you are connection. [10:23] but not getting how to..bcoz i am not able to get even uboot propmt [10:23] Don't trust random values for that unless you have the same display [10:23] persia i have sony bravia tv [10:23] Edit the bootscript (.scr file) [10:23] I don't know the value for a Bravia. [10:24] Does the IGEPv2 output real HDMI, or DVI on an HDMI connector? [10:24] its real HDMI [10:24] i am using hdmi to hdmi cable [10:24] Then, yeah, you just need the settings for your monitor (and to set them in your bootscript). [10:25] As dcordes suggests, this would probably be easier if you had a serial connection to investigate :) [10:25] my system does not have a serial port so any usb to serial port can work with lucid? [10:26] or need a PCI slot serial port [10:26] Most ought. [10:26] ajay: For not having any debug output, you put too many variables. Reduce possible error sources. Download complete tested and known working kernel binary, modules, rootfilesystem. [10:26] which is better usb to serial or PCI slot [10:26] dcordes, ok [10:26] Oughtn't matter. serial is too slow to benefit from alternate connectivity. [10:27] dcordes, Do you happen to know if the lucid omap kernel works on IGEPv2? [10:27] persia: Sorry, before I did not hear about that board [10:27] No worries :) [10:28] dcordes, it is clone of Beagleboard [10:28] * persia wishes rlamiero was around: he has such a board [10:29] I was abscent for nearly two weaks and did not follow the development at all. Have there been any critical changes to the http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/ ? [10:30] I hope the beta freeze is a success [10:30] For maverick? Heaps. It mostly works now. [10:30] Ya. sounds good [10:30] There's a couple bits needing tweaking (and I believe one wants something with higher specs than a Beagle C4), but it's getting really close to good enough to release. [10:32] I am downloading 20100919 right now. Will apply my set of tweaks and run in on the HD2. Maybe I can find some bugs [10:33] Please file them if you do. Patches even more welcome :) [10:35] Will do [10:36] Also need to do further utouch testing [10:41] ajay: Have you found prebuilt stuff to test ? [10:42] dcordes, no they just given source [10:43] if i want ot change uboot parameter while compilation time where i need to change? [10:44] ajay: Do you mean the kernel command line ? [10:45] yes [10:45] dcordes, setenv bootargs console=tty0 console=ttyS2,115200n8 root=/dev/mmcblk0p2 rootwait ro vram=12M omapfb.mode=dvi:hd720 [10:45] ajay: There is a kernel configuration option to hardcode cmdline paramters but it's not common. Not a build time thing [10:46] ajay: Usually you pass it to the bootloader [10:46] ok. [10:46] Ok that is the kernel command line. Did you copy paste "omapfb.mode=dvi:hd720" parameter from a beagle board cmdline ? [10:46] dcordes, yes [10:48] ajay: That assumes you have the exact same framebuffer ( display) driver as the person uses you copied it from [10:48] dcordes, not sure i am new to this porting [10:49] ajay: The problem is you don't even know if your kernel starts :) [10:50] dcordes, i will get a null modem cable [10:52] but have a problem that shall i buy USB-serial or PCI slot serial port connector [10:52] my system doesnot have serial port [10:53] unless you have some leds or so ? [10:53] jkridner: and friends (beagle board devs )used to put tested prebuilt kernel binaries so you could test if your hardwarer setup is fine [10:53] I bet there is something similar for your IGEP board [10:54] get usb serial - less hassle [10:54] ajay, Unless you have some reason to get something special, I'd suggest a one-port serial connection: inexpensive, and easily migrated to other uses in the future. [10:54] Be aware that serial is not serial. There are different voltage levels [10:55] dcordes, which one i should use for getting prompt on minicom [10:55] ajay: I don't know your device well. Search for documentation [10:56] dcordes, it is similar to beagle board [10:58] http://www.google.de/#q=igepv2+serial ? [11:10] dcordes, i am asking for development system should have PCI slot derial oprt or can be usb serial port [11:10] i have IDC10 to DB9 cable [11:11] 11:26 < persia> Oughtn't matter. serial is too slow to benefit from alternate connectivity. === freeflyi1g is now known as freeflying [11:18] ajay, So, I don't know of any RS232 UARTS capable of more than 230,400 bps. USB is twice that. PCI is lots more. In fact, I can't think of *any* interconnect other than IEEE1294 in modern use that is slower. [11:18] And the one on your board is probably 115,200 or slower anyway. [11:19] Err, 1284 (sorry) [11:20] persia, thanks [11:20] Guys I have to run. I will be back once I have tested the recent maverick netbook preinstalled image on my phone (for feedback or to pester you with questions ;) ) [11:20] Oh, and I'm completely wrong: USB is 2000 times that fast, and IEEE1284 is 10 times that fast [11:21] persia, dcordes as well with this board i like to test utouch [11:22] so can you please suggest me which touch screen 10 or 12 inch i should buy [11:34] dcordes, persia thanks i am able to get display on TV [11:35] but i have created filesystem for lucid using rootstock and using kernel 2.6.33.7 [11:35] while booting getting error as no /lib/modules/2.6.28.10/modules.dep file found [11:35] and getting exit [11:36] not able to found from where boot kernel is taking as 2.6.28.10 [11:36] is there any way to chang it to 2.6.33.7 [11:37] My recommendation would be to adjust flash-kernel to handle your hardware, and then install a packaged 2.6.33.7 [11:38] If you want something easier, follow the directions in the article you first linked, create a .deb, install that, and manually copy the kernel and initrd images. [11:40] persia, i have copied manually uImage config [11:41] installed all kernel modules to lib/modules/ [11:41] but doubt is why it is looking for 2.6.28.10 [11:42] Answer is probably somewhere in your dmesg or syslog [11:42] hmm but i am not able to get terminal [11:43] persia, is there any way to describe kernel as 2.6.36.7 [11:43] sorry 2.6.33.7 [11:43] in uboot params [11:47] No idea. [11:52] persia, about touchscreen to test utouch [11:52] which manufacturer touchscreen works? [11:53] Again, no idea. [11:53] I remember reading somewhere that it would benefit from "4-finger" touch. [12:01] persia, yes i got from where it was taking 2.6.28 kernel that is from nand [12:01] and my kernel is 2.6.33.7 in sdcard [12:02] i took boot.src content from beagleboard [12:02] but in http://free-electrons.com/blog/ubuntu-1004-igepv2/ it is mentioned the boot variable and then saved it [12:02] So you adjusted the boot.src, and were able to boot your preferred kernel? [12:02] its not like auto boot [12:03] persia, no [12:03] i adjusted same mentioned above till saveenv [12:03] added extra flag of ompafb.mode [12:04] but getting error as unsupported signal [12:09] Sounds to me like you have a working 2.6.28 kernel and a not-working 2.6.33.7 kernel [12:09] Are you sure the value you selected for omapfb.mode matches your screen? Also, you typed "ompafb.mode" above: are you sure of the spelling? [12:10] 18VABI76S, Now that's an impressive username :) [12:10] Easy to pronounce at least :) [12:11] * ogra_ac sighs ... installing mono on the ac100 trashes my whole chroot [12:11] Silly tomboy [12:12] ogra, Same error as elsewhere, or something else? [12:13] Habgs at installing mono assemblies [12:13] *hangs [12:13] Purging them without a terminal that can copy paste is a pain [12:13] Hrm. Really need to get the Mono folk faster hardware :) [12:13] I'm having some difficulties installing Maverick on my igep2. If someone has any experience and a minute, I could really use some help. I'm following the guide on http://labs.igep.es/index.php/How_to_get_the_Ubuntu_distribution#Ubuntu_10.10_.28Maverick_Meerkat_BETA.29. When I try to create the boot.ini by copying and pasting the command, I just get "$: command not found". [12:13] You don't have openssh-server installed yet? [12:14] Hmm, indeed i could work remotely [12:14] jo-erlend, mkimage isn't found? [12:15] ogra, Makes life lots easier if you can't copy & paste :) [12:15] persia, mkimage is found, but that's the only output it gives me. [12:15] persia, i am not sure about ompafb.mode value . [12:15] but hd720 worked for me when foloowed .src from beagleboard website [12:16] Persia, prob is that i cant install openssh-server due to the install failures, my apt db is totally screwed atm [12:16] ajay, OK. maybe that's correct. I know folks have complained about it not being the same for every display, and so I thought it might be related. [12:16] * ogra_ac tries to dpkg -i [12:16] ogra, dpkg -P is your friend [12:16] Though i dont really know how i would start it, no upstart :P [12:16] manually? [12:17] Pain [12:17] upstart just calls stuff. [12:17] Indeed [12:17] less pain than retyping everything in order to sort it out :p [12:17] * ogra_ac is inclined to give up the chroot idea and rather invest more time into the bootloader [12:17] jo-erlend, "command not found" usually indicates something isn't present. Are you sure you have an mkimage binary? [12:18] persia, quite. [12:18] ogra, Please :) [12:18] Heh [12:18] I dont want to brick it [12:19] Last time I had an ARM laptop I didn't want to brick and suffered through a less-than-ideal environment, I ended up breaking it before I even got to try, and now I wish I'd bricked it, as then there would at least have been a point. [12:19] jo-erlend, I can't imagine why you'd get that output. Maybe your path is funny? Beyond that, I'd suggest grepping the mkimage source to see if you can find any conditions that would result in that output. [12:20] * ogra_ac finally found out why android trashes the wlan at least [12:21] Some silly lib hardcodes an echo into proc that suspends the wifi constantly ... i wonder why they do that [12:22] "power saving"? [12:24] Every 3 seconds ? [12:29] Sure. if you don't have kernel support for making it sleep when unused, you send a poweroff event every three seconds: if there is traffic, this makes it a bit slower, but users don't notice because they expect a phone to be slow, and it there is no traffic, it saves power. [12:29] Silly [12:29] If I did it that way (which would involve much kicking and screaming, and possibly whips and chains), I'd probably do it every 2 seconds. [13:40] Hmm, so i have a loopback vnc session running [13:40] * ogra_ac sees an ubuntu desktop [13:40] :) [13:41] ogra_ac: lp #642465 [13:41] Launchpad bug 642465 in puredata (Ubuntu) "Puredata outputs Very bad sound when using ALSA backend (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/642465 [13:41] But sadly its absolutely unusable [13:41] ogra, Nice ! [13:41] rlameiro, ogra doesn't tend to be too much of an audio guy :) [13:41] Silly vnc client doesnt really hand teh kbd through [13:41] oh welll [13:42] i can try :D [13:42] rlameiro, Better to just watch the bug: someone interested will likely respond. [13:42] persia: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=4adb2744c8cdc1064b657cdffc4c0f91ff7bf442 [13:42] Does puredata by chance use libmad somewhere ? [13:42] check this out [13:42] In fact, I'll try to confirm once I'm not spending all cycles building. [13:42] seems promising [13:43] ogra, No, but it may be the same class of issue. [13:43] for jack [13:43] Assembler crack ? [13:43] rlameiro, That's pure crack, unfortunately. Folk need to focus on making stuff that works for everyone. [13:44] So a better solution would be for JACK to have a more flexible init system, and choose values based on the detected audio interface. [13:44] ogra, Dunno, but possibly. Doesn't sound like pd is doing math correctly. [13:44] persia: well, it may be also a problem with jack [13:45] it seems related to the capture part of the alsa driver [13:45] And it's all sorts of extra odd that doing a software conversion using ALSA OSS emulation makes it kinda work better. [13:46] but since it is working on recent versions of open embedded, maybe the driver in ubuntu isnt recent enough? [13:46] could it be that? [13:46] rlameiro, That entire patch consists of some config adjustments, specifically changing the default number of periods, frames per interrupt, and period length. You can try the same changes with the qjackctl interface. if that makes it better, cool, but it's not a real issue with JACK (the issue being that JACK always sets the same values, hardcoded in the code, rather than having a flexible config system exposed) [13:47] Could be. Could be different toolchain settings. Could be different instruction sets. Lots of possibilities. Needs someone to look carefully at why the sound is wrong. [13:47] persia: well i tried that yesterday with no success, i did tried various setting, and they crasshed qjackctl via ssh -X [13:49] You tried the settings in the patch you just linked? [13:50] persia: all but the sample rate.... [13:51] persia: its weird also that my USB interface doesnt work also the UA-4FX [13:53] USB interfaces not working is *very* odd. Please file a bug on that, as usb-audio should *really* be platform independent. [14:27] * ogra_cmpc curses [14:27] soandroid :( [14:27] *so i trashed [14:28] And this is bad why? [14:28] did you brick it, or just break the default OS? [14:29] well, effectively both [14:29] i messed up the sh link [14:30] neither recovery nor booting works [14:31] i still have a bootloader, kernel and initrd i guess [14:31] but no way to access anything [14:32] ogra: do you have a serial debug to the device? [14:32] i dont know wich device it is.. :) [14:33] rlameiro, one without serial debug [14:34] ogra_cmpc: oops, that is kinda bad.... is it a cell phone? [14:35] ogra, Do you have any means to adjust the bootloader, kernel, or initrd? [14:35] persia, no [14:35] else it would boot ubuntu already [14:35] :P [14:36] ogra_cmpc: the only way I see it is search for the UART pins inside the board and then connect it to a serial-usb converter [14:36] rlameiro, there are none [14:36] ogra_cmpc: really???? who makes a board without a serial debug ????? [14:36] rlameiro, This is in a plastic case: a consumer device. [14:37] even the android phones have it inside.... [14:37] well, that i know, but it is in there [14:37] ogra, So what happens when you boot it? [14:37] if it is bricked you dont loose to much :D [14:38] persia, bootloader hangs at the toshiba screen [14:38] rlameiro, well, i lose 400euro [14:39] ogra_cmpc: well that is expensive, did you tried to see if the usb subsystem is running when you poweron the device? maybe it haves some way of debugin using usb on a very low level [14:39] * rlameiro rlameiro is dreaming to high [14:40] ogra, Ugh. Is there a magic keypress for "recovery" that lets you load recovery from SD? [14:40] yes you are [14:40] persia, it can only install toshiba signed update.zip files [14:40] thats all the recovery mode offers [14:41] but even that nees a shell i guess [14:41] *needs [14:41] it may work as soon it detects an signed file [14:41] Make a support request. [14:41] haha [14:41] I suspect there's a way to fix it (might need to be sent in). [14:41] haha? it7s new enough to still be under warranty. [14:42] warranty already voided persia [14:42] "hello, i bricked a totally locked device that loses all warranty if i root it, can you help me ?" [14:43] No. "Hello, my new laptop won't turn on anymore...Yes, it's plugged in...Yes, the lights are on...Yes, I've pushed the switch, I only see the Toshiba logo....Could you repeat that RMA code?..." [14:43] (the data partiion still has all the crack tools i used to exploit it) [14:44] And you had to agree to some license saying you wouldn't do that? [14:44] yes [14:44] on first boot [14:44] Oh. That's annoying. [14:44] it also has a bunch of sticker on the case [14:44] *stickers [14:45] saying that the warranty is only valid with the unmodified preinstalled os [14:45] So even if you install stuff from the Toshiba store... ? Oh my. [14:46] there is no toshiba store :P [14:46] thats the funniest bit about the preinstalled stuff [14:47] it has a special marker app preinstalled ... which has no apps in it [14:47] hmm, so i did a failed dd of the eMMC [14:47] i wonder if i can pull something out of that, the beginning of the img file should be fine [14:48] and i suspect it is what carries the bootloader [14:49] aha [14:49] partition1 in the img file starts at block 33 [14:50] so i guess 0-33 is bootloader stuff [14:50] * ogra_cmpc wishes he could just access the eMMC directly somehow [15:01] How can I force a USB detection? [15:01] it doesnt appears on dmesg.. [15:02] my board isnt dettecting my usb sound card at all [15:13] well, it dont recognizes my usb hub neither my sound interface [15:13] maybe it only detects USB devices at boot time [15:13] how can i force usb detection [15:13] * rlameiro is feeling like devil advocate.... [15:15] re-plug it [15:15] udev should detect it [15:18] ogra_cmpc: it didnt [15:18] I rebooted the machine now [15:19] ogra_cmpc: well, now i pluged in the sound device and it disconnected the usb hub [15:20] "usb 1-1: USB disconnect, address 2" [15:20] last dmesg enty [15:20] *entry [15:46] hi while testing tv with igep board i did a mistake that uboot.src i changed and in that i wrote saveenv with different parameter [15:46] now i am not getting display on tv [17:54] my board isn't autodetecting devices, how can i force it to detect them? [18:05] -f [18:05] :D [18:28] trying to play using my usb audio interface with aplay shows errors something about sample rate, and with alsaplayer it outputs this error http://pastebin.ubuntu.com/496584/ [19:12] lag: armel cross compiler packages are broken now - fixed versions are on a way [19:31] hrw|gone, heh, he said he was about to board a plane and wanted to use the compiler there