[00:00] dcordes, I think I would have rephrased that question if I were you. [00:01] jo-erlend: Sorry ? [00:02] jo-erlend: I think he is referring to the Toshiba AC100 that ogra_ac is on. And yes, it is a Tegra 2 system. [00:02] ah. :) [00:07] * dcordes is waiting for decent qualcomm smartbook [00:07] Keeping track of one SoC is enough [00:08] Only need to convince somebody I can't make it through the winter term alive without one :D [00:08] ha! [00:09] Doubt it will work but I need to find a job anyway [00:12] Is it possible that it is my monitor that causes this? http://pastebin.com/6hAjUwCW [00:14] I think it's unlikely. I think you have to pass a framebuffer mode argument. Are you booting from recent dailies, or something else? [00:14] beta. It seems to be the only readily available image? [00:15] Oh, if you're using raw beta, that ought be passing a parameter [00:16] http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/ has some more recent images. Mind you, those are entirely unvalidated, so might not work, or not exist, etc. on any given day. [00:17] grand! Thank you. I'll give that a go immediately. I couldn't find that earier today. Perhaps I gave up too quickly. :) [00:18] hm, ureadahead is now using a 8mb buffer, instead of 128mb [00:18] persia: did come out some upgrades to maverick today? [00:18] interesting [00:19] persia: I found my usb-audio device. RCA & SPDIF in/out. Tested ok on my laptop. What were you suggesting I test it on? [00:19] and should fix the oom killer messages at beagleboard [00:19] GrueMaster, Anywhere you can, once per kernel flavour. No point on two devices with the same kernel, but otherwise. [00:20] I'll add it to the to-do list. [00:20] GrueMaster, rlamiero was reporting that on IGEPv2 (omap3) with the kernel from "linux", usb audio didn't work. I wouldn't be surprised to see other kernel flavours affected. [00:21] As it stands, my XM is the only one with spare USB ports, and it is borked. [00:21] well it work on i686 [00:21] No rush, but it would be nice to be able to tell folks they can use usb-audio in the event that we don't have working wiremaps for various boards. [00:21] rlameiro, Sure, but that's a different kernel flavour: linux-generic vs. linux-omap (although it comes from the same source package, which means we're more likely to have it just work) [00:22] yeap [00:22] GrueMaster: so, did you test the audio at the c4? [00:22] it is weird because the problem aplay output was about the bit rate or something like that [00:22] if it's not working this could explain why the same behavior at the xM [00:22] and a regression, as it worked fine for lucid [00:23] different kernel tree though [00:23] rsalveti: On my beagle, I had to tweak alsamixer settings to get the alsa speaker-test to work. [00:23] GrueMaster: same for xM, am I right? [00:23] My xm has other issues. [00:23] (like getting through a boot process). [00:23] but last time you tested audio you said it worked after setting up alsa-mixer correctly, IIRC === amitk is now known as amitk-afk [00:24] I thought I had. With as much testing as I do, unless I find a bug, I tend to forget what I tested where. [00:24] And it is very hard to test everything on everything. [00:25] haha, np [00:25] Especially with special requests coming in fast & furious. [00:25] but yeah, I remember you tested at the xM and you got it working after setting up alsa-mixer [00:26] that's why we discussed it at our call [00:26] and where we got the idea to test it on a normal beagle [00:26] because with lucid it just works [00:27] Right. [00:27] But now I can't get my XM to boot through a known good image. [00:28] These sudden audio issues eat at my soul. [00:28] this is something weird, let me try to boot mine [00:28] GrueMaster: what errors are you getting? [00:31] On which? Audio or XM in general? [00:33] can someone translate this for me? "omapfb.mode=dvi:1024x768MR-16@60". It's the bootargs I've been told to use. :) [00:33] xM in general [00:33] persia: usb-audio device shows up and is configurable on dove. Still no audio output from rythmbox. [00:34] jo-erlend: while setting up omapfb with dvi, this should be your default resolution [00:34] otherwise you'll get 640x480 [00:34] rlameiro, Does GrueMaster's report on dove match your experience on the IGEPv2? [00:34] that's the default one [00:34] rsalveti, what would happen if my monitor didn't support that resolution? [00:34] rsalveti: On my XM, it fails randomly. Sometimes it is a corrupted filesystem on second boot, sometimes it gets most of the way through oem-config before it crashes. [00:34] jo-erlend: will probably give out of sync or something like that [00:35] jo-erlend: use the Digital output of omap framebuffer driver with a res of 1024x768 with a bit depth of 16 bits at 60 hz refresh rate [00:35] I have not seen any errors that could be traceable. [00:35] GrueMaster: sounds like the random mem issues I'm facing with 512 [00:35] Possibly. [00:35] mine just works when I add mem=256M [00:35] persia: i didnt used rythmbox, but yea, it is configured but no sound output [00:35] I'll give it a try. [00:35] Hate hacking the boot.scr [00:36] the interesting thing is that some times it just works fine, but most of the time I try to boot with 512mb it'll give me random segfaults [00:36] I thought that since it fails to load the omapfb module, it might be because the modes were incorrect? [00:36] welcome to the club [00:36] How about lower level sound tests? Let's find the point at which it breaks. [00:37] jo-erlend: default should be safe, then you can try different resolutions [00:37] aplay should be a good ground level [00:37] persia: [00:37] Morning all [00:37] jo-erlend: what's you problem regarding omapfb? [00:37] lag: haha, morning [00:37] crazy timezone [00:37] rlameiro, I don't have an environment to help: tell GrueMaster :p [00:37] australia? [00:37] rsalveti, http://pastebin.com/6hAjUwCW [00:37] taiway [00:37] persia: speaker-test works with usb-audio on dove. [00:37] *taiwan [00:38] jo-erlend: oh, ok, the missing fb0 [00:38] jo-erlend: can you grab your dmesg? [00:38] haha? [00:38] Oh, I see [00:38] :) [00:39] rsalveti, I'm currently overwriting the memorycard with todays daily to see if that works. [00:39] rlameiro: Weren't you having issue with *any* audio? I don't think you were using pulse. [00:39] i dont use pulse [00:39] just lasa [00:39] alsa [00:39] if alsa is broken, pulse will be also [00:40] jo-erlend: could be better, but it's probably not going to fix this weird missing fb0 error [00:41] because we didn't change our kernel much regarding this driver [00:41] rlameiro: too true. [00:41] but please try, let's see what happens [00:41] if still doesn't work, you can try linaro's kernel [00:49] persia: GrueMaster http://pastebin.ubuntu.com/497342/ [00:52] rlameiro: Try "speaker-test -c 2 -t wav -Dhw:1" [00:53] GrueMaster: persia http://pastebin.ubuntu.com/497343/ [00:54] hmm [00:54] seems not using andchat gets me a far more stable network [00:54] i should remind that this usb interface doesnt have software mixer [00:55] rlameiro: try without the -D parameter. [00:55] that will play on the board. [00:56] Worked on usb-audio here. [00:56] the daily image doesn't seem to have anything except the boot partition. Well there is another partition too, but it only contains lost+found. [00:56] the daily should have 2 partitions [00:57] boot and rootfs, that will get resized at the first boot [00:57] GrueMaster: its working on the board output [00:57] yes, it does, but the rootfs is empty. [00:57] jo-erlend: after or before the first boot? [00:57] sorry.... Nautilus fooled me. :) [00:58] uh. I get lots of io errors now. [00:59] rsalveti, what do you mean? Everything should be on the board before the first run? [00:59] GrueMaster: speaker-test says this at the help file --- Recognized sample formats are: S8 S16_LE S16_BE FLOAT_LE S32_LE S32_BE [00:59] jo-erlend: sorry, after you putting the card and giving the first boot or before [01:00] but i am prety sure my board only send 24bit streams [01:00] any way it should work [01:01] * rsalveti brb [01:01] i will try the interface with the advanced mode off [01:02] lol another problem, each time I unplug a usb device, when i plug it back it cant be enumerated [01:02] i need to restart the board [01:11] GrueMaster, persia http://pastebin.ubuntu.com/497351/ [01:12] strange is that the stream of the driver is defined to 44100hz when the board is capable and hardware selected to 48000 [01:13] I forced speaker-test to the driver parameter and it worked, just the problem was that the audio file was at 48khz, so nothing happened [01:13] rlameiro, That looks like a deep bug, and ALSA is probably doing some internal resampling. Probably worth running the testing shell script and sending it upstream. [01:13] humm [01:14] persia: is that test on the repos? [01:16] * persia isn't sure [01:22] found this [01:22] it says nothing special [01:22] persia: unless this ----EDIROL UA-4FX at usb-ehci-omap.0-1.2, full speed [01:23] Hrm. [01:23] this means that the driver is on the specific module usb-ehci-omap.0-1.2 ? [01:24] i dont have the sources of the kernel from rcn-ee so i cant check it out [01:25] rlameiro: Unfortunately, I have a completely different usb-audio device. http://paste.ubuntu.com/497363/ [01:25] And it supports 48000 [01:26] well, then yeah, this driver is completely broken [01:26] I just wanted to know where do i find the sources for it [01:27] Either in the kernel or from http://alsa-project.org. [01:27] What is loaded by Ubuntu is in the kernel. [01:28] alsa-project.org may have newer updates that will be in the next kernel. [01:30] well, but it works on my laptop, i even helped testing the patch made for this device some years ago [01:30] and it went already to alsa release [01:30] and it is fully supported on ubuntu, so i dont get it, it shouldnt be platform dependent should it? [01:32] No, it shouldn't. [01:33] well i gues i will digg into it tomorrow [01:33] /proc/asound/card1/stream0 should be identical across platforms. [01:33] well i will compare it with the one on my laptop [01:33] 1 sec [01:36] GrueMaster: well its definetly an ARM/Board/ubuntu kernel issue or ubuntu driver [01:36] http://pastebin.ubuntu.com/497368/ [01:37] interesting. [01:38] Same kernel on both? [01:38] it seems something is wrong with the driver that doesnt detect the card correctly [01:38] nope [01:38] The omap driver is part of the main stream kernel. [01:38] Linux studiobeta 2.6.32-24-generic #42-Ubuntu SMP Fri Aug 20 14:24:04 UTC 2010 i686 GNU/Linux [01:38] Linux beagleboard 2.6.35.4-l1 #1 PREEMPT Fri Aug 27 22:09:57 UTC 2010 armv7l GNU/Linux [01:39] apples<>oranges. [01:40] Unfortunately. [01:40] ?? [01:40] Completely different alsa versions. [01:41] cat /proc/asound/version [01:41] Say, is Tegra2 going to be an officially supported architecture? [01:41] my is 1.0.23 [01:41] rlameiro: on both? [01:41] Supported as in, has kernels built and won't get illegal-instruction errors? [01:41] this is on arm, let me check the board [01:41] DanaG: If someone pushes a kernel into the pool, sure. [01:41] 1.0.21 [01:42] regression? [01:42] rlameiro: very possibly. [01:42] "has kernels built" could even be unofficial, as rcn-ee's OMAPs are/were/are. [01:42] well, then i have to download the ubuntu beta to test it... [01:42] rlameiro: Try downloading one of the ubuntu i386.iso files and run it live. Compare results. [01:42] A bigger concern is the CPU instruction-set support -- can't use NEON on it. [01:43] rlameiro: I would try a daily. [01:43] http://cdimage.ubuntu.com [01:43] DanaG: Shouldn't be an issue. Marvell doesn't either. [01:43] Most apps don't care. [01:43] Marvell doesn't even do armv7l, at least in any currently available products. =/ [01:44] oversized images... [01:44] rlameiro: Write one to USB. faster and doesn't waste a CD. [01:44] Use usb-creator to make a bootable usb stick. [01:44] well downloading [01:45] i need to buy a usb stick for this stuff :D [01:45] 1438 KB/s not bad... [01:45] rlameiro: WIll your laptop boot from SD? [01:46] dont think so [01:46] I'm jealous btw. My max speed is 468KB/s :P [01:46] GrueMaster: never tried it, it does have an removable disk options [01:46] axel rocks :D [01:46] great downloader for the cli [01:47] Where are you on this great rock in space? [01:47] Portugal [01:47] Ah. Closer to our servers. [01:47] UK? [01:47] I'm in Oregon, USA. Much farther for bits to travel. [01:48] Yes. Millbank, I believe. [01:48] rsalveti, you're right. It didn't change anything. Which files do you want? [01:48] i will test some download from Oregon after the download finish [01:48] do you have some link to a mirror from a uni there or something? [01:50] You can try http://members.dsl-only.net/~tdavis/beagleboard-bug.jpg [01:50] lol [01:50] Not very big, but it is my ISP. [01:50] too small to test [01:51] There is a mirror server at http://ubuntu.osuosl.org/ubuntu/ [01:51] but i will try anyway [01:53] GrueMaster: Downloaded 694,7 megabytes in 8:14 seconds. (1437,95 KB/s) [01:53] I hate you. :P [01:54] I have DSL here. And I am paying premium for the fastest speed available. [01:54] jo-erlend: dmesg would help, for sure [01:54] And I still get pidgeon-net speeds. [01:54] yeap, the connection to there is way worse than to the ubuntu server [01:55] GrueMaster: i have DSL to, my sync is about 17 MB/s [01:56] rsalveti, rbelem has quit (Quit: Leaving) [01:56] * o [01:56] bah. :) [01:56] rsalveti, http://pastebin.ubuntu.com/497379/ [01:57] [ 1.497497] omapfb omapfb: no displays [01:57] [ 1.501312] omapfb omapfb: failed to setup omapfb [01:57] probably because of: [01:57] [ 0.269592] omapdss VENC error: can't get VDDA_DAC regulator [01:57] [ 0.269622] omapdss CORE error: Failed to initialize venc [01:58] which points to bug 607250 [01:58] Launchpad bug 607250 in linux (Ubuntu) "omapdss: VDDA_DAC regulator on IGEPv2 (affects: 1) (heat: 65)" [Undecided,Incomplete] https://launchpad.net/bugs/607250 [02:00] rsalveti, do you think changing the kernel would fix this? [02:01] probably not, as this was reported to linux-linaro [02:01] any ideas at all? [02:02] GrueMaster: rebooting, going to test it now [02:03] ok [02:03] jo-erlend: needs to find the missing patch for it [02:04] Is there a page that that pic goes with? [02:08] jo-erlend: could be this one: http://bazaar.launchpad.net/~beagleboard-kernel/%2Bjunk/2.6-stable/annotate/head:/patches/igepv2/0001-ARM-OMAP3-Add-S-Video-output-to-IGEPv2-board.patch [02:08] will build a new kernel for you to test, give me 10 min [02:08] thanks :) [02:15] GrueMaster: well, back [02:16] the card was detected correctly, speaker test gave the same error but i think it has to do with the format of the streaming "SE24_LE" [02:18] Hmmm. [02:18] the iso is not woking ok, it didnt booted completely after selecting try it.. lol i had to use ctrl+alt+F1 :D [02:18] jo-erlend: please test http://people.canonical.com/~rsalveti/kernel/uImage.igep2 [02:18] jo-erlend: just copy it over your uImage from the first partition [02:18] rsalveti, then I simply replace the uImage with that one? [02:18] right. [02:19] then please get your dmesg for me again [02:21] rsalveti: brasileir? [02:21] brasileiro? [02:21] rlameiro: sim, portugues? [02:21] BOA ! [02:21] :-) [02:22] !pt :p [02:22] Factoid 'pt :p' not found [02:22] !pt [02:22] Por favor, use #ubuntu-br para ajuda em brasileiro. Para entrar no canal por favor faça "/join #ubuntu-br" sem as aspas. Para a comunidade local portuguêsa, use #ubuntu-pt. Obrigado. [02:22] persia: :P [02:23] persia: lol, we got it, it was just that moment of hapiness finding a language brother :D [02:23] I know :) [02:24] And since the two of you have such similar interests, I suspect there's going to be a lot of value to portuguese coordination for ARM audio :) [02:24] rsalveti: also into audio? [02:24] haha, not yet, still fighting with display issues :-) [02:24] lol [02:25] i took that out of the picture, [02:25] Hey, alsa on armel sounds the same, regardless of language. :P [02:25] * rlameiro runs is board headless :D ssh rulezzz [02:25] haha, for sure, but we need to get it working for our users haha [02:25] * rlameiro ssh -X even better :D [02:25] haha :-) [02:26] Not when your armel laptop is closer than your desktop :p [02:26] rsalveti, http://pastebin.ubuntu.com/497388 [02:26] jo-erlend: [ 0.000000] Linux version 2.6.35-22-omap (buildd@satinash) (gcc version 4.4.5 (Ubuntu/Linaro 4.4.4-14ubuntu4) ) #32-Ubuntu Fri Sep 17 09:09:03 UTC 2010 (Ubuntu 2.6.35-22.32-omap 2.6.35.4) [02:26] jo-erlend: still not my kernel [02:27] rsalveti, what I did, was to rename uImage to uImage.old and then rename uImage.igep2 to uImage. [02:27] same one as before [02:27] jo-erlend: hm, weird, could be a bug at u-boot [02:27] we got a similar bug with the older u-boot we used for panda [02:28] I should delete uImage.old as well? [02:28] to make sure, copy your files to your pc, give mkfs.vfat -F 32 /dev/sd and then copy them again [02:28] if you're facing the same bug, then it's not going to work even if you remove the older uImage [02:28] rsalveti, I read somewhere that using -F 32 wouldn't work though. It was old, but still. [02:29] well, don't know much of igep2 [02:29] do it as you can find at your igep2 tutorial :-) [02:29] but recreate it and then copy the new files [02:29] what is the problem with it? [02:29] in this case, just copy the new uImage [02:30] rlameiro: display not working for igep2 using default kernel for omap3 [02:30] omapfb? [02:30] yup [02:30] I found a patch but waiting jo-erlend to test it [02:31] this is mine dmesg [02:31] http://pastebin.ubuntu.com/497390/ [02:32] and boot args, sorry for the flood [02:32] fatload mmc 0:1 0x80000000 uImage [02:32] fatload mmc 0:1 0x82000000 uInitrd [02:32] setenv bootargs-base vram=12M omapfb.mode=dvi:hd720-16@50 [02:32] bootm 0x80000000 0x82000000 [02:33] rlameiro: which kernel? [02:33] rsalveti: Linux beagleboard 2.6.35.4-l1 #1 PREEMPT Fri Aug 27 22:09:57 UTC 2010 armv7l GNU/Linux [02:33] rcn-ee one [02:33] rlameiro: cool, the patch I found was applied at his kernel [02:34] and it's not giving you errors [02:34] so it seems to fix it [02:34] once confirmed will send to the kernel team [02:35] i can try to test it tomorrow on my hdmi tv, my girl is sleeping there now..... she will kill me if I wake her up :D [02:35] rlameiro: haha, don't need to worry, I completely understand you :-) [02:37] rsalveti: lol, and I am Ricardo too [02:37] w00t [02:37] #win [02:37] :-) [02:38] rsalveti, http://paste.ubuntu.com/497395 [02:39] hm, weird, same error [02:39] jo-erlend: could you try rlameiro's boot args? [02:39] vram=12M omapfb.mode=dvi:hd720-16@50 [02:41] well it is for a tv, but should be ok [02:42] but now a different error [02:42] rsalveti, I keep the others, I just add those arguments? [02:42] [ 0.269592] omapdss VENC error: can't get VDDA_DAC regulator [02:42] [ 0.269622] omapdss CORE error: Failed to initialize venc [02:42] and now [02:42] [ 0.255493] omapdss SDI error: can't get VDDS_SDI regulator [02:42] [ 0.255493] omapdss CORE error: Failed to initialize SDI [02:43] jo-erlend: change omapfb.mode=dvi:1024x768MR-16@60 to vram=12M omapfb.mode=dvi:hd720-16@50 [02:44] hm, currently you don't have vram [02:44] mem=512M console=ttyS2,115200n8 console=tty0 omapfb.mode=dvi:1024x768MR-16@60 root=/dev/mmcblk0p2 rw rootwait text [02:44] this is your current boot line [02:45] try just adding vram=12M first [02:45] * GrueMaster is calling it a day. [02:45] while a look for an additional patch [02:45] *I [02:51] rsalveti, no change. [02:52] rsalveti, http://paste.ubuntu.com/497402 [02:53] could be another bug =\, just a moment, looking for other patches [02:54] thank you. I appreciate it. [02:54] well. I did change the omapfb.mode setting. I had already done that when you told me to only add the vram setting. Does that matter? [02:55] nops, seems a different bug [02:55] but wanted to check just to be sure [02:55] rsalveti: do you want my uImage? [02:55] rlameiro: no thanks, I got the sources of it already, trying now to understand what's missing [02:59] seems that it's just missing a regulator definition [03:06] is that something you can fix? :) [03:08] jo-erlend: it seems I found the problem, working for a fix [03:09] seems that the cause is another dss2 patch that we got at our tree [03:10] jo-erlend: give me some minutes and I'll send another uImage to test [03:11] great! :) [03:18] building a new kernel [03:19] :) [03:19] it would be really cool if that worked. I think the board is very cute, but it would be nice to be able to use it for something too :) [03:20] sure :-) [03:22] jo-erlend: http://people.canonical.com/~rsalveti/kernel/uImage.igep2-v2 [03:22] same procedure, please :-) [03:23] hopefully at least a different error message [03:23] :-) [03:25] rsalveti, hey, I have something! [03:25] I have text! It's kinda blue! [03:25] quick Q for those "in the know" ... the default netbook launcher (clutter one) ... what ships there for sound ? that depend on pulse ? or is it also using canberra ? [03:25] jo-erlend: awesome [03:25] oh nvm, I can find out via synaptic [03:25] blue should be fine [03:25] at least I get blue text at my panda [03:25] but it works fine on X [03:26] * rsalveti never tested the clutter based netbook launcher [03:26] jo-erlend: can you send me your dmesg [03:27] jo-erlend: even if you get something, to be able to login at your serial you probably need to create a ttyS2.conf file for your upstart [03:27] with correct parameters [03:27] devilhorns, I believe the default stack should be canberra-on-pulse: which layer a given app hits depends on the app. More use of canberra for simple alerts would be appreciated. [03:27] rsalveti, I have no serial :/ [03:27] oh, that's fine [03:27] jo-erlend: so you should be able to see something at your monitor [03:28] persia, ok, just wanted to check if canberra was the way we want to go wrt the efl one [03:28] I get a fatal error, missing files in /lob/modules/blabla. I noticed that before, I have no /lib/modules directory. Am I supposed to? [03:29] jo-erlend: that's ok for now, that's just because I got you just the uImage [03:29] devilhorns, I think we should go that way, as it makes it easier long-term to have global alert management. [03:29] still missing the modules [03:29] jo-erlend: but does it stops at this fatal errors? [03:29] persia, fair enough :) [03:30] rsalveti, it does. [03:30] jo-erlend: ok, let me get you the kernel modules [03:31] rsalveti, and I get lots of "modules does not exist" in my xorg.log. [03:32] hm, could be another issue [03:32] devilhorns, That said, if you know of some reason why it's better not to use libcanberra, I'd be happy to hear arguments, but based on everything I've heard, it's a sensible direction. [03:32] but can you see something from your X11 session? [03:32] I mean, does it show something at your monitor? [03:32] rsalveti, no. [03:33] one step at a time :-) [03:33] jo-erlend: can you paste me your dmesg? [03:33] yes! I have characters on my screen. That's awesome :) [03:34] rsalveti, http://paste.ubuntu.com/497419 [03:35] hmm... [ 0.000000] Linux version 2.6.35.4+ (rsalveti@evatp) (gcc version 4.4.5 20100824 (prerelease) (Ubuntu/Linaro 4.4.4-9ubuntu2) ) #42 Mon Sep 20 22:15:20 BRT 2010 (Ubuntu 2.6.35-21.31-omap 2.6.35.4) <-- Didn't it say that before too? [03:36] yup, but check the timestamp [03:36] it's now different [03:36] persia, no particular reason comes to mind ... I was just curious if canberra was the "preferred" way to go wrt sounds, or if it was just something the netbook-efl devs dropped in [03:36] actually it isn't. It sais 1/1 2000. [03:36] jo-erlend: hm, with error and etc [03:36] hmm? [03:37] so same dmesg as before [03:37] were you able to login and get the dmesg? [03:37] devilhorns, it's currently used in many flavours (Ubuntu Desktop, Ubuntu UEC live, Edubuntu Desktop, Edubuntu Desktop KDE, Xubuntu Desktop, Mythbuntu Desktop, Ubuntu Netbook), so I think it's safe to say it's well-supported in Ubuntu. [03:37] if not you're probably pasting the older one [03:37] no. I had to eject the card and mount it in my laptop. [03:38]