**** BEGIN LOGGING AT Fri Sep 17 02:59:57 2010 Sep 17 04:35:52 tommd: Well, you could get a copy of jasper-initramfs and read what the scripts do. Then apply them manually to an SD card prior to booting. Sep 17 06:05:42 hi , can we mount a directory like this .. "mount -t nfs /sys/kernel/debug/pm_debug/ /avinash/" Sep 17 07:28:58 morning Sep 17 07:33:17 ogra: Do you have sound working on your 6/8 layer? Sep 17 07:33:26 mopdenacker: What about you? Sep 17 07:34:39 prpplague: Are you around still? Sep 17 07:50:36 ... Sep 17 07:50:52 lag, not with the default setup, no Sep 17 07:50:58 (8 layer) Sep 17 07:51:05 Do you have it working at all? Sep 17 07:51:10 Even with the hacky files? Sep 17 07:51:15 i dont fiddle with it, so no Sep 17 07:51:21 :( Sep 17 07:51:24 Okay Sep 17 07:51:28 let me add the files and see Sep 17 07:51:32 I need someone to do some tests Sep 17 07:51:35 what was the bug# ? Sep 17 07:51:38 That would be great Sep 17 07:52:00 bug 637947 Sep 17 07:52:02 Launchpad bug 637947 in linux-ti-omap4 (Ubuntu Maverick) (and 1 other project) "no sound devices on current ES2.0 boards (affects: 1) (heat: 12)" [High,Confirmed] https://launchpad.net/bugs/637947 Sep 17 07:55:20 Let me know if you get sound working Sep 17 07:56:13 rebooting Sep 17 07:56:26 :) Sep 17 07:57:11 its the bottom plug, right ? Sep 17 07:57:17 Apparently Sep 17 07:58:45 i hear farting sounds ... and pops when the amp goes on and off Sep 17 07:58:53 oh, wait, i forgot the daemon.conf Sep 17 07:59:01 * ogra reboots again Sep 17 07:59:30 How did you test? Sep 17 07:59:38 Command please Sep 17 07:59:44 none, i rebooted Sep 17 07:59:49 expecting the login sound Sep 17 07:59:56 Oh, I see Sep 17 08:00:34 aha Sep 17 08:00:41 sounds were disabled by default Sep 17 08:00:57 enabling them gets me the login sound with distrotion Sep 17 08:01:57 * ogra reboots again to hear Sep 17 08:02:58 yep, works Sep 17 08:03:50 but i always get the farting sound if it initializes Sep 17 08:05:15 dropping the daemon.conf changes still gets me sound Sep 17 08:05:31 so they are not urgetnly needed it seems Sep 17 08:06:27 oh, fun, my USB didnt initialize this time Sep 17 08:06:39 (no NIC, no kbd/mouse) Sep 17 08:11:40 lag, so i obviously need the default.pa and the amixer hacks to make it work, it works fine with the original daemon.conf Sep 17 08:12:22 but initializing the HW is very noisy (popping and farting sounds) Sep 17 08:22:44 I am looking for a ARM board with SATA... Sep 17 08:22:55 good luck :) Sep 17 08:25:30 lag, commented on the bug Sep 17 08:25:47 seems there are a lot unneeded changes in default.pa Sep 17 08:25:54 bah Sep 17 08:26:20 lool: ping Sep 17 08:27:15 ogra: It looks like I had network problems - did you get back to me? Sep 17 08:27:23 yep Sep 17 08:27:29 And? Sep 17 08:27:46 lag, commented on the bug, there are many pointelss changes in default.pa Sep 17 08:28:02 all i need are the three lines i pasted there plus the amixer hack Sep 17 08:28:16 and no changes to daemon.conf at all Sep 17 08:28:54 What application plays the login sound? Sep 17 08:29:03 gnome-session i think Sep 17 08:29:07 ogra: actually, you can probably answer my question too Sep 17 08:29:14 dmart, lets try :) Sep 17 08:29:30 ogra: Do this: aplay -Dplughw:0,0 /usr/share/sounds/ubuntu/stereo/desktop-login.wav Sep 17 08:29:34 ogra: quick one: it seems that the libmad fix works - someone got a good test result Sep 17 08:29:47 ogra: should I submit a merge request, and who's the appropriate maintainer? Sep 17 08:30:04 s/maintainer/approver/ Sep 17 08:30:18 dmart, it would have to be a final freeze exception, not sure we'll get it in Sep 17 08:30:37 final freeze is in effect since yesterday ... i think the release team needs to approve Sep 17 08:30:51 lag, /usr/share/sounds/ubuntu/stereo/desktop-login.wav: No such file or directory Sep 17 08:30:59 * ogra goes to look for a wav Sep 17 08:31:50 lag, aplay -Dplughw:0,0 /usr/share/sounds/speech-dispatcher/dummy-message.wav hangs Sep 17 08:32:03 (with my setup though) Sep 17 08:32:08 ogra: since you have a workaround in place, it can go in post-release... is there a development branch, or do we just have to hold off merging it until natty gets created? Sep 17 08:32:13 lag: I don't have my Panda any more :-( Sep 17 08:32:22 mopdenacker: :( Sep 17 08:32:53 dmart, reopen, mark it as ubuntu-later (as milestone) and turn the bug into an SRU request Sep 17 08:33:20 https://wiki.ubuntu.com/StableReleaseUpdates for more info Sep 17 08:33:25 ogra: hi! where is our daily ? ;-) Sep 17 08:33:36 ndec, waiting for the qeue to settle Sep 17 08:33:48 ogra: argh... Sep 17 08:33:53 ogra: Do this: aplay -Dplughw:0,8 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 08:33:58 ndec, https://edge.launchpad.net/builders Sep 17 08:34:04 ndec, armel 19h behind Sep 17 08:34:36 lag, one sec, my ssh console hangs since the last command Sep 17 08:35:07 ogra: > turn the bug into an SRU request Sep 17 08:35:18 ogra: how does that work? Sep 17 08:35:19 dmart, right Sep 17 08:35:30 dmart, https://wiki.ubuntu.com/StableReleaseUpdates for more info :) Sep 17 08:35:39 aha :) thanks ogra Sep 17 08:36:05 dmart, or more detailed https://wiki.ubuntu.com/StableReleaseUpdates#Procedure Sep 17 08:36:36 lag, i think i have to reboot ... while the mouse moves still the system got unresponsive Sep 17 08:37:24 ndec, sadly thats normal after a freeze, there were 100s of uploads the last two days Sep 17 08:38:02 ogra: until we give you bunch of panda ;-) Sep 17 08:38:12 ndec, yeah, i want 50 ! Sep 17 08:38:28 or one for each package in main ... not sure yet :P Sep 17 08:38:36 ogra: ndec: I am fast coming to the conclusion that my board is incapable of sound Sep 17 08:38:45 lag: 6 or 8 layer? Sep 17 08:38:49 lag, thats bad Sep 17 08:38:52 6 Sep 17 08:39:04 didnt you have sound yesterday for a while ? Sep 17 08:39:06 I'd love to work on this bug, but I need a board that can cope :( Sep 17 08:39:14 Only via HDMI Sep 17 08:39:32 I still have sound via HDMI Sep 17 08:39:34 did you have the speaker output in the sound prefs then ? Sep 17 08:39:49 Sorry? Sep 17 08:39:51 well, HDMI should suffice for hacking on the driver Sep 17 08:40:05 No, the HDMI works Sep 17 08:40:20 right Sep 17 08:40:22 Even without the file hacks Sep 17 08:40:26 so you could work on this Sep 17 08:40:29 oh Sep 17 08:40:31 k Sep 17 08:40:46 aplay -Dplughw:0,6 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 08:40:52 That will _always_ work Sep 17 08:40:56 hangs my board here Sep 17 08:41:07 Not hanging, playing? Sep 17 08:41:14 not here Sep 17 08:41:28 Mine plays the robot voice Sep 17 08:41:49 well, as you see above my board hung by it Sep 17 08:42:16 Oh, you may need the high res timers Sep 17 08:42:27 Hang on, I'll send you my kernel Sep 17 08:42:30 in the kernel ? Sep 17 08:42:39 i should be able to dist-upgrade to it Sep 17 08:42:47 afaik the binary is in the archive Sep 17 08:42:53 one sec Sep 17 08:43:06 Since my commit? Sep 17 08:43:28 linux-ti-omap4 (2.6.35-903.12) maverick; urgency=low Sep 17 08:43:28 [ Lee Jones ] Sep 17 08:43:29 * [Config] Enable CONFIG_HIGH_RES_TIMERS=y Sep 17 08:43:31 ... Sep 17 08:43:37 That's the badger Sep 17 08:46:33 that will take a while ... Sep 17 08:46:41 * ogra goes afk for a moment Sep 17 08:48:08 Dear TI, Sep 17 08:48:08 Please send me a 8 layer ES2.0 pandaboard Sep 17 08:48:08 as soon as possible. I would love to fix Sep 17 08:48:09 your audio issues, but I'm finding it hard Sep 17 08:48:09 without working hardware. Sep 17 08:48:10 Kind regards, Sep 17 08:48:12 Lag Sep 17 08:48:21 ;) Sep 17 08:48:36 :D Sep 17 08:48:43 lag: I heard that at the end of month there will be next set of boards shipped to people Sep 17 08:48:54 \o/ Sep 17 08:49:24 who will get them? no idea Sep 17 08:50:37 lag: ;-) Sep 17 08:51:05 lag: i think HDMI does not work out of the box. i am surprised by your conclusion. Sep 17 08:51:37 lag: do you confirm that flashing 20100914, booting and running the aplay command without any other config works? Sep 17 08:56:36 ndec: I will flash a new card and tell you Sep 17 08:58:43 dd'ing Sep 17 08:58:59 dmart: pong Sep 17 09:15:08 lag, bah, no login sound anymore with your kernel Sep 17 09:15:20 well, with the upgrade from the archive Sep 17 09:16:00 Interesting Sep 17 09:16:16 What if you apply the hacky files again? Sep 17 09:16:19 And run the *.sh Sep 17 09:16:31 geez ! Sep 17 09:16:44 and about 500 kerneloops popups on the desktop Sep 17 09:16:49 * ogra clicks madly Sep 17 09:17:01 lag, which .sh ? Sep 17 09:17:33 omap4*.sh Sep 17 09:17:40 whats that ? Sep 17 09:17:51 The one you ran to get sound working Sep 17 09:17:55 is that different from the amixer hack stuff ? Sep 17 09:17:56 The amixer hacks Sep 17 09:18:01 No Sep 17 09:18:02 ah, its in rc.local Sep 17 09:18:17 i'm lazy ... runs automatically on boot Sep 17 09:18:51 hmm, hangs again Sep 17 09:19:00 (without doing anything) Sep 17 09:19:26 i'm starting to think that this amixer carp is at fault Sep 17 09:19:29 *crap Sep 17 09:20:06 * ogra wanted to at least see dmesg for all the oopses :( Sep 17 09:20:24 ndec: I _can_ confirm that HDMI sound works out-of-the-box Sep 17 09:20:53 ogra: How do you mean 'hangs'? Sep 17 09:21:01 Does the screen go black? Sep 17 09:22:31 no, i was greeted with a ton of kerneloops messages on the desktop ... klicked them away ... Sep 17 09:22:37 then logged in via ssh Sep 17 09:22:45 (to easier get to the logs) Sep 17 09:22:50 talked to you here Sep 17 09:23:00 switched to ssh terminal and noticed it hangs Sep 17 09:23:23 switched to omap desktop, clicked on terminal ... desktop hangs, mouse still moves Sep 17 09:23:44 but i seem to be able to switch consoles, i got the dmesg Sep 17 09:23:54 desktop hangs completely now though Sep 17 09:25:22 lag, http://paste.ubuntu.com/495172/ my dmesg has several 100s of these now Sep 17 09:25:22 I still receive the GPU hang _a lot_ Sep 17 09:26:02 and this one at the end http://paste.ubuntu.com/495174/ Sep 17 09:27:06 Yeah, that doesn't look good Sep 17 09:27:10 right Sep 17 09:27:26 let me restore the configs to default Sep 17 09:27:34 and see whats there after reboot Sep 17 09:28:19 rebooting Sep 17 09:28:41 I'm not testing sound with the kernel from the 14th Sep 17 09:28:55 Without High Res' Timers Sep 17 09:29:14 well, i'm running the lkatest from the archive now Sep 17 09:29:36 i just want to make sure the oopses dont show up without the amixer crap Sep 17 09:30:02 I get this: http://paste.ubuntu.com/495177/ Sep 17 09:30:14 funny, i have the farting sound even without amixer stuff Sep 17 09:30:38 lag, yeah, seems i have both too Sep 17 09:30:59 And that's with the straight 14th kernel Sep 17 09:31:20 With default.pa and the script run Sep 17 09:31:20 http://paste.ubuntu.com/495178/ Sep 17 09:31:24 But still no sound Sep 17 09:31:28 fresh boot (right after booting) Sep 17 09:31:36 with all ubuntu defaults restored Sep 17 09:31:43 now thats worrying Sep 17 09:32:14 sigh ! Sep 17 09:32:26 and a kerneloops message on the desktop for every line there Sep 17 09:33:17 How did you restore the defaults/ Sep 17 09:33:18 ? Sep 17 09:33:30 cp Sep 17 09:33:41 and disabling amixer.sh in rc.local Sep 17 09:33:53 oh, sweet! Sep 17 09:34:04 i see audio HW in the sound prefs Sep 17 09:34:15 Aren't the amixer settings persistent? Sep 17 09:34:17 even without these files Sep 17 09:34:19 You do? Sep 17 09:34:23 yes Sep 17 09:34:34 Oh yeah, me too! Sep 17 09:34:35 hmm, might be that the mixer settings are persistent Sep 17 09:34:41 They are Sep 17 09:34:47 sigh... Sep 17 09:34:49 I think they're persistent Sep 17 09:34:58 Hence why I was shocked that you had them in .local Sep 17 09:35:01 so i need a new image (which we wont have until tomorrow) Sep 17 09:35:14 to see the defaults Sep 17 09:35:28 Or reflash 14 Sep 17 09:35:39 Or reflash 14 and apply the new kernel Sep 17 09:35:57 this is my development setup Sep 17 09:36:03 i dont really want to trash it Sep 17 09:36:21 (takes half a day to restore all the build env and chroots) Sep 17 09:36:33 let me look for another SD Sep 17 09:36:51 argh ... and i only have 0915 here Sep 17 09:36:52 I am missing HDMI from Output though Sep 17 09:37:08 In sound perfs Sep 17 09:37:11 yes, thats only provided by the default.pa Sep 17 09:37:24 But I am using default.pa Sep 17 09:37:30 the one from TI ? Sep 17 09:37:32 Oh, no I'm not Sep 17 09:37:32 or ours Sep 17 09:37:47 I saved it as default.pa.new Sep 17 09:37:52 you effectively only need to add the three lines i posted on the bug Sep 17 09:37:55 Let me try it with our one Sep 17 09:38:15 they shoudl give you all devices and not disable essential services like udev Sep 17 09:38:50 or consolekit (thats a really bad one, not sure why TI added that) Sep 17 09:39:10 Okay Sep 17 09:39:13 I'll do that Sep 17 09:39:17 Do you get audio now? Sep 17 09:39:21 no Sep 17 09:39:28 i get the "farting" on boot Sep 17 09:39:31 Then you need to find out what changed Sep 17 09:39:42 which i guess comes from alsa restoring the mixer stuff Sep 17 09:39:52 Yeah Sep 17 09:39:59 what changed is that i restored the original pulse config Sep 17 09:40:08 Change it back Sep 17 09:40:11 Try to get sound again Sep 17 09:40:15 With the new kernel Sep 17 09:40:16 i wont Sep 17 09:40:22 it will kill the kernel Sep 17 09:40:28 err, the desktop Sep 17 09:40:44 We need to find out why you no longer get sound Sep 17 09:41:24 i switched back to my hacked default.pa and am rebooting now Sep 17 09:41:40 but i get the oopses in both cases with the new kernel Sep 17 09:42:29 aha Sep 17 09:42:35 login sound with some distrotion Sep 17 09:43:04 usual oopses in dmesg Sep 17 09:43:09 and additionally ... Sep 17 09:43:12 [ 51.534149] asoc: interface MultiMedia1 hw params failed Sep 17 09:43:20 5 lines of that at the end Sep 17 09:43:56 oh ! Sep 17 09:44:02 Okay, but the sound is there? Sep 17 09:44:04 oh? Sep 17 09:44:09 in the sound prefs i have two input and three output devices now Sep 17 09:44:27 instead of one input and two output which i should have Sep 17 09:44:27 Lucky you :) Sep 17 09:44:34 What happens when you "Test speakers" Sep 17 09:45:39 I have loads of stuff in Sound Perfs now Sep 17 09:46:16 sigh, clicking around in sound prefs a bit hangs the desktop again Sep 17 09:47:04 [ 156.123229] [] (omap_mcbsp_write+0x8/0x18) from [] (omap_mcbsp_config+0x48/0x128) Sep 17 09:47:27 After changing to HDMI and clicking "Test speakers" Sep 17 09:48:06 ogra: I want you to do something for me Sep 17 09:48:17 ogra: It may take 10 mins and a few reboots Sep 17 09:48:29 ok Sep 17 09:48:38 my board hangs hard now btw Sep 17 09:48:49 no console switching, no mouse movement Sep 17 09:49:07 for i in 0 1 2 3 4 5 6 7 8 9; do aplay -Dplughw:0,$1 /usr/share/sounds/speech-dispatcher/dummy-message.wav; done Sep 17 09:49:12 But you have to do it manually Sep 17 09:49:18 That script won't work Sep 17 09:49:27 k Sep 17 09:49:31 i'll try that Sep 17 09:49:38 The script? Sep 17 09:49:42 You won't Sep 17 09:49:46 Oh, wait Sep 17 09:49:48 It might Sep 17 09:50:06 i'm pretty sure only 0,0 and 0,7 have attached output devices Sep 17 09:50:17 No, this is ALSA Sep 17 09:50:21 (since thats what is defined in default.pa) Sep 17 09:50:29 yes Sep 17 09:50:40 default.pa is Pulse Sep 17 09:50:42 the pulse config applies to the alsa stuff Sep 17 09:50:47 That's at a higher level? Sep 17 09:50:51 device=hw:0,0 Sep 17 09:51:01 means pulse uses this exact alsa device Sep 17 09:51:34 and the three lines i have in default.pa define device=hw:0,0 device=hw:0,7 and device=hw:0,2 Sep 17 09:51:45 0,2 for input Sep 17 09:52:18 they need to correspond to -Dplughw:0,7 and -Dplughw:0,0 Sep 17 09:52:32 else i wouldnt have sound at all Sep 17 09:52:51 * ogra waits for the reboot to be done Sep 17 09:53:33 Then why does -Dhw:0,6 work for HDMI out of the box? Sep 17 09:53:55 because the default.pa entries are wrong ? Sep 17 09:54:05 Which ones? Sep 17 09:54:16 the ones provided by ti Sep 17 09:54:40 Nope, out-of-the-box Sep 17 09:54:51 aplay -Dplughw:0,0 /usr/share/sounds/speech-dispatcher/dummy-message.wav works Sep 17 09:55:11 ogra@panda:~$ aplay -Dplughw:0,7 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 09:55:11 aplay: main:654: audio open error: Device or resource busy Sep 17 09:55:15 With TI's files and *.sh? Sep 17 09:55:23 Try 6 Sep 17 09:55:34 geez !!! Sep 17 09:55:39 ? Sep 17 09:55:43 * ogra tried 0.2 Sep 17 09:55:48 that closed my ssh Sep 17 09:56:14 0,2 is input Sep 17 09:56:16 :) Sep 17 09:56:17 and hung the board hard Sep 17 09:56:27 still, that shouldnt hang hard Sep 17 09:56:35 * ogra reboots Sep 17 09:56:36 You just tried to play a track down an input :) Sep 17 09:56:39 ogra: lag: hdmi is 0,6 on current kernel, and is moved to 0,7 on the future kernel. lag is probably using the latest serie of patches from sebjan. but what is in the archive today is different Sep 17 09:56:42 and ? Sep 17 09:56:42 I told you there may be some inputs Sep 17 09:57:09 still, if i send anything to a device that shouldnt hang the board Sep 17 09:57:20 s/inputs/reboots Sep 17 09:57:28 spill an error is fine, but hard hang isnt Sep 17 09:57:32 Nope, but we can sort that out another time Sep 17 09:57:40 yes Sep 17 09:57:50 I'm fairly sure my board is INCAPABLE of sound now :( Sep 17 09:57:57 Not a dicky-bird Sep 17 09:58:00 :( Sep 17 09:58:05 you just sounded above like it would be acceptable to hang it :) Sep 17 09:58:24 It's none of my concern - yet Sep 17 09:58:34 k Sep 17 09:59:24 so 0,0 and 0,6 work Sep 17 09:59:32 0,7 has no device attached Sep 17 09:59:39 I might try to work on these scheduling while atomic bugs until I receive a working board Sep 17 09:59:43 likely because of what ndec said above Sep 17 09:59:47 Yes Sep 17 09:59:53 0,0 do not work for me Sep 17 09:59:59 any other mnumbers you want me to try ? Sep 17 10:00:41 No, but I would like you to do one more thing for me Sep 17 10:00:56 I think I know the answer, but I'd like to try in any case Sep 17 10:01:01 go Sep 17 10:01:16 Do you have any way of flashing a fresh 14? Sep 17 10:01:23 Spare cards etc Sep 17 10:01:34 i need to downgrade the image i have Sep 17 10:01:41 i only have 15 here Sep 17 10:01:51 I thought 15 was broken? Sep 17 10:01:53 which is totally broken (uninstallable) Sep 17 10:01:57 right Sep 17 10:02:17 http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100914/ Sep 17 10:02:32 dude ! Sep 17 10:02:37 i build them :P Sep 17 10:02:41 I know Sep 17 10:02:46 You're just lazy Sep 17 10:02:47 ;) Sep 17 10:03:01 i'm zsyncing to the image to 14 now Sep 17 10:03:04 Do you want me to write you a script to flash your card too? Sep 17 10:03:05 :) Sep 17 10:03:11 zsyncing? Sep 17 10:03:14 sure Sep 17 10:03:21 takes just a few mins Sep 17 10:03:32 since it only downloads the differences Sep 17 10:03:50 11min for 30% of the image apparently Sep 17 10:03:56 Never done that Sep 17 10:04:09 30% is a lot of changes for a day :) Sep 17 10:04:21 i usually get 5% and 2min or so Sep 17 10:04:29 lag, you should Sep 17 10:04:33 its so much faster Sep 17 10:04:42 and saves badwith too Sep 17 10:04:46 *band Sep 17 10:05:23 the only thing you need to take care of is that you keep the old img.gz around Sep 17 10:05:44 then just call zsync Sep 17 10:05:45 i.e. Sep 17 10:05:53 zsync http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100914/maverick-preinstalled-netbook-armel+omap4.img.gz.zsync Sep 17 10:06:04 in the same dir maverick-preinstalled-netbook-armel+omap4.img.gz lives Sep 17 10:06:19 I think I'll give it a miss Sep 17 10:06:32 My images download without me even knowing about it anyway Sep 17 10:06:50 well, even your scripts will be faster using zsync :) Sep 17 10:06:57 :) Sep 17 10:07:11 It's a non-issue :) Sep 17 10:07:32 it wastes bandwith canonical has to pay for Sep 17 10:08:21 * ogra makes a note to talk to pete about lag's next bonus :P Sep 17 10:08:31 Lol - you goon! Sep 17 10:08:35 heh Sep 17 10:08:59 I'll tell them to take off the £0.05p per week Sep 17 10:09:02 :) Sep 17 10:09:05 heh Sep 17 10:09:35 hmm, actually if i have to write an image anyway ... i always wanted to see if zcat to mmcblk0 works instead of dd Sep 17 10:09:45 * ogra will try that Sep 17 10:09:57 4min to go ... Sep 17 10:10:01 *twiddle* Sep 17 10:10:35 With what? Sep 17 10:10:42 ? Sep 17 10:10:45 the zsync Sep 17 10:11:11 or do you mean what i twiddle ? Sep 17 10:11:16 ;) Sep 17 10:11:16 toes indeed Sep 17 10:11:40 * ogra has mselly feet and needs the fresh air Sep 17 10:11:54 *smeally Sep 17 10:11:56 bah Sep 17 10:12:41 Yuck! Sep 17 10:15:01 zsync done Sep 17 10:17:22 root@osiris:/var/build/images# zcat maverick-preinstalled-netbook-armel+omap4.img.gz >/dev/mmcblk0 ... Sep 17 10:17:25 wow, that works Sep 17 10:18:09 and puts a lot less load on my computer than dd it seems Sep 17 10:19:55 i wonder if there is a sudo way Sep 17 10:20:09 * ogra will try for the next write Sep 17 10:20:24 that should make our install instructions a lot easier Sep 17 10:22:43 hmm, still running Sep 17 10:24:34 ogra: sudo sh -C 'zcat xxx > /dev/mmcblkxx' Sep 17 10:24:44 ndec, yeah Sep 17 10:24:50 i was thinking about tee Sep 17 10:25:11 ogra: zcat xxx | sudo tee /dev/mmcblk ? Sep 17 10:25:17 indeed invoking a subshell will always work Sep 17 10:25:22 no, that doesnt Sep 17 10:25:38 tried that and then resorted to root to not make lag lag Sep 17 10:28:29 ndec, so any idea why your default.pa switches off both module-detect lines and consolekit ? thats really superfluous Sep 17 10:28:52 you really only need to add the sinks to make it work Sep 17 10:30:29 ogra: i think it used to break, but that might have been with older kernel. berco is looking at this. Sep 17 10:31:29 should be fine to just have a three line change Sep 17 10:31:57 consolekit should definitely stay enabled Sep 17 10:32:19 and module-detect wont do any harm since we define the sinks hardcoded anyway Sep 17 10:32:35 ogra: i'm gathering more internal information on this. From my experience, this line can be re-enabled Sep 17 10:33:04 ogra: Have you tried it yet? Sep 17 10:33:07 berco, yeah, from my testing it doesnt change a thing (at least on panda) Sep 17 10:33:15 lag, still waiting for oem-config to finish Sep 17 10:33:35 yeah, except we can get traces in syslog with this line :) Sep 17 10:33:43 k Sep 17 10:34:04 When you have it working, try the aplay commands on 0,0 and 0,6 Sep 17 10:34:20 lag, without any changes to the system ? Sep 17 10:34:29 Correct Sep 17 10:34:32 (old kernel, old pulse, no amixer hack) Sep 17 10:34:43 k Sep 17 10:40:30 ogra@panda:~$ aplay -Dplughw:0,0 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 10:40:30 aplay: main:654: audio open error: Invalid argument Sep 17 10:40:30 ogra@panda:~$ aplay -Dplughw:0,6 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 10:40:30 Playing WAVE '/usr/share/sounds/speech-dispatcher/dummy-message.wav' : Unsigned 8 bit, Rate 8000 Hz, Mono Sep 17 10:40:33 lag ^^^ Sep 17 10:40:52 no output (due to missing amixer settings i guess) Sep 17 10:41:13 but seems HDMI (0,6) is attached by default here Sep 17 10:42:48 oh, intresting Sep 17 10:42:51 ogra@panda:~$ aplay -Dplughw:0,4 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 10:42:51 Playing WAVE '/usr/share/sounds/speech-dispatcher/dummy-message.wav' : Unsigned 8 bit, Rate 8000 Hz, Mono Sep 17 10:42:51 ALSA lib pcm_params.c:2150:(snd1_pcm_hw_refine_slave) Slave PCM not usable Sep 17 10:42:51 aplay: set_params:1031: Broken configuration for this PCM: no configurations available Sep 17 10:43:01 seems there is something attached to 0,4 Sep 17 10:43:19 aha, and to 0,7 too Sep 17 10:44:20 lag, does that help you ? Sep 17 10:44:36 * ogra is brave and tries 0,2 again to see if the old kernel crashes too Sep 17 10:44:51 doesnt :) Sep 17 10:45:03 ogra@panda:~$ aplay -Dplughw:0,2 /usr/share/sounds/speech-dispatcher/dummy-message.wav Sep 17 10:45:03 aplay: main:654: audio open error: Invalid argument Sep 17 10:45:33 Okay Sep 17 10:45:42 oh, i take that back Sep 17 10:45:48 I wonder what ALSA does to connect 0,0 Sep 17 10:45:50 hangs now Sep 17 10:45:58 What kind of hang was it Sep 17 10:46:00 Check the logs Sep 17 10:46:07 quiet hard lock Sep 17 10:46:19 Logs please Sep 17 10:46:33 how ? Sep 17 10:46:39 demsg will be gone after reboot Sep 17 10:47:03 I'm assuming you haven't glued your card into your machine? Sep 17 10:47:34 * ogra nibbles on the glue to get ot out of the slot again Sep 17 10:47:38 bah, you got me Sep 17 10:47:49 ;) Sep 17 10:48:27 omap_mcbsp_write? Sep 17 10:49:41 http://paste.ubuntu.com/495222/ Sep 17 10:49:47 the end of syslog Sep 17 10:49:57 /var/log/dmesg has no errors Sep 17 10:50:34 Sep 17 10:40:06 sycamore kernel: [ 727.309112] Wrapper Enabled... Sep 17 10:50:34 Sep 17 10:40:06 sycamore kernel: [ 727.327941] Start audio transfer... Sep 17 10:50:34 Sep 17 10:40:27 sycamore kernel: [ 747.851928] Wrapper disabled... Sep 17 10:50:34 Sep 17 10:42:14 sycamore kernel: [ 855.361022] omap_device: omap-mcpdm-dai.-1: new worst case activate latency 0: 30517 Sep 17 10:50:42 thats the end of messages Sep 17 10:51:03 (before there is only installer stuff) Sep 17 10:51:34 kernlog please Sep 17 10:51:38 kern.log Sep 17 10:52:01 http://paste.ubuntu.com/495225/ Sep 17 10:52:40 nothing that sticks out there Sep 17 10:53:03 (above that only hdmi init stuff and boot messages) Sep 17 10:55:02 Hmm Sep 17 10:55:26 You should provide yourself with a serial console and try again Sep 17 10:55:55 bah, serial Sep 17 10:56:00 * ogra fiddles Sep 17 10:59:25 wow ! Sep 17 10:59:42 even when loggin in through ssh i already get Sep 17 10:59:44 [ 21.537811] asoc: no valid backend routes for PCM: SDP4430 Media Sep 17 10:59:47 about ten times Sep 17 11:00:06 Yes, I know about that Sep 17 11:00:32 ogra: yes, the driver routes audio over ssh when it sense the tcp connection Sep 17 11:00:36 * amitk ducks Sep 17 11:00:42 lol Sep 17 11:01:03 i wish ... that would have made my live in LTSP a lot easier back then :) Sep 17 11:01:42 lag, hmm, intresting Sep 17 11:02:39 ? Sep 17 11:03:06 [ 93.762878] asoc: no valid backend routes for PCM: SDP4430 Voice Sep 17 11:03:07 [ 100.084808] asoc: no valid backend routes for PCM: SDP4430 Tones Playback Sep 17 11:03:26 and also Sep 17 11:03:29 [ 136.720489] asoc: no valid backend routes for PCM: SDP4430 MODEM Sep 17 11:04:09 0,2 retuns Voice Sep 17 11:04:43 oho, 0,3 gets me an oops Sep 17 11:04:59 http://paste.ubuntu.com/495233/ Sep 17 11:05:17 Perhaps default.pa links back and front ends Sep 17 11:05:23 and the hang too Sep 17 11:05:41 default.pa doesnt link a think in the default setup Sep 17 11:05:51 it uses what udev provides as 0,0 usually Sep 17 11:06:44 or falls back to static alsa detection if udev Sep 17 11:06:48 is missing Sep 17 11:12:29 lag, so [ 28.112762] asoc: no valid backend routes for PCM: SDP4430 Media comes from the desktop coming up Sep 17 11:12:39 which means pulse defaults to 0,0 Sep 17 11:13:56 i can reproduce the message with aplay on 0,0 Sep 17 11:15:56 Sounds about right Sep 17 11:16:41 so getting that connected to something by default would give us headphones by default Sep 17 11:24:45 * ogra goes to find some food ... somehow i missed breakfast again and my friday meeting marathon starts shortly Sep 17 11:26:54 ogra: Do me something else Sep 17 11:27:00 Run the *.sh and try again Sep 17 11:27:11 If that works, it could just be a volume issue Sep 17 11:40:43 It looks like HDMI does _not_ use the CODEC Sep 17 11:40:53 Hence why it works out of the bo Sep 17 11:40:53 x Sep 17 11:41:01 That wasn't a kiss :) Sep 17 11:41:31 aha Sep 17 11:41:42 lag, will test if i'm upstairs again Sep 17 11:43:11 f? Sep 17 11:43:13 if? Sep 17 11:43:20 once Sep 17 11:45:36 when? Sep 17 11:45:49 geez, after i finished eating Sep 17 11:46:29 No Sep 17 11:46:47 I mean "will test _when_ i'm upstairs again" Sep 17 11:46:58 I'm not rushing you Sep 17 11:46:59 :) Sep 17 11:47:33 Just correcting your England Sep 17 11:47:45 ah Sep 17 11:48:00 heh Sep 17 11:48:13 Oooo Sep 17 11:49:04 ? Sep 17 11:49:09 got sound ? Sep 17 11:53:37 re Sep 17 11:53:46 Nope Sep 17 11:54:01 My board doesn't do sound :) Sep 17 11:55:59 soooo Sep 17 11:56:05 i hear sound Sep 17 11:56:18 aplay -Dplughw:0,0 /usr/share/sounds/speech-dispatcher/dummy-message.wav ... Sep 17 11:56:23 after runnign the script Sep 17 11:56:26 Just by running the *.sh? Sep 17 11:56:35 * ogra logs out and in again to see Sep 17 11:56:37 yeah Sep 17 11:56:59 With the old default.pa and daemon.conf? Sep 17 11:57:07 That's exciting Sep 17 11:57:33 That means it's just a volume issue Sep 17 11:58:54 well, while i see the in/output in sound prefs and also the HW i cant get system sounds Sep 17 11:59:05 so its a routing issue too i guess Sep 17 11:59:20 but aplay definitely works with just the mixer stuff Sep 17 11:59:44 We can sort that out separately Sep 17 11:59:59 hmm Sep 17 12:00:02 Surely that's a userspace issue? Sep 17 12:00:08 aplay doesnt spit out anything anymore now Sep 17 12:00:14 Oh? Sep 17 12:00:18 What did you do? Sep 17 12:00:22 seems pulse grabbed the device or something Sep 17 12:00:37 i logged out and in again Sep 17 12:00:51 then trying aplay on the ssh console ... no output Sep 17 12:01:25 aha, because it hangs again Sep 17 12:02:08 http://paste.ubuntu.com/495264/ Sep 17 12:02:16 serial ... Sep 17 12:02:43 I'm working on that right now Sep 17 12:02:55 k Sep 17 12:02:57 So it's just a volume issue then? Sep 17 12:03:02 so i can reboot ? Sep 17 12:03:06 You have to Sep 17 12:03:22 i know, just wanted to know if you need other info before i do :) Sep 17 12:03:31 * ogra reboots Sep 17 12:05:10 davidm: ndec: Can I solder my board? Sep 17 12:05:36 lag, here I am Sep 17 12:05:52 Hi David Sep 17 12:05:55 Thanks for coming :) Sep 17 12:06:10 Right, so we can get sound by only fiddling with amixer Sep 17 12:06:16 okay Sep 17 12:06:18 How does amixer work Sep 17 12:06:26 Who does it talk to? Sep 17 12:06:48 amixer talks to alsa-lib, which makes ioctl's to the kernel Sep 17 12:07:26 Where into the kernel? Sep 17 12:07:39 Basically, I want to do inside the kernel what amixer is doing from userspace Sep 17 12:08:29 okay Sep 17 12:08:46 there is also alsactl init if you want to run amixer on startup Sep 17 12:09:11 I'd like to use that as a fallback Sep 17 12:09:36 but getting things right in the kernel, I assume that would work too Sep 17 12:09:52 I'm guessing there is no way to revert amixer settings? Sep 17 12:10:07 We run a script (which I sent to you) to change them all Sep 17 12:10:26 I'm assuming it's not so easy to change them all back (so I can compare them) Sep 17 12:10:33 revert? Naah, but you can save and restore them with alsactl Sep 17 12:10:58 Cool Sep 17 12:11:04 I need a fresh image then Sep 17 12:11:06 Bear with me Sep 17 12:11:10 in the standard ubuntu distro, this is done for you automatically Sep 17 12:11:23 we noticed :) Sep 17 12:11:44 lag, you might be able to delete the state file and reboot Sep 17 12:11:53 or...let me think...it might be that Daniel T Chen just uploaded a version which *doesn't* do that Sep 17 12:12:01 I think they are persistent over reboot? Sep 17 12:12:07 i have asouns.state here Sep 17 12:12:11 *asound Sep 17 12:12:20 i think thats what stores the settings Sep 17 12:12:48 just before the Maverick freeze Sep 17 12:13:13 ogra@panda:~$ ls -l /var/lib/alsa/asound.state Sep 17 12:13:13 -rw-r--r-- 1 root root 15690 2010-09-17 14:05 /var/lib/alsa/asound.state Sep 17 12:13:13 ogra@panda:~$ date Sep 17 12:13:13 Fri Sep 17 14:12:53 CEST 2010 Sep 17 12:13:19 matches my last reboot Sep 17 12:13:29 diwic: What is the correct way to set ALSA state Sep 17 12:13:58 look in /etc/init/alsa-mixer-save.conf Sep 17 12:13:59 How about if we just had a correct asound.state file? Sep 17 12:14:09 lag, good question. There is a volume database somewhere... Sep 17 12:14:29 lag, wont work, we'd break an existing package with that Sep 17 12:14:37 and it would be more hackish than the script Sep 17 12:14:53 And you think hacking the kernel isn't "hackish"? Sep 17 12:14:58 the proper way for such workarounds are alsa quirks as persia said in the bug Sep 17 12:15:14 if we cant do kernel, we need to do quirks Sep 17 12:15:32 diwic: So it would go: Sep 17 12:15:44 /usr/share/alsa/init Sep 17 12:15:57 right Sep 17 12:16:13 diwic: amixer -> alsalib -> (ioctl) -> ? -> sound chip -> CODEC Sep 17 12:16:14 ? Sep 17 12:16:20 and a line in 00main to sourfce an omap4 setup Sep 17 12:16:59 isnt CODEC before sound_chip ? Sep 17 12:18:03 lag, I'm not into soc that much, but I would say that your question mark should be replaced by "snd" or "snd-soc-core" module (if there is one of the latter) Sep 17 12:18:30 ogra: I don't believe that to be true Sep 17 12:18:51 k Sep 17 12:18:54 I think the volume settings are _on_ the CODEC Sep 17 12:18:56 it depends on what you mean with "sound chip" I guess Sep 17 12:19:06 lag probably meant the controller Sep 17 12:19:16 i meant physical HW Sep 17 12:19:25 They are both physical HW Sep 17 12:19:36 well, attached to the in/output device Sep 17 12:19:42 The sound chip is connected to the CODEC Sep 17 12:20:05 k Sep 17 12:20:11 lag, technically, both the controller and the codec are "sound chips" :-) Sep 17 12:20:28 * ogra doesnt know much ablow alsa Sep 17 12:20:33 *below Sep 17 12:22:39 6-12 months from now there might be an additional component called "Use Case Manager" telling people how they should set up the mixer for different use cases Sep 17 12:22:46 but that's a different story Sep 17 12:23:21 For now I assume alsactl init (i e /usr/share/alsa/init ) should work Sep 17 12:23:57 ndec: Why isn't the sdpXXXX mentioned on either the schematic or datasheet? Sep 17 12:25:50 I guess doing it in the kernel would work as well, but I'm not really familiar with how to do it there. I assume that you could hard-code communication with the codec in the initialization. Sep 17 12:28:10 diwic: alsa scenarios got finally implemented? Sep 17 12:29:05 hrw, no, we're blindly poking sticks into it and see if it survives Sep 17 12:29:12 hrw, scenarios? Well, I assume people have talked about it before, but there is implementation undergoing development, I don't think it is merged into alsa yet Sep 17 12:30:11 hrw, you can search alsa-devel for "UCM" or "Use Case Manager" this month or the month before that Sep 17 12:47:21 ogra: getting a lot of heat from OE-apologists? :) Sep 17 12:48:08 amitk, yeah, enetertaining, isnt it ? Sep 17 12:56:10 i've customised an ubuntu livecd, i would like to know if there's any way to disable that windows which gives us these options: "Try ubuntu" or "Install ubuntu". if it's an off-topic question, please let me know. Sep 17 12:57:28 * rcn-ee has never won an argument with mru either... ;) Sep 17 13:00:18 prpplague: Are you around yet? Sep 17 13:01:03 ogra: just give him an old board and get it over with :-p Sep 17 13:01:23 amitk, i'd love to :) Sep 17 13:01:34 its just to entertaining ... Sep 17 13:01:37 cant stop :) Sep 17 13:11:34 is anyone here using eclipse for cross development? Sep 17 13:12:03 I am wondering which sort of plugins is people using Sep 17 13:15:43 hi, my debug fs is mounted in /sys/kernel ... checked /etc/fstab, not moutned from here .. any hints from where it is moutned ?? script file ?? Sep 17 13:18:51 avinashhm: mount helps? Sep 17 13:18:53 avinashhm: try "sudo mount" (that shows the current mounted file systems) and look at that one Sep 17 13:19:30 pcacjr__: i believe no need sudo for calling mount Sep 17 13:19:37 zumbi: indeed Sep 17 13:19:42 zumbi: my bad Sep 17 13:19:44 sorry Sep 17 13:19:54 no prob :) Sep 17 13:20:00 :-) Sep 17 13:20:06 ichecked mount .. its mounted in /sys/kernel .. don't know from where this is getting moutned .. not fstab ..checke tht ?? Sep 17 13:20:10 rsalveti: wake up lazy brazilian Sep 17 13:21:06 avinashhm: isn't it include in the initrd ? Sep 17 13:21:16 Doh - included Sep 17 13:21:27 pcacjr__, let me check ... Sep 17 13:21:51 avinashhm: sudo mount -o loop initrd /mnt/foo Sep 17 13:25:19 pcacjr__, there was no initrd .. but found the file .. its mountd from mtab ../etc/mtab ... Sep 17 13:25:44 avinashhm: cool, good to know too Sep 17 13:26:21 ogra: How do you run your amixer.sh? Sep 17 13:26:33 sudo sh amixer.sh currently Sep 17 13:26:34 pcacjr__: thanx .... Sep 17 13:26:47 avinashhm: nope Sep 17 13:26:47 So you're using the old version of the script? Sep 17 13:26:48 * ogra was to lazy to make it executable Sep 17 13:26:58 no, i copy pasted from the bug Sep 17 13:27:07 into a file called amixer.sh Sep 17 13:27:49 Which lines? Sep 17 13:28:04 all Sep 17 13:28:17 the whole sript there Sep 17 13:28:21 *script Sep 17 13:28:29 And you don't run amixer.sh -a? Sep 17 13:28:45 just sh amixer.sh Sep 17 13:28:52 And that does what? Sep 17 13:29:20 http://paste.ubuntu.com/495301/ Sep 17 13:29:25 I think that does the same as -a Sep 17 13:29:31 i guess so Sep 17 13:29:38 Yes, same Sep 17 13:29:46 Can you reflash? Sep 17 13:29:52 With 14 Sep 17 13:30:02 Then do sh amixer.sh -s? Sep 17 13:30:31 * ogra has to prepare for a meeting but i'll try to do that on the side Sep 17 13:30:44 Much appreciated Sep 17 13:32:10 zcat running Sep 17 13:32:16 Once you've done it, see if you can get sound without rebooting Sep 17 13:32:34 I think amixer should turn the volume up without the need for a restart Sep 17 13:32:52 yes, it applies immediately Sep 17 13:32:54 Remember to forget to copy over the other files (that we don't need) Sep 17 13:32:54 hm, lots of audio stuff for today Sep 17 13:33:10 * lag has no hair left Sep 17 13:33:13 rsalveti, and flamewars in #beagle Sep 17 13:33:14 what reminds me we should try to ping persia Sep 17 13:33:25 haha Sep 17 13:33:32 * rsalveti looks at #beagle Sep 17 13:33:32 surely #beagle is the more entertaining channel today :) Sep 17 13:33:48 * ogra never saw amitk getting rude before Sep 17 13:34:58 lag, the script definitely applies immediately, i dont need to test that, i can tell you Sep 17 13:35:18 ogra, If you need root to run amisxer.sh something is wrong: it ought just work. Sep 17 13:35:23 the thing is that it clashes with pulse as soon as you restart the session (or reboot) Sep 17 13:35:33 persia, i didnt check :) Sep 17 13:35:40 just ran it with sudo Sep 17 13:35:55 ogra: sometimes an offence is the best defence Sep 17 13:36:04 amitk, yeah ... :) Sep 17 13:36:14 persia: there you go Sep 17 13:36:17 good "morning" Sep 17 13:36:48 * persia claims to be in UTC-7 today Sep 17 13:36:50 amitk, i liked it :) Sep 17 13:38:29 lag, if you need to. Why? Sep 17 13:39:18 prpplague: Thinks my audio filter may be broken Sep 17 13:39:24 Doh Sep 17 13:39:35 davidm: prpplague thinks my audio filter may be broken Sep 17 13:39:53 I need to turn this: http://imagebin.ca/view/C-Otvqp.html Sep 17 13:39:57 NCommander, around already ? Sep 17 13:40:07 Into this: http://imagebin.ca/view/rVkPnEG.html Sep 17 13:40:10 i you are comfortable trying the fix have at it Sep 17 13:40:51 Eh? Sep 17 13:41:05 now i understand why they made the board balck ! Sep 17 13:41:08 *black Sep 17 13:41:17 Well it doesn't work in it's current state Sep 17 13:41:22 It can't get any worse? Sep 17 13:41:28 I need an 8 layer really Sep 17 13:41:38 There is no filter on the 8 layer Sep 17 13:43:09 So you need to remove L28 Sep 17 13:43:42 And C205 Sep 17 13:43:55 ping GrueMaster Sep 17 13:45:09 I think just L28 Sep 17 13:45:15 But I will clarify with prpplague Sep 17 13:47:30 ogra: indeed a flame Sep 17 13:47:32 ogra: don't do this, ubuntu will get blamed if he quits computers ;) Sep 17 13:47:42 heh Sep 17 14:04:40 ogra: yes, that is not good way to clean (improve) your employers name Sep 17 14:10:03 zumbi: I have a minimal kernel for the efikamx Sep 17 14:10:35 zumbi: working on the ethernet support since I like doing nfsroot Sep 17 14:10:43 amitk: those are great news, is it published? Sep 17 14:11:05 zumbi: I'll post patches soon, perhaps this weekend if I get around to hacking Sep 17 14:11:18 amitk: i do not mind to do some early testing Sep 17 14:11:36 the damn thing has ethernet hanging off of usb, so it needs usb initialisation to get ethernet working Sep 17 14:11:52 yes, that is why uboot is borked Sep 17 14:12:14 zumbi: will email you an early bird patch to 2.6.36-rc3 Sep 17 14:12:23 http://wiki.debian.org/EfikaMX#Limitations Sep 17 14:12:33 yeah :-/ Sep 17 14:12:44 amitk: i'll be happy to build and test Sep 17 14:14:03 amitk: aren't all armv7 boards the same? Sep 17 14:14:09 except dove Sep 17 14:14:20 and no clue about omap4 Sep 17 14:14:39 armin76: what are you talking about? Sep 17 14:15:01 zumbi: needing usb init for eth Sep 17 14:15:04 since eth is usb based Sep 17 14:15:29 oh! yeap! I was planning to move to barebox at some point Sep 17 14:15:47 armin76: not really Sep 17 14:16:18 amitk: examples? Sep 17 14:16:20 armin76: most SoCs have ethernet IP on them, e.g. the imx51 has a FEC IP Sep 17 14:16:42 but that isn't being used on the efikamx (it is used on the babbage though) Sep 17 14:16:59 much simpler init when usb is not involved Sep 17 14:17:28 the efika has BT, WLAN and ethernet hanging from usb. So no connectvity unless usb works Sep 17 14:17:43 * amitk gets signs of life from the usb driver Sep 17 14:18:15 uhm... i had no idea, wlan and bt hanged on usb, maybe that explains wlan failures Sep 17 14:18:31 fail Sep 17 14:19:30 well in any case, mainline kernel is good news :) Sep 17 14:31:47 ndec, http://imagebin.ca/view/C-Otvqp.html -> http://imagebin.ca/view/rVkPnEG.html Sep 17 14:32:40 removing is always easier Sep 17 14:32:42 :-) Sep 17 14:35:02 lag, anything else? I'm about to call it a day Sep 17 14:37:01 lag: yea just remove L28 and short across the two sets of pads Sep 17 14:37:22 prpplague, oh, you are up Sep 17 14:37:36 ogra: just got to the office, forgot to log off irc last night Sep 17 14:37:49 we're just discussing possible fixes in the call Sep 17 14:38:07 morning Sep 17 14:38:08 sinbc eit seems ndec's team sees the same issues Sep 17 14:38:11 ogra: ahh ok Sep 17 14:38:36 ogra: basically L28 needs to be removed and the two pads shorted across Sep 17 14:38:53 ogra: i thought all the 6-layer boards had been done that way Sep 17 14:38:55 seems lag's board looks slightly different though Sep 17 14:39:19 What about the surrounding components that you've removed? Sep 17 14:39:20 ogra: than the picture i posted? Sep 17 14:39:42 we mande a fun little discovery this week. u-boot doesn't invalidate dcache .. so if you aren't using the arm monitor (or other very early bootrom to setup the a9) kernel dies when it enables the MMU Sep 17 14:39:51 lag: the other components don't need to be removed, that board was one of my test boards that i was testing some noise issues with Sep 17 14:40:01 lag: L28 is the only item that you should have to change Sep 17 14:40:08 which board is this that needs the rework? Sep 17 14:40:34 Martyn: 6-layer boards Sep 17 14:40:38 --^ Sep 17 14:40:46 * lag warms up the iron Sep 17 14:40:48 prpplague: I meant, which manufacturer/platform Sep 17 14:41:25 * Martyn has a desk full of boards.. everything from those lange boards, to the tegra 250 boards, to the new omap 4xxx boards. Sep 17 14:41:53 Martyn: Panda Sep 17 14:42:03 Martyn: sorry, panda board Sep 17 14:42:59 thanks. Sep 17 14:44:45 STE U8500 arrived today. Sep 17 14:44:54 Can't even think about looking at it until next week Sep 17 14:48:33 https://launchpad.net/builders there are only 9 armel buildds?? Sep 17 14:48:58 no, 7 Sep 17 14:53:40 prpplague: I managed to bridge one pair, but not the other Sep 17 14:53:42 :( Sep 17 14:53:47 How did you do it? Sep 17 15:06:56 lag: just removed the filter, cleaned the area with solderwick, then flux it Sep 17 15:07:11 lag: then solder bridge across Sep 17 15:11:59 * Martyn will use some 0 ohm resistors Sep 17 15:50:07 has anyone been seeing gfx underruns on some of the older builds of ubuntu rootfs? Sep 17 15:50:27 not only on the older ones Sep 17 15:50:45 my screen usually turns black once every two days or so Sep 17 15:50:53 and i have to switch consoles to get it back Sep 17 15:50:54 ogra: anyone looking at the root cause? Sep 17 15:51:17 niocolas said someone at TI is Sep 17 15:53:59 * prpplague wonders who Sep 17 15:55:46 prpplague: maybe mythripk should know better Sep 17 16:02:33 This could be an issue where the monitor is in deep sleep and the system isn't sending a wakeup before reading edid. I'm not seeing it here, but I am only on a DVI monitor. Sep 17 16:03:17 GrueMaster, no, it happens even while i use it Sep 17 16:08:37 Oh. That is strange. Sep 17 16:17:23 argh, apt, dpkg, logrotate, man-db, mlocate, just some of the daily process that consumes all your io Sep 17 16:17:34 can't use my beagle while running these services :-( Sep 17 16:31:32 ogra, got a minute ? Sep 17 16:35:18 devilhorns: not sure if ogra is on-line still, but who knows Sep 17 16:35:24 ahh ok Sep 17 16:35:37 rsalveti, ok, maybe you can answer this quickly for me then ... Sep 17 16:36:11 I have more patches for the netbook-launcher-efl ... but "Bugs" do not exist for these patches ... should I add new bugs ? or just email the patches to someone ? Sep 17 16:36:40 devilhorns: if we plan to add this fixes for the release, than we should have bugs for every fix Sep 17 16:36:49 so we can try to make exceptions and get into our release Sep 17 16:37:03 if they are fixes for the next one, than you could just maintain them at your own tree Sep 17 16:37:07 and let us know about it Sep 17 16:37:13 hrm ... but I don't know what the "plans" are :) Sep 17 16:37:18 devilhorns, Please create new bugs. Sep 17 16:37:18 maybe creating a ppa or something like that to help us testing Sep 17 16:38:13 devilhorns, The "plan" is to make everything shiny and nice. Extra points for feature-equivalence for 3D and 2D solutions. Sep 17 16:38:26 I'd say it all depends on the kind of fixes you have Sep 17 16:38:27 well, in the grand scheme, they are fairly minor really ... (currently does not handle some background options correctly (scale, zoom, wallpaper, etc), and also does not install the schema file either) Sep 17 16:38:43 if you rewrite everything, than just a bug to create a new release :-) Sep 17 16:38:55 rsalveti, well, also doing that on the side :) Sep 17 16:39:03 new version will have both 2d and 3d Sep 17 16:39:09 awesome Sep 17 16:39:15 and now we can test 3d with sgx Sep 17 16:39:26 efl should just fly with sgx Sep 17 16:39:42 at least raster did some work with it Sep 17 16:39:47 rsalveti, efl flies on anything :) ... my toaster is proof :P Sep 17 16:39:58 yeah... you're right :-) Sep 17 16:39:59 procedurally it might be easiest to file two bugs: one for everything that simply *must* be fixed to be release-quality in maverick, and one for an upgrade that should happen when natty opens Sep 17 16:40:32 yep, probably the best option Sep 17 16:40:39 then we can try to still push some fixes for the release Sep 17 16:41:26 wondering if it's really worth it for these 2 small patches tho ... considering that I am rewriting it anyway ... hrm Sep 17 16:41:54 I mean, apparently these 2 minor bugs have not bitten anyone yet Sep 17 16:41:58 the archive is Frozen, so if something isn't release critical it's not going to be approved anyway. Sep 17 16:42:00 persia: am I wrong or we're the only ones who should provide an image with the efl interface? Sep 17 16:42:03 as there are no existing bugs for it Sep 17 16:42:16 rsalveti, Doesn't matter: the same release criteria apply for everything. Sep 17 16:42:32 persia: I know that, just wanted to know if we'd affect anyone else Sep 17 16:42:41 persia, ok, I'll just wait then. If they have not been previously reported then apparently they are not "show stoppers" :) Sep 17 16:42:47 I know there are some folks who keep trying to make an enlightenment flavour, but I haven't heard much from them this cycle. Sep 17 16:43:23 devilhorns, makes sense. Best to focus now on critical fixes *OR* reinventing the world for next time. Sep 17 16:44:23 persia, indeed :) I've been fixing everything ogra has assigned to me so far (plus some things not assigned) ... but I'm fairly new to *buntu development so not sure what proper procedures are, or even what is "critical" Sep 17 16:45:09 but seeing as how these 2 issues are minor, I'll just go back to recreating the wheel :) Sep 17 16:45:21 Much of it is a matter of conscience, but generally "critical" falls into categories like "significant regression", "potential for user data loss", "potential for failed upgrade", "security issue", etc. Sep 17 16:45:23 (for next release) Sep 17 16:46:18 persia, yea, these 2 don't fall into any of those categories imo :) Sep 17 16:47:35 devilhorns, In that case, yeah, don't worry about them if you're rewriting the entire thing anyway. Sep 17 16:47:46 indeed :) Sep 17 16:47:52 new one is much nicer so far :) Sep 17 16:49:25 Excellent. Sep 17 16:50:06 yea :) uses less memory already, is faster, and does things correctly Sep 17 16:50:47 cool Sep 17 16:50:56 devilhorns: any chance to test at a real arm board? Sep 17 16:51:09 last time I saw you requested a beagle but don't know the current status Sep 17 16:51:40 rsalveti, not yet ... I am supposed to receive one, but has not arrived yet ... and I don't know the current status of that "issue" (meaning I don't know if it was sent out or anything) Sep 17 16:52:01 devilhorns: ok, np, just wanted to know Sep 17 16:52:18 so we can then ping davidm later Sep 17 16:52:21 rsalveti, no worries :) believe me, when it gets here I'll be testing it :) Sep 17 16:52:41 Isn't EFL sufficiently ported that we oughtn't expect significant differences? Sep 17 16:52:54 yea, should not be any major issues Sep 17 16:53:12 persia: yep, but it'd be nice to see it running faster on beagle Sep 17 16:53:21 because then you can really know if the code is actually faster Sep 17 16:53:22 :-) Sep 17 16:53:28 I know for fact that some high-up efl developers (myself included) do work for another company that is putting efl on arm based phones Sep 17 16:53:37 I guess. beagle doesn't meet minimum Ubuntu specs anyway (mind you, I'm using one for porting, but still...) Sep 17 16:54:02 so I don't see efl having any major problems on ARM hardware :) Sep 17 16:54:10 devilhorns: yeah, I know many hehe Sep 17 16:54:21 rsalveti, ok, so you already know what's going on then ? :) Sep 17 16:54:41 not every details, but I do sometimes get some updates :-) Sep 17 16:54:57 most of the time when drinking beers with the right people ;-) Sep 17 16:55:20 rsalveti, gotcha :) well sadly I'm under nda and cannot say a whole lot about the issue ... other than "I don't see efl having problems on arm arch" :) Sep 17 16:55:37 devilhorns: yep, understandable :-) Sep 17 16:55:49 rsalveti, ahhhh, so you know some efl devs personally then ? :) Sep 17 16:56:21 devilhorns: for now most of the people at profusion and some others that doesn't actually do much work nowadays Sep 17 16:56:42 rsalveti, ahhh, gotcha Sep 17 16:56:48 persia: I know beagle doesn't meet minimum Ubuntu specs, but would be nice to see it running there Sep 17 16:56:49 I miss rephorm :( Sep 17 16:57:10 as efl should just fly on it, even with lower specs Sep 17 16:57:27 yup :) Sep 17 16:59:14 devilhorns: we had some nice people working on efl at indt, but then nokia decided to buy qt... Sep 17 16:59:27 but now at least we have some other companies putting money on it Sep 17 16:59:36 ahh yea, I remember indt ... started the original webkit-efl stuff Sep 17 17:00:12 yep, canola, webkit-efl and other cool stuffs that didn't turn into "products" :-( Sep 17 17:00:24 now it's just qt hehe Sep 17 17:00:28 everywhere Sep 17 17:00:32 ick Sep 17 17:00:51 if I wanted something to run slow, I'd just stick with Winblows :) Sep 17 17:02:12 hehe Sep 17 17:03:32 devilhorns: and about efl, any news when we're getting the real "1.0" release? Sep 17 17:03:37 maybe then the final e17 Sep 17 17:03:38 hehe Sep 17 17:06:18 well work is happening now to make elm 1.0 alpha Sep 17 17:06:38 after that happens "officially", then it should be a matter of a week or two for "beta" Sep 17 17:07:21 cool Sep 17 17:07:37 and I'd imagine "release" will be shortly (couple weeks to a month) after that ... of course all this is assuming no major issues arise Sep 17 17:08:07 rsalveti, but that's not an "official" timeline or anything, so don't quote me on that ;) Sep 17 17:08:54 yay enlightenment Sep 17 17:09:05 when are we getting eubuntu-desktop? :) Sep 17 17:09:13 indede Sep 17 17:09:35 well, we had one a couple years back w/ gOS .. but bad management killed that Sep 17 17:10:01 that horrible emachines thing? Sep 17 17:10:23 well, we were not responsible for the hardware they chose :) Sep 17 17:10:59 indeed it never got much traction outside of $199 walmart specials.. I did try it and was impressed with how it looked and felt Sep 17 17:11:35 :) and that was with an early efl too Sep 17 17:11:36 is all that work sitting in some git repo or openembedded thing somewhere? Sep 17 17:11:55 or rolled into e17 libs now? :) Sep 17 17:12:29 Actually, I had some non-linux users test that here (before I found Ubuntu), and they had a hard time with some basic operations. That was on real hardware. Sep 17 17:12:30 Neko, no, gOS is as dead as can be ... I have some iso's that we made still lingering around here somewhere but that's about it Sep 17 17:12:51 Neko, some stuff got rolled back into efl, yes Sep 17 17:12:52 the one thing I always hated about enlightenment, despite it being totally awesome, it had a penchant on the default build and themes and things for everything being black on dark grey, with medium grey highlights, 7 point bitmap fonts, and everything being in a 900-item cascading context menu Sep 17 17:13:00 Neko, Making an enlightenment flavour is mostly a matter of coming up with all the right bits: feel free to do it if you're sufficiently interested :) Sep 17 17:13:16 I am sufficiently interested to do it on something other than ubuntu :) Sep 17 17:13:43 GrueMaster, yea, think that was the biggest complaint that we had ... but again, that was with an early efl ( & E) that really wasn't ready for a desktop yet Sep 17 17:15:21 (yes I was on the gOS team too) ... not proud of that, but hey I look @ it like "eh, it was experience in making distros" :) Sep 17 17:15:27 At the time, my main distro was Mandriva. Same group had no complaints with it. I use them for some UNE testing now and again, but haven't in 6 months. Sep 17 17:16:34 I usually do it when one of them hosts a game fest. But the last few fests I have been OOC. Sep 17 17:16:49 * rsalveti out for lunch Sep 17 17:16:50 ahh games ... I miss those Sep 17 17:16:55 haven't had time for em lately Sep 17 17:17:06 Same here. Sep 17 17:17:51 I hope to have a fest post-release. Really feel the need to shoot my fellow man - virtually of course. :P Sep 17 17:18:23 hahaha Sep 17 17:20:48 hum... games Sep 17 17:20:54 my ps3 is off for weeks Sep 17 17:21:15 mine too :( Sep 17 17:21:34 I'd like to play SC2, but gets very slow with wine on my current machine Sep 17 17:21:43 * rsalveti needs a faster machine Sep 17 17:24:03 hehe Sep 17 17:30:41 Great. After my power glitch (affected the neighborhood), my XM has 12 crash reports. Trying to run apport-bug on them...crashed. Sep 17 17:32:08 Extra points! Sep 17 17:33:18 lol Sep 17 17:33:24 you get to keep the extra parts ;) Sep 17 17:33:34 heh. Sep 17 17:34:14 And we have a bot again! Sep 17 17:34:28 ? Sep 17 17:34:53 Oh, I guess we've had it for a while. Sep 17 17:34:59 yea. Sep 17 17:35:04 on this channel. Sep 17 17:35:11 There were a few days we didn't have one, and I just saw the join. Clearly I have to read backscroll more carefully. Sep 17 17:35:26 (note the change from "4" to "2") Sep 17 17:36:14 oh. Sep 17 17:52:18 any beagleboard users able to help with a boot issue? Sep 17 17:54:05 What seems to be the problem? Sep 17 17:57:22 "it doesn't boot" :D Sep 17 17:59:23 Everything was happy until yesterday, where I now get: ** Unable to use mmc 1:1 for fatload ** at boot Sep 17 17:59:35 I'm guessing corrupt SD card.. Sep 17 18:00:16 Hrm. I've not seen that one. Sep 17 18:01:45 this is from a ubuntu-10.04.1-minimal-armel image if that gives any clues Sep 17 18:02:34 ogra, ogra_cmpc, I noticed you were somehow involved with the blueprint etc. on the "make maverick filesystems for boards with crappy storage that are just like the ones the installer makes"? rsalveti too? is there something I can look at that describes the stuff the installer does post package install (like that bit where it restores settings and makes the font terminus or that slightly more wiggly ubuntu one, or whatever..)? Sep 17 18:04:52 Neko, The jasper-initramfs package is what does that. I was to review it at one point, but have been advised that the code is painful to examine. Sep 17 18:05:22 <-- masochist for this stuff Sep 17 18:05:30 It then calls into the oem post-install configuration tool (part of ubiquity) for the final configuration and tweaking. Sep 17 18:05:53 I'm already looking through the oem post-install stuff but it's very light on changing things that aren't usernames, passwords and timezones Sep 17 18:06:10 it's just from the installer /etc/default/console-setup gets tweaked and I don't know where that is Sep 17 18:06:28 there has to be a few things it does besides, too Sep 17 18:07:00 Check the code, but jasper-initramfs does the make-the-image-seem-like-it's-kinda-installed bit, and then calls the oem interface. Sep 17 18:07:07 There shouldn't be any other moving parts. Sep 17 18:07:12 alles klar Sep 17 18:07:30 I'm also kind of looking for the reason oem-config-gtk doesn't run on my rootstocks Sep 17 18:07:47 it just sorta spools up gdm and xfce regardless of what I do.. even oem-config-prepare doesn't do anything Sep 17 18:08:06 Probably something in rootstock that tries to make the image think it's installed. Sep 17 18:08:12 I was thinking, does it require a ramfs? Sep 17 18:08:31 because we don't do that on the efika yet it needs some hardcore uboot changes and I'm scared to even change a variable Sep 17 18:08:33 No, but I think it's mostly untested without one. Sep 17 18:08:54 what I did was grabbed rcn-ee's beagle image and it ran the curses one just fine Sep 17 18:09:12 You're running Ubuntu on the efika? Sep 17 18:09:14 without initramfs or anything but it's ugly as sin Sep 17 18:09:20 since the dark ages yeah Sep 17 18:09:32 Jaunty, Karmic, Lucid, take your pick... Sep 17 18:09:53 I wish I'd known: I wouldn't have supported the removal of the imx kernels if someone who is reliably around was using them. Sep 17 18:10:06 oh believe me, it would have happened anyway Sep 17 18:10:10 Why? Sep 17 18:10:31 Only reason it was dropped was because nobody seemed to be maintaining it, and it wasn't known to work on any retail HW. Sep 17 18:10:33 because I don't think you guys want to maintain a userspace for 2.6.35 where everyone else is, when mx51 is stuck on last year's kernel Sep 17 18:10:55 the amount of work to push efikamx support into the ubuntu kernel would have meant pulling a new bsp patchset too.. you're 9 months behind Sep 17 18:11:18 Just needs someone to do it, really. the 9 months mostly reflected nobody working on it. Sep 17 18:11:22 we're working on 2.6.35.. we'll provide an ubuntuized kernel, then we can work out what we do for Narwhal Sep 17 18:11:33 Who is "we" in that context? Sep 17 18:11:44 <-- product manager for efika mx Sep 17 18:11:53 Oh! Just the right person then :) Sep 17 18:12:31 So, yeah, if you're interested in having Ubuntu work, and you have a kernel that works, it's relatively easy to get that kernel building and into the archive. Sep 17 18:12:45 you'd think. I might have appreciated an email about it to be honest, amit and lool knew about us :D Sep 17 18:13:03 But it does require someone to maintain it, and the processes for granting direct upload rights takes several months of continued work. Sep 17 18:13:32 that's fine, as long as it doesn't cost a ton of money for us unfortunatelt Sep 17 18:13:46 we don't really have the resources of ARM or Freescale to throw a bunch of money at a distro that we won't get for 6 months. Sep 17 18:13:58 There's no monetary cost for bits in the archive: the cost is entirely in engineering. Sep 17 18:14:10 If you can maintain it, it can be there. Sep 17 18:14:16 it basically comes down to how much we want ubuntu and how many people we can throw at it Sep 17 18:14:20 Right. Sep 17 18:14:34 but we have a bunch of guys (well, one main one) who are doing some debian work who are willing to maintain ubuntu stuff too Sep 17 18:14:42 did you see the armhf port? :) Sep 17 18:14:54 I have, and am very excited by it. Sep 17 18:15:11 any chance it'll get into Obtuse Orangutan or whatever it's going to be called? :D Sep 17 18:15:20 No idea. Sep 17 18:16:01 New ports require infrastructure resources, which are mostly limited to Canonical right now, so it's a matter of what ends up being supporting in the Canonical DC. Sep 17 18:16:18 we're pretty stoked that stuff is getting done. we bought the debian-ports guys some big disks for their servers so they had space for armhf Sep 17 18:16:41 Saw that, and lots of talk about hardware being handed around. Sep 17 18:17:23 http://buildd.debian-ports.org/status/architecture.php?a=armhf&suite=unstable lots of packages done :D Sep 17 18:17:41 if we throw another 5 boxes at it we'll rival the ubuntu armel build farm :3 Sep 17 18:17:44 Yep. Prospects look really good for Wheezy Sep 17 18:18:43 Neko: bah, you can't compare debian with ubuntu..ubuntu is the best Sep 17 18:18:49 :) Sep 17 18:19:24 Err, Debian and Ubuntu have such wildly different foci that they can't usefully be compared Sep 17 18:19:34 he was being facetious Sep 17 18:19:41 :D Sep 17 18:20:13 we just figured the ubuntu guys would not bite considering the nightmare of lpia in the past, until it was at least in debian-testing Sep 17 18:20:40 It's not about that: it's all about what ends up in the infrastructure. Sep 17 18:21:22 :) Sep 17 18:21:29 Back when we were working towards feisty, a bunch of us tried to get an "arm" port (this is pre-armel), but ended up having issues because of hardware availability and supportability for delivery to the DC. Sep 17 18:22:01 back then there were no bloody boards :] Sep 17 18:22:33 Things have changed since then, but you'd really have to talk to #canonical-sysadmin about if/whether DC stuff could be supported, and you'd first want to have a significant number of Ubuntu Developers ready to support the port. Sep 17 18:22:39 still waiting for beagle and buying $6000 fsl dev kits was the task du jour.. dark, dark times.. Sep 17 18:22:52 There was plenty of ARM5 hardware back then :) Sep 17 18:23:10 but nothing as cool as a sheevaplug Sep 17 18:23:24 But I'll be the first to admit I'm much happier with the current port than what we would have produced then, as specifications are now much more acceptable. Sep 17 18:23:48 where is jasper-initramfs built from Sep 17 18:24:05 jasper-initramfs source :) Sep 17 18:24:11 no such package Sep 17 18:24:27 also "jasper" is definitely not the same thing, seems some jpeg viewer or so Sep 17 18:24:30 On an Ubuntu maverick system, `apt-get source jasper-initramfs` doesn't work? Sep 17 18:24:58 oh shit okay. you know. I realized I am still running Lucid Sep 17 18:25:00 new thing? Sep 17 18:25:32 Yeah, new for maverick. Lucid beagle images installed to NAND+USB (like the world should be). Sep 17 18:26:10 At maverick UDS, folk were talking about boards shipping with no NAND and no ability to boot from USB, meaning that one had to have install media be the *same* as runtime media. Sep 17 18:26:48 jasper-initramfs was created to handle that. I think it was switched for beagle mostly because nobody had the hardware described when it started getting written. Sep 17 18:28:06 ahaaaaa.. I notice it uses a uboot boot.script Sep 17 18:28:11 man this helps a lot trying to wrangle this work Sep 17 18:28:35 mkdir -p /root/var/lib/oem-config Sep 17 18:28:35 touch /root/var/lib/oem-config/run Sep 17 18:28:39 is that all oem-config really needs? Sep 17 18:29:19 * persia doesn't remember, having not looked at that since ~jaunty, but wouldn't be surprised Sep 17 18:31:50 I don't like the idea of flash-kernel. we want our images on pata, in an ext2 or vfat partition :) Sep 17 18:32:00 Then don't use flash-kernel. Sep 17 18:32:08 it seems to be the default Sep 17 18:32:10 It's a workaround for devices that can't boot off a mountable /boot Sep 17 18:32:19 That's because most devices suck :) Sep 17 18:32:44 do you want to make a list of ubuntu people who would like a current Efika MX with working NEON? :) Sep 17 18:33:20 I have a shortlist of folk I'd like to have one, which I'd be happy to send you :) Sep 17 18:33:36 Neko: You can disable flash-kernel in /etc/kernel-img.conf Sep 17 18:33:46 I know :) Sep 17 18:33:58 But in fairness, I think it would probably be better to have some sort of public offer, and have folk send you a note suggesting what they might do with it, etc. Sep 17 18:34:02 lag: you get your board modified? Sep 17 18:34:14 You can just not install flash-kernel in the first place. Sep 17 18:34:30 (or purge it) Sep 17 18:34:33 persia, done Sep 17 18:34:33 http://projects.powerdeveloper.org/program/imx515/proposed Sep 17 18:35:01 http://projects.powerdeveloper.org/program/imx515/accepted actually that is a better page to look at :D Sep 17 18:35:01 Neko, So I should tell the folks on my list that if they want one, they ought register there? Sep 17 18:35:40 yes at least create an account, give me user id numbers, propose a project.. if it's just "we're ubuntu guys and we want to play around" then get a lead to make a project and I can add developers into the same project group Sep 17 18:36:20 I have to insist on at least a paragraph like a university thesis introduction on the whole thing, one liners make me sad :D Sep 17 18:37:22 Would you prefer several submissions e.g. "working on Ubuntu Haskell porting for armel", "working on improving Ubuntu armel Qt support", etc. or one big one? Sep 17 18:38:20 if it is seperate subsystems or teams I'd prefer seperate Sep 17 18:38:41 for instance the debian guys we special cased, but there is debian-edu, debian-x11 etc. and they should have all made projects if we weren't in a rush to get them boards for debconf Sep 17 18:39:00 I'll say that folks were very happy to all have boards at debconf :) Sep 17 18:39:21 So, yeah, I'll pass the word around to folks who aren't always in this channel to submit stuff then. Sep 17 18:39:37 And anyone in this channel who has a good idea should go get a board if they want one :) Sep 17 18:40:29 we might cut you off at like 10 boards or so :) but we can also drop some older boards with busted NEON for compile farm stuff. Sep 17 18:43:08 All I can do is make sure your program gets noticed by folk I know what HW. You'll have to decide the value of each proposal :) Sep 17 18:46:20 s/what/want/ Sep 17 19:09:25 for anyone that wants to start doing notes/wiki posts on the panda - http://www.elinux.org/PandaBoard Sep 17 19:12:41 can someone add a "buy here" link? :) Sep 17 19:13:29 Hi, Good afternoon. Sep 17 19:13:30 BeagleBoard plugin on a pc (ubuntu) but nothing appears in dmesg Sep 17 19:13:56 USB connection? Which dmesg has no output? Sep 17 19:17:31 yes, usb Sep 17 19:18:25 It has no output Sep 17 19:18:52 On the beagle? On the host? Sep 17 19:19:17 You're running Ubuntu on the host: what are you running on the Beagle? Sep 17 19:21:47 In my notebook using ubuntu. Sep 17 19:21:48 I called the BeagleBoard with USB cable, use dmesg to see what was connected and nothing appears. Sep 17 19:26:27 I called the BeagleBoard with USB cable on notebook (ubuntu), use dmesg to see what was connected and nothing appears. Sep 17 19:26:36 sry 2 msg Sep 17 19:27:20 No, that7s fine. I'd actually like to understand. Sep 17 19:27:29 So, what are you running on the Beagle? Sep 17 19:28:28 Beagle --> USB --> Notebook Sep 17 19:29:00 But there is no recognition Sep 17 19:31:16 I use the command dmesg new hardware plugged Sep 17 19:31:43 but there is no line of response Sep 17 19:32:10 sry , my english =/ Sep 17 19:34:50 What OS do you run on the Beagle? Sep 17 19:35:51 I want to install ubuntu-arm Sep 17 19:36:13 in my / dev shows nothing Sep 17 19:38:08 fredim: well, plugging your beagleboard into your notebook could mean that you want to power it or use it as a gadget Sep 17 19:39:24 and if it is to be used as a gadget, the Beagle needs to have gadget-supporting software running at the time it's attached. Sep 17 19:39:34 If it's just power, no response is expected. Sep 17 19:39:46 yep, do you have any software already running on your beagle? Sep 17 19:39:50 If the goal is to *install* Ubuntu, put the available image on an SD card, and boot the Beagle from SD. Sep 17 19:41:23 fredim: Download http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/beta/ubuntu-netbook-10.10-beta-preinstalled-netbook-armel+omap.img.gz and follow the instructions at https://wiki.ubuntu.com/ARM/OMAP Sep 17 19:41:39 We can help from there, but that should get you started. Sep 17 19:47:44 I have to install some software to appear /dev/mmcblk0 Sep 17 19:48:12 fredim: On your laptop? Sep 17 19:48:19 yes Sep 17 19:49:08 /dev/mmcblk0 isn't the beagle: it's the SD card (if you have one built-in). Sep 17 19:49:13 Look at dmesg. if your SD reader is usb based, it will show up as /dev/sd? (i.e. /dev/sdc). Sep 17 19:49:22 If you are using a USB card reader, you'll end up with /dev/sd${something} Sep 17 19:52:33 My SD card is plugged in beagle ... Still have to appear in / dev / sdx? Sep 17 19:52:45 Why is there no line in dmesg Sep 17 19:57:05 fredim, You need to attach the SD card to your PC *without* the beagle (unless you have special software to make the Beagle an SD reader, but I don't know of any such software) Sep 17 20:00:04 How do I connect SDcard directly on the notebook/computer, what means? Sep 17 20:00:27 my laptop has the card reader but the computer does Sep 17 20:02:07 fredim, Most folk use either a built-in card reader or a USB card reader Sep 17 20:05:10 The beagle won't work until an image is on the SD card. Sep 17 20:06:24 Now I understand, thank you. Sep 17 20:36:03 GrueMaster: ping Sep 17 20:36:12 pong Sep 17 20:36:42 GrueMaster: is the kernel defconfig that is being used for the ubuntu panda builds posted somewhere? Sep 17 20:37:00 I have no idea. lag would know. Sep 17 20:37:02 It's in both source and binary packages. Sep 17 20:37:36 I'd recommend checking the source Sep 17 20:37:48 persia: where can i get the source package? Sep 17 20:37:55 * prpplague is not an ubuntu person Sep 17 20:38:06 apt-get source Sep 17 20:38:41 ok, whats the name of the package? Sep 17 20:38:53 there isn't a git repo somewhere? Sep 17 20:39:05 kernel.ubuntu.com has heaps of git repos Sep 17 20:39:16 * prpplague looks Sep 17 20:39:32 persia: thanks Sep 17 20:39:43 http://archive.ubuntu.com/ubuntu/pool/main/l/linux-ti-omap4/ has the tarball of the source Sep 17 20:40:00 ahh ok thanks Sep 17 20:40:04 The source package is linux-ti-omap4-source-2.6.35. Sep 17 20:40:44 That will pull the tarball, any external patches, and the dsc file to build the .deb with. Sep 17 20:41:41 Also, apt-cache show will show you the name for the source package. Sep 17 20:41:57 (really, it looks stuff up in a database, and downloads all the files from the URL I just posted, but it always has the right place, instead of needing to hunt about) Sep 17 20:42:14 GrueMaster, apt-cache show won't work for packages not available on that architecture Sep 17 20:43:58 prpplague: tarball is easier than just looking for the config file at the git tree Sep 17 20:44:15 as it's broken into common and specific configs depending on the machin Sep 17 20:44:17 *machine Sep 17 20:44:34 but for omap 4 I believe it should be fine, as it's just one machine... Sep 17 20:44:44 1 sec Sep 17 20:44:57 rsalveti: yea grabing it now Sep 17 20:45:23 is there already some package or script to setup ARM cross compiling toolchain easily on ubuntu? Sep 17 20:45:30 that would be awesome :D Sep 17 20:45:33 prpplague: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=blob;f=debian.ti-omap4/config/config.common.ubuntu;h=8d46b556e09697138c3a2b8b975c9f519b901856;hb=refs/heads/ti-omap4 Sep 17 20:45:37 for omap4 Sep 17 20:45:55 as this kernel supports just one machine, there's only the config.common Sep 17 20:46:20 rlameiro: I believe there are packages already for maverick Sep 17 20:46:26 hrw|gone: did the work Sep 17 20:46:31 rsalveti: really:D Sep 17 20:46:35 FTW Sep 17 20:47:02 was in a custom repository, but he was going to push it before the freeze Sep 17 20:47:11 just don't know if it's compiled already Sep 17 20:48:01 probably it is, as we have more builders for i386 Sep 17 20:48:07 se we can compile the kernel like make CROSS-COMPILE ARM=arm bla bla bla Sep 17 20:48:10 ??? Sep 17 20:48:28 *ARCH=arm Sep 17 20:48:40 rsalveti: thanks Sep 17 20:49:17 rlameiro: https://edge.launchpad.net/ubuntu/maverick/+source/gcc-4.5-armel-cross Sep 17 20:49:54 there's also the 4.4 packages Sep 17 20:50:06 rlameiro: yep Sep 17 20:50:26 WOW that is awesome Sep 17 20:50:50 the IGEPv2 comes with a VM for vmware with things almost ready Sep 17 20:51:06 it will be awesome to have it already on the repos Sep 17 20:51:40 they use ubuntu as their SDK Sep 17 20:51:46 8.04.... Sep 17 20:52:19 hehe Sep 17 21:12:47 what is the best filesystem for the rootfs sd card? Sep 17 21:12:55 Depends on the card. Sep 17 21:13:14 well, kingstom sdhc Sep 17 21:13:17 But typically any journalling filesystem will be fine. Sep 17 21:13:18 4gb Sep 17 21:13:35 ext3 or 4? Sep 17 21:13:42 I have no idea which FTL is in that (and I doubt it's enough info, FTLs tend to be swapped out within a single product without model changes) Sep 17 21:14:23 There should be no benefit to ext3 or ext4 specifically for it being on an SD card, unless you have a very fancy FTL that happens to have ext4 support. Sep 17 21:14:40 how can i know that? Sep 17 21:16:52 I don't know of any reliable way :) Sep 17 21:17:04 :) Sep 17 21:18:06 rlameiro, By the way, we tend not to cross-compile stuff: everything in the archive is native-compiled. If you're working on something, it's worth doing a native-compile test every once in a while, even if you usually cross-compile: we've seen a number of issues where something will cross-compile but not native-compile or vice-versa. Sep 17 21:18:31 persia: ok, but that was for the kernel Sep 17 21:18:44 compiling kernel natively takes a long time :D Sep 17 21:18:56 and it's always handy to have a cross compiler :-) Sep 17 21:19:15 rsalveti, Only if your hardware isn't very good :p Sep 17 21:19:16 no need to get them from CS is even better Sep 17 21:19:57 persia: true, the good thing is that this is going to change probably next year :-) Sep 17 21:20:04 panda itself is going to rock already Sep 17 21:20:55 People keep saying "This will change". I purchased a perfectly acceptable-performance laptop running armel about a year ago, but people still talk about the future. Sep 17 21:21:09 The key to having it happen is for someone to just do it. Sep 17 21:21:43 One of my ARM boxes is 1.2Ghz, which is plenty fast. Doesn't happen to support Ubuntu, so it runs Debian. Sep 17 22:14:41 davidm: ping Sep 17 22:27:03 hi jayabharath Sep 17 22:27:10 davidm: greetings Sep 17 22:27:18 Board away, thanks Sep 17 22:27:19 davidm: get your boards picked up? Sep 17 22:27:47 prpplague, yep, Home - TI - FedEx - Home :-) Sep 17 22:28:37 davidm: did you get washed away? Sep 17 22:29:22 *cough* sent one to RoHS, and got nada Sep 17 22:30:00 Nope made it back home before this storm blew in Sep 17 22:30:32 davidm : good to know you got the board. Sep 17 22:31:16 davidm: we were trying to test the audio.. to ensure its working... but could NOT as you came in just on time :( Sep 17 22:31:45 its a new board and tested during production so mpoirier_should have no issue Sep 17 22:33:34 jayabharath, thanks for the heads up, I'll let mpoirier know to check it Sep 17 23:15:49 anyone here has problms with ureadhead? Sep 17 23:16:43 I think rsalveti was talking about some issues with it and OOM in the past Sep 17 23:18:24 well, it still have that problem Sep 17 23:18:57 Im not sure if I used the last rootstock Sep 17 23:19:12 I remeber i needed to change some file somewhere, i will dig it again Sep 17 23:21:34 I think rsalveti was having the issue with jasper images: maybe rootstock needs the same hack Sep 17 23:24:10 rlameiro: Bug 600359 Sep 17 23:24:12 Launchpad bug 600359 in ureadahead (Ubuntu Maverick) (and 1 other project) "ureadahead generating oom messages during boot. (affects: 10) (dups: 1) (heat: 70)" [High,In progress] https://launchpad.net/bugs/600359 Sep 17 23:24:25 Is that what you are referring to? Sep 17 23:24:42 yes Sep 17 23:24:46 it hangs at boot Sep 17 23:25:04 Hangs? Which system? Sep 17 23:26:22 IGEPv2 Sep 17 23:26:34 beagle clone with more ram and wireless Sep 17 23:27:38 rlameiro, anything special with your setup? (mine with IGEPv2's factory u-boot and mainline 2.6.35 is solid..) Sep 17 23:28:42 i build the kernel 2.6.33.7 and custum rootfs made with rootstock Sep 17 23:28:49 this happended to me before Sep 17 23:29:13 it hangs at the boot and it doesnt prompt for login Sep 17 23:29:31 ahh, and the led is blinking green Sep 17 23:29:52 rcn-ee: opps didnt ping you Sep 17 23:29:59 humm weird.. (mine's in the basement so i never see it..) anything special in that 2.6.33.7? Sep 17 23:31:05 well, it had some stuff added for the omap stuff Sep 17 23:31:23 its the igepv2's website Sep 17 23:31:24 reference if nothing too special (i have most of the omap stuff in it.): i'm running: http://rcn-ee.net/deb/maverick/v2.6.35.4-l4/linux-image-2.6.35.4-l4_1.0maverick_armel.deb Sep 17 23:31:51 does audio works aut of the box? Sep 17 23:32:24 rcn-ee: and his it built for the igepv2 board? Sep 17 23:32:27 audio i don't know... video works, one of the igepv2 guys sent me patches when it wasn't working.. otherwise sgx/dspbridge stuff enabled.. Sep 17 23:32:41 it's built for all. ;) (beagle, overo, touchbook, igepv2, panda) Sep 17 23:33:16 why is it a deb? Sep 17 23:33:30 isnt it an uImage file? Sep 17 23:33:33 for rootstock.. ;) Sep 17 23:33:53 ohhhh, i need to build a new rootfs Sep 17 23:33:55 :D Sep 17 23:33:58 otherwise, dpkg extract, mkimage on vmlinuz.. ;) Sep 17 23:34:03 well, the problem will be the same i guess Sep 17 23:34:04 .debs are better anyway: they can have pre- and post- install logic, and be removed cleanly :) Sep 17 23:34:34 rcn-ee: but will rootstock build and install the modules? Sep 17 23:35:56 lol Sep 17 23:35:58 http://webcache.googleusercontent.com/search?q=cache:ZavfyW4gZlwJ:thetechshop.org/showthread.php%3F193-Ubuntu-Howto-Fix-ureadahead-problem-on-10-04+how+to+fix+ureadahead&cd=8&hl=pt-PT&ct=clnk Sep 17 23:36:01 yeap, it'll output a vmlinuz-.... and a uIntrd-.... for first boot and the modules wil be installed. (i'm the one who sent ogra the patches for --kernel-image addon to rootstock) Sep 17 23:36:24 that's easy enough, do that first... ;) Sep 17 23:36:45 so, i copy the vmlinuz and uIntrd to the boot partition Sep 17 23:37:34 yeap, you'll have to run mkimage... the script plays dumb, since u-boot/omap address isn't 100% of the arm boards... Sep 17 23:39:45 hey persia, are you guys tracking/noticing a lot of oem-config/ubiquity bugs in your armel images? Sep 17 23:40:30 I'm not, but I've no idea about others. I've seen ogra trying to work around some things with jasper and ubiquity and debian-installer changes. Sep 17 23:41:18 well, i think i will need to do it your way rcn-ee :D Sep 17 23:41:27 do you have some place explaining the steps? Sep 17 23:41:39 I'm seeing some issues getting the XM to boot into oem-config on the 20100909 image. Sep 17 23:41:52 rcn-ee: ^^^ Sep 17 23:42:01 GrueMaster, it just dies with oem user not found, black screen/etc. ;) Sep 17 23:42:04 Been fighting it all day now. Sep 17 23:42:26 Interesting. My problem is that it just fails to run, or part of it fails. Sep 17 23:42:29 https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bugs?field.tag=armel is the URL for the current set of reported armel/ubiquity bugs. Sep 17 23:42:31 sure rlameiro, where you at with it.. Sep 17 23:42:39 Worked fine when I first tested the 20100909 image. Sep 17 23:43:18 rcn-ee: well, bette, do you have it in some webpage explained ? :D Sep 17 23:43:36 rlameiro, yeap.. elinux.org/BeagleBoardUbuntu ;) Sep 17 23:43:50 rcn-ee: oki Sep 17 23:43:58 those demo images will work on the igepv_2.. ;) just use the latest setup_sd.sh script.. Sep 17 23:44:24 * rcn-ee don't you hate when you cant' find that bug/report log browser tab... Sep 17 23:45:06 rcn-ee: one problem, I dont have HDMI here for the first boot. Sep 17 23:45:16 i run it headless via serial/ssh... Sep 17 23:45:46 ahh crap yeah... i'll have that fixed for the next image.. (defaults ubuntu/temppwd)... Sep 17 23:46:36 GrueMaster, i think that happend between 2.3.18/2.3.19 http://bazaar.launchpad.net/~ubuntu-installer/ubiquity/trunk/changes Sep 17 23:46:45 but if i build it with rootstock will i have the same problem? Sep 17 23:46:52 maybe that is my problem now Sep 17 23:47:05 rlameiro, as long as you put a login and password it'll be fine... Sep 17 23:47:07 it boots but doesnt give me acces to a console Sep 17 23:47:28 actually it might be, if oem-config was installed.. Sep 17 23:47:32 maybe it boots but graphical only, i put a password in it Sep 17 23:47:38 rcn-ee: There is a problem with it not running on newer images. Not sure what it was, but I was told they were working on it. Sep 17 23:48:02 brb Sep 17 23:48:45 i'm just doing nightly builds/tests and it's getting worse.. my workaround with 2.3.3 doesn't work anymore either. ;) Sep 17 23:49:23 The last daily image that built was 20100916, but oem-config fails to run. Prior to that, 20100909 was the last working preinstalled image. Sep 17 23:52:14 I thought people said good things about 20100914 at one point Sep 17 23:52:58 It was good for omap4. Sep 17 23:53:03 Ah. Sep 17 23:53:05 No omap build. Sep 17 23:54:22 as a plus the next release those should be combined.. (well 2.6.37 based ubuntu) Sep 17 23:56:21 The next release is probably .37 or .38, so that's likely. Sep 17 23:56:22 rcn-ee: well i went to test it on the Tv and it passed after the "hang" place Sep 17 23:56:31 rcn-ee: Not sure if we will have 1 kernel to rule them all by then. I know there is active work going on in devicetree, but not sure on status. Sep 17 23:56:41 cool rlameiro that was quick.. ;) Sep 17 23:56:49 GrueMaster, one kernel to rule all omap != one kernel to rule all armel Sep 17 23:56:57 but then the screen changed to black with a coursor blinking at the top right corner Sep 17 23:56:57 Oh. Sep 17 23:56:58 probally not, but one omap kernel atleast. ;) Sep 17 23:57:58 * GrueMaster gives up for the day on beagleXM. 5 reinstalls on three different SD cards, 5 different failures, none reproducible. Sep 17 23:58:02 Now, for linux 3.0, one of the goals should be a single build that works on every architecture. Sep 17 23:58:16 that would be...interesting. Sep 17 23:58:33 Maybe if they shift the code to java? Sep 17 23:58:34 Yep. Think about the board initialisation detection phase :) Sep 17 23:58:36 GrueMaster, random... a1 by chance? Sep 17 23:59:01 This is a premee. P8 board. Sep 17 23:59:12 what 'mpurate=xx' are you using? Sep 17 23:59:19 did you set mem=256 on boot? Sep 17 23:59:36 It worked fine before. With this image. Sep 18 00:00:30 on my pre production boards, i had to limit them to 800mzh with "mpurate=800" and the first 256 of the 512 of memory... Sep 18 00:00:34 I usually don't muck with the bootargs, except to add serial console. Sep 18 00:02:31 i muck with them all the time.. ubuntu could get a 'unfair' speed bump from lucid to maverick by bumping from the default 500mhz to 720 on the c4's like the part is designed ;) Sep 18 00:02:34 What I am seeing is that sometimes it will boot to gdm before oem-config runs, sometimes parts of oem-config will crash. Nothing consistant, which suggests memory issues. Nothing on the serial console. Sep 18 00:02:50 rcn-ee: you make boards? Sep 18 00:03:27 sometimes.. (i didn't make the beagle thou)... i do work at a place that sells them.. ;) Sep 18 00:03:31 My job is to test the images as they are. I usually don't have time to muck with out-of-image settings. Sep 18 00:03:53 which is perfectly allright and better for the end customer Sep 18 00:04:35 but once the voltage scalling patches hit mainline, we need to bump the xm from 500mhz to 1Ghz.. ;) Sep 18 00:05:20 Speed is good. Sep 18 00:05:39 I'd like to see them run at a more usable pace. Sep 18 00:09:25 (at least it would fail faster). :P Sep 18 00:13:59 * GrueMaster thinks 10 hours is enough work for one Friday and wanders off to find a beer. **** ENDING LOGGING AT Sat Sep 18 02:59:57 2010