**** BEGIN LOGGING AT Thu Sep 16 02:59:57 2010 Sep 16 07:24:42 I was told some time ago that the latest CodeSourcery compiler is not the most efficient, but rather older releases work better. Does anyone know anything about this statement? (I do know Ubuntu uses its own native compiler.) Sep 16 07:27:30 morning Sep 16 07:35:34 sveinse, I don't think that is accurate Sep 16 07:36:48 ogra: hi Sep 16 07:37:10 ogra: i would like to be able to easily add console output and console login on a ubuntu image. Sep 16 07:37:38 ogra: today it requires to regenerate the boot.scr and add the proper file in /etc/init/ Sep 16 07:38:35 ogra: can't we find some uboot magic env variable, which is not set by default for the console so that I can easily break and do setenv console ttyS0? Sep 16 07:44:45 rsavoye: That's exactly my question. I just heard about someone doing a performance comparison on ARM between the various CS releases Sep 16 07:45:08 Perhaps someone in this channels did know about it... Sep 16 07:55:18 ndec: Are you still interested in David's audio presentation? Sep 16 07:57:35 ndec, i'll work on that today after i made sure the apturl stuff didnt add any regression (i'd also love to test that one but still havent heard back about the PPA) Sep 16 08:00:31 ndec: I heard rumours that the panda might have facilities to measure current consumption of various subsystems. Do you know anything about this? Sep 16 08:04:40 lag: more than ever! Sep 16 08:05:05 I'll get them for you Sep 16 08:05:07 amitk: i would continue to call this, rumours... Sep 16 08:05:33 amitk: rumours, that I didn't heard... let me try to check with some folks. Sep 16 08:06:57 persia: hi! Sep 16 08:07:12 ogra: thx. for the PPA you should be able to use any other public PPA in the mean time. Sep 16 08:07:54 ndec, since the setup happens on first boot i need to roll an image for each change i want to test so i would prefer to go with the real thing Sep 16 08:07:58 persia: I wrote a /usr/share/alsa/SDP4430.conf file but it doesn't seem to be taken into account. Is there a way to debug this? I don't find useful doc in google or alsa-project.org for this particular item Sep 16 08:08:15 (beyond that we dont have many public armel PPAs :) ) Sep 16 08:08:17 ogra: argh... Sep 16 08:09:41 berco, Did you add a conditional inclusion entry into 00main? Sep 16 08:09:57 persia: no Sep 16 08:10:02 * ogra sighs about 180 new bugs in his inbox overnight Sep 16 08:10:38 berco, If you put it in init/ you need to fiddle with init/00main, if you put it in cards/ you need to fiddle with aliases.conf Sep 16 08:10:46 ogra: you should be happy! more bugs means that more people use your stuff! Sep 16 08:11:25 persia: I put my file in cards/ Sep 16 08:11:53 well, many of them are old cruft i dont really care about anymore due to changing teams etc Sep 16 08:12:05 i just get the CC's Sep 16 08:12:18 berco, And the driver reports "SDP4430" precisely? Sep 16 08:12:22 but i have to go through them nontheless Sep 16 08:13:17 berco, The logic that includes it is in /usr/share/alsa/alsa.conf : cards.@hooks Sep 16 08:13:27 persia: I think so. What command do you recommend me to use to be sure? Sep 16 08:13:41 persia: aplay -l reports "card 0: SDP4430 [SDP4430], device 0: Multimedia null-codec-dai-0 []" Sep 16 08:14:31 I think that ought get included. Sep 16 08:14:54 The amixer output doesn't show your config changes? Sep 16 08:15:27 persia: that's right. My changes are not reflected by amixer Sep 16 08:15:53 I added an entry in aliases.conf which gets included from alsa.conf Sep 16 08:15:59 Is that even needed? Sep 16 08:16:31 I generally recommend an entry in aliases.conf to sort things, but in this case it may not be needed because there is an exact string match Sep 16 08:17:58 persia: is there a way to debug the file is read properly? Like if I have some syntax problems... as this is my first time with this .conf file Sep 16 08:18:09 persia: I pastebin here my file http://pastebin.ubuntu.com/494629/ Sep 16 08:20:50 I don't know of one: I'll look a bit more. There's documentation on the format in /usr/share/doc/libasound2/examples/asoundrc.txt.gz Sep 16 08:22:40 ndec: Am I correct in thinking that the 4430sdp is the dev board and the sdp4430 is the sound chip? Sep 16 08:23:45 lag, I just missed you ~11 hours back: the key is that there is a very high chance that wiremap for the working board and the wiremap for the not working board differ slightly, which is the bit that needs changing. Sep 16 08:24:06 persia: Yes, I received that Sep 16 08:24:18 Ah, OK. Thought maybe you missed it, given the hour. Sep 16 08:24:28 Hence why I'm trying to track down the schematic for the 4430sdp Sep 16 08:24:30 lag: these are just names... which are probably bad in fact... but yes, in the kernel the blaze is called '4430SDP' and the audio chip is called 'SDP4430'... SDP means software development platform Sep 16 08:24:38 berco, Looks like there's some (limited) syslog output for errors. Looking more... Sep 16 08:24:49 persia: thnaks Sep 16 08:25:11 checking do too on my side. but looks similar to what I found on the alsa wiki Sep 16 08:25:17 ndec: I would like the schematic for whatever the 4430SDP was - I need to compare it to the Panda Sep 16 08:26:06 lag: isn't that in the public TRM? Sep 16 08:26:22 lag: oops. sorry boards schematics is not in TRM Sep 16 08:26:36 :) Sep 16 08:27:20 there is a public omap4 trm? Sep 16 08:28:40 suihkulokki: yes: http://focus.ti.com/pdfs/wtbu/OMAP4430_ES2.0_Public_TRM_vJ.pdf Sep 16 08:28:49 * persia suspects that while schematics might be protected IP, the basic wiremap for the driver probably isn't (as we already have the one for 4430SDP in the kernel driver), so someone with internal access can likely compare the wiremaps, and explain the diff for the more public boards. Sep 16 08:30:55 I have already seen that the Panda uses different outputs to it's predecessor Sep 16 08:31:13 snd_soc_dapm_enable_pin(codec->dapm, "Headset Mic"); Sep 16 08:31:13 snd_soc_dapm_enable_pin(codec->dapm, "Headset Stereophone"); Sep 16 08:31:15 lag, If you haven't already, you might want to ask the folks on #alsa-soc (freenode) for more guidance on specifically what you need. Sep 16 08:31:26 These were 'disable' before Sep 16 08:31:37 I made them enable, but it didn't make a difference Sep 16 08:32:10 You might have to raise the volume on the amplifiers, as well as just turning them on. Sep 16 08:35:33 berco, From what I can tell, running amixer is the best way to find out whether the configuration was parsed correctly. Sep 16 08:40:09 berco, If you haven't found it yet, http://www.alsa-project.org/alsa-doc/alsa-lib/conffunc.html (and related) seem to be good for parsing the configuration files. Sep 16 08:45:55 persia: thanks. amixer tells me it wasn't parsed or my conffile is malformed :) Sep 16 08:46:51 Excellent! That means that 1) your file is being included, and 2) it has a bug Sep 16 08:47:00 That's a wonderful place to start fixing it. Sep 16 08:49:33 persia: what I meant is that I don't see any change. So I'm not even sure it is actually parsed Sep 16 08:49:51 hi! Everytime I install the UNE images, I have to create a /etc/init/ttyO2 file to get a login prompt on the serial line. I need this because there is no openssh-server package on UNE. Sep 16 08:50:10 Would it make sense to add such a file to UNE images for OMAP? Sep 16 08:50:27 At least for OMAP4? Sep 16 08:50:43 mopdenacker, It's intentional that it isn't there. You ought be able to apt-get install openssh-server right after install if you want it. Sep 16 08:51:04 Or to install openssh-server... (Not sure whether we can do this, this would impact the desktop). Sep 16 08:51:10 mopdenacker: we discussed this earlier with ogra ;-) Sep 16 08:51:15 persia: right. OK. Sep 16 08:51:36 mopdenacker: ogra will try to make it easy to get a console login, but it won't be in the default image Sep 16 08:52:18 persia: the real trouble is that the Blaze board has a very special keyboard, which at the moment doesn't allow you to type "apt-get". Sep 16 08:52:20 Why do y'all want it? Doesn't the default console (HDMI, HID) work? Sep 16 08:52:35 That's an input problem... Sep 16 08:52:41 mopdenacker, Can you use the mouse? openssh-server is in the software-centre Sep 16 08:52:53 Err, "pointer" :) Sep 16 08:52:59 persia: that's a good idea... Sep 16 08:53:32 persia: oops, sorry for forgetting about the UI way of installing packages :-} Sep 16 08:54:18 Nothing to be sorry about, really, but remember that the netbook images are intentionally focused on UI solutions. Sep 16 08:54:46 Right. This applies to the desktop images as well. Sep 16 08:54:46 Adding a server image would be nicer for those of us who prefer CLI, but someone would have to care for it, and there's never been enough volunteers. Sep 16 08:55:12 Are there desktop images? I thought there were only images for Ubuntu Netbook, Kubuntu Netbook, Kubuntu Mobile. Sep 16 08:55:46 I meant desktop installs. Sep 16 08:56:03 They don't have openssh-server by default, which definitely makes sense. Sep 16 08:56:35 Oh, right. I meant that I didn't think anyone was making Desktop images for OMAP :) Sep 16 08:58:52 This could happen when the Panda is out. It now has enough RAM (and cpu power) to be a real desktop replacement. Sep 16 09:02:15 Mostly needs someone to care for it. Sep 16 09:02:46 Given a large enough volume of interested testers and developers, it's not technically that difficult to make. Sep 16 09:03:31 That said, my expectation is that a fair number of folk who install one of the available images will end up removing the metapackage, and customising it in all sorts of ways. Sep 16 09:05:11 Yes, probably Sep 16 09:09:09 persia: do u know if alsa-utils restart is enough to re-read the config or shall I reboot each time to test? Sep 16 09:12:33 alsa-utils restart ought be enough. Sep 16 09:13:04 I'd probably do "stop" and then "start", but that's be being careful Sep 16 09:13:15 ok Sep 16 09:13:32 Mind you, odd stuff happens on boot, related to HW detection, etc. but in terms of amixer telling you about parse errors, you should be fine. Sep 16 09:17:28 Hey, I have a new error installing the latest pre-installed image for OMAP4. Ubiquity fails to start: Sep 16 09:17:40 http://paste.ubuntu.com/494643/ Sep 16 09:18:03 I didn't see this on Tuesday. May I file a bug? Sep 16 09:18:15 Please do. Sep 16 09:18:34 f your system has internet, please file it *from* the affected system, with `ubuntu-bug ubiquity` Sep 16 09:18:43 This will include lots of useful data about the issue. Sep 16 09:19:08 Even better, if your system is reporting the crash with apport, use apport to file the bug directly. Sep 16 09:20:10 * ogra wonders if that has to do with the broken swap file Sep 16 09:20:17 is that todays or yesterdays image ? Sep 16 09:20:40 Yes, it does have Internet access, but it will be difficult to configure the proxy with the Blaze limited keyboard. Sep 16 09:20:47 Todays Sep 16 09:21:34 I don't have a Panda to reproduce this, unfortunately. Sep 16 09:24:31 looks like its not starting here either Sep 16 09:24:40 gar ! Sep 16 09:24:45 so i cant test the PPA changes Sep 16 09:24:50 * ogra curses Sep 16 09:36:06 how faster is native compilation on panda (es2.0 8layer) compared to normal beagleboard? Sep 16 09:36:51 hrw, likely an order of magnitude for big stuff, down to no improvement for little stuff. Sep 16 09:36:59 RAAAAHHHHHH !!!!!!! Sep 16 09:37:02 * Run oem-config-wrapper as the user oem rather than root. Sep 16 09:37:02 - Fixes issues with pulseaudio applet not working Sep 16 09:37:02 - Prevents root terminals from ctrl-alt-t (LP: #594233) Sep 16 09:37:14 * ogra goes mad Sep 16 09:37:39 persia: fighting ftfbs on bbc3 is pain, so I want to know how better it will be on panda Sep 16 09:38:25 hrw, night and day. I don't have a panda, but my powerpc is similar in specs to what has been announced, and I don't tend to have issues with that. I find bulding stuff on the beagle incredibly painful. Sep 16 09:38:56 hrw, kernel package builds for omap4 take about 6h on the buildds, they take 2.5h on a panda with a similar setup (USB disk) Sep 16 09:39:10 so i'D say twice as fast at least Sep 16 09:39:10 nice Sep 16 09:39:31 ~month of wait then Sep 16 09:39:48 though a kernel build isnt really representative since it doesnt do as much disk operations as most other packages ... i.e. CPU bound Sep 16 09:39:53 ogra, Do you have similar statistics for boost? kernel compiles don't need that much RAM. Sep 16 09:40:04 ogra: run openjdk ;D Sep 16 09:40:13 persia, no, its a stopwatch measure ricardo did Sep 16 09:40:18 openjdk is also a poor example. Sep 16 09:40:40 * ogra cant run *anything* atm due to that super silly oem-config change Sep 16 09:40:42 openoffice.org would be interesting for comparison. Sep 16 09:40:57 ogra, Grab a beta image and apt-get upgrade Sep 16 09:41:11 persia, doesnt help with jasper regression testing Sep 16 09:41:22 i need to find the issues before freeze Sep 16 09:41:23 No. Completely worthless for that. Sep 16 09:41:45 which wont happen since i dont expect any oem-config change today Sep 16 09:41:48 Eeek! File a bug immediately, and mark it critical, just in case. Sep 16 09:41:52 at least none thats in time Sep 16 09:44:01 persia: but I did build of openjdk on bbc3 ;D Sep 16 09:45:02 hrw, You have my sympathy, but really, it's not a fair comparison, as the build time there is hugely dependent on bootstrapping code, rather than CPU/Memory/IO Sep 16 09:45:47 kernel is CPU bound. Boost is typically memory bound. openoffice.org tends to switch back and forth (and requires lots of IO, so can be IO bound on vrey fast high-memory machines) Sep 16 09:49:15 ok Sep 16 09:53:59 ogra: hi there Sep 16 09:54:13 hey dmart Sep 16 09:54:44 I posted a possible fix to bug #587632 Sep 16 09:54:47 Launchpad bug 587632 in libmad (Ubuntu) "Sound very distorted on armel (affects: 2) (heat: 18)" [Undecided,Confirmed] https://launchpad.net/bugs/587632 Sep 16 09:54:56 don't know if you saw... are you in a position to test it? Sep 16 09:55:39 not right now, after the freeze i will be (since it's a bugfix and improvement it should be possible to still get it in) Sep 16 09:56:23 ok Sep 16 09:56:54 Does anyone else have a device running maverick with working audio? Could someone else help verify the fix above sooner? Sep 16 09:57:24 GrueMaster probably ... once he gets up Sep 16 09:57:36 but that might well be after freeze kicks in too Sep 16 09:57:50 Well, lots of folks have been talking about audio, and maybe other folks have beagles or something. Sep 16 09:58:22 yeah Sep 16 09:58:35 I have the board, but unforunately nothing to plug into it (not here, right now, anyway) Sep 16 09:59:02 i only have a panda running atm and there the audio driver is still screwed Sep 16 09:59:09 heh. Clearly you need to adapt to the using-headphones-on-the-mp3-player-all-the-time lifestyle :) Sep 16 09:59:27 ogra, berco's hacks make it work (the wrong way) if you just want audio to work. Sep 16 10:00:30 There was a small bit of Thumb-2 specific code which turned out to have a bug. I _think_ it should be fixed in my branch, but it doesn need testing ;) Sep 16 10:00:42 -> _does_ need testing Sep 16 10:04:16 mopdenacker, save your time, no need for a bug, installer team fixed it Sep 16 10:06:14 argh Sep 16 10:06:20 and the image only has jasper 0.20 Sep 16 10:06:45 which still has the broken SWAP file and no PPA stuff at all Sep 16 10:06:54 * ogra sighs and fires off a rebuild Sep 16 10:09:30 ogra: so current 20100916 is broken? you confirm? I am just flashing it now... Sep 16 10:10:29 ndec, when it fails you can try to get to a shell (hitting left ot right arrow at the splash), create a user called "oem" and reboot, that should get you a bit further Sep 16 10:11:01 ogra: well, we have an internal release tomorrow, and I need a working image... so 0914 is better, right? Sep 16 10:11:14 it didnt fully complete for me but at least you get a graphical login then Sep 16 10:11:25 yes Sep 16 10:11:37 though that will still miss lots of fixes from yesterday Sep 16 10:11:52 ogra: what kind of fixes? Sep 16 10:11:53 i'm not sure when evan will upload the installer fix Sep 16 10:12:02 they usually collect a bunch before uploading Sep 16 10:12:50 jasper saw some extra love ... we use a swapfile again, the PPA stuff was added ... the launcher saw many fixes, the default config too (a set of default launchers in the favorites) etc etc Sep 16 10:13:14 yesterday was possibly the most productive day of this cycle :) plenty of fixes Sep 16 10:14:29 sigh Sep 16 10:14:33 rebuild failed Sep 16 10:14:49 samba and ubuntuone are uninstallable Sep 16 10:14:58 what an annoying day ! Sep 16 10:20:04 ndec, PPAs enabled ! Sep 16 10:20:52 (which sadly doesnt help much with the broken images :/ but at least its one step more) Sep 16 10:27:02 ndec: Did you get around to finding the Blaze schematic for me please? Sep 16 10:28:21 lag, ndec Why do we need the Blaze schematic? Isn't that the one that already works? Sep 16 10:30:36 Exactly Sep 16 10:31:32 I've looked through the driver and nothing jumps out and shouts "change me" Sep 16 10:32:17 Well, that's not strictly true, but I have changed all the things I think need changing and still no audio Sep 16 10:32:55 I would be interested to hear from anyone who has had sound running on their 6 layer board Sep 16 10:33:10 I can't get mine working, even with the hacky files Sep 16 10:34:18 * ogra doesnt touch such old HW anymore :P Sep 16 10:34:28 8 layer FTW ! Sep 16 10:34:47 lag, i'll try later today ... :) Sep 16 10:34:49 Ya know, things'd likely work better if "old" was measured in terms of years rather than weeks. Sep 16 10:34:59 weeks ? Sep 16 10:35:12 I see that there are a few other outputs that are plumbed into the expansion ports Sep 16 10:35:12 i recived them within two days :) Sep 16 10:35:24 * lag wonders if sound is playing out from those instead of the headphone jack Sep 16 10:35:52 iirc we had sound defaulting to HDMI on the ES1 in prague Sep 16 10:35:59 i wonder if thats still the case Sep 16 10:36:05 * lag would love to work on the 8 layer, but no one sent him one Sep 16 10:36:07 * lag is sad Sep 16 10:36:16 * ogra comforts lag Sep 16 10:36:23 I have HDMI plugged in and turned up Sep 16 10:36:28 k Sep 16 10:36:36 As well as an external speaker on the headphone jack Sep 16 10:36:39 lag, Quite possibly. You can check with pamon Sep 16 10:36:40 And ... Sep 16 10:36:40 * ogra doesnt have speakers on his monitor Sep 16 10:36:41 Nothing Sep 16 10:37:29 I didn't think I did either, until one day they came on with white noise and scared the bejesus out of me Sep 16 10:37:39 heh Sep 16 10:37:40 quick'n'dirty trick is to mirror outputs to file sinks, and checking the content. Sep 16 10:37:50 Eh? Sep 16 10:38:01 How does one do that then? Sep 16 10:38:24 pamon lets one feed data into arbitrary sources and pull data from arbitrary sinks in parallel. Sep 16 10:38:42 no manpage, but `pamon --help` gives nice usage instructions. Sep 16 10:39:27 Oh, just packaging error: manpage is the pacat manpage. Sep 16 10:44:24 pamon just churns out lots of garbage Sep 16 10:44:34 Then crashes the Panda Sep 16 10:48:36 Well, that's not ideal. Hrm. Sep 16 10:57:21 Are there any tools I can use to debug? Sep 16 10:57:50 diwic told me about speaker-test, but that doesn't even work on my desktop machine Sep 16 11:01:38 Does anyone know if there an application that will let me play through a specific Card/Device number? Sep 16 11:01:57 I.e. though the devices listed in "aplay -l"? Sep 16 11:02:20 aplay Sep 16 11:02:51 aplay / arecord let you directly interact with ALSA Sep 16 11:02:59 paplay / parecord let you interact with pulse Sep 16 11:03:14 pacat / parec / pamon give you slightly lower level access to pulse Sep 16 11:04:33 Awesome Sep 16 11:04:44 Why don't people tell you these things sooner? Sep 16 11:05:06 I just managed to panic the kernel via aplay Sep 16 11:05:07 did you ask earlier? :) Sep 16 11:05:08 Doh! Sep 16 11:05:20 * persia has the unfortunate habit of being reactive: much information is available on request Sep 16 11:05:25 I've been asking all sorts of stupid questions for the past 24hrs Sep 16 11:05:46 persia: What is the meaning of life ... no, kernel development? Sep 16 11:05:48 ;) Sep 16 11:06:07 42 and trying to clean up after the HW vendors Sep 16 11:06:21 Err, scratch the first one: wrong question Sep 16 11:06:32 The correct answer would be "pies" Sep 16 11:06:44 2 * pies? Sep 16 11:06:59 If you like. One at a time is usually sufficient for me. Sep 16 11:07:00 2 * pies * r2 Sep 16 11:07:49 lag: gst is actually useful to test your audio config: 'gst-launch audiotestsrc ! alsasink' you can replace alsasink with pulsesink Sep 16 11:08:45 * persia likes to leave gst until after ALSA and pulse are working :) Sep 16 11:10:09 persia: of course... but with audiotestsrc you aren't taking any big risk! Sep 16 11:10:49 * lag just POOHed himself again Sep 16 11:10:56 * lag has sound from HDMI Sep 16 11:11:00 Ahhhhhhhhhhhhhhhhhhhhhhhhh Sep 16 11:11:17 Oh, probably not these days. I just had poor impressions from gst from back when PA was polypaudio, and got out of the habit of using it. I understand that most of the frustrating bugs have been fixed in the past couple years, but haven't been fiddling with it. Sep 16 11:11:23 lag, nice! Sep 16 11:11:41 lag: what did you do? are you using our custom config files? Sep 16 11:11:59 lag: which command do you use? Sep 16 11:12:15 * persia suspects lag of hacking the drivers directly Sep 16 11:12:17 I call it 'sound' it was actually white noise Sep 16 11:12:27 That's fine: indicates you connected *something* Sep 16 11:12:44 I have been hacking drivers, but I don't think this is a consequence of that Sep 16 11:12:57 What else would have caused it? Sep 16 11:13:05 unwashed ears ? Sep 16 11:13:16 Do: aplay -Dhw:0,6 -f dat /usr/share/sounds/ubuntu/stereo/desktop-login.ogg Sep 16 11:13:35 * lag tries more HW devices Sep 16 11:13:54 * lag ... and braces one's self Sep 16 11:14:10 lag, Um, does aplay handle ogg? manpage says it doesn't. Sep 16 11:14:52 manpage claims voc, wav, raw, and au Sep 16 11:15:01 (with wav as default) Sep 16 11:15:03 lag: ogg does not work, and would explain the noise... try a wav Sep 16 11:15:19 lag: locate *.wav Sep 16 11:15:47 /usr/share/sounds/alsa has a heap of .wav files for testing Sep 16 11:15:52 lag: aplay -Dxxx foo.wav Sep 16 11:16:23 And leave -f alone: you want to test that ALSA is correctly reporting the preferred format Sep 16 11:20:42 It complained when I didn't use the correct format Sep 16 11:20:43 Wait one Sep 16 11:21:26 lag: it's because aplay does not understand ogg. Sep 16 11:21:42 Grrrrrrrrrrrrrrrrrrrr Sep 16 11:21:49 [ 1013.521209] omapdss DISPC error: GFX_FIFO_UNDERFLOW, disabling GFX Sep 16 11:22:32 If it complains about the wrong format without -F that's a driver bug Sep 16 11:22:38 It really ought correctly autodetect it Sep 16 11:23:03 Perhaps it was because I was using ogg Sep 16 11:23:16 It's the file format ogra said worked out of the box Sep 16 11:23:25 But that was before I started playing with aplay Sep 16 11:23:25 lag, with a proper player Sep 16 11:23:34 use totem or some such Sep 16 11:23:52 No, don't. Sep 16 11:23:58 i doubt it works easily from cmdline without jumpig through some piping hoops Sep 16 11:24:06 *once* we confirm the lower-level layers are working, it's worth testing with .ogg and totem Sep 16 11:24:16 right Sep 16 11:24:42 when i said that i wasnt referring to cmdline tinkering Sep 16 11:24:42 ogra, `ogg123 foo.ogg` is the no-hoops command-line way, but that still depends on the lower level layers working. Sep 16 11:24:46 You can also use alsaplayer if you need a player with full ALSA support. Sep 16 11:24:59 This one supports ogg and other formats IIRC. Sep 16 11:25:02 persia, does that use pulse and gstreamer ? Sep 16 11:25:12 * ogra doubts it Sep 16 11:25:13 ogra, pulse, not gst Sep 16 11:25:23 alsaplayer has a CLI. Sep 16 11:25:39 Channels count not available Sep 16 11:25:45 * ogra tries to find some belated breakfast Sep 16 11:26:10 lag, That indicates a detection failure: are you exporting that from the driver? Sep 16 11:26:56 .name = "HDMI", Sep 16 11:26:56 .playback = { Sep 16 11:26:56 .stream_name = "Playback", Sep 16 11:26:56 .channels_min = 2, Sep 16 11:26:56 .channels_max = 8, Sep 16 11:26:57 .rates = SNDRV_PCM_RATE_48000, Sep 16 11:26:59 .formats = SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S32_LE, Sep 16 11:28:03 It says the same even when I use -c2 Sep 16 11:28:06 I may have found another issue... With the preinstalled images, the rootfs gets resized twice. After reboot, you still get: Sep 16 11:28:30 Begin: Running /scripts/init-premount ... done. Sep 16 11:28:31 Begin: Mounting root file system ... Begin: Running /scripts/local-top ... done. Sep 16 11:28:31 Begin: Running /scripts/local-premount ... [ 9.730651] mmcblk1: p1 p2 Sep 16 11:28:31 Resizing root filesystem. Please[ 13.573852] EXT3-fs: barriers not enabled Sep 16 11:28:31 wait, this will take a moment ... Sep 16 11:28:51 This takes something like 15-20s Sep 16 11:28:57 while it has already been done. Sep 16 11:29:28 Again, I can't test on the Panda... Would this be worth another bug? Sep 16 11:29:52 ogra, can help you with that Sep 16 11:30:02 mopdenacker, Please, file bugs for everything. Worst case they get rejected. Sep 16 11:30:14 If jasper has done it's job properly it will only restart once Sep 16 11:30:26 Personally, I think that one doesn't matter: the resize is to the *same* size, and not doing that means complicated self-modifying code in jasper Sep 16 11:31:04 Right, that's acceptable if this just happens once. Sep 16 11:31:57 persia: thanks! I was right to ask you :-) Sep 16 11:32:11 Sure, but it doesn't hurt to have a wontfix bug in LP to document this for the next person :) Sep 16 11:34:26 All right. On which project? Jasper? Sep 16 11:35:39 mopdenacker: if you test on blaze, that could explain why you see the problem. i suppose that something in the boot.scr tells uinitrd to do the resize. after resize boot.scr is regenerated and flashed, but flash will not work on blaze... so you might end up rebooting with the original boot.scr. ogra, could that be correct? Sep 16 11:35:48 Wooooooooooooooooooooooooo Sep 16 11:35:53 SOUND! Sep 16 11:35:57 mopdenacker, yep: `ubuntu-bug jasper` Sep 16 11:36:15 lag, Nice work. Userspace hacks, or all kernel? Sep 16 11:36:41 persia: that is indeed the right question ;-) Sep 16 11:36:56 ndec, Can you think of any reason not to have some open bugs for Blaze? I'm not sure they can all be closed, given the information available, but tracking them can't hurt Sep 16 11:37:16 lag: with hw:0,0 you should be able to test headset too (this is the lower jack input on the board). Sep 16 11:37:30 persia: no reason, since we do want blaze to be supported ;-) Sep 16 11:38:04 Nope Sep 16 11:38:06 ndec, I think I agree with your diagnosis of the issue, but based on that, let's try to get the discussion in the LP bugs :) Sep 16 11:38:12 Only hw:0,6 works Sep 16 11:38:15 HDMI Sep 16 11:38:39 I will have a look and see what I can do within the driver to get the jack working Sep 16 11:39:04 persia: one problem we've seen is that flash_kernel will use mmcblk0p1, and on blaze this is eMMC, not SD Card. so you end up flashing in eMMC new uboot, boot.scr, uImage, ... but you will boot from SD card, and you end up booting with the old versions. Sep 16 11:39:12 ogra: ^^^^ Sep 16 11:39:18 lag, If you look at the shell script with the `amixer ...` calls, you might be able to figure out which settings are incorrect from autodetect by getting sound item by item. Sep 16 11:39:48 Judging from the schematic, I believe the lower jack to be hw:0,9 Sep 16 11:39:57 ndec, So, how can we differentiate Blaze from Panda? That logic needs to go into flash_kernel (and that needs a bug against flash_kernel detailing the issue) Sep 16 11:39:58 lag: can you pastebin a diff? and can you confirm that you don't use our custom default.pa and amixer.sh Sep 16 11:40:39 persia: blaze and panda are different machine (at kernel level). so they are different in /proc/cpuinfo Sep 16 11:40:51 No, I am using default.pa and amixer.sh Sep 16 11:40:53 ndec, Lovely. Please file a bug :) Sep 16 11:43:25 ndec, and please attach the /proc/cupinfo output there Sep 16 11:43:51 Or even a patch, if you're feeling adventerous :) Sep 16 11:44:07 mopdenacker, what ndec describes above is most likely your bug Sep 16 11:44:45 persia, well, we should ship bleach with jasper if we expect others to look at the code :P Sep 16 11:45:02 persia: against which package? Sep 16 11:45:07 ogra, With a change to flash_kernel, it doesn't need a jasper hack Sep 16 11:45:15 ndec, jasper-initramfs Sep 16 11:45:20 ndec, Ideally a patch to flash_kernel to make it work on Blaze Sep 16 11:45:26 ogra, Why? Sep 16 11:45:26 no Sep 16 11:45:30 Why not? Sep 16 11:45:37 jasper creates the flash-kernel conf Sep 16 11:45:49 which flash-kernel reads the device from Sep 16 11:45:56 But flash-kernel *also* has machine-detection logic Sep 16 11:46:33 yes, it needs to learn about blaze, but the .conf needs to be written by jasper pointing to emmc Sep 16 11:46:35 and, actually, for the Blaze, where one apparently has both SD and eMMC, it becomes *interesting* to use that logic. Sep 16 11:46:44 persia: ogra: I think there are 2 problems. #1) we want the image to work on blaze from SD card, #2) we want the image to work on blaze when flashed from emmc Sep 16 11:46:51 it doesnt need machine detectiuon since NCommanders changes anymore Sep 16 11:46:52 ndec, Please file a bug with *two* tasks (Also Affects Distribution ...) Sep 16 11:46:56 so we use that Sep 16 11:46:59 no Sep 16 11:47:06 please only against jasper Sep 16 11:47:27 * persia remains unconvinced that the unrecognised fallback should be used when we are able to recognize something Sep 16 11:47:35 ogra: persia: we will need machine detection if we want to support flash in eMMC on blaze Sep 16 11:47:52 ndec, right and we have a bug for that already Sep 16 11:48:10 ogra: which one? Sep 16 11:48:18 i cant look it up from here Sep 16 11:48:18 which bug ID, i mean? Sep 16 11:48:37 ill tell you after i fnished eating my breakfast :) Sep 16 11:49:02 ogra_cmpc has exceedingly limited local information: one really needs to ask ogra, who is unfortunately AFK Sep 16 11:49:38 persia, feel free to look search in ubuntu-armel bugs (i dont do anything different) Sep 16 11:49:49 -look Sep 16 11:50:03 bugs.launchpad.net/~ubuntu-armel/+bugs ? Sep 16 11:50:10 Okay, now I can confirm that I am _not_ using default.pa and amixer.sh Sep 16 11:50:16 no idea, i use evolution :) Sep 16 11:50:38 * persia repeats the mutt motto, with gusto Sep 16 11:50:43 But I don't think my changes would have done anything Sep 16 11:50:46 Oh, and without the "less" bit Sep 16 11:50:49 I think HDMI worked before Sep 16 11:50:57 ndec: Please confirm? Sep 16 11:51:44 ndec, Maybe bug #615773 ? Sep 16 11:51:46 Launchpad bug 615773 in initramfs-tools (Ubuntu Maverick) (and 3 other projects) "flash-kernel fails to handle raw boot partitons on eMMC (affects: 1) (heat: 120)" [Undecided,Invalid] https://launchpad.net/bugs/615773 Sep 16 11:51:57 persia, yep, that one Sep 16 11:52:22 lag: with our default.pa and amixer.conf audio works with the daily image. Sep 16 11:52:43 From the lower audio jack? Sep 16 11:53:58 Can you email me all the files I need Sep 16 11:54:19 The only audio I've _ever_ heard was from HDMI Sep 16 11:54:24 from lower headset and hdmi too. Sep 16 11:54:44 lag: berco has the latest files. berco can you please send them to lag Sep 16 11:54:56 It would be advantageous for me to at least ensure the port works with the usespace hacks Sep 16 11:55:25 You don't need those files to retrieve audio from HDMI Sep 16 12:07:13 ndec, "board" is non capitalized on the blaze ? Sep 16 12:07:22 its "Board" on the panda Sep 16 12:14:16 * persia grumbles about ports.ubuntu.com being painfully slow today Sep 16 12:32:09 ogra: mopdenacker can confirm, I am running a old version. Sep 16 12:34:10 Still nothing from berco Sep 16 12:34:12 Is he around? Sep 16 12:39:01 lag: ndec: sorry. I wasn't looking at my desktop screen. Pandaboard is my new desktop :) Sep 16 12:39:25 How do you want the files? In Launchpad? Sep 16 12:40:02 How different are they from the ones you added yesterday? Sep 16 12:40:23 Alternately, should the ones from yesterday be sufficient to allow lag to verify operation before fixing the driver? Sep 16 12:40:28 lag: files attached to bug 637947 are the correct ones Sep 16 12:40:29 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 16 12:42:07 persia: yeop, same files as yesterday Sep 16 12:42:31 Cool. Sep 16 12:43:46 Great ta Sep 16 12:46:32 berco: you can use pidgin on the panda ;-) Sep 16 12:47:22 or install xchat if you want a sane IRC app Sep 16 12:53:16 * persia points out that smuxi can allow someone to use the same IRC client (with the same identify, and simulataneous response) on both machines, if one likes. Sep 16 12:53:37 That said, it might be a bit interesting on armel, as it's Mono, so it could be used to help debug that :) Sep 16 12:57:39 ogra, ndec : I confirm: on the Blaze, /proc/cpuinfo lists "OMAP4430 4430SDP board" ("board" is not capitalized) Sep 16 12:58:13 funny Sep 16 13:01:25 NCommander, can you bump https://edge.launchpad.net/ubuntu/+source/ubiquity/2.3.19/+build/1962841 ? Sep 16 13:07:17 NCommander, and https://edge.launchpad.net/ubuntu/+source/samba/2:3.5.4~dfsg-1ubuntu6/+build/1962200 Sep 16 13:07:47 lag: thanks a lot for your email... Sep 16 13:09:12 No problem Sep 16 13:09:22 I think it would be better if you were there in person :) Sep 16 13:12:19 NCommander, ignore the above, pitti did it already Sep 16 13:19:35 ndec, can anyone from your team upload something to https://edge.launchpad.net/~tiomap-dev/+archive/release ? Sep 16 13:19:51 (it will only then generate the key file i need) Sep 16 13:20:45 (anything will do, even a dummy package) Sep 16 13:21:20 hello-debhelper :) Sep 16 13:21:30 yeah, something like that Sep 16 13:22:02 sigh, 76 packages in queue Sep 16 13:22:08 22h behind Sep 16 13:22:21 i hate freezes ! Sep 16 13:23:09 (or not having 20 pandas as buildds) Sep 16 13:23:36 Um, why do you hate freezes? freezes is when the buildds catch up best. Sep 16 13:24:01 well, i hate the two days before freezes Sep 16 13:24:05 and the freeze day Sep 16 13:24:11 Or do you just dislike your fellow developers habit of uploading *everything* just before freeze? Sep 16 13:24:18 *until* the buuldds have catched up Sep 16 13:24:31 i do the same, so i'm not allowed to complain Sep 16 13:24:49 Obviously the solution is to make sure your favorite architecture is the fastest, and use that to build arch: all stuff :) Sep 16 13:24:57 actually ... i hate queues ! Sep 16 13:25:00 bah Sep 16 13:25:20 i assume you would also hate to wait if there were no queue ;) Sep 16 13:25:47 well, we should have a lottery system ... uploads only happen if qUeue < 2 Sep 16 13:25:58 and you would probably not hate to wait if all was fine ;) so you really hate that things are broken ;) Sep 16 13:26:02 ELSE THEY GET REJECTED Sep 16 13:26:05 OOPS Sep 16 13:26:09 WHATS THAT 1 Sep 16 13:26:11 ogra, Folks would just create loop scripts to work around that. Sep 16 13:26:11 hehe Sep 16 13:26:13 brute force uploads Sep 16 13:26:25 bah Sep 16 13:26:31 i had to re-plug my kbd Sep 16 13:26:36 to get caps off Sep 16 13:26:41 The other option would be to patch launchpad to force publication of all architectures in sync... Sep 16 13:26:52 i should upgrade to maverick, lucid is to unstable :P Sep 16 13:27:15 you run lucid? interesting Sep 16 13:27:36 on my laptop ? sure Sep 16 13:27:51 i have enough maverick machines around me Sep 16 13:27:57 heh Sep 16 13:28:16 they never get stuck in caps though ... Sep 16 13:28:23 (if i use ssh :P ) Sep 16 13:28:49 haha Sep 16 13:31:25 ogra: i uploaded a package. Sep 16 13:34:41 ndec, merci ! Sep 16 13:34:58 ogra: bitte. Sep 16 13:35:05 :) Sep 16 13:46:58 mopdenacker: thanks for checking the rootstock bugs and helping improving it :-) Sep 16 13:47:32 rsalveti: my improvements were tiny. Thanks anyway! Sep 16 13:48:03 all contributions are welcome :-) Sep 16 13:50:52 ndec, btw, your PPA builders live at the very bottom of https://edge.launchpad.net/builders Sep 16 13:51:32 ogra: what do you mean? Sep 16 13:51:45 you have dedicated PPA buildds Sep 16 13:52:01 ogra: how many? Sep 16 13:52:05 if you look for a running build on the builders page you find them at the bottom Sep 16 13:52:08 two Sep 16 13:52:15 shared with other PPA folks Sep 16 13:52:34 ogra: other tiomap-dev PPA, you mean/ Sep 16 13:52:46 before all builds were done by the main builders Sep 16 13:52:58 now all PPAs are separated out to build on these two machines Sep 16 13:54:12 having the PPAs on the main builder meant that builds went in the very end of the queue with a very low prio Sep 16 13:54:15 This is a very good thing :) Sep 16 13:54:21 yes Sep 16 13:55:03 ogra: deb is ready in tiomap-dev/release. Sep 16 13:56:11 ndec, i only needed the key file :) Sep 16 13:56:27 the PPA only generates initially it if an upload happened Sep 16 13:57:00 ogra: ok, so you have it anyways ;-) Sep 16 13:57:14 yeah, already uploaded jasper with the key in it := Sep 16 13:57:16 :) Sep 16 14:02:35 huge queue for armel, and one disabled builder :-( Sep 16 14:02:43 ha, finally all three packages that break builds are buiolding now Sep 16 14:02:53 rsalveti, yeah ... Sep 16 14:03:03 and qt building for lucid Sep 16 14:03:18 cool, we can then have a new image Sep 16 14:03:45 we want PANDA ! * * * we want PANDA ! * * * we want PANDA ! * * * we want PANDA ! * * * we want PANDA ! * * * we want PANDA ! * * * we want PANDA ! * * * Sep 16 14:04:23 * ogra proposes a buildd for every package in main ! Sep 16 14:04:35 no more delays ! never ! Sep 16 14:04:51 haha :-) Sep 16 14:05:07 (probably more power costs than travle costs for the company though) Sep 16 14:08:10 hmm, linux-ti-omap4 seems stuck in NEW Sep 16 14:08:18 :-( Sep 16 14:08:43 well, thats easier than waiting for a build :) Sep 16 14:10:22 ogra: btw, don't know if you saw my test message yesterday, but latest u-boot-linaro also works well for omap3 Sep 16 14:10:39 so we could just start using it instead of maintaining our own Sep 16 14:10:46 and we got a new upload yesterday Sep 16 14:10:49 yeah, saw that (i must admit i forgot about it again, thanks fo rthe reminder) Sep 16 14:10:56 probably still in the queue Sep 16 14:11:06 let me check Sep 16 14:11:29 https://edge.launchpad.net/ubuntu/+source/u-boot-linaro/2010.09~rc1.1-0ubuntu2/+build/1962151 Sep 16 14:11:33 yup, didnt build yet Sep 16 14:12:27 i'll change the build scripts as soon as we have it Sep 16 14:12:34 then we can drop the old cruft Sep 16 14:12:40 yup, cool Sep 16 14:12:51 and we'll finally have the led working at panda Sep 16 14:12:59 yeah Sep 16 14:13:10 did you enable heartbeat too ? Sep 16 14:13:15 yup Sep 16 14:13:18 bah Sep 16 14:13:23 * ogra finds that annoying Sep 16 14:13:28 but it's useful Sep 16 14:13:36 but probably helpful for developers Sep 16 14:13:40 yeah Sep 16 14:13:45 you can easily notice when you got a kernel panic Sep 16 14:13:57 but its the one reason my XM doesnt stay on all the time Sep 16 14:14:27 hm Sep 16 14:14:54 i dont like the blinking on my desk all the time ... but thats only me Sep 16 14:15:11 haha, I think it's kind of cool Sep 16 14:15:16 makes me know that the board is alive Sep 16 14:15:33 sure, but its also distracting when working Sep 16 14:15:37 and that's a good thing when we're running unstable software Sep 16 14:15:50 sure Sep 16 14:15:55 haha, I have another table at my side, so I don't get distracted Sep 16 14:16:05 i do too Sep 16 14:16:18 but the boards i mainly work with live on my desk next to me Sep 16 14:16:29 argh, process update-apt-xapi eating all my beagle Sep 16 14:16:44 mnam mnam Sep 16 14:17:11 but software-center will be breezing fast :) Sep 16 14:17:25 if you get it started and it doesnt OOM indeed Sep 16 14:17:26 hahaha, that's something I don't need now, for sure Sep 16 14:17:34 using swap Sep 16 14:17:40 heh, yeah Sep 16 14:17:41 it'll work, but slooooooooow Sep 16 14:19:00 * rsalveti wants a real working 512MB xM Sep 16 14:40:25 https://lists.ubuntu.com/archives/ubuntu-devel/2010-September/031431.html Sep 16 14:41:31 hm, almost 2 hours :-) Sep 16 14:41:54 17:00? It's almost always been at 0:00 previously. I wonder why... Sep 16 14:43:24 persia, because robbiew was grateful and didnt lock the door at 0:00 last night :) Sep 16 14:44:04 * ogra ponders if he should pull the efl launcher text fix in untested Sep 16 14:44:12 just for the excitement :) Sep 16 14:45:31 Don't. Sep 16 14:45:51 Anything that is really broken can still be uploaded post-FinalFreeze Sep 16 14:45:52 persia, its only two lines :) Sep 16 14:46:08 Sure, but why now vs. after testing? Sep 16 14:46:16 like a jasper rewrite ou mean ? Sep 16 14:46:19 *you Sep 16 14:46:24 Anything. Sep 16 14:46:56 I've seen complete repackaging efforts, with new package splits and new upstream versions happen, as long as they would have to happen anyway (high impact, release critical, security, etc.) Sep 16 15:02:19 * rsalveti lunch Sep 16 15:08:14 hi, i am trying to boot ubuntu fs ... I am getting error when did a sudo .. any one faced this type ... Sep 16 15:08:24 ubuntu@ubuntu:~$ sudo su Sep 16 15:08:24 sudo: can't open /etc/sudoers: Permission denied Sep 16 15:08:24 sudo: no valid sudoers sources found, quitting Sep 16 15:11:32 What filesystem are you using? How did you write it to your rootfs media? Sep 16 15:15:04 i picked up one of the .gz files uploaded in share point ... so not sure how they wrote into rootfs :-( ... i did sudo tar -xzvf fs.gz ... Sep 16 15:15:40 Where did this image come from? What platform are you trying to boot? Sep 16 15:16:29 avinashhm, I'm unsure, and don't tend to trust random stuff in share point :) I've seen that a number of times, but always when someone made a mistake preparing the rootfs, and somehow stripped the setuid bits. Sep 16 15:18:18 GrueMaster, i am picking it from http://ubuntu-omap.tif.ti.com/rootfs/maverick/ .. i am not sure if this link is accessible from outside .. i am booting on OMAP ...OMAP4 .. Sep 16 15:18:43 ndec, Any idea what is happening for avinashhm ? Sep 16 15:19:07 persia, is there any link from where i can pick standard ubuntu minimal FS for arm based SOC .. omap specifically ... Sep 16 15:19:22 No. All the images are big flashy ones. Sep 16 15:19:34 heh. Pretty much. Sep 16 15:19:51 persia, my .gz file is 80mb .. ytpically how much are others ... Sep 16 15:19:54 * ogra_cmpc would call them rather mis-sized Sep 16 15:19:57 err Sep 16 15:19:59 heh Sep 16 15:20:05 *mid-sized Sep 16 15:20:19 The current compressed image is 537M Sep 16 15:20:20 avinashhm, 520M Sep 16 15:20:23 The most well-tested image for omap is probably http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/current/ but today's happens to not boot. Sep 16 15:20:48 the 20100914 should work fine Sep 16 15:20:49 I would recommend http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/20100914/maverick-preinstalled-netbook-armel+omap4.img.gz Sep 16 15:20:56 It is well tested and works. Sep 16 15:22:40 GrueMaster, ogra_cmpc, .. i am downloading this ... let me try ... what i need to do after downloading ... any specific instructions ... Sep 16 15:23:13 avinashhm, there is a link on the page http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-preinstalled/current/ Sep 16 15:23:23 linking to instructions Sep 16 15:23:49 though note that these images are specifically for panda ... Sep 16 15:24:12 blaze or other hardware might or might not work atm Sep 16 15:24:40 oh, and even more specifically ES2.0 omap4 Sep 16 15:24:55 ogra_cmpc,found the link ...i ll follow ... huff .. i am testing on SDP .. :-( ... Sep 16 15:25:35 we were reported an issue with the eMMC today that might bite you Sep 16 15:26:13 i wont' use emmc .. i ll use MMC .. As much as possible, let me try to avoid the biting :-) Sep 16 15:26:19 (the kernel handling script expects the vfat partition to live on mmcblk0p1 (which isnt on the SD on blaze)) Sep 16 15:27:04 what type of kernel handling scripts .. not aware ... Sep 16 15:27:09 avinashhm: you are using TI internal releases... this is not the right place to discuss this... Sep 16 15:28:08 ndec, sorry .. but have been stuck with this from morning .. i could think of this channel only ... Sep 16 15:30:31 GrueMaster, heh, so oregon saved moses ? Sep 16 15:30:38 * ogra_cmpc watches news and giggles Sep 16 15:30:43 Buhh? Sep 16 15:32:51 ogra_cmpc: What are you talking about? Sep 16 15:34:21 * GrueMaster wants whatever ogra_cmpc is taking. Sep 16 15:34:23 seems a camel called moses got stuck in a sinkhole ... Sep 16 15:34:31 Oh. That. Sep 16 15:34:44 was apparently a big enough story to end up on german tv Sep 16 15:35:24 hedalinse was "oregon saves moses" Sep 16 15:38:14 The headline here was "Rescuers dig camel out of sink hole in Oregon City". I never read it because it was not that interesting. Sep 16 15:38:48 Especially with all the other news about animal rescues here almost weekly. Sep 16 15:39:09 yeah, i wouldnt have cared for it either ... but it made it to german news Sep 16 15:40:20 This is far more interesting. http://www.urbandaddy.com/la/food/11245/__:url Sep 16 15:42:21 hey ! i'm trying to lose weight ! Sep 16 15:43:15 drive-by-bacon? Sep 16 15:43:32 fatty. :P Sep 16 15:48:26 Who here has sound working on their ES2.0 boards? Sep 16 15:48:34 rsalveti, i'd say go for DKMS for the kernel mods, its unlikely the kernel team will do an upload just for omap Sep 16 15:48:35 With or without the hacky files Sep 16 15:49:25 Not me. Sep 16 15:49:45 I'll run an update to see if I can pull in any fixes. Sep 16 15:50:55 Note I don't have an HDMI monitor to test HDMI Audio. Sep 16 15:52:51 I'm not interested in HDMI sound Sep 16 15:53:13 grrr. My panda has a crash report with core dump, but apport can't handle it. Sep 16 15:53:16 Basically, I have _never_ heard sound from the Headphone Jack Sep 16 15:53:29 With or without the hacky files Sep 16 15:53:40 So I don't know if I've repaired it or not Sep 16 15:55:32 ndec: berco: If I'm not mistaken, those hacky files are all Pulse Audio related right? Sep 16 15:56:29 lag: no... default.pa and daemon.conf are pulse audio, but you have the amixer.sh shell script that needs to run at least once to configure alsa. Sep 16 15:56:52 lag: still no audio? Sep 16 15:56:59 lag: which board rev do you have? Sep 16 15:57:05 daemon.conf is mostly just a performance enhancement, and ought be ignored (there is a better way to fix it for pulse, already waiting for upload) Sep 16 15:57:06 I have it via HDMI Sep 16 15:57:14 But I am very new to this stuff Sep 16 15:57:25 default.pa is workarounds for the broken driver, because it's not being autodetected, mostly. Sep 16 15:57:28 I'm learning, but still don't really know what's going on Sep 16 15:57:31 lag: what kind of speakers are you using on the headset jack? Sep 16 15:57:50 the amixer.sh script contains all the hints that ought be required to create the machine driver for ASoC Sep 16 15:58:08 lag: you can also switch from headset to hdmi from the Volume Applet in the UI Sep 16 15:58:36 prpplague: I've tried earphones and an external speaker Sep 16 15:59:04 berco: How do you do that then? Sep 16 15:59:30 lag: just so that you know, the audio output on panda is the bottom plug on the connector Sep 16 15:59:44 lag: go with you mouse on the volume icon Sep 16 16:00:12 lag: then sound preferences and you have an output tab where you should see HDMI and Onboard_speakers Sep 16 16:00:20 Nope Sep 16 16:00:25 Just Onboard speakers Sep 16 16:00:29 lag: you can switch dynamically b/w the 2 Sep 16 16:00:50 I don't see HDMI Sep 16 16:00:53 lag: have you run the "amixer.sh -a" script? Sep 16 16:02:17 I ran amixer-settings.sh Sep 16 16:02:22 That robclark sent me Sep 16 16:02:43 lag: and you are using a es2.0 8-layer board?? Sep 16 16:02:54 Nope Sep 16 16:02:55 6 Sep 16 16:03:54 lag: check to see if L28 directly behind the headset jacks is populate with a square chip or if there are two solder bridges there Sep 16 16:05:30 There is nothing between the solder points Sep 16 16:05:48 lag: I think you need "-a" option to run the script Sep 16 16:07:05 prpplague: Don't tell me my board in incapable of sound? Sep 16 16:07:05 * ogra whistels innocently and quickly adds serial getty handling to jasper 1h before final freeze ... to make ndec happy Sep 16 16:07:16 lag: ok so there are 4 solder points and nothing is connected on any of them? Sep 16 16:07:36 ogra: what change exactly are you doing for getty? Sep 16 16:07:38 I only see 2 Sep 16 16:07:51 U22 and L38 Sep 16 16:08:14 ndec, adding an /etc/init/serial.conf script if jasper detects console= on the cmdline Sep 16 16:08:35 lag: L28 is a super small pad, it should have two solder bridges running across the board Sep 16 16:08:44 ogra: ok ... Sep 16 16:08:45 ndec, so if console= is set a getty will come up by default Sep 16 16:09:04 lag: let me take a picture Sep 16 16:09:08 ogra: and console can be set at uboot prompt , right? no need to update boot.scr? Sep 16 16:09:12 as i said, i wont change the defaults for boot.scr ... but that addition should ease your life Sep 16 16:09:46 you should always use /boot/boot.script and run sudo flash-kernel Sep 16 16:10:16 but indeed, if you hack in all the boot.scr commands manually, you can set the cmdline as you like on uboot prompt Sep 16 16:13:21 lag: http://imagebin.ca/view/C-Otvqp.html Sep 16 16:13:27 lag: that is with a filter populated Sep 16 16:13:43 prpplague: I've sent you an email Sep 16 16:15:01 That tiny square chip _is_ on my abord Sep 16 16:15:04 board* Sep 16 16:15:26 lag: ok the filter is populated, i suspect that you have a bad filter Sep 16 16:15:48 lag: we dropped that part from the 8-layer board because we were having constant quality issues with it Sep 16 16:15:58 :( Sep 16 16:16:17 lag: got anyone who can do a board re-work? Sep 16 16:16:58 lag: the filter needs to be removed and soldered across two sets of pads http://imagebin.ca/view/rVkPnEG.html Sep 16 16:17:41 Nope Sep 16 16:18:32 I could probably do it, but I'd need permission from someone up-top Sep 16 16:20:31 ndec, how is the external SD called on blaze ? mmcblk1 ? Sep 16 16:21:09 ogra: what i mean, is that I don't want to modify boot.scr, but at uboot prompt set console to whatever I want. then the default scripts in boot.scr can read the console var. Sep 16 16:21:21 ogra: i forgot to hit enter a few mins ago... Sep 16 16:21:25 that wont work Sep 16 16:21:31 ogra: yes SD is mmcblk1 on blaze Sep 16 16:21:36 boot.src sets the options Sep 16 16:21:44 so it would overwrite Sep 16 16:21:49 ogra: it cannot inherit a variable? Sep 16 16:22:02 we could do such a change but i guess we wont get allowed for mavverick anymore Sep 16 16:22:12 final freeze is in 40min Sep 16 16:22:38 prpplague: Can the kernel auto-detect (un)plugging of the Headphone Jack? Sep 16 16:22:47 lag: no Sep 16 16:23:13 lag: the original design didn't have that support and we were too far along when i came aboard to get that corrected Sep 16 16:23:21 k Sep 16 16:24:21 prpplague: Perhaps you can give me a more full answer Sep 16 16:24:44 prpplague: Why does something that works on Blaze, not work on Panda? Sep 16 16:25:10 Despite the same Audio chip and CODEC being in use Sep 16 16:25:49 The only thing I found that looked incorrect was: snd_soc_dapm_disable_pin(codec->dapm, "Headset Stereophone"); Sep 16 16:26:02 Which I've changed to enable Sep 16 16:28:31 lag: if you have changed that to enable, then your headset will not work Sep 16 16:28:35 lag: on the panda Sep 16 16:28:38 ndec, bug 640406 should be fixed with jasper 0.25 (just uploaded) Sep 16 16:28:39 lag: the same thing does not work on blaze either... you still need to run the custom scripts on blaze too Sep 16 16:28:39 Launchpad bug 640406 in jasper-initramfs (Ubuntu) "Flashing OMAP4 kernel uses the wrong partition on Blaze (affects: 1) (heat: 12)" [Medium,Confirmed] https://launchpad.net/bugs/640406 Sep 16 16:29:25 ogra: thx. we will test when it's in a daily Sep 16 16:29:44 ndec: Ah okay - that makes sense Sep 16 16:29:48 should be tomorrow (if the queue cleans up until then) to saturday Sep 16 16:29:56 s/to/or/ Sep 16 16:29:59 ndec: So has this driver ever worked on its own? Sep 16 16:30:26 lag, no, thats the prob Sep 16 16:30:35 lag: yes, it works. we can have audio on speaker or hdmi. the problem is that it requires the custom config which is not standard. Sep 16 16:30:45 it always needed handholding through alsa hacks Sep 16 16:31:05 That clears lots up Sep 16 16:31:14 (and pulse ... but thats a fallout of the broken alsa config) Sep 16 16:31:15 lag: I just booted 20100914, and added the 2 pulse scripts and ran amixer.sh, and I have audio on the panda HDMI and speaker Sep 16 16:31:45 Does the daily build work? Sep 16 16:31:58 14 is the last one that works Sep 16 16:32:02 k Sep 16 16:32:18 15 didnt build, 16 is screwed due to oem-config being broken Sep 16 16:36:13 * ndec leaving for now... will be back later Sep 16 16:48:03 good morning all (or afternoon, or evening) :) Sep 16 16:55:48 hi, is there any default password for root login ... ?? Sep 16 16:56:20 ubuntu locks root by default Sep 16 16:56:35 if you use the ti image there might be internal info ... Sep 16 16:56:53 for the ubuntu image, make sure to finish the oem-config tool on the screen after first reboot Sep 16 16:59:17 ogra, not sure ... with ubuntu login, used ubuntu password ... with root login, i am not sure ... Sep 16 16:59:32 try sudo Sep 16 16:59:40 and the password you used Sep 16 17:00:59 generally we use sudo for everything that needs root access Sep 16 17:01:04 yeah .. tried .. ubuntu@ubuntu:~$ sudo /etc/cron Sep 16 17:01:04 sudo: can't open /etc/sudoers: Permission denied Sep 16 17:01:04 sudo: no valid sudoers sources found, quitting Sep 16 17:01:04 ubuntu@ubuntu:~$ Sep 16 17:01:05 yeah Sep 16 17:01:26 ls -l /etc/sudoers Sep 16 17:01:29 hm Sep 16 17:01:33 check permissions Sep 16 17:01:42 missing file or missing being into sudo group Sep 16 17:02:02 should be -r--r----- and owned by root.root Sep 16 17:02:34 just FYI everyone ... final freeze is in effect !!! Sep 16 17:02:55 oh sooo cold ... Sep 16 17:02:58 treid .. ubuntu@ubuntu:~$ ls -l /etc/sudoers Sep 16 17:02:58 -r--r----- 1 root root 513 2010-09-16 14:21 /etc/sudoers Sep 16 17:02:58 ubuntu@ubuntu:~$ Sep 16 17:03:36 lag: make sure you un-do that change of the headset plug detect Sep 16 17:03:54 there is chance right ... in sudoers, may be ubuntu isn't a sudoer .. so it's telling permission denied ??? Sep 16 17:04:11 prpplague: Already done :) Sep 16 17:04:19 avinashhm, yeah, that could be Sep 16 17:04:22 lag: okie crokie Sep 16 17:05:10 ogra, wait i ll cat the file in my PC .. just to see ... Sep 16 17:05:32 ubuntu uses the admin group for sudo Sep 16 17:05:37 check thats in there Sep 16 17:05:51 and your user needs to be in that group Sep 16 17:07:02 can i send a file in IRC ???never tried though Sep 16 17:07:19 is the sudo binary correctly setuid? Sep 16 17:08:08 avinashhm, paste.ubuntu.com Sep 16 17:08:16 and paste the link here then Sep 16 17:10:03 checked the sudo binary ... Sep 16 17:10:05 ubuntu@ubuntu:~$ ls -l /usr/bin/sudo Sep 16 17:10:05 -rwsr-xr-x 2 root root 106996 2010-08-31 22:46 /usr/bin/sudo Sep 16 17:10:05 ubuntu@ubuntu:~$ Sep 16 17:10:51 Hrm. Sep 16 17:11:02 And annoyingly, we can't look at sudoers without root access. Sep 16 17:11:03 This is how my sudoers looks ... Sep 16 17:11:05 http://paste.ubuntu.com/494864/ Sep 16 17:11:22 looks good too Sep 16 17:11:24 persia, i am looking it through my PC ubuntu ... catted ... Sep 16 17:11:31 what does the groups command give you ? Sep 16 17:12:06 ubuntu@ubuntu:~$ groups Sep 16 17:12:06 ubuntu adm dialout cdrom floppy audio dip video admin Sep 16 17:12:06 ubuntu@ubuntu:~$ Sep 16 17:12:32 is your filesystem mounted readonly possibly ? Sep 16 17:12:32 ogra, not sure, anything missing ??? Sep 16 17:12:43 no, you are in admin Sep 16 17:12:50 looks all good, should work Sep 16 17:13:32 my only idea it that the card is locked or your fs was remounted readonly because fsck found errors Sep 16 17:14:37 my boot args .. so its RW only ...console=ttyO2,115200n8 root=/dev/nfs rw nfsroot=172.24.136.169:/ Sep 16 17:14:51 nfs ? Sep 16 17:14:55 hm Sep 16 17:14:56 i am not using on a card .. .my filesystem in in nfs ... Sep 16 17:15:05 no_root_squash set on the server ? Sep 16 17:15:20 ogra, not sure ... where to check ??? Sep 16 17:15:32 /etc/exports usually Sep 16 17:15:36 (on the server) Sep 16 17:16:12 yeah .. set properly ... /new-data/nfs-share *(rw,sync,no_root_squash,subtree_check) Sep 16 17:18:36 hmm, out of ideas here Sep 16 17:18:45 i give up for now .. let me try the whole exercise on MMC tomorrow ... it's already 10:30 here ...i ll go home ...hey gusys ... thanks very much for the help and response .. thanks ogra again ... Sep 16 17:18:55 does mount show you its rw mounted ? Sep 16 17:19:20 yeah, 730pm here ... i'm ready to call it a day too :) Sep 16 17:19:27 let me check ... last experiment for day ... Sep 16 17:20:32 out of luck today .. .:-) Sep 16 17:20:33 ubuntu@ubuntu:~$ mount | grep nfs Sep 16 17:20:33 /dev/root on / type nfs (rw) Sep 16 17:20:33 ubuntu@ubuntu:~$ Sep 16 17:21:19 ogra, c u tomorrow .. thx again ... Sep 16 17:21:30 bye Sep 16 17:21:35 avinashhm: I have seen this issue before. Not sure how I got around it. Will look into further. Sep 16 17:21:56 thats promising ... GrueMaster.... thanks Sep 16 17:22:11 bye .. Sep 16 17:25:49 GrueMaster: daily build looking ok on panda with the new u-boot? Sep 16 17:26:13 hmm, u-boot isn't listed in the preinstalled-netbook manifest; how does that get pulled in? Sep 16 17:27:58 slangasek: Can't tell. Image was borked (according to ogra) so haven't tested it. I'll try moving uboot to an older image. Sep 16 17:28:15 ok Sep 16 17:29:17 slangasek: I believe it's still waiting in the queue Sep 16 17:29:26 but not sure Sep 16 17:29:45 at least todays image didn't get it Sep 16 17:30:12 oh drat, let me see what's going on here Sep 16 17:30:18 the source is in, the binaries are not Sep 16 17:30:30 thanks for calling me on it :) Sep 16 17:30:40 * GrueMaster holds off modifying images. Sep 16 17:30:51 slangasek, we dont install the package Sep 16 17:31:00 debian-cd uses just the binary Sep 16 17:31:08 yeah... still in the build queue, that's why I didn't get any mails about its absence Sep 16 17:31:15 it's useful just to create the first partition Sep 16 17:31:23 I assumed it would get built quickly, but I guess the pre-freeze rush applies Sep 16 17:31:29 yep, i'm waiting for it and will change debian-cd accordingly Sep 16 17:31:35 yeah Sep 16 17:31:42 armel is 24h behind or so Sep 16 17:31:45 NCommander, You still about? Could you bump the kde4libs build to just behind anything landing on the image? Sep 16 17:31:59 ogra_cmpc: oh, this is in debian-cd? alrightythen Sep 16 17:32:07 yeah :) Sep 16 17:34:50 slangasek, oh, you asked about panda ... that will be picked up automatically ... i though about omap3 ... that needs the change to linaro u-boot still Sep 16 17:34:59 (thus debian-cd) Sep 16 17:35:45 ogra_cmpc: picked up automatically from where / how? Sep 16 17:35:58 archive Sep 16 17:36:04 ok Sep 16 17:36:17 pulled into the image as a package, or pulled in somewhere else? Sep 16 17:36:22 it unpacks the latest deb and takes u-boot.bin out Sep 16 17:36:37 then we use the bin on the vfat partition Sep 16 17:36:41 debian-cd still, or is debian-cd only in charge for omap3? Sep 16 17:36:50 for all our images Sep 16 17:36:52 ok Sep 16 17:37:21 for omap3 we simply still use our package ... i'm waiting for the new linaro binary here to make the change Sep 16 17:37:38 (need to change package name and .bin location) Sep 16 17:41:20 ` Sep 16 17:45:49 argh, cold coffee Sep 16 17:47:04 ewww Sep 16 18:02:24 ogra_cmpc: oh, you're going to use u-boot-linaro for omap3 as well this cycle? Sep 16 18:02:34 I thought it was deemed safer not to change Sep 16 18:03:00 it works fine for us of course, but that's no guarantee :) Sep 16 18:04:39 slangasek: I tested it yesterday, for beagle B5, C4 and xM p8, it worked fine Sep 16 18:04:47 the idea is to change to linaro's too Sep 16 18:05:04 it seems the new one provided by linaro fixed a but for xM Sep 16 18:05:17 so it's just better to use it instead of maintaining ours Sep 16 18:05:26 easy to debug and fix in case of new bugs :-) Sep 16 18:07:47 ok : Sep 16 18:07:48 ) Sep 16 18:09:33 GrueMaster: if possible, could you test Dave Martin's fix for bug 587632? Sep 16 18:09:34 Launchpad bug 587632 in libmad (Ubuntu) "Sound very distorted on armel (affects: 2) (heat: 48)" [Undecided,Fix released] https://launchpad.net/bugs/587632 Sep 16 18:09:47 needs a working environment to test the sound Sep 16 18:09:53 Sure. Give me a sec. Sep 16 18:10:07 see comment #9 Sep 16 18:11:19 thanks :-) Sep 16 18:27:49 slangasek, well, they are identical now after jcrigby did a big bughunt the last days, so there is no reason to keep ours Sep 16 18:37:31 rsalveti: libmad isn't part of the image. Will have to download & test. Sep 16 18:41:18 libmad0 is the actual package. Looking at what needs it for testing. Sep 16 18:44:11 GrueMaster: just play a mp3 with xine or any other tool that uses mad Sep 16 18:45:06 then you'll need to build the new version, and test Sep 16 18:45:12 Yea, I know how to test it. Just looking to see what is available to install. Our default uses gstreamer, not xine. Sep 16 18:45:37 oh, ok Sep 16 18:45:52 (and beagle isn't a speed deamon when it comes to...much.) Sep 16 18:46:49 yeah :-( Sep 16 18:48:26 Not getting any output on my XM. No audio of any kind. Sep 16 18:49:56 good test, never saw anyone saying it worked for xM :-) Sep 16 18:50:39 possibly bug++ Sep 16 18:50:51 Well, beagle is in the middle of a lucid->maverick upgrade. Didn't want to disturb it. Sep 16 18:51:30 let me boot mine Sep 16 18:51:32 xM Sep 16 18:54:03 Ok, now I have sound on XM. Alsamixer tweaking. Sep 16 18:55:05 oh, cool Sep 16 18:56:35 Hrm. Nothing passing through pulseaudio yet. Sep 16 22:36:00 GrueMaster, Could you remind me of the bug number for the issue where different beagleboards acted differently for audio? Sep 16 22:36:32 I'll have to dig. Not sure if it got filed. Sep 16 22:37:07 probably something GrueMaster just found Sep 16 22:37:14 Heh, he does that. Sep 16 22:37:21 nobody tests audio, that's something we just noticed this week Sep 16 22:37:59 I think it's similar to the issue with the sdp4430 being discussed, and has a similar solution, but I believe the HW docs for the beagleboards are likely miles easier to access, so we can get the right drivers in place. Sep 16 22:38:26 persia: yep, for sure Sep 16 22:38:52 Part of the problem is I made an assumption (partly based on discussions here) that the XM was just an expanded beagle. Sep 16 22:39:12 I didn't think the wiring for audio would change. Sep 16 22:39:42 probably depends on viewpoint :) I have some suspicion that some of the "Beagleboard clones" have different wiremaps. Sep 16 22:40:08 But from a libc perspective, the entire family is likely identical. Sep 16 22:41:15 http://beagle.s3.amazonaws.com/design/xM-A/BB_xM_SRM_A2_01.pdf Sep 16 22:41:17 for xm Sep 16 22:41:35 http://beagleboard.org/static/BBSRM_latest.pdf Sep 16 22:41:37 for beagle Sep 16 22:48:33 rsalveti, Are there also SRMs for other Beagles (e.g. A2, B3)? Do we know if they are the same or different? Sep 16 22:49:07 persia: one problem is that our xM is not the A2 Sep 16 22:49:20 and *I* don't know much about the difference in hardware Sep 16 22:49:26 besides the memory Sep 16 22:49:47 This is true. We could be hitting a non-issue. Sep 16 22:51:32 Well, it doesn't hurt to try to get as many docs in one place as possible so that folks can write the appropriate drivers. Sep 16 22:52:30 Comparing page 79 of Beagle_SRM_C4.doc to page 85 of Beagle_SRM_XM_A2_01.doc, I'm fairly sure these need different handling: most critically the input ports, but it's more complicated than that alone. Sep 16 22:56:02 * rsalveti out for a bit, dinner Sep 17 00:34:20 Yes! kde4libs has made it to the install stage! Sep 17 00:39:13 Sweet. Sep 17 00:39:34 What are you building on? Sep 17 00:44:20 cushaw Sep 17 00:44:25 https://launchpad.net/ubuntu/+source/kde4libs/4:4.5.1-0ubuntu5/+build/1961555 Sep 17 00:44:55 I think it's a Babbage 3.0, but that's guessing based on IRC traffic. Sep 17 00:46:35 I have Alpha 1 (for 10.10) on my Touchbook but it seems to want a password - is there a default? Sep 17 00:47:19 tommd, How did you install? Sep 17 00:47:39 persia: With dd (https://wiki.ubuntu.com/ARM/OMAPMaverickInstall) Sep 17 00:48:34 tommd: You should be able to boot and it should run oem-install on it's own. Sep 17 00:49:00 Where did you get the image? Sep 17 00:49:12 tommd, An image from http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/ ? Sep 17 00:49:12 GrueMaster: It didn't. Is this a result of me using the AI kernel (not the default Ubuntu/Maverick alpha kernel) Sep 17 00:49:29 no Sep 17 00:49:35 persia: http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/10.10/beta/ Sep 17 00:49:44 oops, yes Sep 17 00:49:47 ;-) Sep 17 00:49:50 OK. Sep 17 00:49:59 Where are you seeing the username prompt? Sep 17 00:50:11 Its an X login prompt Sep 17 00:50:18 If you are using a custom kernel and not the provided initrd, it will fail. Sep 17 00:50:23 It's off now and I'm just replacing the /etc/shadow. Sep 17 00:50:50 Hrm. I wonder if there are odd board-support issues. That oughtn't happen. Sep 17 00:50:53 GrueMaster: Ah. How is the kernel involved in first-boot setup? I would have thought this as something user space + a flag/file. Sep 17 00:51:20 the kernel has to have the right modules enabled, and the initrd needs to have the right code to set stuff up for the first-boot experience. Sep 17 00:51:42 So an arbitrary kernel/initrd ends up connecting to a half-installed system, which doesn't work. Sep 17 00:51:46 There are a couple of scripts that run from initrd, one that resizes the image to fill the Sd card, the other to do some minor tweaks, like triggering oem-install to run. Sep 17 00:52:32 The scripts are part of the jasper-intramfs package. Sep 17 00:52:51 tommd, If you must replace the kernel, try not replacing the initrd (as risky as this is). Sep 17 00:52:58 I have a low confidence in that solution. Sep 17 00:53:48 Next, try booting into some other working environment, and mounting the half-installed system as a chroot, and rebuilding the initramfs inside that, and then extracting that for use. Sep 17 00:54:30 ideally, determine what part is failing, and file some bugs so that the touchbook can be supported in the future (although with that RAM and IO, I fear it will never be well supported) Sep 17 00:54:31 Perhaps I can just go with the custom mlo Sep 17 00:55:37 and the Ubuntu kernel... though I bet that won't have much of the hardware supported. Sep 17 00:57:09 tommd, It's mostly a mainline kernel: does mainline support the touchbook yet? Sep 17 00:57:37 Not really. There are a couple issues (accelerometer, I think... some people made noise about the screen) Sep 17 00:57:41 I'm going to try though Sep 17 00:58:02 Once you complete the rootfs install, you can probably swap kernels without such poor results. Sep 17 00:58:56 ok, I'm giving it a go Sep 17 00:58:59 owch Sep 17 00:59:02 ugly screen. Sep 17 00:59:08 Which one? Sep 17 00:59:39 Its a somewhat long story - basically my NAND is a bit screwed up so I need an mlo/u-boot that doesn't use the NAND settings. Sep 17 00:59:49 The latest AI u-boot is sufficient. Sep 17 01:00:13 And it seems trying to boot the 10.10 kernel with the other mlo/u-boot was a bad idea. Sep 17 01:00:34 Unfortuantely, it seems that uboots are even more sensitive to board changes than kernels. Sep 17 01:00:49 * persia wants EFI+grub last year more Sep 17 01:00:53 yeah, I don't know much about u-boot other than I'm unhappy when I hear it said. Sep 17 01:26:31 is there a big package update going on or something? Sep 17 01:27:04 I can't install gnome desktop, smbclient requires samba-common of a very exacting version but the one in the pool is one version better than the smbclient wants Sep 17 01:27:45 Final Freeze was 8.5 hours ago. Sep 17 01:27:58 Lots of folks pushed to the hard limit for their uploads, so the queues got huge. Sep 17 01:28:10 okay it's just I noticed this: Sep 17 01:28:10 Should be sorted in ~30 hours. Sep 17 01:28:19 [ ] samba-common_3.5.4~dfsg-1ubuntu6_all.deb 16-Sep-2010 07:05 386K Sep 17 01:28:19 [ ] samba-common_3.5.4~dfsg-1ubuntu7_all.deb 16-Sep-2010 21:04 385K Sep 17 01:28:49 .. the build times are less than 14 hours apart... which is freakish Sep 17 01:29:05 30 hours is a long time to wait for the counterpart packages :] Sep 17 01:30:13 https://launchpad.net/ubuntu/maverick/armel/+builds?build_text=&build_state=pending is the list of stuff needing building and the order in which it will build. LP guesses about times, but isn't usually that far off. Sep 17 01:31:13 times? you mean the numbers like 1755, 2505? Sep 17 01:31:17 minutes or seconds? Sep 17 01:35:26 oh the links tell you an hours.. and it says 16 at least Sep 17 01:35:28 oh well Sep 17 01:35:29 back in a bit Sep 17 02:13:24 The TouchBook has lots of custom bootargs in the NAND. Not sure how this compares to the BeagleBoard. Should I wipe out the bootargs and try the Ubuntu kernel or should they remain? Sep 17 02:15:42 tommd, last i checked the default ubuntu kernel only has what mainline support exists for the touchbook.. which isn't much, including no screen... Sep 17 02:18:01 rcn-ee: Thanks, I had thought the screen made mainline. Shows what I know. This explains why I'm not getting any stdout... and I thought the kernel wasn't getting loaded. Sep 17 02:18:51 not really any (easy) way to get it to configure then, huh? No screen with Ubuntu. No initrd (no Ubuntu initrd) with the AI kernel. argh Sep 17 02:19:35 tommd, i just happen to be working on it myself too.. other then the dss2 stuff nothing is too useful for end users yet.. http://bazaar.launchpad.net/~beagleboard-kernel/%2Bjunk/2.6.36-devel/files/head:/patches/touchbook/ (slowlying syncing with touchbook's git tree..) Sep 17 02:19:36 Do you have any working armel environments? Sep 17 02:19:50 If so, you ought be able to reconstrut the Ubuntu initrd in a chroot. Sep 17 02:20:43 persia: I had a working OE setup and I have the TB working in conjunction with the AI kernel (and an older Ubuntu) Sep 17 02:20:47 That's it. Sep 17 02:21:16 The AI Kernel doesn't use an initrd, so I'd need to modify that slightly to begin with. **** ENDING LOGGING AT Fri Sep 17 02:59:57 2010