**** BEGIN LOGGING AT Wed May 18 02:59:57 2011 May 18 06:57:53 good morning May 18 06:58:03 morning! May 18 07:32:08 Video from FOSDEM: http://free-electrons.com/blog/fosdem-2011-videos/ May 18 07:32:37 "ELC and ABS videos are coming very very soon" (с) Michael Opdenacker May 18 09:59:57 is it possible to share a tmp/ dir over multiple users .. i tryed to have the users in a own group but it fails on a do_install where it wanted to change permission in tmp/sysroot/xxx/usr/bin/aclocal stuff May 18 10:01:34 so my only solution right now is to have a oe user which everybody uses who builds the tmp/ , right ? May 18 10:19:44 03Koen Kooi  07org.openembedded.dev * r1796f209d8 10openembedded.git/recipes/sdr/dttsp_svn.bb: May 18 10:19:44 dttsp: fix DEPENDS May 18 10:19:44 Signed-off-by: Koen Kooi May 18 10:19:55 03Koen Kooi  07org.openembedded.dev * r359add34ab 10openembedded.git/recipes/sdr/sdrshell-qt4_svn.bb: May 18 10:19:55 sdrshell-qt4: fix DEPENDS and update SRCREV May 18 10:19:55 Signed-off-by: Koen Kooi May 18 11:11:37 broonie, ping May 18 12:56:56 how do I let bitbake reparse the whole tree? is there any command option to do so? May 18 13:13:36 morphis: touch build/conf/local.conf May 18 13:13:56 foerster: ok, thats some workaround I did too May 18 13:14:05 but there is no other option to do so? May 18 13:14:37 my problem is I added a .bbappend file in my overlay which set the COMPATIBLE_MACHINE for my machine May 18 13:15:11 but then the whole tree didn't get re-evaluated and it still tells me u-boot (the one I did the .bbappend for) can't be found for my machine May 18 13:15:52 I've not used bbappend yet, so I'm not sure (I've been using amend.inc, and haven't switched over) May 18 13:16:31 ok May 18 13:16:39 seems to be a general problem May 18 13:18:02 have you previously used bbappend with success? May 18 13:18:15 no May 18 13:18:20 http://sakrah.homelinux.org/blog/2010/11/bblayers-bbappend/ May 18 13:18:23 it's the first time May 18 13:18:32 seems BBFILES needs *.bbappend set as well May 18 13:18:43 else it won't look for them May 18 13:18:53 it has May 18 13:18:58 that not the problem May 18 13:19:08 if I change my conf/local.conf for example May 18 13:19:29 bitbake tolds me then "NOTE: Out of date cache found, rebuilding..." May 18 13:19:42 after it has done the rebuild of the cache everything works fine May 18 13:20:00 oh. May 18 13:20:05 but when I now add only a new bbappend file in my layer directory it doesn't work May 18 13:20:34 bitbake tells me then that there is no provide for u-boot May 18 13:22:41 will write a mail later about that to the ML May 18 13:22:59 sorry, I can't be of any help (I'm using x86, so no u-boot for me) May 18 13:23:22 ok May 18 13:23:35 foerster: but thank you anyway :) May 18 13:23:40 np May 18 16:06:01 Crofton|work: Pong May 18 16:17:49 I actually tried 32khx on an omap3 (alsa) May 18 16:17:54 and it makes funny noises May 18 16:17:57 not the right one May 18 16:18:06 any guesses? May 18 16:18:08 2.6.38 May 18 16:25:49 No idea - not sure what 32kHz on OMAP is. May 18 16:26:01 You're better off asking lrg if he appears. May 18 16:36:53 k May 18 16:39:51 khem: do you happen to know what of linaro's patches have been reviewed / pulled into oe? May 18 16:45:26 It appears psplash comes up before udev. But it seems like when udev is run after psplash the image isn't being rendered on the screen. I think it's because /dev/fb gets overwritten when udev runs. Is this normal? May 18 16:50:00 Does anyone know how to build modphp (for angstrom)? May 18 17:26:04 03Jeff Lance  07master * rb14cb947a3 10openembedded.git/recipes/ti/ (3 files): (log message trimmed) May 18 17:26:04 matrix_gui: Pull in lmbench,multimedia,EZSDK mods May 18 17:26:04 * Update SRCREV to pull in: May 18 17:27:47 http://pastebin.com/ixPjKzJm - this is a new error for me. what does it mean? May 18 17:33:36 it means exactly what it says May 18 17:33:39 two packages installed the same conffile May 18 17:36:37 hm. in Angstrom 2008.x I had kernel-module-rtc-twl and kernel-module-rtc-core. with the latest from dev that gives the error. using only kernel-module-rtc-twl does not, but I have no idea if it is correct :) May 18 18:00:29 kergoth: you mean reviewed in gcc community ? May 18 18:08:00 re May 18 18:28:31 03Philip Balister  07org.openembedded.dev * ra49bd05f40 10openembedded.git/recipes/libosip2/libosip2_3.1.0.bb: May 18 18:28:32 libosip2 : Upgrade from version 3.1 to 3.5 May 18 18:28:32 Signed-off-by: Philip Balister May 18 18:36:43 kergoth: ping May 18 19:05:05 denix, thanks for attending the yocto technical meeting May 18 19:07:06 03Philip Balister  07org.openembedded.dev * r3fc876f03e 10openembedded.git/recipes/tasks/task-sdk-gnuradio-native.bb: May 18 19:07:06 task-sdk-gnuradio-native.bb : GNU Radio (with UHD) no longer needs sdcc. May 18 19:07:06 Signed-off-by: Philip Balister May 18 19:34:00 03Philip Balister  07org.openembedded.dev * r6cc123b617 10openembedded.git/recipes/ortp/ (ortp_0.13.1.bb ortp_0.16.5.bb): May 18 19:34:00 ortp : Upgrade from version 0.13.1 to 0.16.5. May 18 19:34:00 Signed-off-by: Philip Balister May 18 19:46:36 Crofton|work: not a problem - it's rather logistics than technical, but still good May 18 19:46:58 I'd like to see more variety there May 18 19:49:17 hi kergoth, florian, all May 18 19:49:23 * pb_ returns from the big city May 18 19:49:48 hi pb_ May 18 19:50:03 pb_: Bielefeld? ;) May 18 19:51:28 florian: heh, London in this case May 18 20:40:32 hi woglinde May 18 20:41:07 hi ant May 18 20:41:25 have you tried oe-core? May 18 20:41:56 I see pb_ is fiddling with it now May 18 20:42:01 good sign May 18 20:42:20 sorry no time yet May 18 20:42:21 hello May 18 20:42:26 bluelightning: May 18 20:42:33 I talked with khem about a java layer May 18 20:42:36 woglinde: me neither May 18 20:43:22 woglinde: it looks like khem is 'entangled' with koen atm :) May 18 20:43:53 ? May 18 20:44:19 hard working on angstrom layer/oe-core May 18 20:44:56 khem is working on oe-core and meta-oe May 18 20:46:14 now topic on ML is about IMAGE vars May 18 20:46:20 I agree with pb_ May 18 20:46:52 less vars and good documented, overrides in distro-layer May 18 20:53:56 ant__: pong...? May 18 20:54:33 I think maybe I dropped before you sent the actual msg May 18 21:05:46 pb_, seems like github is cinfusing you and me May 18 21:42:33 bluelightning: hey May 18 21:43:07 still OOM booting opie images... May 18 21:43:28 x11 boots May 18 21:43:31 ant__: have you tracked down the cause? May 18 21:43:35 same udev... May 18 21:43:42 yes, is during udev May 18 21:43:50 I have to log udevadm May 18 21:44:00 so, how can opie have an influence on udev? May 18 21:44:17 I checked about specific opie-initscripts but did not find any May 18 21:44:29 conflicting in rcS May 18 21:44:53 but the issue is repeatable: on 32M and 64M devices May 18 21:45:21 fwiw we (Zaurus)don't use udev device cache May 18 21:45:43 because root is still ro (why, still to see how) May 18 21:46:12 ant use udev-168 May 18 21:46:19 or busybox-mdev May 18 21:46:36 in fact I'm thinking about dropping udev May 18 21:46:57 problem with qt-2.3.10 is fonts May 18 21:47:47 woglinde_: I see udev 168 is now preferred May 18 21:47:54 I'll test right now May 18 21:48:18 I had problems with touchscreen on the simpad May 18 21:48:22 and opie May 18 21:48:41 there is a strange issue with tslib May 18 21:48:54 with 2.6.37 headers May 18 21:49:16 okay May 18 21:49:25 there is patch May 18 21:49:43 at least in zaurus-devel ML May 18 21:50:08 jay7 pointer? May 18 21:50:15 but it compiles May 18 21:50:20 well, I have ASSUME_PROVIDED += "linux-libc-headers-native" May 18 21:50:33 because of race-issues on Gentoo May 18 21:50:45 and with 2.6.36 all is ok May 18 21:51:03 woglinde_: http://lists.linuxtogo.org/pipermail/zaurus-devel/2011-May/000507.html May 18 21:51:17 2.6.37 has slightly newer EV_VERSION May 18 21:51:28 and there is a silly check May 18 21:53:23 ah tslib_git is needed May 18 22:18:44 bluelightning: with latest udev 168 opie boots May 18 22:18:56 ant__: good news :D May 18 22:18:58 still stalls after starting udev... May 18 22:19:02 hold breath May 18 22:19:10 and it welcomes :) May 18 22:19:41 argh May 18 22:19:49 loop in ts calibration May 18 22:20:25 still at welcome screen :/ May 18 22:20:29 someone said the ts driver currently had issues on zaurus-devel May 18 22:20:43 well, I see crosshairs and can click those May 18 22:20:43 something about it being too noisy...? May 18 22:20:56 butr calibration never ends May 18 22:21:19 it behaves 'normally' I'd say May 18 22:21:29 maybe cannot write the settings May 18 22:21:30 if calibration never ends then the ts values it's getting back are outside the expected range May 18 22:21:36 ah May 18 22:21:48 I think that's what it means anyway May 18 22:21:54 we have old kernel patch to reduce noise May 18 22:21:55 let me boot x11 now May 18 22:22:00 to compare May 18 22:27:14 on c7x0, no changes: pre-calibrated pointercal values are a bit wrong on my devices and recalibration is impossible May 18 22:35:01 NOTE: package klibc-1.5.22-r0.0: task do_compile: Succeeded May 18 22:35:03 :) May 18 22:40:12 with oe-core? May 18 22:40:19 no, oe-dev May 18 22:40:34 * khem just came back from cheering Tour of California riders May 18 22:40:54 but http://paste.debian.net/117317/ May 18 22:41:23 something must have changed in the sources May 18 22:43:30 ant__: do you know some status info about meta-handhelds? May 18 22:43:51 unfortunately not May 18 22:44:15 Iwe are now chasing the lack-of-modules in rootfs May 18 22:44:38 klibc with oe-core is blocking my reflash on spitz :) May 18 22:44:50 ouch May 18 22:45:08 JaMa: klibc ? May 18 22:45:16 is it committed? May 18 22:46:07 anyway. I added rm klcc/klcc in do_install...how can it be installed? May 18 22:46:07 May 18 22:46:25 JaMa: bluelightning is master of meta-hh :) May 18 22:46:33 ant__: I had it in my meta-shr for month or two, now I have it in meta-smartphone/meta-zaurus which I plan to drop as soon as there is meta-handhelds May 18 22:46:49 great May 18 22:46:59 sorry guys, have not got around to finishing it May 18 22:47:02 Jay7: I know, but haven't seen him online and he doesn't respond to ping now.. May 18 22:47:07 ah, here he is :) May 18 22:47:43 JaMa: any hint about udev/root read-only on Z ? May 18 22:48:04 klibc should go into oe-core as I understand May 18 22:48:06 rootfs read-only? May 18 22:48:12 now udev is after fstab, maybe adding rw could help.. May 18 22:48:22 what I have still currently is what was ripped out of oe-core, which when compared to what's in oe-dev isn't that good May 18 22:48:34 JaMa: yes, somehow we cannot use udev-cache May 18 22:48:41 we need /tmp in tmfs May 18 22:48:47 *tmpfs May 18 22:48:58 then it should work May 18 22:49:18 here it works, I haven't seen any RO issues May 18 22:49:27 do we have tmpfs in /dev/shm already? May 18 22:49:28 but I'm scared by the missing modules in th eend-images... May 18 22:49:31 so sorry no idea what went wrong May 18 22:49:34 at that point May 18 22:50:39 i.e. we lack all snd modules for Z...is not a question of modutils/init_tools or udev...modules are not in the rootfs... May 18 22:50:46 ant__: I prefer explicit module names in MACHINE_EXTRA_RRECOMMENDS anyways May 18 22:51:03 ant__: as upgrading 100 modules just because I bump SRCREV takes too long.. May 18 22:51:08 how could it work with 2.6.26? May 18 22:51:50 we didn't change machine conf May 18 22:52:17 something is fiultering modules out, like for headers May 18 22:52:25 ant__: was sound definitely working with 2.6.26? May 18 22:53:00 well, at least lib/modules was more populated iirc May 18 22:53:21 sound was working fine on my akita May 18 22:53:29 fwiw we moved from linux-rp.inc to linux.inc May 18 22:53:29 I've used it as player long time :) May 18 22:53:43 I'm not positive but I think it was not working on my spitz May 18 22:53:47 tosa was not fine May 18 22:53:59 not sure about other models May 18 22:54:06 corgi was ok, spitz (3200) too May 18 22:54:14 lumag is slowly working on tosa's sound May 18 22:54:16 ant__: http://paste.pocoo.org/show/391510/ May 18 22:54:54 look for /lib/modules/2.6.39-rc7/kernel/sound/soc/pxa/snd-soc-spitz.ko May 18 22:55:02 is not in today's rootfss May 18 22:55:19 only in modules-.tgx May 18 22:55:20 bluelightning: iirc alsa-states were wrong and something was muted for me, but then it worked May 18 22:55:23 I ran into an issue building qt4-native_4.7.3. I get QA staging issues. Anyone who might have a clue on that? May 18 22:55:24 eh May 18 22:55:29 .tgz May 18 22:55:36 JaMa: that was probably it, I didn't dig into it too far May 18 22:55:40 ant__: as I said here it works right as you can see in that pastebin May 18 22:55:50 which distro? May 18 22:55:53 SHR May 18 22:56:16 customized x11-gpe-image May 18 22:56:23 minimal and Angstrom did fail here... May 18 22:56:41 but point is that kernel-modules is installed and it rdepends on all kernel-module-* so it pulls them to rootfs May 18 22:57:35 angstrom May 18 22:57:40 JaMa: I'm in the rootfs.. May 18 22:57:54 i /snd I have only core May 18 22:58:16 hmm May 18 22:58:18 and do you have kernel-modules installed? May 18 22:59:31 no :) May 18 22:59:53 let's grep for it May 18 23:00:54 ant__: ./x11-gpe-image.bb: kernel-modules \ May 18 23:01:02 that's why it's in mine May 18 23:01:38 you can add it to MACHINE_EXTRA_RRECOMMENDS to force it in every image but again: May 18 23:01:39 00:47:46 < JaMa> ant__: I prefer explicit module names in MACHINE_EXTRA_RRECOMMENDS anyways May 18 23:02:03 x11-gpe *only* May 18 23:02:16 I build console- opie- and x11-image May 18 23:02:31 ok, let's add it to zaurus.inc May 18 23:02:40 :( May 18 23:02:53 going to bed, gn8 May 18 23:02:56 sometimes x11-gpe, seldom May 18 23:02:59 ;) May 18 23:03:29 good night May 18 23:03:44 Jay7: mystery almost solved May 18 23:04:00 nice May 18 23:04:16 strange how it was working before, though May 18 23:05:08 and again please dont put *all* modules to MACHINE_EXTRA_RRECOMMENDS just because it's shorter then list of kernel-modules which are really usefull May 18 23:06:43 JaMa|Zzzz: worth discussion on zaurus-devel May 18 23:07:46 I suspect some MACHINE_KERNEL_PR did broke May 18 23:08:12 (some patch about M_K_PR) May 18 23:12:36 MACHINE_KERNEL_PR is not used for Z at all afaik May 18 23:13:01 JaMa|Zzzz: defconfig is customized, almost all modules are useful May 18 23:13:14 many are monolitic May 18 23:13:19 (fs) May 18 23:13:46 btw...you wrote the modularized-defconfig :) May 18 23:13:49 ant__: but defconfig should be "wide" so everybody can find module for his weird usb gadget May 18 23:14:00 ant__: in feeds May 18 23:14:16 ant__: and on rootfs only subset of "usually" usefull modules May 18 23:14:52 ok, apart usb case, the rest should not be compiled at all May 18 23:15:19 different cf wifi cards? May 18 23:15:36 well, how can you decide which ones? May 18 23:15:47 how do you know? May 18 23:15:59 different serial modules like Stanislav Brabec asked? May 18 23:16:17 I'd compile and deploy all those 'pluggable' May 18 23:16:19 if nobody puts kernel-modules to MACHINE_EXTRA_RRECOMMENDS then we can build *all* May 18 23:17:25 because building more modules is not so big overhead as upgrading them on many targets every now and then May 18 23:18:08 building more modules on fast build server... / upgrading on slow target with slow internet connection May 18 23:18:43 3,5 hours to sleep, gtg :) May 18 23:18:47 omg May 18 23:18:54 go away ;) May 18 23:20:59 well.. good idea May 18 23:21:05 * Jay7 gone sleep too May 18 23:25:27 yes, let's finish tomorrow May 18 23:25:32 be May 19 01:20:21 gn **** ENDING LOGGING AT Thu May 19 02:59:58 2011