**** BEGIN LOGGING AT Wed Nov 29 02:59:58 2006 **** BEGIN LOGGING AT Wed Nov 29 04:12:42 2006 Nov 29 05:42:12 hi, all! Nov 29 05:42:36 what is actual purpose of tag= attribute for git:// method? Nov 29 05:51:44 I've been long meaning to understand how tags work in git. Nov 29 06:03:41 It's not the same as in other version control systems? Nov 29 06:11:54 Well, the concept is the same, I hope. I was confused about the application of said tags. Nov 29 06:50:46 well... as I understand in GIT, commit id is tag. am I right? Nov 29 06:51:49 is the bug tracker server down? Nov 29 07:01:31 *yawn* Nov 29 07:01:32 morning guys Nov 29 07:01:56 gm, all! Nov 29 07:02:24 O_Neil: read the channel topic? Nov 29 07:06:31 not yet :) Nov 29 07:10:30 can not login Nov 29 07:10:44 it trying to connect to bugs.openembedded.org Nov 29 07:11:07 yeah Nov 29 07:11:16 this is why you should add the new IP to your /etc/hosts file Nov 29 07:13:31 yes, just did it Nov 29 07:31:47 morning all ! Nov 29 07:44:15 mickey|earlybird: who are you, and what have you done to mickeyl ?!?! Nov 29 07:44:19 good morning all Nov 29 07:44:51 hehehe Nov 29 07:44:53 morning koen Nov 29 07:44:53 what is LEAD_SONAME? Nov 29 07:45:21 and how could I use it in case of multiple .so in a package? Nov 29 07:45:55 koen: the explanation is as simple as that: yesterday i only slept 3 hours and so i was so tired that i went to bed yesterday @ 22:30. got me waking up @ 07:00 today :) Nov 29 07:46:21 slapin_nb: use the most important library name in the package as LEAD_SONAME or package fine granular Nov 29 07:47:40 thanks! Nov 29 07:51:30 mickey|earlybird: I got the python troubles with gnome-python sorted out, it just needs a more recent gnome-vfs(-dbus) to compile against Nov 29 08:04:57 yo Nov 29 08:05:12 koen: ok, cool. i've reviewed your patch, it looks ok Nov 29 08:05:39 mickey|earlybird: yeah, the trouble was in the python-pygtk2 recipe Nov 29 08:05:47 koen, mickey|earlybird: a mention of OpenEmbedded: http://www.linuxdevices.com/news/NS3535328630.html Nov 29 08:07:35 rwhitby: cool, thansk! Nov 29 08:07:35 thanks, even Nov 29 08:07:36 rwhitby: cool Nov 29 08:08:35 mickey|earlybird: maybe the OpenMoko debian port will surpass it one day ;-) Nov 29 08:10:08 s/OpenMoko/Neo1973/ of course ... Nov 29 08:10:57 * rwhitby wonders why people would want to run Debian on a phone, but believes that there will still be people who ask for it. Nov 29 08:11:56 bugs.openembedded is still down and its not possible to login also from bugs.treke.net as it redirects to bus.openembeeded. Any ideas ? Nov 29 08:12:16 rwhitby: people are running gentoo on their PDAs Nov 29 08:12:28 Ifaistos: use your /etc/hosts Nov 29 08:13:13 koen: If I check in that erase block size patch tonight, will those unstable build images be automatically updated somehow? Nov 29 08:13:33 no, but I'll rebuild them :) Nov 29 08:13:43 * rwhitby has a hacking night planned Nov 29 08:13:47 03hrw 07org.oe.dev * radc5941a... 10/ (3 files in 3 dirs): libpcap: bump version of library due to Mickeyl comment. Thx Nov 29 08:32:29 morning all Nov 29 08:33:50 hey RP Nov 29 08:34:34 RP, hi! Nov 29 08:36:09 RP: related to the package arch troubles yesterday: http://svn.o-hand.com/view/poky/trunk/meta/classes/rootfs_deb.bbclass?rev=965&view=diff&r1=965&r2=964&p1=trunk/meta/classes/rootfs_deb.bbclass&p2=/trunk/meta/classes/rootfs_deb.bbclass Nov 29 08:37:09 koen: No, that really is TARGET_ARCH Nov 29 08:37:26 koen: There is deep voodoo going on there ;-) Nov 29 08:39:18 morning Nov 29 08:39:24 hey XorA Nov 29 08:41:21 so what OE feature can I destroy today Nov 29 08:41:52 morning RP Nov 29 08:41:54 yo XorA Nov 29 08:42:51 hey mickey|earlybird Nov 29 08:48:54 /nick hrw|work Nov 29 08:48:56 moring Nov 29 08:50:32 hey hrw|work Nov 29 08:50:56 **yawn** Nov 29 08:51:22 morning hrw|work Nov 29 08:51:41 * hrw|work will ignore fact that each kernel module has package Nov 29 08:52:19 progear.conf ^+kernel-modules Nov 29 08:54:42 rwhitby: I like that sentence: "In the same time it takes to make a baby (nine months), the NSLU2 has taken the ARM architecture from 7th place to 3rd place in Debian architecture popularity." ;D Nov 29 08:54:54 :-) Nov 29 08:56:35 I love how they include all kinds links on the article except where they got the news from Nov 29 08:56:50 oh, I sent it to them directly. Nov 29 08:57:22 oh, ok. It just looked like they would have taken a direct quote from some mailing list Nov 29 08:57:26 well, I sent the bit they quoted - they wrote the first bit. Nov 29 08:57:59 yeah, I basically sent them the same message I sent to the nslu2-linux mailing list. Nov 29 09:03:03 I bet there are more devices running linux on arm than on x86 Nov 29 09:03:51 and mips Nov 29 09:19:44 anyone knowing offhand how to chmod w/ svn? Nov 29 09:19:55 chmod && svn ci isn't detected Nov 29 09:20:28 mickey|earlybird: ha? Nov 29 09:20:32 propedit? Nov 29 09:20:33 svn propedit Nov 29 09:20:37 aha, thanks Nov 29 09:20:41 morning zecke btw. Nov 29 09:21:02 hey zecke Nov 29 09:21:12 * zecke is da chicken Nov 29 09:27:00 hmmm Nov 29 09:27:02 propedit? Nov 29 09:27:27 i can't find the file permissions in the list of available properties Nov 29 09:29:20 good morning Nov 29 09:29:45 ah Nov 29 09:29:47 svn:executable Nov 29 09:29:49 hi florian_kc Nov 29 09:38:01 03xora 07org.oe.dev * r2a4863b0... 10/ (3 files in 3 dirs): tslib/tslib/smdk2440 : add configuration for smdk2440 machine Nov 29 09:53:04 HopsNBarley_: is your commit access working yet? if not, please resend the key to mickeyl@linuxtogo.org and I'll take care of that immediately Nov 29 09:54:48 mickey|earlybird: I hope you didn't add that key Nov 29 09:54:56 mickey|earlybird: it has the wrong ID Nov 29 09:55:12 ah. did you have a chance to request him to change the iD? Nov 29 09:55:32 HopsNBarley_: please generate a key with @openembedded.org as ID Nov 29 09:55:40 I didn't Nov 29 10:04:07 03xora 07org.oe.dev * rcdf35bf5... 10/ (1 packages/fontconfig/fontconfig_2.4.1.bb): Nov 29 10:04:07 fontconfig_2.4.1.bb : add the latest stable fontconfig. Nov 29 10:04:07 Also took opportunity to improve its cross compiling by removing Nov 29 10:04:07 ${HOST_ARCH} stuff from native bits of the build. Nov 29 10:08:59 03rpurdie 07org.oe.dev * r18d45ae1... 10/ (1 packages/maemo-mapper/maemo-mapper-nohildon_svn.bb): maemo-mapper-nohildon: Fix segfault, remap zoom keys, package desktop file Nov 29 10:12:23 http://www.makezine.com/blog/archive/2006/11/the_open_source_1.html Nov 29 10:15:29 fsck.. my devel machine has broken cron or something.. Nov 29 10:15:47 ade|desk: see if that fontconfig makes you happier Nov 29 10:15:58 [hrw@misc1 logs]$ crontab -l Nov 29 10:15:58 20 2 * * * /bin/bash /home/hrw/cron/fetch.sh >/home/hrw/cron/logs/fetch-`date +%Y%m%d`.log Nov 29 10:16:07 this should work - right? Nov 29 10:16:25 at 02:20 run 'command' and put log into file. Nov 29 10:17:01 hrw|work: Id add a 2>&1 in there as well to catch errors Nov 29 10:17:32 XorA: me too, but first need to get why it does not work... Nov 29 10:17:42 and admin will wake up in 6h ;( Nov 29 10:20:08 hrw|work: debian system? Nov 29 10:21:32 rhel Nov 29 10:24:05 XorA: Does the smdk2440 have 2.4 support for a corgi touchscreen? Nov 29 10:25:11 RP: I havent looked at 2.4 support at all Nov 29 10:25:51 XorA: I can near enough guarantee it doesn't use a corgi driver as per the config you just checked into tslib ;-) Nov 29 10:26:14 RP: crappit, I switched that file at last minute and forgot to fix it Nov 29 10:26:22 RP: cheers for that Nov 29 10:27:06 XorA: np, I just want to see those entries *die*, not spread :) Nov 29 10:28:12 corgi_ts is primary implementation now in mainline just like corgi_bl for backlight? Nov 29 10:29:56 hrw|work: Yes and no. Its slightly bad code as it doesn't use the spi subsystem and the standard ads7856(or whatever) driver Nov 29 10:30:29 hrw|work: On the other hand it does work :) (and yes, its in mainline) Nov 29 10:30:39 03xora 07org.oe.dev * rf7da9634... 10/ (1 packages/tslib/tslib/smdk2440/tslib.sh): Nov 29 10:30:39 tslib/tslib/smdk2440/tslib.sh : remove 2.4 stuff that gives away where I Nov 29 10:30:39 stole the file from. Nov 29 10:38:24 03rpurdie 07org.oe.dev * r566766a2... 10/ (3 files in 3 dirs): maemo-mapper: Add missing patch Nov 29 10:40:16 03koen 07org.oe.dev * r2c547e87... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom 2007.1: only install psplash formachines with a 'screen' Nov 29 10:49:20 03koen 07org.oe.dev * refd4f14c... 10/ (26 files in 3 dirs): lots of machines: add 'screen' to MACHINE_FEATURES Nov 29 10:52:16 morning fooks Nov 29 10:52:49 hey likewise Nov 29 10:52:55 hi likewise Nov 29 10:54:30 I am hitting an ipkg problem. All of the armv5te packages cannot be found, ipkgarchs has armv5teb but not armv5te. The .ipk files have armv5te suffixes. Where should I fix this? Nov 29 10:55:18 likewise: It sounds like armv5te was missing from PACKAGE_ARCHS (which used to be IPKG_ARCHS) Nov 29 10:55:39 likewise: If you have files outside OE, make sure you renamed Nov 29 10:55:45 RP: which .bb(class) is responsible for setting that variable? Nov 29 10:56:02 likewise: The machine's conf file normally Nov 29 10:56:25 RP: In general, it is hard for me to find out which variables should be set by which .bb(class), and also, which .bb(class) needs the variables. Nov 29 10:56:37 RP: Thanks. Nov 29 10:57:00 likewise: own machine config? Nov 29 10:57:26 likewise: http://ewi546.ewi.utwente.nl/tmp/ixp4xx.diff is my latest attempt Nov 29 10:57:37 hrw|work: yes, ixp4xx alike/derivative Nov 29 10:58:24 likewise: so you got hit by same as rwhitby probably Nov 29 10:59:23 hrw|work: I should probably manually read the full diff whenever I update my OE checkout, to learn what has changed, and why. Nov 29 10:59:32 sure Nov 29 11:04:37 ~curse monotone Nov 29 11:04:40 May the fleas of a thousand camels infest your most sensitive regions, monotone ! Nov 29 11:11:38 03rpurdie 07org.oe.dev * rbdae1531... 10/ (131 files in 72 dirs): Introduce STAGING_BINDIR_CROSS and STAGING_BINDIR_NATIVE as discussed on the mailing list. There should be no functionality changes yet. Nov 29 11:18:26 re Nov 29 11:19:30 03rpurdie 07org.oe.dev * r0bb5f399... 10/ (5 files in 3 dirs): update-rc.d: Drop old 0.6 version. Set PACKAGE_ARCH = all. Update to reference STAGING_BINDIR_NATIVE Nov 29 11:19:51 RP: dpkg in OE came from poky? Nov 29 11:20:34 RP: dpkg-native depend on update-alternative-native which is provided by update-alternative-dpkg-native which depend on dpkg-native Nov 29 11:20:45 hrw|work: It came from OE -> poky, was updated/altered in poky and synced back (several times) Nov 29 11:21:20 hrw|work: Which bitbake version said that? Nov 29 11:22:02 trunk Nov 29 11:22:37 ERROR: Task /home/hrw/devel/oe/org.openembedded/packages/update-alternatives/update-alternatives-dpkg-native_1.13.22.bb (do_package) has circular dependency on /home/hrw/devel/oe/org.openembedded/packages/dpkg/dpkg-native_1.13.22.bb (do_package) Nov 29 11:23:20 solution was PREF_PROV u-a-n=u-a-cworth-native Nov 29 11:24:16 hrw|work: poky is using that cworth-native too Nov 29 11:25:26 what do you think about adding depends u-a-cworth-native into dpkg-native to break chain? Nov 29 11:27:00 hrw|work: That just works around the problem :-/ Nov 29 11:27:42 hrw|work: hmm, I do re-use the ixp4xx machine, and it has PACKAGE_ARCH set correctly. Any ideas? org.openembedded.dev/conf/machine/include/ixp4xx.conf:5:PACKAGE_ARCH = "armv5te" Nov 29 11:27:45 hrw|work: We removed most of the circular dependency problems with the split packaging but not all :-( Nov 29 11:29:09 likewise: What made it into ipkg's architecture file in your image? (sorry, I can't remember the path or be 100% sure of the filename) Nov 29 11:29:18 RP: kernel coding style question, the samsung i2s unit needs its registers ioremapped into virtual and also needs some clockinfo set that the board driver supplies, should this be in a function that is s3c24x_i2s_init or _probe or nothing like that at all Nov 29 11:30:02 XorA: I'd say _probe Nov 29 11:30:32 XorA: Well, it depends what the context of _init is :-/ Nov 29 11:31:08 RP: well whicher it is would be called from _init in smdk2440 machine driver Nov 29 11:31:19 which ever Nov 29 11:31:57 XorA: Is _probe called once but _init called whenever the audio device is initialized? Nov 29 11:32:36 RP: asoc has neither in its call chain currentl Nov 29 11:33:59 XorA: This i2s code is something like pxa2xx-i2s.c? Nov 29 11:34:11 RP: yes Nov 29 11:35:35 XorA: I'd guess struct snd_soc_cpu_dai needs an optional init function pointer adding Nov 29 11:36:10 XorA: maybe to .ops Nov 29 11:36:10 RP: that makes sense Nov 29 11:36:58 XorA: The idea being startup can get called more than once and is supposed to be quick. Init would be once off and do things like ioremap Nov 29 11:37:32 XorA: On the PXA, we're lucky as the ioremapping is done by the pxa core code Nov 29 11:38:10 I'd check Liam doesn't have an optional init function somewhere in there... Nov 29 11:38:17 RP: I know, it just seems to be a choice Ben Dooks made to do the ioremap inside audio code Nov 29 11:38:46 XorA: That's fair enough, you just want to do it once though :) Nov 29 11:40:21 RP: actually just noticed there is a probe Nov 29 11:40:34 RP: in cpu_dai Nov 29 11:43:07 likewise: show me TARGET_ARCH var Nov 29 11:43:32 RP: cheers for the help Nov 29 11:45:15 XorA: np Nov 29 11:51:35 CIA-4: are you constipated again? Nov 29 11:55:44 Is there a good reason not to include the fix from bug 1138 and set the preferred version for gstreamer? Nov 29 11:57:47 03rwhitby 07org.oe.dev * rb8a61703... 10/ (4 files in 4 dirs): ludeos: removed archaic and deprecated distro (SlugOS supercedes and fully replaces LudeOS) Nov 29 11:57:54 03rwhitby 07org.oe.dev * r7b9bf78a... 10/ (1 conf/machine/include/ixp4xx.conf): ixp4xx.conf: Change default endiannes from be to le Nov 29 11:57:58 03rwhitby 07org.oe.dev * ra0522554... 10/ (1 packages/slugimage/slugimage.bb): slugimage: Updated to latest svn version, which reserves space for the Apex environment. Nov 29 11:59:34 03rpurdie 07org.oe.dev * r36827e88... 10/ (24 files in 4 dirs): linux-openzaurus: Rename to linux-rp Nov 29 11:59:49 RP: finally :) Nov 29 12:00:14 hrw|work: :) Nov 29 12:01:15 hehe Nov 29 12:01:19 making it harder for sash to steal stuff? Nov 29 12:01:56 koen: I wouldn't expect it to make any difference to that... Nov 29 12:02:18 it doesn't have 'z' in its name, so it's unfit for a zaurus Nov 29 12:05:13 03koen 07org.oe.dev * r7537d2b1... 10/ (1 packages/gpesyncd/gpesyncd_svn.bb): gpesyncd: adjust DEPENDS to match svn Nov 29 12:05:49 koen: pdaX is dead Nov 29 12:06:10 hrw|work: Officially or just for those in the know? Nov 29 12:06:11 hrw|work: maybe it just smells that way Nov 29 12:06:36 Is there a good reason not to set CACHE = "${PATH2OE}" where ${PATH2OE} is the root OE directory? Will this put the cache file in ${PATH2OE}? Will this create problems? RP, will this avoid recreation of the cache that we talked about yesterday? Nov 29 12:07:31 Laibsch: It will avoid the recreation of the cache, it shouldn't cause any problems Nov 29 12:07:54 RP: Thanks. Then I will make that change. Nov 29 12:08:32 Laibsch: Was this to change DISTRO? Nov 29 12:08:32 Laibsch: sash does not appear on #cacko, no signs of any activity from him anywhere Nov 29 12:08:32 Laibsch: and there is no other devs in pdax Nov 29 12:08:51 hrw|work: and maybe cacko will be built with OE as well if nicolasfr succeeds Nov 29 12:08:58 koen: so, yes - it smell like that Nov 29 12:09:13 ~curse monotone for not working Nov 29 12:09:23 May the fleas of a thousand camels infest your most sensitive regions, monotone for not working ! Nov 29 12:09:51 RP: Yes, I switch distro sometimes and currently that always recreates the cache. Nov 29 12:10:06 hrw|work: one-man show is never sustainable. Nov 29 12:14:30 fscking shit.. today is a day when I start to hate monotone more Nov 29 12:17:15 hrw|work: what are you cursing about exactly within monotone? Nov 29 12:17:37 likewise: mtn: warning: orphaned node conflict on node 18,372, dead parent 18,331, name files Nov 29 12:17:55 hrw|work: OK, I get the picture :-/ Nov 29 12:18:17 likewise: example of very good information Nov 29 12:18:18 Laibsch: The problem is it will need to recreate the cache when you change DISTRO as you touch machine.conf Nov 29 12:18:33 Laibsch: Also, distro's change a lot of the config data so there is a valid reason for reparsing Nov 29 12:19:45 * RP wonders about CACHE = "${PATH2OE}/${DISTRO}" and using "DISTRO=foobar bitbake something" Nov 29 12:19:57 put ${DISTRO} in ${TMPDIR} Nov 29 12:28:15 koen I have done that already. I was trying to save the time reparsing. I got the impression yesterday this was possible. Nov 29 12:29:13 Laibsch: The trick is to have two cache files which you get with two TMPDIRs but you also must not change the timestamp of local.conf when switching between them Nov 29 12:29:38 Laibsch: You can set variables in the environment to do that but you need to be very careful when doing that Nov 29 12:29:51 Could I just have two local.confs as well? And dynamically switch between them? Nov 29 12:30:13 Laibsch: Just have two build directories Nov 29 12:30:15 set BBPATH Nov 29 12:30:36 * Laibsch will study the possibilities. Nov 29 12:32:02 someone know spanish? I need few curses ;D Nov 29 12:32:17 Laibsch: my website has infos for it Nov 29 12:32:22 hrw|work: putanella Nov 29 12:32:45 * koen tries to find his "curse in 27 languages" book again Nov 29 12:33:03 koen: I had such one with 10 languages but lost ;( Nov 29 12:35:28 RP: could the re-indentation of base.bbclass have had an effect on packages/glibc/glibc-package.bbclass? Nov 29 12:41:00 hrw|work: So I guess you just have build/$DISTRO/conf/local.conf instead of build/conf/local.conf, right? How do you keep the conf files in sync? Manually? Do you care to "sync" the value in them? Nov 29 12:41:30 s/value/values/ Nov 29 12:42:03 Laibsch: BBPATH is your friend Nov 29 12:42:20 conf/need-to-be-insync.conf Nov 29 12:43:52 Laibsch: I have devel/build/BUILDNAME/conf/local.conf where BUILDNAME has nothing common with DISTRO/MACHINE usually Nov 29 12:44:14 zecke: Ah, right. I was thinking about if it is possible to INCLUDE things. But of course, that is not even necessary. Great! Nov 29 12:44:19 Laibsch: and each builddir is separated from others. Nov 29 12:44:35 zecke: have you seen http://landley.net/code/firmware/ ? Nov 29 12:45:36 koen: also, the toybox stuff? Nov 29 12:45:45 koen: hmm doesn't this destroy timesys busines model? Nov 29 12:45:59 zecke: how's your chicken-out option proceeding? Nov 29 12:46:04 zecke: could the re-indentation of base.bbclass have had an effect on packages/glibc/glibc-package.bbclass? Nov 29 12:46:52 likewise: badly, looks I will definately chicken-out. Two tests, little time to prepare :( Nov 29 12:47:08 koen: hmm, sure Nov 29 12:47:20 zecke: we will have to catch up at FOSDEM then. Good luck with your tests! Nov 29 12:47:40 likewise: they are not hard, but force me to prepare... :( Nov 29 12:47:56 zecke: that would explain why binary locales are broken right now Nov 29 12:48:39 zecke: I know, it's good to do a PREFERRED_PRIORITY="100" on tests :-). Nov 29 12:50:19 * koen mov r0,r0 Nov 29 12:50:45 koen: I'm taking a look now... Nov 29 12:51:03 koen: oh putting a bbclass there is not smart... Nov 29 12:51:27 zecke: don't blame me, it has been there for ages Nov 29 12:51:37 zecke: either kergoth or pb put it there Nov 29 13:09:16 koen: oh yeah Nov 29 13:09:24 koen: the anonfunc needs to be spacified Nov 29 13:09:44 this is the reason not to put a bbclass into this dir :) Nov 29 13:10:00 feel free to move it to classes/ Nov 29 13:10:10 should I spacify? Nov 29 13:10:25 if that fixes it, yes Nov 29 13:10:41 took me a week to come with the idea spaces might have broken it Nov 29 13:11:05 koen: what about the other 150 skipped files? Nov 29 13:11:30 zecke: I moved qmake (or moc) to nonworking/foo Nov 29 13:11:39 so I guess it skips all opie stuff Nov 29 13:11:44 what about the other 149 skipped files? Nov 29 13:12:53 * koen dunnos Nov 29 13:22:51 03freyther 07org.oe.dev * rd28ae51c... 10/ (1 packages/glibc/glibc-package.bbclass): Nov 29 13:22:51 packages/glibc/glibc-packages.bbclass: Oh no, tabs :) Nov 29 13:22:51 Use spaces in anonfunc, needed for BitBake trunk Nov 29 13:22:51 to speedup parsing a bit Nov 29 13:23:50 koen: Ultimately, we should adjust all python functions to use spaces... Nov 29 13:24:14 yes, or at least warn about a mix... Nov 29 13:26:40 what is the OZ channel called? Nov 29 13:27:03 lindenle_: #openzaurus Nov 29 13:32:27 * koen NOTE: package glibc-2.5-r3: task do_package: startedERROR: Error in compiling: ERROR: The lines resulting into this error were:ERROR: 6:SyntaxError:' import glob, stat, errno, re'ERROR: Printing the environment of the function Nov 29 13:32:46 koen: progress :) Nov 29 13:32:49 zecke: bitbake doesn't like your change Nov 29 13:34:11 there is no such line in glibc-package.bbclass Nov 29 13:35:36 zecke: http://www.openembedded.org/~koen/glibc.png Nov 29 13:35:46 package.bbclass Nov 29 13:39:54 03freyther 07org.oe.dev * rc9a5d44b... 10/ (1 packages/glibc/glibc-package.bbclass): Nov 29 13:39:54 packages/glibc/glibc-package.bbclass: package.bbclass is not yet spacified... Nov 29 13:39:54 Oh use tabs for this method... Nov 29 13:44:32 zecke: it works now, thanks Nov 29 13:45:40 can anyone tell me why bitbake fails with this in the error log: Nov 29 13:45:41 http://pastebin.ca/261483 Nov 29 13:46:59 TheCompWiz: hehe, well in contrast to mother's BitBake can't be blamed for everything Nov 29 13:47:17 I'm not blaming bitbake... but rather the script bitbake is using... Nov 29 13:47:45 Okay BitBake failed because the script it was asked to execute failed Nov 29 13:47:53 so why did the script it executed failed? Nov 29 13:48:07 because the script failed with "cannot stat `etc/hotplug/{*.{agent,rc},hotplug.functions}': No such file or directory" Nov 29 13:48:17 that I don't know. it's complaining about missing files.. but I don't know why those files would be missing. Nov 29 13:48:34 isn't hotplug dead anyway? Nov 29 13:48:49 I wouldn't know. I'm just trying to get an image together. Nov 29 13:48:59 * TheCompWiz is new to the OE scene... Nov 29 13:49:13 TheCompWiz: take a look at hotplug, check if you find the files manually Nov 29 13:49:27 take a look at *what* in hotplug? Nov 29 13:50:42 TheCompWiz: no idea, how you configured your OpenEmbedded Nov 29 13:51:09 machine = h2200 ... distro = openzaurus-unstable Nov 29 13:51:35 ~lart me for typing reboot in wrong window Nov 29 13:51:35 * ibot_ stuffs xora into a shiny new tin can and vacuum seals it for typing reboot in wrong window Nov 29 13:52:03 TheCompWiz: hehe, e.g. why did you decide to put it in /stuff... you won't get any cookbook answers from me Nov 29 13:52:21 TheCompWiz: BitBake executed a task on htoplug, this task failed, it mentions the path and the script it executed Nov 29 13:52:39 TheCompWiz: this is a good candidate to look into, check why the script failed, what it did Nov 29 13:52:49 where it looked for the files, and where the files are... Nov 29 13:52:50 well... it started off in my ~ folder... but that's another question. Nov 29 13:53:24 it dosn't really make a difference "where" the files are... just as long as you specify the correct paths in the local.conf & bbpath Nov 29 13:53:55 the odd part is... 90% of it already compiled & made packages & such with no problems. Nov 29 13:53:59 actually it does Nov 29 13:54:10 bitbake doesn't like ' in paths Nov 29 13:54:11 o.O Nov 29 13:54:26 there is no ' in a path... o.O Nov 29 13:54:34 over here there is :( Nov 29 13:54:52 over where? Nov 29 13:54:55 wha? Nov 29 13:55:05 TheCompWiz: okay, you are about to fail the turing test, last chanche Nov 29 13:55:13 TheCompWiz: you understood who created the log file? Nov 29 13:55:20 yes. Nov 29 13:55:33 TheCompWiz: so who produced the log file? Nov 29 13:55:52 in this case... bitbake generated the log based on the contents of the .bb Nov 29 13:55:58 right/wrong? Nov 29 13:56:13 TheCompWiz: which package (packaname, packageversion) Nov 29 13:56:33 I'm guessing... but unsure.... "linux-hotplug-20040920" Nov 29 13:57:16 TheCompWiz: actually it will tell you what it build Nov 29 13:57:56 hold on a sec... lemme find it. Nov 29 13:59:00 TheCompWiz: BitBake executed a task, this task failed, which task I don't know as your paste is not complete Nov 29 13:59:13 TheCompWiz: it complained about missing files, the first thing I would do is find them... Nov 29 13:59:20 anyway use the source, you have the tools Nov 29 13:59:40 that's the odd part. 90% of the other packages had already compiled... before this error came into play. Nov 29 13:59:56 if my setup was incorrect... wouldn't those packages have failed as well? Nov 29 14:00:20 TheCompWiz: invalid question Nov 29 14:00:45 TheCompWiz: you think 90% were built correctly but the question is where these packages impacted, heavily impacted by your setup Nov 29 14:02:21 define "impacted"? Nov 29 14:02:53 if they had invalid paths & such as well while compiling... they would also generate errors on compile (i.e. missing files & such) Nov 29 14:02:56 TheCompWiz: no, ask yourself if machine, distro have a different degree of influence on fortune or binutils/gcc Nov 29 14:03:05 or kernel Nov 29 14:03:10 or low level userspace stuff Nov 29 14:03:26 TheCompWiz: but again, what about looking and trying to understand the source of the issue Nov 29 14:03:34 File Not Found -> Check where the file is Nov 29 14:03:51 morning Nov 29 14:03:57 well... in truith ... the file is non-existant... and I have no clue what file it *should* be looking for. Nov 29 14:04:45 in the for statement... says it's looking in "etc/hotplug/{*.{agent,rc},hotplug.functions}" ... but I have no clue what that path is relative to... Nov 29 14:04:59 so in worst case you do find /? Nov 29 14:05:10 what? Nov 29 14:05:24 in best case you do find ${WORKDIR} (where WORKDIR is the configured workdir...) Nov 29 14:05:29 TheCompWiz: man find, or info find Nov 29 14:06:05 find finds nothing. grats. Nov 29 14:06:26 so again... the question is... why would it be trying to "install" non-existant files? Nov 29 14:06:50 define 'it' Nov 29 14:07:19 ... bitbake... Nov 29 14:07:19 BitBake does not install files Nov 29 14:07:46 grrr... the script that bitbake implements :P happy? Nov 29 14:07:46 BitBake implements no scripts Nov 29 14:08:05 TheCompWiz: Check the BitBake manual Nov 29 14:08:05 * TheCompWiz is getting very frustrated here... and you're just being an ass. Nov 29 14:08:19 TheCompWiz: BitBake is a task executor Nov 29 14:08:23 TheCompWiz: It executes tasks, and again I don't know which task you executed Nov 29 14:08:39 initially.... it was bitbake gpe-image .... Nov 29 14:08:54 TheCompWiz: look in tmp/work/${ARCH}-linux/hostplug*/temp/run_task for the scripts run and the variables set Nov 29 14:08:56 and I know gpe-image has several "other tasks" that get performed in the process. Nov 29 14:09:08 TheCompWiz: no these are not tasks Nov 29 14:09:19 do_compile, clean, fetch, patch, install, package are tasks Nov 29 14:09:38 BitBake executed a task on your *hotplug*, again I don't know which task, this task failed Nov 29 14:10:05 take a look at the package it tries to build, in your case hotplug and figure out why it wants to install the files, where the files should be... Nov 29 14:11:24 lemme rephrase that.... bitbake performs SEVERAL do_compile/do_install etc... tasks without error before reaching the package linux-hotplug-20040920-r8 ... where it fails. Nov 29 14:11:42 TheCompWiz: Hint: bitbake != openembedded. Any bug is not in bitbake but in the OE metadata Nov 29 14:11:49 TheCompWiz: which task fails? Nov 29 14:12:06 do_install Nov 29 14:12:16 ERROR: function do_install failed Nov 29 14:12:24 TheCompWiz: wow, okay this is unlikely to be a bitbake issue Nov 29 14:12:36 hence why I ask the question here. Nov 29 14:12:44 TheCompWiz: so OpenEmbedded's metadata, or linux-hotplug itself are wrong Nov 29 14:12:59 TheCompWiz: I wonder why h2200 needs an ancient hotplug but that is not my busines Nov 29 14:13:18 TheCompWiz: so what about going to the directory where linux-hotplug gets baked/build? Nov 29 14:13:30 TheCompWiz: the path is inside your log Nov 29 14:13:39 and from there... doing what? Nov 29 14:14:10 isn't udev all the rage now for hotplugging stuf? Nov 29 14:14:23 TheCompWiz: start understanding hotplug's issue Nov 29 14:14:37 I think so. but I'm not the one maintaing the "machine" for h2200 or h6000 ... both have the same problem. Nov 29 14:14:40 koen: yes, I wonder why a 2.6er kernel based machine uses hotplug... Nov 29 14:15:12 TheCompWiz: find the script/Makefile OF hotplug that fails Nov 29 14:15:15 * RP suspects its being pulled in through task-bootstrap Nov 29 14:15:36 TheCompWiz: once again, check if you have these kind of scripts somewhere else... Nov 29 14:18:48 mm...kay... found them in "tmp/work/armv5te-linux/linux-hotplug-20040920-r8/hotplug-2004_09_20/etc/hotplug" Nov 29 14:21:47 is there an *easy* way to find out what the working directory is while that package is being built? Nov 29 14:21:59 it is deterministic Nov 29 14:22:50 $WORKDIR/${PACKAGE_ARCH}-${TARGET_OS}/${PN}-${PV}-${PR} as configured in bitbake.conf of OpenEmbedded Nov 29 14:23:28 TheCompWiz: Here a guess: This looks like shellcode is failing Nov 29 14:23:45 Do you use edgy/unstable with /bin/sh pointing to dash and this construct is a bashism? Nov 29 14:28:38 zecke: Seems like I managed to royally screw up while trying to split my local.conf. I moved MACHINE and DISTRO into separate files and left the rest in a common file (see http://rafb.net/paste/results/bhqjjF55.html). I set BBPATH to first include the specific stuff, then common and oe.org.dev last. Now bitbake bails out horribly and I cannot fix it. Can you take a look at my paste? Nov 29 14:29:51 Laibsch: Did you call each file local.conf? Nov 29 14:30:04 yes. Not a good idea? Nov 29 14:30:08 Laibsch: same as RP's question Nov 29 14:30:14 Laibsch: no, it will pick the first file Nov 29 14:30:18 Laibsch: no, it will only read one of them. Nov 29 14:30:22 I see. I will rename them. Nov 29 14:30:25 Thanks. Nov 29 14:30:26 Laibsch: in you local.conf you can write "require commonfile" Nov 29 14:31:38 This is some sort of include? Nov 29 14:31:48 site.conf? Nov 29 14:33:09 What is that now? Too many options. I am struggling. Nov 29 14:33:26 no site.conf found via locate. Nov 29 14:33:32 What is it? Nov 29 14:34:18 site.conf sounds good :) Nov 29 14:41:13 Will having naming the files in each build dir as distro.conf, machine.conf and common.conf cause problems? Nov 29 14:41:46 I found that more intuitive than hrw's auto.conf, local.conf and site.conf for what the files are supposed to contain. Nov 29 14:42:22 auto.conf, local.conf and site.conf are the files OE wants Nov 29 14:43:15 Well, in the standard install according to to the getting started guide on the wiki, these three files are not split up last time I looked. Nov 29 14:43:51 I just wanted to make sure these file names are not again used elsewhere in the OE tree. Nov 29 14:52:02 I think everything is back in working order now. Nov 29 15:04:50 sorry zecke... boss needed something.. anyhow... I'm not runny edgy... but still a ubuntu distro. anyhow... bash is pointed to dash... Nov 29 15:04:50 cbrake: 're u here? :> I tested with the generic distro now, same problem Nov 29 15:05:16 so, I guess yes... it is a bashism ;0 Nov 29 15:05:21 zecke: We really need locking in bitbake, *soon*. Two packages say dpkg and dpkg-native with the same SRC_URI = fun :-/ Nov 29 15:07:06 TheCompWiz: Which is the file with the problem and have you a patch to fix it? :) Nov 29 15:08:07 no patch here... I'm just trying to get started with OE atm. Nov 29 15:08:28 TheCompWiz: That answers half my question... Nov 29 15:09:01 Laibsch: remember about BBPATH Nov 29 15:09:03 as for the file system... of my machine? ... um... ext3 Nov 29 15:09:11 the world needs a bash->sh compiler :-) Nov 29 15:10:02 hrw|work: Thank you for the reminder. I set them via a script that creates the necessary environment. Nov 29 15:10:25 It all seems to be working. I am recreating the cache for sharprom on one machine and for openzaurus on another machine. Nov 29 15:11:01 RP ... would the machine's FS really have a problem with how the shell is interpreting commands? Nov 29 15:11:23 Jin^eLD: hmm, can you pastebin source to your test app, and I'll try it here? Nov 29 15:11:44 cbrake: sure, moment Nov 29 15:12:24 RP: locking? as in mutex? Nov 29 15:12:32 RP: After the TARGET_ARCH changes for ixp4xx, should my packages end in *armv5te.ipk or armv6teb.ipk (in my case, big-endian). Nov 29 15:12:56 cbrake: http://pastebin.ca/261507 Nov 29 15:14:15 TheCompWiz: do you have an idea which script might have the bashism included? Nov 29 15:14:56 I'm trying to figure that out now. but it's something to do with the linux-hotplug-20040920-r8 Nov 29 15:15:29 org.openembedded.dev/packages/linux-hotplug/linux-hotplug_20040920.bb Nov 29 15:15:46 there is do_install() {} which will be executed by bitbake Nov 29 15:16:08 yep... but dosn't look like that is the right one. Nov 29 15:16:09 I'm pretty bad with bash/shell, but we run "make install" Nov 29 15:16:25 linux-hotplug 20040920 is buildable Nov 29 15:16:25 TheCompWiz: and sh ${WORKDIR}/fix-net.agent" Nov 29 15:16:37 hrw|work: with /bin/sh pointing to dash? Nov 29 15:16:59 zecke: no - bash. but I can check with dash too Nov 29 15:17:22 but this will take some time Nov 29 15:20:04 verifying ... Nov 29 15:20:05 calculated crc32 = 0xA7E02DCA Nov 29 15:20:05 desired crc32 = 0xA7E02DCA Nov 29 15:20:05 formatting ... done. Nov 29 15:20:05 boot> boot wince Nov 29 15:20:09 then... full stop Nov 29 15:20:22 nothing on display, nothing on console... machine seems completely dead Nov 29 15:20:23 harrydg: #handhelds.org or #oe? Nov 29 15:20:30 ? Nov 29 15:20:37 /ignore harrydg Nov 29 15:20:37 is wince opensource nowadays? Nov 29 15:20:44 nope Nov 29 15:20:44 koen: no it's not Nov 29 15:20:53 but i cant seem to get a linux image booting Nov 29 15:20:56 koen: it is opensource, but not Free Software Nov 29 15:20:59 so i thought, lets try going back to win Nov 29 15:21:14 16:36 < TheCompWiz> harrydg... this chan leans more towards kernel development.... try #oe Nov 29 15:21:20 ==> i joined #oe Nov 29 15:21:38 cbrake: it must be realated to the way my system was built; I do not think the fault is in the application Nov 29 15:21:54 cbrake: but since I took the generic distro, the only thing I could screw up is the machine Nov 29 15:22:18 Jin^eLD: works find on my glibc 2.3.90 system Nov 29 15:22:29 can anyone tell me how i can get my jornada 565 to work again? Nov 29 15:22:46 harrydg: only jornada maintainer probably Nov 29 15:22:47 i've got a working image with 0.7.2 (opie) Nov 29 15:22:52 harrydg... you were asking for an image... not how to fix your PDA Nov 29 15:23:00 but... it fucks my touchpanel Nov 29 15:23:07 Jin^eLD: I'll try a glibc 2.5 system later today -- it is not set up right now. Nov 29 15:23:09 so i can really not use it for anything Nov 29 15:23:11 cbrake: this one has glibc 2.5 I think Nov 29 15:23:15 cbrake: ok thanks Nov 29 15:23:25 and i can't seem to get wince working on it anymore either, so i'm kinda... getting desperate! Nov 29 15:23:37 what do i have to do to get it working? Nov 29 15:23:46 harrydg... well... you shouldn't have messed with your PDA if you were worried about breaking it. Nov 29 15:24:49 it fails with dash. Nov 29 15:25:05 TheCompWiz: i wasn't Nov 29 15:25:06 * cbrake wonders how long it takes to migrate a mtn db? Nov 29 15:25:19 well... now it's b0rke... and you're desperate... Nov 29 15:25:23 i just fixed the touchpanel, had to solder a new connection Nov 29 15:25:36 so it's only working again since friday Nov 29 15:25:40 cbrake: approx the same time as a pull from scratch Nov 29 15:25:55 i'm desperate in the sense that... i dont want to throw it away... i want it to work again Nov 29 15:25:59 is anyone here running some glibc 2.5 distro on arm, and can confirm that pthreads are for sure working? I tried building the generic distro but had exactly the same problem as with my custom one Nov 29 15:25:59 someday ;) Nov 29 15:26:25 Jin^eLD: confirm how? Nov 29 15:26:35 koen: well, launch some app that uses pthread_create Nov 29 15:26:44 a very simple test would be that: http://pastebin.ca/261507 Nov 29 15:27:06 pastebin.ca isn't reachable from here Nov 29 15:27:48 http://www.deadlock.dhs.org/jin/threadtest.c Nov 29 15:28:14 Jin^eLD: can you point me to an armv4 binary? Nov 29 15:28:28 Jin^eLD: my smdk2440 is currently booted with generic Nov 29 15:29:06 XorA: one moment Nov 29 15:29:17 zecke: yes, some kind of mutex os even I guess stamps in the DL_DIR so threads don't trample on each other Nov 29 15:30:16 likewise: I can't comment on what your setup would be doing as I don't know how you've got things setup. Mixing big and little endian sounds bad though Nov 29 15:30:40 XorA: http://www.deadlock.dhs.org/jin/thread_arm_generic.gz Nov 29 15:30:48 this is the binary of the code I pasted above Nov 29 15:31:00 RP: I want to mimic nslu2, except for their recent endianess default to 'le' Nov 29 15:31:04 Jin^eLD: http://rafb.net/paste/results/crKoVx83.html Nov 29 15:31:19 koen: thanks.. Nov 29 15:31:37 glibc 2.5/EABI Nov 29 15:31:39 likewise: armvt5e is little endian Nov 29 15:31:42 that leads me to the next question then... how do I figure out why it is failing here? Nov 29 15:31:51 strace, gdb? Nov 29 15:31:58 I tried it with old ABI, with EABI Nov 29 15:32:13 koen: strace says out of memory when doing mmap2, if I remember correctly, let me retry Nov 29 15:32:55 RP: The ixp4xx runs 'be' as well Nov 29 15:33:03 Jin^eLD: Creating thread 0 Nov 29 15:33:05 etc etc Nov 29 15:33:10 doh... Nov 29 15:33:17 I am faced with a mystery :( Nov 29 15:34:34 cbrake: fetch snapshot for monotone 0.30 Nov 29 15:34:38 ~OE.mtn Nov 29 15:34:40 from memory, oe.mtn is snapshot of OpenEmbedded metadata available on http://www.openembedded.org/snapshots/ as files for monotone <0.30 and for >=0.30 Nov 29 15:37:19 http://rafb.net/paste/results/KfdjNN81.html Nov 29 15:37:43 that's the strace... could you take a look, is there anything suspicious there? Nov 29 15:38:05 I was not able to find any clues Nov 29 15:38:07 Is it possible to specify an either-or dependency? Meaning that either libxine or gstreamer need to be present? Haven't seen such a thing in OE yet. Nov 29 15:38:41 Jin^eLD: seems normal , only the mmap2 that fails and probably crashes it... Nov 29 15:38:48 (stating the obvious prolly) Nov 29 15:38:50 :) Nov 29 15:39:11 if my binary runs on Xora's machine, then something must be wrong with my distro build Nov 29 15:39:31 the question is - how to figure that out Nov 29 15:39:49 hrw|work: yup, much faster than migrate Nov 29 15:40:39 that's my machine config http://rafb.net/paste/results/m6ENpq83.html Nov 29 15:40:42 Jin^eLD: i don't have a lot of time now, but i'm willing to try investigate it... (i've got some mm problems myself, so... ;)) Nov 29 15:40:43 I use the generic distro Nov 29 15:41:21 some weirdness when running java on a pax machine inside vserver stuff etc... ugly shit ;)) Nov 29 15:41:27 heh Nov 29 15:42:20 * harrydg off now, cya'll later Nov 29 15:42:25 cu Nov 29 15:42:40 Jin^eLD: may want to one of the other machine configs. Maybe there is something in tune-arm926ejs.conf that is messing things up? Nov 29 15:43:03 TheCompWiz: mtn pull/update and rebuild linux-hotplug Nov 29 15:43:20 hmm, let me check.. but I did not change anything in the mtune file myself Nov 29 15:43:59 03hrw 07org.oe.dev * r8592f826... 10/ (3 files in 3 dirs): linux-hotplug 20040920: fix bashisms Nov 29 15:44:05 looks ok, just a two liner Nov 29 15:44:42 it came with OE and I guess it is used with other machine configs as well Nov 29 15:49:10 hmm seem to have lost the avian Nov 29 15:54:12 * XorA|gprs cant wait until neo1973 release and a real irc client Nov 29 15:54:49 XorA|gprs: do you have powered hub for connecting external usb keyboard? Nov 29 15:55:27 hrw|work: yes Nov 29 15:56:01 in here even the monitors are hubs Nov 29 15:57:14 03koen 07org.oe.dev * rcdb340b5... 10/ (1 packages/gpesyncd/gpesyncd_svn.bb): gpesyncd: add some more DEPENDS Nov 29 15:57:55 XorA|gprs: monitor hubs are usually 1.1 only ;( Nov 29 15:58:32 hrw|work: you can type at 480Mbps? Nov 29 15:59:03 XorA, no but perhaps he can accellelate his mouse at this speeds :) Nov 29 15:59:26 03koen 07org.oe.dev * rceaca23b... 10/ (1 packages/pango/pango_1.14.8.bb): pango: update 1.14.8 Nov 29 15:59:40 must be bad, i think every it guy has gone into server room Nov 29 16:01:19 XorA|gprs: I can scan at this rate Nov 29 16:02:06 terrier: can use my key board without a hub, is very cool Nov 29 16:05:25 XorA|gprs: I wonder why X works from a fresh install on my ipaqs, by not on your Zs Nov 29 16:06:35 no idea Nov 29 16:08:20 I would switch over to Angstrom if i could get that and loadkeys to work Nov 29 16:09:01 you could alter the initscript to run loadkeys twice Nov 29 16:09:30 koen: thats what i did, but the x things kills me Nov 29 16:12:08 does it just crash, or lock up? Nov 29 16:13:27 koen: lockup with blank screen no keyboard control Nov 29 16:13:44 koen: need to make serial cable Nov 29 16:15:18 that sucks Nov 29 16:16:04 hmm idea, ill put me wireless config in, boot off sd and i should be able to log in Nov 29 16:16:19 later Nov 29 16:16:31 if macine isnt totally locked Nov 29 16:17:57 wonder if the fact ipaqs work point to subtle kernel bug Nov 29 16:19:38 yeah Nov 29 16:19:50 maybe installing X ontop of bootstrap would help Nov 29 16:20:36 wince working now... tomorrow: trying linux again :) Nov 29 16:21:03 koen: i shall try and debug some more, maybe I fire up eabi on smdk2440 in background to work Nov 29 16:21:08 is there sourcecode of the bootloaders? Nov 29 16:26:00 anyway no working network, im off home Nov 29 16:28:23 where is the function "oe_runmake" defined? Nov 29 16:28:57 TheCompWiz: grep classes/ Nov 29 16:37:38 03koen 07org.oe.dev * rd61842e6... 10/ (1 conf/distro/include/sane-srcdates.inc): sane src-dates: add gpesyncd, since r8867 depends on non-released libraries Nov 29 16:37:47 spiffy... adding "SHELL := /bin/bash" to the Makefile for hotplug get's 'er going ;) Nov 29 16:38:15 TheCompWiz: why is that needed? Nov 29 16:38:25 TheCompWiz: hotplug already has been fixed Nov 29 16:38:44 o.O Nov 29 16:38:47 k. fine. Nov 29 16:38:50 didn't know it was fixed. Nov 29 16:39:22 it was needed... becuase someone used bash specific parsing... and my machine is using dash.... which dosn't like it. Nov 29 16:39:27 [17:07] hrw|work: TheCompWiz: mtn pull/update and rebuild linux-hotplug Nov 29 16:39:39 yeah... in the process. Nov 29 16:41:38 03florian 07org.oe.dev * r7e2e60ac... 10/ (1 removal.txt): removal.txt: Add nsqld and gpe-mileage Nov 29 16:41:44 koen... what times is it on your machine? Nov 29 16:41:46 03florian 07org.oe.dev * r57477039... 10/ (1 MAINTAINERS): MAINTAINERS: Add myself Nov 29 16:42:12 nm Nov 29 16:42:26 18:06 Nov 29 16:42:28 TheCompWiz: it was 1 hour ago Nov 29 16:42:33 yup Nov 29 16:42:39 while I was afk ... figures ;) Nov 29 16:42:55 TheCompWiz: switch to irc client which can do hilighting Nov 29 16:43:10 TheCompWiz: or configure the one you are using Nov 29 16:43:40 highlighting... won't help when you're AFK. Nov 29 16:44:13 but thanks. Nov 29 16:44:31 TheCompWiz: really? irssi mark window as 'window with hilight' Nov 29 16:44:48 TheCompWiz: and I can easily query when and who hilighted me Nov 29 16:49:24 Hi! Nov 29 16:49:44 koen, is there Angstrom release going to be out soon? Nov 29 16:50:13 define soon Nov 29 16:50:58 the distro config file has a date-code Nov 29 16:51:00 Nov 29 16:51:04 psokolovsky: test images are available Nov 29 16:51:34 koen, just wonder if my deps on adding new machines are too an eyesore ;-) but just saw deps where removed and added again ;-) Nov 29 16:51:59 they aren't showstoppers for the release Nov 29 16:52:04 hrw|work, well, let users test them ;-) I'm trying to work on getting more images for more machines ;-) Nov 29 16:52:08 and hence have to place in the metabug Nov 29 16:52:42 psokolovsky: each machine in angstrom need to have machine mentors Nov 29 16:53:06 koen, yeah, I guess. that's why I ask. if release is close, tehy should go. otehrwise, why not let them be attached, to have a one big plan seen by showing dependencies? Nov 29 16:53:14 hrw|work, I understand ;-) Nov 29 16:53:53 hrw|work, so I'm going to be a nominated mentor for *snapshots*, if you don't mind. would this work? ;-) Nov 29 16:54:20 is there a wiki || faq for angstrom? Nov 29 16:54:47 ibot: sigh Nov 29 16:54:48 *sigh* Nov 29 16:54:50 ibot: angstrom Nov 29 16:54:52 methinks angstrom is [angstrom] a unit of length (0.1 nm), or a Swedish physicist (Anders Jonas Angstrom), or a new distribution for handhelds utilizing EABI Nov 29 16:56:05 raduga, http://www.angstrom-distribution.org/ - links to wiki, etc Nov 29 16:56:11 bye Nov 29 16:56:12 thanks Nov 29 16:56:59 ibot: angstrom is also http://www.angstrom-distribution.org/ Nov 29 16:57:01 raduga: okay Nov 29 16:57:09 ~botsnack Nov 29 16:57:09 raduga: aw, gee Nov 29 17:09:35 03koen 07org.oe.dev * r50cf42d4... 10/ (1 packages/glibc/glibc_2.4.bb packages/glibc/glibc_2.5.bb): glibc 2.5: bump PR due to class changes this afternoon Nov 29 17:36:15 I did some further debugging regarding my pthread problem; it seems that allocate_stack() function in glibc is failing Nov 29 17:37:45 and that actually tries do to an mmap, so this just confirmed the strace output that mmap failed with out of mem Nov 29 17:41:39 ibot, seen kergoth Nov 29 17:41:57 kergoth is currently on #wowi-lounge (3d 11h 52m 30s) #oe (3d 11h 52m 30s) #wowace (3d 11h 52m 30s) #elinux (3d 11h 52m 30s). Has said a total of 131 messages. Is idling for 16h 52m 42s, last said: 'hehe'. Nov 29 17:47:01 ibot_: test Nov 29 17:47:57 funny thing, when I compile the threadtest program statically, with a different glibc version it works with the generic image; so my glibc is somehow broken I guess? Nov 29 17:52:47 koen, do you have a minute? I tracked the pthread problem a little further Nov 29 17:59:30 doh.. Nov 29 17:59:37 any glibc hackers around? :> Nov 29 18:08:26 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Nov 29 18:30:19 psokolovsky_: ping Nov 29 18:31:06 koen, pong Nov 29 18:31:29 see pm Nov 29 19:02:25 03rwhitby 07org.oe.dev * rdc872c26... 10/ (1 packages/slugimage/slugimage.bb): slugimage: updated to latest svn to fix missing microcode problem Nov 29 19:02:30 03koen 07org.oe.dev * rad36a7c0... 10/ (1 packages/tasks/task-proper-tools.bb): task-propertools: add a task to get proper tools instead of the dumbed down busybox versions Nov 29 19:04:31 koen: s/proper/full Nov 29 19:04:45 no, proper Nov 29 19:05:12 busybox doesn't have a 'half full' ls :) Nov 29 19:05:29 feel free to add tools I missed: http://www.openembedded.org/repo/org.openembedded.dev/packages/tasks/task-proper-tools.bb Nov 29 19:05:30 hi all! there's an OESF forums admin/moderator in? Nov 29 19:05:41 ~oesf Nov 29 19:05:49 oesf is probably rumour has it, oesf is newer name of ZUG - it provides forum for Zaurus/Ipaq/Simpad/etc users (see zug). It has nothing in common with OpenEmbedded (except for the first two letters), see http://www.oesf.org/ Nov 29 19:05:51 can we alias oesf to morons? Nov 29 19:06:08 sorry Nov 29 19:06:20 koen: less, vim, fdisk, e2fsutils-* Nov 29 19:06:32 hi, im looking for help on ipkg OZ 3.5.4 Nov 29 19:06:52 we need to discuss with offroadgeek about oe/oesf Nov 29 19:06:57 Aid: #openzaurus Nov 29 19:07:00 tnx Nov 29 19:08:40 hrw|work: fdisk is in util-linux Nov 29 19:09:01 ok Nov 29 19:11:49 03koen 07org.oe.dev * r293732d5... 10/ (1 packages/tasks/task-proper-tools.bb): task-proper-tools: add more utils Nov 29 19:26:02 03koen 07org.oe.dev * r7b5ec7d9... 10/ (1 packages/tasks/task-proper-tools.bb): task-proper-tools: don't emit bogus packages Nov 29 19:38:27 bugtracker down again :-/ Nov 29 19:42:13 file a bug ;) Nov 29 20:19:28 Bluetooth link keys question... /var/lib is usually mounted on tmpfs for handhelds, which is where link keys are stored. So, on a reboot, all link keys are erased. Should /var/lib/bluetooth -> /usr/lib/bluetooth or something like that? Any thoughts from anyone? Nov 29 20:20:36 Could someone add this line PREFERRED_VERSION_libpng = "1.2.8" to sharprom-compatible.conf. This will fix bug 1624 (I have pb accessing bugtracker for some days...). Thanks! Nov 29 20:28:23 even with 64.136.190.225 bugs.openembedded.org in my host file I often can not access the OE bug tracker, is there another site for OE bug tracking? Nov 29 20:30:04 HEY Nov 29 20:30:09 HEY LRG Nov 29 20:30:16 * lrg turns caps lock off Nov 29 20:30:17 HEY HEY Nov 29 20:30:19 noise in .in? Nov 29 20:30:33 koen: noisy airport ;) Nov 29 20:31:05 koen: I think Graeme and I will also stay in astrid hotel Nov 29 20:31:14 lrg: ah, cool Nov 29 20:31:34 koen: will somebody bring AP ? Nov 29 20:31:38 that solves the "dunno who gets to go" question as well Nov 29 20:32:03 lrg: access point? Nov 29 20:32:34 koen: yep, iirc it was about 7eu per day for access Nov 29 20:32:53 something like that Nov 29 20:33:01 and it didn't work half the time Nov 29 20:33:12 koen: we also need to take power extension (6 way if possible) Nov 29 20:33:26 good point Nov 29 20:33:28 koen: i think the lecture rooms only had 2 points Nov 29 20:33:37 4 Nov 29 20:33:38 cu Nov 29 20:33:40 2 on each side Nov 29 20:33:42 hey lrg Nov 29 20:33:47 hey pH5 Nov 29 20:33:50 hrw|work: cu Nov 29 20:33:54 hey koen Nov 29 20:34:00 need to reserve room too Nov 29 20:34:00 hey pH5 Nov 29 20:34:14 but first need to push one customer to get money for it ;D Nov 29 20:34:23 time to go home - 12h @work Nov 29 20:35:24 pH5: I've had a look at the Monahans data sheet and it descibes the audio clocking a lot better :) I'm making changes now. Nov 29 20:37:17 lrg: great :) Nov 29 20:37:54 lrg: remember I couldn't get the volume down? Nov 29 20:37:58 pH5: fwiw they have made it better for pxa320, but it's still not exact :( Nov 29 20:38:08 lrg: the wince driver doesn't seem to touch the volume registers at all Nov 29 20:38:30 pH5: is there an on board amp ? Nov 29 20:39:44 I don't know. It gets stranger by the day. Still, does this explain why the uda1380 internal mixer doesn't have any effect *at all* (including the mutes)? Nov 29 20:40:01 bugs.treke.net also does not work all the time either.. I'm trying it now and I can't get it to load. Nov 29 20:40:16 pH5: maybe Nov 29 20:52:20 what is up with bugs.treke.net? Why is it so unstable? Any chance it will be working again soon? Nov 29 20:56:12 ask ggilbert(_) Nov 29 21:16:17 Does anyone know how to sign up for the forums at oesf.org? I've tried everything but I still cannot post, even in the "Guest/Validation Forum". Nov 29 21:22:03 re Nov 29 21:22:16 phil284: why you feel that we are able to help you? Nov 29 21:22:59 .oO(it's like love - just a feeling) Nov 29 21:23:36 hrw: the oesf page directed me to go to this channel (among others) Nov 29 21:23:46 phil284: really? where? Nov 29 21:24:01 hrw: www.oesf.org Nov 29 21:24:29 phil284: oesf has nothing common with us Nov 29 21:24:42 phil284: except two first letters Nov 29 21:26:01 hrw: I know, but this seems to be the only contact point, unfortunately. I would contact them any other way if I knew how. Nov 29 21:26:21 phil284: here you will not find ANY oesf admins Nov 29 21:27:07 we need to talk with offroadgeek about that 'oe forums' name... Nov 29 21:27:16 hrw: I suspected so, but again, they provided no other choice for me. Nov 29 21:27:58 hrw: oesf indeed mentions #oe.... Nov 29 21:28:00 phil284: if you want help with oesf then click on 'my assistant' link on forum and then 'admins' and complain to them Nov 29 21:28:06 likewise: FSCKERS Nov 29 21:28:28 hrw: MKFSCKERS FAT Nov 29 21:28:58 hrw: you are still annoyed by monotone I can tell :-) Nov 29 21:29:06 as zaurus user I like ZUG forums. as OE developer I dislike OESF Nov 29 21:29:20 hrw: It says "You are not allowed to use the messenger feature on this board" Nov 29 21:29:22 likewise: I managed to fix that problem which I had Nov 29 21:29:49 phil284: then welcome in club of new users and wait until someone turn you into normal user. Nov 29 21:29:50 hrw: good. you fsck'd inode 16383 and then told his parent to fsck off? Nov 29 21:30:05 phil284: but do not expect help here in this Nov 29 21:30:21 mithro: hi there Nov 29 21:30:38 hi Tim Nov 29 21:31:11 likewise: nearly. monotone get_roster|less, found node, did revert and then update Nov 29 21:31:53 hrw: They haven't verified any registrations since Oct. 30. Sorry if I bothered you; I just tried what they suggested. Please feel free to ignore me; I just wanted to see if they were around here. Again, sorry. Nov 29 21:33:48 phil284: try #zaurus maybe Nov 29 21:34:07 #oe is for discussion related to the Open Embedded system mostly Nov 29 21:34:49 raduga: Thanks Nov 29 21:35:34 we need to add informations to oesf wiki that #oe is *NOT* support channel for them Nov 29 21:44:44 btw - someone know how much does 4GB CF card costs? or 4GB microdrive Nov 29 21:45:08 .oO(how is it called in gb/usa? radioshack?) Nov 29 21:46:08 Not sure about CF, but 4GB SD is about $100US, CF should be roughly the same. Nov 29 22:06:21 hrw: check dealram.com Nov 29 22:06:39 will Nov 29 22:07:07 hrw: right now the lowest price for 4G CF is $72.96 shipped (USD) Nov 29 22:07:25 nice price Nov 29 22:07:57 in .us electronic is cheaper then in europe Nov 29 22:11:18 good nite every1 Nov 29 22:11:35 you pay nearly the same in EUR Nov 29 22:12:17 sadly: 1 EUR = 1.31 USD Nov 29 22:22:50 cu Nov 29 22:30:06 03rpurdie * r708 10bitbake/bin/bitbake: Nov 29 22:30:06 bin/bitbake: Update generateDotGraph to use taskData/runQueue correcting the Nov 29 22:30:06 dependency graphs. depends.dot now includes both rdepends and depends, Nov 29 22:30:06 all-depends.dot isn't generated and a new comprehensive task-depends.dot is Nov 29 22:30:06 generated. Nov 29 22:36:52 looks like 2.6.19 is out Nov 29 22:36:59 psokolovsky_, call your office :-) Nov 29 22:37:53 mreimer, Hi! Yep, pH5 already spread the word ;-). will d/l it soon and start look into the breakage ;-) Nov 29 22:38:01 :-) Nov 29 22:45:06 03rpurdie * r709 10bitbake/bin/bitbake: fix typo Nov 29 22:54:29 03tmbinc 07org.oe.dreambox * r512810e1... 10/ (1 packages/neon/neon_0.25.5.bb): neon-0.25.5: add support for pkgconfig Nov 29 22:54:34 03tmbinc 07org.oe.dreambox * r94dd3dc3... 10/ (1 packages/gstreamer/gst-plugin-fluendo-mpegdemux_0.10.8.bb): gst-plugin-fluendo-mpegdemux: add fluendo mpegdemux plugin as alternative to mpegdemux Nov 29 22:54:38 03tmbinc 07org.oe.dreambox * reedd7dd1... 10/ (1 packages/tuxbox/tuxbox-common.bb): tuxbox-common: update SRCDATE to 20061117 for new terrestrial.xml Nov 29 23:11:07 what's the trick to getting pcmciautils to build for angstrom? Nov 29 23:42:22 03rwhitby 07org.oe.dev * rf7ce1b62... 10/ (4 files in 3 dirs): ixp4xx-kernel: Updated to 2.6.19, and set that version as SlugOS default Nov 29 23:43:02 * rwhitby gets the first 2.6.19 released into OE Nov 30 00:26:17 03rwhitby 07org.oe.dev * r0c0a3921... 10/ (1 packages/linux/ixp4xx-kernel_2.6.19.bb): ixp4xx-kernel: Updated to latest SVN to include disk LEDs swapped patch Nov 30 00:50:51 kergoth: ping Nov 30 01:40:32 hi guys. Little question I do not find answered in the documentation on the website... Nov 30 01:40:49 How can I specify two possible download locations in a .bb file? Nov 30 01:47:11 i saw that in a bunch of bb files. Nov 30 01:48:40 Ket-away: I tried to find some, but I must have been unlucky. You happen to have an example? Nov 30 01:48:46 looking Nov 30 01:49:25 ah, here Nov 30 01:49:26 SRC_URI = "${SOURCEFORGE_MIRROR}/prboom/prboom-${PV}.tar.gz \ Nov 30 01:49:26 file://m4.patch;patch=1 \ Nov 30 01:49:26 file://prboom.png \ Nov 30 01:49:26 file://prboom.desktop" Nov 30 01:49:33 er, wait, wrong one Nov 30 01:50:01 bah, copy from my terminal isnt working across NX, just a sec Nov 30 01:51:14 Ket-away: just the filename would be okay you know :-)( Nov 30 01:52:05 im still looking..i know its in here somewhere because i ran into it like three times messing with stuff. Nov 30 01:53:27 Ket-away: exactly... that is the reason why I asked... I tried something now and it seems to accept it (of course I first updated my monotone so it builds the new stuff first) Nov 30 01:54:23 oh wait, my bad. Nov 30 01:54:38 it's spcifying two seperate required sources, not mirrors Nov 30 01:54:58 ie: file1.tgx file2.tgz Nov 30 01:55:27 Ket-away: well I'll try to find it myself and might add that info on thw wiki (if allowed) Nov 30 01:55:47 hey, wnqana buy a slightly bricked 5150? :P Nov 30 01:55:59 it's got a working bootloader. ;) Nov 30 01:57:24 * philippe has enough toys. Just got a Yopy 3500 and 3700 Nov 30 01:57:28 bah Nov 30 01:57:45 but you /want/ this one..it has personality, like..like an inlaw! or a bratty child. :P Nov 30 01:58:12 ..and this marks failed image number four. bah Nov 30 01:58:31 bah bah bah, screw this shit. g'night. Nov 30 02:00:20 night Ket-away **** ENDING LOGGING AT Thu Nov 30 02:59:57 2006