**** BEGIN LOGGING AT Thu Dec 23 02:59:58 2010 Dec 23 07:41:27 gm Dec 23 07:45:22 hi eFfeM_work Dec 23 07:45:35 gm Dec 23 07:45:41 hi JaMa Dec 23 07:45:51 JaMa: there is a patch on u-boot mailing list for u-boot & gcc 4.5.1 Dec 23 07:46:27 hi JaMa ericben, it is good to be fully operational again; had some lousy flu days Dec 23 07:46:44 03Koen Kooi  07org.openembedded.dev * r73499d3d79 10openembedded.git/recipes/ffmpeg/omapfbplay-dce_git.bb: Dec 23 07:46:44 omapfbplay-dce: bump PR for libdce change Dec 23 07:46:44 Signed-off-by: Koen Kooi Dec 23 07:47:17 JaMa: http://lists.denx.de/pipermail/u-boot/2010-December/084132.html Dec 23 07:47:19 ericben: which one? that build issue? Dec 23 07:47:50 ah this one is different, thanks, trying Dec 23 07:47:52 JaMa: this affect omap3 : http://lists.denx.de/pipermail/u-boot/2010-December/083817.html Dec 23 07:47:59 so this may affect your N900 Dec 23 08:58:28 good morning Dec 23 09:12:11 hi mckoan Dec 23 09:25:08 03Koen Kooi  07org.openembedded.dev * r9c3f54f903 10openembedded.git/recipes/linux/linux-omap_2.6.37rc.bb: Dec 23 09:25:08 linux-omap 2.6.37rc: bump to rc7 Dec 23 09:25:08 Signed-off-by: Koen Kooi Dec 23 09:25:16 03Koen Kooi  07org.openembedded.dev * r53175ec62c 10openembedded.git/recipes/linux/ (9 files in 2 dirs): Dec 23 09:25:16 linux-omap4: add patch to enable DVI Dec 23 09:25:16 Signed-off-by: Koen Kooi Dec 23 09:25:27 03Koen Kooi  07org.openembedded.dev * rf9d1c6273a 10openembedded.git/recipes/linux/linux-omap-psp-2.6.32/am3517-crane/defconfig: Dec 23 09:25:27 linux-omap-psp 2.6.32: update defconfig for craneboard to be more usefull and not break udev Dec 23 09:25:27 Signed-off-by: Koen Kooi Dec 23 09:53:03 good morning Dec 23 09:53:39 florian: hi Dec 23 10:11:41 A short question about overlay: How is the search path working? Is it only for recipes and conf files? And not for file://-files? Dec 23 10:14:01 I tried to do my own distro configuration and wanted to have my local "local/recipes/base-files/base-files/mydistro/issue" file but that is not copied unless I make a local recipes/base-files/base-files_3.0.14.bb Dec 23 10:16:30 With a local recipes/base-files/base-files_3.0.14.bb it complains about not finding the nsswitch.conf file in the local tree but it is in the base openembedded tree Dec 23 10:18:50 Andy_: I think you may need to use a .bbappend file Dec 23 10:19:10 (without copying the base-files recipe into your layer) Dec 23 10:19:10 bluelightning: you shouldn't need to just to override a file Dec 23 10:19:32 foerster: ok Dec 23 10:19:42 if the path is set properly, it should be able to pull files from overlay first. I do it here. Dec 23 10:20:41 foerster: is this bitbake version dependent perhaps? Dec 23 10:21:26 shouldn't be. Dec 23 10:21:37 one just manipulates FILESPATHBASE to look in overlay first Dec 23 10:24:07 foerster: ah, so extra configuration is required then... Dec 23 10:25:47 So, what? Should it work with just having a local "local/recipes/base-files/base-files/mydistro/issue" file? Dec 23 10:26:12 bluelightning: yes. Dec 23 10:26:23 Andy_: FILESPATHBASE needs modified to look in your overlay Dec 23 10:26:40 Andy_, bluelightning: see http://pastebin.com/6R3vKTK9 Dec 23 10:27:09 that would go in build/local.conf Dec 23 10:27:22 with the DEV_BASE and COLLECTIONS set to your environment Dec 23 10:27:57 that way, file:// URIs are searched in the overlay first, then in the oe, so you can override file://s for any package. Dec 23 10:28:21 Ok, I'll check that. Thanx! Dec 23 10:28:25 The order you specify them in COLLECTIONS is the order they are searched. Dec 23 10:28:27 np Dec 23 10:29:25 btw, there was an RFC on the ml recently to build this directly into oe (since it's very common, but not intuitive). Hopefully that gets worked in oe directly. Dec 23 11:10:53 how can i prevent bitbake from unpacking a file it downloaded from SRC_URI? i'd like it to stay compressed in WORKDIR Dec 23 11:12:51 s/bitbake/oe/ Dec 23 11:34:38 i think it isn't possible at all... any objections against a new parameter unpack=yes/no for SRC_URIs? i need to download a jar file and use it as is, without being unpacked. Dec 23 12:10:37 obi, oe.unpack_file doesn't seem to handle .jar files (from the looks it should just copy it, no?) Dec 23 12:11:07 it does Dec 23 12:11:30 elif file.endswith('.zip') or file.endswith('.jar'): Dec 23 12:11:30 cmd = 'unzip -q -o' Dec 23 12:17:36 obi, if you can put that jar into a separate recipe you could overwrite unpack there. Otherwise an unpack=yes/no parm would indeed be in order Dec 23 12:18:16 s/overwrite/override/ ;# even Dec 23 12:22:16 i intend to use it in an .inc file used by gcj. i'll send a patch to the mailing list regarding that param Dec 23 13:29:08 anyone any experience with uclibc++ under OE? Dec 23 15:21:50 e Dec 23 15:22:08 eFfeM_work: sorry no uclibc here and even less uclibc++ :) Dec 23 15:23:45 03Eric BENARD  07org.openembedded.dev * r301be24e66 10openembedded.git/recipes/compat-wireless/ (7 files): Dec 23 15:23:45 compat-wireless: Add latest Linux wireless drivers Dec 23 15:23:45 - provides latest (linux-next) drivers for older kernels Dec 23 15:23:45 Signed-off-by: Eric Bénard Dec 23 15:23:52 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r2e02ec3cfa 10openembedded.git/recipes/compat-wireless/compat-wireless.inc: Dec 23 15:23:52 compat-wireless.inc : fix installation Dec 23 15:23:52 using oe_runmake export LDFLAGS which makes the installation fail, Dec 23 15:23:52 so we use LDFLAGS="" for the installation Dec 23 15:23:52 Signed-off-by: Denis 'GNUtoo' Carikli Dec 23 15:23:52 Signed-off-by: Eric Bénard Dec 23 15:27:35 hmm,, anyone know how to get git-apply to apply with fuzz? Dec 23 15:30:35 kergoth: -C Dec 23 15:31:30 Ensure at least lines of surrounding context match before and after each change. When fewer lines of surrounding context exist they all must match. By default no context is ever ignored. Dec 23 15:32:06 hmm Dec 23 15:32:35 i don't think thats the same as the fuzz factor of patch Dec 23 15:32:45 fuzz has to do with line offsets when looking for hunks Dec 23 15:32:50 not amount of context Dec 23 15:36:51 kergoth: if I understand well, -C was implemented to workaround fuzz problems : http://kerneltrap.org/mailarchive/git/2006/4/10/203439 Dec 23 15:37:44 hmm, okay, will try it out Dec 23 15:37:45 thanks Dec 23 16:54:54 03Alex Ferguson  07org.openembedded.dev * race9e50080 10openembedded.git/recipes/base-files/base-files/ben-nanonote/fstab: Dec 23 16:54:54 base-files: Add fstab for ben-nanonote. Dec 23 16:54:54 Added fstab, which among other things enables swap by default. Dec 23 16:54:54 Signed-off-by: Alex Ferguson Dec 23 16:54:54 Signed-off-by: Kristoffer Ericson Dec 23 16:55:04 03Alex Ferguson  07org.openembedded.dev * r0f2dd9f380 10openembedded.git/recipes/echinus/ (echinus-0.4.3.2/echinus-jlime.patch echinus_0.4.3.2.bb): Dec 23 16:55:04 echinus: Add rule for jamvm. Dec 23 16:55:05 Added a rule for mapping jamvm to a fullscreen window by Dec 23 16:55:05 default. Dec 23 16:55:05 Bumped PR in echinus-0.4.3.2 recipe. Dec 23 16:55:05 Signed-off-by: Alex Ferguson Dec 23 16:55:05 Signed-off-by: Kristoffer Ericson Dec 23 16:55:06 03Alex Ferguson  07org.openembedded.dev * r8cb2a9aa2b 10openembedded.git/recipes/tasks/task-muffinman.bb: Dec 23 16:55:07 task-muffinman: Add jamvm, midpath and xmlpull. Dec 23 16:55:07 This, along with additions to other packages enables a Dec 23 16:55:07 working J2ME environment. Dec 23 16:55:07 Bumped PR. Dec 23 16:55:08 Signed-off-by: Alex Ferguson Dec 23 16:55:08 Signed-off-by: Kristoffer Ericson Dec 23 16:55:20 03Alex Ferguson  07org.openembedded.dev * r5667943cd0 10openembedded.git/recipes/midpath/ (6 files in 2 dirs): (log message trimmed) Dec 23 16:55:20 midpath: Fix midpath-core and midpath-cldc, add patches. Dec 23 16:55:20 midpath-core was slightly refactored, and now installs the Dec 23 16:55:20 configuration files needed which are picked up locally and Dec 23 16:55:20 can be overriden per-platform. A much more elegant solution, Dec 23 16:55:20 in my opinion. Two patches were added to midpath-core, one Dec 23 16:55:21 which changes the default UI colors to a more palatable gray Dec 23 16:55:30 03Alex Ferguson  07org.openembedded.dev * rc7e439d2f1 10openembedded.git/recipes/jlime/jlime-extras_1.0.4.bb: Dec 23 16:55:30 jlime-extras: Update jlime-extras recipe to version 1.0.5. Dec 23 16:55:30 Version 1.0.5 contains some fixes and configuration for jamvm. Dec 23 16:55:30 Signed-off-by: Alex Ferguson Dec 23 16:55:30 Signed-off-by: Kristoffer Ericson Dec 23 17:04:23 03Alex Ferguson  07org.openembedded.dev * r41250d2e67 10openembedded.git/recipes/echinus/echinus-0.4.3.2/echinus-jlime.patch: Dec 23 17:04:23 echinus: Change rule for jamvm to cvm. Dec 23 17:04:23 Changed Echinus*rule3 from jamvm to cvm because we use phoneme now. Dec 23 17:04:23 Signed-off-by: Alex Ferguson Dec 23 17:04:23 Signed-off-by: Kristoffer Ericson Dec 23 17:04:33 03Alex Ferguson  07org.openembedded.dev * r02c7f27aff 10openembedded.git/recipes/phoneme/phoneme-advanced.inc: Dec 23 17:04:33 phoneme-advanced.inc: Fix recipe for mips targets. Dec 23 17:04:33 Fixed recipe for mips targets by implementing oe_phoneme_configmips Dec 23 17:04:33 which specifies -lm for LINK_ARCH_LIBS. Dec 23 17:04:33 Signed-off-by: Alex Ferguson Dec 23 17:04:33 Signed-off-by: Kristoffer Ericson Dec 23 17:04:35 03Alex Ferguson  07org.openembedded.dev * rc0c5d35b7e 10openembedded.git/recipes/midpath/files/jlime/midpath-launcher-j2se: Dec 23 17:04:35 midpath: Add custom midpath-launcher-j2se script for phoneme. Dec 23 17:04:36 The default script assumes a jvm other than phoneme. Since we Dec 23 17:04:36 use phoneme in jlime, we'll need a custom script. Dec 23 17:04:36 Signed-off-by: Alex Ferguson Dec 23 17:04:36 Signed-off-by: Kristoffer Ericson Dec 23 17:04:37 03Alex Ferguson  07org.openembedded.dev * r545fec8616 10openembedded.git/recipes/tasks/task-muffinman.bb: Dec 23 17:05:14 task-muffinman: Use phoneme instead of jamvm. Dec 23 17:05:14 PhoneME is slighly faster than jamvm, so we use that instead. Dec 23 17:05:14 Bumped PR. Dec 23 17:05:14 Signed-off-by: Alex Ferguson Dec 23 17:05:14 Signed-off-by: Kristoffer Ericson Dec 23 17:31:36 gitorious.org seems to have forgotten to pay its DNS Dec 23 17:31:40 provider Dec 23 17:31:43 ouch Dec 23 17:31:44 :) Dec 23 17:31:49 * kergoth hugs github Dec 23 17:32:00 we talk about hosting on external entities Dec 23 17:32:12 khem yes Dec 23 17:32:17 saw this too today Dec 23 17:32:24 so we wait ? Dec 23 17:32:41 * khem better utilize the time taking shower Dec 23 17:33:04 wait for an wonder? Dec 23 17:34:20 gtk+ explodes my minimal build Dec 23 17:34:53 hm.. only for akita Dec 23 17:35:21 http://tinderbox.openembedded.net/packages/1387140/ Dec 23 17:35:29 no log again :( Dec 23 17:36:09 kergoth: Parsing recipes...ERROR: Error parsing /storage/oe/testbuilder/openembedded/recipes/freesmartphone/fsodeviced_git.bb: database is locked Dec 23 17:36:17 specially for you :) Dec 23 17:36:38 ? Dec 23 17:37:03 ah, i bet i know what that is Dec 23 17:41:12 03Chris Larson  07master * r628c5159d1 10bitbake.git/lib/bb/persist_data.py: Dec 23 17:41:12 persist_data: resurrect the lock wait for selects Dec 23 17:41:12 Think this got inadvertantly dropped when switching to the new API. Dec 23 17:41:12 Signed-off-by: Chris Larson Dec 23 17:41:21 :) Dec 23 17:59:15 Jay7: when did u build this Dec 23 18:02:54 Jay7: I doubt that client fails to send Dec 23 18:04:00 the IP for gitorious is 87.238.52.168 Dec 23 18:04:11 if anyone is eager Dec 23 18:07:37 ericben, hi Dec 23 18:08:00 some compat-wireless patches were pushed, was it you? Dec 23 18:09:01 GNUtoo|laptop: yes he did it Dec 23 18:09:11 ok nice Dec 23 18:09:15 GNUtoo|laptop: is something missing Dec 23 18:09:18 no Dec 23 18:09:21 k Dec 23 18:09:31 I was just afraid to have pushed something by error Dec 23 18:10:30 eh Dec 23 18:10:31 ok Dec 23 18:12:57 * khem off to work Dec 23 18:29:24 03Tom Rini  07org.openembedded.dev * r9dd7f1ce76 10openembedded.git/recipes/python/python-pysqlite2_2.4.0.bb: Dec 23 18:29:24 python-pysqlite2: Fix SRC_URI Dec 23 18:29:24 Signed-off-by: Tom Rini Dec 23 18:30:18 We should also really think about moving to 2.6.0 Dec 23 18:30:26 2.4 isn't anywhere except oddball mirror sites Dec 23 18:52:37 Why is it that the xload recipe almost always fails during fetch? Dec 23 18:52:48 then I never remember what I last did to fix it Dec 23 18:53:05 Ahh, | make: *** No rule to make target `omap3530beagle_config'. Stop. Dec 23 18:53:12 That is a valid target, no? Dec 23 18:53:52 IIRC the problem was it was pulling an old version for some strange reason Dec 23 19:32:33 kergoth: is tag=xxxx a valid attribute for src_uri Dec 23 19:32:49 I need to fetch a tag from a branch of git Dec 23 19:32:58 repo Dec 23 19:33:23 again hmm git fetcher chickens out when fetching from gitorious Dec 23 19:34:07 iirc yes Dec 23 19:35:30 well, since git doesn't distinguish for checkouts, you could probably use tag=, or rev= or whatever, or just put the tag in SRCREV Dec 23 19:35:33 * kergoth hates our fetchers Dec 23 20:10:05 * reactor16 Hi all Dec 23 20:16:34 khem: narrowed that gcc issue in u-boot, do you have a sec? Dec 23 20:20:00 khem: http://build.shr-project.org/tests/jama/u-boot/gcc-4.5/ Dec 23 20:30:45 okay, released 1.8.19 and 1.10.2 Dec 23 20:31:02 haven't sent an announcement yet, or uploaded tarballs to berlios, but the git bits are in Dec 23 20:31:37 note that the pristine-tar metadata for the official release tar.gz is on the pristine-tar branch, it can be used to reconstruct the *exact* tarball i created for the release, with matching md5sum Dec 23 20:33:20 really need to make a script for this, ideally including the upload of the tarball and release announcement Dec 23 20:37:01 aargh, somehow libtool and me is an incompatbile match, did a clean rebuild, git head from 2 days ago and I get lots of .la files with lines like: Dec 23 20:37:02 libusb.la:dependency_libs=' =/usr/lib/libusb-1.0.la -lrt' Dec 23 20:37:14 note the = before /usr/lib Dec 23 20:37:29 that looks right Dec 23 20:37:44 JaMa|Off: you mean the = too ? Dec 23 20:38:00 '=' is sysroot Dec 23 20:38:30 read what I said to you and woglinde 2 days ago Dec 23 20:38:32 oh wow, i didn't know about rebase -i's "fixup" and "reword" options Dec 23 20:38:37 Hm. then gmpc has a libtool issue. it says: Dec 23 20:38:39 | libtool: link: cannot find the library `=/usr/lib/libgnutls.la' or unhandled argument `=/usr/lib/libgnutls.la' Dec 23 20:38:50 JaMa|Off: missed that = was for sysroot Dec 23 20:39:09 yes that's whole point of sysroot feature :) Dec 23 20:39:29 ah ok, guess I learned something new Dec 23 20:39:40 replacing hardcoded sysroot in .la files with = which is then replaced with sysroot specified in parameter Dec 23 20:41:18 so the gmpc error I get is because it is using its own libtool which does not know about sysroot Dec 23 20:42:03 looks like it Dec 23 20:42:15 03Martin Jansa  07master * re697ea2df9 10openembedded.git/recipes/vala/vala_0.11.2.1.bb: Dec 23 20:42:15 vala: add version 0.11.2.1 Dec 23 20:42:15 Signed-off-by: Martin Jansa Dec 23 20:42:18 03Martin Jansa  07master * rd486ced03e 10openembedded.git/ (2 files in 2 dirs): Dec 23 20:42:18 dbus: upgrade 1.4.0 to 1.4.1 and make it new default CVE-2010-4352 Dec 23 20:42:18 Signed-off-by: Martin Jansa Dec 23 20:42:18 Acked-by: Khem Raj Dec 23 20:42:24 03Martin Jansa  07master * ra23e98361e 10openembedded.git/ (15 files in 3 dirs): Dec 23 20:42:24 u-boot_git: add support for nokia900 Dec 23 20:42:24 Signed-off-by: Martin Jansa Dec 23 20:44:44 JaMa, can I bug you about fso stuff that gets pulled into console-image/etc? :) Dec 23 20:46:17 Tartarus: yes Dec 23 20:46:54 * Tartarus digs out a log Dec 23 20:47:39 http://pastebin.com/Xqdd82Zd Dec 23 20:48:12 If not clear from the extra long paths in there, DISTRO=minimal-uclibc MACHINE=qemuarm BBPKGS=nas-server-image Dec 23 20:48:25 (also pulled in on other images too, just the first one I found) Dec 23 20:48:49 * Tartarus confirms same deal, BBPKGS=console-image Dec 23 20:48:59 i have committed a fix yesterday Dec 23 20:49:07 when we bump FSO next time, it should work Dec 23 20:49:14 i detect whether we can use backtrace now or not Dec 23 20:49:19 Can we bump before this next testing tag? Dec 23 20:49:31 I'd love an xmas surprise of a bunch of red builds turning blue Dec 23 20:49:35 when is this tag going to happen? Dec 23 20:49:40 (silly hudson, blue==pass?) Dec 23 20:49:50 sometime today Dec 23 20:49:54 i forget the usual hour Dec 23 20:50:48 hmm, that might be too tight. i just merged two months of work in a branch to FSO master and I need to fix a bunch of things before that compiles in OE Dec 23 20:50:56 besides... my day ends now: ) Dec 23 20:50:59 aw, k Dec 23 20:51:03 merry xmas then :) Dec 23 20:51:25 thanks, for you too. chose the timing to merge all that work since there's some spare time over the next couple of days Dec 23 20:51:35 (to fix and repair) Dec 23 20:55:59 kergoth just fyi, i got the feeder error again and now did not do things like ^C; it might be I'm getting this since moving from opensuse 11.2 and python 2.6.2 to opensuse 11.3 and python 2.6.5 Dec 23 21:11:00 eFfeM: hmm, interesting Dec 23 21:11:04 still no idea what would cause it Dec 23 21:11:34 kergoth neither do i Dec 23 21:11:38 03Frans Meulenbroeks  07org.openembedded.dev * r6c7c4f7d53 10openembedded.git/recipes/musicpd/gmpc_0.20.0.bb: Dec 23 21:11:38 gmpc: fix libtool issue, used inherit autotools Dec 23 21:11:38 Signed-off-by: Frans Meulenbroeks Dec 23 21:12:03 GNUtoo|laptop: you said you had problems with gmpc, can you check that they are gone with the patch i just pushed ? Dec 23 21:25:51 eFfeM, ok I'll check thanks a lot Dec 23 21:38:20 hm how could I use ln -s to link to a dir that dont exists Dec 23 21:41:42 args wrong way Dec 23 22:39:22 eFfeM_Zzz, it compiled Dec 23 22:41:27 eFfeM_Zzz: btw in your last gmpc, rm -f would be better as you can run -c configure twice (as iirc koen also said in simalar change) Dec 23 22:41:36 s/last/last commit Dec 23 22:52:39 JaMa: Dec 23 22:52:45 I am here now Dec 23 22:55:01 Can anyone tell me about the history of USE_NLS_glib-2.0 = "yes" ? Dec 23 22:55:18 it makes glib-2.0 pull gettext into the build Dec 23 22:55:46 and i can't see recent versions of glib-2.0 requiring gettext (as opposed to another libintl) Dec 23 22:57:15 so, for uclibc builds, if proxy-libintl is built before gettext, gettext will overwrite libintl.so Dec 23 22:57:27 and glib-2.0 links against the wrong libintl Dec 23 22:58:06 i'm just not sure if i can remove all instances of USE_NLS_glib-2.0 = "yes" from conf files Dec 23 22:58:30 khem: I'm just going to bed :/ but did you noticed that url I've sent you? Dec 23 22:59:13 khem: there is source and generated asm for working -O0 and not-working -Os/-O1 from gcc-4.5 Dec 23 22:59:25 also pushed all 3 versions from gcc-4.1.2 Dec 23 23:01:03 JaMa: hmm Dec 23 23:01:10 OK Dec 23 23:01:21 JaMa: do u know which funciton it is Dec 23 23:01:27 by chance Dec 23 23:02:00 grg: gettext or proxy-libintl both can not reside together Dec 23 23:02:08 no :/ only the result in runtime -> hang while starting kernel, but no idea what was real reason Dec 23 23:02:11 grg: for uclibc we should avoid gettext Dec 23 23:02:28 JaMa: OK I Will try to poke the srcs Dec 23 23:02:29 khem, i know. so having USE_NLS_glib-2.0 = "yes" breaks uclibc builds Dec 23 23:02:42 JaMa: but we need to find hands on time together to work on it Dec 23 23:02:49 JaMa: I am sure we are close to fixing it Dec 23 23:02:58 grg: remove that Dec 23 23:03:13 grg: IIRC I removed it for minimal-uclibc Dec 23 23:03:29 khem, ok. I just want to be sure about why it was there in the first place. (couldn't find a good reason in commit logs) Dec 23 23:03:36 grg: but we cant build everything with proxy-libintl Dec 23 23:04:03 khem: there were some errors with inline in u-boot+gcc-4.5, ericben sent me link in the morning http://lists.denx.de/pipermail/u-boot/2010-December/083817.html Dec 23 23:04:13 grg: glibc specifically needed nls support Dec 23 23:04:32 JaMa: ok is it related Dec 23 23:04:43 khem: but that rx41.c is not included in upstream u-boot, that's probably the reason why nobody reported it yet Dec 23 23:05:01 now really gtg, gnight all Dec 23 23:05:44 JaMa: ok gn Dec 23 23:05:59 JaMa: ah I see Dec 23 23:16:56 obi r u here ? Dec 23 23:17:35 hmmm... given the .conf format is the same as the recipe format, mostly, to truly abstract the file format away, the file format specific code would have to handle loading user configuration as well as finding and loading recipes Dec 23 23:17:39 * kergoth ponders Dec 24 00:04:20 whew, think thats one of the last kinks remaining in the bitbake-master-with-poky-bits branch Dec 24 00:04:37 builds both oe and poky successfully and well, just one issue remaining afaict Dec 24 00:07:20 thats nice....a lot of work to get there Dec 24 00:09:25 was a royal pain, but a relief to be there Dec 24 00:10:14 this last bug is very strange Dec 24 00:10:42 if we save the python functions into the metadata for later use in the signatures, even if those bits are never used otherwise, the build errors out saying it can't call a string object Dec 24 00:10:54 * kergoth scratches head Dec 24 00:12:20 foerster: hmm, i think we may need to check for exit code issues surrounding keyboardinterrupt handling -- seems like its returning 0 sometimes when we interrupt and it cleanly shuts down Dec 24 00:12:24 * kergoth pokes at knotty Dec 24 00:12:55 you just need to check for exit code issues :) Dec 24 00:13:00 ugh, yes. Dec 24 00:13:07 some have already been fixed, but they're a pain to nail down Dec 24 00:13:11 so many possible ways we can fail Dec 24 00:13:57 just return 0 only when all is ok and was done successfully Dec 24 00:13:57 non-0 in all other cases Dec 24 00:14:10 * kergoth rolls eyes Dec 24 00:14:23 like there's only one codepath for an error :) Dec 24 00:14:38 its intertwined with the mess we have for exception handling Dec 24 00:14:51 well.. I understand :) Dec 24 00:15:13 but I wish to have way to know was build succeed or not :) Dec 24 00:15:28 seems I'll grep tail of log for 'failed' Dec 24 00:36:07 how soon until the testing-next tag? Dec 24 00:45:47 Tartarus, 5 minutes after i leave the office... i bet :) Dec 24 00:54:02 which will be about 5 minutes after i'm gone too, heh Dec 24 00:54:10 * Tartarus wants to kick the autobuilder off Dec 24 01:04:57 testing-next is ready Dec 24 01:05:02 sorry for the delay **** BEGIN LOGGING AT Fri Dec 24 01:26:45 2010 Dec 24 02:25:27 03Chris Larson  07master * rb0a3cc0eac 10bitbake.git/lib/bb/data_smart.py: Dec 24 02:25:27 data_smart: don't include functions in name lookups Dec 24 02:25:27 For raw name lookups in python snippets, don't return the values for functions Dec 24 02:25:27 in the metadata. This was necessary to fix issues once the methodpool Dec 24 02:25:27 functions get added to the metadata. Dec 24 02:25:27 Signed-off-by: Chris Larson Dec 24 02:25:36 03Chris Larson  07master * r29f6bceac0 10bitbake.git/lib/bb/runqueue.py: Dec 24 02:25:36 runqueue: simplify next buildable task Dec 24 02:25:36 Signed-off-by: Chris Larson Dec 24 02:32:12 03Chris Larson  07master * rc58ffedd14 10bitbake.git/lib/bb/persist_data.py: Dec 24 02:32:12 persist_data: unbork the resurrect of select lock wait Dec 24 02:32:12 Signed-off-by: Chris Larson Dec 24 02:32:39 * kergoth kicks off another git test-sequence to confirm sanity on the poky-sync branch Dec 24 02:40:45 * kergoth doesn't know how he got along without git test-sequence and bisect run Dec 24 02:55:24 * cbrake looks at git test-sequence ... **** ENDING LOGGING AT Fri Dec 24 02:59:58 2010