**** BEGIN LOGGING AT Tue Sep 30 02:59:57 2008 Sep 30 05:16:51 tharvey: bitbake packagename -f -cTASKNAME Sep 30 05:41:06 moin Sep 30 05:41:13 i gotta question Sep 30 05:41:45 how can i access the release digit variable of the _architecture_ from within the image bb ? Sep 30 05:45:55 fraxi: what do u mean why release digit Sep 30 05:46:31 7 Sep 30 05:48:08 * raster larts OE for not following symlinks right Sep 30 05:50:04 uhm Sep 30 05:50:21 if i use for example an architecture called opendreambox 1.7 Sep 30 05:50:25 then it's about the 1.7 Sep 30 05:50:32 the version number pretty much Sep 30 05:50:56 distro i mean Sep 30 05:50:57 not arch :) Sep 30 05:51:02 PV? Sep 30 05:51:09 Doesnt that work Sep 30 05:51:54 it's a different .bb file that i want to access that from within Sep 30 05:53:05 DISTRO_VERSION ? Sep 30 05:53:49 hmmm i'll try that Sep 30 06:00:01 cool thing Sep 30 06:00:06 this works Sep 30 06:45:19 hm weird Sep 30 06:45:28 khem maybe you know a solution Sep 30 06:45:42 i am trying to get a substring from a variable Sep 30 06:45:50 but the result is always blank Sep 30 06:46:14 shouldn't ${DISTRO_VERSION:0:3} return the first three characters? Sep 30 06:48:23 In what context? Sep 30 06:48:32 in a task in a bb file Sep 30 06:48:55 echo -e "Link:\thttp://dreamboxupdate.com/${DISTRO}/${DISTRO_VER}/${MACHINE}/experimental" >> ${NFO} Sep 30 06:49:35 results in "Link: http://dreamboxupdate.com/opendreambox//dm8000/experimental" Sep 30 06:49:40 Depends on what is expanding the variable, I guess. Sep 30 06:50:20 how would you do it? Sep 30 06:50:38 But you are using DISTRO_VER in the one, and DISTRO_VERSION in the other -- is that the real code? Sep 30 06:52:23 actually that's because i tried getting the substring externally (outside of the task stray in the bb file) Sep 30 06:52:47 it's the same result with /dreamboxupdate.com/${DISTRO}/${DISTRO_VERSION:0:3}/${MACHINE}/... Sep 30 06:56:01 can you reproduce my problem? Sep 30 07:08:30 it works when i use it in a regular bash shell script :/ Sep 30 07:12:45 okay i found a way how to do it if anyone's interested Sep 30 07:12:53 distrover=${DISTRO_VERSION} Sep 30 07:12:53 echo -e "Link:\thttp://dreamboxupdate.com/${DISTRO}/${distrover:0:3}/${MACHINE}/experimental" >> ${NFO} Sep 30 07:13:16 for some reason, the bb internal variables seem to be non-subscriptable Sep 30 07:13:48 morning Sep 30 07:15:21 hi methril Sep 30 07:15:48 is the #@NAME line in the machine.conf file inaccessible and merely a comment? Sep 30 07:27:39 or can i read it from the image bb Sep 30 07:57:49 bonjour! hi! Sep 30 08:09:18 salut Sep 30 08:28:36 !oebug 3053 Sep 30 08:28:37 * * Bug 3053, Status: NEW, Created: 2007-09-22 11:03 Sep 30 08:28:37 * * cesarb(AT)cesarb.net: ruby-1.8.5 fails to build with dash as /bin/sh Sep 30 08:28:38 * * http://bugs.openembedded.net/show_bug.cgi?id=3053 Sep 30 08:33:26 fixed by author Sep 30 08:34:11 !oebug 4508 Sep 30 08:34:12 * * Bug 4508, Status: NEW, Created: 2008-08-19 12:13 Sep 30 08:34:13 * * ronan(AT)aimao.org: new recipe for OpenCV - The Open Computer Vision Library Sep 30 08:34:14 * * http://bugs.openembedded.net/show_bug.cgi?id=4508 Sep 30 08:34:29 thanks to commit :) Sep 30 08:36:09 genesis, do you know if i can access the #@NAME line of the machine.conf from within the image bb file? Sep 30 08:37:12 * * OE Bug 3053 has been RESOLVED (WORKSFORME) by ronan(AT)aimao.org Sep 30 08:37:14 * * ruby-1.8.5 fails to build with dash as /bin/sh Sep 30 08:37:16 * * http://bugs.openembedded.net/show_bug.cgi?id=3053 Sep 30 08:37:25 i donno fraxi Sep 30 08:38:41 i don't use that for my machine conf Sep 30 08:39:30 okidok Sep 30 12:08:03 is there a bitbake developer wiki ? Sep 30 12:08:42 i'd like share some advance in my bitbake code exploration , since there is no doxygen , class diagram etc it could be useful to have a place , perharps OE wiki section Sep 30 12:09:07 i saw some mail about bitbake-ng and need of documentation , but i didn't see some decision Sep 30 12:40:33 <[X]Spot> Hi guys Sep 30 12:43:59 Genesis: there is a bitbake related wiki at the berlios project site Sep 30 12:44:17 <[X]Spot> I am installing OE following instructions http://www.angstrom-distribution.org/building-%C3%A5ngstr%C3%B6m Sep 30 12:44:33 <[X]Spot> I am getting the following error: Sep 30 12:44:44 <[X]Spot> [root@localhost org.openembedded.stable]# bitbake base-image Sep 30 12:44:45 <[X]Spot> ERROR: Importing sqlite3 and pysqlite2 failed, please install one of them. Python 2.5 or a 'python-pysqlite2' like package is likely to be what you need. Sep 30 12:45:02 <[X]Spot> but I have it installed Sep 30 12:45:33 Genesis: however I don't think it is used much by OE people (or anyone) Sep 30 12:45:41 <[X]Spot> [root@localhost org.openembedded.stable]# rpm -qa | grep sqlite Sep 30 12:45:41 <[X]Spot> sqlite-3.4.2-3.fc8 Sep 30 12:45:41 <[X]Spot> python-sqlite2-2.3.3-1.fc7 Sep 30 12:45:41 <[X]Spot> sqlite-devel-3.4.2-3.fc8 Sep 30 12:45:55 Genesis: I suggest you to put your stuff in the OE wiki Sep 30 12:46:16 [X]Spot, does console-image build for you? Sep 30 12:46:54 <[X]Spot> no Sep 30 12:46:55 <[X]Spot> it's the same Sep 30 12:48:15 <[X]Spot> My python version is : Python 2.4 Sep 30 12:52:59 <[X]Spot> Crofton|work I solved this issue Sep 30 12:53:26 <[X]Spot> there was a version mismatch in my python files.. I have some files from older version Sep 30 12:54:35 ah Sep 30 12:54:38 great! Sep 30 13:28:05 does anyone use udev 124 here? Sep 30 13:28:28 I mean the OE package, not on host system :) Sep 30 14:08:12 * * OE Bug 4588 has been created by chrisfriedt(AT)gmail.com Sep 30 14:08:15 * * 'bitbake rxtx' fails on jamvm-initial and then classpath-native Sep 30 14:08:17 * * http://bugs.openembedded.net/show_bug.cgi?id=4588 Sep 30 14:11:43 Crofton|work: can the FPU in the OMAP3 handle double precision floats? (e.g. instructions like stfd)? Sep 30 14:11:58 the crappy one can Sep 30 14:12:03 the NEON unit can't Sep 30 14:12:07 is my understanding Sep 30 14:13:06 Crofton|work: whats the crappy ones official name? Sep 30 14:13:16 vfp Sep 30 14:13:18 I think Sep 30 14:14:40 Crofton|work: we have the 'funny' problem with the cacao jvm that the OMAP throws illegal instruction when executing stfd. can one switch between the neon and the vfp unit at runtime? Sep 30 14:14:56 not sure Sep 30 14:15:00 yes Sep 30 14:15:10 they should both work Sep 30 14:15:17 try asking in #beagle Sep 30 14:15:29 there might be some people who really know this stuff Sep 30 14:15:37 what compiler are you using? Sep 30 14:15:45 gcc 4.3.1 - what OE provides Sep 30 14:15:57 very strange Sep 30 14:17:04 you have luck in #cacao also Crofton|work. I know twisti has been doing some work relating to that recently. Sep 30 14:17:57 kgilmer: twisti just told me that either we have to use softfloat cacao (very bad) or find a different solution Sep 30 14:18:13 urg Sep 30 14:18:17 * Crofton|work wonders why Sep 30 14:18:48 thebohemian, really? I am working with him to verify a build patch that enables vfp on imx31... Sep 30 14:18:56 hope to get that done today Sep 30 14:19:17 that should also work on beagle .... Sep 30 14:19:41 kgilmer: seen this? http://server.complang.tuwien.ac.at/cgi-bin/bugzilla/show_bug.cgi?id=79 Sep 30 14:20:05 kgilmer: do you consider vfp-support in cacao generally broken? Sep 30 14:20:21 (I only have the beagleboard with its strange neon unit to try) Sep 30 14:21:04 thebohemain, let me get back to you on that. i have not seen this bugzilla but i have seen others and twisti has def made some changes in the past 2 - 3 weeks. i think he's waiting on me for verification. Sep 30 14:22:47 morning Sep 30 14:23:26 kgilmer: just curious. is the imx31 the arm that in the bugs? Sep 30 14:23:33 s/bugs/BUGS/? Sep 30 14:23:52 yes, you got it thebohemian Sep 30 14:24:01 Crofton|work: imx31 is arm11 core, omap3 is cortex-a8 core so their vfp units differ Sep 30 14:24:21 ah Sep 30 14:24:24 how much? Sep 30 14:24:36 lot Sep 30 14:24:38 the good news is with rschuster's help i finally have a building oe (dev branch) that builds the java stuff. Sep 30 14:25:06 * kgilmer things it's about this much: ===> <==== Sep 30 14:25:26 arm11 is armv6 and have VFP unit with double precision. cortex-a8 is armv7 and have VFP unit with double precision but slow and NEON unit with single precision but fast Sep 30 14:25:34 thats how I understand it Sep 30 14:27:10 btw: rschuster == thebohemian ;) Sep 30 14:27:49 oh! hi rschuster :) Sep 30 14:28:08 kgilmer: yeah. so many machines, so many logins. :) Sep 30 14:28:46 kgilmer: nice to hear the OE .dev builds for you now Sep 30 14:29:18 i hear ya. yeah i updated from monotone and the patch made it in. i can now finish my java-library howto on the wiki. Sep 30 14:30:50 hrw: there is only one question. what happens when the CPU executes a double precision instruction? to me it looks like it causes an "illegal instruction" fault Sep 30 14:31:22 cacao's output on the beagleboard: Sep 30 14:31:22 LOG: [0x40246210] md_signal_handler_sigill: Unknown illegal instruction 0xed2d8102 at 0x41963d54 Sep 30 14:31:22 0x41963d54: ed2d8102 stfd f0, [sp, #-8]! Sep 30 14:31:35 (the last line is from the builtin dissassembler) Sep 30 14:35:23 thebohemian: do not know - never tested on armv7 Sep 30 14:37:54 thebohemian, any thoughts on my email on the gnome mobile ml? Sep 30 14:40:46 kgilmer: ah yes, I wanted to comment on this but was too busy yet Sep 30 14:41:27 ok np thebohemian, just curious. Sep 30 14:41:53 i want to reproduce twisti's ability to build cacao with openjdk cp. Sep 30 15:44:45 bye Sep 30 15:55:12 what is the proper way to enable UCLIBC_HAS_LOCALE ? Sep 30 15:56:24 if I enable it I get an error about a missing .//include/bits/uClibc_ctype.h when building uclibc Sep 30 16:56:23 did anyone try building .stable/uclibc with enabled locale support? Sep 30 17:59:00 guys, anyone? is USE_NLS="yes" enough to tell OE.stable to build uclibc with UCLIBC_HAS_LOCALE=y or is anything else needed in addition? does not work out for me Sep 30 18:09:00 whats the url to get oe via git? Sep 30 18:11:02 git://git.openembedded.net/org.openembedded.dev Sep 30 18:11:43 thanks Sep 30 19:02:47 03  07org.openembedded.dev * r8180e65cd8 10OE.dev/: Sep 30 19:02:47 merge of '987daa523fa5a5b15abb45b73b13f80aaf25adff' Sep 30 19:02:47 and 'e6663f24d6776704d75695cd7b8583024c162c8c' Sep 30 19:02:52 03  07org.openembedded.dev * rf8fb559d8c 10OE.dev/ (5 files in 2 dirs): frameworkd,fso-gpsd,zhone git update to something like milestone 3.5 Sep 30 19:02:53 03  07org.openembedded.dev * r68d9083861 10OE.dev/packages/mozilla/fennec_hg.bb: fennec: update SRCREVs Sep 30 19:02:55 03  07org.openembedded.dev * r1b03d93c81 10OE.dev/conf/distro/include/sane-srcrevs.inc: Sep 30 19:02:55 sane-srcrevs.inc: bump freesmartphone.org packages again Sep 30 19:02:56 (no need to increase PR, previous bump has not been pushed!) Sep 30 19:02:58 03  07org.openembedded.dev * rf3ddf11490 10OE.dev/packages/linux/ (4 files in 2 dirs): linux-omap git: add patch to fix soft lockups in MUSB code, run dos2unix on remaining musb patches, bump PR, fix PV Sep 30 19:03:02 03  07org.openembedded.dev * r4bba5b4d96 10OE.dev/packages/freesmartphone/fso-gpsd_git.bb: fso-gpsd git actually link the proper initscript Sep 30 19:03:05 03  07org.openembedded.dev * rc452d77cf9 10OE.dev/: Sep 30 19:03:07 merge of '2189b3da22533b6434462b5c6979a74557c37494' Sep 30 19:03:09 and 'd03a9a17ed9edfe3cfe1ae0cf6e589d38d28e23f' Sep 30 19:33:08 <[X]Spot> why bitbake base-image is failing every time with different errors ? Sep 30 19:33:09 <[X]Spot> ex. bitbake_base_image Sep 30 19:33:17 <[X]Spot> ex. http://stancho.com/bitbake_base_image Sep 30 19:34:04 <[X]Spot> Is the server repository overloaded or ? Sep 30 19:34:19 for omap2 you are better of with the dev bracnh Sep 30 19:34:23 and angstrom 2008 Sep 30 19:34:41 <[X]Spot> I am with OMAP3 Sep 30 19:34:47 <[X]Spot> what do u mean ? Sep 30 19:34:53 DISTRO_VERSION = "2007.12-r18" Sep 30 19:36:19 <[X]Spot> ok . I will set it Sep 30 19:36:23 <[X]Spot> current: DISTRO = "angstrom-2007.1" Sep 30 19:36:48 DISTRO = "angstrom-2008.1" Sep 30 19:37:53 <[X]Spot> ok thanks Sep 30 19:37:55 <[X]Spot> I changed it Sep 30 19:38:05 <[X]Spot> and do bitbake base-image again Sep 30 19:43:45 <[X]Spot> libtool fails every time :( Sep 30 19:43:45 <[X]Spot> ERROR: Build of /OE/org.openembedded.stable/packages/libtool/libtool-native_1.5.10.bb do_compile failed Sep 30 19:43:45 <[X]Spot> ERROR: Task 540 (/OE/org.openembedded.stable/packages/libtool/libtool-native_1.5.10.bb, do_compile) failed Sep 30 19:46:40 <[X]Spot> with ERROR: Build of /OE/org.openembedded.stable/packages/libtool/libtool-native_1.5.10.bb do_populate_staging failed Sep 30 19:47:30 I think you are building the stable branch Sep 30 19:47:50 you shuold be seeing libtool-2 Sep 30 19:49:12 <[X]Spot> I have Sep 30 19:49:12 <[X]Spot> #DISTRO = "angstrom-2007.1" Sep 30 19:49:12 <[X]Spot> DISTRO = "angstrom-2008.1" Sep 30 19:49:12 <[X]Spot> DISTRO_VERSION = "2007.12-r18" Sep 30 19:49:20 <[X]Spot> in /OE/build/conf/local.conf Sep 30 19:51:19 <[X]Spot> q Sep 30 19:52:58 <[X]Spot> Do I have to clean something after I changed the DISTRO_VERSION ? Sep 30 19:55:00 it might be a good idea to wipe tmp and start over Sep 30 19:56:04 <[X]Spot> you mean Sep 30 19:56:07 <[X]Spot> ./OE/angstrom-stable/work/i686-linux Sep 30 19:56:07 <[X]Spot> ? Sep 30 19:56:17 <[X]Spot> to remove it ? Sep 30 19:58:02 what is your TMPDIR? Sep 30 19:59:07 <[X]Spot> ok Sep 30 19:59:07 <[X]Spot> ./OE/${DISTRO}-stable/" Sep 30 20:00:45 rm that directory Sep 30 20:02:59 03rpurdie * r1081 10/ (6 files in 6 dirs): utils.py: Move prunedir function to utils.py and add explode_dep_versions function Sep 30 20:05:31 03rpurdie * r1082 10/ (4 files in 4 dirs): fetch/__init__.py: Raise an exception if SRCREV == 'INVALID' (from Poky) Sep 30 20:06:54 The HorstBox Professional would be a nice target for OE... it has three different platforms in a single small case. Sep 30 20:07:17 03rpurdie * r1083 10/ (4 files in 4 dirs): fetch/hg.py: Fix hg fetcher username/password handling and fix crash Sep 30 20:07:35 specs? Sep 30 20:09:55 03rpurdie 07bitbake-1.8 * r1084 10/lib/bb/parse/parse_py/BBHandler.py: BBHandler.py: Backport a small optimisation in PATH handling from trunk Sep 30 20:12:05 03rpurdie * r1085 10/ (4 files in 4 dirs): providers.py: Fix PACKAGES_DYNAMIC handling of packages with '++' in the name (from poky) Sep 30 20:12:32 <[X]Spot> Crofton|work every time I receive different errors Sep 30 20:12:33 <[X]Spot> Now Sep 30 20:12:33 <[X]Spot> NOTE: package shasum-native-1.0-r1: task do_compile: started Sep 30 20:12:33 <[X]Spot> ERROR: function do_compile failed Sep 30 20:12:33 <[X]Spot> ERROR: see log in /OE/angstrom-stable//work/i686-linux/shasum-native-1.0-r1/temp/log.do_compile.13370 Sep 30 20:12:33 <[X]Spot> NOTE: Task failed: /OE/angstrom-stable//work/i686-linux/shasum-native-1.0-r1/temp/log.do_compile.13370 Sep 30 20:12:34 <[X]Spot> NOTE: package shasum-native-1.0-r1: task do_compile: failed Sep 30 20:12:36 <[X]Spot> ERROR: TaskFailed event exception, aborting Sep 30 20:12:38 <[X]Spot> NOTE: package shasum-native-1.0: failed Sep 30 20:12:40 <[X]Spot> ERROR: Build of /OE/org.openembedded.stable/packages/shasum/shasum-native.bb do_compile failed Sep 30 20:12:42 <[X]Spot> ERROR: Task 18 (/OE/org.openembedded.stable/packages/shasum/shasum-native.bb, do_compile) failed Sep 30 20:12:45 <[X]Spot> NOTE: Tasks Summary: Attempted 5 tasks of which 0 didn't need to be rerun and 1 failed. Sep 30 20:12:46 <[X]Spot> ERROR: '/OE/org.openembedded.stable/packages/shasum/shasum-native.bb' failed Sep 30 20:12:48 <[X]Spot> what is so unstable :( :( Sep 30 20:13:41 you need the dev branch Sep 30 20:13:51 org.openembedded.dev Sep 30 20:14:06 no omap3 work is going on in stable Sep 30 20:14:13 03rpurdie * r1086 10/ (2 files in 2 dirs): parse/__init__.py: Fix missing update_mtime function causing shell reparsing breakage (from poky) Sep 30 20:14:26 chouimat|work: of the HorstBox? Sep 30 20:18:08 yes Sep 30 20:20:23 03hrw * r1087 10/ (branches/bitbake-1.8/AUTHORS trunk/bitbake/AUTHORS): AUTHORS: update my email address Sep 30 20:23:48 <[X]Spot> Crofton|work you mean git clone git://git.openembedded.net/org.openembedded.dev.git ? Sep 30 20:24:02 I think so :) Sep 30 20:24:17 I am pretty sure that is correct, I use mtn, because I need to commit Sep 30 20:24:26 <[X]Spot> ; Sep 30 20:24:27 <[X]Spot> ;) Sep 30 20:24:40 <[X]Spot> will the omap3 be stable ? Sep 30 20:24:42 <[X]Spot> or ..? Sep 30 20:26:33 eventually Sep 30 20:26:42 there is a of work on it atm Sep 30 20:27:17 chouimat|work: Its a quite comples ADSL router with support for all sorts of telephony. Sep 30 20:27:34 which cpu? Sep 30 20:29:24 03rpurdie * r1088 10/ (6 files in 4 dirs): cooker.py: Rename __depends to __base_depends after configuration parsing so we don't recheck the validity of the config files time after time. Also bump the cache revision to match the format change (from poky) Sep 30 20:31:34 chouimat|work: There are three in there: One IXP425 (~500MHz, ARM), a TI AR7 (MIPS32) and a ADSP-BF533 (Blackfin) Sep 30 20:33:22 flo_lap: scary :) Sep 30 20:34:31 chouimat|work: yes indeed... they claim to have released everything. I'll investigate this now - the UI is *bad* Sep 30 20:42:24 03rpurdie * r1089 10/ (6 files in 6 dirs): cooker.py: Add parseConfiguration() function and move inheritFromOS call outside the parser and into cooker Sep 30 20:50:16 <[X]Spot> Crofton|work :( :( no success :( Sep 30 20:50:28 <[X]Spot> with the latest version Sep 30 20:50:32 <[X]Spot> http://stancho.com/bitbake_base_image Sep 30 20:50:43 <[X]Spot> would you please take a look Sep 30 20:51:02 how does it fail? Sep 30 20:58:18 <[X]Spot> see it here please.. http://stancho.com/bitbake_base_image Sep 30 20:58:22 <[X]Spot> it's big Sep 30 21:07:48 ERROR: see log in /OE/angstrom_stable//work/i686-linux/shasum-native-1.0-r1/temp/log.do_compile.14072 Sep 30 21:07:54 what does this file say Sep 30 21:07:55 <[X]Spot> I found the issue Sep 30 21:08:06 <[X]Spot> no space left on /tmp ;) Sep 30 21:08:08 heh Sep 30 21:08:10 ok Sep 30 21:08:13 <[X]Spot> it seems ok now Sep 30 21:08:21 it is strange for stuff to fail that fast Sep 30 21:08:34 <[X]Spot> it seems more stable than the first release Sep 30 21:09:32 <[X]Spot> I hope this time everything will be all right .. and I will compile VLC :) Sep 30 21:09:39 03rpurdie * r1090 10/ (10 files in 6 dirs): Sep 30 21:09:39 bin/bitbake: Add better environmental variable handling. By default it will now Sep 30 21:09:39 only pass certain whitelisted variables into the data store. If BB_PRESERVE_ENV Sep 30 21:09:39 is set bitbake will use all variable from the environment. If BB_ENV_WHITELIST Sep 30 21:09:39 is set, that whitelist will be used instead of the internal bitbake one. Sep 30 21:09:40 Alternatively, BB_ENV_EXTRAWHITE can be used to extend the internal whitelist. Sep 30 21:10:39 <[X]Spot> do I have to add any additional packages to compile vlc + httpd ? Sep 30 21:12:20 03rpurdie * r1091 10/ (2 files in 2 dirs): utils.py: Fix error message (from Joshua Lock) Sep 30 21:56:15 * * OE Bug 4589 has been created by dfoley(AT)techsol.ca Sep 30 21:56:17 * * opkg: opkg-cl causes glibc to emit free() errors Sep 30 21:56:19 * * http://bugs.openembedded.net/show_bug.cgi?id=4589 Sep 30 22:20:22 03rpurdie * r1092 10/ (4 files in 4 dirs): fetch/perforce.py: Fix to use commandline options instead of being overriden by the environment from Martyn Welch Sep 30 22:49:18 what directory is do_install in when it runs? Sep 30 22:51:30 nm, found it. Oct 01 01:56:53 03  07org.openembedded.dev * r222e71beb5 10OE.dev/packages/gnome/libgweather_2.24.0.bb: libgweather_2.24 : Add dependency on gtk+. Oct 01 01:56:54 03  07org.openembedded.dev * r7708c5aaba 10OE.dev/: Oct 01 01:56:54 merge of '3771c597611822587b6e7491171f61908927b778' Oct 01 01:56:54 and 'd901f612d19291dc3b05e29b1d430e991e96602d' **** ENDING LOGGING AT Wed Oct 01 02:59:59 2008