**** BEGIN LOGGING AT Wed Nov 04 02:59:57 2009 Nov 04 03:49:26 03Khem Raj  07org.openembedded.dev * r1974e79f65 10openembedded.git/recipes/proxy-libintl/proxy-libintl_20080418.bb: Nov 04 03:49:26 proxy-libintl_20080418.bb: Fix QA error missing GNU_HASH Nov 04 03:49:26 Signed-off-by: Khem Raj Nov 04 05:18:34 Anyone got anything interesting in their /tmp/opkg-intercept-* directories? Nov 04 05:20:47 PS: opkg has code to remove the opkg-intercept directory, but due to an earlier setenv("PATH", "/dev/null"), it fails to remove Nov 04 06:04:44 morning Nov 04 06:08:06 how can i force opkg to install a local version even if it is versionwise older? I already tried remove then opkg install -force-downgrade ./my.ipk Nov 04 06:11:37 eFfeM, try -force-reinstall Nov 04 06:11:48 -force-downgrade doesn't seem to do anything Nov 04 06:12:23 grg, no luck it still wants to download from the web Nov 04 06:12:40 probably a real bug in opkg Nov 04 06:13:14 it reads in your Packages.gz file from the repository and want that version instead Nov 04 06:13:26 i'll have a look into it Nov 04 06:15:00 hm actually it is a qt4 package and we seem to have different flavours of it Nov 04 06:15:42 what is the meaning of the last digit in the package name? Nov 04 06:15:42 can you try a bunch of -force options all together, and see how that goes? Nov 04 06:15:46 ls Nov 04 06:16:24 what I get is this: Nov 04 06:16:25 dunno Nov 04 06:16:25 root@beagleboard:~# opkg -force-reinstall install ./qt*ipk Nov 04 06:16:25 Reinstalling qt4-plugin-sqldriver-sqlmysql (4.5.2-r12.3.3) on root... Nov 04 06:16:25 Downloading http://www.angstrom-distribution.org/feeds/unstable/ipk/glibc/armv7a/base/qt4-plugin-sqldriver-sqlmysql_4.5.2-r12.3.3_armv7a.ipk Nov 04 06:16:40 what I build locally is qt4-plugin-sqldriver-sqlmysql_4.5.2-r12.2.4_armv7a.ipk Nov 04 06:17:33 r12.2 comes from qt4-x11-free whereas r12-3 is generated by qt-x11-free-gles Nov 04 06:17:54 as a workaround, you can comment out your repository in the feed file and do an opkg update Nov 04 06:18:01 then installing the file should work Nov 04 06:18:08 anyone here familiar w/openocd? Nov 04 06:18:16 oh, as workaround I can do a manual install Nov 04 06:20:20 eFfeM, can you -force-reinstall with -V3 and paste the output into a pastebin for me? Nov 04 06:21:25 k Nov 04 06:24:17 grg it became an upload to filebin, file is 10k; http://filebin.ca/qjbwcz/opkg Nov 04 06:26:36 i don't see any mention of your r12.2.4 version Nov 04 06:27:29 well, this was my command line: Nov 04 06:27:31 opkg install -force-reinstall -V 3 ./qt4-plugin-sqldriver-sq Nov 04 06:27:32 lmysql_4.5.2-r12.2.4_armv7a.ipk >/tmp/opkg Nov 04 06:27:49 ofc in one line, terminal window caused line break Nov 04 06:28:31 sure Nov 04 06:28:59 i have to go soon, but I'll look into this tomorrow Nov 04 06:29:16 03Frans Meulenbroeks  07org.openembedded.dev * r7032ae4307 10openembedded.git/recipes/mythtv/mythplugins_0.21+0.22rc2.bb: mythplugins: added zoneminder, cleaned up images in do_stage Nov 04 06:29:38 okie, i'm normally not on this early but in the evenings (gmt+1) you can often find me here Nov 04 06:30:02 under no circumstances should opkg install a different file to the one you specify on the command line. It should install the one you want, or give an error Nov 04 06:30:49 i'm gmt+1030, usually online between 9am-5pm Nov 04 06:31:14 ah ok Nov 04 06:32:33 bye Nov 04 06:34:48 Hello anybody here ? Nov 04 06:35:11 134 people here :-) Nov 04 06:35:16 ?? Nov 04 06:37:09 could not find any discussions about OE here Nov 04 06:38:56 it is 6.38 GMT, normally it is pretty quiet at this time, the US people are gone to sleep and the europeans are not there yet Nov 04 06:39:15 but if you have a Q just ask & wait Nov 04 06:41:31 hmmm Nov 04 06:41:41 i am in the asian time zone Nov 04 06:43:26 in 2 hrs there will probalby be some more activity here Nov 04 06:58:36 ok Nov 04 06:59:06 what are you here for ? Nov 04 07:09:28 gg Nov 04 07:17:53 nin amman Nov 04 07:27:48 ? Nov 04 08:04:44 good morning Nov 04 08:12:13 03Phil Blundell  07shr/merge * r5bf7719db1 10openembedded.git/recipes/eglibc/ (3 files in 2 dirs): Nov 04 08:12:13 eglibc: More armv4t interwork fixes and one typo in existing patch Nov 04 08:12:13 Signed-off-by: Sebastian Spaeth Nov 04 08:19:28 03Koen Kooi  07org.openembedded.dev * r4b01adc581 10openembedded.git/recipes/gnome-mplayer/gecko-mediaplayer/gecko-mplayer-svn.diff: gecko-media-player: add missing patch Nov 04 08:19:30 03Koen Kooi  07org.openembedded.dev * rf33edf7de7 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Nov 04 08:19:30 opkg: bump SRCREV to r240 Nov 04 08:19:30 * tested on angstrom/beagleboard and angstrom/sheeva by upgrading the system after the DISTRO_PR bump Nov 04 08:19:30 03Koen Kooi  07org.openembedded.dev * r37930cdaf9 10openembedded.git/conf/distro/include/sane-srcrevs.inc: efl: bump SRCREV to 43437 Nov 04 08:19:33 03Koen Kooi  07org.openembedded.dev * r4964bb59f5 10openembedded.git/recipes/ (e17/expedite_svn.bb efl1/evas.inc efl1/evas_svn.bb): expedite, evas: don't turn of the GL engine forcefully Nov 04 08:39:57 03Sebastian Spaeth  07shr/merge * rb7c9855c7e 10openembedded.git/recipes/glibc/files/armv4t-interworking.patch: Nov 04 08:39:57 glibc: fix armv4t interworking patch typo. Thanks to Phil Blundell for the hint. Nov 04 08:39:57 Signed-off-by: Sebastian Spaeth Nov 04 09:18:39 any sql wiz around? I try to restart sql, but even if I run /usr/libexec/mysqld as root it complains that it cannot write /var/run/mysqld.pid (athough the dir exists, owned root, mode 755) and others are able to write in it as well Nov 04 09:19:29 03Frans Meulenbroeks  07org.openembedded.dev * r5277f843e6 10openembedded.git/recipes/qt4/qt4-x11-free.inc: Nov 04 09:19:29 qt4-x11-free.inc: bumped PR to force recompilation so it uses the fixed mysql Nov 04 09:19:29 needed to get a working qt4-plugin-sqldriver-sqlmysql Nov 04 09:19:40 03Frans Meulenbroeks  07org.openembedded.dev * rd5270e5061 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 09:21:06 eFfeM: have you strace on target? Nov 04 09:21:24 i.e. strace it to look where it is failing Nov 04 09:23:16 Jay7: it fails at writing the pid file, but strace is a good plan looking to see if it changes to a different user (e.g.user mysql) Nov 04 09:23:38 eFfeM: btw, I'm not sure but yes, it can change uid Nov 04 09:23:55 and may be even chroot Nov 04 09:24:24 eFfeM: you can just set 777 mode on /var/run temporary Nov 04 09:24:32 then look who is owner Nov 04 09:28:49 yeah good plan, file is owned my mysql Nov 04 09:29:55 and strace does not seem to log the change of uid Nov 04 09:43:43 eFfeM: create subidr /var/run/mysql/ and point mysqld to create pid-file there :) Nov 04 09:44:13 chown mysql /var/run/mysql before :) Nov 04 09:44:29 Jay7: yeah guess I'll do that, but want to fix the recipe too Nov 04 09:45:03 actually thought i had another issue namely that it is looking for my.cnf in the wrong dir, but still veryifying that Nov 04 09:45:12 morning all Nov 04 09:45:36 hi RP Nov 04 09:47:21 nah /etc/my.cnf is ok, it is after that that it tries to read /var/mysql/my.cnf and ~/my.cnf Nov 04 10:22:31 What's the 'cleanest' way to split up a date in a recipe? i.e. get from 20090101 to 2009.01.01? Any neat 'bitbake' ways before using pythons regex? Nov 04 10:25:57 DJWillis: you can use pythons left/mids/rights implementation instead of regexp Nov 04 10:27:00 XorA: not a python guy, can you think of an example recipe? That would work, just want a really simple one liner to get me what I need for ${MY_PV} ;-) Nov 04 10:30:02 DJWillis: I cant remeber the syntax offhand, not a python guy either, but I do remeber it being simple Nov 04 10:30:27 03Koen Kooi  07org.openembedded.dev * r4dcaca2917 10openembedded.git/classes/ (e.bbclass efl.bbclass): Nov 04 10:30:27 e* bbclass: fix up PACKAGES (again!) Nov 04 10:30:27 This should why overriding PACKAGES is bad Nov 04 10:30:38 03Koen Kooi  07org.openembedded.dev * r8ac85f656c 10openembedded.git/ (11 files in 5 dirs): exalt: update to current form Nov 04 10:31:48 hrw|gone: do you need mesa-xlib with gallium enabled? there are some cross compiling issues, but without gallium it compiles ok.. Nov 04 10:33:47 XorA: thanks, i'll work it out, just having a thick morning ;-) Nov 04 10:34:26 date[0:4] + '.' + date[4:6] + '.' + date[6:8] Nov 04 10:34:33 or something similar Nov 04 10:35:53 hey XorA Nov 04 10:36:43 XorA: are you now around home? :) Nov 04 10:36:57 Jay7: well at work :-D Nov 04 10:37:16 XorA: I'm about collie and tosa Nov 04 10:37:29 yeah Nov 04 10:37:49 should I send new request to ML? Nov 04 10:38:06 you got a way to get them? Nov 04 10:38:28 we should transfer it to NL anyhow :) Nov 04 10:38:56 Ill speak to koen, might be possible to cheaply post them to him in the UK Nov 04 10:39:23 likewise is from NL too according to hostname :) Nov 04 10:39:43 Jay7: unfortuneately I will not be at OEDEM and I dont know if lrg is going yet Nov 04 10:40:52 yup, .nl here Nov 04 10:41:13 good morning Nov 04 10:41:16 gm Nov 04 10:41:28 what needs shipment from where to where? Nov 04 10:41:42 tosa + collie from UK to NL Nov 04 10:41:43 from UK to NL is the first step :) Nov 04 10:42:17 UK, OEDEM to NL, or UK/XorA to NL? Nov 04 10:42:37 hm.. Nov 04 10:42:47 that is question for XorA :) Nov 04 10:43:03 likewise: getting them to OEDEM would be cheapest if you could carry them, but I dont know yet if LRG is attending Nov 04 10:43:17 otherwise I might as well post them direct to .nl Nov 04 10:45:07 well... Nov 04 10:45:18 what is our current decision? :) Nov 04 10:47:49 XorA: will decide tomorrow, depends if I am feeling better Nov 04 10:48:36 likewise: you are at OEDEM and willing to carry? Nov 04 10:48:36 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r138f792724 10openembedded.git/classes/package.bbclass: Nov 04 10:48:36 Revert "Revert "package.bbclass: Cleanup getVar syntax to use True, not 1 as third argument (from Poky)"" Nov 04 10:48:36 Oops, accidently reverted a patch, reverting the revert. Sorry for any inconvenience Nov 04 10:48:36 This reverts commit 056193c6638892ddde7bddbfc6f6f06a2288ba64. Nov 04 10:48:39 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r056193c663 10openembedded.git/classes/package.bbclass: Nov 04 10:48:42 Revert "package.bbclass: Cleanup getVar syntax to use True, not 1 as third argument (from Poky)" Nov 04 10:48:44 This reverts commit fcee557e1b0551d1743bbb3ad61102b9ccc4cba8. Nov 04 10:48:46 03Thomas Kunze  07martin_jansa/xorg-7.5 * r2ec047c78a 10openembedded.git/conf/machine/include/tune-strongarm.inc: Nov 04 10:48:53 tune-strongarm.inc: Nov 04 10:48:55 remove setting of oabi. With gcc 4.4 we can use eabi. Distros can reset this if they want. Nov 04 10:48:57 03Leon Woestenberg  07martin_jansa/xorg-7.5 * r91403056b1 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Nov 04 10:49:00 linux-kirkwood-2.6.31: Add PCI Express support for openrd-base. Nov 04 10:49:02 Note there are upstream problems involving devices using PCI IOMEM. Nov 04 10:49:04 Signed-off-by: Leon Woestenberg Nov 04 10:49:06 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r9fd25648fc 10openembedded.git/ (4 files in 4 dirs): linux-kirkwood: updated defconfigs (more modules, more filesystems in kernel) Nov 04 10:49:14 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * rbdc1df7330 10openembedded.git/recipes/mythtv/mythtv_0.22rc1.bb: mythtv: added staging Nov 04 10:49:14 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r117419bac6 10openembedded.git/recipes/mythplugins/ (8 files in 2 dirs): mythplugins: created recipe, not fully tested yet Nov 04 10:49:23 hrw|gone: pushed Nov 04 10:49:23 03Khem Raj  07martin_jansa/xorg-7.5 * r9458828947 10openembedded.git/recipes/eglibc/ (4 files in 2 dirs): Nov 04 10:49:26 eglibc: Backport IO-acquire-lock-fix.patch Nov 04 10:49:30 * Move the svn recipe to latest git revision. Nov 04 10:49:34 Signed-off-by: Khem Raj Nov 04 10:49:38 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r0c5891b4c9 10openembedded.git/conf/checksums.ini: checksums: added mythplugins checksum Nov 04 10:49:41 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r8b32d20a07 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 10:49:56 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r1377125138 10openembedded.git/recipes/mythplugins/mythplugins_0.21+0.22rc1.bb: mythplugins: added a few lines of comment for clarification Nov 04 10:49:59 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r2d365ab53b 10openembedded.git/recipes/mythtv/ (mythtv-0.21+0.22rc1/configure.patch mythtv_0.22rc1.bb): mythtv: updated staging, fixed configure; renamed recipe Nov 04 10:50:03 03Koen Kooi  07martin_jansa/xorg-7.5 * rf1a20d6191 10openembedded.git/recipes/xorg-lib/ (pixman/neon-24bpp.patch pixman/prefetch.patch pixman_git.bb): pixman git: add more NEON patches, bump SRCREV Nov 04 10:50:06 03Koen Kooi  07martin_jansa/xorg-7.5 * r5ccd6effbe 10openembedded.git/recipes/mythplugins/ (8 files in 2 dirs): mythplugins: move into mythtv directory Nov 04 10:50:11 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r13355c1620 10openembedded.git/ (17 files in 5 dirs): mythtv, mythplugins: updated to 0.22rc2 Nov 04 10:50:18 03David-John Willis  07martin_jansa/xorg-7.5 * r8ad7a9b528 10openembedded.git/ (conf/checksums.ini recipes/xchat/xchat_2.8.6.bb): xchat: Add 2.8.6 release (and fix package to include d-bus service files). Nov 04 10:50:21 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * ra26da59582 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 10:50:25 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r24c8f2b9e0 10openembedded.git/recipes/mythtv/ (5 files in 2 dirs): mythplugins: added more plugins, packaging is not complete yet Nov 04 10:50:28 03David-John Willis  07martin_jansa/xorg-7.5 * r34616c490d 10openembedded.git/recipes/libgtkstylus/ (files/gtkstylus.sh libgtkstylus_0.3.bb): libgtkstylus: Fix session script to support loading of more than one GTK_MODULES. Nov 04 10:50:34 03Holger Hans Peter Freyther  07martin_jansa/xorg-7.5 * r9393ac7170 10openembedded.git/recipes/binutils/binutils-cross-sdk.inc: Nov 04 10:50:37 Revert "binutils-cross-sdk: Avoid QA error on building" Nov 04 10:50:40 This reverts commit b294033116a6a4f33d1b1ec9f8256adaf7374298. This Nov 04 10:50:41 patch is no longer needed as .a files are no longer stripped. Nov 04 10:50:44 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * rd88287ca8a 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 10:50:48 03David-John Willis  07martin_jansa/xorg-7.5 * r648409319f 10openembedded.git/conf/machine/omap3-pandora.conf: omap3-pandora.conf: Line endings. Nov 04 10:50:51 (65 lines omitted) Nov 04 10:53:55 can soeone try to bitbake gs and see if he can fix the errors ? Nov 04 10:56:05 XorA: not at OEDEM, I had to cancel due to new house/office becoming ready. I am not aware of any .nl ppl attending btw. Nov 04 10:56:31 likewise: ok, so posting direct to Jay7 contacts is best plan then Nov 04 10:56:48 XorA: well.. /msg BaT then :) Nov 04 10:57:15 BaT??? Nov 04 10:57:34 yes :) Nov 04 10:57:48 did you know him? Nov 04 10:57:52 no Nov 04 10:58:24 I can ask him to /msg you if you want :) Nov 04 10:58:36 Jay7: can you send emails to introduce us, I really need to be coding at the moment Nov 04 10:58:42 ah, ok Nov 04 11:10:58 hm, i'm doing opkg install gcc and after that I have no binary gcc only /usr/lib/gcc and /usr/libexec/gcc Nov 04 11:11:22 was hoping for a package build-essential but that seems not to be there either Nov 04 11:13:35 eFfeM: with OE? you will need to install the -symlinks packages too Nov 04 11:13:47 zecke ahh Nov 04 11:14:08 thought that would be a package deal :-) Nov 04 11:15:28 hm, we need a collection package or so, after installing gcc-symlinks I get the message that cc1 can't be exec'd Nov 04 11:16:46 * eFfeM is trying to chase down an issue with iconv_open and wants to compile a local test program Nov 04 11:16:49 eFfeM: do an strace... for me it searched in /usr/lib/gcc/amd64---... Nov 04 11:17:08 I'm running Angstrom on Beagle. How/where can I change the options it mounts sd-card partitions with? Nov 04 11:17:22 tasslehoff: /etc/udev/scripts/mountall.sh Nov 04 11:18:51 XorA: mail sent Nov 04 11:18:57 zecke it is looking at a.o. /usr/libexec/gcc/arm-angstrom-linux-gnueabi/4.3.3/cc1, this one does have a cc1plus and g++ tst.c indeed compiles; gcc tst.c does not as cc1 is not there (also not in /usr/lib.gcc/.... Nov 04 11:19:43 03Koen Kooi  07org.openembedded.dev * r2f0ed05940 10openembedded.git/conf/machine/ (10 files in 2 dirs): omap3 machines: don't drag in omap3-sgx-modules by default Nov 04 11:19:54 XorA: can I ask for CE-RH2 too? :) Nov 04 11:20:19 Jay7: as long as you have a c1000/c30x0 device Nov 04 11:20:27 I have akita Nov 04 11:21:35 btw zecke, found the problem I had yesterday building qt4 database drivers and the -L iwthout dir; mysql_config did not return the proper values, it is repaired Nov 04 11:21:53 still messing with qt4 though as now it complains about iconv_open Nov 04 11:23:22 eFfeM: glibc-locales are installed? Nov 04 11:23:28 XorA: Do you mean I should edit it on target? I have mountall.sh in /etc/init.d/mountall.sh. I need to do it in the code so that my image behaves like this by default. Nov 04 11:23:58 tasslehoff: /etc/udev/scripts/mount.sh sorry, and yes edit it on target Nov 04 11:24:19 tasslehoff: when you tested your edit, fiddle with udev recipe Nov 04 11:24:51 XorA: thanks Nov 04 11:26:48 zecke: no, didn't think of that one, will try Nov 04 11:27:08 btw a small test program did work (just doing an open) Nov 04 11:27:23 is there a way to add this in a depends ? Nov 04 11:28:24 pb__: once again I want to benefit from your knowledge... I want to do something like memset but with a 32bit value. I have copied this value to eight registers and then do a 4x stmgeia after each other... can you think of something better? Nov 04 11:28:41 eFfeM: oh, normally i built with -no-iconv to use the built-in unicode tables Nov 04 11:29:18 zecke: i used the std qt4 did not mess with it Nov 04 11:29:52 eFfeM: Qt complains but it will not exit/abort due it Nov 04 11:31:56 zecke ah ok, then the next error is not related, btw the iconv error is because it tries to find things in /lusr/lib/gconv but cannot find them Nov 04 11:32:42 lets see if glibc-gconv improves things Nov 04 11:32:55 nope, it is installed and up to date Nov 04 11:35:48 zecke found it, needed glibc-gconv-utf-16 Nov 04 11:40:22 need advice on how to fix an issue with mysql: in the postinstalll of mysql-server it does a chmod go+rw /var/run Nov 04 11:40:47 however on beagle this is a symlink to /var/volatile/run so it does change the mode of the symlink not of the dir Nov 04 11:41:05 eFfeM: its totally wrong anyway Nov 04 11:41:15 eFfeM: see above you were given the solution Nov 04 11:41:27 XorA, i was getting to that Nov 04 11:41:46 eFfeM: youll need to create a mysql volatiles file in /etc/volatiles Nov 04 11:41:47 want to fix the recipe and make a dir, but that also needs to change my.cnf is that a problem? Nov 04 11:42:05 eFfeM: doing it right is never a problem Nov 04 11:42:21 eFfeM: you should not change access rights on /var/run.. that is bad idea Nov 04 11:42:52 easy way is change pid path Nov 04 11:42:53 Jay7: XorA I agree, note that *I* did not write this bb file, the chmod is there already Nov 04 11:43:07 eFfeM: I know this Nov 04 11:43:12 eFfeM: then you can improve it ;) Nov 04 11:43:26 yeah and change pid path will require change my.cnf and that will break things if people do an upgrade Nov 04 11:43:38 that is why i want some opinion Nov 04 11:43:48 otherwise i would have fixed it already :-) Nov 04 11:43:55 eFfeM: its so badly wrong I think you can safely change it Nov 04 11:44:16 eFfeM: my.cnf should be marked as a config file as well, then opkg will warn on differences Nov 04 11:44:51 XorA, ah ok, Nov 04 11:45:35 last stupid q before lunch :-) where/how do i mark as a config file ? Nov 04 11:45:55 * XorA cant remeber offhand, its a variable you set in .bb Nov 04 11:46:18 ah will look at how samba or so do it Nov 04 11:47:23 CONFILES_${PN} Nov 04 11:47:41 cool, thanks Nov 04 11:51:51 made mysql changes, will test and if ok, commit Nov 04 11:55:46 eFfeM: awesome Nov 04 12:02:52 After doing a change to to mount.sh in recipes/udev/files, what curses do I need to make sure that it rebuilds udev for my image? Nov 04 12:04:12 03Marcin Juszkiewicz  07shr/merge * rea6f6abebd 10openembedded.git/recipes/openvpn/openvpn_2.0.9+2.1rc20.bb: openvpn: added 2.1rc20 and enabled reading passwords from file Nov 04 12:04:13 03Frans Meulenbroeks  07shr/merge * rc206dabdc8 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 12:04:14 03Koen Kooi  07shr/merge * r4b01adc581 10openembedded.git/recipes/gnome-mplayer/gecko-mediaplayer/gecko-mplayer-svn.diff: gecko-media-player: add missing patch Nov 04 12:04:15 03Chris Larson  07shr/merge * r5c4f5bec44 10openembedded.git/recipes/autofs/ (autofs.inc autofs_3.1.7.bb autofs_4.0.0.bb autofs_4.1.4.bb): Nov 04 12:04:17 autofs: switch to a .inc. Nov 04 12:04:19 Signed-off-by: Chris Larson Nov 04 12:04:21 03Frans Meulenbroeks  07shr/merge * r1632c44d94 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 12:04:24 03Marcin Juszkiewicz  07shr/merge * rfa73b45447 10openembedded.git/conf/checksums.ini: checksums.ini: added usbutils and openvpn Nov 04 12:04:27 03Chris Larson  07shr/merge * rf1d0fab3ac 10openembedded.git/recipes/opkg/ (opkg.inc opkg/isatty.patch): Nov 04 12:04:36 opkg: Don't prompt for user input from stdin if it's not a tty. Nov 04 12:04:38 Also fixes a minor memory leak while we're at it, as the code was Nov 04 12:04:40 strdup'ing the malloc'd string read from stdin without ever Nov 04 12:04:42 freeing the original. Nov 04 12:04:44 Signed-off-by: Chris Larson Nov 04 12:04:46 03Koen Kooi  07shr/merge * r37930cdaf9 10openembedded.git/conf/distro/include/sane-srcrevs.inc: efl: bump SRCREV to 43437 Nov 04 12:04:49 03Koen Kooi  07shr/merge * r4964bb59f5 10openembedded.git/recipes/ (e17/expedite_svn.bb efl1/evas.inc efl1/evas_svn.bb): expedite, evas: don't turn of the GL engine forcefully Nov 04 12:04:52 03Koen Kooi  07shr/merge * rf33edf7de7 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Nov 04 12:04:55 opkg: bump SRCREV to r240 Nov 04 12:04:57 * tested on angstrom/beagleboard and angstrom/sheeva by upgrading the system after the DISTRO_PR bump Nov 04 12:05:07 03Frans Meulenbroeks  07shr/merge * r5277f843e6 10openembedded.git/recipes/qt4/qt4-x11-free.inc: Nov 04 12:05:10 qt4-x11-free.inc: bumped PR to force recompilation so it uses the fixed mysql Nov 04 12:05:12 needed to get a working qt4-plugin-sqldriver-sqlmysql Nov 04 12:05:14 03Koen Kooi  07shr/merge * r8ac85f656c 10openembedded.git/ (11 files in 5 dirs): exalt: update to current form Nov 04 12:05:19 03Frans Meulenbroeks  07shr/merge * rd5270e5061 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 12:05:22 03Jeremy Lainé  07shr/merge * rd62c9c5244 10openembedded.git/recipes/ntp/ (ntp-4.2.4p7/ntpdate ntp_4.2.4p7.bb): Nov 04 12:05:25 ntp-4.2.4p7: invoke ntpdate using if-up.d, not an init script Nov 04 12:05:27 * provide an if-up.d script for ntpdate (based on Debian's) Nov 04 12:05:33 * remove invocation of update-rc.d busybox-cron, this should Nov 04 12:05:35 definitely not be part of the ntpdate package Nov 04 12:05:37 03Chris Larson  07shr/merge * rfbccdee191 10openembedded.git/recipes/prelink/ (prelink_20061027.bb prelink_20071009.bb): Nov 04 12:05:37 tasslehoff: -c rebuild Nov 04 12:05:40 (66 lines omitted) Nov 04 12:05:44 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * r7bd8c333e2 10openembedded.git/recipes/mythtv/ (2 files in 2 dirs): mythplugins: fixed remaining staging issues Nov 04 12:05:47 03Frans Meulenbroeks  07martin_jansa/xorg-7.5 * rf0792ef5d7 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 12:05:50 03Jeremy Lainé  07martin_jansa/xorg-7.5 * rb8f658eb91 10openembedded.git/recipes/linux/ (linux-2.6.31/boc01/005-091008-isl12024.patch linux_2.6.31.bb): linux-2.6.31: update boc01 RTC driver for upstream submission Nov 04 12:06:00 03Sebastian Spaeth  07martin_jansa/xorg-7.5 * re972a049e4 10openembedded.git/recipes/connman/connman_0.46.bb: Nov 04 12:06:03 connman: include connman_0.46 (mv'd from connman_0.19) Nov 04 12:06:09 Signed-off-by: Sebastian Spaeth Nov 04 12:06:11 03Sebastian Spaeth  07martin_jansa/xorg-7.5 * rf5701cca05 10openembedded.git/conf/checksums.ini: Nov 04 12:06:15 connman: add checksums for 0.46 Nov 04 12:06:16 Signed-off-by: Sebastian Spaeth Nov 04 12:06:18 03Matthew Dombroski  07martin_jansa/xorg-7.5 * rc510866574 10openembedded.git/classes/qmake2.bbclass: Nov 04 12:06:21 qmake2.bbclass: Add lupdate and lrelease. Nov 04 12:06:23 http://patchwork.openembedded.org/patch/987 Nov 04 12:06:25 Signed-off-by: Matthew Dombroski Nov 04 12:06:27 Acked-by: Holger Hans Peter Freyther Nov 04 12:06:33 03Holger Hans Peter Freyther  07martin_jansa/xorg-7.5 * rc724ad44d6 10openembedded.git/classes/base.bbclass: Nov 04 12:06:36 base.bbclass: Better error message in case the Config Parser fails Nov 04 12:06:42 http://patchwork.openembedded.org/patch/1088/ Nov 04 12:06:44 Tested-by: Ed Nelson Nov 04 12:06:46 03Ricardo Salveti  07martin_jansa/xorg-7.5 * r14caede54c 10openembedded.git/ (conf/checksums.ini recipes/eject/eject_2.1.5.bb): Nov 04 12:06:51 eject-2.1.5: adding HOMEPAGE and fixing SRC_URI Nov 04 12:06:53 As geocities is now down, changing the SRC_URI to point to OE Nov 04 12:06:55 mirror. Nov 04 12:06:57 Signed-off-by: Ricardo Salveti Nov 04 12:06:59 Acked-by: Holger Hans Peter Freyther Nov 04 12:07:01 (69 lines omitted) Nov 04 12:20:43 XorA: I tried "bitbake -c rebuild udev", followed by "bitbake myimage", but that apparently did not do the trick. Nov 04 12:21:26 tasslehoff: not sure why, but sometimes is safer to use -c clean instead of rebuild... Nov 04 12:22:32 JaMa: cleaning udev, and than building a package that uses it should work? or do I have to clean my own package as well? Nov 04 12:26:36 JaMa: it did a lot more after a -c clean, and I think it picked up the new mount.sh. thanks Nov 04 12:32:40 tasslehoff: rebuild just rebuilds (probably starts with do_compile) so new changes are not unpacked from source or patched with new patch Nov 04 12:33:51 I need to really understand how bitbake works. and also to debug it. For example, in http://pastebin.com/m18aed5f , I have an already known error. That I fixed somewhere, but I didn't remember how. It is possible to investigate where is the rootfs the I'm launching ? Nov 04 12:35:13 bitbake -i is a good starting point? Nov 04 12:37:21 recalcati: http://docs.openembedded.org/usermanual/usermanual.html Nov 04 12:39:26 03Koen Kooi  07org.openembedded.dev * r2598835e7f 10openembedded.git/recipes/opencv/ (opencv-samples_svn.bb opencv_svn.bb): opencv: bump SRCREV to just after 2.0.0 release Nov 04 12:39:27 03Koen Kooi  07org.openembedded.dev * re55a3b9423 10openembedded.git/recipes/u-boot/ (3 files in 2 dirs): u-boot: fix beagle fw_env.conf Nov 04 12:45:27 is it possible to have OE use distcc with its self-built toolchain? Nov 04 12:55:59 mike_cw: no, but we have icecream support Nov 04 13:13:04 zecke: are there any docs for that? Nov 04 13:13:34 03Richard Purdie  07rpurdie/work-in-progress * r1fab79ff0b 10openembedded.git/classes/base.bbclass: Nov 04 13:13:34 base.bbclass: Correct autotools reference Nov 04 13:13:34 Signed-off-by: Richard Purdie Nov 04 13:22:01 03Koen Kooi  07org.openembedded.dev * r0f926962a2 10openembedded.git/recipes/gnome/ (3 files): gnome-themes: delete obsolete versions Nov 04 13:34:24 zecke: no, that's about as good as it gets. Nov 04 13:35:05 zecke: the exact tradeoffs for maximum performance are different (and slightly complex) for different cpus, and also depend on what type of memory region you are writing to. Nov 04 13:35:47 for example, if you are writing to main memory then that's going to be the bottleneck and the instruction mix is largely irrelevant. Nov 04 13:54:31 03Sebastian Spaeth  07shr/merge * refa02ae309 10openembedded.git/recipes/connman/ (connman_0.46.bb files/no_system_user_perms.patch): Nov 04 13:54:31 connman: don't give user 'system' permissions by default. We might not have one. (connman 0.46) Nov 04 13:54:31 Signed-off-by: Sebastian Spaeth Nov 04 13:56:10 03Sebastian Spaeth  07shr/merge * rb604f12f33 10openembedded.git/recipes/connman/connman_0.46.bb: Nov 04 13:56:10 connman 0.46: bump PR Nov 04 13:56:10 Signed-off-by: Sebastian Spaeth Nov 04 14:03:41 zecke: Any objection if I cherrypick some of the parser rework branch into 1.10 and master? Nov 04 14:04:12 zecke: get the "fixes"/reorg out of the way and merged, then maybe we can rebase the branch? Nov 04 14:18:12 03Martin Jansa  07org.openembedded.dev * r8fedaafb39 10openembedded.git/recipes/eglibc/ (4 files in 2 dirs): Nov 04 14:18:12 eglibc: updated armv4t-interworking.patch Nov 04 14:18:12 * Updated patch to make SYSCALL_ERROR_HANDLER interworking safe. Nov 04 14:18:12 * Phil Blundell found and fixed another issue in eglibc armv4t builds. Nov 04 14:18:12 Signed-off-by: Martin.Jansa Nov 04 14:37:47 03Koen Kooi  07shr/merge * r2fcc81d7ce 10openembedded.git/recipes/opencv/ (opencv-samples_svn.bb opencv_svn.bb): opencv: bump SRCREV to just after 2.0.0 release Nov 04 14:37:48 03Martin Jansa  07shr/merge * r209b308dee 10openembedded.git/recipes/eglibc/ (2 files in 2 dirs): Nov 04 14:37:48 eglibc: sync eglibc_2.10.bb with version from oe.dev Nov 04 14:37:48 Signed-off-by: Martin Jansa Nov 04 14:37:50 03Koen Kooi  07shr/merge * rff1b2a5e02 10openembedded.git/recipes/gnome/ (3 files): gnome-themes: delete obsolete versions Nov 04 14:37:53 03Koen Kooi  07shr/merge * re0581f0d59 10openembedded.git/recipes/u-boot/ (3 files in 2 dirs): u-boot: fix beagle fw_env.conf Nov 04 14:37:56 03Sebastian Krzyszkowiak  07shr/merge * r1777ecc500 10openembedded.git/recipes/tangogps/tangogps_0.9.8.bb: Nov 04 14:38:03 tangogps: add recipe for new version 0.9.8 Nov 04 14:38:07 Signed-off-by: Martin Jansa Nov 04 14:38:07 03Martin Jansa  07shr/merge * rb134e6348c 10openembedded.git/conf/checksums.ini: Nov 04 14:38:07 checksums.ini: add PackageKit-0.5.4 and tangogps-0.9.8 Nov 04 14:38:09 Signed-off-by: Martin Jansa Nov 04 14:38:12 03Martin Jansa  07shr/merge * raa2a7cc7c6 10openembedded.git/recipes/eglibc/ (eglibc_2.9.bb eglibc_svn.bb files/armv4t-interworking.patch): Nov 04 14:38:15 eglibc: updated armv4t-interworking.patch Nov 04 14:38:17 * Updated patch to make SYSCALL_ERROR_HANDLER interworking safe. Nov 04 14:38:19 * Phil Blundell found and fixed another issue in eglibc armv4t builds. Nov 04 14:38:21 Signed-off-by: Martin.Jansa Nov 04 14:52:57 JaMa: I can't see any obvious interworking problems in eglibc with your patch applied. Do you have a register dump corresponding to your python crash? Nov 04 14:53:23 lunchtime now but I will look at the crash a bit later if you send me the details Nov 04 14:53:43 XorA: got sidetracked but i'm now back to mysql, I added CONFFILES_${PN} += "${sysconfdir}/my.cnf" to the recipe, but the build fails with: *** Error: CONTROL/conffiles mentions conffile /etc/my.cnf which does not exist; although I have image/etc/my.cnf and install/mysql-server/etc/my.cnf Nov 04 14:53:47 any idea ? Nov 04 14:54:26 (or anyone else ofc :-) Nov 04 14:55:03 eFfeM: _${PN} != mysql-server ? Nov 04 14:55:52 RP, ah ofc, no, PN = mysql Nov 04 14:56:18 got it, ty Nov 04 14:57:07 RP solves problems quicker than I can read :-D Nov 04 14:57:12 rofl Nov 04 14:57:39 btw do we have someone who feels responsible for the mysql recipe ? Nov 04 14:58:43 tha's a good Q to make people silent :-) Nov 04 15:01:28 ok, let me coin my Q: for mythtv I need mysql5, I can work on creating a recipe, np, but I am not a mysql wiz so no idea about the compatibility and whether other pacakges require mysql4. So the Q is do I create a recipe for mysql 5 in the mysql dir or do i create a new dir mysql5 (like done for qt4) so both can coexist if desired) Nov 04 15:02:00 eFfeM: Why can't they coexist in the mysql dir? Nov 04 15:02:55 RP, they can but as I said I am not an sql wiz, so I cannot really guarantee that everything that works with 4 will work with 5 Nov 04 15:03:09 and as several packages probably depend on mysql Nov 04 15:03:24 don't want to break things, but can't really be sure I don't Nov 04 15:03:44 eFfeM: You;re not planning on changing mysql4 though, just adding 5? Nov 04 15:03:47 eFfeM: that doesn't seem to have anything to do with what directory the recipes should be in. Nov 04 15:03:48 mysql5_1.2.3.bb in the mysql/ directory then Nov 04 15:04:13 if you are talking about making mysql5 be a new version of mysql.bb, rather than a separate PN, that would another question Nov 04 15:04:29 ok, great, as a starter I might push it with DEFAULT_PREFERENCE -1 Nov 04 15:04:44 gah, ccache finally broke for me :) Nov 04 15:05:46 pb__ i thought indeed of making it a new version of mysql.bb: mysql_4.1.22.bb mysql_5.1.40.bb Nov 04 15:07:01 btw also not too sure why we have mysql-native Nov 04 15:07:26 pb__ would my 5.1.40 recipe cause a problem or break anything ? Nov 04 15:07:55 eFfeM: if it were me, I would make it be a separate recipe (named "mysql5" or something) and arrange for both it and mysql4 to provide some suitable virtual. Nov 04 15:08:43 I doubt that just calling mysql_5.1.40.bb as you propose would do any real harm, but mysql4 and mysql5 are sufficiently different that maybe it is worth taking some extra care. Nov 04 15:08:47 pb__ understood, will consider, can't judge the complete impact Nov 04 15:09:34 pb__ and agree with taking care, especially since I'm not an sql wiz Nov 04 15:11:02 there definitely are some user-visible compatibility issues when going from 4 to 5. Nov 04 15:11:17 most notably, any ISAM tables will be completely inaccessible after upgrading because 5 doesn't support that storage engine at all. Nov 04 15:11:32 03Richard Purdie  07rpurdie/work-in-progress * r1905d572a9 10openembedded.git/classes/native.bbclass: Nov 04 15:11:32 native.bbclass: Fix for packages using autotools_stage_all directly Nov 04 15:11:32 Signed-off-by: Richard Purdie Nov 04 15:11:41 and there are various SQL dialect changes as well that might break some applications. Nov 04 15:11:44 ah ok, good point Nov 04 15:12:16 i'll name it mysql5, store in mysql dir, and when done and committed mail to the list Nov 04 15:12:27 good-o Nov 04 15:12:37 will get back on the virtual, that is uncharted territory for me, but will do as i get there Nov 04 15:13:22 eFfeM: The virtual can be fitted to it later Nov 04 15:13:29 agree Nov 04 15:14:03 actually was alittle bit surprised mysql5 was not there yet Nov 04 15:15:15 * kergoth yawns Nov 04 15:16:53 morning kergoth Nov 04 15:17:30 pb__: wasnt register dump in gdb.txt? mmt check Nov 04 15:17:52 pb__: http://jama.homelinux.org/org.openembedded.shr/python/gdb.txt Nov 04 15:20:58 hey pb__, RP, others Nov 04 15:21:54 Hi all, Nov 04 15:22:44 If I understand corectly, the qte SDK is only available in the .dev branch (not in stable ?) Nov 04 15:24:36 hi kergoth Nov 04 15:26:31 is there a -dev i can install that includes all required dev files, i just installed g++ , libstdc++-dev and tried a hello wolrd but there appears to be alot of include files still missing Nov 04 15:27:16 mango_make: no Nov 04 15:28:21 :( Nov 04 15:30:16 mango_make: if you want it learn oe and bitbake and create a package similar to debian build-essential Nov 04 15:34:22 03Thomas Kunze  07martin_jansa/xorg-7.5 * rf67339544b 10openembedded.git/conf/machine/include/tune-strongarm.inc: tune-strongarm.inc: add EXTRA_FEED_ARCHS Nov 04 15:34:28 I have done some debug, but rootfs fails anyway: http://pastebin.com/m2b87c4a6 . ipk opn problem. I'm in stable/2009 branch Nov 04 15:35:47 03Martin.Jansa  07shr/merge * r1b8747f37e 10openembedded.git/ (33 files in 7 dirs): Nov 04 15:35:47 xorg: update recipes of other freedesktop projects Nov 04 15:35:47 * Update versions to latest available at Nov 04 15:35:47 http://xorg.freedesktop.org/releases/individual Nov 04 15:35:47 when X11R7.5 was released Nov 04 15:35:48 03Martin.Jansa  07shr/merge * rdd4c41a775 10openembedded.git/conf/distro/include/ (2 files): Nov 04 15:35:53 xorg: add preferred versions for projects not included in X11R7.5 Nov 04 15:35:55 * Versions are defined in preferred-xorg-versions-X11R7.5-latest.inc Nov 04 15:35:57 * Included in preferred-xorg-versions-X11R7.5.inc Nov 04 15:35:59 03Martin.Jansa  07shr/merge * rd357db7cf9 10openembedded.git/recipes/xorg-xserver/xorg-xserver-common.inc: Nov 04 15:36:02 xorg-xserver-common: drop lbxutil dependency, add dri2 package Nov 04 15:36:04 * Do we need lbxutil for older xorg-xserver? Nov 04 15:36:06 03Martin.Jansa  07shr/merge * re00b3df48c 10openembedded.git/ (229 files in 13 dirs): Nov 04 15:36:09 xorg: add recipes for X11R7.5 release Nov 04 15:36:11 * Update conf/distro/include/preferred-xorg-versions-X11R7.5.inc Nov 04 15:36:13 * Only recipe versions specified in X11R7.5 and available at Nov 04 15:36:15 http://www.x.org/releases/X11R7.5/src/ Nov 04 15:36:17 03Martin.Jansa  07shr/merge * r987a56d5d2 10openembedded.git/recipes/xorg-driver/xf86-video-glamo_git.bb: xf86-video-glamo: enable KMS as kms branch is now merged to master Nov 04 15:36:28 03Martin.Jansa  07shr/merge * r5cc05e8f6f 10openembedded.git/recipes/xorg-lib/libxaw_1.0.7.bb: Nov 04 15:36:31 libxaw: disable docs building as workaround Nov 04 15:36:33 * It tries to use groff from buildhost and there is no groff-native Nov 04 15:36:35 * If groff on host is new enough it works ok, but fails with older grof Nov 04 15:36:37 * We can create groff-native and enable docs again Nov 04 15:36:41 03Martin.Jansa  07shr/merge * r50e12cd830 10openembedded.git/recipes/xorg-lib/libxfont_1.4.1.bb: libxfont: disable docs as workaround for problems on some buildhosts Nov 04 15:36:42 dpkg-deb -I /app/opt/sources/OE_OMAP/tmp/deploy/glibc/ipk/all/alsa-state-dbg_0.2.0-r7.1_all.ipk correctly open the file Nov 04 15:36:44 03Martin.Jansa  07shr/merge * r6e7101764a 10openembedded.git/contrib/source-checker/ (2 files): (log message trimmed) Nov 04 15:36:47 contrib: scripts for easy preferred-xorg-versions generation Nov 04 15:36:49 * live.sh generates conf/distro/include/preferred-xorg-versions-live.inc with Nov 04 15:36:55 latest versions available at http://xorg.freedesktop.org/releases/individual/ Nov 04 15:36:57 * 7.5.sh generates preferred-xorg-versions-X11R7.5-live.inc with latest Nov 04 15:36:59 versions available at http://www.x.org/releases/X11R7.5/src/ Nov 04 15:37:01 * Temporary files are created in script directory and contains info Nov 04 15:37:03 03Martin.Jansa  07shr/merge * r21eb97a87a 10openembedded.git/recipes/xorg-xserver/xserver-xorg-conf/om-gta02/xorg.conf: om-gta02: xorg.conf with evdev drivers for input Nov 04 15:37:08 03Martin.Jansa  07shr/merge * rdd568d575e 10openembedded.git/recipes/xorg-app/ (xdriinfo_1.0.1.bb xdriinfo_1.0.2.bb): xdriinfo: depend on virtual/libgl instead of mesa directly Nov 04 15:37:11 03Khem Raj  07shr/merge * r017c8c5568 10openembedded.git/recipes/eglibc/ (4 files in 2 dirs): Nov 04 15:37:14 eglibc: Backport IO-acquire-lock-fix.patch Nov 04 15:37:18 * Move the svn recipe to latest git revision. Nov 04 15:37:20 Signed-off-by: Khem Raj Nov 04 15:37:22 03Martin.Jansa  07shr/merge * r82981db13f 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom: use latest glproto, libx11, mesa, util-macros and xserver-xorg Nov 04 15:37:25 03Thomas Kunze  07shr/merge * r868551cb53 10openembedded.git/conf/machine/include/tune-strongarm.inc: Nov 04 15:37:28 (19 lines omitted) Nov 04 15:40:12 03Frans Meulenbroeks  07org.openembedded.dev * rcea3d55fd2 10openembedded.git/recipes/mythtv/mythtv_0.21+0.22rc2.bb: Nov 04 15:40:12 mythtv: added dependency to glibc-gconv-utf-16 Nov 04 15:40:12 still has DEFAULT_PREFERENCE = "-1" as this needs mysql5 Nov 04 15:40:12 03Frans Meulenbroeks  07org.openembedded.dev * rbfb8287d87 10openembedded.git/recipes/mysql/ (files/my.cnf mysql_4.1.22.bb): Nov 04 15:40:14 mysql: changed postinstall: pid file goes to /var/run/mysql i.s.o /var/run as mysqld cannot write to /var/run Nov 04 15:40:16 this alo changes /etc/my.cnf Nov 04 15:40:18 03Frans Meulenbroeks  07org.openembedded.dev * r82721f8688 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 15:40:21 03Frans Meulenbroeks  07org.openembedded.dev * r945c5c95b3 10openembedded.git/recipes/mythtv/mythplugins_0.21+0.22rc2.bb: mythplugins: add zoneminder package Nov 04 15:40:37 eFfeM: really strange, but I remember I already had a simila pb. Nov 04 15:41:26 RP, all for mysql I'll use mysql5_5.1.40.bb, that way at least the recipe can remain to use PV Nov 04 15:41:38 eFfeM: /app/opt/sources/OE_OMAP/tmp/deploy/glibc/ipk/all/alsa-state-dbg_0.2.0-r7.1_all.ipk is good, but inside rootfs I have some pb to open it. Nov 04 15:43:12 recalcati: i do not understand your problem how do you open it Nov 04 15:43:36 I have done some debug, but rootfs fails anyway: http://pastebin.com/m2b87c4a6 . ipk open problem. I'm in stable/2009 branch Nov 04 15:44:41 recalcati: i've seen the pastebin, but I have no idea what you are doing or what command you executed and I am not a python or opkg wiz Nov 04 15:45:06 bitbake bootstrap-image Nov 04 15:45:37 and /app/opt/sources/OE_OMAP/tmp/deploy/glibc/ipk/all/alsa-state-dbg_0.2.0-r7.1_all.ipk is good, dpkg-deb -I open it correctly . Nov 04 15:45:56 thx anyway Nov 04 15:46:34 sry, can't help you with this, maybe someone else can Nov 04 15:47:34 no pb, I'll fix it, now I do some diff with dev branch Nov 04 15:49:21 JaMa: hrm, seems to be libpthread again. could you put a copy of your libpthread.so.0 (unstripped) there as well? Nov 04 15:51:10 pb__: sure... it will take some time because i have inherited rm_work, so need to rebuild Nov 04 15:51:16 ah Nov 04 15:51:24 if you need to rebuild, please capture a new register dump as well Nov 04 15:51:36 it is fairly important that all the data comes from the same set of binaries, otherwise the addresses may not match up Nov 04 15:51:58 if you have the stripped version of libpthread.so then that might be enough in the meantime Nov 04 15:51:58 yaay! C++ helloworld built on z :) Nov 04 15:54:04 pb__: ok its there Nov 04 15:54:41 pb__: btw libpthread from -dbg would be better? or is it unstripped for sure there? Nov 04 15:55:30 morning Nov 04 16:02:33 JaMa: could you also put your libreadline there? Nov 04 16:02:38 stripped is probably ok Nov 04 16:03:30 pb__: its libreadline.so.4 or so.5? Nov 04 16:03:38 imho there was so.4 Nov 04 16:04:00 .4, it seems Nov 04 16:04:27 both there.. Nov 04 16:04:32 tahnks Nov 04 16:04:33 thanks, too Nov 04 16:09:17 pb__: eglibc-2.10-r7.5/build-arm-oe-linux-gnueabi/nptl/libpthread.so.0 is there as libpthread-2.10.1.so.unstripped Nov 04 16:12:04 pb__, RP: either of you guys know if anyone has a script to create an initial recipe from an srpm/spec file? obviously it would need serious manual fixups, but i could see such a thing being useful as a starting point.. bitbake used to have basic code for that ages ago, but i think a separate script would be better Nov 04 16:12:13 kergoth: not that I know of Nov 04 16:12:52 hmmm, k Nov 04 16:13:07 JaMa: hm, this one is more of a pain. might need debugging interactively. Nov 04 16:13:14 do you have an image I could flash on my gta01 conveniently? Nov 04 16:13:28 tools to help people transition away from other tools to OE or mvl6 could be useful Nov 04 16:13:32 * kergoth googles Nov 04 16:13:44 I do have some other v4t hardware here but it isn't entirely straightforward for me to run up the shr bits on there. Nov 04 16:14:16 kergoth: that seems like it would pretty straightfoward to write such a script Nov 04 16:14:35 yeah, most of the metadata maps 1-to-1 Nov 04 16:14:37 pb__: yes.. wait few minutes.. after eglibc build finishes there should be new image.. Nov 04 16:14:44 just hte emitted recipe wouldn't properly take advantage of our bbclasses Nov 04 16:14:57 unless you analyze the commands being executed in the steps Nov 04 16:15:05 pb__: do you need some extra kernel config for gta01, hmm I guess so.. so a bit longer then.. Nov 04 16:15:53 hm, dunno. mickey|office, any idea? Nov 04 16:16:32 I'll try to find my gta01 in the meantime :-} Nov 04 16:16:59 pb__: or I can give you ssh access to my neo.. Nov 04 16:17:16 oh, that would be easier Nov 04 16:17:31 ? Nov 04 16:17:31 mmt then Nov 04 16:17:39 (just joined) Nov 04 16:19:17 mickey|office: ah, sorry. I was going to ask, do you know what JaMa needs to do to build a shr image that will run on my gta01. but apparently we don't need that anymore, so don't worry about it. Nov 04 16:19:59 pb__: ah, ok. adjusting the MACHINE plus chosing the proper (small) image should do Nov 04 16:21:23 * XorA really should do something with his gta03 one day for a laugh Nov 04 16:22:40 mickeyl: thanks Nov 04 16:22:52 mickey|office: thanks to you too Nov 04 16:22:56 :) Nov 04 16:23:17 XorA: is usbgadget done for the 6410 now? Nov 04 16:23:25 if so i might revive mine as well Nov 04 16:23:37 although i rather work with devices than with development boards Nov 04 16:25:06 mickey|office: Ive only got the 2442 based gta03 Nov 04 16:26:05 ah Nov 04 16:26:37 I should ebay it really and see if there any stupid OM adicts still around Nov 04 16:27:16 heh Nov 04 16:27:19 good plan Nov 04 16:27:33 * mickey|office digs into his cupboard and looks for "collector's items" Nov 04 16:27:51 gta01 (no B)... Nov 04 16:28:05 R4R3 Openmoko GTA03 NEO FREERUNNER IPHONE ANDROID (as ebay adverts seem to go these days) Nov 04 16:28:28 mickey|office: I think broonie still has a gta01 as well Nov 04 16:29:17 anyway is time I escaped!!!! Nov 04 16:31:42 gm Nov 04 16:32:31 my google skills are failing me! what is the conf file setting to have bitbake delete the source files after a build? Nov 04 16:33:42 sakoman_: INHERIT += "rm_work" Nov 04 16:33:55 kergoth: I don't think it exists but would be interesting to have Nov 04 16:34:13 mickey|office: I found a nice collector item here ... a SGI O2 Nov 04 16:34:21 * RP has a GTA01 somewhere Nov 04 16:34:52 RP: thanks! Nov 04 16:34:55 RP: don't you also have an A780 that you didn't use as advertised? :) Nov 04 16:34:59 chouimat: nice :) Nov 04 16:35:15 can't run linux on it :( but netbsd works fine Nov 04 16:35:33 mickey|office: An A780? I don't even know what one is? :) Nov 04 16:36:12 RP: hmm, really? d'oh, mixed that up in my mind. i thought we got you a MOTO A780 in order to get some PXA kernel goodness Nov 04 16:36:28 mickey|office: No, I never had one of them Nov 04 16:36:57 morning Nov 04 16:36:58 hmm, k Nov 04 16:37:00 hi hrw Nov 04 16:37:20 cool found my blackdog and my yoggie opensoho Nov 04 16:37:24 hi hrw Nov 04 16:37:47 mickey|office: Did you have any thoughts on the import bb problem? Nov 04 16:37:58 XorA|gone: two transmission changes for you Nov 04 16:38:07 03Marcin Juszkiewicz  07org.openembedded.dev * r9f162c68f7 10openembedded.git/recipes/transmission/transmission.inc: Nov 04 16:38:07 transmission: improve packaging Nov 04 16:38:07 transmission package contains GUI client Nov 04 16:38:07 transmission-client contains CLI and Remote clients Nov 04 16:38:07 transmission-daemon contains daemon and Web interface Nov 04 16:38:08 mickey|office: I'm thinking we need to change the use of __builtins__ in methodpool Nov 04 16:38:17 03Marcin Juszkiewicz  07org.openembedded.dev * r266aa909d6 10openembedded.git/ (conf/checksums.ini recipes/transmission/transmission_1.76.bb): transmission: added 1.76 Nov 04 16:38:42 mickey|office: But I pushed the __builtins__ version into bitbake 1.8 so we can start to ditch all the "import bb" statements that are around Nov 04 16:40:25 didn't have time to play more with it Nov 04 16:40:40 ok, we can always change it to something less intrusive if we find a method Nov 04 16:40:46 meh, we sure have a lot of old ugly remants from old ways of doing things hanging around Nov 04 16:41:19 kergoth: I'm hoping to have a spring clean in poky :) Nov 04 16:41:27 thatd be nice Nov 04 16:41:38 kergoth: Which stuff just caught your attention? Nov 04 16:42:06 kergoth: Mind if I cherrypick bits of the parserwork branch into 1.10 and rebase it btw? Nov 04 16:42:10 in this particular case, the image stuff. IMAGE_CMD_... ugh.. why didn't we just create nice little shell functions that took the path, etc as arguments? Nov 04 16:42:12 go for it Nov 04 16:42:14 sounds good to me Nov 04 16:42:46 kergoth: I'm basically trying to put together something in the 1.10 branch that looks more like master Nov 04 16:43:02 * kergoth nods Nov 04 16:43:23 hopefully we can work out any remaining kinks with master and get that out shortly after 1.10, get it more exposure Nov 04 16:43:34 maybe do an rc of it as soon as we can Nov 04 16:43:40 kergoth: I see 1.10 being a single thread version of master Nov 04 16:44:16 I'm toying with the idea of tossing xmlrpc and using dbus too. That needs some thought :/ Nov 04 16:44:32 hmm Nov 04 16:45:00 I'm not sure yet though, just thinking Nov 04 16:46:22 We don't half have some junk in old recipes :/ Nov 04 16:47:32 xmlrpc? Nov 04 16:47:49 distributed IPC w/ dbus is horror Nov 04 16:48:08 mickey|office: I was talking to ross about this. There could be ways around it Nov 04 16:48:28 anyone an idea how to bypass a configure check for "checking whether pthread_t can be used by GCC atomic builtins" ? Nov 04 16:48:34 you don't seem to like xmlrpc Nov 04 16:48:35 why not? Nov 04 16:48:57 mickey|office: Anyone who has tried interfacing to bitbake through it has run off screaming :( Nov 04 16:49:06 outside of python anyway Nov 04 16:49:46 well Nov 04 16:49:57 that may not be the fault of the communication protocol Nov 04 16:50:06 dbus from C is to run away and scream as well Nov 04 16:50:25 DIPCs have a certain complexity, so you need good libraries Nov 04 16:50:33 (or programming languages) Nov 04 16:50:41 mickey|office: Right, I know that. I'd just like to at least check something like dbus wouldn't work better Nov 04 16:51:10 mickey|office: I'm not saying we should switch, just that I think some checks wouldn't hurt Nov 04 16:52:58 would be good to get a good discussion / brainstorming session going about possibilities for future directions for bitbake and the oe core classes in particular, and perhaps sit down and lay out real use cases for our userbase to see how we can meet those needs better Nov 04 16:53:23 kergoth: I'd love to have that discussion Nov 04 16:53:36 kergoth: Could do with you there for that one though :) Nov 04 16:54:22 * kergoth needs to email his boss about trying to get to oedem next year Nov 04 16:54:55 Volunteer to host it in your office! Nov 04 16:55:11 broonie: US? nack Nov 04 16:55:13 hehe Nov 04 16:55:40 hrw: Given the limited attendance in .uk... :) Nov 04 16:56:08 broonie: attendance in the UK isn't looking so bad... Nov 04 16:56:52 Including all the dropouts? Nov 04 16:57:12 broonie: Its just Liam and XorA so far isn't it? Nov 04 16:57:27 Everyone always talks crap about the US, yet everybody seemd to make it to Portland just fine Nov 04 16:58:19 CosmicPenguin: If I didn't work for a company with offices nearby there I can assure you I'd never have made the US yet Nov 04 16:58:20 I do not want to go though 'get US visa or not get US visa' procedure Nov 04 16:58:34 RP: I thought there were more people (and a bunch that didn't come in the first place). Nov 04 16:58:56 Come to boulder - we have beer and rocks and snow! Nov 04 16:58:59 CosmicPenguin: I'm not sure I'd want to go there for non-work purposes. Nov 04 17:03:09 broonie: There are some noticable absences, particularly kergoth and zecke but I doubt the location influenced that too much Nov 04 17:03:18 and Koen Nov 04 17:09:57 RP: I agree, location is good. I'm just feeling ill atm :( Will decide tomorrow. Nov 04 17:10:17 lrg: :( Nov 04 17:10:29 lrg: something is going around atm :( Nov 04 17:11:01 yeah, I know lots of folks who have been ill lately :( Nov 04 17:16:29 hmm Nov 04 17:19:48 flu ? Nov 04 17:19:57 Plague. Nov 04 17:26:06 I think there have been four oedem drop-outs in total, plus two last-minute registrations. Nov 04 17:26:20 so we're two attendees down on balance which is not too bad. Nov 04 17:33:09 who else is not coming? Nov 04 17:33:19 (apart from [perhaps] Liam and XorA) Nov 04 17:33:36 girdwood and gregory are diseased, gardich is going to a cardiac pacing session (!), and smith has a conflicting work trip Nov 04 17:34:56 sounds like lrg might still decide to come after all though Nov 04 17:35:55 heh Nov 04 17:36:02 ok, doesn't sound too bad Nov 04 17:36:44 yeah, I think that's fine. some level of cancellations is always inevitable, you can't expect everyone to drop all other activities. Nov 04 17:37:17 *nod* Nov 04 17:37:39 as far as locations go, I can't really think of anywhere else that would have made it easier for most folks to attend. Nov 04 17:38:13 sure Nov 04 17:38:13 if we wanted zecke to come then we would probably have to hold oedem in taipei, which might be fun but would probably rule out lots of other people :-} Nov 04 17:38:17 hehe Nov 04 17:38:27 pb__: will see how I feel tomorrow, sir ! ;) Nov 04 17:38:35 lrg: righto Nov 04 17:38:40 get well soon Nov 04 17:38:52 as much as i love berlin, it's good to have multiple locations to distribute the workload of organizing Nov 04 17:39:28 yeah. Nov 04 17:39:58 next time, maybe lrg and rp could invite us all to the frozen north. Nov 04 17:41:08 hmm Nov 04 17:41:16 don't we have someone on the bermudas? Nov 04 17:41:18 or hawaii? Nov 04 17:41:22 for me Dublin or London airports are perfect + everything in ~1h from there Nov 04 17:41:28 heh Nov 04 17:41:34 can't think of anyone in those locations :-} Nov 04 17:44:37 There's direct flights from .pl to EDI Nov 04 17:45:49 broonie: .pl is big country Nov 04 17:46:15 hrw: I know, but it might be from somewhere convenient for you - the flights started fairly recently. Nov 04 17:47:57 broonie: SSZ -> STN,Oslo,Dublin Nov 04 17:48:09 for other flights I need to go to Berlin Nov 04 17:48:09 also Milan (BGY) have a plenty of no-frills airlines Nov 04 17:56:49 hmm, i kind of like ptxdist's project based setup Nov 04 17:58:19 pb__: newcastle or edinburgh would make nice venues but the transport links aren't as good as London/Cambridge or Berlin... Nov 04 17:59:10 kergoth: I have never tried this but the idea sounds good. Current OE setup procedure with environment variables and such things is not what i call a good usability Nov 04 17:59:19 florian: indeed Nov 04 17:59:25 flexible, but not easy Nov 04 17:59:41 RP: mm, true. how long does it take on the train from london to newcastle? Nov 04 17:59:54 pb__: 3-3.5 hours Nov 04 18:00:00 a project (==single directory) based configuration would be better Nov 04 18:00:14 in ptxdist you ./configure; make; make install their pxtdist script, then you run ptxdist menuconfig to configure it, and ptxdist go to build it Nov 04 18:00:17 many people like rwhitby's makefiles. although i don't agree with mapping each and every bitbake command to make targets, i think they have a use in setup though. Nov 04 18:00:34 There is no reason we shouldn't have a wrapper around bitbake to handle these kinds of things. Its kind of how poky has approached it Nov 04 18:01:26 That's true, would be a valid solution for OE too. Nov 04 18:02:31 * kergoth ponders Nov 04 18:03:00 laibsch did some great things with his bootstrap virtual package as well Nov 04 18:03:23 ah, how does that work? Nov 04 18:03:31 just deps on all our requirements? Nov 04 18:03:36 apt-get install openembedded-essential? Nov 04 18:03:41 kergoth: How about defining a file which turns that directory into something when bitbake parses it Nov 04 18:04:17 i don't know the details, but iirc it was a combination of dependencies and some clever postinstsa Nov 04 18:04:53 Something like PROJPATH which takes a collection of directories and then BBPATH and BBFILES get set automagically from that Nov 04 18:05:11 along with collections variables, prorities and so on Nov 04 18:06:09 You can then distribute "project" packs which you'd just decompress and they'd expand the metadaya Nov 04 18:06:14 metadata Nov 04 18:07:16 bye Nov 04 18:07:25 is ASSUME_PROVIDED += "virtual/kernel" correct to stop a kernel recipe from being built? it seems to still pull in every kernel under the sun to satisfy module dependencies Nov 04 18:07:38 i've thought about having a file that sits in the root of the project which bitbake knows how to search upward and find, which defines what collections of metadata are to be used.. a sort of bootstrap config file and runs prior to bitbake.conf, to tell it how to find bitbake.conf Nov 04 18:07:49 * kergoth shrugs, lots of possibilities Nov 04 18:08:03 kergoth: Thats kindfof what I'm describing badly :) Nov 04 18:08:09 ah, right :) Nov 04 18:12:10 kergoth: yes that might be even better Nov 04 18:24:33 03Koen Kooi  07org.openembedded.dev * r733d54f07b 10openembedded.git/recipes/mythtv/mythtv_0.21+0.22rc2.bb: mythtv 0.22rcX: only add glibc-gconv stuff for glibc build Nov 04 18:27:30 RP: http://projects.linuxtogo.org/projects/zaurus/ Nov 04 18:27:37 RP: join us :) Nov 04 18:30:08 we are looking for active zaurus (former) developers :) Nov 04 18:34:38 03Khem Raj  07org.openembedded.dev * ra0afb38137 10openembedded.git/recipes/eglibc/eglibc_svn.bb: Nov 04 18:34:38 eglibc_svn.bb: Update to latest svn tip. Nov 04 18:34:38 Signed-off-by: Khem Raj Nov 04 18:34:46 03Khem Raj  07org.openembedded.dev * rd26c8da511 10openembedded.git/recipes/uclibc/ (uclibc-nptl/uClibc.distro uclibc_nptl.bb): Nov 04 18:34:46 uclibc-nptl: Update to latest tip of branch Nov 04 18:34:46 * Compile uclibc nptl is ARM mode. Thumb has some issues for now. Nov 04 18:34:46 * Add new config option UCLIBC_SUSV4_LEGACY=y Nov 04 18:34:48 Signed-off-by: Khem Raj Nov 04 18:41:41 kergoth: I used to use ptxdist Nov 04 18:42:30 kergoth: my goal would be to merge the best of ptxdist into oe... and more Nov 04 18:44:06 mckoan: aye.. i have a lot of things i'd like to see happen, but i haven't quite gotten a handle on how best to make it happen... Nov 04 18:44:10 heh Nov 04 18:44:48 i'd still like to see good support for project areas with directly available sources rather than using the fetch/unpack/patch idiom, but haven't quite worked out the best way to do that Nov 04 18:47:44 I'm implementing support for splitting up a filesystem into separate images, as partitions, so you can make, say, an ext2 /usr and a jffs2 /. so the size limitations will be per partition.. i'm thinking PARTITIONS = "usr base" .. PARTITION_base_PATH = "/", PARTITION_base_FS = "jffs2", PARTITION_base_SIZE = "+50M".. that seem reasonable to people? Nov 04 18:50:25 kergoth: looks like a difficult goal Nov 04 18:51:38 kergoth: BTW looks like your company alreadi achieved many of my ideas ;-) Nov 04 18:51:40 which? :) Nov 04 18:52:17 kergoth: difficult is splitting up a filesystem into separate images Nov 04 18:52:42 the split and image creation is easy, but trying to automate the update of the fstab less so ;) Nov 04 18:53:16 heh :) mvl6 does make some things easier, with their source / pstage mirror, project creation / update tool, separate toolchains with installers, and bitbake+tools installer.. not to mention their eclipse plugin, though its integration with the integration platform is still in its infancy Nov 04 18:53:28 kergoth: of course the difficult part is automation :-) Nov 04 18:54:02 well, the issue is that the default fstab could vary from distro/machine, etc.. might just leave it alone, make the user handle the fstab change Nov 04 18:56:28 kergoth: we are also working on eclipse plugin integration platform, looks like twice the same work :-( Nov 04 18:57:07 kergoth: will you be at oedem? Nov 04 18:58:06 bbl Nov 04 19:02:30 mckoan: sadly, no Nov 04 19:04:15 did you want to participate by phone? I imagine we could probably arrange that. Nov 04 19:05:20 hmm, could do that Nov 04 19:05:42 what the hell, anyone seen an error 'd' undefined in an expansion of a ${@}? Nov 04 19:06:12 hm, never seen that Nov 04 19:06:36 shouldn't be possible, since d is injected into the locals, iirc.. Nov 04 19:06:38 * kergoth grumbles Nov 04 19:07:29 yeah Nov 04 19:11:45 What is the general feeling about recipes for release candidate kernels? Worth getting support into the tree for a board asap, or waiting for next stable release? Nov 04 19:12:51 okay, now this thing will modify an fstab if it finds an entry for the path going into the partition in question.. hmm Nov 04 19:13:37 jic23_: if you have a particular need for a release candidate then it would be fine to add a recipe for it. Nov 04 19:14:17 I wouldn't suggest that you add rc kernel recipes routinely, "just because you can", but if there is a particular kernel thing that is holding you up and the rc is the only way to get it, then go ahead. Nov 04 19:14:58 Cool. Nov 04 19:17:24 Silly question, but is there a standard approach to pulling a recipe at a later date. Afterall, not going to really want an rc recipe kicking around for one board when it is now supported by stable relase. Nov 04 19:18:28 No, there is no real policy for that. Nov 04 19:18:39 cbrake u there Nov 04 19:18:53 In the particular case of kernels, if you are the board/distro maintainer and you are fairly sure that nobody else is using your recipe, you can probably just delete it when you are bored with it. Nov 04 19:19:16 In the general case, though, this is a somewhat tricky (and controversial) subject and there is no clear "right answer". Nov 04 19:20:09 rob_w: hello Nov 04 19:20:48 Some people, including me, take the view that old recipes should never be deleted unless they are clearly and demonstrably broken to such a point where they provide no conceivable value (i.e. could not be compiled under any circumstances for any platform). Other folks feel that it is not desirable to have lots of old versions cluttering up the tree and prefer to delete them when they are no longer selected by any in-tree distro. Nov 04 19:20:54 hey, i managed to setup a working ijs chain with imagemagick , ijs_client_example and hpijs Nov 04 19:21:19 JaMa: fwiw, I sent that interworking patch to glibc-ports Nov 04 19:21:30 so now i could draw into a png and put it inside this loop and print .. is that basicly what you did ? Nov 04 19:21:55 hello Nov 04 19:22:57 pb___: great, thank you Nov 04 19:23:22 JaMa: are there any binaries still crashing for you, or have those problems now gone away altogether? Nov 04 19:23:42 rob_w: yes, that sounds pretty similar Nov 04 19:23:52 pb___: everything works now.. Nov 04 19:23:54 very good Nov 04 19:24:16 any of you using the bsd-mailx / liblockfile recipes? Nov 04 19:24:46 unfortunately i dont have ghostscript yet Nov 04 19:24:54 pb__: you could have used existing macrso Nov 04 19:24:58 but i can live without it for some time Nov 04 19:28:41 khem: I don't think there are any existing macros with the appropriate meaning. Nov 04 19:29:17 pb__: yeah I see Nov 04 19:29:37 __USE_BX__, which you mentioned on the list, is defined for ARMv5T so that doesn't help. Nov 04 19:30:10 pb__: although USE_BX could be used I guess Nov 04 19:30:36 even for other architectures that support bx Nov 04 19:31:03 I'm not quite sure I understand what you're saying. Nov 04 19:31:11 Is there a USE_BX macro that's different from __USE_BX__? Nov 04 19:31:14 is mov pc or ldr pc or ldm pc any better Nov 04 19:31:34 no its same I was referring to Nov 04 19:32:01 okay, yeah, so that is not appropriate. Nov 04 19:32:14 the code sequences as they stood are fine for ARMv5. Nov 04 19:32:31 it's only v4T that needs the alternative. Nov 04 19:32:50 and, only then when interworking is enabled. Nov 04 19:33:04 urg Nov 04 19:33:07 zombie Nov 04 19:33:22 so I think that "if defined(__ARM_ARCH_4T__) && defined(__THUMB_INTERWORK__) is exactly the condition that's needed Nov 04 19:33:31 and, afaik, there is no existing macro with that meaning Nov 04 19:33:42 admittedly I could have added one, but it didn't really seem worth it for those three instances Nov 04 19:33:51 yes Nov 04 19:34:46 testing __USE_BX__ would have resulted in v5 getting the alternative sequences as well, which is not what we want. Nov 04 19:34:47 however I was proposing to change for all interworkable architectures to use BX Nov 04 19:34:54 must remain awake .... Nov 04 19:35:34 Crofton: in EU? Nov 04 19:35:40 re Nov 04 19:35:45 why would you want to do that? Nov 04 19:36:03 hrw, yeah Nov 04 19:36:07 at my Uncle's Nov 04 19:36:07 the code sequences with BX are longer and slower. Nov 04 19:36:21 * kergoth ponders Nov 04 19:36:29 Crofton: 8h difference? Nov 04 19:36:35 5 I think Nov 04 19:36:41 I do not dwell on these things Nov 04 19:36:49 WebKitBuilds/Debug//lib/libWebKitGtk.so: undefined reference to `WebCore::RenderImage::updateAltText()' - someone got it? Nov 04 19:43:20 hmm Nov 04 19:43:45 pb__: yeah true Nov 04 19:45:18 one extra intruction to load the branch address into a temporary Nov 04 19:46:37 right Nov 04 19:46:48 pb__: addressing specific case as you did is the right thing here Nov 04 19:47:05 anybody able to give me a clue as to which of the qt4 packages may contain uic, i thought i had them all installednow but i still have no uic binary Nov 04 19:49:43 I would like to have ipkg/opkg to follow build prefs... Nov 04 20:00:09 PREF_PROV_psplash = "bsplash" makes bitbake builds bsplash as psplash. but even when resulting package Provides: psplash I got psplash in rootfs (if it is in deploy) or bsplash (if psplash is not present in deploy) Nov 04 20:05:57 Sorry to all for patch dump on mailing list. Been sitting on most of that lot awaiting kernel work merging for far too long! Nov 04 20:07:35 bye all Nov 04 20:07:49 *laughs* and one of the patches just merged with v4l 20 seconds after I sent it it to oe-devel. Nov 04 20:07:53 bye all. Nov 04 20:07:56 khem: right, I think so Nov 04 20:07:57 jic23_: heh Nov 04 20:15:20 re Nov 04 20:38:41 florian: how is going with ML's on LTG? :) Nov 04 20:40:07 mmm strange strange Nov 04 20:40:08 hi Nov 04 20:40:10 | + echo 'export HISTSIZE=1000' Nov 04 20:40:17 is the last line of failure Nov 04 20:40:24 after it doesn't say it failed Nov 04 20:41:56 the branch is SHR/merge Nov 04 20:42:11 and the commit is just before the distro abi change Nov 04 20:42:53 it was bitbaking shr-image and I bet it's the do_rootfs or something like this Nov 04 20:42:57 I'll look in the logs Nov 04 20:53:05 does anyone recall what the situation is with the -common x packages, for kdrive and xorg? Nov 04 20:53:11 it seems xorg use isn't ideal at the moment Nov 04 20:55:36 kergoth: I was asking about that the other day, and playing with the xserver-common 1.30 package I was advised to test, the whole thing is a mess right now with most of the kdrive stuff supporting Xorg and the like. Nov 04 20:55:44 meh Nov 04 20:56:51 kergoth: trying to tie it all up but I am not sure I like the idea of the scripts being on an external SVN (xserver-common) when they are really just for OE? Seems odd but that's just me. Nov 04 20:57:20 Hangover from GPE I guess but they ideally would suit being more generic (well they are) Nov 04 21:00:02 Xorg compiles fine for me.... Nov 04 21:00:43 its a good thing we only need to it to compile, not actually be useful :) Nov 04 21:01:09 Xorg works fine for me Nov 04 21:01:18 even 7.5(ish) Nov 04 21:05:34 kergoth: ;-) Nov 04 21:07:19 03Koen Kooi  07org.openembedded.dev * r3a26240e09 10openembedded.git/recipes/boost/ (6 files in 2 dirs): Nov 04 21:07:19 boost 1.36: various ARM fixups: Nov 04 21:07:19 * changed intrinsics patch to work for armv6 and armv7a Nov 04 21:07:19 * disabled atomic ops for arm versions <6 Nov 04 21:07:21 * fix associated build damage Nov 04 21:07:23 * pass in correct gcc version Nov 04 21:07:25 03Koen Kooi  07org.openembedded.dev * r11b63303a0 10openembedded.git/recipes/asio/asio_1.1.1.bb: asio: bump PR to get boost fixes Nov 04 21:24:01 ugh, this partition stuff is awfully ugly.. hmm Nov 04 21:26:36 who am i kidding, image.bbclass is ugly as hell period Nov 04 21:27:54 03Martin Jansa  07shr/merge * r33ec688b6e 10openembedded.git/ (2 files in 2 dirs): Nov 04 21:27:54 linux-openmoko-2.6.31: add new version Nov 04 21:27:54 Signed-off-by: Martin Jansa Nov 04 21:27:55 03Frederik Sdun  07shr/merge * r836e4ce0b1 10openembedded.git/recipes/gypsy/ (files/remove-werror.patch gypsy_svn.bb): Nov 04 21:27:55 Remove -Werror from gypsy CFLAGS Nov 04 21:27:58 Signed-off-by: Frederik 'playya' Sdun Nov 04 21:28:00 Signed-off-by: Martin Jansa Nov 04 21:28:02 03Frederik Sdun  07shr/merge * r13ac8a5ca9 10openembedded.git/recipes/navit/navit.inc: Nov 04 21:28:05 Navit: depend on libgypsy instead of gypsy Nov 04 21:28:07 Signed-off-by: Frederik 'playya' Sdun Nov 04 21:28:09 Signed-off-by: Martin Jansa Nov 04 21:28:11 03Martin Jansa  07shr/merge * r8994eb3e99 10openembedded.git/recipes/navit/ (files/use-fso-raw.patch navit.inc): (log message trimmed) Nov 04 21:28:14 navit: small fixes Nov 04 21:28:16 * Request only GPS and CPU instead of Display resource with fsoraw, its Nov 04 21:28:17 * kergoth grumbles Nov 04 21:28:18 useless if phone suspends, but if its used ie on bike or just for trip Nov 04 21:28:20 tracking than its better to save battery life with Display off. Nov 04 21:28:29 * Drop default freerunner config for now, it needs update. Nov 04 21:28:30 * Add fsoraw to RDEPENDS for shr. Nov 04 21:28:32 03Frederik Sdun  07shr/merge * rdab67aa682 10openembedded.git/recipes/gypsy/gypsy.inc: Nov 04 21:28:35 Split gypsy into libgypsy and gypsy Nov 04 21:28:37 Signed-off-by: Frederik 'playya' Sdun Nov 04 21:28:39 Signed-off-by: Martin Jansa Nov 04 21:28:41 03Martin Jansa  07shr/merge * r4dba1c5bc0 10openembedded.git/recipes/gypsy/gypsy.inc: Nov 04 21:28:46 gypsy: remove PROVIDES, s/configdir/sysconfdir/ Nov 04 21:28:48 Signed-off-by: Martin Jansa Nov 04 21:28:50 03Martin Jansa  07shr/merge * redd7019778 10openembedded.git/recipes/navit/navit.inc: Nov 04 21:28:53 navit: drop navit.png, navit.service files Nov 04 21:28:54 really need the ability to share the same fakerootkey across multiple fakeroot tasks in some fashion Nov 04 21:28:55 * We can use icon and service file from sources Nov 04 21:28:57 * Old recipes should be addapted to this change or better dropped Nov 04 21:29:03 Signed-off-by: Martin Jansa Nov 04 21:29:05 03Martin Jansa  07shr/merge * r23325b7a1a 10openembedded.git/recipes/mesa/ (mesa-dri_6.5.2.bb mesa-dri_7.0.3.bb mesa-dri_git.bb): Nov 04 21:29:08 mesa: update git version, set default preference -1 for older mesa-dri too Nov 04 21:29:10 Signed-off-by: Martin Jansa Nov 04 21:29:12 03Martin Jansa  07shr/merge * rc4c6e68d98 10openembedded.git/conf/distro/shr.conf: Nov 04 21:29:19 shr: set bluez-utils and libs provider to bluez4 Nov 04 21:29:21 Signed-off-by: Martin Jansa Nov 04 21:29:23 03Martin Jansa  07shr/merge * rce9ca03337 10openembedded.git/recipes/ (4 files in 3 dirs): Nov 04 21:29:26 mplayer: add git with patches for om-gta02 Nov 04 21:29:28 (4 lines omitted) Nov 04 21:29:44 oh bite me, bitbake Nov 04 21:30:04 getting an undefined 'd' again Nov 04 21:30:06 damnit Nov 04 21:31:24 * kergoth twitches Nov 04 21:31:50 sometimes i get so annoyed at things, but lack any real plan to implement to get from here to where i'd rather be, and just lose all motivation to work on it at all.. Nov 04 21:31:51 heh Nov 04 21:32:56 NOTE: exceptions.NameError:global name 'd' is not defined while evaluating: Nov 04 21:32:56 ${@base_dep_prepend(d)} task-base-extended ${@' '.join(d.getVar('IMAGE_DEPENDS_%s' % (d.getVar('PARTITION_%s_FS' % partition, True) or ''), True) for partition in 'base'.split())} Nov 04 21:32:58 hrmph Nov 04 21:38:16 03Stanislav Brabec  07org.openembedded.dev * r276e2fbec4 10openembedded.git/recipes/linux/ (3 files in 2 dirs): linux-rp-2.6.26 for spitz and akita: Added netfilter support. Disabled two deprecated networking options. Nov 04 21:44:11 03Stanislav Brabec  07org.openembedded.dev * r68cdc74da7 10openembedded.git/recipes/gpsd/ (gpsd.inc gpsd_2.38.bb): gpsd: Removed custom do_stage. Nov 04 21:46:30 btw, we have dejitter patch now for zauruses Nov 04 21:47:03 meh Nov 04 21:48:02 florian: can I help anyhow with LTG? Nov 04 21:49:46 03Stanislav Brabec  07org.openembedded.dev * r292709ae7e 10openembedded.git/recipes/mplayer/mplayer_svn.bb: mplayer-svn: Worked around internal compiler error on iwmmxt. Nov 04 21:52:53 03Stanislav Brabec  07org.openembedded.dev * r691030b524 10openembedded.git/conf/distro/angstrom-2008.1.conf: angstrom-2008.1.conf: Must use mplayer-svn, anything older is incompatible with linux-libc-headers-2.6.31. Nov 04 21:58:28 03Stanislav Brabec  07org.openembedded.dev * r846588de96 10openembedded.git/recipes/prismstumbler/ (2 files in 2 dirs): prismstumbler-0.7.3+0.7.4pre1: Use much simper and more robust libgps check. Nov 04 22:02:46 03Martin Jansa  07shr/merge * r689fad2143 10openembedded.git/recipes/python/ (python-dbus_0.83.0.bb python-epydoc-native_3.0.1.bb): python-dbus: needs python-epydoc-native to build documentation Nov 04 22:04:07 Jay7: Nasty... FusionForge shows the list as created but this seems to be a lie. Nov 04 22:04:43 florian: seems mailman does not see it Nov 04 22:05:25 I can create that list in mailman by hands Nov 04 22:05:35 when I have enough rights Nov 04 22:05:50 Jay7: I suppose it is a problem of the FusionForge installation. Nov 04 22:06:14 may be just upgrade to 4.8? Nov 04 22:06:19 03Stanislav Brabec  07org.openembedded.dev * re0e13253cd 10openembedded.git/recipes/balsa/ (balsa-2.4.1/gnome-icon-theme.patch balsa_2.4.1.bb): balsa-2.4.1: Depends on gnome-icon-theme. Nov 04 22:06:32 or did you mean configuration? Nov 04 22:07:45 cd fou Nov 04 22:07:45 gi Nov 04 22:07:46 t Nov 04 22:07:47 bah Nov 04 22:07:56 * kergoth shakes fist at his macbook Nov 04 22:08:57 Jay7: hmm maybe, but I would have expected it to work before too. Creating it manually isn't too complicatd but not really a solution :) Nov 04 22:09:16 yeah.. Nov 04 22:09:27 we are just waiting for that ML :) Nov 04 22:10:01 * florian just asked in #fusionforge :) Nov 04 22:10:08 well.. say if I can help :) Nov 04 22:10:52 Jay7: ltg always needs help :-} Nov 04 22:12:00 hehe.. Nov 04 22:25:27 khem: hey there. just wanted to share that with gcc4.4 the bitbake manages to compile where 4.3 was segfaulting. so it seems opensuse gcc4.3 is kinda broken (on my machine at least) :-) Nov 04 22:25:53 there you go Nov 04 22:26:22 qsp: do they have gcc 4.4 available for your suse version Nov 04 22:26:39 yes, i installed from fficial reps Nov 04 22:26:43 *official Nov 04 22:26:56 khem: can you imagine gcc4.4 could miscompile 'selectively' some 2.6.3x kernel bits, so that it won't boot from mtd? Nov 04 22:27:47 ant__: It could.. it has happened in past with other versions of gcc Nov 04 22:27:58 but I wont know more unless you give more details Nov 04 22:28:55 sometimes its a coding problem too which just gets exposed by newer compilers sometimes its opposite Nov 04 22:29:37 booting fine from SD/CF after kexec Nov 04 22:30:05 khem: it fails immediately if launched from nand. LAst line is from bootloader: Jump to linux Nov 04 22:30:18 then death Nov 04 22:30:45 older kernels compiled with same toolchain are fine, though Nov 04 22:30:48 hmm same kernel image ? Nov 04 22:30:55 so I suspect mtd changes in kernel tree Nov 04 22:30:57 03Koen Kooi  07shr/merge * r2598835e7f 10openembedded.git/recipes/opencv/ (opencv-samples_svn.bb opencv_svn.bb): opencv: bump SRCREV to just after 2.0.0 release Nov 04 22:30:59 03Koen Kooi  07shr/merge * re55a3b9423 10openembedded.git/recipes/u-boot/ (3 files in 2 dirs): u-boot: fix beagle fw_env.conf Nov 04 22:31:00 03Koen Kooi  07shr/merge * r0f926962a2 10openembedded.git/recipes/gnome/ (3 files): gnome-themes: delete obsolete versions Nov 04 22:31:02 03Martin Jansa  07shr/merge * r8fedaafb39 10openembedded.git/recipes/eglibc/ (4 files in 2 dirs): Nov 04 22:31:02 eglibc: updated armv4t-interworking.patch Nov 04 22:31:04 * Updated patch to make SYSCALL_ERROR_HANDLER interworking safe. Nov 04 22:31:06 * Phil Blundell found and fixed another issue in eglibc armv4t builds. Nov 04 22:31:08 Signed-off-by: Martin.Jansa Nov 04 22:31:10 btw 2.6.26 and 2.6.31 for armv5te Nov 04 22:31:10 03Frans Meulenbroeks  07shr/merge * rbfb8287d87 10openembedded.git/recipes/mysql/ (files/my.cnf mysql_4.1.22.bb): Nov 04 22:31:13 mysql: changed postinstall: pid file goes to /var/run/mysql i.s.o /var/run as mysqld cannot write to /var/run Nov 04 22:31:15 this alo changes /etc/my.cnf Nov 04 22:31:17 03Frans Meulenbroeks  07shr/merge * r945c5c95b3 10openembedded.git/recipes/mythtv/mythplugins_0.21+0.22rc2.bb: mythplugins: add zoneminder package Nov 04 22:31:20 03Frans Meulenbroeks  07shr/merge * r82721f8688 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 04 22:31:27 03Marcin Juszkiewicz  07shr/merge * r266aa909d6 10openembedded.git/ (conf/checksums.ini recipes/transmission/transmission_1.76.bb): transmission: added 1.76 Nov 04 22:31:30 03Frans Meulenbroeks  07shr/merge * rcea3d55fd2 10openembedded.git/recipes/mythtv/mythtv_0.21+0.22rc2.bb: Nov 04 22:31:33 mythtv: added dependency to glibc-gconv-utf-16 Nov 04 22:31:35 still has DEFAULT_PREFERENCE = "-1" as this needs mysql5 Nov 04 22:31:39 03Koen Kooi  07shr/merge * r733d54f07b 10openembedded.git/recipes/mythtv/mythtv_0.21+0.22rc2.bb: mythtv 0.22rcX: only add glibc-gconv stuff for glibc build Nov 04 22:31:42 03Khem Raj  07shr/merge * rd26c8da511 10openembedded.git/recipes/uclibc/ (uclibc-nptl/uClibc.distro uclibc_nptl.bb): Nov 04 22:31:45 uclibc-nptl: Update to latest tip of branch Nov 04 22:31:47 * Compile uclibc nptl is ARM mode. Thumb has some issues for now. Nov 04 22:31:51 * Add new config option UCLIBC_SUSV4_LEGACY=y Nov 04 22:31:53 Signed-off-by: Khem Raj Nov 04 22:31:55 03Marcin Juszkiewicz  07shr/merge * r9f162c68f7 10openembedded.git/recipes/transmission/transmission.inc: Nov 04 22:31:58 transmission: improve packaging Nov 04 22:32:00 transmission package contains GUI client Nov 04 22:32:02 transmission-client contains CLI and Remote clients Nov 04 22:32:04 transmission-daemon contains daemon and Web interface Nov 04 22:32:08 03Khem Raj  07shr/merge * ra0afb38137 10openembedded.git/recipes/eglibc/eglibc_svn.bb: Nov 04 22:32:11 eglibc_svn.bb: Update to latest svn tip. Nov 04 22:32:13 Signed-off-by: Khem Raj Nov 04 22:32:15 03Koen Kooi  07shr/merge * r3a26240e09 10openembedded.git/recipes/boost/ (6 files in 2 dirs): Nov 04 22:32:18 boost 1.36: various ARM fixups: Nov 04 22:32:20 * changed intrinsics patch to work for armv6 and armv7a Nov 04 22:32:26 (5 lines omitted) Nov 04 22:33:14 khem: still there after the flood? Nov 04 22:33:50 heh yeah Nov 04 22:34:20 I could force my irc client to filter messages from CIA Nov 04 22:34:20 khem: the fact that the kernel booted from SD/CF detects properly the mtd partitions leads me to the conclusion the bug can only be in init code Nov 04 22:35:45 yeah seems so Nov 04 22:35:47 exactly same kernel Nov 04 22:36:35 you could use jtag to debug it Nov 04 22:37:47 well, I see the 5 contacts.. Nov 04 22:37:57 mm.. seems the problem i'm having. the fact that angstrom kernels wont boot in my c700 Nov 04 22:37:59 never heard before about jtagging a Zaurus :) Nov 04 22:38:30 qsp: no, the kernel faiuling on your device is booting fine on mine Nov 04 22:38:36 same kernel Nov 04 22:38:52 you just have 32M ram vs. 64M Nov 04 22:38:59 ah. sry :-) Nov 04 22:39:06 need to buy a dimm :-D Nov 04 22:39:53 anyway, i'm 50 tasks away from testing linux-rp (praying) Nov 04 22:40:08 ...I sent you linux-rp :/ Nov 04 22:40:11 ant__: heh I did not realize that this is on zaurus Nov 04 22:40:40 there is jtag pinout, thoug Nov 04 22:40:43 03Thomas Zimmermann  07shr/merge * r29f03c5d5c 10openembedded.git/recipes/webkit/webkit-efl_git.bb: Nov 04 22:40:43 webkit-efl: depend on sqlite3 instead of sqlite Nov 04 22:40:43 Signed-off-by: Martin Jansa Nov 04 22:41:08 khem: iirc that angel bootloader is even client-server Nov 04 22:41:21 but one would need special sw, I guess Nov 04 22:51:02 libxfont and libxfont-native are both providing xfont Nov 04 22:53:27 can anyone shed light on how ${oe}/tmp/deploy/${libc}/sources is supposed to work? it looks like source code should be grouped by license, which is what i need, but not all packages seem to be there. Nov 04 22:57:40 where can i find instrcutions on building Angstrom for Beagleboard using OE .... i am a not much familiar with OE ... any suggestions Nov 04 22:59:25 http://code.google.com/p/beagleboard/wiki/HowToGetAngstromRunning Nov 04 23:00:39 i already tried that .... i want to build complete new package using open embedded .. any help on that Nov 04 23:04:12 03Stanislav Brabec  07org.openembedded.dev * ra74f86eab3 10openembedded.git/recipes/lame/ (lame/ldflags-qa.patch lame_3.96.1.bb): lame: Fix possible linking QA error. Nov 04 23:04:13 03Stanislav Brabec  07org.openembedded.dev * r3d79f309ec 10openembedded.git/ (4 files in 3 dirs): lame: Update to version 3.98.2. Nov 04 23:08:23 amitshah1: read oe manual Nov 04 23:11:15 Heinervdm: where do you see it Nov 04 23:13:17 khem: in shr/merge but it should be the same in org.oe.dev Nov 04 23:14:02 libxfont-native includex libxfont and libxfont has PROVIDES = "xfont" Nov 04 23:14:58 i think it's the same for libxext, there i changed it in shr/merge. but i don't know if i did it the correct way Nov 04 23:17:05 Heinervdm: it gets redefined after including Nov 04 23:17:40 for PROVIDES = "${P} ${PF} ${PN}" Nov 04 23:17:51 in libxfont-native_1.3.0.bb Nov 04 23:18:20 should override whatever it is set to in libxfont Nov 04 23:19:31 khem: oh, ok then it's just an error in shr/merge Nov 04 23:20:25 i will fix it there Nov 04 23:26:22 03Khem Raj  07org.openembedded.dev * r8bfbb10c98 10openembedded.git/conf/machine/ (alix.conf guinness.conf progear.conf x86.conf): Nov 04 23:26:27 x86.conf: Fix typo in comments Nov 04 23:26:29 * Write NPTL instead of NTPL Nov 04 23:26:31 Signed-off-by: Khem Raj Nov 05 01:24:08 has anyone had success with src_distribute_local added to INHERIT? Nov 05 01:44:03 ah nevermind, looks like i didn't clean /tmp well enough Nov 05 01:50:03 kergoth: where is the tmp/deploy/etc dir? i'm not finding one in me OE tree anywhere. Nov 05 01:50:28 TMPDIR defaults to ${TOPDIR}/tmp Nov 05 01:50:37 topdir defaults to $PWD, where you ran the bitbake command Nov 05 01:52:06 http://docs.openembedded.org/usermanual/html/gettingoe_configuring_oe.html Nov 05 01:52:42 to use that example structure, it would be $OEBASE/build/tmp/deploy/images/, assuming you run in build, which is the usual case Nov 05 01:53:44 nullpuppy: http://docs.openembedded.org/usermanual/html/usage_workspace.html describes this. Nov 05 01:54:00 hmm Nov 05 01:55:21 ok. i wonder if i rm'd tmp, cause i'm not finding one. *sigh* Nov 05 01:57:34 tmp is created when you do a build, relative to the dir you run the command in Nov 05 01:57:58 can always explicitly set TOPDIR or TMPDIR in local.conf to make it always emit relative to the build dir Nov 05 01:58:00 many do so Nov 05 01:59:09 yeah. i still had the screen session open that i ran bitbake in, (which, since i was following angstrom build instructions, was in $(OE)/openembedded) Nov 05 02:00:02 and the setup script isn't changing TOPDIR or TMPDIR so.. i dunno what happened. Nov 05 02:00:57 * kergoth shrugs Nov 05 02:01:17 yep, pretty much. Nov 05 02:01:22 rebuilding now. Nov 05 02:03:04 *sigh* screen -x can be so unreliable sometimes. Nov 05 02:04:58 hmm, and i don't see a tmp being generated. /me pokes around a bit. Nov 05 02:10:36 kergoth: ok... if its not creating a tmp dir in pwd, have any idea what i should do or look for/at? Nov 05 02:10:55 oooh Nov 05 02:10:56 bitbake -e | grep \^TMPDIR= Nov 05 02:10:59 ignore that. Nov 05 02:11:35 the local.conf is setting it ;) Nov 05 02:12:11 however, i don't have a deploy/images. i do have a deploy/ which only has glibc in it. Nov 05 02:12:30 ah ha! ok. Nov 05 02:12:54 things make a lot more sense now. thanks. **** ENDING LOGGING AT Thu Nov 05 02:59:56 2009