**** BEGIN LOGGING AT Sat Nov 08 03:00:02 2008 Nov 08 03:44:34 03Mike Westerhof  07org.openembedded.dev * rc636a602e0 10openembedded.git/packages/images/fso-image-nox.bb: Nov 08 03:44:34 fso-image-nox: a small no-x11 fso image for the GTA01 and GTA02 Nov 08 03:44:34 (suitable for use as a rootfs for Qtopia or Qt Extended) Nov 08 03:44:45 03Mike Westerhof  07org.openembedded.dev * r081088f32f 10openembedded.git/packages/ (3 files in 2 dirs): meta-toolchain-fso: bitbake recipes for a toolchain and SDK for fso. Nov 08 04:35:45 Hello; do someone has any idea why xorg can be ignoring xorg.conf keyboard layout? Nov 08 04:38:21 Oh! it is using HAL Nov 08 06:18:07 03Otavio Salvador  07org.openembedded.dev * r1303e1f35b 10openembedded.git/ (2 files in 2 dirs): xkeyboard-config: update to 1.4 Nov 08 08:48:37 03Koen Kooi  07org.openembedded.dev * r78c51f1083 10openembedded.git/packages/linux/ (14 files in 3 dirs): linux-omap git: add patches for new DSS (the nokia version, not the TI one), update beagle defconfig and bump to 2.6.28rc3 Nov 08 09:13:11 03Tom Talpey  07org.openembedded.dev * r091b88f6df 10openembedded.git/packages/gnome/gnome-games_2.24.0.bb: gnome-games: better granular packaging Nov 08 09:13:12 03Tom Talpey  07org.openembedded.dev * ra7986a7530 10openembedded.git/packages/gnome/gnome-python-desktop_2.23.0.bb: gnome-python-desktop: Add DEPENDS for gnome-python and libgnomeprint*, needed for gnome-games. Nov 08 09:13:13 03Tom Talpey  07org.openembedded.dev * rb8fc3e9a04 10openembedded.git/packages/gnome/gnome-python_2.22.1.bb: gnome-python 2.22.1: new package Nov 08 09:13:14 03Tom Talpey  07org.openembedded.dev * rfe18b91c58 10openembedded.git/packages/gnome/gnome-desktop_2.24.0.bb: gnome-desktop: add libxrandr to DEPENDS Nov 08 09:13:17 03Tom Talpey  07org.openembedded.dev * rf230fbcfc5 10openembedded.git/packages/guile/guile_1.8.5.bb: guile: Add libtool dependency Nov 08 09:13:20 03Tom Talpey  07org.openembedded.dev * r14074e4e13 10openembedded.git/packages/gnome/pyorbit_2.14.3.bb: pyorbit 2.14.3: new recipe Nov 08 10:09:07 * * OE Bug 4815 has been created by  Nov 08 10:09:09 * * wireless-tools-29-autobuild Nov 08 10:09:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4815 Nov 08 10:37:12 morning folks Nov 08 11:31:49 anyone doing regular builds w/ 4.3.x? Nov 08 11:31:52 (gcc, that is) Nov 08 11:40:14 mickeyl: I did some with Poky Nov 08 11:40:39 mickeyl: pb told once that 4.3 generated bigger binaries then 4.2 (for target=arm) Nov 08 11:41:40 right, i have observed this as well Nov 08 11:41:55 they claim that the binaries are faster though Nov 08 11:42:00 need to run some benchmarks Nov 08 11:55:05 indeed Nov 08 11:56:01 I will do new builds for some hardware - some are present in OE already (progear, alix), some are not yet Nov 08 11:56:20 but all angstrom based so rather 4.2 then 4.3 Nov 08 11:56:58 hmm.. but angstrom uses 4.3 for some targets... so alix will switch to 4.3 to get -march=geode ;D Nov 08 11:57:01 bbl Nov 08 12:22:00 mickeyl: there is a fairly grave optimiser bug in 4.3 which causes sucky code generation for arm in many situations. Nov 08 12:22:12 I think you are probably best advised to stick to 4.2.4 for most purposes right now Nov 08 12:24:15 PR31849, if you're interested Nov 08 12:24:48 (good morning, by the way) Nov 08 12:53:27 re Nov 08 12:54:03 pb_: in case you're interested...glib patch for thumb http://rafb.net/p/ALa9ku20.html Nov 08 12:55:52 back to baby-sitting now... Nov 08 13:01:57 pb_: good morning. thanks for the heads up Nov 08 13:02:19 pb_: do you think there's any benefit in moving from 4.1.3 up to 4.2.x? Nov 08 13:02:26 4.1.3 is what we're currently using Nov 08 13:07:13 mickeyl: I have found that 4.2.4 seems to give better code in general but the benefits are only modest. Nov 08 13:07:32 if you plan on using thumb, I think 4.2.4 is probably obligatory; 4.1.x seems to miscompile some important packages in that mode. Nov 08 13:08:39 I suppose the ideal thing would be to use 4.3.x but revert the particular change which triggered PR31849. There are some other improvements in 4.3.x (e.g. the section anchor thing) which would be worth having. Nov 08 13:08:53 I'm not sure how easy it would be to back out just that one change though. Nov 08 13:09:14 hi pb Nov 08 13:09:22 hi woglinde Nov 08 13:17:45 pb_: ah, good to know. thanks Nov 08 13:18:21 pb and binutils-2.19? Nov 08 13:20:33 have fun mickeyl Nov 08 13:21:23 thanks. league game Nov 08 13:27:05 * * OE Bug 4816 has been created by  Nov 08 13:27:07 * * keylaunch-conf-2.0.10-autobuild Nov 08 13:27:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4816 Nov 08 14:19:05 Hello, this http://rafb.net/p/H5h2zU57.html is the gdal-config file of gdal library. My problem is that it is staged unchanged even with binconfig class included and thun the paths are incorrect. How can I fix it please? Nov 08 14:20:30 Note that it's locaced in ${S}/apps/gdal-config if it matters. Nov 08 14:25:49 pb_: ping Nov 08 14:26:53 bluez-utils is failing now (using thumb) http://tinderbox.openembedded.net/public/logs/1906571.txt Nov 08 14:27:21 hi woglinde Nov 08 14:27:42 ant_: hello Nov 08 14:27:50 hi Nov 08 14:28:46 ah, bluez-utils_3.36.bb is teh suck. Nov 08 14:28:57 anyone familiar with bitbake? Nov 08 14:29:15 ant_: you need to fix it to not clobber SRC_URI, or alternatively to manually import sbc-thumb.patch Nov 08 14:29:24 pb_ I see a sbc-thumb.patch Nov 08 14:29:33 he, ok Nov 08 14:29:53 hi mrdata Nov 08 14:30:12 Blackhex I am just trying to verify it Nov 08 14:30:19 hi woglinde Nov 08 14:30:44 zecke: good afternoon Nov 08 14:30:53 woglinde: great thanks, but it's newer version of gdal I'make bb file for. I'll past it too Nov 08 14:31:08 Blackhex hm problem should be the same Nov 08 14:31:35 woglinde: yes but current package won't compile anyway Nov 08 14:33:49 blackhex then please pastebin it Nov 08 14:33:52 woglinde: http://rafb.net/p/eBNvPt30.html and http://rafb.net/p/eoDgaQ96.html Nov 08 14:35:37 Blackhex names are missing Nov 08 14:36:14 woglinde: ah sorry, gdal.inc and gdal-1.5.3.bb Nov 08 14:36:49 hm and why you delete most of the options? Nov 08 14:36:53 and depends? Nov 08 14:37:10 curl is needed Nov 08 14:37:42 pb_: bluez-utils_3.36.bb need to be purged...why SRC_URI after including ? Nov 08 14:37:48 woglinde: I just removed postgres Nov 08 14:38:11 are we talking about the same package? Nov 08 14:38:21 packages/gdal/gdal_1.5.2.bb Nov 08 14:39:46 woglinde: this is original gdal-1.3.2.bb http://rafb.net/p/B1Tgez76.html Nov 08 14:40:05 woglinde: I'm working on openmoko version of openembedded Nov 08 14:41:09 Blackhex that do intrest me much Nov 08 14:41:17 in oe is a newer one Nov 08 14:41:24 I will fix this and good Nov 08 14:41:27 take over the packge Nov 08 14:41:43 woglinde: ok Nov 08 14:43:54 ant_: indeed Nov 08 14:44:06 and FILESPATH seems broken too Nov 08 14:47:45 pb_: both patches are failing...I see why some lazy dev skipped these... Nov 08 14:48:09 and rewrote SRC_URI Nov 08 14:51:35 well, actually 'sbc-thumb.patch' applies...is hciattach-ti-bts.patch who is failing Nov 08 14:52:13 sorry for leave, my laptop went down Nov 08 15:12:36 is there any way to completely clean bitbake's knowledge of a package? Nov 08 15:12:49 pb_: glib-20_2.18.1 and bluez-utils-3.36 were the only recipes to patch in order to compile thumb-mode (built console-, x11-, opie-image) Nov 08 15:13:11 my kernel package seems to have become completely broken and it refuses to do anything with it Nov 08 15:13:23 ant_: ok, very good Nov 08 15:14:58 pb_:gcc-cross-4.2.4 Nov 08 15:16:25 righto Nov 08 15:16:31 so, do the resulting images work ok for you? Nov 08 15:16:53 I'm checking now... Nov 08 15:20:23 pb_: yes..booting from SD Nov 08 15:20:53 (usual logs :-) Nov 08 15:21:24 very good Nov 08 15:21:33 pb_: I can confirm all quirks remain as before ;-) Nov 08 15:21:46 heh Nov 08 15:22:11 right, compiling as thumb will not make opie any better :-) Nov 08 15:23:57 he, I mean before opie starts...udev and other initscripts Nov 08 15:24:05 now let see x11 image Nov 08 15:24:49 pb_: is it normal : export CXX="ccache arm-angstrom-linux-gnueabi-g++ -march=armv5te -mtune=xscale -mthumb-interwork -mthumb -mthumb-interwork -mthumb" Nov 08 15:24:57 twice -mthumb Nov 08 15:25:13 and before was : export CC="ccache arm-angstrom-linux-gnueabi-gcc -march=armv5te -mtune=xscale -mthumb-interwork -mthumb -mthumb-interwork -mthumb" Nov 08 15:31:35 I can confirm that gdal_1.5.2 suffers with same problem Nov 08 16:18:39 ant_: the repeated -mthumb is a bit weird, but it won't do any harm Nov 08 16:22:08 re Nov 08 16:37:36 Blackhex still there? Nov 08 16:37:43 woglinde: yep Nov 08 16:38:09 Blackhex are the /usr/share/ from gdal needed for runnig? Nov 08 16:38:33 otherwise I put them in an extra package for save space Nov 08 16:38:33 woglinde: I'm not sure, never get so far Nov 08 16:38:41 hm okay Nov 08 16:38:59 which packages needs gdal? Nov 08 16:39:09 I want to test Nov 08 16:39:14 the -config Nov 08 16:39:29 woglinde: I don't know which one in OE but I can post you my package of grass Nov 08 16:40:35 yes would be okay Nov 08 16:40:46 I can check in too if you want Nov 08 16:40:48 pb_ : tune-xscale.inc already requires tune-thumb.inc. Nov 08 16:40:56 ant_ hehe Nov 08 16:40:58 http://rafb.net/p/17Dp6p91.html gras_6.2.3.bb Nov 08 16:41:37 woglinde: great thanks, getting GRASS on openmoko is my priority now, but I'm just starting with bitbake and git so I don't have needed knowlegde yet Nov 08 16:43:51 Blackhex hm DEPENDS only is okay Nov 08 16:44:01 because gdal depends on proj4 Nov 08 16:44:18 woglinde: ok Nov 08 17:01:45 ~lart grass authors for not understanding autotools Nov 08 17:01:45 * ibot turns grass authors into a lifesized tux doll for not understanding autotools Nov 08 17:13:15 pb_: I'm checking about that broken hciattach-ti-bts.patch, then I'll post my git diffs Nov 08 18:37:23 Good morning! Nov 08 18:37:48 * hubar notes it is 1:30 already, not exactly "morning"... Nov 08 18:37:59 hi Nov 08 18:42:28 hi hubar Nov 08 18:42:35 I am still fighting with apache Nov 08 18:42:42 hi wog! Nov 08 18:42:47 but not I have to switched for grass Nov 08 18:42:51 args now Nov 08 18:43:15 Anything I can help with? Nov 08 18:43:31 hm not really Nov 08 18:43:41 woglinde: maybe you can instruct me what should i do? Nov 08 18:43:43 apache suckz Nov 08 18:43:51 Blackhex nope not yet Nov 08 18:46:19 I noticed that gdal sets gdal-config paths according to --path configure option but all packages in OE uses --prefix=/usr or so... Nov 08 18:51:46 ~seen rp Nov 08 18:51:48 rp is currently on #maemo (10d 2h 10m 17s) #handhelds (10d 2h 10m 17s) #oe (10d 2h 10m 17s) #openmoko (10d 2h 10m 17s), last said: 'morning all'. Nov 08 21:01:39 are the recipes in packages/images/ supposed to make a complete system, or is it assumed that you sometimes have to add things to them? Nov 08 21:01:49 ????? Nov 08 21:01:59 what you mean with complete system? Nov 08 21:02:04 i'm trying to use xfce4-image Nov 08 21:02:20 bitbake opie-image builds an image which you can boot and use Nov 08 21:02:20 but it tries to use lots of bits of X11 that it hasnt included Nov 08 21:02:35 oliford when you should peseter otavio Nov 08 21:02:44 use, i've been using x11-gpe-image and that works great Nov 08 21:02:48 oliford: in which way not included? libraries? runtime binaries? Nov 08 21:02:48 *yes, Nov 08 21:03:07 ? Nov 08 21:03:09 packages presumably Nov 08 21:03:35 sorry, my lisp AI can not compute that Nov 08 21:04:05 like it wants xset, which is included in x11-gpe-image etc from task-gpe.bb Nov 08 21:04:13 but xfce-image never includes it Nov 08 21:07:04 oliford: I still don't follow. But if xfce-foo-app invokves xset, but xset is not installed, then you want to add 'xset' to the RDEPENDS of xfce-foo-app Nov 08 21:08:23 I expected xfce-image to build a system which at the very least boots in xfce given that opie-image and x11-gpe-image etc create a system which at least gets you into opie and gpe Nov 08 21:08:55 but it doesn't.. It doesn't even try to start x and when doing it manually it fails Nov 08 21:09:24 I wanted to know if it is suppose to, but just isnt complete, or if I'd misunderstood what the images were supposed to do Nov 08 21:09:57 oliford you know the line "work in progress"? Nov 08 21:10:03 oliford: images should boot :) Nov 08 21:10:06 ah ok Nov 08 21:10:07 thats what is oe too Nov 08 21:10:26 thats fine, i'll probably try to help make it work then Nov 08 21:10:28 oliford: the question is if you want to help, then you should look at the issue and propose patches Nov 08 21:10:40 just wanted to make sure i wasnt getting something very wrong Nov 08 21:10:56 oliford: where do you want to start? A xserver is installed at all? Nov 08 21:11:29 yes, it installs xserver-kdrive-Xfbdev I think Nov 08 21:12:45 but doesnt start it, and the 'startxfce4' script it provides tries to run 'xinit' which doesnt exist Nov 08 21:13:31 but I'll play about with it, I should have more time for higher level stuff now I've almost finished the kernel Nov 08 21:15:12 oliford write a bureport Nov 08 21:15:20 I think otavio will fix this Nov 08 21:16:08 oliford: okay, does installing xserver-nodm-init change anything? Nov 08 21:17:13 i'll give it a try Nov 08 21:17:16 thanks Nov 08 21:58:31 hmm well, with some fiddling that starts x. unfortunately it crashes everytime you run anything Nov 08 21:58:53 glibc detected Xfbdev: munmap_chuck() Nov 08 22:00:12 pb_: ping Nov 08 22:19:06 zecke: may i request your help once more? do you have a little time to spare? Nov 08 22:21:09 hmm Nov 08 22:21:16 maybe someone else reads this Nov 08 22:21:29 i have a build-failure of libtool-native Nov 08 22:21:40 it somehow thinks it's configured already Nov 08 22:21:59 same with libtool Nov 08 22:22:45 http://rafb.net/p/zNaJ0j53.html <--- that's the log.do_configure of libtool Nov 08 22:24:18 tomboy64: hehe, well, read, see line 40, open config.log and search for "C compiler cannot create executables Nov 08 22:24:22 " Nov 08 22:24:36 uhm Nov 08 22:24:38 heh Nov 08 22:24:40 -.- Nov 08 22:25:26 zecke: what about this one? Nov 08 22:26:54 tomboy64: no idea, somehow your system compiler does not get executed... consult config.log for the failure ;) Nov 08 22:27:20 grrr Nov 08 22:27:39 i meant this one: http://rafb.net/p/oeKNyQ28.html Nov 08 22:27:50 (checking out the config.log now Nov 08 22:27:52 ) Nov 08 22:29:22 pb_: hot! thumb-patches in git-format: http://www.pastebin.ca/1249060 http://www.pastebin.ca/1249061 Nov 08 22:37:08 zecke: and here is the config.log: http://rafb.net/p/ok3GQi17.html Nov 08 22:37:28 tomboy64: sorry, you need to read it yourself... I have a cold and quite some headache Nov 08 22:37:55 zecke: oh, i'm sorry. have a good rest then :) Nov 08 22:38:26 tomboy64: have you mounted your data partition no exec? Nov 08 22:38:53 arm-angstrom-linux-gnueabi-gcc: Permission denied Nov 08 22:38:57 line 592... Nov 08 22:39:01 zecke: no. i've run into this error before because of that problem. since then it's mounted with exec-rights Nov 08 22:39:32 * tomboy64 doublechecks Nov 08 22:39:53 /dev/sda10 on /home/tboy/dev type ext3 (rw,nosuid,nodev) Nov 08 22:39:57  Nov 08 22:40:58 zecke: i've recently uninstalled a cross-compile that i've installed earlier. might that be the problem? Nov 08 22:41:03 a cross-toolchain even Nov 08 22:41:28 or does bitbake come with its own native toolchain? Nov 08 22:41:36 tomboy64: no idea, a) you removed the exec permission of the toolchain Nov 08 22:41:49 tomboy64: OE builds a toolchain (normally) Nov 08 22:42:12 bitbake -cdevshell libtool Nov 08 22:42:21 * tomboy64 shrugs Nov 08 22:42:31 which arm-angstrom-linux-gnueabi-gcc Nov 08 22:43:07 which: no arm-angstrom-linux-gnueabi-gcc Nov 08 22:43:21 i'll exec setup-env Nov 08 22:43:58 hmm Nov 08 22:44:02 weird Nov 08 22:44:14 still no gcc Nov 08 22:45:18 http://rafb.net/p/ZLSafb19.html <--- the devshell Nov 08 22:52:06 zecke: does it sound right to you that, compiling using thumb, the kernel size is slightly increased while the image size is at least 5-10% smaller? Nov 08 23:45:42 is there a program/package/something for X to emulate right clicking like winCE does? Nov 08 23:50:28 oliford look into libxtst and write your own programm Nov 08 23:50:53 cheers Nov 08 23:55:36 oliford: there are various ways, e.g. there is a tap and hold module for gtk Nov 08 23:55:53 yea, thats what i was thinking Nov 09 00:08:43 03Stefan Schmidt  07org.openembedded.dev * rb71fd3c563 10openembedded.git/conf/checksums.ini: checksums.ini: Add sums for xkeyboard-config-1.3 **** ENDING LOGGING AT Sun Nov 09 02:59:57 2008