**** BEGIN LOGGING AT Fri Sep 19 02:59:57 2008 Sep 19 04:44:56 03  07org.openembedded.dev * r4d936f868b 10OE.dev/packages/images/fso-image.bb: fso-image: fixed typo Sep 19 04:45:01 03  07master * r4d936f868b 10OE.dev/packages/images/fso-image.bb: fso-image: fixed typo Sep 19 05:22:16 <_12ways_> hi Sep 19 05:23:08 <_12ways_> how can i do when workpaces conflict due to unversioned patch? Sep 19 05:23:35 <_12ways_> should i do clear or rollback something? Sep 19 05:46:20 * * OE Bug 4554 has been RESOLVED (FIXED) by xjqian(AT)gmail.com Sep 19 05:46:22 * * typo in fso-image.bb Sep 19 05:46:24 * * http://bugs.openembedded.net/show_bug.cgi?id=4554 Sep 19 05:46:49 RP: ping - who knows most about the git fetcher? Sep 19 06:10:29 Hi Sep 19 06:11:08 Where can I see all the environment variables in OE Sep 19 06:17:09 I am totally new to OE Sep 19 06:17:16 Please respond Sep 19 06:42:08 in a device with gpe, how do i turn on the wlan? Sep 19 07:13:10 hi, anybody familiar with FLAT file format and how kernel load it? Sep 19 07:48:15 anybody remember what does: APPTYPE and APPDESKTOP mean in some of the old OPIE recipes? Sep 19 07:48:29 bonjour Sep 19 08:01:38 03  07org.openembedded.dev * r212ba0727d 10OE.dev/: Sep 19 08:01:38 merge of '56d2135bf97d27ef385220e2f6f36071d007d2c1' Sep 19 08:01:38 and '93613724e08c89b20a49d8619284f2f1ee5f7898' Sep 19 08:01:39 03  07org.openembedded.dev * rfb277a17b3 10OE.dev/packages/gcc/ (gcc-package-target.inc gcc_4.3.2.bb): gcc 4.3.2: fix packaging of crt*.o and include-fixed Sep 19 08:01:42 03  07master * rfb277a17b3 10OE.dev/packages/gcc/ (gcc-package-target.inc gcc_4.3.2.bb): gcc 4.3.2: fix packaging of crt*.o and include-fixed Sep 19 08:01:53 03  07master * r212ba0727d 10OE.dev/: Sep 19 08:01:53 merge of '56d2135bf97d27ef385220e2f6f36071d007d2c1' Sep 19 08:01:53 and '93613724e08c89b20a49d8619284f2f1ee5f7898' Sep 19 08:08:20 * * OE Bug 4555 has been created by miska(AT)atrey.karlin.mff.cuni.cz Sep 19 08:08:22 * * Circular dependency in opie-i18n Sep 19 08:08:24 * * http://bugs.openembedded.net/show_bug.cgi?id=4555 Sep 19 08:50:34 03  07org.openembedded.dev * r4fe990d230 10OE.dev/packages/gtk+/ (gtk+_2.12.0.bb gtk+_2.12.11.bb gtk+_2.12.3.bb): Sep 19 08:50:34 Increased PR to reflect change in do_stage in gtk+.inc by koen Sep 19 08:50:34 (see http://gitweb.openembedded.net//OE.dev.git/?a=commit;h=b2322f422be6264a3c9424a69d7bb46e67634dfa) Sep 19 08:50:34 gtk+ 2.12.0: Increase PR. Sep 19 08:50:35 gtk+ 2.12.3: Increase PR. Sep 19 08:50:37 gtk+ 2.12.11: Increase PR. Sep 19 08:50:41 03  07master * r4fe990d230 10OE.dev/packages/gtk+/ (gtk+_2.12.0.bb gtk+_2.12.11.bb gtk+_2.12.3.bb): Sep 19 08:50:44 Increased PR to reflect change in do_stage in gtk+.inc by koen Sep 19 08:50:46 (see http://gitweb.openembedded.net//OE.dev.git/?a=commit;h=b2322f422be6264a3c9424a69d7bb46e67634dfa) Sep 19 08:50:48 gtk+ 2.12.0: Increase PR. Sep 19 08:50:50 gtk+ 2.12.3: Increase PR. Sep 19 08:50:52 gtk+ 2.12.11: Increase PR. Sep 19 08:57:20 03  07org.openembedded.dev * rca1b613481 10OE.dev/: Sep 19 08:57:20 merge of '86aec156c9005ac5b6e746f4efd0249468f7a04e' Sep 19 08:57:20 and 'f09f8f0cf40e82fd7a6a914eb75442c70a3afdc9' Sep 19 08:57:22 03  07master * r0c4ad63c14 10OE.dev/packages/zbedic/libbedic_1.1.bb: (log message trimmed) Sep 19 08:57:22 libbedic: fix build Sep 19 08:57:22 * add DEPENDS = sqlite3 Sep 19 08:57:25 * add -DVERSION, introduced in 1.1, used as a string to set a db Sep 19 08:57:27 property Sep 19 08:57:29 in dynamic_library.cpp Sep 19 08:57:31 * we don't want mkbedic and xerox on embedded devices Sep 19 08:57:33 03  07org.openembedded.dev * r0c4ad63c14 10OE.dev/packages/zbedic/libbedic_1.1.bb: (log message trimmed) Sep 19 08:57:36 libbedic: fix build Sep 19 08:57:38 * add DEPENDS = sqlite3 Sep 19 08:57:40 * add -DVERSION, introduced in 1.1, used as a string to set a db Sep 19 08:57:42 property Sep 19 08:57:44 in dynamic_library.cpp Sep 19 08:57:46 * we don't want mkbedic and xerox on embedded devices Sep 19 08:57:48 03  07master * rca1b613481 10OE.dev/: Sep 19 08:57:50 merge of '86aec156c9005ac5b6e746f4efd0249468f7a04e' Sep 19 08:57:52 and 'f09f8f0cf40e82fd7a6a914eb75442c70a3afdc9' Sep 19 09:09:07 re raster Sep 19 09:19:44 is there an easy way to check which package drags in a certain dependency to another package? Sep 19 09:23:51 cortez bitbake -g mootarget paints you a nice graph with dependencies Sep 19 09:24:16 unfornatly the graph is to big for human readibly Sep 19 09:24:40 but you can search the two .dot files for the depends Sep 19 09:24:54 another way is bitbake -i Sep 19 09:25:04 and in the bitbake shell type Sep 19 09:25:20 peek packagename DEPENDS Sep 19 09:26:19 aha, I'll try that. Thanks! Sep 19 09:30:54 hi thesing Sep 19 09:33:45 morning all Sep 19 09:35:43 03  07org.openembedded.stable * rf48bb1e941 10OE.dev/packages/gob2/gob2_2.0.14.bb: gob: adjust SRC_URL Sep 19 09:43:43 morning Sep 19 10:00:41 cortez: you can have a look at the testlab directories in /deploy Sep 19 10:01:30 I'll do that, thanks Sep 19 10:01:36 cortez: or you can check the deps here for single packages: http://tinderbox.openembedded.net/ Sep 19 10:04:14 cortez: have you tested the latest kexecboot-kernels btw? Sep 19 10:04:29 iirc you were pioneering kexec... Sep 19 10:59:24 03  07org.openembedded.dev * r164e29492b 10OE.dev/packages/schroedinger/schroedinger_1.0.5.bb: schroedinger: add staging Sep 19 10:59:29 03  07org.openembedded.dev * r39c6dadfc6 10OE.dev/packages/ffmpeg/ffmpeg_git.bb: ffmpeg: enable schroedinger Sep 19 10:59:29 03  07master * r164e29492b 10OE.dev/packages/schroedinger/schroedinger_1.0.5.bb: schroedinger: add staging Sep 19 10:59:29 03  07master * r39c6dadfc6 10OE.dev/packages/ffmpeg/ffmpeg_git.bb: ffmpeg: enable schroedinger Sep 19 11:00:38 mickeyl: good morning Sep 19 11:01:39 morning pb_ Sep 19 11:01:39 he pb Sep 19 11:01:45 hi woglinde Sep 19 11:07:10 gm Sep 19 11:07:23 he crofton Sep 19 11:23:53 hey pb_ short-tailed Sep 19 11:24:51 03  07org.openembedded.dev * rc44bc230cc 10OE.dev/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions.inc: prefer 2.14.2 version of python-pygobject Sep 19 11:24:55 03  07org.openembedded.dev * r998b73e0fc 10OE.dev/: Sep 19 11:24:55 merge of '02c1db88e98d999e6fbb3fda096727c1fc8ad6aa' Sep 19 11:24:55 and '3a233f347017316b01a5399291d49424a3a4ba9f' Sep 19 11:24:56 03  07master * rc44bc230cc 10OE.dev/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions.inc: prefer 2.14.2 version of python-pygobject Sep 19 11:24:57 03  07master * r998b73e0fc 10OE.dev/: Sep 19 11:24:59 merge of '02c1db88e98d999e6fbb3fda096727c1fc8ad6aa' Sep 19 11:25:01 and '3a233f347017316b01a5399291d49424a3a4ba9f' Sep 19 11:31:47 thesing: ping Sep 19 11:41:40 ant_work: pong Sep 19 11:41:51 hey Sep 19 11:41:57 what about fixing kexecboot (angstrom-dev ML)? Sep 19 11:42:34 next angstrom-stable was planned for end-september, hopefully could incorporate kexecboot Sep 19 11:43:18 are you in touch with koen? What are the reason for the delay/no reply ? Sep 19 11:44:35 is kexecboot stuff in dev also? Sep 19 11:44:42 dev only Sep 19 11:44:46 atm Sep 19 11:46:18 now including linux.inc, more machines can be added Sep 19 11:47:34 (machines with insane bootloaders, mostly) Sep 19 11:48:14 ant_work: I don't have time atm. I have to do my thesis first. Sep 19 11:48:36 he, priority #1 Sep 19 11:49:39 but I don't ask for further development now, only asking to apply the patches Sep 19 11:50:02 (nonmworking now for c7x0, akita, spitz) Sep 19 11:52:15 thesing, what is your thesis on? Sep 19 11:54:38 Crofton|work: its about a middleware to add offline capabilities to webshops. Sep 19 11:55:09 Crofton|work: its a diploma thesis so not too complex Sep 19 11:55:41 this is equivalent of US BS degree? Sep 19 11:57:36 Crofton: more to a master degree Sep 19 11:57:44 ok Sep 19 12:03:21 ant_work: which bug are those patches on? Sep 19 12:04:50 so I will head of to the maemo summit Sep 19 12:14:05 re Sep 19 12:14:18 thesing: without console= kexecboot fails Sep 19 12:14:35 (koen broke that) Sep 19 12:15:14 and then that nasty if_down message Sep 19 12:15:37 http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2008-September/002536.html Sep 19 12:16:50 is harmless, but destroys the nice screen background... Sep 19 12:21:29 thesing: btw I did not open a new bug, rather I attached patches to #4543 Sep 19 12:23:27 ah, well, I see now is probably broken for collie/poodle/tosa too :D Sep 19 12:25:00 every machine is now needing a specified CMDLINE_xyz = "console=ttyS0,115200n8 console=tty1" (defconfig only contains 'debug') Sep 19 13:02:04 03  07org.openembedded.dev * r1869672013 10OE.dev/ (6 files in 3 dirs): Sep 19 13:02:04 enlightenment: PV went backwards *again* so bump PE *again* Sep 19 13:02:04 * I know people don't care about QA, but having PV go backwards twice? Come on! Sep 19 13:02:10 03  07org.openembedded.dev * r7e6f480837 10OE.dev/: Sep 19 13:02:10 merge of '40f7de1bbc9465cdefd2658e7f34b24d7d121a82' Sep 19 13:02:10 and '5d50023ad7c56a1839a196c074ca30c9200fbd15' Sep 19 13:02:10 03  07master * r7e6f480837 10OE.dev/: Sep 19 13:02:11 merge of '40f7de1bbc9465cdefd2658e7f34b24d7d121a82' Sep 19 13:02:13 and '5d50023ad7c56a1839a196c074ca30c9200fbd15' Sep 19 13:02:20 03  07master * r1869672013 10OE.dev/ (6 files in 3 dirs): Sep 19 13:02:20 enlightenment: PV went backwards *again* so bump PE *again* Sep 19 13:02:20 * I know people don't care about QA, but having PV go backwards twice? Come on! Sep 19 13:08:57 03  07org.openembedded.dev * r63c4c6c6e9 10OE.dev/packages/efl1/ (25 files): efl: bump PR since PE doesn't get encoding into package filenames Sep 19 13:08:58 03  07org.openembedded.dev * r77c50bf1d7 10OE.dev/: Sep 19 13:08:58 merge of '32e3cb7e99864d52a06bdc489a128fc6d02960c2' Sep 19 13:08:58 and '9f20eba616b60c443a76d8de23b3759590c868c8' Sep 19 13:09:05 03  07master * r63c4c6c6e9 10OE.dev/packages/efl1/ (25 files): efl: bump PR since PE doesn't get encoding into package filenames Sep 19 13:09:06 03  07master * r77c50bf1d7 10OE.dev/: Sep 19 13:09:06 merge of '32e3cb7e99864d52a06bdc489a128fc6d02960c2' Sep 19 13:09:06 and '9f20eba616b60c443a76d8de23b3759590c868c8' Sep 19 13:48:51 re Sep 19 14:31:43 hello Sep 19 14:32:06 is there any easy way to force using libusb instead of libusb-compat? Sep 19 14:32:34 ant_work: which bugs are the kexec patches at? Sep 19 14:33:35 at? the patches are attached to @4543 Sep 19 14:33:52 03  07org.openembedded.dev * r3cb2337765 10OE.dev/ (3 files in 2 dirs): webkit-gtk: bump SRCREV Sep 19 14:34:16 there is not a specific bug, just koen wiped cmdline for all machines updating defconfigs Sep 19 14:34:24 03  07master * r3cb2337765 10OE.dev/ (3 files in 2 dirs): webkit-gtk: bump SRCREV Sep 19 14:40:33 Hello, according to the machine configurations files...oe works on several mobiles phones...but would it work on voip phones with some adaptations(own machine configuration and own distro if needed)? i'd like to have openmoko->openvpn->asterisk<-openvpn<-voip-phone...which ones are cheap and avaliable(i do not need fancy stuff...only ethernet and phone function...but there musn't be some proprietary kenrel modules...) Sep 19 14:48:29 Gnutoo : work on all device you support Sep 19 14:48:58 ok...but i don't know which device to buy... Sep 19 14:49:36 thx you gave me an buisness idea Sep 19 14:49:36 :o) Sep 19 14:50:02 jeremy_laine, what package is having issues with libusb-compat? Sep 19 14:50:25 Crofton: librfid looks like it's having issues Sep 19 14:51:03 Crofton: it builds fine, but my applications no longer work and I think the last time they worked was when using libusb instead of libusb-compat Sep 19 14:51:18 for now I'd just like to be able to test whether this is indeed the issue Sep 19 14:52:07 there is no easy way Sep 19 14:52:30 Crofton: arg Sep 19 14:52:42 but you can change the dependency in a couple of packages and it should built the old way Sep 19 14:52:59 basically I am running a 2.6.26 kernel and I read some messages saying libusb1 is broken on these kernels Sep 19 14:53:03 I have to do this with gnuradio , as they use libusb private structs Sep 19 14:53:52 edit bluez and usbutils Sep 19 14:53:58 and whatever you use Sep 19 14:54:10 and it should build Sep 19 14:54:42 you might also check with the libusb guys Sep 19 14:55:01 anything appearing on oe.dev may be second/third hand :) Sep 19 14:55:48 thesing: found it? Sep 19 14:56:49 * mwester thinks we may indeed be forced to do the virtual/libusb thing. Sep 19 14:57:07 I would be happier with that Sep 19 14:57:24 I am just not good at virtual stuff Sep 19 14:57:41 I'm not good at anything that's not a localized change. Sep 19 14:57:50 that too :) Sep 19 14:58:17 ant_work: yes but not sure about them. I would rather add a the CMDLINES (taken from defconfigs) for all supported machines to linux-kexecboot.inc Sep 19 14:58:21 ;) And if it might bleed through to Angstrom, I have to ask myself if I'm prepared to have my head handed back to me on a platter. Sep 19 14:58:27 I also need to figure out how to fix gnuradio, because I think it would benefit from the newer libusb Sep 19 14:58:49 thesing: on angstrom-dev ML koen said this is wrong place Sep 19 14:58:55 well, if you break things for loads of people a beating is in order :) Sep 19 14:59:33 thesing: and indeed I moved to linux-kexecboot.inc Sep 19 14:59:42 Well, yes, but most times it's better if one is told what the beating is for. Sep 19 15:00:22 thesing: please restore the *normal* CMMDLINE_c7x0 if you edit linux-kexecboot.inc Sep 19 15:00:57 But I think something needs to be done about libusb; the "fix" for SlugOS is not a fix at all -- we just exclude the packages that will cause both libusb and libusb-compat to be built. Rather a fragile solution, to be sure. Sep 19 15:01:26 thesing: and don't forget to change 'char *kexec_exec_argv[] = { NULL, "-e -x", NULL};' Sep 19 15:01:30 ant_work: he said that linux.inc is the wrong place. And you only added the cmdlines for for some machines. And only the console= part. (but maybe they are this short) Sep 19 15:01:51 it is enough (for akita/c7x0) Sep 19 15:02:17 supposedly collie/poodle/tosa need a cmdline too Sep 19 15:03:25 thesing: ^^ and split the argv +const char *kexec_exec_argv[] = { NULL, "-e", "-x", NULL}; Sep 19 15:03:42 (though it works with an unified arg) Sep 19 15:03:57 I will only add the cmdlines. Sep 19 15:04:01 that's all Sep 19 15:04:45 thesing: whitout -x the screen is corrupted before kexec... Sep 19 15:10:12 is anyone building for sh4 with OE? Sep 19 15:15:02 ant_work: thats a bug but not critical. Sep 19 15:21:39 the normal z? Sep 19 15:21:43 should work with dosbox Sep 19 15:21:46 oops wrong window Sep 19 15:23:44 thesing: no, is not critical, but was not evident with your defconfigs (net enabled) Sep 19 15:23:58 why let it there ? Sep 19 15:24:40 because I don't have time to test it. Sep 19 15:24:54 fair enough :-) Sep 19 15:25:41 (BTW I have logs with and without errmsg, after fix, if you need) Sep 19 15:27:41 then, the open thing remains +require conf/machine/include/kexecboot.inc in machine.conf (or even zaurus-2.6.inc) Sep 19 15:27:55 I committed some cmdlines and hope that it works again now. Sep 19 15:28:03 great, thx Sep 19 15:29:41 in a recipe, how would i make sure the recipe is only compatible with a range of linux kernel versions? Sep 19 15:31:23 tharvey: I don't think there is an easy way of doing this. The best thing you can do is adding a comment about it. Sep 19 15:32:55 tharvey: in angstrom-2008.1.conf there are some hints I think Sep 19 15:33:06 KERNEL = "kernel26" Sep 19 15:33:15 MACHINE_KERNEL_VERSION = "2.6" Sep 19 15:33:52 and even an #Horrible workaround for armv7a Sep 19 15:33:53 thanks, I'll look into it more Sep 19 15:36:04 but thats only for 2.4 vs 2.6 not about specific kernel versions Sep 19 15:36:05 thesing: btw there are two references to initramfs in that angstrom-2008.1.conf (I think legacy/bitrot from pfalcon work) Sep 19 15:36:35 ah, was the uclibc try... Sep 19 15:37:10 before klibc ;-) Sep 19 15:37:25 I think they are still needed for linux-hh kernel Sep 19 15:38:04 thesing: I think you made an outstanding work! The world must know ;-D Sep 19 15:39:07 basically every device (even with sane u-boot) could have a nice graphical bootmenu in ~1Mb image... Sep 19 15:39:48 ant_work: thanks. But I only rewrote the stuff from pfalcon and added some candy. Sep 19 15:40:17 well, let say entirely rewritten... Sep 19 15:40:49 btw do you think it would be possible to boot two jffs2 images on the same SD partition? Sep 19 15:41:08 s/boot/distinguish at boot/ Sep 19 15:42:59 distinguishing would be no big problem. But I don't know if linux can boot with an images as rootfs. Sep 19 15:43:12 well, as from nand Sep 19 15:44:27 Crofton: I have started putting together a patch for virtual/libusb Sep 19 15:45:11 jeremy_laine: hey, I have something for you too (oestats) Sep 19 15:45:13 Crofton: one question though, if I put PROVIDES += "virtual/libusb" in libusb_0.1.12.bb, libusb-native_0.1.12.bb gets it too, so I have to do a PROVIDES = "" Sep 19 15:45:21 Crofton: is there a better way? Sep 19 15:45:36 ant_work: honestly recently I've been too busy to touch oestats much Sep 19 15:45:51 ant_work: I found time to switch it to django 1.0 yesterday but that's about it Sep 19 15:45:58 I see, it's about a missing log/wrong link Sep 19 15:46:58 http://tinderbox.openembedded.net/packages/134157/ Sep 19 15:47:25 ah Sep 19 15:47:34 could be a side effect of my recent changes Sep 19 15:47:46 actually the package builds... Sep 19 15:47:57 ant_work: not sure. You would need to mount the image with loopback device. Sep 19 15:48:12 as in altboot Sep 19 15:48:15 I see now Sep 19 15:48:56 but even, how to reach the kernel in /boot? Sep 19 15:49:23 one would need to mount it with -o loop Sep 19 15:49:27 thesing: workaround: multi-partition on SD (works fine) Sep 19 15:49:56 and uncompressed ext2 is fast enough Sep 19 15:50:21 03  07org.openembedded.dev * rf47d150d3c 10OE.dev/packages/linux/ (linux-2.6.25/mpc8313e-rdb/defconfig linux_2.6.25.bb): linux-2.6.25: resync mpc8313e-rdb defconfig with 2.6.26 Sep 19 15:50:22 03  07org.openembedded.dev * ra4209bde48 10OE.dev/: Sep 19 15:50:22 merge of '5245fa92de31b9af071626ab0aa8ee1699126429' Sep 19 15:50:22 and 'd589ae71d47d05c2754a75c5d7d6af2d12e8c58f' Sep 19 15:50:22 03  07org.openembedded.dev * rb66f3f5009 10OE.dev/packages/kexecboot/ (3 files): linux-kexecboot: add cmdlines Sep 19 15:50:27 03  07master * rf47d150d3c 10OE.dev/packages/linux/ (linux-2.6.25/mpc8313e-rdb/defconfig linux_2.6.25.bb): linux-2.6.25: resync mpc8313e-rdb defconfig with 2.6.26 Sep 19 15:50:29 03  07master * ra4209bde48 10OE.dev/: Sep 19 15:50:31 merge of '5245fa92de31b9af071626ab0aa8ee1699126429' Sep 19 15:50:35 and 'd589ae71d47d05c2754a75c5d7d6af2d12e8c58f' Sep 19 15:50:40 03  07master * rb66f3f5009 10OE.dev/packages/kexecboot/ (3 files): linux-kexecboot: add cmdlines Sep 19 15:51:57 thesing: with the cheap 4Gb sd cards out now, one normal user won't need compressed fs I think... Sep 19 15:52:58 it's even simpler for customization (/etc/fstab) Sep 19 15:53:38 yes, I forgot the issue: one would comment out /, /home and perhaps /dev/mmcblk0p1 in cmdline Sep 19 15:53:49 s/cmdline/fstab Sep 19 15:54:13 otherways you boot e.g. GPE with Opie /home...very bad! Sep 19 15:55:31 yes. we need to think about a solution for this. Sep 19 15:56:00 this must be deided at compile time Sep 19 15:56:13 in linux.inc there are exemples (cmdlines) Sep 19 15:56:44 03  07org.openembedded.dev * r4bb03f91f3 10OE.dev/packages/dosbox/ (dosbox_0.72.bb files/gcc-4.3.patch): dosbox: fix gcc 4.3.x issues Sep 19 15:56:50 03  07master * r4bb03f91f3 10OE.dev/packages/dosbox/ (dosbox_0.72.bb files/gcc-4.3.patch): dosbox: fix gcc 4.3.x issues Sep 19 15:56:58 thesing: I think all should be in machine.conf Sep 19 15:57:24 a new flag e.g. 'BOOT_FROM' Sep 19 15:57:41 or even in local.conf Sep 19 15:58:38 It would be better if you could use the same images for everything. Sep 19 16:08:10 thesing: then you have to hack fstab after install, or you make assumptions (/home in root) Sep 19 16:10:07 well, we already assume /boot in root Sep 19 16:10:53 yes. I'm not sure if other machines have a different partition for /home Sep 19 16:11:33 if you boot other distros from SD it does not make sense imho Sep 19 16:13:39 * ant_work is used to having distinct partitions for /boot /home /var /usr /usr/local on production servers Sep 19 16:17:53 For devices with enough flash (say enough to flash an x11-image) I think we should assume flash as default root and have fstab accordingly. For other devices (like collie) we only support root on sd and fstab should reflect this. Sep 19 16:18:25 +1 from me Sep 19 16:19:12 even on collie it might be interesting to use internal flash as /home (and sd as / ) Sep 19 16:19:46 what happens if you boot let say debian on another SD? Sep 19 16:20:18 home would be horribly overwritten Sep 19 16:21:18 what if I want a minimal/console image for collie? can it be in flash? Sep 19 16:21:21 I was discussing with JaMa about SL-C3000: it has only 16M flash (7+5+4) Sep 19 16:22:39 supposedly a minimal image with busybox could be ~2.6 M Sep 19 16:22:51 denix: yes. But current flash driver can't write to flash. There is a kernel patch to add an old driver that works better, but its not in oe. Sep 19 16:23:24 denix: its easier to use sd or cf as root. Sep 19 16:23:46 sd is fine, I guess, but I need cf for wifi... Sep 19 16:24:52 tthesing: I'm thinking about using dm_crypt for the nand space and have my reserved data there Sep 19 16:25:01 denix: http://www.linuxtogo.org/~thesing/ has a prebuild image. Sep 19 16:25:16 tharvey: thanks! Sep 19 16:25:54 ant_work: bad idea. your flash will die. Sep 19 16:26:30 I'm pretty sure equally aged SD will die before... Sep 19 16:27:01 ant_work: encrypt a file, change the (original) file, encrypt it again and compare the two files. Sep 19 16:27:18 you will find out that the encrypted versions are completely different. Sep 19 16:28:11 This only matters if the file is bigger than one eraseblock of the flash but one should have it in mind. Sep 19 16:29:46 thesing: oneday someone will suggest to use it as swap ! Sep 19 16:29:56 what task/step creates the .debug directories off files placed in image? Sep 19 16:30:30 tharvey: the installation of foo-dev packages Sep 19 16:30:46 I'm placing some pre-built unit-tests in image and packaging them but something is adding .debug dirs causing insane.bbclass to complain Sep 19 16:36:07 are you sure it's debug dirs? Sep 19 16:36:15 And not say, foofile.debug ? Sep 19 16:36:24 ya, its .debug/ Sep 19 16:36:46 the stage where we do the stripping is what does that Sep 19 16:36:55 you can kludge around with INHIBIT_PACKAGE_STRIP=1 or so Sep 19 16:37:13 so what I have is a set of unit tests from vendor that contain scripts, exe's and ko's - so I'm trying to create a package within the kernel recipe for them (as they are dependent on that kernel) Sep 19 16:37:44 hmm, ok Sep 19 16:37:52 and you're 100% positive it's not say a Kconfig.debug file? Sep 19 16:38:01 I hit that when I hacked in locally a kernel-src package Sep 19 16:38:18 had to fixup insane.bbclass to only complain on ends in .debug and is a dir Sep 19 16:39:05 (I didn't submit that as there can be foo.debug debug files that should in fact be errored out on) Sep 19 16:39:45 bbl, bye Sep 19 16:43:38 Tartarus, ya, it looks like for every .ko and every exe a .debug directory is created perhaps with debug versions of those - what I really want is for that directory (image/unit_tests) to not be touched Sep 19 16:44:53 i'd just not strip in the kernel Sep 19 16:45:01 since you can't really do that anyhow (and OE doesn't really) Sep 19 16:46:05 Tartarus, ok I'll try that Sep 19 16:46:36 I'm thinking that perhaps putting the package in the kernel recipe is not the right thing to do anyway - perhaps I should have a stand-alone recipe that requires that specific kernel version or something Sep 19 16:50:25 mr_nice: hi Sep 19 16:50:37 thesing hi Sep 19 16:53:49 thesing: do you know if it is possible to change a line of code in a sourcefile inside oe temp workdir and recompile it without overwrite the source? Sep 19 16:54:56 mr_nice: it is. bitbake -f -c compile foo Sep 19 16:55:40 mr_nice: that will compile it. After that "bitbake foo" will to everything else needed Sep 19 16:59:05 thesing: ok thanks Sep 19 17:07:36 Tartarus, worked fine to strip in my kernel recipe thx Sep 19 17:29:26 thesing: can you take a look at http://wiki.openembedded.net/index.php/Bitbake_cheat_sheet ? I just added this page maybe you know some other useful commands? Sep 19 17:31:45 mr_nice: not really. I corrected an error (missing compile). Maybe "bitbake -c clean foo; bitbake foo" should be there. But it is quite obvious. Sep 19 17:33:18 thesing: thanks. I will add the others as well Sep 19 17:40:22 * * OE Bug 4556 has been created by gregoiregentil(AT)gmail.com Sep 19 17:40:24 * * gst-ffmpeg build error due to missing dependency Sep 19 17:40:25 * * http://bugs.openembedded.net/show_bug.cgi?id=4556 Sep 19 17:42:22 * * OE Bug 4557 has been created by gregoiregentil(AT)gmail.com Sep 19 17:42:24 * * Clutter do_configure error due to missing dependency Sep 19 17:42:26 * * http://bugs.openembedded.net/show_bug.cgi?id=4557 Sep 19 17:43:01 ok added Sep 19 17:49:21 * * OE Bug 4558 has been created by gregoiregentil(AT)gmail.com Sep 19 17:49:23 * * Pidgin-2.5.0-r0 do_configure fails because of missing dependency or bad OECONF Sep 19 17:49:26 * * http://bugs.openembedded.net/show_bug.cgi?id=4558 Sep 19 18:21:27 03  07org.openembedded.dev * r2045ec3a28 10OE.dev/: Sep 19 18:21:27 merge of '26da1e390ef5f5b958b6660965006568272d8f67' Sep 19 18:21:27 and '4e8627d0e74554ef7b775c83e08f37f0fa0921a1' Sep 19 18:21:27 03  07org.openembedded.dev * rbb6e363435 10OE.dev/packages/python/python-pygobject_2.14.2.bb: python-pygobject 2.14.2 increase PE to fix feeds, 2.15.x was premature Sep 19 18:21:28 03  07master * rbb6e363435 10OE.dev/packages/python/python-pygobject_2.14.2.bb: python-pygobject 2.14.2 increase PE to fix feeds, 2.15.x was premature Sep 19 18:21:31 03  07master * r2045ec3a28 10OE.dev/: Sep 19 18:21:33 merge of '26da1e390ef5f5b958b6660965006568272d8f67' Sep 19 18:21:35 and '4e8627d0e74554ef7b775c83e08f37f0fa0921a1' Sep 19 18:28:21 * * OE Bug 4559 has been created by jin(AT)mediatomb.cc Sep 19 18:28:23 * * building for sh4 / uclibc causes an exception in insane.bbclass Sep 19 18:28:25 * * http://bugs.openembedded.net/show_bug.cgi?id=4559 Sep 19 18:30:41 how does it work again, I was away for quite a while, am I supposed to send an email to oe-stable ml if I submit a bug vs. it? Sep 19 18:31:17 probably a good idea Sep 19 18:32:14 k Sep 19 18:34:10 is there some doc on the keywords for the subject? i.e. - anything special I have to [add] so ppl know its a bugreport/patch? Sep 19 18:38:58 never mind, found something useful in the archives Sep 19 18:51:46 re Sep 19 18:53:39 hi woglinde Sep 19 19:05:41 hello, i have a very simple package: xboard...the only thing to do is to change the install dir,i tried EXTRA_OECONF += "--prefix=${D}" but it still wanted to install it in /usr/bin of gentoo Sep 19 19:06:51 Gnutoo: you'll have to inspect the makefile to see what controls the installation directory. Sep 19 19:06:59 normally there's a variable like DESTDIR for that. Sep 19 19:07:07 pb___, ok thanks Sep 19 19:07:14 --prefix is, generally, the runtime thing which you don't want to change Sep 19 19:08:03 ah ok Sep 19 19:08:04 thanks Sep 19 19:10:13 i think it's bindir...thanks Sep 19 20:09:23 re Sep 19 20:09:29 i have thinfgs such as $(INSTALL_PROGRAM) -s xboard $(bindir)/xboard Sep 19 20:09:46 and there is no DESTDIR variable Sep 19 20:09:59 only --exec-prefix and --prefix Sep 19 20:10:25 re Sep 19 20:16:50 woglinde, hello, do you know what to do in case of $(INSTALL_PROGRAM) -s xboard $(bindir)/xboard in the Makefile? Sep 19 20:20:49 delete the Makefile and write an new one Sep 19 20:20:55 this I mean serious Sep 19 20:22:05 ok i'll patch it cause i do not know how to write makefile...and it uses autotools Sep 19 20:23:26 ca i do a patch during do_make() ? Sep 19 20:25:01 hm no Sep 19 20:25:12 if it uses autotools it should be easy to fix Sep 19 20:25:32 mostly Sep 19 20:25:47 ok i'll look in configure.in Sep 19 20:25:56 but I know there are some perverted they dont understand the right the way of using autotools Sep 19 20:26:06 *shrug* Sep 19 20:26:19 that means old and rotten autotool code Sep 19 20:26:31 oops... Sep 19 20:26:42 made a confusion sorry...Makefile.in Sep 19 20:26:43 hm $(INSTALL_PROGRAM) -s xboard $(bindir)/xboard is rather coming from Makefile.in or Makefile.am Sep 19 20:26:48 args Sep 19 20:26:59 Makefile.in means this package is really evil Sep 19 20:27:03 From Makefile... Sep 19 20:27:34 s/made a confusion sorry...Makefile.in/ok i'll look in Makefile.in not in configure.in Sep 19 20:29:59 http://pastebin.com/m6cfb3820 Sep 19 20:30:19 ok i'll patch it... Sep 19 20:30:40 what is xboard for? Sep 19 20:31:03 haha Sep 19 20:31:11 I would suggest run autoscan Sep 19 20:31:22 and write a new Makefile.am Sep 19 20:34:24 ok...xboard is for chess...there is already phalanx,gnuchess etc...in oe Sep 19 20:34:29 but no frontend Sep 19 20:34:51 hm are i remeber Sep 19 20:34:54 yes Sep 19 20:58:22 * * OE Bug 4560 has been created by denis(AT)denix.org Sep 19 20:58:24 * * gcc-cross-*_csl-* cleanup after "include-fixed" fix in glibc Sep 19 20:58:26 * * http://bugs.openembedded.net/show_bug.cgi?id=4560 Sep 19 21:08:07 re ant Sep 19 21:08:44 hi Henning Sep 19 21:10:44 huh, attachment #9999 - that's nice :) Sep 19 21:10:57 denix *g* Sep 19 21:16:12 mickey|sofa: do you know what speed the main memory in the gta01 runs at? just out of curiosity I did an lmbench run on my phone and I was a bit surprised to find that the results were significantly worse than for my old h3600 back in 2002. Sep 19 21:16:38 not sure if that's an instrumentation problem, or slow hardware, or something to do with the openmoko kernel, but it'd be interesting to know what's going on with it. Sep 19 21:17:31 pb hm I guess kernel Sep 19 21:22:14 woglinde: no DSME on garage yet Sep 19 21:22:27 woglinde: but the wifi driver is there Sep 19 21:22:31 yes Sep 19 21:22:34 as I said Sep 19 21:22:40 even specifications (at linuxwireless.org) Sep 19 21:22:43 very nice Sep 19 21:22:48 git archive will follow Sep 19 21:23:28 too bad. when the new omap3-based hw will be released they have a new non-free component: the 3d graphics driver Sep 19 21:24:00 hm Sep 19 21:24:04 thats the game Sep 19 21:26:30 woglinde: at least this driver isnt critical Sep 19 21:27:30 and additionally the kernel module for the chip will be free with a non-free userspace driver. so at least we are not nailed to one kernel version Sep 19 21:28:34 pb_: I wonder if the clocks are slower to save power..... Sep 19 21:28:51 hi bluelightning Sep 19 21:28:55 lrg lol Sep 19 21:29:04 neo and powersaving Sep 19 21:29:07 good joke Sep 19 21:29:08 heh Sep 19 21:29:23 lrg: yah, I wondered that, but the cpu main clock doesn't seem to be any slower. Sep 19 21:30:26 hi all, woglinde Sep 19 21:30:37 lmbench says 261MHz, /proc/cpuinfo says 132 bogomips, so I think it probably is running at 266. Sep 19 21:31:33 pb_: iirc memclk is derived from cpuclk via a divider Sep 19 21:31:43 yeah, it is Sep 19 21:31:52 I guess they could be running the memory at cpu/4 Sep 19 21:31:59 yeah Sep 19 21:33:24 I'm not quite sure how to interpret the lmbench memory numbers to figure that out. Sep 19 21:33:35 pb_: btw, I'm off to visit Jamey tomorrow Sep 19 21:34:28 lrg: ah, very good Sep 19 21:34:37 are you going to Boston? Sep 19 21:38:15 hm vfat15 is slow on my pc Sep 19 21:38:19 fat16 Sep 19 21:50:18 lrg: hm, no, I just noticed this in dmesg: Sep 19 21:50:19 S3C2410: core 266.000 MHz, memory 133.000 MHz, peripheral 66.500 MHz Sep 19 21:50:46 pb seems okay Sep 19 21:52:06 I guess I should benchmark one of our own s3c2410 systems and see how those compare. Sep 19 21:55:49 hi gregoiregentil Sep 19 21:56:24 http://www.fsf.org/news/thank-you-sgi Sep 19 21:56:33 mesa is free ... again :) Sep 19 21:57:25 Hello woglinde Sep 19 21:59:06 I have posted a few bugs (4556, 4557, 4558) with a few suggested fixes. as I said yesterday, I want to help the community. I still have a bunch of packages that don't compile or with some errors (thunar, firefox, gnome-games) Sep 19 22:05:24 good nite Sep 19 22:14:55 gregoiregentil: those look like low hanging fruits Sep 19 22:15:03 I'll check them Sep 19 22:17:44 could we update wesnoth with this bb file? http://pastebin.com/mdb90183 Sep 19 22:19:17 rschuster: Thanks! I have more complex reports, especially for thunar, gnome-games, firefox and xfdesktop. I will post them later Sep 19 22:19:41 if i remember well it was woglinde that helped me a lot (he modified libsdl and others libs to make them bettter ) Sep 19 22:23:08 gregoiregentil: I am interested in the problem report about gnome-games Sep 19 22:23:37 I fumbled a bit to get it working lately (mainly disabling aisleriot because I coult not get guile to build) Sep 19 22:24:40 Gnutoo: does it build? Sep 19 22:25:02 and install, and run correctly? Sep 19 22:25:19 yes it builds i'm testing it...i'm playing against my laptop Sep 19 22:25:21 online Sep 19 22:25:29 Gnutoo: fine :) Sep 19 22:25:50 Gnutoo: I really want to break down wesnoth into smaller pieces like Debian is doing it Sep 19 22:26:01 ok Sep 19 22:26:03 feel free Sep 19 22:26:06 but that is a lot of work and we can do this incrementally Sep 19 22:26:24 ok Sep 19 22:26:44 --enable-tinygui should be dependent on the display size Sep 19 22:26:55 yes...i just realised it Sep 19 22:27:03 it's not perfect...it's my first bb file... Sep 19 22:27:09 beagleboard has 1024x768. :) Sep 19 22:27:17 wow... Sep 19 22:28:27 i'm running it on the openmoko freerunner Sep 19 22:29:05 * rschuster is tempted to build a deb for maemo ... Sep 19 22:29:17 Has anyone tried to compile microb (the lite browser of N810) on OE? Sep 19 22:29:33 and then blog about how cool OE is ;) Sep 19 22:30:20 yup! Sep 19 22:30:23 gregoiregentil: I don't see a recipe for it in OE. so I guess not Sep 19 22:30:42 gregoiregentil: is it autotools based? if you it should be easy to add Sep 19 22:30:49 yup. I searched for it and was not able to find any. Strange because there is a maemo image... Sep 19 22:30:53 (and if all dependencies are already there) Sep 19 22:31:02 Well, I'm taking a look right now from their svn Sep 19 22:31:15 gregoiregentil: there is no image. just a compatibility distro configuration Sep 19 22:31:54 it's really debian focused. I'm on Gentoo so it's a little bit mroe difficult to compile it for me... They have some scripts for non debian but they look strange Sep 19 22:32:21 http://browser.garage.maemo.org/docs/build_howto.html Sep 19 22:32:40 rschuster, wow!!! it even exit correctly... Sep 19 22:32:51 Gnutoo: do you have a nice .xpm image for wesnoth? Sep 19 22:33:12 rschuster, i have ipk but not xpm...what's xpm? Sep 19 22:34:04 Gnutoo: an image format. wesnoth should have an icon for its starter Sep 19 22:34:22 rschuster: it seems that it's autotools based indeed. I would like to try to write a bb file. Can you help me? Give me some directions. Is there a similar package I could use the bb as an example? Sep 19 22:34:29 rschuster, it has an icon Sep 19 22:34:59 rschuster, but the first time you have to launch it from the command line as it guess a 1024x748 display...and i've a 640x480 one Sep 19 22:36:03 Gnutoo: looks like we need a small patch for wesnoth (check for available display size and take one of those instead of hardcoded ones) Sep 19 22:36:30 gregoiregentil: have a look at the recipe Gnutoo just posted Sep 19 22:36:37 http://pastebin.com/mdb90183 Sep 19 22:36:47 wesnoth is also autotools-based Sep 19 22:37:07 just replace every entry with what you need Sep 19 22:37:49 if you take the sources from an SVN you need a special svn-URI inside SRC_URI Sep 19 22:38:10 i prefer using sources that have been released... Sep 19 22:38:42 looks like this: Sep 19 22:38:42 SRC_URI = "svn://svn.gna.org/svn/wesnoth;module=trunk;proto=http" Sep 19 22:39:14 rschuster, that's the old one... Sep 19 22:40:04 rschuster, ah you're talking to gregoiregentil Sep 19 22:40:07 Gnutoo: this was just an example for an SRC_URI with svn in it Sep 19 22:41:40 gregoiregentil: here another one. In case you need more svn arguments: Sep 19 22:41:40 svn://phoneme.dev.java.net/svn/phoneme/builds/phoneme_advanced-mr2-dev-${BUILDREV};module=cdc;proto=https;localdir=cdc;user=guest;pswd='' \ Sep 19 22:59:23 * * OE Bug 4561 has been created by gregoiregentil(AT)gmail.com Sep 19 22:59:25 * * Pidgin 2.5.0-r0 compilation error Sep 19 22:59:27 * * http://bugs.openembedded.net/show_bug.cgi?id=4561 Sep 19 23:18:23 * * OE Bug 4556 has been RESOLVED (FIXED) by thebohemian(AT)gmx.net Sep 19 23:18:25 * * gst-ffmpeg build error due to missing dependency Sep 19 23:18:27 * * http://bugs.openembedded.net/show_bug.cgi?id=4556 Sep 19 23:19:51 rschuster: regarding 4556, thanks! Sep 19 23:23:47 i must go...bye Sep 19 23:28:10 hey, i am trying to get gumstix to boot of sd card, but i get some errors while compiling Sep 19 23:28:33 i added mmcroot to conf/machine/gumstix-custom-verdex Sep 19 23:28:56 and tryed to rebuild task-base-gumstix, but it fail Sep 19 23:30:34 http://rafb.net/p/2Op7XX53.html << error log Sep 19 23:32:08 has anyone set up or is there any examples in OE for setting up a device as a bluetooth NAP? Sep 19 23:33:36 I've done it in the past using PAND but that is not really the correct way to do it these days. I have checked bluez wiki but there is no real good example. Sep 20 00:21:54 03  07master * r6d0b9e510c 10OE.dev/packages/libxml/ (libxml2-native_2.6.32.bb libxml2_2.6.32.bb): Add new recipes. Old recipes do not compile with gcc 4.3. They need backports. Sep 20 00:21:58 03  07master * rae580be15b 10OE.dev/: Sep 20 00:21:58 merge of '1d03ea3d490cbb5935a7a927a9bfd95a70c8599f' Sep 20 00:21:58 and '2551aa2928c2ff58083738f1e4e839273afde6d4' Sep 20 00:21:58 03  07org.openembedded.dev * r6d0b9e510c 10OE.dev/packages/libxml/ (libxml2-native_2.6.32.bb libxml2_2.6.32.bb): Add new recipes. Old recipes do not compile with gcc 4.3. They need backports. Sep 20 00:21:59 03  07org.openembedded.dev * rae580be15b 10OE.dev/: Sep 20 00:22:01 merge of '1d03ea3d490cbb5935a7a927a9bfd95a70c8599f' Sep 20 00:22:03 and '2551aa2928c2ff58083738f1e4e839273afde6d4' Sep 20 01:29:33 * denix is away: I'm not here **** ENDING LOGGING AT Sat Sep 20 02:59:57 2008