**** BEGIN LOGGING AT Tue Jan 08 02:59:56 2008 Jan 08 03:13:55 http://pastebin.com/m391b936a Jan 08 04:25:55 Can anybody pls help me with O.E. Jan 08 08:22:54 rushadrenaline: you probably need to say what kind of help you need Jan 08 09:12:17 SyNko: ciao Jan 08 09:12:30 morning all Jan 08 09:18:05 gm Jan 08 09:21:22 yo ! Jan 08 09:30:09 * openembedded-issues@lists.openembedded.org seems dead ... Jan 08 09:31:29 Any hint about oebug 3635? Jan 08 09:32:28 xserver-kdrive-imageon-1.2.0-r3 compiles only in stable branch Jan 08 09:34:52 apart this angstrom-2008.1 compiles ok for c7x0 Jan 08 09:57:42 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 09:58:56 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:00:02 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:01:14 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:02:03 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:02:56 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:05:08 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:07:28 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:08:16 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:09:08 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:09:46 can anybody tell me ,,what actually is the concept of angstrom and c7x0 Jan 08 10:09:48 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:09:58 I mean what do they specify Jan 08 10:10:49 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:11:33 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:11:40 rushadrenaline: Angstrom is the distribution, c7x0 is a machine it runs on Jan 08 10:11:41 hi Jan 08 10:11:46 yo Bernardo Jan 08 10:11:59 hi XorA Jan 08 10:12:17 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:12:33 I might see you guys at fosdem - I have a job interview coming up in brussels, with some luck... Jan 08 10:13:37 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:15:05 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:15:58 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:16:50 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:17:30 I am still unable to build glibc locales on stable, with the exact same config that worked on dev Jan 08 10:18:48 http://pastebin.ca/84648 Jan 08 10:19:38 looks a bit like #2683, but happens using akita or simpad targets Jan 08 10:21:32 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:22:38 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:23:38 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:23:47 Bernardo: I solved the qemu/glibc issues just bitbaking gcc-native and then recompiling qemu Jan 08 10:24:18 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:25:13 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:26:09 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:28:00 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:29:13 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:31:42 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:33:11 ant|work: ehyla' Jan 08 10:33:29 morning Jan 08 10:33:38 hi there Jan 08 10:34:08 OE devs present on LinkedIn: I created OpenEmbedded group. You can join here: http://www.linkedin.com/e/gis/49269/0321B4146D12 Jan 08 10:34:15 Jan 08 10:34:18 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:35:02 ant|work: for BB issues ive studied a little the code yester night.. hey seems alienized! really a cool code.. Jan 08 10:36:05 ant|work: now im locked in undestanding the BB works.. nice! if you want help me , you are wellcome :) Jan 08 10:36:30 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:36:49 hi hrw Jan 08 10:37:16 that is only for devs or for users too? Jan 08 10:37:31 Bernardo: feel free Jan 08 10:37:34 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:40:45 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:41:42 hrw: thanks, done Jan 08 10:43:07 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:44:03 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:46:18 03freyther 07org.oe.dev * r24407917... 10/ (1 contrib/mtn2git/mtn2git.py): contrib/mtn2git: Start with only using the manifests Jan 08 10:49:38 in what recipe could I add a module alias definition for my machine (like alias g_ether usb0) ? Jan 08 10:50:16 kernel one would be best Jan 08 10:58:15 hrw: do you have an example where this is done ? Jan 08 10:59:02 Longfield: packages/linux/linux-rp.inc maybe? Jan 08 11:00:46 .XorA : Suppose that in my case i want to build a package for mipsel , and my host arch is x86 . So would Angstrom, c7x0 work in may case. Jan 08 11:01:07 rushadrenaline: no, c7x0 is not mipsel Jan 08 11:01:17 rushadrenaline: and I cant offhand remeber which machines are Jan 08 11:03:07 wrt54 is mipsel Jan 08 11:03:50 And what bout angstrom Jan 08 11:03:55 is it ok for my case Jan 08 11:04:16 it is well tested Jan 08 11:04:24 not for mipsel anyway Jan 08 11:07:25 So hrw what to use for mipsel Jan 08 11:07:59 hrw: there it defines modules that are autoloaded ... I could already do this in the machine.conf file. It would be cleaner if I could only add an alias, so that networking service load the right kernel module Jan 08 11:08:23 You have an idea if this was already done in OE ? Jan 08 11:08:32 (I'm pretty sure that yes) Jan 08 11:12:11 rushadrenaline: angstrom anyway Jan 08 11:12:25 Longfield: checked classes/kernel*? Jan 08 11:13:28 hrw: no. I found also that there is a files directory in the modutils package, this could also be a good place to put this Jan 08 11:14:39 ok hrw i have some problems understanding the working of OE Jan 08 11:14:47 suppose that i do bitbake nano Jan 08 11:15:35 Then what actualy is o.e. doing,,,is it downloading all dependencies for nano,,and the building one single nano.rpm Jan 08 11:16:06 rushadrenaline: it builds everything needed to build nano and nano itself Jan 08 11:16:33 so the dependencies are built seperately Jan 08 11:16:38 rushadrenaline: result will be ipkg packages Jan 08 11:16:44 during that build Jan 08 11:16:45 yes Jan 08 11:17:03 what is ipkg Jan 08 11:18:12 hrw: there is also something in kernel.bbclass that's true Jan 08 11:19:07 module_conf_rfcomm = "alias bt-proto-3 rfcomm" Jan 08 11:19:27 and what about root file system Jan 08 11:20:53 ~ipkg Jan 08 11:20:54 ipkg is probably the Itsy Package Management system: http://www.handhelds.org/z/wiki/iPKG or find the source in the handhelds.org CVS http://cvs.handhelds.org/cgi-bin/viewcvs.cgi/familiar/dist/ipkg or see the howto at http://docs.zaurus.com/ipkg_howto.shtml Jan 08 11:22:16 for instance ... so this could become module_conf_g_ether = "alias usb0 g_ether" Jan 08 11:25:28 Longfield: iirc yes Jan 08 11:25:44 rushadrenaline: look into packages/images/ - those are recipes for images (rootfs) Jan 08 11:26:18 hrw: and modutils won't work for me: I have a 2.6 kernel and iirc modutils was for 2.4 kernels right ? Jan 08 11:26:57 yes Jan 08 11:27:03 module-init-tools are for 2.6 Jan 08 11:27:20 I have to admit that I never played with modules configs Jan 08 11:27:48 So while generating rootfs using o.e. ,,we are actually just populating the whole file system hierarchy Jan 08 11:27:51 ?? Jan 08 11:29:47 rushadrenaline: OE builds software, package it and then create rootfs using those packages Jan 08 11:30:15 sorry for short answers and lag but I am at work and have few megabytes diff to adapt Jan 08 11:31:17 ok thanx man . Are the packages rpm in format Jan 08 11:33:21 no - they are in ipkg format (which is same as Debian one) Jan 08 11:33:27 ok Jan 08 11:33:57 there is rpm support in OE but no one use it so it is hard to tell does it works at all Jan 08 11:34:22 rpm was Really Poorly Maintained in last few years... Jan 08 11:34:46 hrw: we should just make an alien-native :-D Jan 08 11:34:55 ;D Jan 08 11:35:03 XorA: alien-native use rpm-native Jan 08 11:35:14 and then comes problem... which rpm? Jan 08 11:35:32 Fedora? Mandriva? RHEL? Debian? PLD? Jan 08 11:35:48 each distro has own rpm version/patchset Jan 08 11:36:11 http://blog.haerwu.biz/2008/01/07/how-to-replace-zaurus-pda/ Jan 08 11:36:43 hrw: your pimping that article :-) Jan 08 11:37:47 yes ;D Jan 08 11:38:09 want to gather misc opinions before will try to spend cash Jan 08 11:39:21 hrw: nokia suck ass, n810 uses miniSD Jan 08 11:39:50 XorA: nokia? Jan 08 11:40:22 hrw: them who designed n810 Jan 08 11:40:48 ah.. Finnish company ;) Jan 08 11:41:12 main importer of toilet paper to Irland in 80s Jan 08 11:50:43 what is an opie / gpe image Jan 08 11:55:21 anybody??? Jan 08 11:55:34 what is an opie / gpe image Jan 08 11:59:42 rushadrenaline, i really suggest you to read http://www.openembedded.org/wiki/GettingStarted Jan 08 12:00:09 and anyway take some time to read something about OE in the site ... Jan 08 12:01:04 Yeah i was reading Jan 08 12:01:18 but theres nothing about opie/gpe image Jan 08 12:06:55 It's something you install on your PDA which provides a usable environment, using the OPIE or GPE graphical user interfaces. Jan 08 12:07:21 Although the definition of "usable" differs from person to person. Jan 08 12:07:32 *looks at OE* Jan 08 12:08:27 ~opie Jan 08 12:08:28 well, opie is the Open Palmtop Integrated Environment. More info at http://opie.handhelds.org/ or in the #opie channel. Installation instructions at http://opie.handhelds.org/wiki/index.php?OpieInstallation or respect, API http://handhelds.org/~zecke/apidocs/, or a nose-hair trimming device invented by Steve Ballmer in his high-school days Jan 08 12:08:28 ~gpe Jan 08 12:08:29 well, gpe is the Great Palmtop Environment and can be found at http://gpe.handhelds.org - Join #gpe to chat about it. Jan 08 12:08:56 ibot: no, gpe is the Great Palmtop Environment and can be found at http://gpe.linuxtogo.org - Join #gpe to chat about it. Jan 08 12:08:56 And the images provide them. Jan 08 12:08:56 okay, hrw Jan 08 12:09:14 haha Jan 08 12:09:25 ibot has gone to the other side! Jan 08 12:13:03 ~corporate Jan 08 12:13:04 i guess corporate is a mispelling of corrupt Jan 08 12:13:37 ibot: no, corporate is an evil system that transforms human in slave Jan 08 12:13:38 Genesis: okay Jan 08 12:13:41 :)) Jan 08 12:15:26 good morning Jan 08 12:15:38 hi all Jan 08 12:16:18 could someone have a look at the two small patches and whether they are ok for commit? http://bugs.openembedded.org/show_bug.cgi?id=3384 Jan 08 12:18:36 rschuster: sorry but thats related to one of Finnish companies. I cant review it Jan 08 12:19:03 rschuster: Whats the difference between our libtool and the osso one? Jan 08 12:19:55 RP: its much older ;) Jan 08 12:20:33 rschuster: Its patched by maemo or is it a standard libtool? Jan 08 12:21:04 urg Jan 08 12:21:17 why do we need an older libtool? Jan 08 12:21:46 Crofton|work: Note its not libtool-native or libtool-cross Jan 08 12:22:01 Crofton|work: not we. people who use OE to build for Finnish tablets would need it Jan 08 12:22:04 RP, good point .. Jan 08 12:22:25 and I suppose it would only be used be people that try really hard Jan 08 12:22:38 maintenance of older versions of stuff is a feature of OE Jan 08 12:23:44 rschuster: Can you ask about this on the mailing list please, I'm basically ok with it but I think a few things need to be changed. It might me time to create a packages-maemo directory and those files look like they need help from a ,inc file... Jan 08 12:23:58 s/me/be/ Jan 08 12:24:01 RP: Its patched. not so much by the osso team but more by debian already (thats the -6 after 1.5.6) Jan 08 12:24:07 One question, when building mono, where do I need to untar the results of mono-mcs-intermediate? Jan 08 12:24:14 My staging dir? Jan 08 12:24:25 Or the host PC's directories? Jan 08 12:24:38 I get error: No JIT or interpreter support available or selected. Jan 08 12:24:44 * RP suspects the time has come to split up packages a bit Jan 08 12:24:49 re Jan 08 12:25:12 * Crofton|work tends to agree with RP Jan 08 12:25:23 how is a question ... Jan 08 12:25:44 RP: meta/ meta-extras/ meta-openmoko/ meta-xyz/ like in Poky? Jan 08 12:25:50 Crofton|work: packages-gpe, packages-openmoko, packages-maemo, packages-opie ? Jan 08 12:26:02 is their overlap? Jan 08 12:26:09 acakages-core? Jan 08 12:26:16 packages/ simply Jan 08 12:26:33 Crofton|work: any overlap would be acceptable in packages I guess Jan 08 12:26:51 If there is a lot of overlap a new packages dir could be created Jan 08 12:28:04 re Jan 08 12:28:18 wb mickeyl Jan 08 12:28:24 mickeyl: Any thoughts on the above? Jan 08 12:29:21 Our biggest problem is we can't share a .inc file between two packages directories :( Jan 08 12:29:32 or patches Jan 08 12:32:48 how to quickly have my rootfs rebuilt (so that I can check that my changes for my alias are working - see above) ? Jan 08 12:34:31 Longfield: bitbake my-image ? Jan 08 12:35:32 RP: well, in this case, my change did not work :( Jan 08 12:39:43 I have added module_conf_g_ether = "alias usb0 g_ether" to my kernel recipe, but I did not get a file where this alias is present Jan 08 12:41:00 i have no problem with splitting up packages Jan 08 12:41:06 as long as it stays two-dimensional Jan 08 12:41:12 err Jan 08 12:41:27 as long as we don't introduce a deeper hierarchy Jan 08 12:41:27 that is Jan 08 12:42:20 This should probably be rfc'd Jan 08 12:42:45 what I sort of don't like is how will rschuster specific case work Jan 08 12:43:02 yeah Jan 08 12:43:10 where we have libtool-really-old-version under packages-osso Jan 08 12:43:25 and the "real" libtool under packages Jan 08 12:43:52 would you need to treat them as collections with priorities ... Jan 08 12:44:20 RP: as you do it in your kernel recipes, what is needed to add an alias for a MACHINE kernel recipe ? Jan 08 12:47:14 Crofton|work: normal devs will not have -osso in bbfiles Jan 08 12:47:38 true Jan 08 12:52:05 amd gcc-cross-4.2.2 still does build for me Jan 08 12:53:53 * Crofton|work is still at the pull daily and try again stage though Jan 08 12:58:32 Longfield: offhand I'm not sure what you're trying to do or what the solution is, sorry Jan 08 13:00:06 ibot: seen koen? Jan 08 13:00:10 koen was last seen on IRC in channel #oe, 26d 2h 53m 47s ago, saying: 'I forgot to commit that portion yesterday'. Jan 08 13:00:16 "Please install following missing utilities: GNU Fortran Compiler" Jan 08 13:00:18 hmm Jan 08 13:00:19 which one? Jan 08 13:00:22 77 or 95? Jan 08 13:00:39 looks like I will stick with Poky Jan 08 13:01:03 * RP is rather against having that as a dependency :( Jan 08 13:01:11 me too Jan 08 13:01:15 who added that and why? Jan 08 13:01:24 and will wonder when OE will start demanding 3 virgins to run one build Jan 08 13:01:35 mickeyl: You can't cross compile fortran without having a native fortran compiler Jan 08 13:01:39 a5c7b007eaa675cc904f5285291d2f8a597efc84 ? Jan 08 13:01:45 and octave needs fortran Jan 08 13:01:49 right Jan 08 13:01:59 and having a native fortran does not always solve the problem ... Jan 08 13:02:14 We could possibly "solve" this with gcc-native... Jan 08 13:02:33 * Crofton|work is lucky he is working on something esle this week .... Jan 08 13:02:50 RP: sorry, I was unclear. I want to add an alias for a module for my machine, and the kernel recipe seems to be a good place to do it using such a line : module_conf_g_ether = "alias usb0 g_ether" Jan 08 13:03:02 At the very least I'd like to be able to set IHATEFORTRANWITHAPASSION=1 somewhere though... Jan 08 13:03:08 :) Jan 08 13:03:40 just run octave through f2c... ;-) Jan 08 13:03:46 I am trying to use qemu for troubleshooting on ubuntu gutsy. "qemu-arm zImage-qemuarm.bin sonkei-dualGUI-image_0.0.1+WIP-20080108-r0_qemuarm.rootfs.ext2" tells me "Error loading zImage-qemuarm.bin" Why is that? Jan 08 13:03:53 rp, that is the default setting of that variable Jan 08 13:04:03 Do I need a special qemu? Jan 08 13:04:26 Laibsch: That is the user mode emulator, try qemu-arm-system Jan 08 13:04:33 or qemu-system-arm Jan 08 13:04:43 or poky-qemu script Jan 08 13:05:14 You really do need a script to call qemu nicely... Jan 08 13:05:31 sudo qemu-system-arm -kernel tmp-poky/deploy/images/zImage-qemuarm.bin -append "psplash=false root=/dev/sda console=ttyAMA0 console=tty0" -net nic,vlan=0 -net tap,vlan=0,ifname=tap0,script=/home/hrw/devel/OH/poky/scripts/poky-qemu-ifup -M versatilepb -usb -usbdevice wacom-tablet -serial stdio -hda poky-image-sdk-qemuarm-20071227233704.rootfs.ext2 -cpu arm1136 Jan 08 13:05:40 Longfield: You changed the .inc file for the kernel, what else did you do? Jan 08 13:06:54 Laibsch: thats one of my start scripts Jan 08 13:07:15 well, good to see that I already have poky-scripts installed ;-) Jan 08 13:07:50 yes, it seems to work now Jan 08 13:08:02 But is it absolutely necessary to run the script as root? Jan 08 13:08:35 Laibsch: It needs root for two things - the frequency it needs timer interrupts at and the network interface Jan 08 13:08:39 if you want networking... Jan 08 13:09:34 yes, I do Jan 08 13:09:45 Well, I guess poky can be trusted Jan 08 13:09:55 So don't mess up my system, OK? Jan 08 13:09:57 ;-) Jan 08 13:13:35 RP: no, I only have simple kernel recipe (.bb) for my machine and only added module_conf_g_ether = "alias usb0 g_ether" in it Jan 08 13:13:42 Is it normal that I get lots of errors from failed accesses to the HD sda? Jan 08 13:15:35 or maybe I should add this to the machine.conf file ? Jan 08 13:16:20 hrw: I think you suggested that I try to use wrt54.conf for a BCM3302 V0.6 based device? Jan 08 13:17:00 Jin^eLD: yes Jan 08 13:17:01 my binary produces an "illegal instruction".. does that sound like - toolchain is not matching the system? Jan 08 13:17:13 maybe Jan 08 13:18:13 hrw: http://pastebin.mozilla.org/299080 Jan 08 13:18:39 this is the output of the "file" tool, the only difference I see is in the version of whatever they mean by "GNU/Linux" Jan 08 13:18:49 does that give any hints? Jan 08 13:18:55 or is this info more or less useless? Jan 08 13:19:44 busybox: ELF 32-bit LSB executable, MIPS, version 1 (SYSV), dynamically linked (uses shared libs), corrupted section header size Jan 08 13:19:46 Longfield: Did you bump the PR of the recipe to rebuild ot? Jan 08 13:19:51 its my wrt54 Jan 08 13:20:04 Jin^eLD: I am not familiar with mipsel Jan 08 13:20:10 first line from my paste is the busybox from target Jan 08 13:20:28 Laibsch: That isn't normal. Which kernel version was that? Jan 08 13:20:38 mhm Jan 08 13:20:43 Laibsch: I have a nasty feeling you need a more recent qemu Jan 08 13:21:02 hrw: and your box also says it's cpu model : BCM3302 V0.6 ? Jan 08 13:21:31 Jin^eLD: BCM3302 V0.7 Jan 08 13:21:46 RP: hum no ... Jan 08 13:21:59 but I have rebuilt it manually Jan 08 13:22:10 mhm Jan 08 13:22:29 Longfield: I really don't remember how the alias stuff works :( Jan 08 13:22:34 guys, anyone here using a BCM3302 V0.6 based hardware? Jan 08 13:23:06 RP: ok, no problem, thank you for your help, I will try to figure out by myself Jan 08 13:24:15 ha! the asus wl500gp should be the same Jan 08 13:25:04 it is Jan 08 13:25:10 OK, now it has worked with the definition in the machine conf file Jan 08 13:25:43 RP: host-kernel 2.6.22-14-generic, qemu 0.9.0+cvs20070613-0oh7fei, OE kernel 2.6.23. The poky images seem to run fine the host qemu. Jan 08 13:25:49 let's see what toolchains openwrt folks use for this asus model Jan 08 13:26:52 seems I can get the toolchai directly from asus Jan 08 13:58:57 ok, back to the fortran situation Jan 08 13:59:02 what do we do about it? Jan 08 13:59:15 and more important... which version is this referreing to? Jan 08 13:59:17 77 or 95? Jan 08 14:01:08 mickeyl: gfortran Jan 08 14:01:13 mickeyl: rather than g77 Jan 08 14:01:35 you don't have to install fortran on host Jan 08 14:02:02 it's just looking for libmpfr, which can be found in staging_libdir Jan 08 14:02:13 well, now you have to, according to the sanity checker Jan 08 14:02:39 blame Laibsch :) Jan 08 14:03:13 which package did he add that for? Jan 08 14:03:26 !oebug 3363 Jan 08 14:03:28 * * Bug 3363, Status: NEW, Created: 2007-11-24 10:32 Jan 08 14:03:29 * * mwester(AT)dls.net: Build fails on gcc-cross_4.1.2 - errors with gfortran Jan 08 14:03:30 * * http://bugs.openembedded.org/show_bug.cgi?id=3363 Jan 08 14:07:27 he add that for exactly this problem, which could be solved other ways Jan 08 14:08:43 then i prefer the other way Jan 08 14:08:43 are we talking about the same thing? http://lists.linuxtogo.org/pipermail/openembedded-commits/2008-January/011432.html Jan 08 14:14:38 adding gfortan does not solce the problem for me ... Jan 08 14:16:11 xjqian|away, where you you apply the "ugly solution"? Jan 08 14:16:34 hi all ... which package contain the scripts to autostart X (Xorg) when in runlevel 5 ??? Jan 08 14:24:20 xjqian|away: You are mistaken. The sanity checker was changed by zecke IIRC. I just made a corresponding change to openembedded-essential Jan 08 14:25:18 and more importantly I have gfortran, but gcc-cross does not compile Jan 08 14:25:29 I am trying the ugly hack added to the gcc-cross recipe Jan 08 14:25:36 morning Jan 08 14:25:39 gm Jan 08 14:26:46 maybe we need to look at/update the gfortran patch Jan 08 14:26:59 * Crofton|work needs to do real work first though Jan 08 14:29:39 zecke, I think the gfortran thing is not required Jan 08 14:29:51 and it does not work on my machine anyway Jan 08 14:30:16 I am trying the ugly hack from bug 3363 Jan 08 14:30:19 Crofton|work: you do? configure of gcc-cross-4.2 fails without it (new angstrom) Jan 08 14:30:45 I am on new ANgstrom Jan 08 14:31:05 the build of gcc-cross fails for me, with gfortran Jan 08 14:31:15 maybe there are two problems? Jan 08 14:32:12 Crofton|work: sadly I'm forced to work now :} Jan 08 14:32:20 yeah me too Jan 08 14:32:27 I am procrastinating Jan 08 14:32:47 I'll update 3363 with my experience though Jan 08 14:33:21 zecke: Doing real work your last month? Jan 08 14:33:31 Crofton|work: yeah, this would be appreciated. your build failures interests me Jan 08 14:33:41 stefan_schmidt: yeah, like finding my password for linkedin Jan 08 14:33:52 zecke: heh Jan 08 14:34:43 bah new failuer ,.. Jan 08 14:35:04 I'm going to have to come back to this .... Jan 08 14:37:20 my daily question: is anyone in the mood of being annoyed by http://bugs.openembedded.org/show_bug.cgi?id=3593 ? :) Jan 08 14:38:15 mailed OEML? Jan 08 14:38:42 hrw: me? Jan 08 14:38:55 you Jan 08 14:39:15 uhm, no, should I? I thought asking around in IRC would kind of be enough Jan 08 14:39:28 Jin^eLD: we are lame Jan 08 14:39:33 zecke :) Jan 08 14:40:09 well, I probably would have mailed if I had found a clean solution, but since nothing came to my mind I hoped to get it solved by annoying people :) Jan 08 14:40:21 /kickban Jin^eLD Jan 08 14:40:25 doh Jan 08 14:40:48 so you want me to be annoying on the ml as well then? :P Jan 08 14:40:59 /kill Jin^eLD :-D Jan 08 14:41:03 hehe Jan 08 14:41:50 well, I hacked in a workaround that does the trick for my build but its very ugly, would be nice to have a clean fix for this Jan 08 14:42:23 and the other problem of course - people who use a static dev might simply not realise that their device table is not being used in the way they thought it woul dbe Jan 08 14:42:35 did cost me some time... Jan 08 14:44:55 post ugly one even Jan 08 14:45:01 so it will be seen what you mean Jan 08 14:45:33 ok Jan 08 14:47:47 hrw: have you merged you tons of diffs? Jan 08 14:47:58 not all Jan 08 14:48:07 building takes time too Jan 08 14:48:08 * Bernardo is still trying to understand why he can't generate glibc locales Jan 08 14:48:13 hrw: ok, later Jan 08 15:13:17 bye all Jan 08 15:13:21 gn Jan 08 15:14:41 rschuster: hey :) Jan 08 15:15:18 zecke: hi Jan 08 15:15:42 rschuster: any idea what they do to their libtool (besides adding osso to the name)? Jan 08 15:17:59 zecke: those are mainly debian patches, the changes made by nokia and not relevant to OE since it affects the debian metadata only Jan 08 15:18:19 one of the patches is: * Apply patch from upstream to allow five digits in version-info. Jan 08 15:18:25 rschuster: why do we need their version then? just the version itself? Jan 08 15:18:37 zecke: yeah, just the version Jan 08 15:18:47 rschuster: then don't name it osso :) Jan 08 15:19:01 rschuster: and go with the DEFAUL_PREFERENCE of -1 (like RP said) Jan 08 15:19:07 zecke: I could live with plain 1.5.6 either but it is not in OE either. would that cause less trouble? Jan 08 15:19:29 rschuster: we had 1.5.6 and moved to 1.5.10 Jan 08 15:20:12 rschuster: (re)add plain 1.5.6, maybe even add the nokia patch, add my cross compile hack and set DEFAULT_PREFERENCE :) Jan 08 15:20:24 zecke: ok Jan 08 15:21:17 zecke: is there a way to retrieve the old 1.5.6 recipe via mtn (like cvs update missing-file)? Jan 08 15:21:47 rschuster: indirectly (using bonsai) Jan 08 15:25:39 rschuster: I think we never had 1.5.6 in OE Jan 08 15:26:28 :) ok Jan 08 15:26:54 http://openembedded.bkbits.net:8080/openembedded/packages/libtool/?PAGE=dir Jan 08 15:26:55 hmm Jan 08 15:27:37 my tool just hated me :) Jan 08 15:27:42 http://www.openembedded.org/bonsai/query/file/libtool-cross_1.5.6/ Jan 08 15:28:25 rschuster: http://www.openembedded.org/bonsai/view/rev/16855/ to see when it was removed Jan 08 15:30:41 otavio: I fixed my mtn2git script, now my ETA is two days :) Jan 08 15:31:25 rschuster: find out the parent revision for d8844564ee3505aa7cebcc46960dfb8c54457ab9 and check out packages/libtool/ for that revision Jan 08 15:32:40 otavio: it is roughly at 3-4 seconds per rev Jan 08 15:57:46 Laibsch: thats easy. mtn 0.37 allows saying p:REVISION to specify a parent :) Jan 08 15:58:53 * Laibsch is still on .34 and .35 Jan 08 16:06:13 * Crofton|work crosses fingers Jan 08 16:19:36 stupid question but task do_package_write_ipk is not build when i make a bitbake foo and i should add explicitly foo to my image recipe to have the ipk Jan 08 16:19:51 what's wrong ? Jan 08 16:27:19 Genesis: that's a feature Jan 08 16:28:55 perharps i could add some option somewhere to do do_package_write_ipk when i bitbake something Jan 08 16:33:28 bbl Jan 08 16:34:04 Genesis: try setting BB_DEFAULT_TASK = Jan 08 16:34:04 "buildall" in local.conf Jan 08 16:34:35 oki i try , thanks Jan 08 16:36:27 I think you can also do: bitbake foo -c buildall Jan 08 16:36:56 not sure it's ok for the dependance Jan 08 16:37:02 with this method Jan 08 16:37:21 !oebug 2952 Jan 08 16:37:23 * * Bug 2952, Status: RESOLVED (FIXED), Created: 2007-09-07 13:16 Jan 08 16:37:24 * * : Some tasks are skipped during build Jan 08 16:37:25 * * http://bugs.openembedded.org/show_bug.cgi?id=2952 Jan 08 16:37:29 Genesis: check that bug Jan 08 16:37:41 ho i look Jan 08 16:55:34 zecke: haha! ok >:-) Jan 08 16:55:41 zecke: thank you very much! Jan 08 17:33:22 mickey|sports: Hey, I've got a couple questions for you when you have a moment Jan 08 18:24:08 !oebug 3647 Jan 08 18:24:10