**** BEGIN LOGGING AT Wed Nov 07 02:59:57 2007 Nov 07 05:34:10 03tmbinc 07org.oe.dreambox * r85a1e0ce... 10/ (3 files in 3 dirs): perl-native: gcc-4.2 buildfix Nov 07 05:34:16 03tmbinc 07org.oe.dreambox * r8f1ea263... 10/ (3 files in 3 dirs): glibc_2.3.2+cvs20040726: fix bashism Nov 07 07:20:07 good morning! Nov 07 07:20:57 steliosk : when you commited the change to the MPC8323E-RDB machine features, you seem to have commited a local change too Nov 07 07:21:18 steliosk : the TARGET_FPU="soft" line is commented out! Nov 07 07:31:52 morning Nov 07 07:34:42 aye' Nov 07 07:36:46 hrm.. adding linux-libc-headers to a angstrom 2007 based image fails.. Cannot find package linux-libc-headers ... hrw?koen? anyone got an idea? :) Nov 07 07:43:15 gilligan_: add 'task-native-sdk' if you want to build on device Nov 07 07:44:43 hrw: ah.. bleh.. had no idea about that one Nov 07 07:45:13 hrw:thanks Nov 07 07:47:28 ERROR: No providers of runtime build target libgcc-s-dev (for ['/home/pflug/projects/oedev/org.openembedded.dev/packages/tasks/task-native-sdk.bb']) Nov 07 07:47:57 *_* Nov 07 07:48:12 now what is that about Nov 07 07:48:33 well, I also have this problem.... Nov 07 07:50:29 hrw: how come 'bitbake linux-libc-headers' works (although it does for some reason not build the prefered version - 2.6.20 instead of 2.6.23) but including it as an image dependency fails? Nov 07 07:53:37 and I am not overly surprised that task-native-sdk won't work because there simply is no libgcc-s-dev package Nov 07 08:00:00 shit. need to update then Nov 07 08:00:08 remove this dep from it Nov 07 08:00:15 ave to go Nov 07 08:48:10 re Nov 07 08:55:27 hey likewise Nov 07 08:55:33 hey Nov 07 09:16:07 good morning Nov 07 09:16:12 moin Nov 07 09:17:05 hey lrg Nov 07 09:17:16 hey XorA Nov 07 09:17:40 XorA: meant to go for Visa this mroning but accident after bridge :( Nov 07 09:17:52 may have to go tomorrow Nov 07 09:21:21 Any idea why in base-files the distro override gets ignored but if i add a machine override gets included ? Nov 07 09:34:14 03ifaistos 07org.oe.dev * r3ceed45c... 10/ (1 conf/machine/mpc8323e-rdb.conf): Nov 07 09:34:14 conf/machine/mpc8323e-rdb.conf : Fix a previus commit where TARGET_FPU was commented. Nov 07 09:34:14 Set TARTGET_FPU to softfloat only if distro/local.conf has not set any other value Nov 07 09:46:18 steliosk: did you try to build a meta-toolchain for powerpc? Nov 07 09:47:11 i tried to build for dht-walnut and it first failed at binutils-sdk with a QA error (fixed that) Nov 07 09:47:36 then it failed at gcc-cross-sdk packaging with the /nof problem Nov 07 09:58:18 did you try with hard-float? Nov 07 09:58:54 nop Nov 07 10:22:05 * * OE Bug 3265 has been created by joaquinduran(AT)adtelecom.es Nov 07 10:22:07 * * Update the version of glibmm library Nov 07 10:22:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3265 Nov 07 10:24:04 * * OE Bug 3266 has been created by joaquinduran(AT)adtelecom.es Nov 07 10:24:06 * * new version of library libsigc++-2.0 Nov 07 10:24:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3266 Nov 07 10:25:07 steliosk: if you have time, I would like to know what the result. Nov 07 11:29:10 mogning Nov 07 12:04:41 hi Nov 07 12:46:20 morning all Nov 07 12:46:33 hey Nov 07 12:49:42 is there an embedded linux distro for the nintendo ds? Nov 07 12:49:48 hey hrw... can you remove libgcc-s-dev from task-native-sdk ? can't push stuff from office and by the time i get home i'll prolly forget about it :) Nov 07 12:50:04 Phoix: googled for it? Nov 07 12:50:14 ok Nov 07 12:51:32 gilligan_: done Nov 07 12:55:05 ;] Nov 07 12:56:33 guys... http://pastebin.org/7321 I was trying to solve this by adding FILES_${PN}-dbg = "${bindir}/.debug/*" but that does not seem to help Nov 07 12:56:36 what am I missing? Nov 07 12:58:25 Jin^eLD: The files aren't installed into bindir, they're installed into /usr/local/... Nov 07 12:58:30 SDK_PREFIX Nov 07 12:58:38 oh, crap Nov 07 12:58:54 That is by design Nov 07 12:59:47 hmmm...... this does not seem right. Nov 07 13:00:01 rm -rf tmp ; bitbake meta-toolchain Nov 07 13:00:22 esben: ah, you also meta toolchaining? :) Nov 07 13:00:25 and after meta-toolchain is build, there are still tasks in progress.... Nov 07 13:00:40 esben: for me the binutils-cross-sdk is failing with the error I pastebinned Nov 07 13:00:59 RP: there is no reference to SDK_PREFIX in the whole packages/binutils/ directory Nov 07 13:01:03 where is it supposed to come from? Nov 07 13:01:29 meta-toolchain does not fail, but I cannot understand why it should pull in recipes which are then allowed to be completed AFTER meta-toolchain is done. Nov 07 13:01:34 Jin^eLD: binutils-cross-sdk defintely uses it. It comes from sdk.bbclass Nov 07 13:01:44 ok let me look there.. Nov 07 13:02:03 03hrw 07org.oe.dev * r58f9cc3c... 10/ (1 packages/tasks/task-native-sdk.bb): task-native-sdk: dropped libgcc-s-dev which is now part of gcc Nov 07 13:02:18 hmm Nov 07 13:02:24 sdk.bbclass:prefix = "${SDK_PREFIX}" Nov 07 13:02:32 so sdk.bbclass also just uses it Nov 07 13:02:39 maybe its set in distro hmm Nov 07 13:02:52 nope Nov 07 13:03:08 ah, its set in bitbake.conf Nov 07 13:03:14 I have almost verified my problems with gcc_4.1.2.bb for ppc603e. It really looks like a bug in libgcc2.c. Nov 07 13:04:26 But luckily, gcc-4.2.2 seems to work :-) Nov 07 13:06:24 hm.. just curious.. does anyone here maintain a local ipkg feed for easy installation/upgrading of packages on the target platform? don't want to reflash my rootfs for each non-trivial change Nov 07 13:06:50 gilligan_: I do, I also use angstrom feeds Nov 07 13:07:08 gilligan_: I tried that once.. Nov 07 13:07:12 and it did not work out for me Nov 07 13:07:18 because of ipkg weirdness Nov 07 13:07:52 hrw: ah of course.. angstrom feeds.. god knows why i didn't think of that right away Nov 07 13:09:16 someone can suggest good amd64 cpu cooler? Nov 07 13:09:55 hrw: I have a zahlman 7000 CuAl or something Nov 07 13:10:02 Very quiet Nov 07 13:10:55 yeah i think those zahlman (or so) are good Nov 07 13:11:12 CM: Orb ones will not fit - I have northbridge cooler near cpu Nov 07 13:11:40 hrw: http://www.zalman.co.kr/eng/product/code_list.asp?code=005 Nov 07 13:11:57 I'm sure at least one model will fit.. they are very good indeed Nov 07 13:13:30 hrw: maybe i am blind.. but neither on the FAQ nor on the angstrom wiki can I find any info on what the feed urls in ipkg.conf should look like... ? :) Nov 07 13:14:20 gilligan_: angstrom adds them into images Nov 07 13:14:39 dammit, which directory is it? http://pastebin.org/7321 I already tried setting various combinations for FILES_${PN}-dbg but still can't get it right :P Nov 07 13:15:31 hrw: well, looking at my ipkg.conf all I see is `lists_dir ext /var/lib/ipkg' (and root definition) Nov 07 13:15:46 hrw: ah..oops.. Nov 07 13:15:49 hrw: subdir.. Nov 07 13:17:21 * cbrake_away wonders if heat is the reason my AMD workstation shuts itself off about once a month ... Nov 07 13:17:26 oh,hm.. apache is only available in unstable Nov 07 13:17:42 cbrake: X_X Nov 07 13:20:04 cbrake_away: my machine cpu fan do 4500rpm now and when cores temp reach 65°C it steps back to 1GHz Nov 07 13:21:55 at 65 already? Nov 07 13:23:17 gilligan_: 60°C now Nov 07 13:32:12 cbrake: I suppose I can change the kernel cmdline from " ... -(root);cm-x270-nand:64m(app),-(data)" to "-(root);cm-x270-nand:128m(app)" in order to reserve all 128M for one partition and i shouldn't have to do other changes elsewhere or would I ? Nov 07 13:34:23 After I typed the command "reparse x11-image", bitbake die with error messages: Nov 07 13:34:29 File "/var/lib/python-support/python2.5/bb/shell.py", line 290, in fileReparse parse.update_mtime( bbfile ) Nov 07 13:34:29 AttributeError: 'module' object has no attribute 'update_mtime' Nov 07 13:34:47 (bitbake 1.8.8 , 1.8.9) Nov 07 13:35:31 How come the update_mtime() was not defined...? Nov 07 13:59:35 doh, this binutils-cross-sdk stuff is driving me nuts Nov 07 13:59:38 hello all Nov 07 13:59:54 it seems such a trivial issue and I still can't figure it out Nov 07 14:00:01 hey likewise Nov 07 14:00:03 Jin^eLD: hacking the toolchain bakes? Nov 07 14:00:16 yeah.. I wanted to use the meta-toolchain package Nov 07 14:00:31 ah yes, haven't tried myself yet - waiting for the dust to settle Nov 07 14:00:33 which seems to work, but the binutils-cross-sdk fails on QA because it packages a .debug directory into a non debug package Nov 07 14:00:35 seems so trivial Nov 07 14:00:44 but goddamn, I think I spent over an hour already Nov 07 14:01:19 I spent 4 hours trying to get a USB -> RS485 dongle to work (but now suspect a bug in that freaking kernel module). Nov 07 14:01:31 heh.. Nov 07 14:01:39 The problem is, I fixed this earlier and even sent a patch somewhere Nov 07 14:01:47 cannot find it though... grr. Nov 07 14:01:59 Jin^eLD: FILES_${PN}-dbg = "${prefix}/whatever/.debug/* Nov 07 14:02:28 RP: either I am an idiot or I never hit the right "whatever", well, probably a bit of both Nov 07 14:03:30 I think whatever = ${HOST_SYS}/bin Nov 07 14:04:19 ok let me try that... Nov 07 14:06:48 RP: nope, that did not help Nov 07 14:07:48 oh wait I maybe missed something Nov 07 14:07:49 Jin^eLD: ok, run bitbake -b /path/to/.bb -e | grep FILES_ Nov 07 14:07:57 Jin^eLD: See what bitbake turns it into Nov 07 14:08:39 that's a good hint indeed, thanks Nov 07 14:16:50 FILES_${PN}-dbg = "${prefix}/${TARGET_SYS}/bin/.debug/*" Nov 07 14:16:53 that was the magic line Nov 07 14:17:03 should go to binutils.inc Nov 07 14:17:24 shall I submit a bugreport or will someone take a look and add it now? Nov 07 14:23:44 cbrake: oh, hm.. no nfs support in kernel? Nov 07 14:25:21 cbrake: did you disable nfs specifically for some reason? Nov 07 14:26:54 morning Nov 07 14:28:07 cbrake: other than the fact that it increases the kernel by 27kb .. Nov 07 14:46:09 good morning all Nov 07 14:46:31 hi koen Nov 07 14:47:03 yo koen Nov 07 14:49:08 morning koen Nov 07 14:49:10 hm.. I added nfs support to the kernel but mount from util-linux doesn't seem to know '-t nfs' ? Nov 07 14:49:13 hey koen Nov 07 14:50:17 could someone please take a look and fix this? http://bugs.openembedded.org/show_bug.cgi?id=2482 it's a one line fix, solution is posted :> Nov 07 14:51:07 seems like something is wrong though.. as nfs does not appear in /proc/nfs and I think it should Nov 07 14:53:13 gilligan_: cat /proc/filesystems Nov 07 14:54:11 hi koen Nov 07 14:54:13 koen: eeh.. typo.. of course i meant /proc/filesystems and not nfs.. and it doesn't appear there. maybe i accidently downloaded the wrong kernel image. rechecking now Nov 07 14:55:35 Jin^eLD: ah, TARGET_SYS. All becomes suddenly clear :) Nov 07 14:56:10 :) Nov 07 14:56:28 I was close :) Nov 07 14:56:32 yes :) Nov 07 15:00:42 koen: seems like I indeed just accidently had flashed the wrong kernel image.. Nov 07 15:05:27 oops.. anyone using portmap on arm/xscale here? I get a segfault on /etc/init.d/portmap start Nov 07 15:06:05 gilligan_: I've been using portmap -- other than taking forever to start, I've not had any problems Nov 07 15:06:27 root@cm-x270:~$ start-stop-daemon --start --quiet --exec /sbin/portmap Segmentation fault Nov 07 15:06:56 that's from NOR/uclibc Nov 07 15:08:02 gilligan_: ahh, I don't use portmap in my uclibc builds Nov 07 15:08:04 of course i have no strace/gdb or whatever installed so right now I can't say much about what is going on hehe Nov 07 15:09:14 cbrake: yeah well.. glibc images were starting to get too big so i thought i'd just go for nfs and access images from there Nov 07 15:09:56 gilligan_: sounds good. So you have used up all the cm-x270 flash? Nov 07 15:12:21 cbrake: well at least it didn't fit into 64M anymore.. and as soon as you get close to that size using tftp to grab it to your ramfs won't work that well because linux will run out of memory Nov 07 15:12:59 gilligan_: IC, yes, that is a problem. Nov 07 15:13:53 cbrake: i'll see if I can find out what's going on with portmap.. otherwise i'd have to fall back to USB.. but i'd like nfs to work Nov 07 15:14:02 RP: you understand oprofile, right? is there any reason I'd have trouble getting callgraphs on an N800? Nov 07 15:16:08 cbrake: i sent you the kernel patch for pcmcia support btw Nov 07 15:23:14 robtaylor: Yes, I wrote the callgraph support for ARM ;-). Your binaries probably lack frame pointers Nov 07 15:24:37 RP: ah, do I need to -O0? Nov 07 15:24:50 robtaylor: No, you need framepointers though Nov 07 15:25:00 -fno-omit-framepointers Nov 07 15:25:02 oh, no just -fno-omit-framepointers.. Nov 07 15:25:04 heh Nov 07 15:28:02 cbrake: okay.. found out what's happening.. strace executes mmap with a NULL pointer as start address argument --> *BOOM* Nov 07 15:28:10 eeh Nov 07 15:28:15 s/strace/portmap Nov 07 15:28:48 and actually it's mmap2 and not mmap, but that doesn't make the NULL pointer any better :) Nov 07 15:29:14 gilligan_: the problem I'm having with uclibc builds is alsa does not work Nov 07 15:29:55 gilligan_: I'm working on some diagnostics that run from NOR flash, so I need to get that working. Nov 07 15:29:56 RP: thanks :) Nov 07 15:30:23 hmm, sdk packges an empty "/usr" directory Nov 07 15:30:37 cbrake: but actually I suppose that this is uclibc setting up the process.. open("/etc/ld.so.cache", O_RDONLY) = 4 ; mmap2(NULL, 624, PROT_READ, MAP_SHARED, 4, 0) = 0x40006000 Nov 07 15:33:38 cbrake: http://pastebin.com/m5bff98ab Nov 07 15:34:19 cbrake: looks like actually close() is crashing.. but i'm sure start must not be NULL for mmap2() Nov 07 15:37:25 hm.. looks like mmap2 with start=NULL is okay after all.. even if the manpages seem to suggest the opposite Nov 07 15:44:13 cbrake: oh.. i might have exceeded flash size in nor causing the crash. checking.. Nov 07 15:45:33 cbrake: nope.. that wasn't the prob Nov 07 15:46:10 gilligan_: we should add some checks for NOR image sizes -- I've run into that issue several times, but I think it always failed earlier Nov 07 15:47:10 cbrake: yeah - some sanity check for that would make a lot of sense Nov 07 15:48:59 cbrake: I take it the code putting together rootfs cpio/tar archives is in the image class ? ... Nov 07 15:49:35 gilligan_: + some vars in bitbake.conf Nov 07 15:50:44 i see.. hm.. there is a $(MACHINE_POSTPROCESS_COMMAND} Nov 07 15:51:21 and also $(IMAGE_POSTPROCESS_COMMAND) .. both are called at the end of do_rootfs Nov 07 15:53:21 but that seems to be at the wrong level Nov 07 15:55:16 cbrake: was thinking of using a size limit variable in the cmx270 machine config and checking created images against that.. however you could only check .cpio images easily for their actual size Nov 07 15:56:08 cbrake: then again there is no need to check the actual image but just the staged rootfs directory Nov 07 16:02:14 RP: i think we should apply the AUTOREV for the time being, even if it's ugly Nov 07 16:02:57 mickeyl: It'll just never get fixed properly :/ Nov 07 16:05:12 RP: i understand this concern, but i doubt whether just leaving it out -- making it easy for actual users to fall into the trap -- is enough of a motivation to fix it... Nov 07 16:05:52 mickeyl: It does mean I keep noticing it... Nov 07 16:06:22 mickeyl: Lets say commit it in one week Nov 07 16:06:34 deal :) Nov 07 16:06:48 I just don't have the time I need to fix these things :( Nov 07 16:07:09 i know and i don't want to impose any pressure. just thinking about actual users here Nov 07 16:08:31 * mickeyl launches the first OE build on his new machine+distro Nov 07 16:08:36 *crossing fingers* Nov 07 16:08:38 cbrake: maybe we can discuss this some other time. you seem to be busy and I really want to go home now :) see you.. Nov 07 16:08:41 laters everyone Nov 07 16:08:46 cu gilligan_ Nov 07 16:10:35 re Nov 07 16:10:44 mickeyl: so what monster you bought? Nov 07 16:11:17 Intel(R) Core(TM)2 Duo CPU E6850 @ 3.00GHz, 4G RAM, 500G harddrive Nov 07 16:11:34 Dell 2707WFP Nov 07 16:11:41 that baby _flies_ Nov 07 16:12:14 and matrox g550 pci version? Nov 07 16:12:50 heh Nov 07 16:12:57 no, ATI Radeon 2600 XT Nov 07 16:13:00 giving me a headache Nov 07 16:13:07 ati suxx under linux Nov 07 16:13:08 1900x1200 w/ vesa driver = no go Nov 07 16:13:25 w/ proprietary driver = freeze Nov 07 16:13:31 mickeyl: replace it with nvidia someting - atleast binary driver works Nov 07 16:13:36 running proprietary driver + disabled all accel atm. Nov 07 16:14:16 maybe one day amdati will learn how to write drivers Nov 07 16:14:54 or they will open all docs and say: hi community - here are docs, write us drivers because we cant Nov 07 16:15:59 ya, either one would be nice Nov 07 16:16:43 mmhr Nov 07 16:16:45 File "tmpHandler(e)", line 31, in tmpHandler Nov 07 16:16:45 ValueError: substring not found Nov 07 16:16:48 patches gladly accepted over at xf86-video-radeonhd Nov 07 16:16:59 CosmicPenguin: what's the state of radeonhd? Nov 07 16:17:11 usable Nov 07 16:17:16 no accel Nov 07 16:19:32 no accel, as in 2d or 3d? Nov 07 16:19:34 or both? Nov 07 16:20:35 hrw: They have opened all the docs Nov 07 16:21:09 CM: for x200 too? Nov 07 16:21:44 RP: OT: does the n8x0 really do USB2.0? Nov 07 16:21:57 onboard gfx which I have here (RS482/SB400 chipset) + free driver was catastrophe not using Nov 07 16:22:06 zecke: not sure tbh Nov 07 16:22:06 Hmm.. Maybe not all the old ones Nov 07 16:22:26 now I do not care about ati part of amd - nvidia cards just works with binary driver (fglrx was not so good) Nov 07 16:48:14 hmm Nov 07 16:48:22 local.conf.sample does not document the multithreading options Nov 07 16:52:09 heh, noticed that, every time i fire up an oe environment on a new box i have to dig into bitbake Nov 07 16:52:14 's code to remember what they are Nov 07 16:54:21 yea Nov 07 16:54:23 * mickeyl will fix that Nov 07 16:54:32 just need to find out how they're called again.... Nov 07 16:54:33 ;) Nov 07 16:54:36 :) Nov 07 16:56:30 * cbrake observes a PXA270 system brought to its knees by a SD card problem -- mmcqd and IRQ-32 taking up lots of CPU time, and system very sluggish. Replace SD card and everything is fine. Nov 07 17:07:05 * * OE Bug 3267 has been created by ikkakujyu(AT)gmail.com Nov 07 17:07:07 * * e-wm fails to start after install Nov 07 17:07:10 * * http://bugs.openembedded.org/show_bug.cgi?id=3267 Nov 07 17:23:51 oooh Nov 07 17:23:52 http://bec-systems.com/web/index.php?option=com_content&task=view&id=85&Itemid=9 Nov 07 17:24:01 angstrom bootsplash :) Nov 07 17:25:09 koen|away: =D Nov 07 17:25:59 also, they have a neo. Interesting Nov 07 17:27:42 CoreDump: bec-systems == cbrake :) Nov 07 17:27:55 oh, no big wonder then Nov 07 17:28:07 ~praise cbrake Nov 07 17:28:08 All hail cbrake! Nov 07 17:28:20 in todays news: installing leopard trashes your partitions Nov 07 17:28:58 openmoko-agpsui2 ?!?! Nov 07 17:29:07 hehe Nov 07 17:29:26 I would settle for agps1...w/ a working driver ;) Nov 07 17:29:37 "The author is Ken Yale from Global Locate (now Broadcom)." Nov 07 17:31:04 http://svnweb.openmoko.org/developers/olv/openmoko-agpsui2/src/gllin.cpp?rev=3371&view=markup is funny Nov 07 17:55:42 koen|away: what is it? Nov 07 17:56:45 hrw: gllin 'driver' Nov 07 17:57:01 execvp('gllin', ''); Nov 07 17:58:04 * koen|away rediscovers the netbook charger Nov 07 17:58:13 so it is useless as there is no gllin-eabi Nov 07 17:58:21 hrw: correct Nov 07 17:58:39 * koen|away still hasn't tried gypsy Nov 07 17:58:55 I looked at 'mokostat' project recently - another useless tool.. Nov 07 17:59:18 * RP keeps meaning to play with gllin Nov 07 17:59:53 RP: as you are one of phase0 devs I can provide you working chroot for gllin Nov 07 18:02:23 * cbrake hopes koen|away does not look to closely at his photography skills ... Nov 07 18:05:49 * koen|away discovers that pulling files from encrypted loop images isn't supported in OSX Nov 07 18:06:09 that is a loop image inside a corrupted partition Nov 07 18:38:04 * * OE Bug 3268 has been created by autobuild(AT)openembedded.org Nov 07 18:38:06 * * neod-0.1.0+svnr2956-r5-do_compile Nov 07 18:38:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3268 Nov 07 18:40:05 anyone know if dos2unix is in a package available to OE? Nov 07 18:40:23 let me see Nov 07 18:40:35 actually, I think it's in busybox Nov 07 18:40:42 now that I think of it Nov 07 18:41:19 jsheldon but I think its not activated Nov 07 18:41:29 so you have to customize the busybox conf Nov 07 18:41:32 right Nov 07 18:41:48 thanks for checking Nov 07 18:43:58 or build textutils iirc Nov 07 18:45:22 yeah for angstroem it is comment out Nov 07 18:45:24 # CONFIG_DOS2UNIX is not set Nov 07 18:45:24 # CONFIG_UNIX2DOS is not set Nov 07 20:31:05 * cbrake is resolved to try a little harder not to let customers put SD cards in industrial devices ... Nov 07 20:35:05 * * OE Bug 3269 has been created by autobuild(AT)openembedded.org Nov 07 20:35:07 * * gnash-0.8.0-r0-do_populate_staging Nov 07 20:35:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3269 Nov 07 20:53:21 hello Nov 07 21:57:06 mickeyl: ping? Nov 07 22:08:31 anyone here played with a gp2x? Nov 07 22:15:54 gn Nov 07 23:55:35 03florian 07org.oe.dev * ra8532082... 10/ (3 files in 3 dirs): libgemwidget: add 1.0 Nov 08 00:46:45 mickey|zzZZzz, zecke_: If you get a chance could you glance at r3101 in poky and see if you think its ok for bitbake or if its still too hacky (http://svn.o-hand.com/view/poky?rev=3101&view=rev) Nov 08 00:47:02 haha, just wanted to go to bed Nov 08 00:48:43 You even remove ' # ***Nasty hack***' Nov 08 00:49:57 zecke_: I add a whole 5 lines of warning elsewhere though Nov 08 00:50:16 Line 96 in git.py. Is the indention right? Nov 08 00:50:47 hg is completely untested? Nov 08 00:51:14 zecke_: Yes to the first question Nov 08 00:51:24 I think the submitter did test the hg fetcher Nov 08 00:51:38 If did have some srcrev support originally but I removed it Nov 08 00:52:01 I'll add it when hg can check the revision of a remote url without pulling it Nov 08 00:52:28 I need to be more awake to review the __init__.py change :) Nov 08 00:53:02 night Nov 08 00:53:42 zecke_: fair enough :) Nov 08 00:53:48 'night Nov 08 00:53:56 * RP -> Zzzz too Nov 08 01:31:32 I am trying to reconfigure the kernel and I am trying to get so i can do a 'make menuconfigu' to update the defconf file. I run: bitbake virtual/kernel -c devshell and it fails. I've pasted the output to: http://pastebin.com/m1e64b319 any help would be Nov 08 01:31:41 nice and thanks a head of time :) Nov 08 02:36:14 Good morning Nov 08 02:44:51 A question for packaging . I think a default FILE_${PN} should be set for each .bb according to conf/bitbake automatically Nov 08 02:45:13 s/FILE/FILES/ Nov 08 02:46:19 However, in my build environment, FILES_busybox was not set. Therefore , many files are not included (e.g. /bin/busybox ) Nov 08 02:46:38 It is rather stange... how can I trace out the source of error? Nov 08 02:47:00 peek busybox FILES_busybox Nov 08 02:47:05 /usr/share/udhcpc **** ENDING LOGGING AT Thu Nov 08 02:59:57 2007