**** BEGIN LOGGING AT Fri Apr 02 02:59:57 2010 Apr 02 05:20:56 03Martin Jansa  07org.openembedded.dev * rc2c7abf724 10openembedded.git/recipes/openmoko-3rdparty/mcnavi_0.2.5.bb: mcnavi: replace link to (sometimes) non-existent INSTALL file with empty file Apr 02 06:31:44 which recipe provide libebook-1.2.so? should some thing similar to evolution-data-server Apr 02 07:22:38 morning Apr 02 07:23:54 RP: I had a new kind of ERROR building linux-rp. I'm using bitbake master and pullet yesterday evening, after the packaged-staging commits Apr 02 07:23:59 http://www.pastebin.ca/1854693 Apr 02 07:24:37 ERROR: Exception: Message:too many values to unpack Apr 02 07:25:39 he ..is my old friend... Apr 02 07:25:43 # Apr 02 07:25:45 NOTE: :too many values to unpack while evaluating: Apr 02 07:25:47 # Apr 02 07:25:48 ${@base_version_less_or_equal("KERNEL_VERSION", "2.6.17", "", Apr 02 07:25:50 # Apr 02 07:25:52 "apm-wifi-suspendfix", d)} Apr 02 07:26:40 (pls fix vercmp) Apr 02 07:29:45 IIRC this is broken for a while :/ I've already dropped that line in my work branch and forgot that later :/ Apr 02 07:29:58 hi JaMa Apr 02 07:30:04 then is bitbake Apr 02 07:30:43 I did never got this with 1.8.18 Apr 02 07:33:23 ..anyway base_version_less_or_equal can't cope with special kernel names (e.g. linux_2.6.33+2.6.34-rc3) Apr 02 07:34:41 ant_work: do you have linux_2.6.33+2.6.34-rc3? if yes.. please commit it, I got new serial cable.. so I would like to debug the problems Pavel reported for .34* Apr 02 07:35:36 ok, I thought it would be beter starting with linux-kexecboot Apr 02 07:37:00 ant_work: well 2nd stage is easier/faster to change.. so as you wish (I'll be gone till Wed skiing..) Apr 02 07:37:31 Carpazi or Alps? Apr 02 07:37:58 Alps Apr 02 07:38:15 Stubai, leaving tomorrow early morning Apr 02 07:38:34 ah, Stubai Tal..Neuss Apr 02 07:39:00 long ago I was training there..it's only 250 km from here Apr 02 07:48:16 03Martin Jansa  07org.openembedded.dev * r2db6137033 10openembedded.git/recipes/shadow/ (files/shadow.automake-1.11.patch shadow_4.1.4.2.bb): Apr 02 07:48:16 shadow: fix do_install with automake-1.11 Apr 02 07:48:16 Signed-off-by: Martin Jansa Apr 02 07:48:17 03Michael Pilgermann  07org.openembedded.dev * rfe67969ea2 10openembedded.git/ (3 files in 2 dirs): Apr 02 07:48:18 pisi: Add version 0.5, move checksums to recipe Apr 02 07:48:18 Signed-off-by: Martin Jansa Apr 02 07:48:18 03Tom  07org.openembedded.dev * rb0ab62e6b4 10openembedded.git/recipes/shr/shr-settings_git.bb: Apr 02 07:48:18 shr-settings: Changed the broken use of * in the CONFFILE section. It seems that opkg does not respect *s there. Wrote all the files manually instead. Apr 02 07:48:19 Signed-off-by: Martin Jansa Apr 02 07:48:34 good morning Apr 02 07:51:53 hi mckoan Apr 02 08:01:20 03Martin Jansa  07org.openembedded.dev * rccd4ab4bab 10openembedded.git/recipes/mokoko/mokoko_svn.bb: mokoko: add autopoint work-around Apr 02 08:35:07 hillct, how can i update ipk package feed list? ~tmp/deploy/glibc/ipk/mipsel/Packages.gz Apr 02 09:04:33 Hi, I was triyng to build busybox but i always get this error: NOTE: The SHA256Sums did not match. NOTE: Task failed: Checksum of... ERROR: TaskFailed event exception, aborting. I didn't found any info how to fix it. Any ideas? Apr 02 09:06:07 J_: remove from downloads dir (also with ARCHIVE.md5 file) and try to redownload Apr 02 09:07:22 http://wiki.openembedded.net/index.php/Dealing_with_checksum_mismatch Apr 02 09:07:46 is outdated Apr 02 09:27:29 Openfree`: bitbake update-packates or something like that Apr 02 09:29:26 package-index? Apr 02 09:30:14 JaMa: redownloaded it same error, so it will be outdated is there any way how to bypass it? I have looked into checksum.ini but haven't found anything usefull. Apr 02 09:30:46 JaMa, seems right Apr 02 09:30:55 bitbaking... Apr 02 09:31:20 J_: only sha256 is failing and md5 is correct? strange Apr 02 09:31:53 J_: busybox checksums are already in recipes.. see SRC_URI[tarball.md5sum] Apr 02 09:33:57 J_, also you could set OE_ALLOW_INSECURE_DOWNLOADS=1 to ignore the checksum, or update the official checksum.ini file Apr 02 09:34:04 J_, better file a bug.. Apr 02 09:34:11 or. Apr 02 09:54:22 while I using "opkg install xx_package" command, where does the download package store? Apr 02 09:58:43 http://www.libcpu.org/wiki/index.php/Main_Page Apr 02 09:59:07 hi woglinde Apr 02 09:59:36 he ant!!!!!!!!!!!! Apr 02 09:59:45 hm fixed klibc stuff Apr 02 09:59:51 but didnt commit Apr 02 09:59:59 do you remeber the u8 bug? Apr 02 10:00:04 just rename it to __u8 Apr 02 10:00:08 hmm.. Apr 02 10:00:12 taht worked Apr 02 10:00:21 or kexecbla Apr 02 10:00:34 iirc they say old kernels(headers) will be not supported Apr 02 10:00:39 after 1.5.15 Apr 02 10:00:50 hm worked here for qemuarm Apr 02 10:01:01 nice Apr 02 10:01:09 I need your kexec-stuff for my n810 Apr 02 10:01:18 so I am abled to boot diffrent kernels Apr 02 10:01:21 I need a new tester ;) Apr 02 10:01:32 just include conf/machine/include/initramfs-kexecboot.inc Apr 02 10:01:50 I hope the n810 can do kexec fine in recent omap kernel Apr 02 10:01:52 and create a defconfig for n810 Apr 02 10:02:03 ah okay Apr 02 10:02:10 I worked around it Apr 02 10:02:20 build the initamfs with oe Apr 02 10:02:27 he he Apr 02 10:02:30 and then build the kernel outside Apr 02 10:02:46 I know you know how-to :D Apr 02 10:02:50 with pointer to the initramfs in oetmp Apr 02 10:02:52 *g* Apr 02 10:03:14 unfornatly qemu not run s for n810 Apr 02 10:03:22 only versatile Apr 02 10:03:24 but that worked Apr 02 10:03:34 didnt test kexec yet Apr 02 10:04:33 hm let me see how big the kernel is with gz cpio Apr 02 10:04:46 I think I saw several kernel patches for OMAP Apr 02 10:05:02 hopefully they unbreak kexec Apr 02 10:05:04 ant I am working directly with the ompa kernel tree ;) Apr 02 10:05:09 ups omap Apr 02 10:05:28 iirc koen could kexec one year ago with beagle Apr 02 10:05:28 818844 2. Apr 02:15 /devel/archive/git/linux-omap/arch/arm/boot/zImag Apr 02 10:05:35 then got broken :/ Apr 02 10:05:36 I stripped the kernel down Apr 02 10:05:46 disabeld most stuff Apr 02 10:05:48 is lzma? Apr 02 10:05:51 thats a nice kernel size Apr 02 10:05:52 no Apr 02 10:05:55 gz Apr 02 10:05:59 OMG Apr 02 10:06:11 I got around 800kb for zaurus Apr 02 10:06:37 only jffs2 and vfat and mtd is enabled Apr 02 10:06:39 and if I don't lzma-compress the initramfs is 100kb bigger... Apr 02 10:06:41 debug is off Apr 02 10:07:13 unfornatly I dont works with printk Apr 02 10:07:18 args it Apr 02 10:07:29 try to lzma-compress both kernel + initramfs...you should gain 20-30 percent Apr 02 10:07:37 ant sorry Apr 02 10:07:51 I think the n810 bootloader will not handle lzma Apr 02 10:08:05 ? Apr 02 10:08:13 even zaurus one boots Apr 02 10:08:16 hm Apr 02 10:08:18 okay Apr 02 10:08:27 be careful with alignment Apr 02 10:08:33 I thought it depends on the bootloader Apr 02 10:08:51 lets see Apr 02 10:09:05 iirc the header is th same, just the last bytes differs (lzma signature I think) Apr 02 10:09:36 what about lzo? Apr 02 10:09:39 initramfs Apr 02 10:10:07 well,lzo it should be new standard for arm afaik Apr 02 10:10:49 lzma is slightly slower on initial decompression Apr 02 10:10:55 but with these sizes... Apr 02 10:11:00 below 1 mb Apr 02 10:11:04 hm I only have lzo Apr 02 10:11:17 >2.6.32 in linux-kexecboot Apr 02 10:11:22 > 2.6.33 in linux Apr 02 10:11:33 only gzip and lzo Apr 02 10:11:39 no lzma Apr 02 10:11:39 I ported the patches to 2.6.32-kexecboot Apr 02 10:11:49 for kernel compression Apr 02 10:11:49 and to 2.6.33 linux-kexecboot Apr 02 10:12:13 check in recipes/kexecboot Apr 02 10:12:39 okay Apr 02 10:12:44 initramfs with lzma works Apr 02 10:12:57 hm, anyone an idea what is wrong if bitbake works, but if I build an image in the do_rootfs Ik get the message that it cannot open bitbake.conf; this seems to come from bbimage Apr 02 10:13:02 791500 2. Apr 12:12 /devel/archive/git/linux-omap/arch/arm/boot/zImage Apr 02 10:13:17 under 800k nice Apr 02 10:14:35 now lzma-compress initramfs-kexecboot cpio Apr 02 10:14:37 too Apr 02 10:14:45 yeah mom Apr 02 10:14:46 I know it seems silly.. Apr 02 10:15:09 but strangely enough is better to compress the cpio separately Apr 02 10:15:18 its 34-rc3 from linux-omap tree Apr 02 10:15:42 yes...the sucker seems not booting on aurus.. Apr 02 10:16:10 I'll commit this evening.. Apr 02 10:17:02 hm Apr 02 10:17:08 patch dont applys clean Apr 02 10:17:12 but its only Makefile Apr 02 10:17:27 ups Kconf Apr 02 10:18:12 woglinde, the recipe will look like here http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=eef155ea30aa97e913d0fe52cd1d9564db369aa0 Apr 02 10:18:20 just updated Apr 02 10:18:59 /bin/sh: lzop: command not found Apr 02 10:19:11 ah..I think I know why.. Apr 02 10:19:15 mom Apr 02 10:19:22 lzop - Ein schnelles Kompressionsprogramm Apr 02 10:19:33 Have anyone got an idea how I can compile an SDK for Angstrom@Beagleboard? I managed to build an SDK, but linux-libc-headers doesn't give me the linux-omap kernel with the necessary headers and patches. My *.bb-fu is not strong enough that I'm sure what to do. Apr 02 10:20:37 ant worked Apr 02 10:20:47 see bitbake.conf Apr 02 10:21:11 lacks EXTRA_IMAGECMD_cpio.lzo Apr 02 10:21:30 someone should add lzo here Apr 02 10:22:05 ah..while you're there... Apr 02 10:22:23 Stanislav had same problem with lzma: he did not have on host Apr 02 10:22:29 ant but the kernel is bigger Apr 02 10:22:33 but in bitbake.conf... Apr 02 10:22:42 or do I have to run anothe target Apr 02 10:22:46 like zImage? Apr 02 10:22:58 839872 2. Apr 12:20 /devel/archive/git/linux-omap/arch/arm/boot/zImage Apr 02 10:23:08 no, just menuconfig and select both lzma Apr 02 10:23:16 yes Apr 02 10:23:18 for initramfs and kernel itself Apr 02 10:23:24 kernel is 440k bigger now Apr 02 10:23:30 hm.. Apr 02 10:23:33 than with gzip Apr 02 10:23:46 size of cpio.gz and cpio.lzma ? Apr 02 10:24:15 from oe? Apr 02 10:24:26 of initramfs.kexecboot-image Apr 02 10:24:30 hm Apr 02 10:24:42 are you sure you embedded it previously Apr 02 10:24:43 no I meant its 400k bigger with lzma enabled Apr 02 10:24:46 for kernel Apr 02 10:24:48 initramfs Apr 02 10:24:53 was already lzma Apr 02 10:24:55 it seems too strange... Apr 02 10:25:00 okay I have to leave Apr 02 10:25:10 will be home in 1 1/2 weeks Apr 02 10:25:13 I gained 250kb Apr 02 10:25:18 than I can test it again Apr 02 10:25:21 bye Apr 02 10:25:24 bye Apr 02 10:29:48 I have deleted sha256sum from recipe but, and it worked but now I get NOTE: Applying patch 'udhcpscript.patch' NOTE: Task failed: Command Error: exit status: 127 Output: sh: quilt: command not found. Apr 02 10:34:23 03Martin Jansa  07org.openembedded.dev * r510979b065 10openembedded.git/recipes/xorg-xserver/ (2 files in 2 dirs): Apr 02 10:34:23 xorg-xserver: add new 1.8.0 version Apr 02 10:34:23 Signed-off-by: Martin Jansa Apr 02 10:37:24 JaMa: if you have 2 mins, pls commit linux-kexecboot-2.6.33+2.6.34-rc3...we have an OMAP tester...I'd make him wait until this evening ;) Apr 02 10:38:26 ant_work: no time today.. sorry Apr 02 10:38:43 already decided 4 times to start daywork :) Apr 02 10:38:44 np..have a good WE Apr 02 10:38:58 thanks Apr 02 11:06:38 J_ not sure what you are baking but seems you are missing some steps, normally you should have gotten quilt Apr 02 11:07:21 if you are missing the util to do the sha256 sum might be that that causes the checksum error Apr 02 11:11:28 eFfeM: I have fixed that already but now it gives me http://pastebin.com/Jz5sxXNA Apr 02 11:13:48 J_: don't use bitbake -b you'll be missing all the dependency compilations Apr 02 11:13:54 do bitbake busybox Apr 02 11:14:31 J_ btw the cause of your compile is in /home/john/lin/stuff/build/tmp/work/armv5te-oe-linux-gnueabi/busybox-1.15.3-r26.0/temp/log.do_compile.5350 Apr 02 11:14:51 but it will definitely be a missing header file or so from a package that is needed by busybox Apr 02 12:14:51 03Michael 'Mickey' Lauer  07org.openembedded.dev * r8f5f909a77 10openembedded.git/recipes/freesmartphone/mioctl_git.bb: mioctl: new recipe; command-line utility to call ioctls Apr 02 12:14:53 GNUtoo: ^^ Apr 02 12:15:05 thanks Apr 02 12:15:17 np, you can check /usr/include/vt.h for the necessary constants Apr 02 12:15:34 err Apr 02 12:15:35 fb.h Apr 02 12:15:40 vt.h is virtual terminal Apr 02 12:16:15 I already did Apr 02 12:16:26 good Apr 02 12:16:33 mioctl supports decimal and hexadecimal Apr 02 12:16:37 your image is packaging Apr 02 12:16:39 just remember to use 0x as prefix for hex Apr 02 12:16:44 cool Apr 02 12:17:11 ~praise Vala for enabling mickey to write C utilities Apr 02 12:17:12 All hail Vala for enabling mickey to write C utilities! Apr 02 12:17:30 doing a do_package_write_ipk for most of packages Apr 02 12:54:53 03Martin Jansa  07org.openembedded.dev * red61ecec2a 10openembedded.git/recipes/xorg-driver/ (xf86-video-glamo_git.bb xorg-driver-common.inc): xf86-driver-common: PR bump for newer xserver in SHR Apr 02 14:32:17 heh, everyone must be eating chocolate eggs today Apr 02 14:32:30 Or hunting rabbits. Apr 02 14:32:42 Are you guys back in Edinburgh yet? Apr 02 14:32:45 no Apr 02 14:32:52 tomorrow night Apr 02 14:35:16 It's actually sunny again now, I was wondering if you'd brought back the weather :) Apr 02 14:36:21 heh Apr 02 14:36:25 probably snow tomorrow then Apr 02 14:48:04 did you bring back run? Apr 02 14:48:16 * XorA isnt back Apr 02 15:14:07 03Tom Rini  07org.openembedded.dev * r973169bb71 10openembedded.git/classes/insane.bbclass: (log message trimmed) Apr 02 15:14:07 insane.bbclass: Switch to endswith() and fix la/pc check Apr 02 15:14:07 This changes insane.bbclass to use s.endswith("str") rather than s[-3:] == Apr 02 15:14:07 "str". The primary motivation for this is that we have tests that check for Apr 02 15:14:07 'la' and 'pc' files when we really want '.la' and '.pc' to be what we check Apr 02 15:14:08 for. Also, our -dev symlink test does if test1: if test2 and test3: action, Apr 02 15:14:09 when it could just be if test1 and test2 and test3. Apr 02 15:18:41 morning Apr 02 15:21:33 gm Apr 02 15:30:14 hello bluelightning Apr 02 15:30:17 hi kergoth Apr 02 15:30:23 hi ant_home Apr 02 16:09:28 03Andrea Adami  07org.openembedded.dev * radadb03430 10openembedded.git/recipes/kexecboot/ (5 files in 5 dirs): linux-kexecboot: initial commit of 2.6.34-rc3 Apr 02 16:12:45 hi bluelightning Apr 02 16:12:57 hi mickeyl Apr 02 16:14:45 join #qt Apr 02 16:14:49 doh Apr 02 16:23:11 khem, we've got uClibc-0.9.31. Please bump; TIA Apr 02 16:24:59 bbl Apr 02 16:50:21 RP: in today? Apr 02 17:29:58 03Martin Jansa  07org.openembedded.dev * r6b2b312e24 10openembedded.git/recipes/openmoko-3rdparty/thone_0.6.bb: Apr 02 17:29:58 thone: move to version 0.7 Apr 02 17:29:58 Signed-off-by: Martin Jansa Apr 02 17:57:11 i don't care. khem, I just want something sane as opposed to the nothing that it does ATM Apr 02 17:57:40 khem, and stuff like that belongs to #here and not #there, but you know the game, i suppose Apr 02 18:01:12 khem, it apparently takes all that, yes. But i won't, for the stuff i need. I'm very happy to let the rest screw themselves but i'll never ever accept that sloppiness unto /me -- and yea call me freak because of that. I simply don't care about stuff like that. Sanity, maintainability, the rest fux0res up themselves on their own Apr 02 18:23:49 JaMa: did you commit your SRCREV change ? Apr 02 18:51:20 hi. are there any plans for a stable 2010 release? Apr 02 19:13:40 re Apr 02 20:11:38 nik0n, that might happen if/when dev slows down for a bit :) Apr 02 20:12:36 Crofton: kthx Apr 02 20:13:18 btw, a bit of an OT question. anyone aware of a SOHO linux router with gigabit ethernet (and maybe oe support)? Apr 02 20:47:13 ka6sox: ping Apr 02 21:31:41 khem|vacation: yes I did Apr 02 21:32:20 khem|vacation: after koen's comment.. and those fixes we talked about Apr 02 22:43:36 03Michael 'Mickey' Lauer  07org.openembedded.dev * rc1b7233ffd 10openembedded.git/recipes/linux/linux-leviathan_git.bb: linux-leviathan: bump to master to incorporate GNUtoo's panel resume fixes **** ENDING LOGGING AT Sat Apr 03 02:59:57 2010