**** BEGIN LOGGING AT Mon May 04 02:59:57 2009 May 04 04:11:20 * rwhitby wonders what we will call the next distro when 'micro' get bloated .. May 04 06:43:22 good morning May 04 07:25:53 good morning May 04 08:41:23 hello all i'm building my own root filesystem (based on angstrom-minimal-image) and i have problem with vsftpd May 04 08:41:27 in pkg_postinst() { addgroup ftp May 04 08:42:17 i have addgroup ftp but in root filesystem tarball there is no group named ftp in /etc/group where's the problem ? May 04 08:46:49 good morning May 04 08:58:12 Guest86954: addgroup can only run on the real device May 04 09:12:24 Hi all! May 04 09:12:59 Need help while bitbaking openmoko: May 04 09:13:06 /tmp/staging/i686-linux/usr/bin/install-sh: No such file or directory May 04 09:13:16 What goes wrong? May 04 09:25:34 szsoftware: some changes in that area recently. install-native got replaced by coreutils-native May 04 09:26:29 hi May 04 09:27:53 i have a lots of jffs2_scan_eraseblock meassages when i mount the nand on an zaurus c-3200 , and it took very long to mount it , any ideas ? May 04 09:31:27 pwgen: wrong block size? May 04 09:32:13 pwgen: EXTRA_IMAGECMD_jffs2 = "--eraseblock=0xabc" May 04 09:32:22 substitute abc with your eraseblock size May 04 09:47:20 ah... May 04 09:47:43 pwgen: see zaurus-2.6.inc May 04 09:47:57 try adding ERASEBLOCKSIZE_spitz = "0x20000" May 04 09:48:24 (like for akita) May 04 09:48:37 ..default is ERASEBLOCKSIZE = "0x4000" :/ May 04 09:48:46 akheron: good hint May 04 09:49:23 XorA: nobody bothered to flash spitz before.... May 04 09:49:48 XorA: I hope apart size, the flash is the same on 3000 vs. 3100/3200 May 04 09:50:27 * JaMa has c-3000 May 04 09:50:35 you want me to test something? May 04 09:52:03 It would be great when someone write ubifs detection module for klibc's fstype routine May 04 09:52:15 kexecboot using it May 04 09:53:09 JaMa: sometimes new kexecboot commits should be tested on misc machines :) May 04 09:53:28 we are currently testing on Andrea's C-7x0 :) May 04 09:53:31 hm.. see the bootlogs http://digit.que.ne.jp/visit/index.cgi?59414d41#i24 May 04 09:53:34 and in qemu-system-arm May 04 09:53:48 3000 NAND device: Manufacture ID: 0xec, Chip ID: 0x73 (Samsung NAND 16MB 3,3V) May 04 09:54:20 3100/3200 NAND device: Manufacture ID: 0xec, Chip ID: 0xf1 (Samsung NAND 128MB 3,3V) May 04 09:55:12 same as akita! May 04 09:55:39 it seems older pxa255 has another May 04 09:55:41 NAND device: Manufacture ID: 0xec, Chip ID: 0x79 (Samsung K9D1G08V0M) May 04 10:02:12 XorA: Thank you for that advice. I now removed coreutils-native stamp files and bitbaked it again. Although install.sh is still missing. What can I do next? My oe environment is really fresh (some days old with currently git pulled) May 04 10:02:50 szsoftware: I dont think we should be looking for install.sh anymore, but I havent actually seen problems May 04 10:03:46 XorA: but om-image can't be bitbaken. Some tasks fail because of that issue May 04 10:04:26 XorA: what does that install.sh do? Can I fake it? May 04 10:04:40 szsoftware: install.sh is shell version on install May 04 10:04:43 AFAIk May 04 10:06:04 XorA: so install.sh is called by recipes, they dont' know it is replaced during the install-native/coreutils-native replacement? May 04 10:06:29 szsoftware: possibly May 04 10:06:41 JaMa: have you ever tried to flash a small image to nand? Something less than 5mb? I'm unsure about erasesize on 3000 May 04 10:06:56 if we put ERASEBLOCKSIZE_spitz = "0x20000" May 04 10:07:08 we possibly break 3000 May 04 10:07:36 XorA: install is a executable that resides too within /tmp/staging/i686-linux/usr/bin/? May 04 10:07:57 szsoftware: silly question, will look for.. May 04 10:08:35 ant_work: the same jffs2 worked on all 3x00 devices for sharp May 04 10:09:48 XorA: seems rather strange then May 04 10:10:28 iirc spitz was just pivot-rooting without touchng nand May 04 10:10:30 XorA: ok, it does. Will make a symlink install-sh to install May 04 10:10:41 XorA: and wait what happens May 04 10:11:19 szsoftware: I don't think the recipes should explicit search for installer.sh... May 04 10:14:41 szsoftware: ok, so it's a debate if it is a bug of oe, that should we workarounded/patched (symlinking helps) or a bug of dozens of recipes. May 04 10:15:04 ant_work: ok, so it's a debate if it is a bug of oe, that should we workarounded/patched (symlinking helps) or a bug of dozens of recipes. May 04 10:15:37 ups, I seem to be a little schizophren, talking to my self, haha May 04 10:18:22 szsoftware: see, I encountered a ton of issues after introduction of newer coreutils May 04 10:19:06 szsoftware: fwiw the packages in console-image, opie-image, x11-image are *immune* from the installer.sh issue May 04 10:19:34 Hello everybody May 04 10:19:47 so, I'd say there are buggy packages around... May 04 10:21:10 szsoftware: which package is failing? May 04 10:22:03 Hello everybody May 04 10:22:27 I have a quick noob question... May 04 10:24:41 How can I change kernel options and reubild it? I have found a file named 'defconfig' with the default configuration under oe/stuff/build/tmp/work/beagleboard-angstrom-linux-gnueabi/linux-omap-2.6.28-r23. If I modify the options and bitbake, it looks like the kernel isn't recompiled. May 04 10:25:32 If I bitbake -cclean virtual/kernel, then the entire linux-omap-2.6.28-r23 directory is deleted, and when I bitbake virtual/kernel, the file is restored with the default flags May 04 10:25:47 morning May 04 10:25:57 Where can I modify the kernel options? May 04 10:26:08 * hrw wants u-boot instead of grub on laptop May 04 10:26:28 ~curse debian for breaking grub-pc May 04 10:26:30 May you be reincarnated as a Windows XP administrator, debian for breaking grub-pc ! May 04 10:27:51 doragasu: you can change them manually with 'make menuconfig', running 'bitbake -c devshell virtual/kernel' May 04 10:28:53 Thanks tsjsieb, I'm going to try that right now May 04 10:30:32 good morning May 04 10:30:39 yo florian May 04 10:31:50 ant_work: last time I flashed emergency.img for u-boot, but it didn't work (not sure why as I still don't have serial cable).. May 04 10:33:14 Jay7: I've zImage-kexecboot-2.6.26 prepared for flash, maybe this evening... May 04 10:33:53 JaMa: here is #kexecboot channel, btw :) May 04 10:44:16 So I 'bitbake -c devshel virtual/kernel', then edit the configuration with make 'menuconfig', and when I finish, what do I need to do, 'bitbake virtual/kernel'? May 04 10:53:18 XorA: libart-lgpl for example but now workarounded May 04 11:07:12 some of the zaurus zubuntu people here ? May 04 11:07:27 doragasu: you can: 1) bitbake -c do_compile and finish the remaining tasks 2) create a patch / copy your new defconfig in appropriate place May 04 11:08:00 thanks a lot ant_work May 04 11:08:01 doragasu: it's in your WORKDIR/machine/linux-xy May 04 11:08:12 I suggest 2 May 04 11:08:28 because you'll easily end up overwriting :-) May 04 11:08:35 I'll have a look at it, but I have another question... May 04 11:09:10 : make a bitbake -c configure ; cd tmp/xxxx/xx/Xx/kernsourcedir ; export ARCH=arm ; make menuconfig ......; bitbake -c compile May 04 11:09:19 I want to enable 'CONFIG_I2C2_OMAP_BEAGLE' May 04 11:09:50 but when I 'make menuconfig', that option is not available! May 04 11:09:55 ? May 04 11:10:01 just bitbake -c menuconfig May 04 11:10:12 It is available in the defconfig file May 04 11:10:31 there is a special 'menuconfig' task for kernel recipes May 04 11:10:36 but the corresponding option doesn't appear when I 'make menuconfig' May 04 11:10:56 ah..perhaps is a patch then May 04 11:10:58 will it work if I enable it only in the defconfig file and skip the 'make menuconfig' step¿ May 04 11:11:46 doragasu zou have to change into the kernelsourcedirectory where your kernel is extracted. Set the architecture with export ARCH=arm May 04 11:13:49 OK, I'm trying right now. Thanks a lot May 04 11:13:54 Hope it works May 04 11:14:09 doragasu: do a bitbake -c patch your-kernel and see if the option is in Kconfig... May 04 11:14:11 pwgen: isn't that already done by devshell? May 04 11:14:27 tsjsieb: yes May 04 11:21:31 The option doesn't appear :( May 04 11:21:50 I 'bitbake -c devshell virtual/kernel' May 04 11:22:19 doragasu: bitbake -c menuconfig virtual/kernel May 04 11:22:25 then in the shell, if I write 'echo $ARCH' I get 'arm' May 04 11:22:44 ant_work: I have also tried that May 04 11:22:53 and? May 04 11:22:55 and the option doesn't appear May 04 11:24:32 I search it under Device drivers/I2C Support, but the option isn't there May 04 11:28:55 doragasu: I see your option under Device drivers->I2C support->I2C Hardware Bus support->Enable I2C2 for OMAP3 BeagleBoard May 04 11:29:47 I think it's because I have not properly set some dependencies. I'm trying to set it up right now May 04 11:33:34 doragasu: I think you're using old/wrong kernel or you're missing patches May 04 11:33:42 hi ant May 04 11:33:57 hello woglinde May 04 11:36:21 FinallyI have found the option. I had to enable some dependencies. I'm suspecting there is something wrong because it's a bit strange it isn't properly configured by default for the beagleboard May 04 11:56:26 hey May 04 11:57:21 could you tell whether specific mountpoints aren't disk/flash, but ram per default? e.g. /tmp, /var? May 04 11:58:37 mirko you want to know the location inside oe? May 04 11:59:08 woglinde: no, i wanna know if ram is used instead flash for some special mountpoints May 04 11:59:36 some distros mount a ramdisk to /tmp May 04 11:59:50 to increase speed and avoid high i/o on flash May 04 12:00:00 same with /var (mind log-files here) May 04 12:00:02 mirko hehe so you want to know where in oe the files are May 04 12:00:10 to look it up yourself May 04 12:00:29 woglinde: hehe - that would be the right wa yi think ;) May 04 12:01:39 woglinde: can you give me a hint where i would find these information? ;) May 04 12:02:30 * XorA wonders why mirko want to know May 04 12:03:09 mirko it is under recipes/basefiles May 04 12:03:12 fstab May 04 12:03:18 woglinde: thanks! May 04 12:03:24 hi xora May 04 12:03:45 XorA: some apps create caches in /var May 04 12:04:06 and -in this case- /var would be wiped out after every reboot May 04 12:04:12 i wonder if anyone has managed to get the usb networking with 2.6.26 running ? May 04 12:04:13 mirko: wrong May 04 12:04:22 but (re)creating the cache would not be necessary May 04 12:04:25 mickey|bbq: ? May 04 12:04:32 oh May 04 12:04:34 XorA: ? May 04 12:04:42 mirko: only LSB compliant directories in /var are in ram May 04 12:04:56 mirko: as long as your app is compliant you shouldnt need to care May 04 12:05:15 look in /etc/volatiles/default May 04 12:06:52 XorA: shiny nice world, hm? :) May 04 12:07:35 but ok - thanks - i see the concept now May 04 12:08:58 XorA: hey, I got some further problems. Such like this: May 04 12:09:04 XorA: NOTE: Checking Package: dbench May 04 12:09:05 ERROR: QA Issue: No GNU_HASH in the elf binary: May 04 12:09:22 help appreciated. thanks in advance May 04 12:09:24 szsoftware please look at the commits May 04 12:09:26 szsoftware: you should be able to find the fixes for that in mailing list May 04 12:09:39 there are plenty of fixes for this kind of errors May 04 12:09:43 been like 1000 of those fixes in the last couple of weeks May 04 12:09:53 xora *g* May 04 12:09:53 sorry, will do my best.. May 04 12:10:12 xora we should synchronize somehow May 04 12:10:33 woglinde: SYN May 04 12:11:00 woglinde: XorA: yeah, so the punishment for me is only the half :-) May 04 12:26:21 gm May 04 12:40:21 woglinde: XorA: I found something out, that the "No GNU_HASH" issue has something todo with LDFLAGS, but I don't know where and how to workaround/fix that problem. In every bb-file, that causes this issue? In one *.conf? May 04 12:44:44 hi crofton May 04 12:45:07 Hi, I have ./configure --build=${BUILD_SYS} on my bb-file. It gives error about missing x86_64-linux-gcc, because it should be x86_64-linux-gnu-gcc, I can hardcode it to bb-file, but what is right solution? May 04 12:45:46 szsoftware the easiest way is May 04 12:45:53 go to http://cgit.openembedded.net/cgit.cgi?url=openembedded/log/ May 04 12:45:53 szsoftware: in every .bb file unfortuneately May 04 12:46:04 type GNU_HASH into the search mask May 04 12:46:15 hit enter and wonder May 04 12:49:23 woglinde, I do not understand May 04 12:53:36 woglinde: I love your sarcasm.. May 04 12:54:26 aapo what software are you trying to compile? May 04 12:55:07 woglinde, newer version of openttd May 04 12:55:53 florian, ping May 04 12:56:01 Crofton: pong May 04 12:56:11 aapo and why the older recipes dont works if you bump the versionumber of the .bb? May 04 12:56:21 he florian May 04 12:56:27 hi woglinde May 04 12:56:48 bye May 04 12:57:15 bye hrw May 04 12:57:43 woglinde, good question, I do not even test it, because I has my own version of it allreyde which containing OpenGraphicsReplacement May 04 12:57:58 http://marcin.juszkiewicz.com.pl/2009/05/04/cursing-intel/ May 04 12:59:15 florian, don't forget to send email for linnuxtag May 04 12:59:31 hrw|gone: its due to lack of docs only bios can change mode on that chipset which kernel doesnt support May 04 12:59:51 Crofton: yes, I try :) May 04 13:00:00 I try to find that old and test it May 04 13:00:09 hi kgilmer May 04 13:00:14 aapo hint May 04 13:00:23 hey florian :) May 04 13:00:30 aapo -> recipes/openntpd/ May 04 13:00:34 hi kgilmer May 04 13:00:44 hi woglinde May 04 13:01:37 * kgilmer is trying to find something that will work like a mousepad in the wife's parent's kitchen. May 04 13:03:33 don't they have vending machines with mousepads in Japan? May 04 13:03:42 I hear they have everything else in them :) May 04 13:05:05 hi cbrake May 04 13:05:16 florian: hello May 04 13:11:44 they probably do in Tokyo or Osaka Crofton. but not in the sticks :) May 04 13:12:16 i can't even get a proper bicycle tire tube here. May 04 13:15:59 kgilmer: try with a sushi pad ;-) May 04 13:17:09 hmmm, not sure where they are keeping that mckoan :) May 04 13:17:21 i found a towel that seems to be working May 04 13:33:16 Hi, my ${BUILD_SYS} is reported wrongly with openttd, it is managed with this: http://cgit.openembedded.net/cgit.cgi?url=openembedded/tree/recipes/openttd/files/openttd_0.6.3-buildfix.patch There are reported bug in upstream, but they sayd it is because of compiling system May 04 14:00:03 Some basic questions about oe. How can I build or work with different distros. When I change DISTRO in local.conf, I need a new tmp dir, but this contains the whole toolchain also. Is't this independent from target distro? May 04 14:00:31 different distros may select different toolxchains May 04 14:02:04 Crofton|work: ok this explains a lot. So I have to hold a lot of disk-space.. May 04 14:02:26 szsoftware, right :) May 04 14:04:14 Crofton|work: I did not understand the distro_pr hint regarding tmp_dir, so the distro tmp dirs can reside beside each other. Where/in which way I have to set the right tmp_dir for those purposes? May 04 14:04:27 not sure about that May 04 14:04:48 szsoftware: you could follow this tree model http://www.kaeilos.com/?q=download May 04 14:07:49 i've been having an issue building madwifi, is this a known issue? May 04 14:08:43 szsoftware: or this http://marcin.juszkiewicz.com.pl/2006/11/22/my-openembedded-environment-ii/ May 04 14:10:08 mckoan: thanks, will check this out. May 04 14:10:44 mckoan: don't wanna be pedantic but found a typo on your website: http://www.kaeilos.com/?q=kaeilos-open-embedded-linux May 04 14:10:57 anyone successfully built qt-embedded-4.5.1? I get do_compile error while building qtdesigner... anyone? May 04 14:11:09 mckoan: in the heading "KaeilOS and OpenEmbedded commecial support" May 04 14:11:20 mckoan: self explaining :-) May 04 14:11:36 booxter please show me the output May 04 14:11:55 booxter I build qt-embedded even with gles stuff May 04 14:12:01 for beagleboard May 04 14:12:51 woglinde: http://rafb.net/p/2Zu5JL99.html May 04 14:13:46 szsoftware: thx May 04 14:13:54 booxter are you compiling with profiling support? May 04 14:14:03 or why -pg is in cflags? May 04 14:14:03 woglinde: I built qt-embedded-4.5.0 with oe stuff before (with mine custom recipe - edited 4.4.3 a little). And now I tried to use qt-embedded-4.5.1 recipe found in oe itself - but the error occured May 04 14:14:15 woglinde: yes, profiling I think May 04 14:15:06 woglinde: perhaps that's what I work for day long - profiling libqt*E :) May 04 14:15:17 szsoftware: at last I found why nobody was wsking for commercial support (LOL) May 04 14:16:06 booxter hm the compilier cannt find this one symbole at linking time May 04 14:16:20 but that has to come from qt itself May 04 14:17:21 woglinde: I searched google for the error and found another person pasted the same error on pastebin with comment like: "qt-4.5.1 failed compilation" May 04 14:17:29 booxter but anyway how powerful is your mips box May 04 14:17:38 that you want to run designer on it May 04 14:17:45 I would suggest for quickfixing May 04 14:17:58 disbale the build of desginer inside the recipe May 04 14:18:22 woglinde: hm, of course I can disable it but that's not the right solution. that's a quick-hack :) May 04 14:18:39 yeah May 04 14:18:46 but the problem might be May 04 14:18:53 is it while you are trying profiling May 04 14:19:06 is it while your trying a mips build May 04 14:19:12 or just another build May 04 14:19:20 problem inside qt May 04 14:19:50 so first try to build without profiling May 04 14:20:15 woglinde: maybe a quick help like - what flags are for disabling designer? May 04 14:20:45 ./conifugre --help May 04 14:20:56 hm was soemthing like -no-designer or so May 04 14:21:53 mckoan: LOL May 04 14:23:23 woglinde: can't find :-\ May 04 14:24:10 woglinde: all I can do is -nomake tools May 04 14:24:31 but I think that's not what I really want to do (not sure though) May 04 14:27:12 have bitbake "--only-download" parameter ? May 04 14:27:50 no packages will be build, i just want only download sources May 04 14:28:16 Guest86954: use bitbake -cfetchall May 04 14:28:38 mckoan: thx May 04 14:29:51 booxter hm May 04 14:36:29 morning May 04 14:59:39 hi kergoth May 04 15:22:14 re May 04 15:26:24 hrw: welcome back! :-) May 04 15:39:59 hrw: greetings May 04 15:42:46 guys May 04 15:42:47 http://thread.gmane.org/gmane.comp.handhelds.openembedded.user/94/focus=96 May 04 15:42:54 did this ever get resolved? I'm still hitting it 2 years later.. May 04 16:01:31 Goodbye and thanks everyone! May 04 16:06:18 re May 04 16:19:10 how should I configure an angstrom distro /etc/network/interfaces for bringing up a bridge? seems like busybox ifup doesn't like the debian bridge statements May 04 16:21:23 03Marcin Juszkiewicz  07org.openembedded.dev * r4303a5f5c6 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev May 04 16:21:23 03Marcin Juszkiewicz  07org.openembedded.dev * r13f97da771 10openembedded.git/classes/base.bbclass: base.bbclass: added 'gcc-3.4' to list of gcc3 versions (Debian systems) May 04 16:21:23 03Marcin Juszkiewicz  07org.openembedded.dev * r85f0c2d5e5 10openembedded.git/conf/bitbake.conf: May 04 16:21:23 bitbake.conf: make MACHINE_KERNEL_PR defaults to PR May 04 16:21:23 By default MACHINE_KERNEL_PR defaults to 'r0' which makes build of May 04 16:21:25 kernel for machines which not use MACHINE_KERNEL_PR harder May 04 16:21:27 (kernel.bbclass uses M_K_P instead of PR). May 04 16:21:29 03Marcin Juszkiewicz  07org.openembedded.dev * r58104ee0f8 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev May 04 16:21:45 tharvey: install full versions of net tools? May 04 16:22:29 * Tartarus shakes fist at hrw May 04 16:22:33 * Tartarus does a pull and push again May 04 16:23:04 03Tom Rini  07org.openembedded.dev * r501e733081 10openembedded.git/conf/distro/ (26 files in 2 dirs): May 04 16:23:04 All distro conf files: Use LIBC to pick or set your libc. May 04 16:23:04 Acked-by: Leon Woestenberg May 04 16:23:04 Acked-by: Mike Westerhof May 04 16:23:05 Signed-off-by: Tom Rini May 04 16:23:07 03Tom Rini  07org.openembedded.dev * r69cc656f42 10openembedded.git/recipes/glibc/ (11 files): May 04 16:23:12 glibc: In various old recipes add RPROVIDES virtual-libc-dev, bump PR. May 04 16:23:14 In some cases we also add in 'libc-dev' if nothing was being provided and May 04 16:23:17 add RDEPENDS += linux-libc-headers-dev if that was missing. May 04 16:23:18 Acked-by: Leon Woestenberg May 04 16:23:20 Acked-by: Mike Westerhof May 04 16:23:22 Signed-off-by: Tom Rini May 04 16:23:25 03Tom Rini  07org.openembedded.dev * r19641ab1b2 10openembedded.git/conf/distro/include/ (eglibc.inc glibc.inc uclibc.inc): May 04 16:23:28 Add distro inc files for eglibc, glibc and uclibc. May 04 16:23:29 This gives us the libc into OVERRIDES, of glibc for glibc/eglibc and uclibc May 04 16:23:31 for uclibc. We also ?= for libintl/libiconv and = set the rest of the virtuals. May 04 16:23:33 Acked-by: Leon Woestenberg May 04 16:23:35 Acked-by: Mike Westerhof May 04 16:23:37 Signed-off-by: Tom Rini May 04 16:23:49 hrw, your commit is bogus btw May 04 16:23:49 03Tom Rini  07org.openembedded.dev * r47a04595e6 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev May 04 16:23:52 03Tom Rini  07org.openembedded.dev * rd59fc6ce47 10openembedded.git/conf/distro/ (33 files in 2 dirs): May 04 16:23:55 Debian gcc-3.4 doesn't work, that's the problem May 04 16:23:58 All distros: Bring in conf/distro/include/${LIBC}.inc May 04 16:24:00 With minimal/minimal-uclibc we have to do a little rework so that May 04 16:24:02 minimal-uclibc can override the main distro. May 04 16:24:04 Acked-by: Leon Woestenberg May 04 16:24:06 Acked-by: Mike Westerhof May 04 16:24:10 Signed-off-by: Tom Rini May 04 16:24:37 I suppose it's ok as those older 3.3/3.2s also probably don't work May 04 16:24:46 Tartarus: worked so far here May 04 16:25:41 hrw, debian or ubuntu and what version? May 04 16:25:50 The reason for my changes this weekend was yet another 'gcc-3.4 didn't compile qemu' report May 04 16:25:58 debian unstable May 04 16:26:04 gcc -v on it? May 04 16:26:08 mom May 04 16:26:16 Using built-in specs. May 04 16:26:16 Target: x86_64-linux-gnu May 04 16:26:16 Configured with: ../src/configure -v --with-pkgversion='Debian 4.3.3-8' --with-bugurl=file:///usr/share/doc/gcc-4.3/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --enable-shared --enable-multiarch --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --enable-nls --with-gxx-include-dir=/usr/include/c++/4.3 --program-suffix=-4.3 --enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc - May 04 16:26:24 Thread model: posix May 04 16:26:27 gcc version 4.3.3 (Debian 4.3.3-8) May 04 16:26:28 Reading specs from /usr/lib/gcc/x86_64-linux-gnu/3.4.6/specs May 04 16:26:28 wrong one? :) May 04 16:26:30 ops May 04 16:26:33 Configured with: ../src/configure -v --enable-languages=c,f77 --prefix=/usr --libexecdir=/usr/lib --with-gxx-include-dir=/usr/include/c++/3.4 --enable-shared --with-system-zlib --enable-nls --without-included-gettext --program-suffix=-3.4 --enable-__cxa_atexit x86_64-linux-gnu May 04 16:26:37 Thread model: posix May 04 16:26:39 gcc version 3.4.6 (Debian 3.4.6-9) May 04 16:26:45 ok May 04 16:27:04 * Tartarus installs jaunty's one real quick May 04 16:27:14 that's an alleged 3.4.6 too May 04 16:27:29 but really, fuckit, lets get qemu 0.10.x working for what we need qemu for instead May 04 16:27:41 yep May 04 16:28:15 and yay, 2 bits of green on my buildbot now May 04 16:51:40 03Angus Ainslie  07fso/milestone5.5 * rfe2812e471 10openembedded.git/recipes/connman/ (6 files): May 04 16:51:40 connman : rename dbus file connman to org.moblin.connman.service May 04 16:51:40 bump PR on all connman files for fix May 04 16:53:55 Question: Is there a way to temporarily override rmwork in the INHERIT variable? I want to leave the work for just 1 build, but not all the builds I do afterward May 04 16:54:56 why not remove it? May 04 16:55:04 from config May 04 16:55:13 I can, but I figured I'd ask if there's a way to override on the commandline first. May 04 16:56:54 meh....I just removed it from my local.conf temporarily. Thanks. May 04 16:57:21 * hrw -> off May 04 16:59:08 shadowland: some vars in the environment are allowed into the bitbake build, but config files *always* override what came in via the environment, not vice versa May 04 17:01:21 kergoth: Oh, okay, I didn't know that. That explains why I had to remove MACHINE from my local.conf when I was trying to override it on the commandline May 04 17:03:22 <[denix]> re May 04 17:04:38 hoj: ping May 04 17:05:54 shadowland: make behaves the same way when it comes to env stuff, but make provides a way to set vars on its actual commandline, as arguments, which bitbake does not. i could see that being a worthy addition to bitbake 2.0 though May 04 17:05:56 shadowland: heh May 04 17:06:17 (if you hadnt noticed, we do take a variety of bits, syntactically, from make. :=, ?=, etc) May 04 17:06:55 denix: hello May 04 17:08:21 hoj: I was going through the logs and saw you had some questions about arago last week, while I was on vacation... May 04 17:10:04 hoj: feel free to ping me. I also joined #arago in case you don't want to bother #oe guys, but they are very helpful and covered for me nicely! :) May 04 17:13:47 denix: hey May 04 17:13:57 denix: about akita and udev cache May 04 17:14:02 Jay7: hi! how's it going? May 04 17:14:25 Jay7: yeah, I just replied to angstrom-dev... May 04 17:14:37 denix: fine, we have found that w/o your commit it is working ok :) May 04 17:14:45 denix: yep, I've too :) May 04 17:15:08 Jay7: sorry for breaking it for you :) May 04 17:15:08 denix: we have no separate /tmp and rootfs is r/o at that moment May 04 17:15:32 denix: np :) just looking for way to fix this May 04 17:15:35 Jay7: basically, my point is that /etc is also read-only and it doesn't make sense to create a dev cache tarball in /etc... May 04 17:16:01 I'm surprised /tmp is not writable... May 04 17:16:18 denix: can we move caching later? May 04 17:16:27 Jay7: Do you know if /var/tmp is? May 04 17:16:53 Jay7: hmm. it's part of udev startup - depending on the cache it starts udev differently... May 04 17:17:18 well May 04 17:17:25 can we create tarball in /dev May 04 17:17:44 then move it to somewhere else by init.d script? May 04 17:18:29 /var is placed on rootfs too May 04 17:18:44 we have /var/volatile on tmpfs May 04 17:18:55 basically, if cache exists and valid, use it and don't enumerate some devices May 04 17:19:08 (at least on my old angstrom stable) May 04 17:19:30 /var/volatile can be used for temp files May 04 17:19:41 /dev/shm too :) May 04 17:19:59 or just /dev May 04 17:20:38 we can create archive in some tmpfs then move to appropriate place on rootfs by some init.d script May 04 17:20:46 udev can read tarball from there May 04 17:20:47 is root remounted r/w later? May 04 17:20:51 yes May 04 17:21:15 yeah, I guess we can split it in 2 steps May 04 17:21:31 denix: Welcome back. I was wondering about if arago even had an irc channel and about the state of the 2.6.29 kernel for omxp35x evm (specifically USB). May 04 17:21:52 strange that no one else have found this behaviour.. May 04 17:22:00 seems all have writeable /tmp May 04 17:22:09 I wonder how devcache used to work before my change... I can only guess since it was failing to create tarball in /etc it wasn't even used May 04 17:23:11 denix: have you time now to split this into two steps? May 04 17:23:28 Jay7: it's not just writable /tmp, but also /etc - otherwise it just can't create the cache May 04 17:24:00 we have bluelightning on channel for tests ;) May 04 17:24:06 and I can test in qemu May 04 17:25:24 hoj: you are the first to request the arago irc channel (which is now live). as of omap35x kernel support - it depends which kernel you use (upstream, oe or TI PSP) May 04 17:25:49 hi denix May 04 17:26:02 yes, happy to test stuff :) May 04 17:26:51 Jay7: I'm still catching up with stuff while I was out... but I guess this requires fixing asap May 04 17:26:52 denix: I used the default kernel that the arago-base-image built May 04 17:26:55 bluelightning: hey May 04 17:28:14 hoj: arago-base-image doesn't build any kernels :) anyway, I'll switch to #arago... May 04 17:38:16 Jay7, bluelightning: can confirm dev cache was not used before? have you seen any /etc/dev.tar files before? May 04 17:39:49 denix: I've never looked, but there are none now at least... May 04 17:41:08 denix: moment May 04 17:42:17 denix: no, I have no such file May 04 17:43:28 that's the point - dev cache never worked for your platform, as it failed to create /etc/dev.tar May 04 17:43:50 so, it should be safe to just disable it completely in /etc/default/udev May 04 17:44:15 yep :) May 04 17:44:22 now, it can be made to work, as Jay7 suggested by copying the cache in /etc later on... May 04 17:45:39 my suggestion would be to have akita-specific default file with devcahe disabled for now, until we come up with proper solution May 04 17:45:54 well.. zaurus should not be rebooted frequently because of hw suspend/restore :) May 04 17:46:03 denix: akita, c7x0 and spitz May 04 17:46:26 may be I'll look at this.. May 04 17:46:52 anyone know why bridge-utils doesn't install the up/down scripts in /etc/network/ - do they simply not work with busbox ifupdown applet? May 04 17:46:54 bluelightning: ok, thanks. that temporary solution will at least make those platforms to boot again... :) May 04 17:47:59 denix: can you commit this temporary fix? May 04 17:48:44 or there is no way to easy select zauruses? May 04 17:49:31 Jay7: need to list them one by one, I guess May 04 17:50:08 ccc May 04 17:52:28 bluelightning: what about osa? May 04 17:52:33 e.g. tosa May 04 17:52:53 denix: I'm not sure, I don't know much about tosa May 04 17:54:58 denix: should be the same.. May 04 17:55:20 but better to ask XorA May 04 17:55:30 * Jay7 is still waiting his tosa to arrive May 04 18:01:08 03Denys Dmytriyenko  07org.openembedded.dev * rf9b60e552a 10openembedded.git/recipes/udev/ (5 files in 5 dirs): May 04 18:01:08 udev 124: disable dev cache on akita, c7x0, spitz and tosa May 04 18:01:08 above platforms have /tmp and /etc on read-only rootfs during udev start May 04 18:01:32 thanks denix :) May 04 18:01:39 03Angus Ainslie  07fso/milestone5.5 * r64e0efda59 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs : bump frameworkd May 04 18:02:39 bluelightning: let me know if it works - I'm lazy to build entire fs for akita to verify if it picks the right default file... :) May 04 18:02:57 denix: I will try it now May 04 18:03:49 bluelightning: thanks May 04 18:06:00 I'll try this too May 04 18:07:15 hmm, since recent pull I'm now getting "maximum recursion depth exceeded while calling a Python object" errors during bitbake cache update May 04 18:08:13 yeah I get that too May 04 18:08:29 I am so glad someone else noticed, I thought I had screwed my config over May 04 18:08:41 doesn't sound good :/ May 04 18:08:58 I'm gonna try downgrading to 1.8.12 release and see if that fixes anything May 04 18:09:25 which bb are you using? the one from stable? May 04 18:09:32 Neko: hmm, well I've been using 1.8.13 for a few weeks now, haven't had this issue until I just pulled May 04 18:09:34 I just pasted in the line from the wiki :] May 04 18:09:50 yeah me too but it doesn't read local.conf for some reason so I had to hack around a little bit May 04 18:10:03 and then I got all these parse errors and recursion stuff and "cmp" failures May 04 18:10:13 I should probably not have pulled May 04 18:15:00 I have such messages too May 04 18:15:14 NOTE: :maximum recursion depth exceeded in cmp while evaluating:${MACHINE_KERNEL_PR} May 04 18:15:27 while parsing May 04 18:22:47 downgrading now to see if it fixes it May 04 18:25:36 1.8.12 still does the recursion thing May 04 18:26:13 maybe the openembedded repo classes are the problem May 04 18:28:39 hm origin/stable/2009 does the same thing May 04 18:29:05 gremlin[it]: hi May 04 18:29:29 hi mckoan ! May 04 18:29:35 is there any 'official' information about this error? :) May 04 18:30:50 gremlin[it]: have you build kaeilos distro recently ? May 04 18:31:33 mckoan, building right now ... i'm bacl to OE world just now (ho firmato il preliminare per la casa) May 04 18:33:24 hm.. May 04 18:33:32 commit 85f0c2d5e5646795506e5966a9c95c7dfa685d8f May 04 18:33:43 * Jay7 suspect this ^^ May 04 18:34:42 gremlin[it]: great news! best wishes for your new home :-.) May 04 18:35:21 Jay, what exactly is it> May 04 18:35:35 gremlin[it]: which branch/machine/recipe are you using? May 04 18:36:22 I can't find that commit somehow May 04 18:36:32 mckoan, stable/2008 qemux86 consoleimage (kaleios) May 04 18:36:53 mckoan, stable/2009 ! May 04 18:37:00 Neko: http://cgit.openembedded.net/cgit.cgi?url=openembedded/commit/&id=85f0c2d5e5646795506e5966a9c95c7dfa685d8f May 04 18:37:08 but I'm not sure May 04 18:37:15 just first found candidate May 04 18:37:37 aha okay May 04 18:37:39 gremlin[it]: my daughter says hello to you May 04 18:37:44 well that's a simple thing to test out May 04 18:38:02 mckoan, hahah reply to her ! May 04 18:39:17 gremlin[it]: I suspect console-image would build smoothly. The problem is x11-image May 04 18:39:37 mckoan, after i build also x11-image May 04 18:40:07 Jay7, nah that doesn't change a thing. May 04 18:41:01 well.. looking deeper :) May 04 18:41:11 gremlin[it]: ok, please let me know about results May 04 18:41:30 * Jay7 suspects that hrw|gone can know something about this :) May 04 18:42:31 morning May 04 18:42:36 whats up? May 04 18:42:39 hrw: hi May 04 18:42:50 MKP? May 04 18:43:01 03Roman I Khimov  07org.openembedded.dev * r1a77f2ae75 10openembedded.git/ (8 files in 2 dirs): May 04 18:43:01 qemu: add version 0.10.3, initial May 04 18:43:01 Finally not requiring GCC3. Added with masked preference, need to carefully May 04 18:43:01 sort out and apply OE's patches. May 04 18:43:01 Acked-by: Tom Rini May 04 18:44:21 hrw: can you say something about NOTE: :maximum recursion depth exceeded in cmp while evaluating: ${MACHINE_KERNEL_PR} May 04 18:44:55 hrw: three peoples with todays pull have this at least May 04 18:45:16 I wonder if this is to do with PV containing PR May 04 18:45:23 reparsing May 04 18:45:45 fsck.. my .dev parsing again halt on 70th recipe May 04 18:46:01 NOTE: Handling BitBake files: \ (0070/6721) [ 1 %]DEBUG: Cache: /home/hrw/devel/oe/openembedded/recipes/at76c503a/at76c503a-modules_0.17.bb is not cached May 04 18:47:11 btw May 04 18:47:28 ok, will revert May 04 18:47:35 this is somehow affected by require modules.inc May 04 18:47:43 I'm getting it with stable/2009 and 1.8.12 too though, maybe I am screwing SOMETHING up here May 04 18:47:44 grr. no May 04 18:48:00 inherit module May 04 18:48:24 I found that recipes which inherits module affected May 04 18:48:44 03Marcin Juszkiewicz  07org.openembedded.dev * ra15c4725b2 10openembedded.git/conf/bitbake.conf: May 04 18:48:44 Revert "bitbake.conf: make MACHINE_KERNEL_PR defaults to PR" May 04 18:48:44 This reverts commit 85f0c2d5e5646795506e5966a9c95c7dfa685d8f. May 04 18:48:44 It was pushed by mistake, sorry. May 04 18:49:25 Jay7: that would agree with hrw crash above :-) May 04 18:50:04 yes that fixes thing :) May 04 18:50:11 hrw: thanks :) May 04 18:52:45 hrw, Hi! I have still trouble with rpm build ... on kaelios at least ... May 04 18:53:38 gremlin[it]: it was working for me when I pushed. tested under debian/sid and ubuntu/8.04 (or 8.10). never tried later May 04 18:54:30 hrw, i suspect some interaction with native Fedora rpm/yum ... it fail on createrepo step ... May 04 18:54:52 hrw, http://tinderbox.openembedded.net/public/logs/4920618.txt May 04 18:56:01 gremlin[it]: I prefer to stay with 'I have no knowledge about rpm/yum' May 04 18:56:26 hi eFfeM May 04 18:56:57 hrw ... ghhghg ok ... but as soon i'll be able to investigate i'll ask for some info about OE ;) May 04 18:57:19 hi hrw May 04 18:59:22 gremlin[it]: I did part of 'lets generate working rpm packages and build working images from them and make upgrade/install in them working fine' job just because there was a possibility that Poky will get used in Moblin and I was paid to do that. Then I decided that it would be shame if such stuff will get wasted in Poky and worked on merging it into OE. once merged I prefer to forget that rpm/yum exists. May 04 19:00:04 hrw, poky people can help me ? May 04 19:00:17 gremlin[it]: you mean RP? May 04 19:00:27 gremlin[it]: I think that he is overloaded May 04 19:00:38 hrw, yes and few others ... May 04 19:01:25 gremlin[it]: rpm/yum was mostly RP and me May 04 19:01:26 hrw, ok ... as sook kaleios x11-image will build fine i'll check rpm ... and let's see May 04 19:05:40 gremlin[it]: a silly question: why don't you use .deb or .opk? May 04 19:06:30 * mckoan is wearing his Debian/rules T-shirt bought at fosdem May 04 19:06:39 maybe he is used to rpms? May 04 19:06:50 reverting that commit didn't change a thing here?? May 04 19:06:59 mckoan, cause on all my desktop machine I use Fedora/CentOS ... and i'm close to try Ferora ARM I'll like to have rpm also on embedded devices ... May 04 19:07:01 I am wearing OpenedHand t-shirt which I got at first OEDEM May 04 19:07:06 also MACHINE_KERNEL_PR isn't in any of my configs May 04 19:07:15 hrw: I also was used to rpms in 90s but I overcomed them May 04 19:07:16 Neko: for me all is working now May 04 19:07:24 what branch are you on? :/ May 04 19:07:27 Neko: but it is in one of main classes which is also used in your build May 04 19:07:35 Neko: it was in .dev May 04 19:08:04 Neko: try .dev and try killing the cache in tmp May 04 19:08:17 gremlin[it]: I understand May 04 19:08:19 denix: touch conf/local.conf is enough May 04 19:08:32 hmn May 04 19:08:49 well.. x11-image built May 04 19:08:57 denix: trying now :) May 04 19:09:03 hrw: true :) May 04 19:09:05 * Neko tiptaps as it checks out May 04 19:10:18 Jay7: akita? May 04 19:10:28 denix: akita image in qemu-system-arm :) May 04 19:10:39 ok May 04 19:14:19 /etc/init.d/rc: /etc/rcS.d/S37populate-volatile.sh: line 171: can't create /var/tmp/tmpdefined.252: nonexistent directory May 04 19:14:19 hehe.. May 04 19:14:29 denix: udev part is working ok May 04 19:14:49 but now I have something strange with volatile :) May 04 19:18:38 have a nice rest of day all May 04 19:31:51 mckoan, there is something strange also on console-image May 04 19:33:37 gremlin[it]: what? May 04 19:34:38 mckoan, some error about "no space left" ... May 04 19:34:51 mckoan, example when dropbear start May 04 19:35:59 mckoan, "cd /etc; touch foo --> touch: foo: No space left on device" ... mhhhh May 04 19:36:14 gremlin[it]: do you mean during qemu boot ? May 04 19:36:26 yes, but also after the login May 04 19:37:04 I have been facing too it too May 04 19:45:01 mckoan, solution ? May 04 19:47:33 gremlin[it]: not yet, I have been working on it May 04 20:04:38 hi mckoan May 04 20:06:08 Gnutoo: hi May 04 20:07:04 mckoan, I've a friend that is doing the port of the linux kernel on the eten m810(similar to the m800 that harald welte ported the linux kernel on it) and he's interrested in the jtag that is here: http://www.koansoftware.com/it/prd_peedi.htm May 04 20:07:42 the differences between m800 and m810 are the screen,the keyboard,the gpio that are different etc... May 04 20:08:04 he has now the screen partially working(problem with the touchscreen) May 04 20:08:27 how much does this jtag thing cost? May 04 20:09:32 Gnutoo: ok I'm selling it, email me privately May 04 20:09:46 mckoan, ok May 04 20:11:51 thanks May 04 20:12:52 Gnutoo: do you want to use peedi jtag debugger with eten m810? May 04 20:13:04 yes he wants to May 04 20:13:23 what is the cpu? May 04 20:13:39 the same as the openmoko...so it should be a samsung one May 04 20:13:44 I'll check May 04 20:13:56 ah CPU not SOC May 04 20:14:42 and does it have a jtag connector? May 04 20:15:49 ARM920T I think May 04 20:15:53 yes it's under the battery May 04 20:16:46 ok so an ARM7/ARM9 model should be ok May 04 20:17:57 ok May 04 20:42:35 good night May 04 21:01:08 denix: seems to be working fine here May 04 21:02:23 bluelightning: good, thanks for testing it! I should look on how to enable it properly for those platforms... May 04 21:40:14 * Jay7 -> sleep May 04 21:40:22 * florian too May 04 21:40:25 good night May 04 21:40:37 good night too :) May 04 23:09:20 03Andrea Adami  07org.openembedded.dev * r2b9b2a2ca9 10openembedded.git/recipes/teleport/ (5 files in 2 dirs): May 04 23:09:20 teleport: remove unused 0.33 version (broken atm - may 2009) May 04 23:09:20 /oe/org.openembedded.dev/conf $ grep -R teleport . May 04 23:09:20 ./checksums.ini:[http://gpe.linuxtogo.org/download/source/teleport-0.33.tar.gz] May 04 23:09:20 ./checksums.ini:[http://gpe.linuxtogo.org/download/source/teleport-0.34.tar.gz] May 04 23:09:23 ./distro/include/preferred-gpe-versions-2.8.inc:PREFERRED_VERSION_teleport ?= "0.34" May 04 23:09:25 ./distro/include/preferred-om-2008-versions.inc:PREFERRED_VERSION_teleport ?= "0.34" May 04 23:24:57 03Andrea Adami  07org.openembedded.dev * rbf111405e1 10openembedded.git/recipes/gpe-su/ (gpe-su_0.11.bb gpe-su_0.14.bb gpe-su_0.15.bb): (log message trimmed) May 04 23:24:57 gpe-su: remove unused versions May 04 23:24:57 /oe/org.openembedded.dev $ andrea@mizar May 04 23:24:57 /oe/org.openembedded.dev/conf $ grep -R gpe-su . May 04 23:24:57 ./checksums.ini:[http://gpe.linuxtogo.org/download/source/gpe-su-0.11.tar.gz] May 04 23:25:01 ./checksums.ini:[http://gpe.linuxtogo.org/download/source/gpe-su-0.14.tar.gz] May 04 23:25:03 ./checksums.ini:[http://gpe.linuxtogo.org/download/source/gpe-su-0.15.tar.gz] May 04 23:25:08 03Andrea Adami  07org.openembedded.dev * r25e3f8d497 10openembedded.git/conf/checksums.ini: teleport_0.33.bb: remove from checksums.ini May 04 23:25:10 03Andrea Adami  07org.openembedded.dev * r6923d7073a 10openembedded.git/conf/checksums.ini: gpe-su: remove unused entries from checksums.ini May 05 00:26:42 * denix is away: I'm not here May 05 02:43:06 Hi everyone May 05 02:45:06 Does anyone have an idea what this error means: undefined reference to `__nldbl_fprintf'? **** ENDING LOGGING AT Tue May 05 02:59:57 2009