**** BEGIN LOGGING AT Thu Mar 17 02:59:57 2011 Mar 17 03:02:38 do_rootfs fails in oe-core? http://pastebin.com/qeResjq4 Mar 17 03:20:44 those loop messages are warnings not errors.. Mar 17 03:21:02 what is your host system type? it looks a lot like you have the same behavior as Khem was reporting on Ubuntu 11.04-alpha.. Mar 17 03:21:07 I don't have access to a host like that to figure out the issue.. Mar 17 03:21:28 good to know we're not the only one that say ERROR when we mean warning ;) Mar 17 03:21:32 s/one/ones/ Mar 17 03:22:09 kergoth_ ya, I had a patch to remove the message.. (I probably should litterly change it to "warning"...) but I decided to leave the message because it is a real looping problem, something we do eventually need to resolve Mar 17 03:22:20 * kergoth_ nods Mar 17 03:22:22 ---but---- that should not cause an install failure ;) Mar 17 03:25:16 denix -- if you see this and I'm gone.. could you email (or file a bug in the bugzilla.yoctoproject.org) about your machine setup (host type, processor type..) Mar 17 03:28:24 fray: Ubuntu 10.04.2 Mar 17 03:31:07 why is it I see so many problems with building on Ubuntu (other than the bash/dash thing)? Mar 17 03:32:41 no problems with OE though Mar 17 03:37:02 ahh back now Mar 17 03:37:25 denix, as a work around, I suggest you setup your configuration to use opkg instead of RPM for the time being.. Mar 17 03:37:39 I've not seen any reported failures on Ubuntu 10.04 before.. so something odd is happening.. Mar 17 03:38:15 fray: I left default rpm+opkg, I'll disable rpm for now Mar 17 03:38:24 (I do want RPM to work btw...) but I need to step through it and figure out what is going wrong.. it might or might not be the same problem that khem is seeing on his machine.. Mar 17 03:38:44 should I still fill out the bugreport? Mar 17 03:39:36 denix if you have time to experiment.. I'd suggest you edit meta/classes/package_rpm.bbclass Mar 17 03:39:40 yes please file a bug Mar 17 03:39:49 in package_rpm.bbclass, search for: Mar 17 03:40:00 # Attempt install Mar 17 03:40:00 ${RPM} --root ${target_rootfs} \ Mar 17 03:40:13 add "-vv" after the ${RPM}, and attach the log to the bug.. Mar 17 03:40:43 From Khem's log.. it appears that RPM started processing and then simply stopped as if nothing needed to be done and quit.. Mar 17 03:40:47 (very confusing) Mar 17 03:55:47 fray: http://bugzilla.pokylinux.org/show_bug.cgi?id=890 Mar 17 03:55:54 thanks Mar 17 03:59:31 denix -- this looks like the same problem as Khem.. Mar 17 04:00:06 I need to get in there with a debugger and figure out what it's doing.. the behavior appears to be that it's about to install and suddenly it gets a catestrophic failure and exits.. (usually this is due to a failed chroot or similar) Mar 17 04:00:15 which if thats the case, means pseudo isn't working right.. Mar 17 04:00:36 I'll see what I can do about this in the morning Mar 17 04:04:29 fray: thanks. I'll disable rpm for now. Mar 17 04:04:50 hopefully I'll be able to reproduce, file and fix the issue.. Mar 17 04:04:59 g'night all Mar 17 04:06:51 gn Mar 17 04:07:21 nite fray Mar 17 05:02:53 denix: I am using angstrom+oe-core now Mar 17 05:02:58 denix: works well Mar 17 05:03:56 denix: http://sakrah.homelinux.org/blog/2011/03/using-openembedded-core-to-build-angstrom-for-qemu/ Mar 17 05:04:00 works like a charm Mar 17 05:04:39 it does not use rpm as default package manager so I am avoiding the problem Mar 17 07:38:55 03Koen Kooi  07org.openembedded.dev * rb0cb26dddb 10openembedded.git/recipes/linux/linux-omap-psp_2.6.32.bb: Mar 17 07:38:55 linux-omap-psp 2.6.32: bump PR Mar 17 07:38:55 Signed-off-by: Koen Kooi Mar 17 07:39:11 03chase maupin  07org.openembedded.dev * r55f8ed6f98 10openembedded.git/ (6 files in 5 dirs): (log message trimmed) Mar 17 07:39:11 dm368-evm: Add configuration for DM368 machine Mar 17 07:39:11 * Added the configuration for the dm368-evm machine type. Mar 17 07:39:11 * Updated the sort.sh script Mar 17 07:39:12 * Updated the Linux kernel and u-boot recipes to support this Mar 17 07:39:12 new device. Mar 17 07:39:13 Signed-off-by: Chase Maupin Mar 17 07:39:19 03chase maupin  07org.openembedded.dev * r9849191099 10openembedded.git/recipes/ti/ti-pru-sw-examples_1.00.00.bb: (log message trimmed) Mar 17 07:39:19 ti-pru-sw-example: fix building with new toolchain sysroot Mar 17 07:39:19 * Patch taken from Arago overlay where it was added by Mar 17 07:39:19 Denys Dmytriyenko Mar 17 07:39:19 * PR incremented by two to align recipe PR values between Arago Mar 17 07:39:19 and OE. Mar 17 07:39:19 Signed-off-by: Chase Maupin Mar 17 07:39:20 03chase maupin  07org.openembedded.dev * re65b9e6061 10openembedded.git/recipes/u-boot/u-boot.inc: (log message trimmed) Mar 17 07:39:21 u-boot.inc: allow variable make targets Mar 17 07:39:22 * Later versions of u-boot sometimes have top level make targets Mar 17 07:39:22 such as u-boot.imx, u-boot.kwb, and for upcoming TI devices Mar 17 07:39:22 u-boot.ti. Mar 17 07:39:23 * These targets define different calls to the mkimage tool to Mar 17 07:39:23 change the images created, such as changing the load address. Mar 17 07:39:24 03chase maupin  07org.openembedded.dev * rbbe144e058 10openembedded.git/recipes/ti/ (matrix-gui-common.inc matrix-gui-e.inc matrix-gui.inc): Mar 17 07:39:24 matrix-gui: add support for dm386-evm machine Mar 17 07:55:26 gm Mar 17 08:50:13 morning Mar 17 08:58:50 hiho Mar 17 09:01:39 morning Mar 17 09:01:39 hi woglinde Mar 17 09:01:52 well.. new bugs Mar 17 09:01:52 woglinde: libecj-bootstrap-3.6.1-r0: task do_compile: Failed same problem Mar 17 09:02:41 ericben Mar 17 09:02:45 hrms Mar 17 09:02:48 thats strange Mar 17 09:02:56 I will check my logs of jikes here Mar 17 09:03:04 ericben do you have a host jikes installed? Mar 17 09:03:17 hm.. Mar 17 09:03:30 libgrypt still failing with ICE Mar 17 09:03:37 libgcrypt Mar 17 09:03:48 I was sure that khem fixed this Mar 17 09:04:51 ericben can you pastebin it again? Mar 17 09:04:57 I want to show it robert Mar 17 09:06:17 woglinde: I think I don't have jikes on the host apt-cache search jikes returns nothing but maybe it's in an other package name ? Mar 17 09:06:26 hm Mar 17 09:06:32 which jikes in cmdline? Mar 17 09:06:54 woglinde: I checked and no I don't have any exec names jikes Mar 17 09:07:34 hm have a lot of warnings too Mar 17 09:07:37 but not errors Mar 17 09:08:57 ericben the pastebin? Mar 17 09:09:16 woglinde: it's coming Mar 17 09:09:48 woglinde: http://pastebin.com/j0e9PtGz Mar 17 09:10:41 woglinde: btw, I get the exact same thing on minimal distro (target is armv5) Mar 17 09:10:54 but at this stage I think target is not a problem Mar 17 09:13:24 ericben do you have another buildhost? Mar 17 09:13:49 woglinde: this one is a ubuntu x86 and I have a debian amd64 on which I can test Mar 17 09:13:55 and a FC14 also Mar 17 09:14:03 which one do you prefer ? Mar 17 09:14:37 try the fc14 Mar 17 09:15:07 ok the slowest build machine :) Mar 17 09:15:23 I keep you informed Mar 17 09:42:25 with OE 2011.03, I'm getting undefined reference to getaddrinfo in util-linux-ng, ideas? Mar 17 09:46:11 akheron: uclibc? which distro/machine? Mar 17 09:46:22 uclibc Mar 17 09:46:43 my own distro and machine, but basically minimal-uclibc and mpc8315e-rdb Mar 17 09:47:06 maybe I should test with those first... Mar 17 09:48:24 hmm, ah, my uClibc.machine and uClibc.distro may be wrong Mar 17 09:48:41 according to this table http://www.openembedded.org/index.php/Release-2011.03 Mar 17 09:48:45 it should be ok Mar 17 09:48:47 that would explain it Mar 17 09:49:01 ynezz: yes, I also looked at that table Mar 17 09:49:24 well it didn't build for me, but for khem yes Mar 17 09:50:10 combination of efikamx/minimal-uclibc/console-image was ok for khem, but ts72xx/minimal-uclibc/console-image didn't build for me Mar 17 09:50:45 but khem didn't supplied git hash for his build so who knows what he had build :) Mar 17 09:51:40 and tom rini build mpc8315e-rdb/minimal-uclibc/console-image succesfully Mar 17 09:52:20 but I don't know if util-linux-ng is part of console-image Mar 17 09:59:52 true Mar 17 10:01:01 ynezz: testbuilder 0.6.1 should have all you need wrt ANGSTROMLIBC :) Mar 17 10:01:04 re Mar 17 10:03:45 is there a smaller package to check than 'bitbake nano'? Mar 17 10:04:15 bitbake gcc-cross Mar 17 10:04:25 bitbake hello? Mar 17 10:04:50 also if i change the MACHINE:= just after a broken build, do i need cleanup things? Mar 17 10:05:57 changed from 'Beagleboard' to 'at91sam9263ek' Mar 17 10:06:48 | Error: alsa check failed Mar 17 10:06:50 | Make sure to have the alsa libs and headers installed. Mar 17 10:06:54 qemu-0.14.0 Mar 17 10:07:13 jay7 good catch Mar 17 10:07:19 woglinde: seems to work on FC14 x86_64 Mar 17 10:07:22 we dont need alsa stuff either on host qmeu Mar 17 10:07:28 woglinde: without jikes installed Mar 17 10:07:29 args qemu-native Mar 17 10:07:41 ericben hm I am bit puzzeld Mar 17 10:07:42 NOTE: package qemu-native-0.14.0-r2: task do_configure: Failed Mar 17 10:07:46 woglinde: yes, -native Mar 17 10:07:56 woglinde: build is not yet finished but I see libecj-bootstrap-3.6-r0.do_compile in stamps Mar 17 10:08:07 ericben sure Mar 17 10:08:11 it will last a while Mar 17 10:08:15 2 hours Mar 17 10:08:21 min Mar 17 10:08:45 woglinde: I assume more as this is on a slow machine here Mar 17 10:08:51 Jay7: ah ok, thanks, will try it soon Mar 17 10:08:58 woglinde: so this may be a problem with the build host configuration ? Mar 17 10:10:00 ynezz: seems UCLIBC_HAS_IPV4 or 6 needs to be set for getaddrinfo et al, I'll need to update my uclibc config Mar 17 10:11:34 ericben as I said I never saw it before Mar 17 10:11:40 and have no idea Mar 17 10:11:45 all buildhost I have Mar 17 10:11:46 it works Mar 17 10:11:58 woglinde: ok now testing on debian amd64 Mar 17 10:12:37 that should work Mar 17 10:17:52 NOTE: Running task 4731 of 8627 (ID: 4017, /storage/oe/testbuilder/openembedded/recipes/pango/pango_1.28.3.bb, do_populate_sysroot) Mar 17 10:17:53 ERROR: fork failed: 12 (Cannot allocate memory) Mar 17 10:18:02 ben-nanonote failed second night.. Mar 17 10:18:26 only build failed with such message Mar 17 10:18:51 I am trying to build a small distribution with a given (binary) kernel and toolchain. Where should I start? I've successfully build the distribution for qemu, but need it on a specific device. Mar 17 10:20:16 Jay7: add more RAM :) Mar 17 10:20:45 ynezz: yeah :)) 4Gb isn't sufficient to build image for device with 32Mb RAM :) Mar 17 10:20:52 or lower number of threads? Mar 17 10:21:07 yeah I got the same issue out of memory issue after I updated bitbake Mar 17 10:21:16 builds for akita, collie, efikamx and qemux86 passed Mar 17 10:21:32 but it only seems to happen when bitbake recompiles its cache or whatever at the start Mar 17 10:21:36 well, not passed entirely but Mar 17 10:21:39 the time after that it uses less memory Mar 17 10:22:04 Jay7: and if you continue that build, it builds fine? Mar 17 10:22:32 ynezz: I don't know.. it's clean build in testbuilder so it is removed after :) Mar 17 10:22:38 ah Mar 17 10:22:39 I'll try to debug tonight Mar 17 10:22:56 maybe it's some race of compiling some huge c++ stuff concurently? who knows Mar 17 10:23:01 or no.. this night should be busy Mar 17 10:23:16 I've remember one fork-bomb we has some time ago :) Mar 17 10:23:34 well, I've build from scratch on master/master tonight and it was ok Mar 17 10:23:37 on 4GB also Mar 17 10:23:51 I mean bitbake master/oe master Mar 17 10:23:55 same here Mar 17 10:24:07 I'm sure this is only current build problem Mar 17 10:24:13 hm Mar 17 10:24:26 'bitbake hello' breaks Mar 17 10:24:34 NoProvider: hello Mar 17 10:24:38 angstrom-2010.x/ben-nanonote/console-image+x11-image+opie-image Mar 17 10:24:53 nataraj: helloworld iirc Mar 17 10:25:46 angstrom-2010.x/ts72xx and image similar to minimal-image (eglibc) Mar 17 10:27:02 so no pango Mar 17 10:29:36 and lot of warnings about multiple providers: http://pastebin.com/mmfj8VGS Mar 17 10:29:59 just as reminder :) Mar 17 10:31:23 Jay7: is this new after updating to bitbake 1.12.0 by any chance Mar 17 10:31:28 cos I think that's what happened to me Mar 17 10:41:16 woglinde: debian amd64 : virtual:native:/scratch/oe/cpu9krea/openembedded/recipes/jakarta-libs/log4j1.2_1.2.15.bb, do_compile) failed with exit code '1' Mar 17 10:41:53 seems java doesn't like debian based build hosts :) Mar 17 10:42:00 ????? Mar 17 10:42:05 I build it here all the time Mar 17 10:42:25 we build it 10 times on squeeze system in a row Mar 17 10:42:37 woglinde: ok, I may be unlicky but that fails on ubuntu x86 & debian amd64 for me. Mar 17 10:42:38 can you please pastebin the error log Mar 17 10:42:47 woglinde: of course Mar 17 10:44:24 woglinde: http://pastebin.com/Q0cjJbkG Mar 17 10:45:04 woglinde: btw I have 4 bbthreads and -j8 : can this be a problem here ? Mar 17 10:45:20 ericben oh Mar 17 10:45:23 maybee Mar 17 10:45:44 trying with 1 / 1 Mar 17 10:45:59 javax.mail.Transport cannpt be resolved Mar 17 10:46:00 hm Mar 17 10:47:18 hm Mar 17 10:47:20 okay Mar 17 10:48:28 ericben can please lookup to which binary the java exe points in your sysroot/x64 dir? Mar 17 10:49:09 either there is really a problem with gnumail producing sometimes classes with 0 bytes Mar 17 10:49:24 java -> /scratch/oe/cpu9krea/tmp_angstrom-2010.x_eukrea-cpu9260/sysroots/x86_64-linux/usr/bin/jamvm javac.jikes -> /scratch/oe/cpu9krea/tmp_angstrom-2010.x_eukrea-cpu9260/sysroots/x86_64-linux/usr/bin/jikes Mar 17 10:49:25 or you have jamvm-native as the java compiler Mar 17 10:49:29 yes Mar 17 10:49:41 why the hell you have jamvm-native again Mar 17 10:49:47 it should be cacao-native Mar 17 10:49:58 didnt you update to latest stuff? Mar 17 10:50:28 woglinde: oops not on this host Mar 17 10:50:36 hehe Mar 17 10:50:48 try to update Mar 17 10:50:49 woglinde: sorry Mar 17 10:51:02 and bitbake -c clean virtual/java-native Mar 17 10:51:04 on the ubuntu one I'm at b365c5f805a8fdf2ddc51b56e2feee0c69f166ec Mar 17 10:51:09 bitbake virtual/java-native Mar 17 10:51:12 woglinde: more sure : rm -rf tmp ;) Mar 17 10:51:22 see if it builds cacao-native Mar 17 10:51:33 jamvm segfaults from time time Mar 17 10:51:46 we didnt find the error Mar 17 10:52:00 but cacao-native from hg repos works fine Mar 17 10:52:07 btw. Mar 17 10:52:08 root@beagleboard:~# java -version Mar 17 10:52:09 java version "1.6.0_18" Mar 17 10:52:09 OpenJDK Runtime Environment (IcedTea6 1.8.5) (6b18-1.8.5) Mar 17 10:52:09 OpenJDK Shark VM (build 14.0-b16, mixed mode) Mar 17 10:52:43 un/lib/libc.so.0 -> libuClibc-0.9.32-rc2-git.so Mar 17 10:53:08 strange, got 4.4G space Mar 17 10:53:27 patch: **** write error : No space left on device Mar 17 10:53:55 hehe.. Mar 17 10:53:58 sorry Mar 17 10:54:07 thats not enough Mar 17 10:54:18 you will need 20 Mar 17 10:55:16 my home dir is just 20G Mar 17 10:55:40 df -h Mar 17 10:55:44 you will see Mar 17 10:55:54 nataraj: what you trying to build? Mar 17 10:55:55 /dev/sda9 20G 15G 4.4G 77% /home Mar 17 10:56:07 bitbake helloworld Mar 17 10:56:11 btw. qemu needs now over 800 mb in /tmp for linking Mar 17 10:56:35 any way to disable qemu? Mar 17 10:56:59 its not qemu I think Mar 17 10:57:02 i can put the arm binary in the target and checkout Mar 17 10:57:06 patch somewhere else Mar 17 10:57:07 we should really consider about lowering HDD requirements.. Mar 17 10:57:15 yocto have this in 1.1 features Mar 17 10:57:21 we may collaborate here Mar 17 10:58:57 any other suggestions? Mar 17 10:59:48 mv sources somewhere else? Mar 17 11:00:09 use BB_NUMBER_THREADS=1 (or how this var was named) Mar 17 11:00:22 INHERIT += "rm_work" Mar 17 11:01:12 and do not build all at once but step by step Mar 17 11:01:37 "bitbake qemu-native" "bitbake glibc" "bitbake gcc-cross" Mar 17 11:04:29 do i need qemu at all? Mar 17 11:04:59 hi hrw btw. Mar 17 11:05:55 Hi everybody Mar 17 11:05:58 nataraj: glibc uses it to generate locales Mar 17 11:07:57 ericben: I have build problems with Java/Cacao, too. On Ubuntu 10.04 / amd64 I had no luck at all (segmentation fault in ecj). On Ubuntu 10.04 / x86 it kind of works, but I have to restart the build several times. Mar 17 11:17:19 hrw: glibc wasn't converted to use cross-localedef-native? Mar 17 11:20:01 hi florian, mickeyl, all Mar 17 11:20:18 gm pb Mar 17 11:20:23 hi woglinde Mar 17 11:20:40 mkneissl we already changed it Mar 17 11:21:00 no idea Mar 17 11:21:01 mkneissl use cacoa-native as virtual/java-navit Mar 17 11:21:19 woglinde: That means, I should update via git? Mar 17 11:21:43 sure Mar 17 11:21:55 woglinde: Thanks. Mar 17 11:25:37 does anybody know if there is a free toolchain which targets Freescale 9S08? Mar 17 11:25:50 our hardware guys seem to have run amok and designed one of those into some or other product. Mar 17 11:26:33 what? Mar 17 11:26:52 yeah, quite Mar 17 11:27:08 worse, there is also an 8031 on there. Mar 17 11:27:16 SMP :-) Mar 17 11:28:12 what the hw guys did? Mar 17 11:28:18 invented own compiler? Mar 17 11:28:45 no, they just use the freescale-provided CodeWarrior(tm) for windows Mar 17 11:29:59 which sounds reasonable Mar 17 11:30:19 yeah, just annoying for me because I don't have a windows pc on my desk Mar 17 11:30:33 .o(virtualbox) Mar 17 11:31:01 yeah, I could use virtualbox but it's a bit slow and tedious Mar 17 11:31:19 hm Mar 17 11:31:28 also I think I would need to find a windows licence from somewhere, my current copy is not activated :-} Mar 17 11:31:32 yeah could be for codewarrior Mar 17 11:31:57 hm time for launch Mar 17 11:32:04 today I will made some pizza Mar 17 11:32:12 launch? crumbs Mar 17 11:32:15 ignition! Mar 17 11:32:18 args Mar 17 11:34:59 hm, the mailing list seems to be lagging again. I think there is something a bit weird going on with discovery. Mar 17 11:38:57 yes Mar 17 11:39:01 was yesterday too Mar 17 11:50:39 he, we have to clean minimal wrt multiple providers Mar 17 11:50:42 NOTE: multiple providers are available for runtime hotplug (hotplug-ng, hotplug2, udev, linux-hotplug) Mar 17 11:50:48 a bit too many... Mar 17 11:52:03 and the gtk/gdk too Mar 17 11:52:30 yeah, this is sort of a policy collision between minimal and oe. Mar 17 11:52:45 I'm not totally sure what the right answer to that is. Mar 17 11:56:23 hi obi Mar 17 11:57:55 hi! Mar 17 11:59:24 ant__: but those are runtime providers.. not so easy to fix as to define P_P Mar 17 12:08:27 I'll send patch for gtk+ providers to list Mar 17 12:42:55 why bitbake -i doesn't work anymore? Mar 17 12:43:51 0.10.2 Mar 17 12:44:25 ant__: gst-plugins also fixed Mar 17 12:47:16 hmm, it's just unsupported nowadays? Mar 17 12:47:34 akheron yes Mar 17 12:47:39 :( Mar 17 12:47:40 try bitbake -E Mar 17 12:47:46 or -e Mar 17 12:47:57 its not that brilliant as -i Mar 17 12:47:59 but works Mar 17 12:48:16 what -E does? Can't see it in --help Mar 17 12:48:30 -e sounds useful though Mar 17 12:48:38 that's what I usually used -i for :) Mar 17 12:48:44 also, should I use 1.10 or 1.12? Mar 17 12:48:49 or does it matter? Mar 17 12:49:10 use 1.12 its faster at parsing Mar 17 12:53:17 ok, let's try Mar 17 13:02:41 any idea why some machine set # tmpfs /var/volatile tmpfs defaults 0 0 and others Mar 17 13:02:41 tmpfs /var/volatile tmpfs mode=0755 0 0 ? Mar 17 13:03:06 or even volatile type tmpfs (rw,mode=755) Mar 17 13:03:32 why default is not enough? hmm.. Mar 17 13:08:58 woglinde: debian amd64: build ok, FC14: failed in icedtea6-native-1.7.6-r3.0 with /usr/bin/ld: cannot find -ljvm Mar 17 13:10:04 hi, CharDevice.h is from java? Mar 17 13:10:31 for me (old ubuntu based distro) java seemed to have built fine Mar 17 13:10:55 but....I've some com.buglabs issues (recipes are in recipes/buglabs-osgi/ Mar 17 13:10:57 ) Mar 17 13:11:36 I've spoted 2 kinds of errors until now: Mar 17 13:11:38 | Accelerometer.cpp:39:24: fatal error: CharDevice.h: No such file or directory Mar 17 13:11:50 | SerialImp.o: could not read symbols: Bad value Mar 17 13:12:00 I'll look with binutils for the second one Mar 17 13:17:34 http://www.pastie.org/1682520 that's for the symbol issue Mar 17 13:19:25 maybe I should follow the advise Mar 17 13:19:30 ( use -fPIC ) Mar 17 13:28:55 ericben???? it should 1.7.10 Mar 17 13:29:07 woglinde: forget it, the fc14 host was not on the right branch Mar 17 13:29:11 I updated it yesterday Mar 17 13:29:51 03Andrea Adami  07org.openembedded.dev * rbcf73715cf 10openembedded.git/recipes/base-files/ (6 files in 6 dirs): Mar 17 13:29:52 base-files: use default fstab for Zaurus devices. Bump PR. Mar 17 13:29:52 * Zaurus boots from different media and we can't make any guess. Mar 17 13:29:52 Signed-off-by: Andrea Adami Mar 17 13:56:13 why is glibc-2.9-r37.4 trying to install stuff to my root? | make[2]: Entering directory `/home/dropbox/Dropbox/jedahan/bug/rootfs/tmp/work/armv7a-angstrom-linux-gnueabi/glibc-2.9-r37.4/glibc-2.9/locale' Mar 17 13:56:13 | /home/jedahan/bug/rootfs/oe-buglabs/bin/install -c -m 644 locale.h /usr/include/locale.h Mar 17 13:56:13 | install: cannot remove `/usr/include/locale.h': Permission denied Mar 17 13:56:35 mickey|office: good afternoon Mar 17 13:58:37 good afternoon, pb__ Mar 17 13:58:41 lol Mar 17 13:58:47 cairo developer Mar 17 13:58:55 seems this comes from historic reasons Mar 17 13:59:07 int atomic_add(int i) { return __sync_fetch_and_add (&i, 1); } Mar 17 13:59:09 pb__: wrt gtk+.inc I didn't put it there because _virtclass-native override doesn't make sanse in recipes not using BBCLASSEXTEND and to make it "nicer" one should put at least all 3 vars BBCLASSEXTEDN/PACKAGES_DYNAMIC/PACKAGES_DYNAMIC_virtclass-native there (and I don't care about messy gtk+ dir enough to do it Mar 17 13:59:18 int atomic_cmpxchg(int i, int j, int k) { return __sync_val_compare_and_swap (&i, j, k); } Mar 17 13:59:36 when this compiles fine -> cairo_cv_atomic_primitives="Intel" Mar 17 13:59:38 is set Mar 17 13:59:40 muahaha Mar 17 13:59:43 JaMa: the _virtclass-native override won't do any harm to recipes that don't set BBCLASSEXTEND, it'll just have no effect. Mar 17 14:00:11 I know but PACKAGES_DYNAMIC is compied in every gtk+ recipe now Mar 17 14:00:26 next test is for libatomic-ops Mar 17 14:00:27 I don't see how one PACKAGES_DYNAMIC_virtclass-native line in gtk+.inc will make it better Mar 17 14:00:43 s/compied/copied/ Mar 17 14:00:45 but this is from cairo 1.10 Mar 17 14:00:46 JaMa: pardon? Mar 17 14:00:52 1.11.2 is out Mar 17 14:01:21 if you put the line in gtk+.inc then all the recipes which do set BBCLASSEXTEND will see it and your problem will go away. Mar 17 14:01:39 since the set of recipes which set BBCLASSEXTEND is a subset of those which inherit gtk+.inc Mar 17 14:01:59 yes I know it will be enough and I'll be saving 3 lines Mar 17 14:02:18 right. and you will avoid duplicating that override for every future version of gtk. Mar 17 14:02:22 but if someone removes gtk+ recipes nobody is using then we could save hudrets of lines Mar 17 14:02:35 well, sure, but that is a separate issue. Mar 17 14:02:41 damn it Mar 17 14:02:46 its still in 1.11.2 Mar 17 14:02:50 * woglinde shakes fists Mar 17 14:03:11 there's no reason to make things worse than they are. Mar 17 14:03:31 if you put the new override in gtk+.inc then at least you aren't adding any further duplication. Mar 17 14:03:49 pb__: from my POV it will be worse.. when you see PACKAGES_DYNAMIC in recipe and then had to check .inc file that there is also PACKAGES_DYNAMIC_virtclass-native Mar 17 14:03:54 if you wanted to go further and remove some of the duplication that already exists then of course that would be fine too, but there is no reason the two things need to be linked. Mar 17 14:03:58 hm lets see if I find a page where I can see which atomic-ops are in gcc now Mar 17 14:04:25 JaMa: surely that's exactly the same issue that you always have with any variable which might be defined in an .inc file Mar 17 14:04:30 pb__: that's what I meant that at least those 3 lines should be moved together to .inc file Mar 17 14:05:16 pb__: yead but PACKAGES_DYNAMIC/PACKAGES_DYNAMIC_virtclass-native are more close together then DESCRIPTION not in recipe but in .inc.. Mar 17 14:05:30 anyways I don't care about that problem anymore.. feel free to NACK it :) Mar 17 14:05:40 *g* Mar 17 14:06:02 ah here we go Mar 17 14:06:03 http://gcc.gnu.org/wiki/Atomic Mar 17 14:06:17 JaMa: pretty sure we just need PACKAGES_DYNAMIC = "" and PACKAGES_DYNAMIC_virtclass-native = "" in native.bbclass. everyone i've talked to so far as agreed that it's a sane approach, including RP. Mar 17 14:06:43 kergoth_: even better Mar 17 14:07:01 it's also consistent, as thats what we do to PACKAGES Mar 17 14:07:05 do we support sh? Mar 17 14:07:37 kergoth_: heh, true, though I have been trying to stop us doing that to PACKAGES. I think it would make more sense to track PACKAGES per HOST_ARCH or something for that purpose. Mar 17 14:07:56 yeah, agreed, it would be nice Mar 17 14:08:28 if we get to that point, though, we really do need to make it possible to specify preferences for runtime packages explicitly Mar 17 14:08:37 yeah Mar 17 14:11:10 the reason I would like to stop squashing PACKAGES for native is that, if you let the native stuff generate output packages for itself, you can (in theory) eliminate the dedicated staging packages and just use the regular output ones for sysroot construction. Mar 17 14:11:38 which, though not to everyone's taste I suspect, would be nice in a bunch of ways. Mar 17 14:12:28 yeah. i think oe-lite does that, actually, along with elminating build time dependencies entirely Mar 17 14:12:36 yeah Mar 17 14:12:41 definitely cleaner Mar 17 14:18:27 till later Mar 17 14:28:41 I'm struggling with using 2.6.38 kernel headers from a userspace app's recipe Mar 17 14:29:19 kernel.bbclass in sysroot_stage_all_append() copies a bunch of kernel headers into ${SYSROOT_DESTDIR}${STAGING_KERNEL_DIR} (rather than using headers_install) Mar 17 14:30:06 gstreamer-ti.inc passes this dir to its Makefiles, and do_compile gets an error for trying to use kernel headers from userspace Mar 17 14:32:04 03Andrea Adami  07org.openembedded.dev * rde2721fde6 10openembedded.git/recipes/linux/linux-kexecboot-2.6.38/collie/defconfig: Mar 17 14:32:04 linux-kexecboot: update 2.6.38 defconfig for collie (fonts 4x6) Mar 17 14:32:04 Signed-off-by: Andrea Adami Mar 17 14:32:47 In Oct, Koen said on the mailing list that kernel.bbclass does this because some recipes use private kernel api Mar 17 14:34:06 but then regular userspace apps can't include the correct kernel headers. There must be a correct way to do this? Mar 17 14:37:17 how do I tell glibc-2.9.bb to not install crap to / ? Mar 17 15:13:25 woglinde: FC14 : package icedtea6-native-1.7.10-r3.0: task do_compile: Failed - /usr/bin/ld: cannot find -ljvm Mar 17 15:23:12 Why does helloworld have an infinite loop?!? At least it builds and runs on my target :) Mar 17 15:41:44 woglinde: it seems that the freescale codewarrior works ok under wine Mar 17 15:41:50 so I guess that is my best option for now Mar 17 17:15:49 ericben hm Mar 17 17:17:41 woglinde: so in the end ubuntu : fails on libecj-bootstrap, fc14 : fails on icedtea6-native Mar 17 17:18:05 woglinde: ubuntu in x86, fc14 is x86_64 Mar 17 17:19:12 denix: florian: Crofton: meeting in 45 minutes from now, right? Mar 17 17:19:52 ericben my ubuntu at work dont fails Mar 17 17:20:06 mickeyl: yes :-) Mar 17 17:20:06 ericben can you mail me the log? Mar 17 17:20:09 mickeyl, denix and I thought an hour off Mar 17 17:20:29 if we can start early, it would be helpful, I have a telecon 1430 local Mar 17 17:20:58 I guess its because they changed to DST already... Mar 17 17:20:59 shall we say in 15 minutes? Mar 17 17:21:29 ok for me Mar 17 17:21:46 ok Mar 17 17:21:46 morning florian mickeyl Crofton|work woglinde Mar 17 17:21:51 hi ka6sox-home Mar 17 17:21:55 hi ka6sox-home Mar 17 17:22:00 morning pb__ Mar 17 17:22:08 hi ka6sox-home Mar 17 17:22:14 * florian is mack to the office after more than a week off... still catching up Mar 17 17:22:27 pb__, you need anything from me to help you? Mar 17 17:22:43 ka6sox-home: don't think so. I need to get with cbrake to talk about git. Mar 17 17:23:40 pb__, okay good...let me know... Mar 17 17:23:55 how goes the dodgy floor joists? Mar 17 17:24:15 joists all done now, I spliced the last bit of timber for that last night Mar 17 17:24:32 heating pipework is next Mar 17 17:24:54 also been experimenting with fermacell dry lining. Mar 17 17:25:08 heating pipework Mar 17 17:25:12 need to look up this Mar 17 17:25:32 woglinde: just pipes that circulate warm water in the floor Mar 17 17:25:44 pb__, I'd like to try the fermacell stuff here. Mar 17 17:26:10 pb its called "fußbodenheizung" here Mar 17 17:26:54 ka6sox-home: yeah, I'm quite pleased with it so far. my wife not so pleased, since she had to unload a delivery of boards this morning. Mar 17 17:27:08 for some reason the supplier sent them out on a truck without a hiab. Mar 17 17:27:30 woglinde: aha, right. Mar 17 17:27:45 pb__, my wife would have just called me and said "get here to deal with this" Mar 17 17:28:08 ka6sox ;) Mar 17 17:28:16 yeah, that would have been the smart plan Mar 17 17:28:22 ka6sox my wife would build it one day alone Mar 17 17:30:48 I guess that's the main downside of fermacell for me, the boards are a pain to move around because they're so heavy. Mar 17 17:32:55 hi there Mar 17 17:33:12 im emulating an akita in qemu and i dont know how to network it with my host pc Mar 17 17:33:20 does anyone know how to do that? Mar 17 17:35:28 dromede read the manpage Mar 17 17:35:33 of qemu Mar 17 17:35:45 lol Mar 17 17:35:48 already did Mar 17 17:35:56 mom Mar 17 17:36:14 i dont think i quite understand all the methods of connecting vm's to host pc Mar 17 17:36:38 -net nic[,vlan=n][,macaddr=mac][,model=type] [,name=name][,addr=addr][,vectors=v] Mar 17 17:36:48 ok Mar 17 17:36:54 but which one? Mar 17 17:37:00 al the emulated ones are pci bases Mar 17 17:37:04 *based Mar 17 17:37:13 pb__, yup...she would just call me. Mar 17 17:38:11 dromede I give you a pointer Mar 17 17:38:16 no google the rest Mar 17 17:38:22 or you use the cdc_usb stuff Mar 17 17:38:56 does cdc_usb work on 2.6.36+ kernels? Mar 17 17:39:12 sure Mar 17 17:39:25 I have working the g_composite here for beagle board Mar 17 17:45:14 till later Mar 17 18:09:37 03Tom Rini  07master * ra159104809 10openembedded.git/recipes/util-linux-ng/util-linux-ng.inc: Mar 17 18:09:37 util-linux-ng.inc: Add pkgconfig-native to native DEPENDS Mar 17 18:09:37 Signed-off-by: Tom Rini Mar 17 19:18:34 hi all Mar 17 19:19:23 I get this error when i execute my bitbake file " ERROR: The SRCREV_FORMAT variable must be set when multiple SCMs are used" Mar 17 19:20:55 this is how my recipe looks like " SRCREV = "${AUTOREV}" PV = "1.0+svnr${SRCREV}" PR = "r1" SRCREV_axonclnt = "2474" SRC_URI = " svn://taser.unfuddle.com/svn/taser_axon/axonapp/branch/;module=axonctrl_fiji;proto=https \ svn://taser.unfuddle.com/svn/taser_axon/axonapp/trunk/axon_units/;module=AxonClient;proto=https;rev=2474;name=axonclnt \ " " Mar 17 19:21:52 as it says, you're using mulitple scms in a single SRC_URI, so need to set SRCREV_FORMAT. i don't know what the syntax of that variable is offhand, though Mar 17 19:22:19 ok Mar 17 19:22:20 re Mar 17 19:22:34 ~pastebin Mar 17 19:22:35 [~pastebin] A "pastebin" is a web-based service where you should paste anything over 3 lines so you don't flood the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://bin.cakephp.org/ , http://asterisk.pastey.net/ , or install pastebinit with yum or aptitude. Mar 17 19:23:43 kergoth : ibot: i ll try to find that , thanks for informing about paste bin i will use pastebin here after Mar 17 19:25:42 brave ibot Mar 17 19:51:20 hi woglinde, which ubuntu versiion are you using ? Mar 17 19:51:34 10.04 Mar 17 19:51:38 at work Mar 17 19:51:40 amd64 Mar 17 19:52:04 ericben you didnt send me the logfile Mar 17 19:52:45 woglinde: I'm just back on the computer, was with my sons until now :) Mar 17 19:54:22 ;) Mar 17 19:55:34 hi, all Mar 17 19:55:43 have one quick ? Mar 17 19:55:56 what? Mar 17 19:56:10 in the x11-gpe-image there is "setting"->"screen setup" Mar 17 19:56:45 I can choose "Portrait" or "Landscape", however it will be preserved for current session only Mar 17 19:57:00 when I reboot it boots up with default one Mar 17 19:57:10 which is portrait Mar 17 19:57:27 how can I preserve "Landscape" settings? Mar 17 19:58:06 sorry dont know this much about gpe Mar 17 19:58:19 woglinde: mail sent Mar 17 19:58:26 okay Mar 17 19:58:28 woglinde: do you also want the FC14 log ? Mar 17 19:58:32 yes Mar 17 19:58:36 please Mar 17 20:00:54 woglinde: it's x11-gpe-image, not just gpe ;-) Mar 17 20:00:57 woglinde: for fc14, I've found the problem : /usr/bin/ld: cannot find -lstdc++ Mar 17 20:01:05 ericben *g* Mar 17 20:01:25 do we have no c++ apps for native Mar 17 20:02:00 woglinde: seems that's the case Mar 17 20:02:27 http://openembedded.net/index.php/OEandYourDistro libstdc++ is only listed for OpenSuSe Mar 17 20:02:48 which makes no sense Mar 17 20:03:02 could you edit it Mar 17 20:03:06 when you are at it Mar 17 20:05:32 anyone know how tom make import sipconfig work in python? Mar 17 20:05:38 I install python-sip Mar 17 20:05:41 and -dev Mar 17 20:05:47 crofton you are really trying it? Mar 17 20:06:04 yes Mar 17 20:06:09 crofton I can send you my stuff if you want Mar 17 20:06:10 I have my orders Mar 17 20:06:14 please Mar 17 20:06:19 poor bosses Mar 17 20:06:47 the python-sip recipe seems to not install sipconfig Mar 17 20:08:51 ericbe which email did you use? Mar 17 20:09:06 woglinde: in fact icedetea tries to link libstdc++ static ... so under FC installing libstdc++-static seems to fix the problem Mar 17 20:09:23 woglinde: the one you sent in private Mar 17 20:09:52 h, Mar 17 20:10:03 woglinde: so libstdc++ static seems needed to build icedtea6-native Mar 17 20:10:37 anybody knows the answr 2 my ? ? Mar 17 20:10:44 thank u Mar 17 20:11:12 crofton which email? Mar 17 20:11:22 igork1 florian maybee Mar 17 20:11:23 IgorK1: sorry I'm not using gpe. You can check the script which launch X this may be here that there is a defautl setting Mar 17 20:14:32 hm still no email here Mar 17 20:14:47 ericben how big was the log? Mar 17 20:14:51 maybee it was to larfe Mar 17 20:14:53 large Mar 17 20:15:05 florian: everybody points to u ;-) any clues? Mar 17 20:15:37 woglinde: sent to your other email Mar 17 20:15:51 woglinde: 97KB large Mar 17 20:16:18 IgorK1: is gpe-confd running? Mar 17 20:16:42 ericben hm thats the ecj Mar 17 20:17:18 woglinde: yes that's the blocking part on my ubuntu setup Mar 17 20:18:01 I'm going to launch it into an other ubuntu to see if I get the same problem Mar 17 20:24:21 woglinde: ok icedtea6-native 's build finished on the FC14 host Mar 17 20:24:30 ;) Mar 17 20:24:57 do you think libstdc++-static should be added to required packages ? Mar 17 20:25:21 yes Mar 17 20:25:45 ok once I find my wiki password I'll u pdate it Mar 17 20:26:09 libstdc++-dev for debian/ubuntu Mar 17 20:28:21 woglinde: yes also found this one missing Mar 17 20:28:41 florian: hold on let me check.... Mar 17 20:29:09 * florian will go home really soon Mar 17 20:34:00 woglinde: libecj-bootstrap builds fine on theh other ubuntu (a 10.04.2 LTS desktop x86) Mar 17 20:34:12 lol Mar 17 20:34:13 quite strange Mar 17 20:34:18 as I said didnt saw that before Mar 17 20:34:21 jo kergoth Mar 17 20:35:05 woglinde: the difference is that one is headless server, the other a desktop edition in a vm Mar 17 20:35:37 ericben sorry no clue Mar 17 20:36:15 woglinde: I'm looking at it, thanks Mar 17 20:37:32 florian: yes it is running, anything else 2 check? Mar 17 20:43:54 03Martin Jansa  07master * ra1f41f4d00 10openembedded.git/recipes/freesmartphone/ (5 files): Mar 17 20:43:55 freesmartphone: bump cornucopia, msmcomm, libfso-glib, fso-specs, libgisi SRCREV Mar 17 20:43:55 Signed-off-by: Martin Jansa Mar 17 20:43:55 03Martin Jansa  07master * rbfd451f014 10openembedded.git/conf/distro/shr.conf: Mar 17 20:43:55 SHR: blacklist some packages causing multiple runtime providers Mar 17 20:43:55 Signed-off-by: Martin Jansa Mar 17 20:44:25 03Martin Jansa  07master * ra8804b6cae 10openembedded.git/recipes/gstreamer/ (17 files in 5 dirs): Mar 17 20:44:25 gstreamer: move old versions to obsolete Mar 17 20:44:25 Acked-by: Koen Kooi Mar 17 20:44:25 Signed-off-by: Martin Jansa Mar 17 21:44:12 florian: welcome back Mar 17 21:44:32 re Mar 17 21:44:52 the gpe-confd is running, anything else 2 check? Mar 17 21:48:28 hm was there a cvs SRC_URI for only checkout a subdir? Mar 17 21:59:46 I have a problem with building perl-native, it complains about missing references to libm Mar 17 21:59:57 hm uh? Mar 17 22:00:24 florian: anything else 2 check? Mar 17 22:00:39 see http://www.pastie.org/1684297 Mar 17 22:01:20 IgorK1: nothing really obvious... maybe just change the X server script Mar 17 22:01:39 woglinde: any idea? Mar 17 22:04:29 morphis you are missing pthread stuff too Mar 17 22:04:42 looks like you dont have libc-dev installed Mar 17 22:04:42 hm Mar 17 22:05:29 libc6-dev is installed Mar 17 22:09:18 hm it's using -L/usr/local/lib rather than -L/usr/lib Mar 17 22:09:31 hm Mar 17 22:09:42 cc -L/home/morphis/oe/machines/palmpre/tmp/sysroots/x86_64-linux/usr/lib -Wl,-rpath-link,/home/morphis/oe/machines/palmpre/tmp/sysroots/x86_64-linux/usr/lib -Wl,-rpath,/home/morphis/oe/machines/palmpre/tmp/sysroots/x86_64-linux/usr/lib -Wl,-O1 -fstack-protector -L/usr/local/lib -o miniperl Mar 17 22:09:57 it's a clean build Mar 17 22:10:28 sorry Mar 17 22:10:35 dont know at the moment Mar 17 22:10:46 need to concentrate on libatomic-ops Mar 17 22:10:52 ok Mar 17 22:14:38 is there some way to avoid using perl-native and use system perl? Mar 17 22:15:58 morphis: you can't cross-compile perl without a native one of the *exact same version*, and to ensure we're consistent during the build (e.g. one app doesn't build with it and one without), we use it for everything. Mar 17 22:16:01 afaik, anyway Mar 17 22:19:33 kergoth: I have a installed system perl Mar 17 22:19:38 but building perl-native fails Mar 17 22:20:35 system-perl is 5.10.1-17ubuntu3 and oe-perl is 5.10.1 too Mar 17 22:22:01 kergoth: look at http://www.pastie.org/1684297 for the build failure Mar 17 22:32:45 denix I know exactly what your and khem's problem is Mar 17 22:33:09 and I have no idea how this can possible work for others.. :P I'm working on a fix Mar 17 22:33:41 :) Mar 17 22:34:16 (if you use the poky bitbake your problem will go away.. the issue is what I suspected, pseudo isn't being run during the rootfs install, and chroot fails and causes RPM to shutdown.. why there is no error message, I don't know.. but thats what is happening..) Mar 17 22:34:46 (PSEUDO isn't starting because the bitbake in OE is missing a patch that setups the "fakeroot" environment variables -before- the fork in "fork_off_task".. Mar 17 22:35:21 pseudo is activated (since it's already in memory) by either a fork() or exec*()... no exec happens within the rootfs construction.. so pseudo isn't running at all points) Mar 17 22:39:19 any plans on pushing that change to bitbake? or is it after-1.0 activity? Mar 17 22:39:41 that should definitely get into master sooner rather than later, i'd think Mar 17 22:39:53 I'm working on the patch right now.. Mar 17 22:40:00 ideally, along with anything else *other than* the server bits Mar 17 22:40:02 I'm digging through 4 months of commits and trying to reconcile with bitbake Mar 17 22:40:07 since clearly thats a big can of worms Mar 17 22:40:44 fray: heh, that's what i did when doing the sync work i did, went through poky's changes to bitbake since the last bitbake sync one by one, along with some direct file diffing Mar 17 22:40:51 * kergoth shakes head and sighs Mar 17 22:41:13 its one thing git really sucks at, management of cherry picks between long lived branches Mar 17 22:41:33 hm? Mar 17 22:41:59 I could have cherry-picked this stuff.. but I think it would have been more confusing due to the change in designs along the way Mar 17 22:42:12 ah long lived branches Mar 17 22:42:18 take a commit in branch one, cherry pick it to branch two, hit a conflict, resolve the conflict and commit, and git now has zero knowledge that the two commits are the same / originated from the other branch Mar 17 22:42:19 * khem remembers of uclibc-nptl branch Mar 17 22:42:26 so in the future you can't eliminate that from the ones you're going through Mar 17 22:42:31 he khem Mar 17 22:42:36 woglinde: hi Mar 17 22:42:48 khem another task disable alsa support in qemu-native Mar 17 22:43:45 woglinde I have a patch to the oe-core qemu that disables the GL and alsa (actually all audio) Mar 17 22:43:52 I need it for my server install Mar 17 22:44:10 woglinde: yes I have to do that Mar 17 22:44:48 hm Mar 17 22:44:54 but wait Mar 17 22:45:01 why not add alsa to deps Mar 17 22:45:04 than some features cannt be test Mar 17 22:45:14 do we have alsa-native? Mar 17 22:45:43 no we dont Mar 17 22:46:19 but shouldnt be that hard to include support Mar 17 22:46:23 we don't really want alsa-native.. thats really leaking into to odd places Mar 17 22:46:55 fray but when you will test your qemu-image with sound? Mar 17 22:47:03 never Mar 17 22:47:07 :) Mar 17 22:47:08 you Mar 17 22:47:14 but there might be others Mar 17 22:47:17 testing alsa Mar 17 22:47:21 inside qemu Mar 17 22:47:27 yes, thats why it's not the default for qemu.. Mar 17 22:47:27 its useless using sword to kill a fly :) Mar 17 22:47:41 the default needs to be more useful then the patch I have.. but I need to build on systems that don't have video or sound.. Mar 17 22:47:49 (and there is no reason they ever will or should) Mar 17 22:48:07 fray: you mean building qemu-native itself ? Mar 17 22:48:18 fray: It could be one if use flags Mar 17 22:48:21 yes.. on these systems I might run qemu, but only to verify an image boots.. Mar 17 22:48:27 and yes, it could (and likely should be).. Mar 17 22:48:28 I understand Mar 17 22:50:11 woglinde: can you try this patch please http://paste.ubuntu.com/581835/ Mar 17 22:50:25 hm not this evening anymore Mar 17 22:50:40 woglinde: no you can not leave ... Mar 17 22:50:50 khem I can Mar 17 22:50:54 today is St. Patricks day Mar 17 22:51:38 ah hm Mar 17 22:51:53 * fray is planning to go to mother-in-laws and get cornbeef and cabage.. (skip the cabbage) Mar 17 22:52:08 I have to make a special trip to my daughters school to hand her the green pull over Mar 17 22:52:25 woot! I have a fix Mar 17 22:52:34 kergoth whats the proper place to send the fix to (email wise)? Mar 17 22:52:38 She is totally american Mar 17 22:52:54 fray: bitbake-dev@lists.berlios.de Mar 17 22:53:13 or is it bitbake-devel I confuse Mar 17 22:53:19 its bitbake-dev i think Mar 17 22:53:25 we need to move it off of there Mar 17 22:53:58 oh no Mar 17 22:54:03 not another gcc fix Mar 17 22:55:53 woglinde: unfortunately there will be yet another one Mar 17 22:56:00 which I am currently investigating Mar 17 22:56:44 gmail is wonky today Mar 17 22:57:10 Oops.. a server error occurred Mar 17 22:57:33 I bet my wife has been on call by now :) Mar 17 22:59:30 khem: have you seem my reports about same ICE in gcc on libgcrypt and fail of qemu do_configure on alsa libs? Mar 17 22:59:47 Jay7: yes Mar 17 22:59:51 Jay7: I am on it Mar 17 22:59:56 florian: what kind of changes u r talking about? I'm running standard image, where I didn't change anything.. Mar 17 23:00:02 I know the culprit Mar 17 23:00:15 ah, I see now :) Mar 17 23:00:34 IgorK1: take a look at /etc/default/xserver Mar 17 23:05:41 03Andrea Adami  07org.openembedded.dev * r5f1473a737 10openembedded.git/recipes/base-files/ (5 files in 5 dirs): Mar 17 23:05:41 base-files: remove obsolete usbdmodule entries for Zaurus. Mar 17 23:05:41 * bump PR Mar 17 23:05:41 * grep reveals $usbdmodule is only used by Zaurus and openmn (?). Mar 17 23:05:41 * only use is in linux-hotplug/files/usbd.agent, for 2.4 kernels... Mar 17 23:05:42 Signed-off-by: Andrea Adami Mar 17 23:16:37 good nite Mar 17 23:23:49 03Khem Raj  07master * ra7b6843c11 10openembedded.git/recipes/gnutls/gnutls_2.10.4.bb: Mar 17 23:23:50 gnutls_2.10.4.bb: Revert the -Os -g workaround. Mar 17 23:23:50 We have taken care of this in GCC Mar 17 23:23:50 Signed-off-by: Khem Raj Mar 17 23:24:00 03Khem Raj  07master * rd765a71812 10openembedded.git/recipes/gcc/ (gcc-4.5.inc gcc-4.5/more-epilogues.patch): Mar 17 23:24:00 gcc-4.5: Fix the ARM ICE regression at -Os -g Mar 17 23:24:00 This holds off the patches that cause the ICE at -Os -g Mar 17 23:24:00 for now. The other workaround it to use -fno-shrink-wrap Mar 17 23:24:00 but it has to go into all affected recipes Mar 17 23:24:01 Signed-off-by: Khem Raj Mar 17 23:24:02 03Khem Raj  07master * r91d4e598bf 10openembedded.git/recipes/qemu/qemu_0.14.0.bb: Mar 17 23:24:02 qemu_0.14.0.bb: Do not enable sound emulation. Mar 17 23:24:03 Drop gnutls from DEPENDS we disable tls Mar 17 23:24:03 Signed-off-by: Khem Raj Mar 17 23:24:39 Jay7: I have pushed the changes to gcc which should hopefully fix the ICE you are seeing could you do me a favor by doing bitbake whatever you were doing after upgrading to latest tip on master Mar 17 23:25:30 khem: hi Mar 17 23:25:47 would you add MACHINE_FEATURES += " iwmmxt" to pxa270 or only to specific apps? Mar 17 23:26:14 ant__: which example Mar 17 23:26:20 khem: I'll run testbuilder tonight again on master Mar 17 23:26:30 iirc only mplayer was using it Mar 17 23:26:44 can't run right now - doing presentation Mar 17 23:26:50 I have two zaurus with pxa270 Mar 17 23:26:57 well.. composing presentation :) Mar 17 23:27:03 ant__: I think you could add it to MACHINE_FEATURES yes Mar 17 23:27:06 ./zaurus-2.6.inc:MACHINE_FEATURES = "kernel26 apm alsa pcmcia irda usbgadget keyboard touchscreen screen vfat ext2" Mar 17 23:27:21 for the 2 devices I'll add iwmmxt, isn't? Mar 17 23:29:46 florian: there is no such file "/etc/default/xserver" Mar 17 23:42:08 kergoth ok.. email sent Mar 17 23:43:30 drat.. I think I sent it to the wrong add.. looks like it should have been bitbake-dev Mar 17 23:44:39 ericben: i see. and which phone stack do you want to run with the modems? Mar 17 23:57:57 florian, u still here? Mar 17 23:58:06 there is no such file... **** ENDING LOGGING AT Fri Mar 18 02:59:57 2011