**** BEGIN LOGGING AT Thu Jan 27 02:59:57 2011 Jan 27 03:04:52 03Mike Westerhof  07org.openembedded.dev * r8fe839a9fa 10openembedded.git/ (conf/distro/include/slugos.inc recipes/tasks/task-slugos.bb): Jan 27 03:04:52 SlugOS: slugos.inc, task-slugos.bb - Debugging improvements Jan 27 03:04:52 Adds devshell, rootfs directory is qualified by machine and libc, Jan 27 03:04:52 and add kexec to the base image. Jan 27 03:04:52 Signed-off-by: Mike Westerhof Jan 27 06:14:08 does anyone help me how to configure network in QEMU Jan 27 06:21:08 try #qemu Jan 27 06:45:46 gm Jan 27 06:49:50 hiii ..........can anyone help me how to configure network in QEMU Jan 27 06:52:42 morning eFfeM_work Jan 27 06:55:07 sanket: try #qemu Jan 27 06:56:17 eFfeM_work, qemu-system-arm -M versatilepb -kernel vmlinuz-2.6.26-2-versatile -initrd initrd.img-2.6.26-2-versatile -hda debian_lenny_arm_small.qcow2 -append "root=/dev/sda1" Jan 27 06:56:32 eFfeM_work, I m using this command Jan 27 06:57:08 sanket: what makes you think I have any qemu experience at all Jan 27 06:57:15 sanket: and this is definitely not a qemu channel Jan 27 06:57:26 your question is totally inappropriate fro this room Jan 27 06:58:06 you're not even tyring to boot anoe based image on qemu Jan 27 07:36:24 eFfeM_work, hey I had configured my qemu network. And first of all I had not personally asked u abt qemu. Since u had suggested me abt qemu, I think u r the right person. But I think you dont know anything. I had already tried many oe images on my development board. Jan 27 07:55:04 good morning Jan 27 09:15:24 hi, does anybody have a PXA270 based board available on the desk? Jan 27 09:15:40 I am facing to a weird behavior with serial ports (and udev) Jan 27 09:16:23 I have 3 ports ttyS[012] but I see only ttyS0 in /dev Jan 27 09:41:32 mckoan: I have pxa270-based device, unfortunately no uart on it, but I've one question to you: what kernel do you use and is cpufreq works for you? :) Jan 27 10:04:50 anarsoul: Linux spitz 2.6.34-kaeilos #13 Tue Jan 18 17:18:17 CET 2011 armv5tel Jan 27 10:05:19 oh, ok Jan 27 10:05:42 cpufreq is working for me on 2.6.36, but does not on 2.6.37 and later Jan 27 10:06:37 anarsoul: I don't use cpufreq Jan 27 10:15:12 mckoan: why? Jan 27 10:19:28 anarsoul: because I don't need it :-) Jan 27 10:42:08 hi Jan 27 10:42:35 we're discussing the NOR / NAND debate Jan 27 10:43:08 NOR seems much more reliable, but it is smaller & more expensive Jan 27 10:43:30 NAND is bigger, cheaper, but not reliable anymore. Jan 27 10:43:37 any thoughts on this? Jan 27 10:43:52 tdebrouw: wrong channel? Jan 27 10:44:32 mckoan: openembedded => embedded => small footprint => flash? Jan 27 10:45:15 mckoan: i mean, people who are using openembedded probably use flash? Jan 27 10:48:05 if it's not possible/not allowed to talk about this topic in this channel, then I'm sorry to bring it up Jan 27 10:48:08 :-/ Jan 27 10:52:42 tdebrouw: it is not really the right topic for this channel, but in brief: your analysis on nor/nand diffs is ok, note that the unreliability in nand only shows if you are updating it, so if you do not write to the nand and it works then it will remain working Jan 27 10:52:56 and/or use a bad block mapping fs (like ubi) Jan 27 10:54:01 eFfeM_work: is there a better channel to discuss this? Jan 27 10:54:41 tdebrouw emmc is the cheapest Jan 27 10:55:24 tdebrouw: elinux and edev might be better places. You have NAND SLC and NAND MLC -> makes a BIG difference with Linux Jan 27 10:55:50 thx all Jan 27 10:55:53 one thing i've always been unsure of - is NOR faster than NAND? Jan 27 10:56:05 is there a #mtd for these thigns? Jan 27 10:56:18 dont know Jan 27 10:56:57 robtaylor: I guess it depends on the speed grade of both Jan 27 10:57:06 The IRC channel is #mtd on irc.ipv6.oftc.net (which includes Legacy IP addresses despite its name) Jan 27 10:58:07 robtaylor did you read booting qt in 1 second Jan 27 10:58:10 robtaylor: nor is byte accessible to typically for small reads/writes it will be faster, for larger chunks maybe nand read is faster but writes probalby will remain slower (due to the erase being needed), but as tdebrouw says it depends on the speed too Jan 27 10:58:20 woglinde: oooh, that sounds interesting, where? Jan 27 10:58:52 woglinde: yeah that demo was very impressive, I joined the presentation Jan 27 10:58:54 eFfeM_work: tdebrouw: ah, that makes sense, thanks! Jan 27 10:58:57 http://www.embedded-bits.co.uk/2011/1-second-linux-boot-to-qt/ Jan 27 10:59:31 woglinde: thanks for the link Jan 27 10:59:37 woglinde: thanks Jan 27 10:59:41 * robtaylor reads Jan 27 10:59:46 woglinde: wrt my lirc Q yesterday evening: I get Jan 27 10:59:47 ERROR: QA Issue with lirc: non -dev package contains symlink .so: lirc path '/work/calamari-oe-linux-gnuspe/lirc-0.8.7-r0/packages-split/lirc/usr/lib/liblirc_client.so' Jan 27 10:59:58 effem yes Jan 27 11:00:13 yeah, the function mapping to specific blocks was awesome. :) Jan 27 11:00:23 recipe says: Jan 27 11:00:24 FILES_${PN}-dev += "${libdir}/liblirc_client.so" Jan 27 11:00:24 FILES_${PN} = "${bindir} ${sbindir} ${libdir} ${sysconfdir} ${exec_prefix}/var" Jan 27 11:00:30 effem yes I looked it up Jan 27 11:00:33 thought the first lien would pick up the so Jan 27 11:00:39 ah ok Jan 27 11:00:42 so, are there any OE guys foling/involved with linaro? Jan 27 11:00:46 but not deeper Jan 27 11:01:04 hrw Jan 27 11:01:55 anyone seen this error (or similar): tmp/work/ppce500v2-angstrom-linux-gnuspe/pixman-0.21.4-r5.0/temp/run.staging_packager.15645: line 367: stage-manager-opkg-build: command not found Jan 27 11:01:57 | ERROR: Function staging_packager failed Jan 27 11:02:18 likewise latest dev? Jan 27 11:02:25 woglinde: astonishing Jan 27 11:02:28 kergoth pushed some stuff abouit Jan 27 11:02:35 woglinde: ah yes.. Jan 27 11:03:01 woglinde: yes, Wed Jan 26 13:53:24 2011 +0100 Jan 27 11:03:11 likewise: got that earlier this week, pull git head, changes were pushed yesterday evening Jan 27 11:03:36 eg this http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=0bf1d00fb4e2b36dc3c25194bdddc6d95300caf4 Jan 27 11:03:46 looks like i'll be helping them out with their verification stuff, it's good to know which old-school guys there are around.. Jan 27 11:04:50 eFfeM_work: thanks. Another commit mesg that does tell what has changed but not why... Jan 27 11:05:12 ask kergoth Jan 27 11:06:20 likewise: PM Jan 27 11:09:50 robtaylor: I am @linaro Jan 27 11:10:20 robtaylor: by 'verification stuff' you mean 'validation team'? Jan 27 11:10:31 robtaylor: there is ~100 people in linaro now Jan 27 11:13:56 hrw sounds like more than in ubuntu Jan 27 11:15:13 hrw: yup, validation team, been talking with Paul L. Jan 27 11:15:32 hrw: just getting started seeing where I can be of help though Jan 27 11:22:59 ieehks sf still broken Jan 27 11:24:03 robtaylor: I have no idea what they do there Jan 27 11:40:54 hrw: what are you working on? Jan 27 11:41:07 robtaylor: cross toolchains for ubuntu mostly Jan 27 11:41:53 ah, fun Jan 27 11:57:59 03Koen Kooi  07org.openembedded.dev * rf1243d659f 10openembedded.git/recipes/x-load/x-load-git/beagleboard/name.patch: Jan 27 11:57:59 x-load-git: add header to beagle name patch Jan 27 11:57:59 Signed-off-by: Koen Kooi Jan 27 12:11:58 03Stefan Schmidt  07master * r734e8852af 10openembedded.git/recipes/felix/ (9 files in 2 dirs): felix: Add recipes for the felix OSGi implementation Jan 27 12:12:00 03Stefan Schmidt  07master * rf1c6fc7bc5 10openembedded.git/conf/distro/include/angstrom-jalimo.conf: Jan 27 12:12:00 angstrom-jalimo.conf: Bump icedtea6-native version to 1.7.6 Jan 27 12:12:00 Signed-off-by: Stefan Schmidt Jan 27 12:12:00 Acked-by: Koen Kooi Jan 27 13:02:57 woglinde_: I wonder how much is still linux that "1 second boot" system Jan 27 13:08:04 there are some slides about a 1 second demo: http://www.slideshare.net/andrewmurraympc/elce-the Jan 27 13:08:21 (i'm not sure if someone posted them already) Jan 27 13:08:54 tdebrouw: yes, that's the topic I have been examining Jan 27 13:21:05 where is the kernel load address (for ARM) defined in OE? machine or linux recipes? I have a hard time finding this (but will backtrace now) Jan 27 13:40:47 likewise: kernel load address is usually defined into the kernel Jan 27 13:50:45 Need some hints on do_install error of unifdef-native-2.4 ... http://pastebin.com/xbE56EGu Jan 27 13:51:15 It seems that shr-unstable/openembedded has a new install script that is miss functioning Jan 27 13:52:27 what to do ? Jan 27 14:00:01 whois shaz Jan 27 14:00:03 ups Jan 27 14:01:01 shaz: http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=af5f4debf19be227c63d5cc89e4acd2961469724 Jan 27 14:01:12 JaMa|wrk if u remember I was working on the security related work on openmoko specifically shr Jan 27 14:02:10 I need to do some modifications in sysvinit and getting this error ... Jan 27 14:02:49 looks like OE bin/install needs different shell than what you have now, like bash Jan 27 14:04:43 hmm ... not sure what to do but trying to try out the branch you mentioned Jan 27 14:05:15 I just pointed to commit which changed it Jan 27 14:05:36 you can ask kergoth why he used such install script when he joins Jan 27 14:05:48 Ok Jan 27 14:39:48 Hello, I recently built an Arago Linux image for my (very old) iPaq h3000 (havent twested it yet). I am wondering, if I know my Bitbake machine conf name, is there a Bitbake command that will show mw the names of all buildable images for it? Jan 27 14:40:20 typos galore, sorry. Jan 27 14:44:22 ls recipes/images Jan 27 14:58:54 under oe, i have openembedded/recipes/images, arago/recipes/images and oe_at91sam/recipes/images -- among these which are compatible with a new machine conf I need to build for with the cross-compiler on my PATH? do I have to try one by one and see if it builds? Jan 27 15:02:58 bertszoghy, ? Jan 27 15:03:10 I don't understand the sentence Jan 27 15:03:32 ahhh Jan 27 15:03:51 a coma is missing Jan 27 15:04:20 could you refrase Jan 27 15:04:26 this part: Jan 27 15:04:37 among these which are compatible with a new machine conf I need to build for with the cross-compiler on my PATH? Jan 27 15:09:01 engineers keep changing the target board for the project Jan 27 15:09:09 here Jan 27 15:09:33 so what was your questions, how to add a new machine? Jan 27 15:09:52 i am wodering if there is a bitbake command shortcut Jan 27 15:09:59 ah for changing machine Jan 27 15:10:00 yes Jan 27 15:10:02 there is Jan 27 15:10:04 that will use the cross-compiler on my PATH Jan 27 15:10:10 and list Jan 27 15:10:18 MACHINE="foo-machine" bitbake console-image Jan 27 15:10:23 compatible bb image recipes Jan 27 15:10:25 with MACHINE not defined in local.conf Jan 27 15:10:46 hmmm Jan 27 15:11:01 could you explain more Jan 27 15:11:13 did i mention I am going to have a third cross-compiler soon? ;) Jan 27 15:11:20 you want a command that list machine<->recipe compatibility? Jan 27 15:11:35 yes, depending on Jan 27 15:11:41 the problem is the construction of your phrases Jan 27 15:11:49 the cross-compiler on the path Jan 27 15:11:58 I've a hard time understanding what they means because of that Jan 27 15:12:02 ah ok Jan 27 15:12:30 so you want the list of compiler<->machine<->recipes Jan 27 15:12:35 compatibility Jan 27 15:12:41 SELECT recipes FROM bitbake WHERE cross-compiler='A' AND machine='foo' Jan 27 15:12:56 ;) Jan 27 15:13:04 ok that's more clear Jan 27 15:13:09 or, better Jan 27 15:13:23 SELECT recipe_images FROM bitbake WHERE cross-compiler='A' AND machine='foo' Jan 27 15:13:25 that doesn't exist in bitbake Jan 27 15:13:28 but.... Jan 27 15:13:39 people do branches that are supposed to build Jan 27 15:13:45 so they have that info Jan 27 15:13:56 it's put in the wiki Jan 27 15:14:08 let me find it Jan 27 15:14:30 http://wiki.openembedded.org/index.php/Testing Jan 27 15:14:33 I understand that most people work for one target machine with one cross-compiler, such is not my lot Jan 27 15:14:48 with that you'll have a lot of combinaison Jan 27 15:15:07 many target machines, many cross-compilers -- our engineers are artists. Jan 27 15:15:09 ;) Jan 27 15:15:18 cross-compiler is usally integrated in openembedded Jan 27 15:15:25 so they are not listed Jan 27 15:15:29 but the rest is Jan 27 15:15:53 machine + image + distro + workstation + bitbake branch + oe branch + build type Jan 27 15:15:57 bertszoghy: OE normally builds an appropriate toolchain automatically. there are include files that make it somewhat easy to point to an external one, but the problem is the layout of toolchains can vary slightly, andi ts not enough to just use it to build, we have to extract things like the libc from it to package for the rootfs Jan 27 15:16:12 so there are multiple external toolchain recipes available, depending on the layout Jan 27 15:16:16 not sure if that helps Jan 27 15:16:45 yes that's a more verbose explanation than mine, and more clear :) Jan 27 15:24:32 i need to know what i'm doing Jan 27 15:25:59 that helps, yes :) Jan 27 15:28:59 bertszoghy: recipes don't have an associated crosscompiler. a recipe can state what machines/archs its compatible with, but not toolchains Jan 27 15:31:31 I like that answer kergoth, thanks. Jan 27 15:32:22 bertszoghy: the thing is, bitbake generates its runqueue (what tasks to execute in what order) up front, after parsing, but before things have been built. a recipe can only indicate compatibility (buildability) at parse time. and since oe builds its own compiler, the compiler version, etc isn't available at that time Jan 27 15:36:03 kergoth: thank you, I'll ask for more project time from the engineers. ;) Jan 27 15:36:08 heh :) Jan 27 15:52:51 03Chris Larson  07master * r76390d3224 10openembedded.git/classes/cml1.bbclass: Jan 27 15:52:51 cml1: fix the case of the variables in menuconfig Jan 27 15:52:51 Signed-off-by: Chris Larson Jan 27 16:01:36 03Chris Larson  07master * rf1ea2f35f4 10openembedded.git/classes/siteinfo.bbclass: Jan 27 16:01:36 siteinfo: add linux-gnu os Jan 27 16:01:36 Aside: should really shift most of the bits from this class into a python Jan 27 16:01:36 module in the oe package, which would make it possible to manipulate the data Jan 27 16:01:36 from the metadata if necessary. Jan 27 16:01:37 Signed-off-by: Chris Larson Jan 27 17:08:37 03Chris Larson  07master * rf16f443fd8 10openembedded.git/recipes/file/ (file-5.04/reloc.patch file_5.04.bb): Jan 27 17:08:37 file: make 5.04 relocatable Jan 27 17:08:37 This uses binreloc so that libmagic finds its magic file relative to its own Jan 27 17:08:37 location, and directly uses $ORIGIN for file to find libmagic. Jan 27 17:08:37 Signed-off-by: Chris Larson Jan 27 17:12:15 kergoth: shr-unstable/openembedded/bin/install is giving me some problem ... http://pastebin.com/xbE56EGu Jan 27 17:12:39 I am using /bin/sh Jan 27 17:12:51 still? i haven't heard any reports at all Jan 27 17:12:56 erm, misread Jan 27 17:12:58 * kergoth looks Jan 27 17:13:05 "using /bin/sh" tells me nothing Jan 27 17:13:11 what is your /bin/sh? Jan 27 17:13:16 dash? bash? busybox sh? ash? what? Jan 27 17:14:08 I don't suppose anyone has a script handy for mass PR bumping Jan 27 17:14:46 kergoth: /bin/bash Jan 27 17:15:38 mine too, works fine here, but i'll push a workaround Jan 27 17:16:21 03Chris Larson  07master * rfc837dd1e9 10openembedded.git/bin/install: Jan 27 17:16:22 bin/install: use '.' not source, and use -eq not == Jan 27 17:16:22 Signed-off-by: Chris Larson Jan 27 17:16:23 shahbaz: try now Jan 27 17:16:35 ok Jan 27 17:18:48 gm Jan 27 17:18:53 hey eFfeM Jan 27 17:20:30 Has anyone tried imote2 image? Jan 27 17:31:38 * Tartarus does some lazy for looping Jan 27 17:33:44 kergoth: the staging stuff worked like a charm for me, but it was already pushed before I could ack; thanks a lot for it Jan 27 17:35:33 but I do still have the problem with the bb that won't continue after printing the build config, recipe exist, try to do a clean Jan 27 17:35:50 weird Jan 27 17:36:03 let me know if you find a way to reproduce it reliably Jan 27 17:36:09 well i ^C'd the previous build Jan 27 17:36:14 can't dive in much until we've gotten to that Jan 27 17:36:51 i've touched local.conf, see if rebuilding the cache helps Jan 27 17:36:55 k Jan 27 17:37:07 i ^C builds pretty often, haven't seen that yet, must be more to it Jan 27 17:38:37 there is something odd with the OE wiki Jan 27 17:38:46 when i follow it and download the git bitbake Jan 27 17:39:00 it seems bitbake doesnt find the neccesary variables to work Jan 27 17:39:10 then you didn't yet the variables it needs to work Jan 27 17:40:01 s/yet/set/ Jan 27 17:42:38 kergoth: touch local.conf reparses cache but does not seem to change things Jan 27 17:48:19 kergoth: did a retest and reparse, but actually it does not even barf with the mistake I made: bitbake -cclean local.conf Jan 27 17:48:20 :-) Jan 27 17:48:35 (taht was wrt the mistake, not wrt the bug) Jan 27 17:48:48 kergoth: i did Jan 27 17:48:59 my printenv showed the BBPATH according to the wiki Jan 27 17:49:16 kergoth: if there is something you want me to test let me know Jan 27 17:49:20 pinting to both build:openembedded Jan 27 17:49:30 unsolo: have you updated local.conf ? Jan 27 17:49:40 set a machine etc Jan 27 17:52:28 yea Jan 27 17:52:40 i did this on two different systems Jan 27 17:52:58 but the guide sais to get the bitbake using wget to a version above the .18 Jan 27 18:03:32 Hi I am new to OE , I was trying to compile linux kernel 2.6.22 with distribution=arago and MACHINE=omap3evm with CSL external tool chain using the command " MACHINE=omap3axon bitbake virtual/kernel -D" after compiling i am seeing "uImage-2.6.28.10+xxxxxxxx.bin " getting generated in "deploy/glibc/images/omap3axon/ " where i was expecting "uImage-2.6.22.10+xxxxxxxx.bin" as I am usin g only 2.6.22 version of kernel can any onehlep Jan 27 18:05:57 unsolo: if you use git head you need at least bitbake1.10.2 Jan 27 18:06:03 but it will tell you Jan 27 18:06:38 kergoth__: removed cache, still no success Jan 27 18:08:08 cleaned tmpdir see if that helps Jan 27 18:19:50 did anyone notice, that opkg silently replaces any call to depmod in pkg-postinst with "depmod -A", which does nothing if modules.dep has a newer timestamp than the installed modules? Jan 27 18:25:23 effectively that means, that you can install exactly one module from a feed Jan 27 18:25:43 after that, modules.dep will have a newer timestamp than the modules Jan 27 18:26:12 for any following module, modules.dep won't get updated any more Jan 27 18:27:56 03Tom Rini  07org.openembedded.dev * r8e4ad8fe12 10openembedded.git/recipes/openswan/ (2 files in 2 dirs): Jan 27 18:27:57 openswan: Fix a build race in libipsecconf Jan 27 18:27:57 This has been upstreamed for the 2.6.33 release. Jan 27 18:27:57 Signed-off-by: Tom Rini Jan 27 18:28:00 03Ryan D Phillips  07org.openembedded.dev * r7298be1f35 10openembedded.git/recipes/bind/ (4 files): Jan 27 18:28:00 bind: Add DISTRO_FEATURES support for ipv6 Jan 27 18:28:00 Signed-off-by: Ryan D Phillips Jan 27 18:28:00 Signed-off-by: Tom Rini Jan 27 18:32:12 eFfeM: i did.. ill give it a full retry later on Jan 27 18:33:13 ok, gl Jan 27 19:00:40 kergoth: do_patch openembedded/recipes/linux-libc-headers/linux-libc-headers-2.6.32/export_sa_family_t.patch fails. And some other failures as well. Jan 27 19:00:57 Why so many issues? Last time it was pretty stable Jan 27 19:01:39 why are you asking me? Jan 27 19:02:25 Whom should I ask? Jan 27 19:02:58 no one. ask in the channel, wait for someone to respond Jan 27 19:03:03 there are over 100 people in here Jan 27 19:03:20 i'm not the sole maintainer of OE, i'm one of many developers, that's all Jan 27 19:03:26 nor am i the maintainer of the linux-libc-headers recipe Jan 27 19:06:20 kergoth__: Hi, today someone reported also missing "save" in install wrapper, is it resolved by your last install commit? Jan 27 19:06:49 kergoth__: ah save is function in wrapper.sh so I guess so (. instead of source) Jan 27 19:07:11 yeah, its strange that some people encountered a problem and not others. my /bin/sh handled source fine Jan 27 19:07:14 either way it should be fixed Jan 27 19:07:35 erm, was fixed, to put it more clearly :) Jan 27 19:08:36 eFfeM: ERROR: Could not inherit file classes/rootfs_${IMAGE_PKGTYPE}.bbclass while parsing /OE/openembedded/recipes/images/xfce46-image.bb Jan 27 19:10:00 think i need to define IMAGE_PKGTYPE somewhere.. Jan 27 19:10:32 most distros do that for you.. Jan 27 19:11:24 anstrom201.x Jan 27 19:12:15 +g Jan 27 19:13:33 Hi I am new to OE , I was trying to compile linux kernel 2.6.22 with distribution=arago and MACHINE=omap3evm with CSL external tool chain using the command " MACHINE=omap3axon bitbake virtual/kernel -D" after compiling i am seeing "uImage-2.6.28.10+xxxxxxxx.bin " getting generated in "deploy/glibc/images/omap3axon/ " where i was expecting "uImage-2.6.22.10+xxxxxxxx.bin" as I am usin g only 2.6.22 version of kernel can any onehlep Jan 27 19:14:24 did you specify a prefered_virtual_kernel ? Jan 27 19:14:52 yes Jan 27 19:15:33 and it exists as a recepie and bitbake doesnt complain about not finding it Jan 27 19:15:43 before starting pulling 2.6.29 Jan 27 19:16:05 pastebin the log its easier to read Jan 27 19:17:25 i have modified the KVER= 2.6.22 in my .bb file Jan 27 19:17:36 should it replace the degault KVER ? Jan 27 19:18:38 should nt it override the default KVER value Jan 27 19:18:40 dont think thats how its intndd Jan 27 19:18:44 intended Jan 27 19:18:52 there is a 2.6.32 bb already Jan 27 19:26:16 do_patch openembedded/recipes/linux-libc-headers/linux-libc-headers-2.6.32/export_sa_family_t.patch fails. And some other failures as well in other packages. What can be the problem? Jan 27 19:29:24 kergoth__: I'm not sure what happened, keep having that error, in the end did rm TMPDIR cleaned /tmp and refetched bitbake; now it is buildign again Jan 27 19:29:51 is anyone here having write access to the opkg repository? Jan 27 19:30:30 unsolo : I am using PREFERRED_PROVIDER_virtual/kernel = "linux-omap3xxx" in my .bb file Jan 27 19:30:52 in your conf file you mean ? Jan 27 19:30:55 kergoth__: maybe it was waiting for oetstats or so, i noticed (with -v -DDD) a long time of inactivity between printing the build config and preparing runqueue Jan 27 19:31:03 no cpu usage, no disk activity Jan 27 19:31:07 sorry in my comf file Jan 27 19:31:13 conf file Jan 27 19:32:24 but where is linux-omap3xxx ? is that your own bb file ? Jan 27 19:32:41 i am building for a different machine but which is compatible with omap3 so replace the opmap3 --> opmap3xxx in my local.conf file Jan 27 19:32:52 eFfeM: seems angstrom 2010.x doesnt work very well in OE :) Jan 27 19:33:19 yes i have my own .bb file in recipeies /linux linux-opma3xx .bb file Jan 27 19:34:11 ahh Jan 27 19:37:55 btw its sreddy so you changed the PREFERRED_PROVIDER_virtual/kernel = "linux-omap-psp" Jan 27 19:38:07 oh a bit to much there before sreddy Jan 27 19:38:27 which is in omap3evm machine config ? Jan 27 19:39:21 http://www.securityfocus.com/bid/46027 Jan 27 19:39:27 MuPDF 'closedctd()' PDF File Handling Remote Code Execution Vulnerability Jan 27 19:39:37 yes i i changed PREFERRED_PROVIDER_virtual/kernel = "linux-opam3xxx" in my machine xxx.conf file Jan 27 19:40:31 unsolo, I use minimal distro :-) Jan 27 19:40:47 i've never build angstrom 2010.x Jan 27 19:41:21 well its no big thing im making a new distro anyways.. Jan 27 19:42:00 do you have a linux-omap3xxx directory sreddy Jan 27 19:42:17 * unsolo is trying to learn OE by having sreddy do the wrong thing Jan 27 19:42:22 yes i created one and place the my defconfig Jan 27 19:43:06 well i am also learning ;) Jan 27 19:47:09 im trying to figure out how to get 2.6.34.. Jan 27 19:55:39 i know the vanilla 2.6.34 boots nicely on the omap Jan 27 19:55:55 btw also running evm sreddy Jan 27 19:56:13 guess i have to add the vanilla and start there.. Jan 27 19:56:52 only thing that the vanilla 2.6.34 doesnt do well on the evm is display and pm from what i can see Jan 27 19:57:22 obi: for opkg ask grg when he is online Jan 27 19:57:27 obi: he is in OZ Jan 27 19:57:45 khem: thanks Jan 27 20:02:29 Hmm i see that there is a omap3-2.6.34 but it is for the 1.5GHZ arm+dsp thing Jan 27 20:02:52 ti816x Jan 27 20:03:08 which is a omap with somewhat more power.. Jan 27 20:07:15 03Leon Woestenberg  07master * r9b5df85143 10openembedded.git/ (23 files in 5 dirs): Jan 27 20:07:15 adb4000.conf: New machine, kernel and u-boot support. Jan 27 20:07:15 STK ADB4000 is a development board with a SAM9G45 SODIMM CPU module. Jan 27 20:07:15 These patches are from the manufacturer (In-Circuit). Jan 27 20:07:15 Resulting kernel boots on the board. Jan 27 20:07:16 Signed-off-by: Leon Woestenberg Jan 27 20:07:34 gm Jan 27 20:12:06 likewise_: howdy Jan 27 20:12:16 lol Jan 27 20:12:50 howdy khem, all Jan 27 20:13:53 likewise_: ADB4000 is it cortex bases ? Jan 27 20:14:05 khem: no, arm9 Jan 27 20:14:45 ah SAM9G45 Jan 27 20:14:56 that should be armv5te isnt it Jan 27 20:15:04 are you using oe/master for it Jan 27 20:15:40 khem: armv5te, tune-arm926ejs.inc, yes Jan 27 20:16:08 khem: had Qt running on it today Jan 27 20:16:35 cool Jan 27 20:16:41 ARM926 it it Jan 27 20:17:05 are you running with angstrom2010 Jan 27 20:17:08 or minimal ? Jan 27 20:17:35 khem: the build doesn't work for minimal, at least not one or two weeks ago: http://bugs.openembedded.org/show_bug.cgi?id=5498 Jan 27 20:18:58 has anyone in here run an opkg upgrade since sometime last night on angstrom? Jan 27 20:19:08 it looks like boost libs 1.45 got pushed, but the headers are still 1.41 Jan 27 20:21:02 likewise_: hmm 2.22 is old you should get atleast 2.24 Jan 27 20:24:33 khem: I think I tried, but I think it's a distro bug, as I didn't override any version numbers Jan 27 20:32:29 likewise_: ok right now my machine is not accessible to me but I will take a look Jan 27 20:32:50 likewise_: I build minimal quite often never saw this issue Jan 27 20:34:24 likewise_: strange for me minimal distro builds glib-2.0-native-2.26.1-r3 Jan 27 20:34:26 Crofton|work: ping? Jan 27 20:34:32 and builds is successfully Jan 27 20:34:57 likewise_: git pull to latest and retry Jan 27 20:35:02 it should work Jan 27 20:37:14 likewise_: hang on what version of bitbake do u use ? Jan 27 20:37:34 likewise: hmmm I see "1.11.0" Jan 27 20:38:29 master :) Jan 27 20:38:36 khem: now starting 1.10.2 and minimal Jan 27 20:38:37 likewise: there is glib-2.0-native_2.22.1.bb but then glib-2.0_2.26.1.bb has BBCLASSEXTEND="native" Jan 27 20:38:52 so somehow your build picks glib-2.0-native_2.22.1.bb over BBCLASSEXTEND Jan 27 20:39:08 Tartarus: please backout the bind patch thats not needed Jan 27 20:39:12 on master Jan 27 20:39:29 khem, oh? Jan 27 20:39:38 I wonder why do we have to keep glib-2.0-native_2.22.1.bb Jan 27 20:39:58 khem: yup, correct analysis, I could not find the reason why though Jan 27 20:40:18 khem, why isn't it needed? Did that change go into autotools.bbclass? Jan 27 20:40:52 Tartarus: read my response to thread in ml Jan 27 20:40:55 yeah, ok, oops, thanks :) Jan 27 20:41:06 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r9c66f0a21b 10openembedded.git/recipes/xorg-xserver/ (2 files in 2 dirs): (log message trimmed) Jan 27 20:41:06 xserver-xorg-conf: htcdream: switch to htcdream version of the US keyboard Jan 27 20:41:06 The htcdream has an hardware keyboard with some key binded to Jan 27 20:41:06 alt+key,for instance alt+q should produce a Tab. Jan 27 20:41:06 Before we didn't activate that feature, and instead we used Jan 27 20:41:07 the US keyboard default. Jan 27 20:41:08 This feature depend on xkeyboard-config_git which is the default Jan 27 20:41:24 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r0e80cde9f0 10openembedded.git/recipes/xfce-base/xfdesktop_4.6.2.bb: Jan 27 20:41:24 xfdesktop 4.6.2: fix compilation Jan 27 20:41:24 Without that fix we have: Jan 27 20:41:24 libtool: link: cannot find the library `=/usr/lib/libICE.la' or unhandled argument `=/usr/lib/libICE.la' Jan 27 20:41:25 Note that libtool is not the cross libtool. Jan 27 20:41:25 Signed-off-by: Denis 'GNUtoo' Carikli Jan 27 20:42:13 likewise_: you can send a patch to remove recipes/glib-2.0/glib-2.0-native_2.22.1.bb Jan 27 20:42:18 no one pins it Jan 27 20:42:46 all distros have moved to 2.24 minimum Jan 27 20:42:50 03Tom Rini  07org.openembedded.dev * r4b632377f1 10openembedded.git/recipes/bind/ (4 files): Jan 27 20:42:50 Revert "bind: Add DISTRO_FEATURES support for ipv6" Jan 27 20:42:50 As Khem noted this is no longer needed. Jan 27 20:42:50 This reverts commit 7298be1f35f88c4646a47a59e2ac96e5efa5b39a. Jan 27 20:52:29 bitbake gets stuck for me very often (every third time I run it) after parsing (both 1.10.12 and master) Jan 27 20:53:05 ah, I have oestats enabled, seems a server problem Jan 27 20:53:28 one CTRL-C mostly gets me out of it Jan 27 20:56:03 yes Jan 27 20:56:04 and of course I run into the daily amount of race conditions (they take ages to debug): http://pastebin.com/t4H3Ufth Jan 27 20:56:16 disable oestats Jan 27 20:56:59 woglinde: I love it for sharing build problems though... Jan 27 21:00:56 khem: hmm, old glib selected again... What is your host system? Jan 27 21:21:13 anyone have any thoughts on https://github.com/kergoth/openembedded/compare/master...autotools ? needs rebasing against recent OE, but conceptually.. Jan 27 21:27:51 kergoth__: yes, the commit message doesn't tell me why this is refactored :-) Jan 27 21:28:33 https://github.com/kergoth/openembedded/commit/206cd21 Jan 27 21:28:39 is that message somehow unclear? Jan 27 21:28:53 guess i could elaborate on why i split out oe_autoreconf Jan 27 21:34:02 kergoth__: Ah yes, commit messages seem ok to understand what is happening there, for me it seems important future breakage can be related to commit messages. I cannot really comment on the actual code changes. Jan 27 21:37:34 np, thanks Jan 27 22:27:51 So.. target perl Jan 27 22:27:55 What's a good test? Jan 27 22:28:01 perl -e 'exit 11'; echo $? gives '11' Jan 27 22:28:05 rm -rf ? Jan 27 22:28:28 (actually there is a perl test suite that is part of the LSB.. it's a really good way to verify perl functionality) Jan 27 22:29:03 Don't have those in OE yet, heh Jan 27 22:29:39 I suspect you can run download and run the binaries.. ;) Jan 27 22:29:56 perl itself I think also has a test suite, but I don't believe it's as extensive (more perl language based then modules and functional based) Jan 27 22:30:29 * Tartarus ponders sneaking this into poky first and waiting :) Jan 27 22:30:50 I know in Poky we've been running some, if not all of the LSB perl test harness Jan 27 22:31:09 it's not exactly 100% though cause there are some config changes needed to pass LSB Jan 27 23:26:33 03Andrea Adami  07org.openembedded.dev * rf626c0217a 10openembedded.git/recipes/klibc/ (20 files in 2 dirs): Jan 27 23:26:33 klibc: move to latest 1.5.21 Jan 27 23:26:33 * temporarly remove the klcc binary for target (broken paths) Jan 27 23:26:33 * needs to be fixed before recipe refactor (debian-style) Jan 27 23:26:33 * reset INC_PR and PR Jan 27 23:26:34 Signed-off-by: Andrea Adami Jan 28 00:17:09 likewise_: mine is different ubuntus Jan 28 00:17:21 likewise_: today I installed f14 on one Jan 28 00:17:22 lets see Jan 28 00:32:21 khem: at least when I fix the PREFERRED_VERSION_glib-2.0-native="2.26.1" it seems to work on two Ubuntu 32-bit boxes. Jan 28 00:32:37 ok Jan 28 00:32:50 I think I will get rid of 2.20 native recipe Jan 28 00:35:34 I wonder if the version gets selected "at random" if none is set. glib-2.0- has both BBCLASSEXTEND as well as 2.0 in its name... **** ENDING LOGGING AT Fri Jan 28 02:59:56 2011