**** BEGIN LOGGING AT Sat Oct 16 02:59:58 2010 Oct 16 03:18:23 wtf Oct 16 03:18:24 ERROR: ? while parsing /Users/kergoth/Code/Devel/oe/openembedded/recipes/quilt/quilt-native_0.48.bb Oct 16 03:18:29 that's helpful.. Oct 16 06:03:18 03Khem Raj  07master * r6901dc958c 10openembedded.git/recipes/curl/ (7 files in 2 dirs): Oct 16 06:03:18 curl: Fix build for native BBLASSEXTEND Oct 16 06:03:18 Delete unused recipes. Oct 16 06:03:18 Signed-off-by: Khem Raj Oct 16 06:08:45 03Henning Heinold  07master * reb05920272 10openembedded.git/recipes/libsdl/ (3 files in 2 dirs): Oct 16 06:08:46 libsdl-ttf: fix version 2.0.10 by removing the libtool-m4-macros Oct 16 06:08:46 * bump PR Oct 16 06:08:46 Signed-off-by: Martin Jansa Oct 16 07:16:51 gm Oct 16 08:02:28 anyone an idea where the Packaged contents lines are constructed/outputted? Oct 16 08:02:30 Packaged contents of module-init-tools-cross-dev into /home/hudson/jobs/FM_TEST/workspace/tmp/deploy/ipk/armv5te/module-init-tools-cross-dev_3.2.2-r5_armv5te.ipk Oct 16 08:02:35 like that one Oct 16 08:13:26 oh my got, why do I deserve this Oct 16 08:13:31 s/got/god/ Oct 16 08:13:50 bitbake -c clean lua Oct 16 08:14:00 NOTE: package lua-5.0.2-r5: task do_clean: Succeeded Oct 16 08:14:12 wtf? there's lua-5.1.4 clearly Oct 16 08:14:21 lua-5.1.4-r6 Oct 16 08:14:52 looks like using bitbake master wasn't a good idea Oct 16 08:14:53 :) Oct 16 08:16:53 or maybe it's the bbappend stuff causing me such a headaches, who knows Oct 16 08:18:42 addded DEFAULT_PREFFERENCE = "-1" to lua_5.0.2.bb makes no difference, it's still using it, arrrgh Oct 16 08:31:21 found it, last line of ./recipes/stage-manager/files/stage-manager-ipkg-build Oct 16 08:53:21 03Thomas Zimmermann  07org.openembedded.dev * r9598964b05 10openembedded.git/recipes/bluez/obexd_0.34.bb: Oct 16 08:53:22 obexd_0.34.bb: DEPEND on libical Oct 16 08:53:22 * This dependencie was added with 0.24 Oct 16 08:53:22 Signed-off-by: Thomas Zimmermann Oct 16 08:57:09 ynezz, I don' t really understand your USERDISTRO patch, I see only assignments of the type: USERDISTRO := "${DISTRO}" Oct 16 08:57:14 with and without : Oct 16 08:57:23 I would expect them to be the same Oct 16 08:57:39 look at tinderbox Oct 16 08:57:52 http://tinderbox.openembedded.net/builders/ynezz/ Oct 16 08:57:59 the last one is without the patch Oct 16 08:58:18 angstrom is changing it Oct 16 08:58:29 look at conf/distro/include/angstrom.inc Oct 16 08:58:30 i understand what you want to achieve, but I would expect the vars to be the same Oct 16 08:58:40 no, they're not Oct 16 08:58:53 actually got to this because I wanted to check if this does not break things for other distros Oct 16 08:59:14 i believe you when you say they are not the same, but just can't find it in the code Oct 16 08:59:53 rep USERDISTRO -A2 conf/distro/include/angstrom.inc Oct 16 08:59:56 grep USERDISTRO -A2 conf/distro/include/angstrom.inc Oct 16 09:00:03 should be clear then Oct 16 09:00:25 yeah, understood Oct 16 09:00:35 maybe it does apply only to angstrom Oct 16 09:00:37 will ack Oct 16 09:00:42 ok, thanks :) Oct 16 09:01:06 done Oct 16 09:01:57 i saw that in bitbake.conf it always gets a value; which was my main concern Oct 16 09:02:29 afk for a while Oct 16 09:03:04 tharvey: ^ Oct 16 11:02:13 what is the preferred way of solving libtool 2.4 version mismatch errors? (in gnutls in my case) Oct 16 11:02:40 woglinde told me to rm that libtool.m4 Oct 16 11:03:04 effem see my fix for libsdl-ttf Oct 16 11:03:21 woglinde: thanks for the pointer Oct 16 11:03:29 unfornatly it cannt got to autotools.bbclass Oct 16 11:03:42 because some packages have the macros in another dir Oct 16 11:03:48 like sdl-ttf Oct 16 11:12:47 woglinde: I'd say it can! just have a var to specify the dir and give the var a sensible default, then we could just redefine the dir var in the recipes that need it! Oct 16 11:12:52 seems a lot more simpler Oct 16 11:13:11 effem hm and ovveride it when needed Oct 16 11:13:13 okay Oct 16 11:13:28 would made sense Oct 16 11:13:51 yes Oct 16 11:14:01 one small issue, it could be that there are multiple dirs Oct 16 11:14:11 hm yes Oct 16 11:14:19 but I didnt see that yet Oct 16 11:14:36 one can solve it with DIRS = "" and iterate over it Oct 16 11:14:39 e.g. I have an issue with gnutls, this one has m4 libextra/m4 and lib/m4 Oct 16 11:14:58 woglinde: are it always the 5 files you mentin in the libsdl-ttf patch ? Oct 16 11:15:16 i'm not really an auto wizz (and don't plan becoming one) Oct 16 11:15:57 effem only for libtool 2.2 older libtools has other Oct 16 11:16:07 effem you could grep the recipe Oct 16 11:16:26 I did a few patches for same issues when 1.5 -> 2.2 libtool transition Oct 16 11:16:53 can I find out from the files themselves ? Oct 16 11:17:04 btw guess that would make it more difficult for autotools Oct 16 11:17:12 bbclass Oct 16 11:17:32 gnutls m4 dir has Oct 16 11:17:33 gtk-doc.m4 guile.m4 libtool.m4 ltoptions.m4 ltsugar.m4 ltversion.m4 lt~obsolete.m4 pkg.m4 Oct 16 11:17:54 I was just wondering if it could be fixed in more generic way by some modifications in autotools class Oct 16 11:19:42 libtool.m4 ltoptions.m4 ltsugar.m4 ltversion.m4 lt~obsolete.m4 Oct 16 11:19:50 you need to remove Oct 16 11:26:11 jo zecke Oct 16 11:26:29 ok, have done that will test, then submit patch Oct 16 11:28:12 03Martin Jansa  07master * rd52cc61e95 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Oct 16 11:28:12 SHR: prefer libtool 2.4 Oct 16 11:28:12 Signed-off-by: Martin Jansa Oct 16 11:28:30 effem okay Oct 16 11:28:32 03Martin Jansa  07master * ra514e32630 10openembedded.git/recipes/xorg-font/ (font-util-native_1.1.1.bb font-util_1.1.1.bb): Oct 16 11:28:32 font-util_1.1.1: replace -native recipe with BBCLASSEXTEND Oct 16 11:28:32 Signed-off-by: Martin Jansa Oct 16 11:28:33 03Martin Jansa  07master * r823b452286 10openembedded.git/recipes/xorg-font/font-util_1.1.2.bb: Oct 16 11:28:33 font-util: add version 1.2.0 Oct 16 11:28:34 Signed-off-by: Martin Jansa Oct 16 11:28:34 03Martin Jansa  07master * r23c06fa864 10openembedded.git/recipes/xorg-font/xorg-font-common.inc: Oct 16 11:28:35 xorg-font-common: use BPN instead of XORG_PN Oct 16 11:28:35 Signed-off-by: Martin Jansa Oct 16 11:28:36 03Martin Jansa  07master * rbaf37a1dcb 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Oct 16 11:28:37 EFL: bump SRCREV a bit more Oct 16 11:28:37 Signed-off-by: Martin Jansa Oct 16 11:28:38 03Ɓukasz Pankowski  07master * rf3a3b438a3 10openembedded.git/recipes/ffalarms/ (ffalarms/ical.timezone.patch ffalarms_git.bb): Oct 16 11:28:38 ffalarms: bump SRCREV, remove patch applied upstream Oct 16 11:28:39 Signed-off-by: Martin Jansa Oct 16 11:28:39 03Martin Jansa  07master * rc2802927e5 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Oct 16 11:28:40 SHR: use newer python-pycairo Oct 16 11:28:40 Signed-off-by: Martin Jansa Oct 16 11:29:35 shr: bump DISTRO_PR after gcc changes Oct 16 11:29:35 Signed-off-by: Martin Jansa Oct 16 11:29:35 03Martin Jansa  07master * r7c5e37db7a 10openembedded.git/ (6 files in 3 dirs): Oct 16 11:29:35 xserver-xorg: move from 1.9.0.901 to 1.9.0.902 Oct 16 11:29:36 Signed-off-by: Martin Jansa Oct 16 11:29:36 03Andreas Oberritter  07master * r68abc46555 10openembedded.git/classes/utils.bbclass: Oct 16 11:29:37 utils.bbclass: add name to SRC_URI[{md5, sha256}sum] Oct 16 11:29:37 * When a checksum is missing or invalid, print strings which can be Oct 16 11:29:38 copied into a recipe without modification. Oct 16 11:29:38 (2 lines omitted) Oct 16 11:30:37 hlipbt Oct 16 11:31:16 jtuuujfzuruzzzrtzevcebceg Oct 16 11:59:57 woglinde: gnutls patched for libtool, builds fine; patch submitted to ml Oct 16 12:00:06 could use an ack from someone Oct 16 12:00:46 we might also consider expiring the 2.8 versions, I only fixed and tested 1.10.1 Oct 16 12:04:55 effem why? Oct 16 12:05:01 if its working Oct 16 12:05:04 push it Oct 16 12:15:13 woglinde: gnutls is not really a recipe i maintain or am familiar with and no idea how widespread it is, so decided to do it the safe way Oct 16 12:15:58 git log should give you an idea of who to talk too Oct 16 12:16:16 I thnik gnutls has no maintainer Oct 16 12:16:28 but is used by a wide varity of apps Oct 16 12:17:15 jama added 10.1, khem did an update for the libtool 2.4 macros on oct 5, then again I still got the 2.2.6 or so message Oct 16 12:17:22 woglinde: thought it was widely used Oct 16 12:17:44 that's why i decided to be careful, core changes should be reviewed Oct 16 12:18:38 and the recipe itself builds fine, but the image itself is still building, wanted to be sure that it wait for that one too Oct 16 12:18:47 if it was just an ordinary recipe I would have pushed it thogh Oct 16 12:19:25 (and i am not an autotools/libtool/etc hero) (and I am :-) with that) Oct 16 12:31:36 hmm something broke unpacking from FILESPATH.. Oct 16 12:31:43 NOTE: Unpacking dev/recipes/linux/linux/defconfig to tmpdir-shr/work/spitz-oe-linux-gnueabi/linux-2.6.35+2.6.36... Oct 16 12:32:02 it should use dev/recipes/linux/linux/spitz/defconfig, when I remove dev/recipes/linux/linux/defconfig it does again Oct 16 12:32:11 is it because of reversed OVERRIDES? Oct 16 12:38:24 could be Oct 16 12:39:36 jama hm intressting Oct 16 12:41:20 maybe just def machine_paths(d) bug :/ Oct 16 12:41:52 pity that I've just finished rebuild from scratch :/ hard to grep all fscked recipes Oct 16 12:46:38 woglinde: zecke does not seem to like the gnutls patch and approach Oct 16 12:47:38 bah Oct 16 12:47:43 I will strap him Oct 16 12:47:49 next time I see him Oct 16 12:49:18 jama maybe your issie is called due to the overrides change and the fact that linux is both a recipe and an os name Oct 16 12:49:37 because TARGET_OS is also in teh OVERRIDES Oct 16 12:50:59 args Oct 16 12:51:13 I will strap zecke a second time for not reading the patch Oct 16 12:55:23 woglinde: I see '+' in ltoptions.m4 Oct 16 12:55:50 o.O? Oct 16 12:55:50 woglinde: the whole point is that the patch is so big... just by the size it can not be right Oct 16 12:55:59 look at do_configure_prepend Oct 16 12:56:20 the patch is so big because effem removes the stupid patch Oct 16 12:56:29 which patches the m4 macros Oct 16 12:56:36 * woglinde will strap zecke a third time Oct 16 12:56:54 woglinde: I think every patch of this size is wrong Oct 16 12:57:04 zecke *sigh* again Oct 16 12:57:18 someone ALREADY patched the m4 macros Oct 16 12:57:21 the question is, why 'autoreconf'/'libtoolize' does not override the old files Oct 16 12:57:24 effem is revertung that Oct 16 12:57:33 ensc the files get copied Oct 16 12:57:39 and then you have mixed stuff Oct 16 12:58:01 its same situation as with libtoll 1.5 -> 2 transition Oct 16 12:58:07 woglinde: well... if you remove a file it is a ---+++ 1 January 1970. but this patch is thousands of lines Oct 16 12:58:23 zecke *sigh* Oct 16 12:58:33 please look at the gnutls dir Oct 16 12:58:44 there is NOW the patch which patches m4 macros Oct 16 12:58:50 but removing the m4 macros Oct 16 12:58:58 the patch is not NEEDED anymore Oct 16 12:59:31 .o(need to think of a new punishment) Oct 16 13:00:33 woglinde: I got lost in the diff of a diff Oct 16 13:00:38 yeah Oct 16 13:00:43 thats tricky Oct 16 13:00:49 all + Oct 16 13:00:52 and all - Oct 16 13:01:01 woglinde: would be nicer if git format-patch would show "Delete old junk" Oct 16 13:01:40 aeh Oct 16 13:01:47 look at all the ------------------------------------------ Oct 16 13:01:54 on summary Oct 16 13:01:55 woglinde: yeah, the dirstat even says deleted Oct 16 13:02:20 okay Oct 16 13:02:22 woglinde: new files are copied only when you remove the old files Oct 16 13:02:37 ensc hm might bee Oct 16 13:02:45 didnt looked 100% in the autoconf code Oct 16 13:03:10 finding out, why libtoolize does not work as expected might be the better longtime strategy instead of patching every recipes which fails Oct 16 13:03:33 depends Oct 16 13:03:41 woglinde: with git.. it could just say... delete this from manifest, current blob is... XYZ. anyway.. i should have seen the patch is deleted Oct 16 13:05:10 ensc for its faster Oct 16 13:13:40 zecke, not that I have looked at the patch, but also, Oct 16 13:13:54 if it is possible do deletions in a seperate commit :) Oct 16 13:14:20 crofton if its not working Oct 16 13:14:26 you have to revert 2 commits Oct 16 13:17:39 i'd say keep things together that go together (and indeed I only removed an existing patch file that is not needed any more) Oct 16 13:19:33 zecke: woglinde if you agree that the patch is ok, please add an ack to it Oct 16 13:20:17 woglinde, i SAID i HADN'T LOOKED AT THE PATCH IN QUESTION :) Oct 16 13:20:21 URG Oct 16 13:20:26 stupid caps lock Oct 16 13:20:31 more coffee Oct 16 13:20:39 eFfeM: okay, and sorry for the noise. Oct 16 13:21:00 crofton are you under some presaure? Oct 16 13:21:06 heh Oct 16 13:21:06 zecke: no problem; I'm known to make noise too Oct 16 13:21:08 yeah Oct 16 13:21:16 eFfeM: you have libtool-2.4 but without LIBTOOL_HAS_SYSROOT set to yes? Oct 16 13:21:16 need to ship hw in Nov Oct 16 13:21:55 zecke and I'd rather have things fully clarified with a discussion upfront than revering things afterward Oct 16 13:22:09 JaMa: I think so, let me check Oct 16 13:22:23 eFfeM: if yes, then why? Oct 16 13:22:29 eFfeM: well, it is not about you at all. I hope git could generate a removal in one line. Oct 16 13:23:16 JaMa, i only set 2.4 as preferred in my local.conf haven't set the has_sysroot thing Oct 16 13:23:59 eFfeM: that's why gnutls failed for you.. Oct 16 13:24:00 wanted to mvoe forward without creating too much problems by making too big steps, i have an image to build thatt I want to use too Oct 16 13:24:22 JaMa, should it not work in that case too ? Oct 16 13:24:49 probably yes.. but I would say it's not much expected scenario Oct 16 13:25:47 if 2.4 and LIBTOOL_HAS_SYSROOT should go together it should me made in such a way that they do go together Oct 16 13:26:24 and in any cas as I could delete a big patch file, I feel this is better Oct 16 13:26:26 how would you implement it? Oct 16 13:27:04 I agree.. just wanted to point that with LIBTOOL_HAS_SYSROOT set you would get that big patch at least applied and then gnutls build tok Oct 16 13:27:08 JaMa, not sure but I feel it would be possible in python to combine the two Oct 16 13:27:19 biab need to replace kbd batteries Oct 16 13:27:47 re Oct 16 13:28:37 JaMa would the patch I did work with LIBTOOL_HAS_SYSROOT? if so, I'd say it is an improvement Oct 16 13:28:58 yup agreed on that Oct 16 13:29:06 basically it removes an almost 11000 lines patch file Oct 16 13:30:00 ok, afk repairing my bike so I can go cycling tomoroow Oct 16 13:30:08 back in 30-60 mins Oct 16 13:44:48 anyone have any thoughts on how I should install a sane vimrc file? Oct 16 13:45:02 should I just add it to the main vim install Oct 16 13:45:15 or make a vim-conf package that installs it Oct 16 13:51:07 crofton hm Oct 16 13:51:29 if you have plenty space Oct 16 13:51:35 use the second option Oct 16 13:52:19 that has the advantage of not breaking old school mode for existing people that use it Oct 16 13:52:32 although using it like that is crazy :) Oct 16 15:10:17 Hm, is there a way how to force order of loaded kernel modules? Oct 16 15:10:36 do I would like to find some way without using the udev Oct 16 15:10:41 s/do// Oct 16 15:10:58 so it works with mdev also Oct 16 15:11:44 maybe blacklist the modules and load them manualy in some rc script? Oct 16 15:17:23 JaMa: rebuild failed at eglibc: http://shr.pastebin.com/tpBrATyZ Oct 16 15:21:31 JaMa: scripts/gen-libc-abis looks quite bogus Oct 16 15:21:32 PATIBILITY IDEOGRAPH-2F905 Oct 16 15:21:32 /xf0/xaf/xa4/x86 CJK COMPATIBILITY IDEOGRAPH-2F906 Oct 16 15:21:36 are the first two lines Oct 16 15:31:00 heyho Oct 16 15:43:12 03Philip Balister  07org.openembedded.dev * r937cc3dae2 10openembedded.git/recipes/uhd/ (uhd.inc uhd_git.bb): uhd: Add recipe for Ettus Research Universal Hardware Driver. Oct 16 16:19:53 rew Oct 16 16:41:39 hi Oct 16 16:41:44 hi gnutoo Oct 16 16:41:48 I've that: Oct 16 16:41:59 | cp: target `/home/gnutoo/embedded/oe/oetmps/bug/work/bug-angstrom-linux-gnueabi/linux-bug-2.6.27.2+svnr10746-r32/image/kernel//include/asm/' is not a directory Oct 16 16:42:06 hi woglinde Oct 16 16:53:03 is there something I missed for my kernel issue? such as some conversion that was not made? Oct 16 17:08:52 hmmm Oct 16 17:08:56 there is a do_install_append Oct 16 17:09:52 somewhre in install -d missing and orderchange d? Oct 16 17:10:47 http://www.pastebin.ca/1964136 Oct 16 17:11:02 ok Oct 16 17:11:04 possible Oct 16 17:12:31 hmmm I think I get it Oct 16 17:12:56 you copy to $kerneldir/include/asm but that is not guaranteed to exist, ust add an install -d (install -d is better than mkdir as it will also create inbetween dirs if needed) Oct 16 17:13:36 I'll correct to that: /home/gnutoo/embedded/oe/oetmps/bug/work/bug-angstrom-linux-gnueabi/linux-bug-2.6.27.2+svnr10746-r32/image/usr/src/linux-2.6.27.2/include/asm Oct 16 17:13:40 guess something changed due to staging being directly done from ${S} to stage dir Oct 16 17:13:50 indeed Oct 16 17:13:56 because there is no kernel dir anymore Oct 16 17:14:10 just add the install Oct 16 17:14:13 I bet that's caused from the changes with the long stagin process Oct 16 17:14:21 are you sure? Oct 16 17:14:37 I should rather change the path Oct 16 17:14:59 let me look Oct 16 17:15:00 again Oct 16 17:15:42 ah ok Oct 16 17:15:44 sorry Oct 16 17:15:53 I didn't look attentively enough Oct 16 17:15:57 it's the dest that it can't find Oct 16 17:15:59 not the src Oct 16 17:16:18 so I'll do as you said Oct 16 17:16:51 yes, this patch is proably the cause http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=b6cc4bc217e695801db060a1366915a870cab5db Oct 16 17:17:52 it moved stuff from kernel_do_install to sysroot_stage_all_append, guess before this patch the dir was created in the do_install code, that now moves things directy to staging Oct 16 17:18:34 I saw the mailing list and suspected now that it had to do with NOTE after sawing the absence of .../kernel Oct 16 17:18:48 yes Oct 16 17:20:01 do you need to have the file to go to staging eventually? guess you want to, so maybe adding to do_install is not the proper way any more Oct 16 17:20:25 back in 30 mins or so Oct 16 17:21:15 Question: OE has a defconfig in /recipes/linux/linux-msm7/htckaiser but fails to compile virtual/kernel, but the msm7 kernel from git.linuxtogo.org/home/groups/mobile-linux/kernel.git repository has one that works, I just copied the .config over the one OE provides. My question, What is the proper way for patching this in OE. should I diff them and patch OE's copy before compile time and summit it that way or change the way OE does t Oct 16 17:21:15 his and make it always use the one from the repo? Oct 16 17:24:34 overwrite the defconfig with the working .config Oct 16 17:24:40 eFfeM, I don't know, I just hit the install problem Oct 16 17:24:59 and yes that's what I suspected Oct 16 17:25:08 I must look for the proper way Oct 16 17:25:39 benjamin_: bump PR of the kernel recipe, build it, boot it and it works, then send the patch Oct 16 17:25:51 s/and it/and if it/ Oct 16 17:28:53 ynezz - Wish it was that simple newest kernel will not boot from haret so right now im working out how to get a KAISER.nbh file with the new kernel and initrd like android does just thought I should get some input before I start summiting patches no one wants in OE because its not the "proper way" one day though yes exactly what you just said Oct 16 17:29:36 benjamin_: sorry, I don't follow you Oct 16 17:29:55 benjamin_: and I think, that you can try to send the patches in any case with [RFC] in the subject Oct 16 17:31:13 mrmoku: never seen this one and buildhost is just finishing eglibc atm Oct 16 17:32:01 ynezz - sorry nand it needs to boot from nand haret just bombs out when i boot it up from winmo the KAISER.nbh is a flash file the android guys use to install it on the nand Oct 16 17:33:04 I would ask the author of that recipe in the OE, git log recipe and you'll see Oct 16 17:34:22 althought I know how haret works(send few patches there also), used it on the Loox720, but I've never seen device with android and don't know anything about it's internals Oct 16 17:34:44 so I can't help you without knowing the platform details, sorry Oct 16 17:36:02 It's alright just wanted to know about patching defconfig really Oct 16 17:36:21 that I told you Oct 16 17:36:31 and thanks Oct 16 17:36:54 I copy the working .config over the defconfig, commit it and send a patch to the mailing list, that's all Oct 16 17:37:02 no magic Oct 16 17:43:49 JaMa: after clean_toolchain the build of eglibc worked fine Oct 16 17:43:56 dunno what happened Oct 16 17:44:08 maybe I was doing too many things in parallel on my poor laptop :P Oct 16 17:52:42 03Martin Jansa  07master * rfbb79ab6b1 10openembedded.git/recipes/linux/ (4 files in 3 dirs): Oct 16 17:52:42 linux(-kexecboot)_git: bump SRCREV for 2.6.36-rc8 Oct 16 17:52:42 Signed-off-by: Martin Jansa Oct 16 17:52:50 03Justus Winter <4winter@informatik.uni-hamburg.de> 07master * r09b2b1abf7 10openembedded.git/recipes/dns2tcp/dns2tcp_0.5.2.bb: Oct 16 17:52:50 dns2tcp: add recipe for dns2tcp-{client, server} Oct 16 17:52:50 * dns2tcp is a network tool designed to relay TCP connections through DNS traffic Oct 16 17:52:50 Signed-off-by: Martin Jansa Oct 16 17:52:51 03Martin Jansa  07master * r081c4d779b 10openembedded.git/recipes/tasks/task-shr-feed.bb: Oct 16 17:52:51 task-shr-feed: add dns2tcp Oct 16 17:52:51 Signed-off-by: Martin Jansa Oct 16 18:26:23 jo kergoth Oct 16 18:27:22 GNUtoo|laptop: what kernel did you have this issue with Oct 16 18:28:32 i see the asm copying mostly for older kernels Oct 16 18:29:21 mmvsta git, bug 2.6.27, omap1 2.6.25 rest is 2.6.23 or older Oct 16 18:32:51 khem, i saw your name in the log for libpam 1.1.1; get lots of notes with it about files installed but not shipped Oct 16 18:33:13 how could I remove /boot/* (kernels) from the image? Oct 16 18:33:14 all but one in /usr/lib/security/.debug Oct 16 18:33:26 yness post processing in the image build Oct 16 18:33:28 tried all pre/post process commands, but neither seems to work Oct 16 18:34:04 IMAGE_POSTPROCESS_COMMAND += "rm -f ${IMAGE_ROOTFS}/boot/?Image*" Oct 16 18:34:36 ROOTFS_POSTPROCESS_COMMAND > is what is used in images/* Oct 16 18:35:54 ynezz try something like this: Oct 16 18:36:05 ROOTFS_POSTPROCESS_COMMAND += " rm -rf ${IMAGE_ROOTFS}/boot/*; " Oct 16 18:36:21 works for me Oct 16 18:38:07 ynezz: if you need it only for some FSTYPE then look at IMAGE_CMD_jffs2 in recipes/images/shr-image.inc Oct 16 18:39:41 btw this weeks testing build seems to work quite nicely Oct 16 18:40:48 good that it was branched before reversing OVERRIDES Oct 16 18:42:16 yeah, noticed you found a sneaky bug with the patch search order Oct 16 18:42:40 eFfeM: I swear I had tested ROOTFS_ also Oct 16 18:43:39 and I wonder where it's used Oct 16 18:43:57 I can't find ROOTFS_POSTPROCESS_COMMAND in image.bbclass Oct 16 18:44:44 neither in tasks/* nor images/* Oct 16 18:44:53 so I wonder how it works Oct 16 18:46:01 ynezz, guess your tree is wrong; see http://www.pastebin.ca/1964212 for what I have Oct 16 18:47:00 if you use ipk it is executed from ./rootfs_ipk.bbclass Oct 16 18:47:19 ah, thanks Oct 16 18:47:35 now I see it Oct 16 18:47:38 line 83 to be precise, (if you are on git head) Oct 16 18:47:40 :-) Oct 16 18:48:04 yes, it's there, just missed it Oct 16 18:48:24 I should wear contact lenses even at home :D Oct 16 19:20:57 :-) Oct 16 19:28:10 tec zecke Oct 16 19:28:18 woglinde: JaMa, zecke (or someone else) can I have a 2nd ack for the gnutls patch ? Oct 16 19:28:33 so I can push it Oct 16 19:30:14 hms Oct 16 19:30:29 shouldnt delete the email Oct 16 19:31:01 woglinde: if that was to me, feel free to express your ack here and I'll add it manually Oct 16 19:31:38 btw python-twisted fails to install for me, anyone else encountered that? Oct 16 19:31:53 last try succeed Oct 16 19:32:06 only made slight local change for packaging Oct 16 19:33:48 so I hope this ack is right Oct 16 19:38:04 eFfeM: let me ack it Oct 16 19:38:05 hms why is ulf still supporting kdrive Oct 16 19:38:05 woglinde: thanks Oct 16 19:38:20 zecke: ok (was about to push) Oct 16 19:41:03 hrms anyone offhand knows which component generates the .h.in Oct 16 19:41:09 from autotools Oct 16 19:41:42 h ah autoheader Oct 16 19:42:38 he why frans get the praise Oct 16 19:42:42 for the autotools Oct 16 19:42:46 args libtool Oct 16 19:42:50 from khem Oct 16 19:42:54 *g* Oct 16 19:43:12 ah hehe Oct 16 19:43:21 I schould read thread to the end Oct 16 19:45:12 03Frans Meulenbroeks  07org.openembedded.dev * r470e6b0f64 10openembedded.git/recipes/gnutls/ (gnutls-2.10.1/libtool-update.patch gnutls_2.10.1.bb): Oct 16 19:45:13 gnutls_2.10.1: remove the libtool-m4-macros Oct 16 19:45:13 Signed-off-by: Frans Meulenbroeks Oct 16 19:45:13 Acked-by: Khem Raj Oct 16 19:45:13 Acked-by: Henning Heinold Oct 16 19:45:13 Acked-by: Holger Hans Peter Freyther Oct 16 19:49:03 do we really want FEED_ARCH as preppended to OVERRIDES? Oct 16 19:53:38 03Frans Meulenbroeks  07org.openembedded.dev * r71ca5dc3be 10openembedded.git/recipes/python/ (2 files in 2 dirs): Oct 16 19:53:38 python-twisted: removed 2.5.0 version Oct 16 19:53:38 we already have 8.2.0 for 1.5 year Oct 16 19:53:38 no one pins this Oct 16 19:53:38 some recipes depend on python-twisted-native but this version does not have a native counterpart Oct 16 19:53:38 Signed-off-by: Frans Meulenbroeks Oct 16 19:58:37 jama, your patches look good to me but my python is not really good enough to confidently ack it Oct 16 20:03:38 anyone an idea on python-twisted, this used to install for me but today I get Oct 16 20:03:39 | error: option --single-version-externally-managed not recognized Oct 16 20:07:03 hiyo Oct 16 20:07:12 I am trying to use mono on beagleboard with OE Oct 16 20:07:24 use ubuntu Oct 16 20:07:30 I feel like I'd have easier time if I installed OE as a hostin a VM to do my programming Oct 16 20:07:34 oh? Oct 16 20:07:39 isn't OE lighter? Oct 16 20:07:42 mono recipes are not well maintained Oct 16 20:08:26 oh err Oct 16 20:08:30 I see 2.6.3 on here Oct 16 20:08:36 which is fine for me so figured I'd try it Oct 16 20:09:03 but is it possible to isntall OE as a host in a VM? Oct 16 20:09:06 err guest I mean Oct 16 20:11:09 sure why not Oct 16 20:11:20 how would that work? Oct 16 20:11:23 you can even use qemu as machin Oct 16 20:11:23 e Oct 16 20:11:36 ( sorry not terribly good at that level of OS ) Oct 16 20:34:33 hm, python-twisted now installs fine (after rebuilding python-setuptools, previously it was restored from pacakged staging Oct 16 20:38:13 03Andreas Oberritter  07org.openembedded.dev * rbeec49df85 10openembedded.git/recipes/cdtextinfo/cdtextinfo_0.1.bb: Oct 16 20:38:13 cdtextinfo: initial recipe Oct 16 20:38:13 * cdtextinfo queries information about an Audio-CD from CD-Text Oct 16 20:38:13 or CDDB. It outputs queried data as XML. Oct 16 20:38:13 Signed-off-by: Andreas Oberritter Oct 16 20:38:13 Signed-off-by: Frans Meulenbroeks Oct 16 20:42:57 03Andreas Oberritter  07org.openembedded.dev * re23a8ff9c4 10openembedded.git/recipes/cdfs/ (2 files in 2 dirs): Oct 16 20:42:57 cdfs-2.6.27: initial recipe Oct 16 20:42:57 Signed-off-by: Andreas Oberritter Oct 16 20:42:57 Signed-off-by: Frans Meulenbroeks Oct 16 20:55:07 hm, what is the best way to enable an additional module in lighty? postinst with sed script on /etc/lighttpd.conf ? Oct 16 23:45:05 03Frans Meulenbroeks  07org.openembedded.dev * rf6fa929835 10openembedded.git/recipes/php/php_5.2.13.bb: Oct 16 23:45:05 php: version 5.2.13 removed --without-iconv Oct 16 23:45:05 5.2.13 had an inconsistency: there was a DEPENDS on virtual/iconv Oct 16 23:45:05 yet there was a configure argument --without-iconv Oct 16 23:45:05 as my project needed iconv, I removed the --without-iconv Oct 16 23:45:05 (works like a charm :-) ) Oct 16 23:45:05 Signed-off-by: Frans Meulenbroeks Oct 17 02:49:45 hey Oct 17 02:50:23 this overrides/filespath thing is yet another item highlighting how much shit just works by accident, or depends upon implementation details in such a way that the slightest change makes the world explode Oct 17 02:50:27 which makes me sad Oct 17 02:50:31 heh **** ENDING LOGGING AT Sun Oct 17 02:59:58 2010