**** BEGIN LOGGING AT Tue Feb 16 02:59:57 2010 Feb 16 03:03:53 is there a bitbake target that can be used to regenerate Packages, Packages.gz, etc? Feb 16 03:09:58 package-index Feb 16 03:11:25 denix, cheers :) Feb 16 03:15:22 goddamnit, external-toolchain-csl can bite me Feb 16 03:22:54 kergoth: I can bite you instead :) what's the problem? Feb 16 03:32:33 kergoth: ? Feb 16 03:33:04 keeps choosing it despite the preferences otherwise, acting. think it might be specific to master, might have to bisect. trying to confirm Feb 16 03:33:06 odd behavior. Feb 16 03:35:23 strange. is it in your deps tree? Feb 16 03:37:17 are you building an internal toolchain? it probably does not provide something, hence picking up external-toolchain-csl... Feb 16 08:19:35 I'm packaging tcl8.5. Its install target in the Makefile runs a built binary to "install message catalogs" Feb 16 08:19:40 What's the general approach to this in OE? Feb 16 08:20:42 i.e the Makefile directly calls "./tclsh $(TOOL_DIR)/installData.tcl" Feb 16 08:21:59 Is there a way to use the host tclsh perhaps? Feb 16 09:09:13 Hi, if I enable DEBUG_BUILD = "1" in locale.conf, it's sufficent to run "bitbake base-image" to enable debugging support for my whole distribution ? Feb 16 09:13:34 good morning Feb 16 09:15:12 hi, I built some toolchain with meta-toolchain and i get the build/tmp/deploy/sdk/minimal-.dev-snapshot-20100216-armv5te-linux-gnueabi-toolchain.tar.bz2 Shouldn't the archive name reflect what host the toolchain is build for? Feb 16 09:22:54 ao2: no not really... you shuldn't get device specific stuff into the toolchain. Therefore the name has the target architecture only. Feb 16 09:23:26 well, but the actual crosscompiler binaries are host, specific Feb 16 09:24:41 (forget the second comma) I mean I should be able to see from the archive name, where I am supposed to run the cross-compiler, am I missing anything? Feb 16 09:25:19 easy-comma today, sorry Feb 16 09:27:57 ao2: that's true - this would be a little bit cleverer Feb 16 09:28:56 adding HOST_ARCH to TOOLCHAIN_OUTPUTNAME in meta-toolchian.bb should be enough I'll propose a patch to oe-devel Feb 16 09:32:56 Hi, sorry I built base-image using OE.. but now if I enable DEBUG_BUILD = "1" in locale.conf, should I delete build/tmp directory ? or bitbake will notice the changes and rebuild the distribution with debugging enabled ? Feb 16 09:40:11 good morning Feb 16 09:44:03 vadmeste: I think if you want to go this way, you should wipe the tmp directory. But instead of enabling debug on the whole system you could just use the *-dbg packages generated by oe ? Feb 16 09:44:04 hi mckoan Feb 16 09:44:17 ao2: yes good point Feb 16 09:46:06 mlip2, it's okey, I don't know what should I do exactly.. I am trying to use oprofile for my kernel but I've got some errors I can't run it.. I'll return for further questions ;-) Feb 16 09:46:12 florian, canadian-sdk.bb is using SDK_SYS, I'll copy from there and see how it looks like Feb 16 09:48:01 ao2: yes that's because it isa ble to build sdks independent from the build host Feb 16 09:54:29 i686-linux-minimal-.dev-snapshot-20100216-armv5te-linux-gnueabi-toolchain.tar.bz2 that's better but maybe SDK_SYS looks even better just after DISTRO_VERSION Feb 16 09:55:12 Hello Feb 16 09:55:26 I am trying to cross compile dbus for my omap processor Feb 16 09:55:30 can anyone help me Feb 16 09:56:22 sharief_: you alread got answers for that in #beagle Feb 16 09:56:26 Hello Feb 16 09:56:32 no Feb 16 09:56:36 i didnt get the answer Feb 16 09:57:21 during configuration i am getting this error "checking for XML_ParserCreate_MM in -lexpat... no configure: error: Could not find expat.h, check config.log for failed attempts " Feb 16 09:58:05 i dont know what to do Feb 16 09:58:10 confused Feb 16 09:58:22 i struck with these problem from last 3 days Feb 16 09:59:45 what will be the problem? Feb 16 10:01:12 sharief_: did you check config.log? Feb 16 10:01:36 yes i checked Feb 16 10:01:43 in that what i get is Feb 16 10:02:23 conftest.c -lexpat >&5 /mnt/omap/toolchain/bin/../lib/gcc/arm-none-linux-gnueabi/4.4.1/../../../../arm-none-linux-gnueabi/bin/ld: cannot find -lexpat collect2: ld returned 1 exit status Feb 16 10:03:07 did you check in your staging directory to see if libexpat.so is actually there? Feb 16 10:03:16 if not, that would be a place to start. Feb 16 10:04:38 sharief@sharief-desktop:/mnt/omap/nfs-develop/usr/local/lib$ ls libexpat* libexpat.a libexpat.la libexpat.so libexpat.so.1 libexpat.so.1.5.2 Feb 16 10:04:45 i have that already Feb 16 10:05:36 sharief_: you are not using OE, are you ? Feb 16 10:05:55 :) Feb 16 10:05:59 ys i am using OE Feb 16 10:06:05 yes i Ma using OE Feb 16 10:06:43 strucked from last 3 days Feb 16 10:07:33 sharief_: in that case it sounds as though the linker is looking in the wrong place for some reason. if you run it by hand with --verbose, what does it say it is using as the SEARCH_DIR? Feb 16 10:07:34 sharief, I don't think you are using OE, this is not a path you'd see in OE: Feb 16 10:07:42 * pb___ afk now Feb 16 10:07:43 /mnt/omap/toolchain/bin/../lib/gcc/arm-none-linux-gnueabi/4.4.1/../../../../arm-none-linux-gnueabi/bin/ld: Feb 16 10:08:36 actually i am cross compiling the things Feb 16 10:08:51 OE is for cross compiling and you need to set it up properly Feb 16 10:09:11 pb__:you mean to add --verbose in ./configure? Feb 16 10:09:38 sharief_: XorA already told you in #beagle Feb 16 10:09:41 eFfeM-work: sorry i didnt understand whta you mean Feb 16 10:10:03 sharief_: have you read this: http://wiki.openembedded.org/index.php/Getting_Started Feb 16 10:10:05 eFfeM-work:he only told me to consult in this OE irc Feb 16 10:10:22 sharief_: and have you done this: http://www.angstrom-distribution.org/building-angstrom ? Feb 16 10:11:25 why i should do that? i am cross compilaing in my host system Feb 16 10:11:36 I don't think he suggested consulting this irc: http://www.pastebin.ca/1798193 Feb 16 10:12:28 yes see,XorA: we normally suggest OE to cure cross compile headaches Feb 16 10:12:32 sharief_: one more time, this is an openembedded developer channel, if you do not want to use openembedded as cross compiling environment you are asking your questions in the wrong channel Feb 16 10:12:56 sharief_: he meant using OE Feb 16 10:13:28 this is not an omap cross compile support channel, omap questions are better off in #beagle Feb 16 10:13:31 this is great, I employ eFfeM-work as my new PR guru Feb 16 10:13:46 rofl Feb 16 10:13:53 oh i see Feb 16 10:14:04 ok fine Feb 16 10:14:05 XorA: had to wait until a git clone to finish Feb 16 10:14:08 * XorA did actually make a suggestion abou the non OE case as well on #beagle Feb 16 10:15:50 good morning pb___ Feb 16 10:30:25 03Martin Jansa  07org.openembedded.dev * r8179a444c0 10openembedded.git/recipes/kexecboot/ (5 files in 5 dirs): Feb 16 10:30:25 linux-kexecboot: move from 2.6.33-rc8 to 2.6.33-rc8, enable DEVTMPFS Feb 16 10:30:25 Signed-off-by: Martin Jansa Feb 16 10:30:26 03Martin Jansa  07org.openembedded.dev * re9634bcd7f 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Feb 16 10:30:26 linux-2.6.33-rc6: move to linux-2.6.33-rc8 Feb 16 10:30:26 Signed-off-by: Martin Jansa Feb 16 10:30:26 03Martin Jansa  07org.openembedded.dev * r2b914d1c7f 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs: bump shr-launcher for app rescan fix Feb 16 10:30:27 03Martin Jansa  07org.openembedded.dev * r3a090cd3e1 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Feb 16 10:30:27 sane-srcrevs-fso: bump mdbus2 for latest fixes Feb 16 10:30:27 Signed-off-by: Martin Jansa Feb 16 10:30:29 03Martin Jansa  07org.openembedded.dev * rbf450860bf 10openembedded.git/recipes/kexecboot/ (7 files): (log message trimmed) Feb 16 10:30:29 linux-kexecboot: use INC_PR for kernels, PR should get bump with every kexecboot change Feb 16 10:30:29 * Andrea Adami wants only one INC_PR shared between kexecboot and Feb 16 10:30:29 linux-kexecboot recipes, maybe we could use sth like Feb 16 10:30:30 INC_PR = "${KEXECBOOT_PR}.19" in linux-kexecboot.inc Feb 16 10:30:34 and set KEXECBOOT_PR in some new .inc file Feb 16 10:30:35 * For now use INC_PR only for linux-kexecboot Feb 16 10:35:04 morning all Feb 16 10:35:23 * RP has just proposed a new way for bitbake to find configuration files on the OE/bitbake lists Feb 16 10:35:37 It should make overlay handling much nicer Feb 16 10:36:22 mickeyl: good morning Feb 16 10:38:47 http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-February/017428.html Feb 16 11:07:00 RP: your proposal sounds good Feb 16 11:09:39 * mckoan was preparing his new configuration set with overlays ;-) Feb 16 11:21:55 -> bbl Feb 16 11:50:12 pb_: Any opinion on the proposal? Feb 16 11:50:22 XorA, mickeyl: Likewise :) Feb 16 11:52:32 RP: ?? Feb 16 11:52:43 XorA: http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-February/017428.html Feb 16 11:53:14 RP: working in my sysadmin job today havent caught up with that yet Feb 16 11:54:03 XorA: np, I'm just feeling keen ;-) Feb 16 11:54:25 Now I've actually written a patch for that problem - its been a long time in coming :/ Feb 16 11:57:28 looks interesting Feb 16 12:09:44 03Thomas Zimmermann  07org.openembedded.dev * rf34fc00871 10openembedded.git/recipes/tangogps/ (tangogps.inc tangogps_0.99.3.bb): Feb 16 12:09:44 tangogps: add version 0.99.3 Feb 16 12:09:44 Signed-off-by: Martin Jansa Feb 16 12:13:33 I wonder, why bitbake always rebuilds cache if I change something in recipe in overlay Feb 16 12:14:30 it's standard behaviour? Feb 16 12:18:59 RP: explanation looks good, cant test it now Feb 16 12:29:10 ynezz: depends what you change but it should always notice Feb 16 12:29:30 XorA: np, I'm most interested in whether the idea works in theory :) Feb 16 12:30:32 I think layer.conf should be uneeded in a default case where just add overlay/recipes/* to bbpath Feb 16 12:31:39 XorA: How do you make that work without hardcoding assumptions into bitbake? Feb 16 12:33:13 it just seems to be adding lots of identicle files for the simple case :-( Feb 16 12:34:51 XorA: On the other hand a two line file describing what an an overlay contains isn't such a bad thing? Feb 16 12:35:39 just people cant get BBPATH right first time :-) Feb 16 12:36:30 XorA: This would most likely avoid people ever having to deal with BBPATH, other than copy another layer.conf ;-) Feb 16 12:50:34 morning Feb 16 13:10:24 hrw: gm Feb 16 13:11:25 hi marco Feb 16 13:11:32 mckoan: how goes new board? Feb 16 13:12:35 hrw: fine thx, and your one? Feb 16 13:12:49 only booted not tested anything on it yet Feb 16 13:39:16 RP: What do you mean by 'notice' ? Cache rebuild? It's quite time consuming if you change something in recipe like kernel, bumps PR and it rebuilds cache from scratch Feb 16 13:39:55 RP: is it possible to find out somehow what's causing the rebuild? Feb 16 13:49:14 hi folks need libstdc++ on verdex how can I get it ? Feb 16 13:49:32 the verdex compiler makes the binary fine Feb 16 13:50:01 actually the nastiest thing I find the cache rebuild if you accidently hit ^C while still parsing Feb 16 13:53:36 eFfeM-work: o yes Feb 16 14:05:45 ynezz, touching a single recipe should only be resulting in a reparse of that recipe. Feb 16 14:09:43 I think so Feb 16 14:09:59 but it isn't and I don't know why Feb 16 14:36:37 ok, have a nice est of day Feb 16 15:01:11 RP: whenever you have a chance, if you would, please revert ee9afccf33b220a21b74fab279925eeb4771249b in bitbake master. it hoses obeying of preferred providers, I didn't test it sufficiently before applying their patch, apparently. Feb 16 15:01:21 RP: macbook is out of commission atm, don't have my ssh key :) Feb 16 15:26:12 pb_: master? Feb 16 15:27:38 zecke: at your service Feb 16 15:28:37 pb_: awesome. I have something i need a hint. I have a Unix Domain socket. I connected to it. And I'm trying to do something like write(fd, &len, sizeof(len)); write(fd, data->data, data->len)); Feb 16 15:29:09 pb_: and at the other end I doa read(fd, len, sizeof(len)); read(fd, data, len); (actually I write it with network order and translater properly) Feb 16 15:29:45 pb_: and somehow I do read wrong stuff out of the socket. would you have a wild guess? Feb 16 15:31:18 zecke: hm, nothing obviously comes to mind. are you checking the return values from read() and write()? Feb 16 15:32:20 is this a stream socket or a datagram socket? Feb 16 15:33:35 pb_: stream, yeah. I check short write and short read... and it does not occur Feb 16 15:35:34 hm, strange. what exactly is wrong with the data you receive? Feb 16 15:37:10 kergoth: Ah, I did wonder about that patch Feb 16 15:37:48 didn't seem too horrible at first glance, but.. *shrug*, definitely broken right now :) Feb 16 15:37:48 ynezz: look at the debug logs from bitbake - its verbose about why it rebuilds Feb 16 15:38:54 kergoth: reverted Feb 16 15:39:00 thanks. Feb 16 15:40:31 pb_: the receive has a value it can not have... and a value that was not at the end of the other data... Feb 16 15:43:57 pb_: and on top of that... i just verified that i only write from the method i should write into.. Feb 16 15:46:11 zecke: very strange. but, sorry, I have no idea what might be wrong there. Feb 16 15:51:01 pb_: thanks. I will probably read everything... dump it to a file and see what is going on..w Feb 16 15:51:31 right, good idea Feb 16 15:52:33 Hello Feb 16 15:52:45 I want to know some basic things Feb 16 15:52:48 can anyone help me Feb 16 15:53:12 whether i can use OE for cross compilaing all the packages for OMAP processor? Feb 16 15:53:25 is it easy? Feb 16 15:54:19 I have to do. tell me guys Feb 16 15:54:42 hawk_: yes, OE can cross compile for arm cpus Feb 16 15:56:11 but, where is the details about cross compilaing packages. I didnt see anywhere Feb 16 15:56:44 hawk_: You would have to convince it _not_ to compile cross Feb 16 15:57:22 Sorry i didnt understand Feb 16 15:57:59 hawk_: OE usually compiles everything cross. Feb 16 15:58:20 Before this it build the toolchain to do this. Feb 16 15:58:50 What you have to do is to tell OE about your target devices in order to select the correct settings for it. Feb 16 15:58:54 oh i see Feb 16 15:59:15 ok where to select the packages? Feb 16 15:59:19 OE is used quite a lot for building for OMAP, in fact its the basis for the TI PSP Feb 16 16:00:12 XorA,Oh i see. good Feb 16 16:00:39 * XorA has OE built gnome desktop on all 3 of my omaps :-) Feb 16 16:01:08 ok fine Feb 16 16:01:33 sounds like we sold 10 copies to hawk_ :-) Feb 16 16:02:05 XorA, Ha ha ha Feb 16 16:59:31 Hi, I have this error when trying to build oprofile : http://paste.pocoo.org/show/178877/ ! can anyone help me ? Feb 16 18:25:37 kergoth, RP, the theory behind ee9afccf33b220a21b74fab279925eeb4771249b was that a distro could set PREFERRED_PROVIDER_virtual/foo and have that picked. Without that patch that didn't work for me? Feb 16 18:59:46 kergoth, RP, on a second thought it's odd that i needed it. Sorry for the breakage |( Feb 16 19:02:59 btw.. where are the bitbake CIA messages in #here (but then my VT may be blind)? and (since i see mixed style) does bitbake use the 'file.ext:' or the '[file]' subject prefix? Feb 16 20:45:04 03Koen Kooi  07org.openembedded.dev * r5fa112f983 10openembedded.git/classes/kernel.bbclass: kernel.bbclass: prepare for 2.6.33 header move and simplify logic a bit Feb 16 21:24:20 re Feb 16 21:24:51 guys, is it possible to enable additional debug output for the patch stage? -DDD is not enough, I'd like to see the actual command being run and to know the path from where it is running it Feb 16 21:25:19 have a very, very strange case where the patch applies manually but no matter what I do fails via the "default" way Feb 16 22:04:31 jo Feb 16 22:06:03 hi woglinde Feb 16 22:06:26 03Tom Rini  07org.openembedded.dev * r09f94fefc6 10openembedded.git/recipes/openswan/ (openswan_2.2.0.bb openswan_2.4.7.bb): openswan: Set USERLINK=${LDFLAGS} to fix GNU_HASH errors, bump PR Feb 16 22:43:59 03Andrea Adami  07org.openembedded.dev * r55004d44d3 10openembedded.git/recipes/images/initramfs-kexecboot-image.bb: initramfs-kexecboot-image: standardize image (remove PR, reorder) Feb 16 22:44:10 03Andrea Adami  07org.openembedded.dev * r4dd2c2666c 10openembedded.git/conf/machine/include/initramfs-kexecboot.inc: initramfs-kexecboot.inc: apply sane defaults Feb 17 00:14:13 hmm, i should revisit the autotools/autotools_base branch bits i was playing with Feb 17 00:16:11 configure: creating ./config.status Feb 17 00:16:11 ./config.status: line 1: :t: command not found Feb 17 00:16:12 hmm. Feb 17 00:16:13 that's a new one. Feb 17 00:16:17 anyone run into this one before? Feb 17 00:29:38 kergoth did you examine the config.status? Feb 17 00:34:13 sounds like a regex goof Feb 17 00:34:23 git log -p, search on :t ? :) **** ENDING LOGGING AT Wed Feb 17 02:59:58 2010