**** BEGIN LOGGING AT Fri Mar 02 03:00:00 2007 Mar 02 04:58:17 Hello all Mar 02 04:58:56 hey likewise Mar 02 04:59:00 hey hvontres|home Mar 02 04:59:35 Can someone look at this http://pastebin.com/891964 Mar 02 05:00:04 I understand the error but I can not find the Makefile that defines that -DVERSION variable Mar 02 05:00:35 That sounds familiar.. which package? Mar 02 05:01:06 apmd Mar 02 05:01:34 the gcc command has -DVERSION=\"Sun JDK 1.4.2.10\" Mar 02 05:01:44 which I think is wrong because they are escaping the " in the statement Mar 02 05:01:57 causing the shell to parse as seperate strings, not one Mar 02 05:02:56 Hmm Mar 02 05:03:21 03mwester 07org.oe.dev * r16412457... 10/ (1 conf/distro/unslung.conf): Mar 02 05:03:21 unslung.conf - change mkfs.jffs2 flag to -n from --no-eraseblock-headers to Mar 02 05:03:21 accomodate the long-option name change with the update mtd-utils package. Mar 02 05:03:51 cyrus__: It could be make got this variable passed through over its command line, and not from a Makefile. Mar 02 05:04:02 Looking at the .bb file, it doesn't look like it's an autoconfed package.. Mar 02 05:04:29 likewise - okay, and what would pass these variables through the command like Mar 02 05:05:04 cyrus__: Take a look at apmd-3.2...orig/Makefile under XTRACFLAGS Mar 02 05:05:06 usually variables like EXTRA_CFLAGS are passed on, in this wat Mar 02 05:05:14 way Mar 02 05:06:00 NAbyss_ - i see it Mar 02 05:06:11 NAbyss_ beat me by a second *and* looked into the sources whereas I just guessed :-) Mar 02 05:06:18 cyrus__: What happens if you remove the \ and bitbake it? Mar 02 05:06:28 Hehe Mar 02 05:06:38 HopsNBarley: Good evening, are you there? Mar 02 05:06:48 trying now Mar 02 05:07:01 cyrus__: I can see why they should be there.. I'm just confused as to why they're being pushed through to the shell Mar 02 05:07:08 but this is my problem with oe, every thing i try to build just fails. Never once have I had it successfully build anything Mar 02 05:07:35 cyrus__: bitbake ampd -c clean first, so that the faulty sources are not re-used Mar 02 05:08:04 okay..it made it past, I didn't remove the \ I put \"" instead Mar 02 05:08:09 and it built that package Mar 02 05:08:16 moving onto the next Mar 02 05:09:14 One thing that comes to mind.. are you using GNU make for it, or a different toolchain? Mar 02 05:09:24 the weird thing was, at the top of the make file VERSION was defined at 3.2.1 and in that XTRACFLAGS it was set -DVERSION=\"$(VERSION)\" so I am not sure how VERSION got replaced with Sun JDK 1.4.2 Mar 02 05:09:48 cyrus__: make -e prioritises upstream (environment) variables over those defined in the makefile Mar 02 05:10:14 cyrus__: Might be worth starting a shell with a clean environment Mar 02 05:44:55 morning all Mar 02 06:41:07 good morning here Mar 02 07:04:31 hello Mar 02 07:05:56 I'm trying to compile for "sharprom-compatible" and I put cross-2.95.3 in /usr/local/arm. What am I renaming to arm-linux-gcc-2.95? Mar 02 07:21:11 morning Mar 02 07:21:20 ~zauruskernels Mar 02 07:21:27 somebody said zauruskernels was by default Zaurus used 2.4.18 or 2.4.20 - now we have 2.6 working on all clamshells and work is progressing for tosa, poodle, collie. More info in OpenZaurus wiki: http://wiki.openzaurus.org/Kernel_Information Mar 02 07:21:37 Glombool: read that Mar 02 07:21:55 thanks Mar 02 07:23:04 hrw: read the kernel_information? Mar 02 07:28:20 yes Mar 02 07:28:32 there are links to 2.4 situation Mar 02 07:29:09 oe for MIPS ? Mar 02 07:29:24 crink: mips is supported Mar 02 07:29:54 hrw: i set db1200 and generic Mar 02 07:30:34 it failed glibc_2.3.5 Mar 02 07:32:00 how failed? Mar 02 07:32:10 | In file included from ../sysdeps/mips/libc-tls.c:21: Mar 02 07:32:10 | ../sysdeps/generic/libc-tls.c: In function ‘__libc_setup_tls’: Mar 02 07:32:10 | ../sysdeps/generic/libc-tls.c:191: warning: implicit declaration of function ‘INTERNAL_SYSCALL_DECL’ Mar 02 07:32:10 | ../sysdeps/generic/libc-tls.c:191: error: ‘err’ undeclared (first use in this function) Mar 02 07:32:10 | ../sysdeps/generic/libc-tls.c:191: error: (Each undeclared identifier is reported only once Mar 02 07:32:11 | ../sysdeps/generic/libc-tls.c:191: error: for each function it appears in.) Mar 02 07:32:13 | ../sysdeps/generic/libc-tls.c:191: warning: implicit declaration of function ‘INTERNAL_SYSCALL’ Mar 02 07:32:15 | ../sysdeps/generic/libc-tls.c:191: error: ‘set_thread_area’ undeclared (first use in this function) Mar 02 07:32:17 | ../sysdeps/generic/libc-tls.c:191: warning: implicit declaration of function ‘INTERNAL_SYSCALL_ERROR_P’ Mar 02 07:32:23 OE Build Configuration: Mar 02 07:32:23 BB_VERSION = "1.6.6" Mar 02 07:32:25 OE_REVISION = "7ecf0e32974264619fa996b2de3f674cfe2ef705" Mar 02 07:32:27 TARGET_ARCH = "mipsel" Mar 02 07:32:29 TARGET_OS = "linux" Mar 02 07:32:31 MACHINE = "db1200" Mar 02 07:32:33 DISTRO = "generic" Mar 02 07:32:35 DISTRO_VERSION = ".dev-snapshot-20070302" Mar 02 07:32:37 TARGET_FPU = "" Mar 02 07:32:37 ~pastebin Mar 02 07:32:40 somebody said pastebin was a place to paste your stuff without flooding the channel - try http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/, or http://pastebin.com is usually painfully too slow and unresponsive to use, use one of the other pastebin sites, or dpaste.com is a very nice pastebin as well Mar 02 07:32:42 sorry Mar 02 07:32:59 that' all Mar 02 07:33:00 tried to goodle bug? Mar 02 07:33:11 what is that? Mar 02 07:33:49 * crink considering trying cross-lfs on oe partly.. Mar 02 07:33:56 glib-2.4 part Mar 02 07:35:53 tried to find that bug using google? Mar 02 07:36:04 I'm a bit busy now ;( Mar 02 07:36:12 of course i do Mar 02 07:36:51 i can handle internal of oe, so be it. Mar 02 08:12:04 hi zecke greentux Mar 02 08:12:30 hi hrw... money is on the way Mar 02 08:14:21 greentux: great Mar 02 08:14:21 thx Mar 02 08:24:23 morning Mar 02 08:24:28 !oebug 1898 Mar 02 08:24:30 * * Bug 1898, Status: NEW, Created: 2007-02-21 03:09 Mar 02 08:24:31 * * oliver.fels(AT)gmx.net: libts patches for OpenMoko break touchscreen handling Mar 02 08:24:32 * * http://bugs.openembedded.org/show_bug.cgi?id=1898 Mar 02 08:24:40 morning all Mar 02 08:25:10 ahah, 1898 is the bug I see Mar 02 08:25:16 ~lart bad openmoko Mar 02 08:25:16 * ibot hits bad openmoko with an anvil and laughs with a contralto voice ... Haha Ha HA Ha Mar 02 08:59:38 morning all Mar 02 09:01:56 hi dirk Mar 02 09:05:12 hey Marcin Mar 02 09:15:06 hi liam Mar 02 09:17:07 hi Marcin, morning all Mar 02 09:19:47 hi Liam Mar 02 09:25:16 hi Liam Mar 02 09:31:14 koen: your latest proposal gets me vote Mar 02 09:31:21 s/me/my Mar 02 09:32:50 ade|desk: but I dare not commit it within a week Mar 02 09:33:03 since people might want to comment on it. *slowly* Mar 02 09:36:43 you mean wait 12 mins and then do it ;) Mar 02 09:37:14 koen: you missed the ipkg-make-index thinks X.X+cvs < X.X Mar 02 09:37:27 right Mar 02 09:38:07 I still think it's insulting and disrepectfull to users to use deploy/ipk as a proper feed when knowing the problems Mar 02 09:38:23 even *I* wouldn't do that Mar 02 09:39:52 :-D Mar 02 09:41:16 and coredump can have his commit access back when he stops being an ass Mar 02 09:41:41 03daniel 07org.oe.dev * rfea55c0e... 10/ (1 files/device_table-ezx.txt files/device_table-minimal.txt): device tables: document format and add a table for motorola ezx phones Mar 02 09:41:46 03koen 07org.oe.dev * r3221506b... 10/ (1 conf/machine/a1200.conf conf/machine/a780.conf): a780, a1200: remove duplication and use new device tables Mar 02 09:41:51 03daniel 07org.oe.dev * rab801558... 10/ (3 files in 3 dirs): opentapi: start up real early to initialize the BP properly to stop the phone shutting of because udev is so slow Mar 02 09:41:59 drat, I missed a --author on the machine.confs Mar 02 09:42:41 the module_do_compile function uses a MAKE_TARGETS variable. However when I try to set it into a do_compile_prepend function for a module recipe I get an error telling me MAKE_TARGETS was not found. Mar 02 09:43:10 where is MAKE_TARGETS defined? Mar 02 09:43:20 that's for a module where I need to use (short way): make arm Mar 02 09:43:53 koen: in do_compile_prepend Mar 02 09:44:01 you could try 'make ${TARGET_ARCH}' Mar 02 09:45:10 koen: yes, that was my first guess, but I would like to keep the "genercity" that I get from the module_do_compile function Mar 02 09:45:26 s/genercity/genericity Mar 02 09:46:01 but sure I could rewrite a do_compile function Mar 02 09:46:25 the thing is just that affection MAKE_TARGETS would be easier and cleaner Mar 02 09:53:24 Ifaistos: the ieee_sqrt unresolved symbols might be due to using -Os instead of -O2 Mar 02 09:57:08 ade|desk: we did discover a new zaurus model at fosdem Mar 02 09:57:13 the "sharp tosser" Mar 02 09:58:37 03koen 07org.oe.dev * r2fe55428... 10/ (1 conf/distro/include/angstrom-glibc.inc): angstrom: don't use -Os on sparc, glibc won't compile due to some libgcc weirdness Mar 02 10:02:09 Hi! is there a task to rebuild Packages.gz files only? angstrom-bootstrap-image does rebuilding them, but that's too much overhead Mar 02 10:02:18 gena_: package-index Mar 02 10:03:12 that is useful information! :) Mar 02 10:03:32 I missed it somehow too here http://www.openembedded.org/wiki/UsefulTargets Mar 02 10:07:20 then add it? Mar 02 10:12:32 hrw: it's already there - I just wasn't able to read it whole :D Mar 02 10:12:54 who the hell reads all that guides? :b Mar 02 10:16:16 here is my solution for rt2570 arm module build: http://www.pastebin.ca/378222 It's ugly but it is because the original driver needs EXTRA_CFLAGS for arm, and you need to call make arm Mar 02 10:16:25 good morning Mar 02 10:17:49 ljp: got a second? Mar 02 10:17:59 koen: glibc-intermediate compiles! Mar 02 10:18:23 oxo: NOTE: package glibc-2.5: completed (for sparc) Mar 02 10:18:38 oxo: I think it's a libgcc problem Mar 02 10:18:46 I'm 85% sure Mar 02 10:18:57 vlo: moment Mar 02 10:18:57 euhm, but I only recompiled glibc Mar 02 10:19:11 I did the same and it didn't work :( Mar 02 10:19:21 with FULL_OPTIMIZATION_pn-glibc BUILD_OPTIMIZATION_pn-glibc FULL_OPTIMIZATION_pn-glibc-intermediate and BUILD_OPTIMIZATION_pn-glibc-intermediate Mar 02 10:19:42 since I'm building for supersparc cq sparc, I was not sure to do _supersparc or _sparc Mar 02 10:19:54 and the iterations are quite lengthy Mar 02 10:20:08 what did you do in order to get it to compile? Mar 02 10:20:24 FULL_OPTIMIZATION_sparc = "-fexpensive-optimizations -fomit-frame-pointer -frename-registers -O2" Mar 02 10:20:36 since I use MACHINE="sum4cdm" Mar 02 10:20:45 I can fix supersparc later :) Mar 02 10:20:47 weird Mar 02 10:21:15 vlo: http://pastebin.ca/378228 Mar 02 10:21:17 I have that, but with Os instead of O2, and in the mentioned FULL_/BUIL_* O2 again Mar 02 10:21:35 so everything with Os, instead of glibc and friends Mar 02 10:21:53 but then we're doing the same for the failing part Mar 02 10:23:07 btwm, I think you can cram Xfbdev into that 8MB Mar 02 10:23:07 vlo: test it Mar 02 10:23:12 -m Mar 02 10:23:17 and what's the preferred way? BUILD_OPT_yaddi_sparc in conf/distro/* or BUILD_OPT_yaddi in confi/machine/krups ? Mar 02 10:23:28 or none of the above? Mar 02 10:23:35 I'm not sure Mar 02 10:23:52 BUILD_OPT is policy and should be in distro, but it makes a machine fail Mar 02 10:24:27 koen: do you have arm machine ON? Mar 02 10:24:35 hrw: yes, multiple Mar 02 10:24:46 you can choose between pxa255 and px270 atm Mar 02 10:25:24 ok, make that 1 Mar 02 10:26:30 ok confirmed, the openmoko "fix" to tslib in fact breaks tslib for at least zaurus machines Mar 02 10:26:33 hrw: what do you want to test? Mar 02 10:26:56 XorA: hmmm, it works on my ipaqs, is it due tskeys trying to access it as well? Mar 02 10:26:58 koen: had one binary. but c7x0 started so I will check locally Mar 02 10:27:20 koen: I'll put it in conf/machine/krups, then Mar 02 10:27:33 koen: /dev/input/* are supposed to be multi open capable, once tskeys opens the device, tslib prevents anything else doing so after that change Mar 02 10:27:37 in that way it does not matter if one decides to do supersparc instead of sparc Mar 02 10:28:07 koen: it was the bug that I found on saturday morning at FOSDEM when my Z suddenly didnt work :-) Mar 02 10:31:49 hrw: so many drivers for the same chip ... which version is the one you gave right now ? Mar 02 10:33:13 vlo: vendor one Mar 02 10:33:21 vlo: did not tested it Mar 02 10:33:51 I can confirm, ts devices are meant to be multiple open Mar 02 10:34:11 hrw: ok ... so it's supposed to be another recipe than rt2570 to be coherent ? Mar 02 10:34:36 vlo: that one builds for arm Mar 02 10:35:47 hrw: the other one did also build for arm and could be loaded .. anyway, giving this one a test Mar 02 10:48:56 QA error for prismstumbler: http://rafb.net/p/7MIsO753.html Mar 02 10:49:12 The gpsd path doesn't look like a host path for me. Mar 02 10:49:49 hrm, sounds like i need to work on this toy a little bit more :) Mar 02 10:50:45 florian: If you mean a780, the answer is yes. :) Mar 02 10:53:14 hrw: there are two different drivers: http://www.ralinktech.com/ralink/Home/Support/Linux.html I'm really using rt2570, you're using rt2571. see here: http://ralink.rapla.net/ Mar 02 10:54:16 wasn't there something in bugzilla for ralink drivers? Mar 02 10:54:55 stefan_schmidt: this time it was related to prismstumbler, but "luckily" it looks like a bug in QA Mar 02 10:56:09 So workaround and open a bug report about it? Mar 02 10:56:43 the gpsd configure probably has -I/usr/include somewhere Mar 02 10:57:42 koen: That would mean qa would shout if I recompile gpsd, right? Mar 02 10:57:59 ehm Mar 02 10:58:03 it should Mar 02 10:58:08 * stefan_schmidt tries Mar 02 10:58:23 but there's a bug in QA that fails to notice the error if you try a second time Mar 02 10:58:35 koen: you're right, there's a bug report for exactly what I've done for rt2570 recipe Mar 02 10:59:17 koen: oh, ok. Mar 02 10:59:53 koen: I read up how to diable qa for this package and open a bugreport for it. Mar 02 11:00:26 fwiw, I think prismstumbler should link against the system gpsd Mar 02 11:00:31 this just wastes space Mar 02 11:01:02 NOTE: Checking sanity of the config.log file Mar 02 11:01:02 NOTE: package gpsd-2.34-r0: task do_qa_configure: completed Mar 02 11:01:20 But this can be about the second build bug Mar 02 11:01:33 well Mar 02 11:01:42 and the implementation is nicer than my try ... but they were able to affect MAKE_TARGETS, like my first try ... looks like you cannot affect MAKE_TARGETS in do_compile_prepend() but you can in do_compile() Mar 02 11:01:44 the qa bug is about prismsstumbler's copy of gpsd Mar 02 11:04:13 koen: hmm, so what's the preferred way to deal with this? Mar 02 11:04:28 spank florian to not include a broken gpsd Mar 02 11:05:51 koen: I prefer polite suggestions. :) Mar 02 11:08:28 !oebug 1711 Mar 02 11:08:30 * * Bug 1711, Status: NEW, Created: 2006-12-22 10:15 Mar 02 11:08:31 * * jack-bugs-openembedded-org(AT)mudshark.org: Small update to packages/rt2x000/rt2570_1.1.0-b2.bb to support arm Mar 02 11:08:32 * * http://bugs.openembedded.org/show_bug.cgi?id=1711 Mar 02 11:09:01 the above bug could attachment could be pushed in (working for me) Mar 02 11:11:03 florian: Can you have a look at the prismstumbler / gpsd problem? Mar 02 11:11:33 koen : hmmm strange Mar 02 11:12:46 koen : Does gcc for efika build if you changed it ? Mar 02 11:26:56 03xora 07org.oe.dev * r661f7545... 10/ (3 files in 3 dirs): Mar 02 11:26:56 prism3-support_1.0.0.bb : correct thinko/typo in hostap_fw_load which prevented prism3 firmware ever loading unless we werent a wireless card. Also thanks to Mar 02 11:26:56 pH5 snuck in a small optimisation. Mar 02 11:28:12 hmm, is there a 'mtn phrase' which shows me which new files live in de working copy? Mar 02 11:28:31 'mtn status' only shows known files in de WC Mar 02 11:28:52 mtn list unknown Mar 02 11:28:57 re Mar 02 11:29:04 tnx hrw Mar 02 11:32:06 stefan_schmidt: i'll try... but i won't promise that i manage to do this today. i have to catch up at work first... Mar 02 11:33:13 florian: OK, I disabled QA for now. Should I put it into a bugreport as reminder? Mar 02 11:33:54 who is the maintainer of the openssl packages? Mar 02 11:34:03 * oxo has a patch to make it work on (super)sparc Mar 02 11:34:04 http://rafb.net/p/Zt9B4f17.html Mar 02 11:34:37 stefan_schmidt: i'll try to keep it in mind :-) Mar 02 11:34:43 shall I put that in bugzilla? Mar 02 11:34:50 florian: fine with me Mar 02 11:35:08 oxo: sute Mar 02 11:35:09 ehm Mar 02 11:35:10 sure Mar 02 11:42:00 !oebug 1927 Mar 02 11:42:01 * * Bug 1927, Status: NEW, Created: 2007-03-02 04:20 Mar 02 11:42:02 * * jorik(AT)kippendief.biz: Add (super)sparc target to OpenSSL Mar 02 11:42:03 * * http://bugs.openembedded.org/show_bug.cgi?id=1927 Mar 02 11:42:07 :-) Mar 02 11:43:32 hrw, I added a 2.6.20 for the OSK Mar 02 11:43:42 I saw that Mar 02 11:43:59 I boots, and I converted the file system to angstrom Mar 02 11:52:15 great Mar 02 12:09:22 just updated my oe checkout and libx11 is now failing to compile (cross compile badness) Mar 02 12:09:23 http://bugs.openembedded.org/show_bug.cgi?id=1928 Mar 02 12:10:49 tkp: now look at config.log and figure out why it did include /usr/include/X11/Xtrans? Mar 02 12:13:24 03xora 07org.oe.dev * rdf7aec92... 10/ (6 files in 3 dirs): Mar 02 12:13:24 mplayer_0.0+1.0rc1.bb : add start of the pxafb overlay driver from Tim Chick Mar 02 12:13:24 This is a good start to fully working overlay video. Mar 02 12:14:05 well, I'm not too sure _why_, but it has: Mar 02 12:14:08 pkg_cv_X11_CFLAGS='-D_BSD_SOURCE -DHAS_FCHOWN -DHAS_STICKY_DIR_BIT -I/usr/include/X11/Xtrans -I/home/sts/trunk/oe-sts/build/tmp-stc/staging/i686-linux/include ' Mar 02 12:14:19 which looks wrong Mar 02 12:15:08 tkp: see above? Mar 02 12:42:20 what is it that generates the pkg_cv_X11_CFLAGS variable? Mar 02 12:42:41 PKG_CHECK_MODULE or similiar Mar 02 12:43:50 in what packet would I find udevinfo ? because it it not included with udev Mar 02 12:49:39 vlo: udev-utils? Mar 02 12:51:32 hrw: doesn't exist in my OE tree Mar 02 12:51:44 vlo: its part of udev recipe Mar 02 12:52:09 hrw:ok going to have a look Mar 02 12:54:15 so... what, it pulls the info out of a .pc file? Mar 02 12:54:20 like... oe-sts/build/tmp-stc/staging/i586-sts-linux/share/pkgconfig/xtrans.pc Mar 02 12:56:00 just, whats in that .pc file doesn't seem to tally up with what config.log has Mar 02 12:57:57 tkp: welcome to autoconf Mar 02 13:00:45 ~lart gstreamer people a bit Mar 02 13:00:45 * ibot pours hot grits down the front of gstreamer people a bit's pants Mar 02 13:01:44 ok, well clearly it's looking at the wrong pkg-config stuff Mar 02 13:02:05 since it's got .../staging/i686-linux/include in there Mar 02 13:02:15 i686-linux is my host machine, not the target Mar 02 13:13:49 ok... well, a fix is: Mar 02 13:13:49 export PKG_CONFIG_PATH=${STAGING_LIBDIR}/../share/pkgconfig Mar 02 13:13:59 but surely I shouldn't have to do that Mar 02 13:14:25 tkp: feel free to paste the .pc content Mar 02 13:14:56 tkp: also PKG_CONFIG_PATH should be set, or at least pkg-config it self has a default path it looks into Mar 02 13:21:49 ok. http://rafb.net/p/eZJnP359.html Mar 02 13:22:00 the top bit is what I get without altering the bb file Mar 02 13:22:20 the bottom is what I get after setting PKG_CONFIG Mar 02 13:22:28 it's always looking for i586-sts-linux-pkg-config Mar 02 13:22:34 which it doesn't find.. Mar 02 13:22:42 is this supposed to exist? Mar 02 13:26:45 tkp: okay, that is interesting now I need to remember Mar 02 13:26:54 koen: in which directory do we look for pkg-config files? Mar 02 13:27:08 tkp: we might have a regression here Mar 02 13:27:09 also... right before do_configure, when it prints out the env, I can see Mar 02 13:27:10 export PKG_CONFIG_PATH=/home/sts/trunk/oe-sts/build/tmp-stc/staging/i586-sts-linux/lib/pkgconfig Mar 02 13:27:21 so, why isn't that being passed through Mar 02 13:30:13 zecke: {libdir}/pkgconfig Mar 02 13:30:50 koen: so in tkp's xtrans.pc is in the right place Mar 02 13:31:24 and anyhow... none of the xtrans.pc files contain a reference to /usr/include/X11/Xtran Mar 02 13:31:44 I have 4 versions Mar 02 13:31:56 /staging/i586-sts-linux/lib/pkgconfig/xtrans.pc Mar 02 13:32:00 /staging/i586-sts-linux/share/pkgconfig/xtrans.pc Mar 02 13:32:07 /staging/i686-linux/lib/pkgconfig/xtrans.pc Mar 02 13:32:14 and /staging/i686-linux/share/pkgconfig/xtrans.pc Mar 02 13:32:52 ah Mar 02 13:32:53 wait Mar 02 13:33:01 zecke: no idea, I'm busy stabbing matlab Mar 02 13:33:02 this one Mar 02 13:33:03 staging/i586-sts-linux/lib/pkgconfig/xtrans.pc Mar 02 13:33:18 http://rafb.net/p/cnBKHq32.html Mar 02 13:34:02 tkp: md5sum /staging/i586-sts-linux/lib/pkgconfig/xtrans.pc Mar 02 13:34:06 so it's the xtrans.pc in lib/pkgconfig Mar 02 13:34:13 tkp: md5sum /staging/i586-sts-linux/share/pkgconfig/xtrans.pc Mar 02 13:34:36 they are different Mar 02 13:34:49 the one in share looks right Mar 02 13:35:44 so.. the problem lies with the Xtrans recipie? or with some pkg-config stuff? Mar 02 13:37:15 tkp: maybe (is the best I can say) Mar 02 13:53:08 odd... I cleaned xtrans, removed both pkg-config files and rebaked Mar 02 13:53:28 this time it only put a .pc file in lib/pkgconfig, and it is correct Mar 02 13:53:44 there is no .pc file in share/pkgconfig now Mar 02 13:53:52 no idea where that came from before Mar 02 14:03:14 hi Mar 02 14:03:18 seems there are a LOT of broken pkgconfig files in staging/.../lib/pkgconfig Mar 02 14:03:22 take a look at this: Mar 02 14:03:30 http://rafb.net/p/R9dOtm40.html Mar 02 14:04:29 is there a smapl bitbake sample repository out? Mar 02 14:04:57 i want to use it for my own software and the OE repository is very bit Mar 02 14:05:05 big sry Mar 02 14:05:38 tkp: we symlink nowadays, is that a clean rebuild? Mar 02 14:06:12 zecke: yes, fresh build... although that is from aan oe checkout from a couple of weeks back Mar 02 14:06:32 monsti: http://www.openembedded.org/~mickeyl/tools/oe-iwmmxt.tar.bz2 is an OE which only builds a iwmmxt toolchain Mar 02 14:06:45 monsti: alternatively have a look at poky, which is a special purpose distribution based on OE Mar 02 14:06:48 ~poky Mar 02 14:06:51 poky is probably http://projects.o-hand.com/poky, or a subset of OpenEmbedded, created by o-hand.com to testdrive some of their technology Mar 02 14:08:01 zecke: thx! Mar 02 14:08:21 zecke: i need something very similar to OE for the psp toolchain Mar 02 14:09:10 anybody knows which package contains "ldd" program? Mar 02 14:09:21 monsti, add support to OE for PSP? Mar 02 14:09:30 Crofton|laptop: no - that won't work Mar 02 14:09:38 why not? Mar 02 14:09:43 Crofton|laptop: the OS is very different to linux Mar 02 14:09:48 monsti: why won't it work? Mar 02 14:09:55 hmmm Mar 02 14:09:55 monsti: so this sounds like fun to support Mar 02 14:09:58 no fork, no kill ... Mar 02 14:10:17 well, it should be possible to use OE anyway Mar 02 14:10:21 uClinux is in a very early stage Mar 02 14:10:22 but most of the packages wouldn't work :) Mar 02 14:10:28 currently i am porting mono Mar 02 14:10:29 ah uClinux Mar 02 14:10:42 uClinux has no interrupts atm :) Mar 02 14:10:44 ucLinux support in OE would be useful Mar 02 14:10:55 I am only slightly familiar with uclinux Mar 02 14:11:11 uclinux is nice - but uclinux for PSP is not usable Mar 02 14:11:24 you need hardware connected to the serial port to get a console Mar 02 14:11:34 so this will take a very long time Mar 02 14:11:48 well, good luck! Mar 02 14:11:52 but i want to get a tool to recompile the toolchain with different compiler fo the psp os Mar 02 14:12:00 i think bitbake can handle this? Mar 02 14:12:10 I suspect it could Mar 02 14:12:34 there are some issues for gcc 4.0.2 <-> 4.1.2 at current SDK Mar 02 14:12:43 monsti: it should be, feel free to ask questions Mar 02 14:12:52 also i want to add some custom patches to the sdk Mar 02 14:12:56 OE is very good at maintinaing the recipe required to build things Mar 02 14:13:01 according to the license i can use the eclasses? Mar 02 14:13:08 so it would easily keep track of needed patches Mar 02 14:13:52 I am wondering if you could use OE, and just creat your own packages directory with uclinux focused bb files Mar 02 14:14:06 I am nt smart enough to know if that is possible though Mar 02 14:14:17 Crofton|laptop: come on Mar 02 14:14:29 will OE work with mingw? Mar 02 14:14:40 monsti: good question, I don't know Mar 02 14:15:21 i will still need the satelite tools like wget, svn, cvs i guess Mar 02 14:15:48 monsti: right, and initial host compiler Mar 02 14:16:00 is there a list? Mar 02 14:16:20 monsti: yes, take a look at sane.bbclass for tools we test presence for Mar 02 14:16:46 hi, I'm trying to bitbake freeciv, but fail because it can't find file-4.18.tar.gz, how can I solve this? Mar 02 14:17:27 monsti: http://www.openembedded.org/repo/org.openembedded.dev/classes/sanity.bbclass Mar 02 14:17:33 YoG: find file-4.18.tar.gz? Mar 02 14:18:00 zecke, ok... I have file-4.18.tar.gz (from another place) Mar 02 14:18:11 YoG: place it in your DL_DIR Mar 02 14:18:21 YoG: and the do touch file-4.18.tar.gz.md5 Mar 02 14:18:43 and me mumbles and complains why file is removed again... Mar 02 14:18:48 zecke: any idea why OE doesn't use prortage from gentoo? Mar 02 14:19:18 monsti: no, has the portage data the concept of packages? Mar 02 14:19:46 packages as in ".deb/.rpm"? Mar 02 14:19:53 for "binary distribution"? Mar 02 14:21:31 zecke: this time I refuse updating file Mar 02 14:21:44 thanx zecke Mar 02 14:22:24 zecke, do I really have to compile all this stuff? I just want freeciv... Mar 02 14:22:55 monsti: yes and also one sourcefile, generating like 1000 packages? Mar 02 14:23:13 YoG: feel free to set ASSUME_PROVIDED to say which software you have Mar 02 14:23:47 hrw: let us move file-4.18 to one of our sites Mar 02 14:23:54 zecke: no - gentoo uses a "depends on" concept with some "fixed" entries e.g. the compiler/core libraries Mar 02 14:24:01 zecke: yes Mar 02 14:24:05 zecke: it also has some meta packages e.g. X11-meta Mar 02 14:24:31 zecke: you can install xorg or other X servers Mar 02 14:24:48 monsti: someone use other X server then X.org? Mar 02 14:24:50 zecke: but i guess OE doesn't need that complexity Mar 02 14:25:00 hrw: was just an example Mar 02 14:25:26 monsti: we talk about packages, e.g. to avoid that your 6mb of flash is populated with doc and man pages Mar 02 14:25:58 hrw: NOTE: package file-4.20: completed Mar 02 14:26:06 monsti: or header files, e.g. if you compile glibc we split the files up to many different packages. one for the runtime, headers, debug packages, timezone data (for each city), locales Mar 02 14:26:08 zecke: for psp i just want to use oe for the toolchain Mar 02 14:26:15 monsti: just to allow tiny distributions Mar 02 14:26:30 monsti: well, you asked why we don't use portage and there is your argument? Mar 02 14:28:10 zecke: i only know portage - i was just asking why bitbake was invented - but you have a good point for splitting ... portage cant' handle splitting very well Mar 02 14:28:47 XorA: good Mar 02 14:28:51 is there something similar to the USE flags? e.g. USE="zlib" build foo Mar 02 14:29:51 monsti: no USE flags, but we have no alternative concept for that Mar 02 14:30:09 so just what configure detects? Mar 02 14:30:10 monsti: the issue with USE flags, is Gentoo ened up with too many and it is hard to track/document Mar 02 14:30:27 monsti: I think currently we have sane defaults and META (virtual/foo) depends Mar 02 14:31:01 so you have foo and foo-zlib? Mar 02 14:32:46 monsti: sometimes we do, but mostly we stick to sane defaults. Personally I like the variants of macports but this is one open topic Mar 02 14:33:22 zecke: can you give me a kick-off - i downloaded oe-iwmmxt.tar.bz2 Mar 02 14:33:28 how can i compile gcc? Mar 02 14:33:39 cd oe-iwmmxt Mar 02 14:33:48 bitbake gcc doesn't work Mar 02 14:34:39 monsti: let me check Mar 02 14:35:35 monsti@sam ~/projects/bitbaketest/oe-iwmmxt $ ls Mar 02 14:35:35 conf toolchain Mar 02 14:35:36 monsti: you need to have bitbake installed Mar 02 14:35:39 thats where i am Mar 02 14:35:40 yes Mar 02 14:35:44 monsti: you need to set BBPATH=$PWD Mar 02 14:35:50 $ bitbake --version Mar 02 14:35:50 BitBake Build Tool Core version 1.6.6, bitbake version 1.6.6 Mar 02 14:36:22 $ bitbake gcc Mar 02 14:36:22 ERROR: Unable to open conf/bitbake.conf Mar 02 14:36:28 i have to copy local.conf? Mar 02 14:36:32 BBPATH wrong? Mar 02 14:36:45 BBPATH=$PWD:path-to-OE-metadata Mar 02 14:36:55 monsti: wait a second, BBPATH should also point to the toolchain dir Mar 02 14:37:34 monsti: once again, bitbake error message is misleading Mar 02 14:37:36 zecke: not toolchain but metadata Mar 02 14:38:08 http://www.xkcd.org/c221.html Mar 02 14:38:39 chouimat|tv: ;D Mar 02 14:39:22 03xora 07org.oe.dev * r449ea8e4... 10/ (3 files in 2 dirs): file-native_4.20.bb,file_4.20.bb : new version of everyones favourite util. Mar 02 14:40:01 monsti: give me some minutes Mar 02 14:40:17 zecke: thx Mar 02 14:42:05 monsti: okay two issues with the conf/local.conf Mar 02 14:42:13 monsti: remove the ASSUME_PROVIDED line Mar 02 14:42:18 monsti: remove the INHERIT line Mar 02 14:42:23 zecke: can you fix the tar.bz2 and add a readme? Mar 02 14:42:42 i might consider just giving the URL you posted to a psp dev Mar 02 14:43:02 monsti: http://rafb.net/p/AJgnGR63.html Mar 02 14:43:15 monsti: oh well, need to talk to mickeyl but he is currently packing for a skiing trip Mar 02 14:43:43 ok - just one second let me try to compile Mar 02 14:44:05 so just fix "oe-iwmmxt/conf/local.conf" ? Mar 02 14:44:06 eek, wait a second Mar 02 14:44:13 I will repack Mar 02 14:44:30 ok thx :) Mar 02 14:47:10 finished? Mar 02 14:47:18 monsti: packing it up Mar 02 14:50:32 monsti: openembedded.org/~zecke/oe-iwmmxt.tar.bz2, but mostly only a README and fix to the local.conf Mar 02 14:50:45 ok thx a lot Mar 02 14:51:06 that is OE' package system? Mar 02 14:51:14 monsti: this will create a gcc 3.3 iwmmxt toolchain Mar 02 14:51:28 monsti: I don't get that question Mar 02 14:52:00 that's the result for a package? Mar 02 14:52:07 some .rpm? a .deb? Mar 02 14:52:12 what do you create for executables Mar 02 14:52:25 monsti: it depends. The host tools install to the staging or cross area Mar 02 14:52:47 monsti: for the target system we can create packages. mainly tar.gz and ipk, and not so well maintained support for rpm Mar 02 14:53:37 oe-iwmmxt $ export BBPATH=$PWD Mar 02 14:53:43 $ bitbake gcc Mar 02 14:53:43 ERROR: Unable to open conf/bitbake.conf Mar 02 14:55:10 monsti: see the README Mar 02 14:55:34 find oe-iwmmxt -name "README" Mar 02 14:55:34 monsti: or better rtfm Mar 02 14:55:37 no readme :( Mar 02 14:55:51 openembedded.org/~zecke/oe-iwmmxt.tar.bz2 <- i downloaded that file Mar 02 14:56:21 Length: 149,935 Mar 02 14:56:46 monsti: and md5? Mar 02 14:57:09 one second Mar 02 14:57:23 $ md5sum oe-iwmmxt.tar.bz2 Mar 02 14:57:23 0853e83cb35ab4465ed2f0692707e007 oe-iwmmxt.tar.bz2 Mar 02 14:57:34 that is the correct answer, then unpack it? Mar 02 14:57:43 now i have a readme Mar 02 14:57:52 musst bne some proxy caching Mar 02 14:58:14 thx - works no! Mar 02 14:58:22 +w Mar 02 14:59:31 actually bitbake virtual/arm-linux-gcc would generate the iwmmxt toolchain Mar 02 14:59:55 where is virtual/arm-linux-gcc fefined? Mar 02 15:00:29 monsti: nowhere, it is magic. The gcc files say they provide virtual/${TARGET_ARCH}-gcc Mar 02 15:00:53 ah ok - so every bb file is scaned for "who provides ffoo" Mar 02 15:01:40 right, DEPENDS, RDEPENDS; PROVIDES not being a gentoo user I think this should be familiar Mar 02 15:02:23 yes but gentoo has some "world" file with essential tools that are not part of the DEPENDS of the .ebuilds Mar 02 15:02:37 e.g. "everything" needs gcc - but there is no depend on gcc in every ebuild Mar 02 15:03:25 but thx a lot! i have to read the .bb doc now Mar 02 15:04:18 monsti: now you can use openembedded.org/~mickeyl/tools/oe-iwmmxt.tar.bz2 again Mar 02 15:05:27 can you send the md5? :) Mar 02 15:05:35 09aac49bd80160500c12efb3441ff5e5 oe-iwmmxt.tar.bz2 Mar 02 15:05:59 no readme inside :))) Mar 02 15:06:03 it should be the same, as above, mickeyl should just have copied the file Mar 02 15:06:40 0853e83cb35ab4465ed2f0692707e007 <- should be the sum Mar 02 15:07:14 downloaded gagin 09aac49bd80160500c12efb3441ff5e5 Mar 02 15:07:19 and no README inside Mar 02 15:07:36 monsti: then fix your proxy? Mar 02 15:07:48 it's ok with your site? Mar 02 15:07:58 monsti: yes Mar 02 15:09:13 ok works from a different host Mar 02 15:09:17 << UMTS user :( Mar 02 15:13:01 http://www.openembedded.org/viewmtn monotone is dead somehow Mar 02 15:15:33 indeed Mar 02 15:16:08 * florian has no idea how to fix this, maybe koen|away or zecke Mar 02 15:16:23 remove the sqlite lock Mar 02 15:16:46 dead in which way? Mar 02 15:16:51 http://www.openembedded.org/viewmtn/diff.psp?id1=df7aec92a4539183334adae4278b754f14e68bb7&id2=449ea8e45f6398be01dd6f681846b2c1f8a10c67 Mar 02 15:17:17 ah works again Mar 02 15:17:27 *sith power* Mar 02 15:18:54 http://www.openembedded.org/~zecke/sith.jpg Mar 02 15:21:46 zecke: scary Mar 02 15:22:11 ;) Mar 02 15:27:31 zecke: so is http://www.openembedded.org/~zecke/greenphone.jpg what you have in mind for "open" hardware ? Mar 02 15:28:11 hvontres|poodle: lol Mar 02 15:28:18 hvontres|poodle: if it gets delivered with a screw-driver (torque), yes Mar 02 15:28:39 zecke: how is the phone btw? Mar 02 15:28:40 it looks like the new pango_1.16.0 isn't building pango-module-basic-fc, although it does RRECOMMEND it. This is making my gpe-image fail. Mar 02 15:28:58 chouimat: it feels like a normal phone (which is good) Mar 02 15:29:12 chouimat: have not played with it much, I wait for the WebKit port to be usable for it Mar 02 15:29:34 zecke: too bad I don't have enough money to get one ... look like a nice toy Mar 02 15:30:07 zecke: so I got the next best thing : ipod 30gb ... rockbox work well on it and it's fun to hack too :) Mar 02 15:31:03 chouimat: which ipod generation is that? Mar 02 15:31:32 5,5 Mar 02 15:42:13 zecke: which phone? the neo1973? Mar 02 15:50:19 summatusmentis: greenphone Mar 02 15:50:44 hrw: is that GSM too? Mar 02 15:51:17 oh, and hrw: I noticed you were looking for developers to get another release of OZ out, what sort of stuff do you need? I'm not all that great, but I can try Mar 02 15:51:33 does anyone know anything about the recent pango update? Mar 02 15:53:07 summatusmentis: greenphone is gsm Mar 02 15:53:14 03hrw 07org.oe.dev * recf40970... 10/ (4 files in 3 dirs): glib-2.0-native: add 2.12.4 from Poky (needed for python-pygobject 2.12.x) Mar 02 15:53:38 summatusmentis: any one who know how to use OE and how to integrate patches into OE Mar 02 15:54:22 hrw: I'm working on learning that, currently, I assume you're looking for coding though? not just someone to administer patches? Mar 02 15:55:29 summatusmentis: current OZ team is XorA and me. even if we provide images for testing there must be some more people to work on fixing Mar 02 15:56:36 hrw: hrm... I'd be willing to try, my coding experience is nothing more than the basics, but I'd like to help Mar 02 15:58:36 ok Mar 02 15:59:07 hrw: what do you need from me? Mar 02 15:59:30 summatusmentis, long hours :) Mar 02 16:00:08 Crofton|home: heh, I've got some hours Mar 02 16:00:54 summatusmentis: currently you can look at distro/openzaurus-3.5.5.conf and clean it ;D Mar 02 16:01:59 hrw: ok, I'll figure out how to get it, then figure out how to read it, and start there Mar 02 16:02:24 or you could ditch OZ and help angstrom getting a release :) Mar 02 16:06:34 * chouimat is sad his 5000d died ... well the battery kind of exploded Mar 02 16:07:51 03koen 07org.oe.dev * r21f4ee6d... 10/ (1 packages/openmoko-pim/openmoko-today_svn.bb): openmoko: add openmoko-today Mar 02 16:08:22 now OE has more openmoko recipes than openmoko svn Mar 02 16:08:54 what does today_svn do? Mar 02 16:09:05 a 'today' screen Mar 02 16:09:15 with summarrie sof your appointements and stuff Mar 02 16:09:25 ah Mar 02 16:09:36 like opie-today ;) Mar 02 16:09:39 so taoday has nothing to do with svn SRCDATE+"now" Mar 02 16:09:46 no :) Mar 02 16:11:26 koen: please update glib-2.0-native in angstrom Mar 02 16:12:07 where? Mar 02 16:12:41 hrw: please check before asking :) Mar 02 16:13:25 koen: updated to 2.12.4 which I added? Mar 02 16:13:34 what's there to update? Mar 02 16:13:49 angstrom isn't in a branch Mar 02 16:14:03 and it has no PREFERRED_VERSIONS for glib native Mar 02 16:14:11 koen: ah sorry - I was thinking that it has Mar 02 16:14:26 :) Mar 02 16:14:31 my grep disagrees :) Mar 02 16:14:52 I wanted to say "sure", but I couldn't find where it was pinned Mar 02 16:15:10 when can we expect elisa and flumotion in OE? Mar 02 16:16:06 elisa? flumotion? Mar 02 16:17:13 gstreamer stuff :) Mar 02 16:17:23 I'm trying to guess what you are working on Mar 02 16:17:48 later all Mar 02 16:18:00 ah Mar 02 16:50:09 cu Mar 02 17:08:21 re Mar 02 17:25:24 RP: ping? http://www.pastebin.ca/378679 Mar 02 17:26:02 NAiL: update-altternatives --set git... Mar 02 17:26:08 a Mar 02 17:26:28 There is only 1 program which provides git Mar 02 17:26:28 (/usr/bin/git.transition). Nothing to configure. Mar 02 17:27:00 HopsNBarley: ping Mar 02 17:27:14 NAiL: then set git to git.scm? Mar 02 17:27:19 Hi all Mar 02 17:27:54 hey likewise Mar 02 17:28:01 Jin^eLD: Hi there Mar 02 17:30:22 zecke: got it. Mar 02 17:30:43 now it fails compiling instead: http://www.pastebin.ca/378682 Mar 02 17:31:14 likewise, ping Mar 02 17:31:31 HopsNBarley: pong Mar 02 17:31:43 linux kernel 2.6 for $20: what system calls can delete a directory? just unlink/rmdir? Mar 02 17:32:13 HopsNBarley: yes, AFAIK Mar 02 17:32:49 as near as i can tell, that is at the root of this fakeroot problem i'm tracking down. Mar 02 17:33:08 a directory is chmod'ed, okay says fakeroot, and remembers this. Mar 02 17:33:08 NAiL: Looks like something is wrong with the libs paths for mtd-utils. I admit I've tested mtd-utils-native more than mtd-utils :/ Mar 02 17:33:43 much later, in a different command (now mkfs.jffs2) the same inode is stat'ed. fakeroot remembers it as a directory, though now it is a file. Mar 02 17:33:57 so somehow, a directory has been deleted w/o fakeroot knowing about it. Mar 02 17:34:14 but, fakeroot catches unlink/rmdir, so i'm at a bit of a loss as to how this could happen. Mar 02 17:34:17 any ideas? Mar 02 17:34:42 HopsNBarley: could you show a list of batch commands to reproduce this? Mar 02 17:35:21 it seems installation dependent. only a few people have seen it. Mar 02 17:38:11 RP: Any fix? Mar 02 17:40:08 RP: so lzo is another fast realtime compression algorithm? Mar 02 17:41:53 zecke: yes Mar 02 17:42:14 NAiL: Not that I know of at the moment. I'd need to look at it Mar 02 17:44:28 HopsNBarley: Just out of curiosity, what FS do you have on you build machine? Mar 02 17:46:14 lolo Mar 02 17:46:26 Hrw: you there ? Mar 02 17:46:28 hvontres|poodle, ext3 Mar 02 17:46:48 hvontres|poodle, i've booted/fscked with both 2.6.17 and 20. Mar 02 17:46:59 HopsNBarley: hmmm...so nothing special there... Mar 02 17:47:52 hrw|gone: is the openzaurus-3.5.5.conf in the .dev tree? Mar 02 17:48:53 back in a bit. Mar 02 17:48:59 HopsNBarley: rename(2) ? Mar 02 18:41:24 can someone remind me off the gentoo users and olpc joke? Mar 02 18:54:54 luke-jr_: feel free to join the Month of the Qtopia Patch Mar 02 19:08:03 Crofton: How many gentoo users does it take to screw in an olpc? Something along those lines? :) Mar 02 19:08:35 * chouimat feels insulted by this boring joke Mar 02 19:09:08 heh Mar 02 19:31:24 anyone has any experience with g729 codec ? Mar 02 19:51:16 zecke: What is this Qtopia Patch Month thingy? Mar 02 19:54:33 hvontres|poodle: Send issues best with patch attached to qtopia-interest Mar 02 19:54:47 hvontres|poodle and at the end of the month judge how much feedback was achieved Mar 02 19:56:13 zecke: sound fun :) Mar 02 20:16:33 hi Mar 02 20:16:38 gerwinin: pong Mar 02 20:17:03 summatusmentis: yes Mar 02 20:19:11 hrw: two questions:) Mar 02 20:19:32 go Mar 02 20:19:52 hrw: Did you speak with the guy from nokia ? I want to arrange vouchers for 10-20 oe people to buy n800 at 99 euro Mar 02 20:20:11 hrw: Do you still have enough paid work ? Some dude called me and wants to hire some people Mar 02 20:20:44 gerwinin: I did not - we spoke for <10 minutes in total. without koen I would not even know who that guy was Mar 02 20:21:21 hrw: org.openembedded.dev/conf/distro ? Mar 02 20:21:23 gerwinin: I'm fulltime busy to August. Then month break for vacations and later not known yet. Mar 02 20:21:33 summatusmentis: yes. read comments inside of it Mar 02 20:21:48 I don't see that file, I just updated Mar 02 20:22:02 hrw : okay Mar 02 20:22:16 hrw: I will tell the guy I don't know anybody :) Mar 02 20:22:37 gerwinin: no one answered yet to that question here? Mar 02 20:22:37 gerwinin, what des he want people to do? Mar 02 20:23:17 summatusmentis: ops. I did not pushed it... sorry Mar 02 20:24:56 gerwinin: there are more people working on OE then just coreteam Mar 02 20:25:46 hrw: no worries Mar 02 20:26:26 hrw: I know Mar 02 20:26:40 Crofton: he wants people to make some kind of medical device Mar 02 20:26:57 Crofton: In holland they is a huge shortage of it-people Mar 02 20:27:23 Hello Mar 02 20:27:23 Crofton: For example I get paid nowadays 100 euro per hour for making some kind of stupid java app Mar 02 20:27:50 intriguing Mar 02 20:28:04 but likely tricky for me since I am in US and it is medical Mar 02 20:28:17 sometimes that stuff can be very onvolved Mar 02 20:28:32 gerwinin: how hard is it to get into holland? Mar 02 20:28:46 Crofton: I had an offer from the us to become the asterisk dude for some company 100k dollar a year Mar 02 20:28:55 100 eur/h - nice price Mar 02 20:29:04 granted the market will probably change in 4 years Mar 02 20:29:14 Maybe hrw is less busy than he thinks :) Mar 02 20:29:15 summa : depends on where you from Mar 02 20:29:28 Crofton: I have half year contract and I like it. Mar 02 20:29:34 gerwinin: us Mar 02 20:30:38 ~change 100 euro to usd Mar 02 20:30:44 100.00 Euro (EUR) makes 131.941 United States Dollar (USD) (from http://www.xe.com/) Mar 02 20:30:52 ouch Mar 02 20:30:58 ~lart usd Mar 02 20:30:58 * ibot squishes usd like a bug Mar 02 20:32:10 hrw, take the contract and hire someone for half the rate to do the work :) Mar 02 20:32:40 Crofton: it would be possible to find someone. but he would not be so good Mar 02 20:32:46 :) Mar 02 20:32:50 RP: Hello? Mar 02 20:32:58 03hrw 07org.oe.dev * r224b2627... 10/ (1 conf/distro/openzaurus-3.5.5.conf): (log message trimmed) Mar 02 20:32:58 OpenZaurus 3.5.5 distro config - need work Mar 02 20:32:58 It contain stuff from: Mar 02 20:32:58 - OpenZaurus 3.5.4.x Mar 02 20:32:58 - Ångström 2007.1 Mar 02 20:32:59 - UmbaUmba 17.3 Mar 02 20:33:01 This needs to be cleared and split into includes, some stuff even updated. Mar 02 20:33:23 summatusmentis: go with it Mar 02 20:33:31 hrw: thanks Mar 02 20:34:04 * Crofton needs to edit thesis.tex Mar 02 20:34:31 * gerwinin going in ze pub :) Mar 02 20:41:28 ~change 100 euro to cad Mar 02 20:41:36 100.00 Euro (EUR) makes 155.269 Dollar (CAD) (from http://www.xe.com/) Mar 02 20:43:40 http://pastebin.com/892321 Mar 02 20:43:45 can someone look at that and help me out Mar 02 20:43:58 I have jsut started to get into OE but to be honest, I have never had one thing compile without failing Mar 02 20:44:03 RP: I'm afraid that the minimum address where we could map external memory is 0x200000. Not all the bits are settable. Mar 02 20:44:11 hrw: I'm unfamiliar with the way OZ is built, what needs cleaning? is there a reference somewhere I can look at? Mar 02 20:44:25 i have been able to get through some of it (such as undefined constants, etc) but this one is a little more difficult Mar 02 20:44:48 summatusmentis: compare with .oz354x ones Mar 02 20:45:33 RP: I tried to set both registers, and the masks are 0xFFFFF800 for MC_FB_LOCATION and 0xFFFFE000 for MC_EXT_MEM_LOCATION. Mar 02 20:45:34 hrw: ok Mar 02 20:45:55 hrw: thanks for pushing the 3.5.5 config... I'm going to try and take a look at it this weekend myself. BTW, I did some initial testing of 2.6.20 on poodle and ran into a probelem with the screen getting corupted on resume. Mar 02 20:46:55 hvontres|poodle: that config needs cleaning before use Mar 02 20:47:19 hrw: I saw. But at least it's a start :) Mar 02 20:47:50 cyrus__, that is an error in the code Mar 02 20:48:13 something like you do not have an implementation for one of the class members Mar 02 20:48:59 http://gcc.gnu.org/faq.html#vtables Mar 02 20:49:09 The message is useless :) Mar 02 20:52:35 Crofton__: no, it says the linker is crap :) Mar 02 20:53:21 well Mar 02 20:53:29 that is an laternate explanantion :) Mar 02 20:53:41 you make use a dictionary Mar 02 20:53:56 you make my dictionary look bad Mar 02 20:53:59 speel checker Mar 02 20:54:01 you meant alternate Mar 02 20:54:05 yes Mar 02 20:54:14 which makes my dictionary look good :) Mar 02 20:54:58 That error almost drove me insanse once Mar 02 20:55:05 insane Mar 02 20:58:56 XScale DSP coprocessor detected. Mar 02 20:59:25 I think I never saw that message before 2.6.20. What is that coprocessor supposed to be? Mar 02 21:20:47 cu Mar 02 21:40:08 zecke: best place for patches is not the mailing list, but task tracker Mar 02 21:43:14 ljp: can I use mails to send stuff to the task tracker **** BEGIN LOGGING AT Fri Mar 02 21:50:13 2007 Mar 02 22:09:59 where can i find a description of a package. For instance, I am building linux-rp-2.6.20. I know that is linux kernel 2.6.20 but what does the -rp mean Mar 02 22:10:53 it means RP Mar 02 22:11:07 /whois RP Mar 02 22:11:14 ;-) Mar 02 22:11:34 cyrus__: it means Richard Purdie (RP) made this wonderfull thing work for us :) Mar 02 22:12:00 from what I can tell, .bb packages are kind of like ebuild in gentoo. Is there a tool in oe like esearch in gentoo to search the pakages for what you are looking for Mar 02 22:17:19 hi Mar 02 22:17:22 cyrus__: find | xargs grep Mar 02 22:18:18 cyrus__: bitbake interactive also does command completion which sometimes helps Mar 02 22:19:29 Hey CoreDump|home :) Mar 02 22:34:00 hi everyone, I want to know how to get around building stuff using gcc 4.1, more specifically QEMU Mar 02 22:34:34 (I am rather new at this so I am following the getting started instructions on the website) Mar 02 22:34:50 work_donato: what distro are you on? usually you have older gcc versions in separate packages that you can install Mar 02 22:34:56 suse Mar 02 22:35:00 opensuse 10.2 Mar 02 22:35:11 hmm, look if they offer gcc-compat packages, with gcc 3.x Mar 02 22:35:27 on most distros those can live together with gcc 4.x Mar 02 22:35:37 Ill give it a try, but, nonetheless, it can be installed in parallel with 4.1, right? Mar 02 22:35:42 ok Mar 02 22:35:46 yes Mar 02 22:35:52 thx very much Mar 02 22:36:14 work_donato: on some distros (like fedora IIRC) it's compat-gcc Mar 02 22:37:08 work_donato: if you still have problems with that you could alternatively try to disable binary locale generation by putting ENABLE_BINARY_LOCALE_GENERATION = "0" into your local.conf Mar 02 22:37:33 I had to do this because there were problems with the compat-gcc packages and building qemu on FC6 Mar 02 22:37:41 'night everyone Mar 02 22:39:44 > "If SUSE (or any other vendor) started to ship compilers for every variant Mar 02 22:39:44 > of broken code floating around in the world they would end up shipping all Mar 02 22:39:44 > versions of gcc ever released or even those that were never released." Mar 02 22:39:55 this came from suse itself Mar 02 22:40:07 (apparently) Mar 02 22:40:22 Nice attitude. Mar 02 22:40:55 The differences between gcc3 and gcc4 are dramatic, esp. for kernel work. Mar 02 22:41:01 mwester: Yet it sounds strangly familiar... Mar 02 22:41:31 And it's hardly because of "broken" code, either... Mar 02 22:41:54 hvonres|poodle: :-D Mar 02 22:41:55 imho it is "outdated" code Mar 02 22:42:29 * hvontres|poodle didn't even mean to throw in an extra pun... Mar 02 22:45:00 work_donato: depends on how you define the average lifetime of a software release, I guess. By comparison, I'm sure that when Canada switched to the metric system, the wise mechanic kept his old 9/16" wrench for quite some time... Mar 02 22:45:16 * CoreDump|home tries to build OpenMoko from .dev Mar 02 22:52:02 * CoreDump|home notes that do_install of openmoko-panel-gsm-0.0.1+svn20070302-r0: is b0rked Mar 02 22:53:13 zecke: ?? Mar 02 23:15:20 sirfred: ah. Interesting to know I guess Mar 02 23:15:42 sirfred: I guess it doesn't matter where you remap it since you can work that out in userspace... Mar 02 23:26:39 * CoreDump|home humms Mar 02 23:27:26 03coredump2 07org.oe.dev * r971174e3... 10/ (1 packages/openmoko-panel-plugins/openmoko-panel-gsm_svn.bb): openmoko-panel-gsm_svn.bb: Add missing dependency to libgsmd Mar 02 23:28:51 * CoreDump|home wonders if there might be a "policy" against fixing depends somewhere. Mar 02 23:29:54 * mwester rolls his eyes Mar 02 23:30:14 * CoreDump|home couldn't resist Mar 02 23:30:34 CoreDump|home: please try ;-) Mar 02 23:30:48 * RP can't take much more today Mar 02 23:31:18 * CoreDump|home is pretty proud of himself that he didn't respond to koens latest BS on the ML Mar 02 23:32:26 I still need to respond to that. It had batter wait until tomorrow now though as I don't think I'm in the right frame of mind... Mar 02 23:32:31 this u-boot thingy sure takes a while to compile Mar 02 23:32:46 That's a bootloader Mar 02 23:32:51 I know =) Mar 02 23:33:04 that was meant as in "surprise" Mar 02 23:33:07 :) Mar 02 23:33:18 ;) Mar 02 23:33:25 funny name though. Mar 02 23:33:38 Second-stage bootloader, is it replacing the bootloader in the target? Mar 02 23:34:06 * mwester is wondering, since it seems risky to replace as a matter of course... Mar 02 23:35:09 mwester: Not really replace I guess. Dunno why OM rebuilds it with every image. As I understand it, kernel and rootfs can be "flashed" into neo w/o touching u-boot. Flashing u-boot is optional Mar 02 23:35:38 A search of the openmoko wiki didn't show anything obvious - thanks. Mar 02 23:35:52 I personally would settle for a fixed SVN date for it =) Mar 02 23:36:38 There's a number of things that probably need to be locked down before the OpenMoko stuff gets released, I would expect :) Mar 02 23:36:59 You probably need part of uboot to generate the kernel image? Mar 02 23:37:30 mwester: Did mtd-utils-native work out btw? Mar 02 23:37:32 Oh? I didn't realize u-boot and kernel are that close Mar 02 23:37:49 mwester: indeed =) Mar 02 23:37:52 CoreDump|home: uboot can use a uImage from the kernel instead of a zImage Mar 02 23:37:57 For example, I'm accustomed to a tool-chain being locked down and immutable at the time the base firmware is released, and then all packages and maintenance work done using that toolchain -- which is not the standard model for OE. Mar 02 23:38:29 RP: Yes, all works. un-scientific test on SlugOS seems to indicate that the rootfs is 4 blocks larger with the new compressor. Mar 02 23:38:31 mwester: that's probably the reason for OM SVN ;) Mar 02 23:38:48 RP: I guess I need to do some reading. what's uImage? Mar 02 23:38:49 I thought everything in OM SVN was moving into OE. Mar 02 23:38:56 mwester: Thats only because toolchains are tradionally difficult to build :) Mar 02 23:39:35 mwester: Ok, thats expected as lzo doesn't compress quite as well, the benefit is much faster decompression Mar 02 23:39:47 mwester: OM SVN is a controlled environment for a fixed .dev DB snapshot (or IOW: your locked down toolchain) Mar 02 23:39:52 mwester: I'm assuming you enabled lzo? Mar 02 23:40:25 CoreDump|home: A special version of the kernel for use by uboot as I understand it Mar 02 23:40:36 I don't know if omoko use it or not Mar 02 23:41:05 bbimage specified "-x lzo" on the mkfs.jffs2 command line, and I verified that commandline was used by checking the correct log... Mar 02 23:41:36 mwester: Right, that means exclude lzo... Mar 02 23:42:04 * mwester should read man pages more often. Mar 02 23:42:07 -m favourlzo will be what the zaurii will use Mar 02 23:42:50 Well, then I can't tell you if the lzo compressor worked, but I can tell you that it all builds fine. :) Mar 02 23:43:28 That was my main worry. I'm sure the rest will be fine Mar 02 23:47:45 openmoko does use uimage Mar 02 23:49:03 CoreDump|home: mokomakefile has patches so that it doesn't rebuild every time Mar 02 23:49:11 rwhitby: =) Mar 02 23:49:40 but that's offtopic in this channel probably ;-) Mar 02 23:50:11 the DISTRO OpenMoko build is coming quite far this time Mar 02 23:50:27 * rwhitby will add openmoko to the nslu2-linux Master Makefile instead :-) Mar 02 23:50:44 =D Mar 02 23:50:51 * rwhitby goes swimming with the kids ... Mar 02 23:51:03 have fun! Mar 02 23:55:40 * mwester walks in, covered in snow, having been scraping ice off of a vehicle in his driveway... Mar 02 23:56:20 'night all Mar 02 23:56:40 n8 RP Mar 03 00:01:18 !oebug 1869 Mar 03 00:01:20 * * Bug 1869, Status: NEW, Created: 2007-02-13 11:48 Mar 03 00:01:21 * * manuel.teira(AT)telefonica.net: New package: GPE frontend for the snes9x emulator Mar 03 00:01:22 * * http://bugs.openembedded.org/show_bug.cgi?id=1869 Mar 03 00:28:10 !oebug 1869 Mar 03 00:28:11 * * Bug 1869, Status: RESOLVED, Created: 2007-02-13 11:48 Mar 03 00:28:12 * * manuel.teira(AT)telefonica.net: New package: GPE frontend for the snes9x emulator Mar 03 00:28:13 * * http://bugs.openembedded.org/show_bug.cgi?id=1869 Mar 03 00:31:38 03coredump2 07org.oe.dev * r9115424b... 10/ (1 packages/gpe-snes packages/gpe-snes/gpe-snes_0.1.bb): gpe-snes: Add new .bb from Bugzilla, closing bug #1869 by Manuel Teira Mar 03 00:31:43 03coredump2 07org.oe.dev * r9382fc4a... 10/ (1 MAINTAINERS): MAINTAINERS: Change reciepes maintained by Joaquim Duran, closing bug #1922 Mar 03 00:31:52 03coredump2 07org.oe.dev * r1b2e4b30... 10/ (1 packages/qpegps/qpegps_0.9.3.bb): qpegps_0.9.3.bb: Add new .bb from Bugzilla, closing bug #1806 by Henry von Tresckow, thanks! Mar 03 00:34:19 03coredump2 07org.oe.dev * rd884993f... 10/ (1 packages/aircrack/aircrack-ng_0.7.bb): aircrack-ng_0.7.bb: Add new .bb from Bugzilla, closing bug #1932 by Zero_Chaos, thanks! Mar 03 00:42:15 03coredump2 07org.oe.dev * r4b823086... 10/ (3 files in 3 dirs): kismet: Add new .bb from Bugzilla, closing bug #1933 by Zero_Chaos, thanks! Mar 03 00:42:43 Was today the bugday? Mar 03 00:42:56 * NAiL hasn't even looked at the list of bugs Mar 03 00:42:56 not really **** BEGIN LOGGING AT Sat Mar 03 01:07:02 2007 Mar 03 01:45:39 !oebug 1934 Mar 03 01:45:40 * * Bug 1934, Status: NEW, Created: 2007-03-02 18:06 Mar 03 01:45:41 * * r.farina(AT)adelphia.net: mdk2 v33 Mar 03 01:45:42 * * http://bugs.openembedded.org/show_bug.cgi?id=1934 Mar 03 01:45:44 !oebug 1936 Mar 03 01:45:46 * * Bug 1936, Status: NEW, Created: 2007-03-02 18:13 Mar 03 01:45:46 * * r.farina(AT)adelphia.net: Raw FakeAP Mar 03 01:45:47 * * http://bugs.openembedded.org/show_bug.cgi?id=1936 Mar 03 01:45:53 anyone be a pal and commit those for me? Mar 03 01:54:16 emte: you have commit access buddy? Mar 03 01:59:40 Zero_Chaos: Have you tested them? Mar 03 02:03:21 repvik@apollo:~/slug/openembedded/packages/fakeap$ cat /home/repvik/slug/openprotium/tmp/work/fakeap-0.2-r1/temp/log.do_install.22004 Mar 03 02:03:25 install: cannot stat `rfakeap': No such file or directory Mar 03 02:05:08 mdk2 fails to compile for me Mar 03 02:07:42 Zero_Chaos: I've updated the bugs with info and added myself to CC. Feel free to update the .bb's and update the bugs. Mar 03 02:23:43 !oebug 1907 Mar 03 02:23:45 * * Bug 1907, Status: NEW, Created: 2007-02-23 11:08 Mar 03 02:23:46 * * angus(AT)handheldshell.com: Toshiba e740 kernel recipe Mar 03 02:23:47 * * http://bugs.openembedded.org/show_bug.cgi?id=1907 **** ENDING LOGGING AT Sat Mar 03 02:59:59 2007