**** BEGIN LOGGING AT Thu Mar 16 10:59:56 2006 Mar 16 11:00:56 pth_mctx.c:476:2: #error "Unsupported Linux (g)libc version and/or platform" Mar 16 11:00:59 bizarre Mar 16 11:20:11 morning all, should there be parsing errors with the latest version of bitbake and the yesterday's org.openembedded.dev? Mar 16 11:20:58 Like this: http://pastebin.ca/45904 Mar 16 11:21:41 ah, ignore me, the local.conf file has changed again Mar 16 11:21:54 * france is back (gone 11:05:33) Mar 16 11:22:30 hey Liam Mar 16 11:22:36 hey Liam Mar 16 11:22:48 Hey Dirk, got your email, thanks for the patch Mar 16 11:23:01 hey lrg, lardman & do13_ Mar 16 11:23:06 lrg: thx for your work Mar 16 11:23:08 hey koen Mar 16 11:23:14 Hey koen Mar 16 11:25:32 hi koen Mar 16 11:25:50 lrg: hey Mar 16 11:26:05 Hey XorA Mar 16 11:26:16 anyone else having problems compiling pth? Mar 16 11:27:58 XorA: will check Mar 16 11:28:00 XorA: someone on the mailinglist Mar 16 11:32:11 Attempted builds: 0 Mar 16 11:32:37 so I built pht-2.0.2 before without problems Mar 16 11:33:39 hrw|work: I have built it before, it doesnt build today Mar 16 11:34:45 /home/dp/zaurus/build-gcc/tmp/work/armv5te-linux/pth-2.0.0-r1/temp/log.do_compile.30872 Mar 16 11:34:54 http://article.gmane.org/gmane.comp.handhelds.openembedded/9676/match=pth Mar 16 11:35:01 is the same problem I see Mar 16 11:35:19 * XorA starts to go mad combining windows clipboard with X one Mar 16 11:35:49 koen: what is the correct sollution on OE if one CVS-Package dosen't compile? put the error here on chat or send email or ??? Mar 16 11:36:27 but the errror in pastebin and paste the link here Mar 16 11:36:51 if that doesn't give a solution file a bug at bugs.openembedded.org Mar 16 11:36:59 NOTE: package pth-2.0.2: started Mar 16 11:37:02 will look Mar 16 11:37:39 | ERROR: QEMU requires SDL or Cocoa for graphical output Mar 16 11:37:39 | To build QEMU with graphical output configure with --disable-gfx-check Mar 16 11:37:39 | Note that this will disable all output from the virtual graphics card. Mar 16 11:37:39 | FATAL: oe_runconf failed Mar 16 11:37:39 NOTE: Task failed: /opt/contec/ptxdist-projects/ptxdist-micro9/micro9-15/tmp/work/i686-linux/qemu-native-0.8.0+cvs20060316-r0/temp/log.do_configure.18214 Mar 16 11:37:40 NOTE: package qemu-native-0.8.0+cvs20060316-r0: task do_configure: failed Mar 16 11:37:42 ERROR: TaskFailed event exception, aborting Mar 16 11:37:44 NOTE: package qemu-native-0.8.0+cvs20060316: failed Mar 16 11:37:55 that is my error on bitbake task-boostrap Mar 16 11:37:55 ~pastebin Mar 16 11:37:57 ~flood Mar 16 11:37:58 methinks pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Mar 16 11:37:59 somebody said flood was spewing loads of output into a channel; *very* rude in most channels and often grounds for banning. If you want to show a lot of output to someone, ask them to join you in #flood and paste the output there. Mar 16 11:38:19 XorA: NOTE: package pth-2.0.2-r0: task do_build: completed Mar 16 11:38:20 grma: set ENABLE_BINARY_LOCALE_GENERATION ?= "" in your local.conf Mar 16 11:38:27 ok, i have understand, sorry Mar 16 11:39:05 sorry for puting this here on chat Mar 16 11:39:15 grma: or installed qemu on your host and set ASSUME_PROVIDED += "qemu-native" in your local.conf Mar 16 11:41:07 what is qemu, a emulator ? Mar 16 11:41:27 grma: yes, it can run ARM binaries on your host Mar 16 11:41:57 ahh, i thought i have heard about qemu :-) Mar 16 11:44:01 hrw|work: crappit it must be a host info leakage thing then Mar 16 11:51:23 hrw|work: can youpastebin you staging/arm-linux/bits/setjmp.h file? Mar 16 11:51:46 koen: bitbake has always tried to build another provider in case of failure Mar 16 11:52:15 zecke: are different versions also providers? Mar 16 11:52:55 koen: IIRC yes Mar 16 11:53:04 OE/bitbake used to error out if the build of a package failed Mar 16 11:53:23 right now it picks another version/provider if you 'restart' the build Mar 16 11:53:30 koen: that is still the case (and no this is not a contradiction) Mar 16 11:54:05 koen: if you ask it to build xyz (preferred version 1.3) anf that fails, it will fail Mar 16 11:54:29 I agree it can create issues (as seen by mickeyl) Mar 16 11:54:45 but if you say 'bitbake xyz' after that it will build version 1.2 Mar 16 11:55:00 without stating that 1.3 failed to build Mar 16 11:55:00 does it? Mar 16 11:55:06 XorA: cat: tmp/staging/arm-linux/bits/setjmp.h: No such file or directory Mar 16 11:55:11 that's the problem we have been seeing Mar 16 11:55:15 (just if it is in the already_failed cache) Mar 16 11:55:49 the already_failed cache is causing more harm than good Mar 16 11:55:57 it makes build unpredictable Mar 16 11:56:00 build* Mar 16 11:56:51 tmp/staging/arm-linux/include/bits/setjmp.h sorry, I missed a bit on path Mar 16 11:57:30 koen: I'm currently in the process of creating testcases and documenting bitbake Mar 16 11:57:38 (idling at my uni project) Mar 16 11:58:45 ibot: botmail for ph5: try http://ewi546.ewi.utwente.nl/OE/eabi/gcc-no-fortran.diff to get around the fortran problems Mar 16 11:59:14 ibot: botmail for pH5: try http://ewi546.ewi.utwente.nl/OE/eabi/gcc-no-fortran.diff to get around the fortran problems Mar 16 12:00:03 NOTE: package glibc-2.4-r1: task do_populate_staging: started Mar 16 12:00:12 * koen gets a warm and fuzzy feeling Mar 16 12:03:40 today I have a 'fast fixes' day Mar 16 12:04:03 and my home machine died during building ;( Mar 16 12:04:43 :( Mar 16 12:05:07 I lack ideas why that happens.. Mar 16 12:05:30 memtest86 passed, not overheating, two PSU tested Mar 16 12:05:50 I can suspect hdd but no cash for another 120G one Mar 16 12:19:17 for recipe in pkg/cfxga/*.bb;do bbrebuild $recipe;done Mar 16 12:19:25 how to not like shell ;) Mar 16 12:20:02 hrw|work: that mean you have no access to that file I want anymore? Mar 16 12:20:25 XorA: I do not have that file in staging Mar 16 12:21:34 hrw|work: try tmp/staging/arm-linux/include/bits/setjmp.h the corrent path :-) Mar 16 12:21:54 better now Mar 16 12:22:08 http://pastebin.com/605243 Mar 16 12:23:17 hrw|work: interesting, yours is the same as mine, which begins to make me suspect on your host it pulls in /usr/include/bits/setjmp.h from somewhere Mar 16 12:23:22 NOTE: package glibc-2.4: completed Mar 16 12:25:18 ${KERNEL_CC} -D__KERNEL__ -I ${STAGING_KERNEL_DIR}/include -DMODULE -c -o memory_cs.o memory_cs.c Mar 16 12:25:39 its enough to build kernel module but I probably need to add all CFLAGS etc? Mar 16 12:37:12 * france is away: Away Mar 16 13:36:19 morning Mar 16 13:40:50 hey chouimat Mar 16 13:41:36 03frederic 07org.oe.oz354fam083 * r86ec76fe... 10/packages/xserver/xserver-kdrive_20050207.bb: xserver-kdrive: simpad doesn't need disable-apm.patch after all Mar 16 13:43:20 * chouimat kicks monotone ... Mar 16 13:59:53 03hrw 07org.oe.oz354fam083 * r10b3ccdd... 10/packages/meta/gpe-image.bb: Mar 16 13:59:53 gpe-image: OpenZaurus 3.5.4x does not have separate feed for GPE Mar 16 13:59:53 - one feed for X11 stuff like in Familiar Mar 16 14:00:00 03hrw 07org.oe.dev * r891ff1a9... 10/packages/sharp-binary-only/ (2 files in 2 dirs): sharp-aticore-oss: patch makefile to get it working with make 3.81 Mar 16 14:03:14 time to remove some old builddirs... Mar 16 14:03:19 no more 3.5.3 upgrades Mar 16 14:15:54 hi folks Mar 16 14:17:46 03koen 07org.oe.dev * r39d6becf... 10/packages/gcc/gcc-cross-sdk_4.1.0.bb: gcc-cross-sdk: add 4.1.0 Mar 16 14:35:44 03florian 07org.oe.oz354fam083 * r62d94bad... 10/conf/distro/preferred-gpe-versions-2.7.inc: preferred-gpe-versions-2.7.inc: Update libgpepimc, add libeventdb. Mar 16 14:35:48 03florian 07org.oe.oz354fam083 * rdd70dce4... 10/conf/distro/familiar-0.8.3.conf: familiar-0.8.3.conf: Add SDK compiler version. Mar 16 14:35:52 03florian 07org.oe.oz354fam083 * r1978e2f9... 10/ (7 files in 2 dirs): gpe-filemanager: Update 0.25 fixing several minor issues. Drop obsolete releases. Mar 16 15:02:14 moo moo Mar 16 15:50:35 moo Mar 16 15:57:08 ~curse httrack Mar 16 15:57:15 May you be reincarnated as a Windows XP administrator, httrack ! Mar 16 16:06:21 03pH5 07org.oe.dev * rf395d2cb... 10/site/arm-linux-gnueabi: site/arm-linux-gnueabi: add some ac_cv_sizeof to make rxvt-unicode and lzo build Mar 16 16:34:43 Any ideas about this error (I presume it's a simple explanation)?: http://pastebin.ca/45930 Mar 16 16:37:52 lardman, do you set BBPATH ? Mar 16 16:38:22 leoncamel: yep, BBPATH="/home/enpsgp/dev/build-c7x0-gpe:/home/enpsgp/dev/mt/org.openembedded.dev" Mar 16 16:38:41 hmm, should have a packages on the end of that...? Mar 16 16:39:35 ah, it looks like I've not set it in my local.conf Mar 16 16:40:28 though it's not set in the example local.conf either. Any suggestions? Mar 16 16:42:26 lardman, your BBFILES is setted correctly ? Mar 16 16:42:45 lardman: its ok Mar 16 16:42:49 lardman, paste your local.conf please . Mar 16 16:43:51 http://pastebin.ca/45932 Mar 16 16:45:47 ~botmail Mar 16 16:46:31 botmail read Mar 16 16:46:36 ~botmail read Mar 16 16:47:20 ~botmail read Mar 16 16:49:08 cya Mar 16 17:01:14 where do i tell OE to buit the hole build process ? Mar 16 17:01:27 s/built/put/ Mar 16 17:01:37 s/bult/put/ Mar 16 17:01:43 cu Mar 16 17:06:48 by default it's built in the ./tmp directory in the directory you call bitbake from Mar 16 17:12:14 lardman: can i specify that in a config file ? Mar 16 17:12:24 local.conf :-) Mar 16 17:13:32 yes, there is a setting: TMPDIR = /usr/local/projects/oetmp Mar 16 17:14:08 thanks Mar 16 17:14:16 np Mar 16 17:14:43 reading the file would also help, sorry Mar 16 17:15:15 don't worry, I've no idea why mine's not working so I've read it a few times ;) Mar 16 17:15:46 :-) Mar 16 17:17:34 What is diffstat used for (as in why is it a runtime dep, which I don't seem to have)? Mar 16 17:21:21 hi, folks, I build bootstrap-image failed while compiling glibc . Mar 16 17:21:24 I have "Nothing provides" errors for diffstat-native, autoconf-native, pkgconfig-native & ipkg-native. Are these supposed to be provided by .bbs now? How do I remove these errors? Mar 16 17:21:25 here is the log : http://pastebin.com/605740 Mar 16 17:22:18 I see that there is no -native package for diffstat, but the others have -native packages afaics Mar 16 17:22:47 lardman: your local.conf is broken Mar 16 17:23:11 compare it to local.conf.sample Mar 16 17:23:22 koen: it looked pretty similar to me Mar 16 17:23:47 koen: but I'll have a closer look Mar 16 17:24:31 hi, koen. could you have a look at http://pastebin.com/605740 ? Mar 16 17:24:34 pay special attention to stuff starting with ASSUME_PROVIDED ;) Mar 16 17:25:30 leoncamel: I don't know why that happens Mar 16 17:27:33 koen: The example local.conf has no ASSUME_PROVIDED lines uncommented... Mar 16 17:28:38 and your local.conf has? Mar 16 17:28:56 Mine has one: ASSUME_PROVIDED = "virtual/arm-linux-gcc-2.95" Mar 16 17:29:02 yup Mar 16 17:29:04 that wrong Mar 16 17:29:07 in two ways Mar 16 17:29:21 1) it's a crappy gcc which deserves to die Mar 16 17:29:23 ah, += Mar 16 17:29:35 2) it overrules ASSUME_PROVIDED = "foo" in bitbake.conf Mar 16 17:29:42 thanks Mar 16 17:31:05 * koen watches the dual opteron rip through an EABI gpe-image build Mar 16 17:32:05 * lardman looks forward to EABI -> vfp soft float -> Octave working correctly Mar 16 17:33:04 So, supposedly X6.9 and X7.0 are identical except for the build systems, right? Mar 16 17:33:16 So if a driver isn't in 6.9, it won't be in 7.0 either? Mar 16 18:08:25 hi everyone Mar 16 18:08:36 greets Mar 16 18:09:06 koen: thanks for the GFORTRAN hint, I'll try that during the next rebuild. Mar 16 18:09:50 koen: is there a reason that PREFERRED_PROVIDER_libx11=diet-x11 is commented out in angstrom-2006.9.conf? Mar 16 18:09:58 hey CosmicPenguin Mar 16 18:13:07 hi, which quilt version is neccessary for OE ? i get often: Usage: quilt [--trace[=verbose]] [--quiltrc=XX] command [-h] ... Mar 16 18:15:59 pH5: I suspect I wanted to build something with a 'real' x11 Mar 16 18:16:05 pH5: I didn't mean to commit that Mar 16 18:16:14 grma: iirc 0.42 Mar 16 18:17:14 koen: ok then is my error somewhere else :-) Mar 16 18:18:49 zecke: I found another problem with bitbake: 'bitbake gpe-image' -> rxvt breaks, 'bitbake gpe-image' again -> it starts with do_rootfs while a big part of the RDEPENDS are missing Mar 16 18:20:11 koen: oh :( Mar 16 18:21:07 zecke: with r384 Mar 16 18:26:24 hey gizmo1 CosmicPenguin: correct, X0rg 6.9 and 7.0 should be identical (minus the buildsystem) Mar 16 18:26:50 Bug 636 doesn't seem to be fixed in .dev Mar 16 18:26:51 hrm Mar 16 18:26:59 * koen has trouble typing today Mar 16 18:27:21 * pH5 just had to enter his password three times because of similar problems Mar 16 18:28:20 03pH5 07org.oe.dev * r316d86b6... 10/conf/distro/angstrom-2006.9.conf: uncomment PREFERRED_PROVIDER_libx11 = "diet-x11" Mar 16 18:29:26 hmm, ignore me, looks like there's some other error with it Mar 16 18:29:56 it seems most stuff in .dev likes eabi Mar 16 18:30:01 except osb-nrcore Mar 16 18:30:57 pH5: Paul Brook (of CSL fame) mentioned that gcc 4.1 doesn't have real EABI Mar 16 18:31:07 it doesn't turn on thumb-interworking Mar 16 18:31:25 but their ueber-cool CSL toolchain will have it working Mar 16 18:31:47 to quote lennert: "If gcc would work, CSL would be out of business" Mar 16 18:32:06 Any ideas what this error indicates?: http://pastebin.ca/index.php Mar 16 18:32:10 oops Mar 16 18:32:20 koen: is rxvt broken? (so I can test) Mar 16 18:32:24 http://pastebin.ca/45938 even Mar 16 18:32:28 zecke: it isn't anymore Mar 16 18:32:42 koen: how do I get it broken? Mar 16 18:32:55 koen: build failure? Mar 16 18:33:04 build for eabi and revert pH5 addition to the site files Mar 16 18:33:20 EABI is unsupported by bitbake ;) Mar 16 18:33:26 lol Mar 16 18:33:33 zecke: or just cp rxvt-unicode_5.6.bb rxvt-unicode_7.2.bb Mar 16 18:34:26 lardman: you probably have a stale checkout in your DL_DIR Mar 16 18:34:29 koen: I plan to push some test cases to the bitbake module Mar 16 18:34:41 koen: ok, cool Mar 16 18:35:26 hi ! I have still the problem with pkgconfig-native: http://de.pastebin.ca/45939, has someone a quick help ? Mar 16 18:36:23 * koen tries to remember what caused that error Mar 16 18:36:38 grma: could you remove the tmp/cache directory? Mar 16 18:36:43 ok Mar 16 18:37:40 pH5: with eabi you should be able to use iwmmxt tuning without problems Mar 16 18:38:18 koen: yay, gentoo style. Mar 16 18:38:26 heh Mar 16 18:38:49 * koen wonders why gcc-cross-sdk_4.1.0.bb fails in strange ways Mar 16 18:38:50 bye guys Mar 16 18:38:57 cu lardman Mar 16 18:38:59 * pH5 includes tune-iwmmxt in ipaq-pxa270.conf Mar 16 18:41:14 does anybody know why the samba packages have "S := ${WORKDIR}/${P}/source" instead of "S = ..."? Mar 16 18:41:27 no idea Mar 16 18:42:40 This breaks with my modified multimachine.inc. It installs the package to work/arm-none-linux-gnueabi and then tries to patch it in work/arm-linux-gnueabi... Mar 16 18:43:51 hrm Mar 16 18:44:19 * koen still wants a clean ldflags fix for gcc Mar 16 18:44:25 before I resort to sed Mar 16 18:44:46 after that a nice fix for the glibc "can't find headers" problem Mar 16 18:47:15 real 0m8.728s Mar 16 18:47:17 hmm Mar 16 18:54:24 koen: did we have this -lmpfr problem with both gcc-cross-initial and gcc-cross? Mar 16 18:54:48 pH5: iirc only in -cross Mar 16 18:54:57 initial doesn't build some stuff Mar 16 18:55:00 bbl Mar 16 19:02:20 hi all Mar 16 19:23:26 hey Mardy Mar 16 19:31:23 koen: http://en.pastebin.ca/45945 Mar 16 19:32:00 pH5: nice! Mar 16 19:32:25 I suspect the same should be added to gcc-cross-sdk Mar 16 19:37:58 pH5: I just pushed a cset to disable fortran in gcc 4.1.0 Mar 16 19:39:52 03koen 07org.oe.dev * rd86e2f15... 10/packages/gcc/gcc_4.1.0.bb: gcc 4.1.0: disable fortran since it is breaking the build Mar 16 19:40:40 koen: OE sucks because we have no working fortran? Mar 16 19:41:02 zecke: no, OE sucks regardless Mar 16 19:41:17 OE sucks because it isn't made my mvista Mar 16 19:41:19 koen: right, but now they have an argument Mar 16 19:41:32 koen: "OE sucks because hey look there is monkey" Mar 16 19:41:47 "look behind you a 3 headed monkey" Mar 16 19:42:09 ah, the memories. Mar 16 19:42:27 zecke: http://en.wikipedia.org/wiki/Monkey_Island Mar 16 19:42:52 matthew garret used it in his hal talk quite nicely Mar 16 19:43:06 hijacker Mar 16 19:43:31 was a pretty bad Nokia talk :} Mar 16 19:44:06 the topic was interesting, but they didn't have any news Mar 16 19:44:57 bbl Mar 16 19:44:58 koen: "HAL without any requirements (... beside HAL,DBUS,libexpat,libxerces-c,libusb,libpci,libscsi,xml data en mass) Mar 16 19:45:46 hal depends on itself? Mar 16 19:45:56 koen: sure Mar 16 19:46:07 sounds like mickeylware Mar 16 19:47:10 koen: you need n HAL's installed Mar 16 19:47:20 koen: because not all features are working in one Mar 16 19:47:21 where n = 2001? Mar 16 19:47:31 koen: n is a power of two Mar 16 19:48:06 it took me a while to figure out what NPOT meant Mar 16 19:48:19 koen: if you have version 2^(n-1) you need 2^n versions as well to recover from the breakage Mar 16 19:48:30 * RP returns Mar 16 19:48:34 hey REdOG Mar 16 19:48:36 ehm Mar 16 19:48:38 hey RP Mar 16 19:48:46 minus my luggage again :-/ Mar 16 19:48:56 D'oh! Mar 16 19:48:58 hey woglinde Mar 16 19:49:09 he koen Mar 16 19:49:32 Apparently this is normal for Air France... Mar 16 19:50:07 RP: you should be happy that they have not relabeled your luggage Mar 16 19:50:13 RP: they could have returned anything Mar 16 19:50:20 Air France those have the most legroom during flights Mar 16 19:50:41 s/those/does/ Mar 16 19:51:24 which they compensate by using tiny aircraft which are too low for people > 1.75m Mar 16 19:51:36 zecke: That might still happen :-/ Mar 16 19:51:52 woglinde: we should organize a OE event Mar 16 19:51:55 koen: I'm not sure I'd agree with that as I found their aircraft uncomfortable :-/ Mar 16 19:52:02 woglinde: and play Hacker Jeapardy all night long Mar 16 19:52:05 RP: in other news, pH5 booted an EABI rootfs Mar 16 19:52:11 zecke at your house? Mar 16 19:52:16 koen: How far did it get? Mar 16 19:52:33 till login Mar 16 19:52:38 woglinde: did your flat burn down? or where was the fire? Mar 16 19:52:47 koen: cool :) Mar 16 19:52:53 zecke do you have children? Mar 16 19:52:59 *g* Mar 16 19:53:22 RP: http://en.pastebin.ca/45835 Mar 16 19:53:25 that's the bootlog Mar 16 19:53:28 woglinde: Al didn't mention such things Mar 16 19:53:56 koen: "UFF"? Mar 16 19:54:29 probably garbage on seria; Mar 16 19:54:31 koen: very nice. I'll have to experiment with this :) Mar 16 19:55:08 Sadly I never found the time I thought I might have for OE hacking whilst away :-/ Mar 16 19:55:23 I promised mallum to discuss how to handle eabi with zaurus kernels Mar 16 19:56:42 let's hope the old-abi-compat bits work as expected Mar 16 19:56:55 koen: Lets hope so, yes Mar 16 19:59:05 it *almost* builds without intervention in OE Mar 16 20:01:51 "almost"s have been giving me lots of problems recently :) Mar 16 20:04:21 * RP wishes the computer monitor would stay still. I think I'd best go to bed or something... Mar 16 20:04:43 RP: cya! Mar 16 20:05:06 sorry to ask but i'm the only one who have problem building boostrap-image with org.dev.oe, I got error on glibc-2.3.5+cvs20050627-r3 Mar 16 20:08:18 03koen 07org.oe.dev * rcc41aded... 10/packages/gcc/gcc-cross_4.1.0.bb: gcc-cross 4.1.0: add fix for the mpfr problem, courtesy Philipp Zabel Mar 16 20:09:11 according to my notes only the kernel header problem is left Mar 16 20:12:29 RP: from today I'm free from work, so I'll have time to try implementing accels for the w100 Mar 16 20:13:21 Mardy: in kernel or X? Mar 16 20:14:03 koen: in kernel first, for the console Mar 16 20:14:40 yes, faster scrolling! Mar 16 20:14:44 then I'd like to add them to X, and maybe directfb and sdl Mar 16 20:16:52 Mardy: I would be interested in the performance of DirectFB Mar 16 20:17:56 I'm still unsure, whether to load ATI's microcode in the FPGA, or to program the chip by direct register access Mar 16 20:18:06 zecke: I'd think that the optimized arm memcpy would be also a nice speed boost Mar 16 20:18:09 I guess the FPGA is faster Mar 16 20:19:16 koen: on StrongARM two years ago glibc was pretty fast already :} Mar 16 20:19:33 koen: or sandman failed - unlikely Mar 16 20:20:15 pb_: I need your englizh skillz later tonight Mar 16 20:20:29 according to a followup glibc doesn't have optimized mem* for arm Mar 16 20:20:37 haven't checked though Mar 16 20:29:04 hey pH5 Mar 16 20:31:02 time to watch 'Prison Break' Mar 16 20:31:59 Enjoy your hour of crap Mar 16 20:35:07 hi Mar 16 20:38:03 hey hrw Mar 16 20:38:17 someone has cfxga card? Mar 16 20:39:47 hi greentux Mar 16 20:39:52 hi greentux Mar 16 20:41:38 http://pdashop.nl/images/product/image.php?id=18794 Mar 16 20:41:43 that would be nice to have Mar 16 20:42:30 ARGH!!!!!! Mar 16 20:42:42 'clean task-bootstrap' cleaned all its dependencies ;( Mar 16 20:42:55 * hrw send some hate letters to bitbake Mar 16 20:42:56 that's a 'feature' of the new bitbake Mar 16 20:43:04 * koen|tv hates it with a vengeance Mar 16 20:43:16 mickeyl: please add a send hate letter in bitbake shell Mar 16 20:43:21 and group of dress-waring neckless people Mar 16 20:43:23 mickeyl: it seems to be a popular request Mar 16 20:44:12 we need fileclean, clean, thisoneonlyclean Mar 16 20:44:18 I'm rewriting the buildscripts anyway Mar 16 20:45:01 back to old days when rm tmp/stamps/ was norm... Mar 16 20:48:37 hrw: koen|tv : ATM I'm documenting the fetcher, I will write more tests tonight Mar 16 20:48:54 hrw: koen|tv : Hopefully including tests for the issues we currently see (RDEPENDS) Mar 16 20:49:02 great Mar 16 20:49:28 hrw: we now have a bitbake-tests module. Where we can create all the tests we want Mar 16 20:49:43 currently my favorite bitbake version is r329 (before rdepends) Mar 16 20:49:57 RP: We need to change that! Mar 16 20:50:09 hrw: mine is r666 but we are getting there Mar 16 20:50:19 their... Mar 16 20:50:21 I did lot of 666 today Mar 16 20:50:37 hrw: Mddi Mar 16 20:50:40 Doing the devils work again? Mar 16 20:50:41 Mein Benteley? Mar 16 20:50:41 ~curse KT badly Mar 16 20:50:43 May you be reincarnated as a Windows XP administrator, KT badly ! Mar 16 20:51:00 KT is Polish distributor of VW & Audi Mar 16 20:54:36 poor guy Mar 16 21:08:32 pH5: what do you think about copying the kernel headers into place inside the glibc .bb? Mar 16 21:09:57 is that due to the pending death of linux-libc-headers? Mar 16 21:10:17 koen|tv: then it should drop the dependency on llh Mar 16 21:10:28 true Mar 16 21:10:33 pH5: or we make our own llh-2.6.16-rc6-fake package Mar 16 21:10:55 that's also an option Mar 16 21:11:09 iirc it's the last 'problem' for EABI Mar 16 21:11:19 (and kernel support) Mar 16 21:14:11 hi zecke :) Mar 16 21:14:25 CosmicPenguin: 'crap' is a bit too strong, 'predictable' is more like it Mar 16 21:14:52 CosmicPenguin: hehe, it survived a long time :} Mar 16 21:16:19 Indeed Mar 16 21:16:31 There is now talk about keeping it alive if SuSE and Redhat promise to use it Mar 16 21:27:31 * koen|tv gets reminded to the smelliness of rpms Mar 16 21:33:02 NOTE: package osb-nrcore-0.5.0+cvs20060212-r0: task do_compile: failed Mar 16 21:33:07 yep Mar 16 21:34:37 pH5: like this http://pastebin.com/606225 ? Mar 16 21:35:15 koen: exactly like this. Mar 16 21:36:20 stooopid c++ Mar 16 21:36:57 hehe Mar 16 21:45:12 koen: could this help: http://en.pastebin.ca/45958 ? Mar 16 21:46:02 looking at c++ makes my eyes bleed Mar 16 21:46:07 * koen gets a tissue Mar 16 21:46:23 koen: don't look then, just feel it. Mar 16 21:47:16 if it builds....... Mar 16 21:50:11 pH5: did you enable eabi in the hx4700 defconfig? Mar 16 21:51:21 koen: locally http://en.pastebin.ca/45959 Mar 16 21:51:33 actually, this looks like it's safe to submit. Mar 16 21:52:37 with old-abi compat turned on, yes Mar 16 21:53:01 won't help much until we have eabi/oabi patches or 2.6.16-rc6 in cvs, though. Mar 16 21:53:23 or in OE Mar 16 21:54:52 osb-nrcore seems to build fine now. as always, it takes roughly infinite time to finish. Mar 16 21:58:44 pH5: that should be all that's needed for an eabi gpe-image Mar 16 21:59:42 koen: sounds good. only given the speed of my cpu, I won't finish this today. Mar 16 22:00:46 NOTE: package osb-nrcore-0.5.0+cvs20060212: completed Mar 16 22:00:54 we're just two patches away from building angstrom testing images :) Mar 16 22:03:44 koen: which two patches do you have in mind? Mar 16 22:07:03 pH5: the webcore one and the one to stage the eabi headers Mar 16 22:10:58 NOTE: package gpe-image-1.0-r19: task do_rootfs: started Mar 16 22:11:23 looks like I might be wrong after all Mar 16 22:11:57 you might have hit a bitbake bug Mar 16 22:18:12 NOTE: package gpe-image-1.0: completed Mar 16 22:21:31 hi Mar 16 22:22:32 hi CoreDump|home Mar 16 22:23:14 hey CoreDump|home Mar 16 22:27:14 03pH5 07org.oe.dev * r8fed4c5c... 10/packages/gtk-webcore/ (files/KWQKURL-urlcmp.patch osb-nrcore_20060212.bb): osb-nrcore_20060212: add build fix for gcc 4.1 Mar 16 22:39:17 * ArjanS looks forward to mt 0.26 (have to update for 302 new revs now :( ) Mar 16 22:52:53 koen: not quite there yet. my kernel doesn't like to load modules and doesn't have have socket support among others: http://en.pastebin.ca/45963 Mar 16 22:53:16 good night all Mar 16 23:01:02 * koen heads to bed as well Mar 16 23:01:06 'night all Mar 16 23:02:31 night koen Mar 16 23:04:22 03koen 07org.oe.dev * r366e0044... 10/packages/gobby/net6_1.2.2.bb: net6: add 1.2.2 Mar 16 23:48:36 koen|sleep: bitdoc works now again Mar 16 23:48:46 I need proof reading and fixes for the fetcher documentation Mar 17 00:03:14 ~lart scap badly for fscking c7x0 scaps Mar 17 00:03:14 * ibot strangles scap badly with a 9-pole serial cable for fscking c7x0 scaps Mar 17 00:03:51 its useless on that machines Mar 17 00:04:08 hrw: scap itself or the Opie plugin? Mar 17 00:04:27 scap itself Mar 17 00:04:42 nevermind do I use 'scap' in shell or opie-screenshotapplet Mar 17 01:05:32 cu all Mar 17 02:07:54 woohoo Mar 17 02:08:14 i now know which commit killed things :) Mar 17 02:08:28 now to reverse the badness :) Mar 17 02:26:24 is anyone using oe for powerpc? Mar 17 02:30:39 building or target? Mar 17 02:35:07 target Mar 17 02:35:10 what is ac_cv_sizeof... hmmm Mar 17 02:35:14 (sorry for not specifying) Mar 17 02:36:48 emte: which commit killed the goodness? Mar 17 02:36:50 jacques, i cannot think of a project offhand that is specifically for ppc, so taht doesnt mean one doesnt exist Mar 17 02:37:10 jnc, it was a merge commit Mar 17 02:37:19 emte, there is only one powerpc in machines dir, and it might be old - I am having trouble getting a powerpc toolchain to build Mar 17 02:37:25 as I would expect if it's unused Mar 17 02:37:27 i am currently sorting out which part did it Mar 17 02:37:54 jacques, what distro are you trying to build? Mar 17 02:38:11 and how far do you get? Mar 17 02:38:52 emte, it dies in building glibc-initial (I don't even know why it tries to built that, as the arm dists don't have it) Mar 17 02:39:19 I was trying to build a variant of slugos, but that doesn't matter much as it's not even getting through the toolchain Mar 17 02:40:13 jacques, the arm dists build glibc-initial Mar 17 02:40:34 they do? are you sure? Mar 17 02:40:39 but if your getting that far i susupect it might be the version of glibc you have specified Mar 17 02:40:47 yes Mar 17 02:40:59 glibc currently builds three times Mar 17 02:41:17 its one of the cyclic compile packages Mar 17 02:41:30 but there is no glibc-initial dir in tmp/work Mar 17 02:41:36 you need glibc to compile glibc to compile a target glibc Mar 17 02:42:04 ? Mar 17 02:42:07 but wouldn't there be a glibc-initial dir in tmp/work then? Mar 17 02:42:11 what is your specific error? Mar 17 02:42:25 I am looking at a completed debianslug build tree Mar 17 02:42:37 there is no glibc-initial dir in tmp/work Mar 17 02:42:52 * emte wonders how old a repo you r looking at Mar 17 02:43:04 march 2nd Mar 17 02:43:13 newer ones are devided up more now Mar 17 02:43:52 emte: *which* one Mar 17 02:44:13 ls tmp/work/ Mar 17 02:44:13 arm-linux/ h3600-linux/ i686-linux/ Mar 17 02:44:24 is specific for mine Mar 17 02:44:29 ok that's new to me Mar 17 02:44:41 I just have tmp/work and all the package build dirs Mar 17 02:44:55 but still no glibc-initial Mar 17 02:44:57 i think you may need to update bitbake Mar 17 02:45:23 I have watched slugos/debianslug build probably 80 times, and I don't even remember seeing glibc-initial Mar 17 02:45:38 unless your using an sdk i find it very odd you have the old setup Mar 17 02:46:44 the nslu2/slugos/debianslug stuff is kept current Mar 17 02:46:50 jnc, 2310202db82b311ad297541e20f94d8bf5706860 is the merge ... the history was split for quite a while which is taking me a bit to sort out Mar 17 02:47:23 jacques, i suspect it is, but when was the last time you updated your bitbake? Mar 17 02:48:08 emte: what goodness was killed? Mar 17 02:48:12 the nslu2 stuff is self-contained, and I always start from scratch - I have learned what can happen with mismatched versions Mar 17 02:48:45 jnc, something introduced libgcc issues Mar 17 02:48:52 oh hey yeah Mar 17 02:48:55 i noticed that too today Mar 17 02:48:59 and i made a post on the ML Mar 17 02:49:06 didn't you get the TPS reports? Mar 17 02:49:16 bitbake:__version__ = "1.3.2" Mar 17 02:49:32 jnc, there are a bunch of posts from me on the oe mailing list Mar 17 02:49:54 as well bug 747 is filed for it Mar 17 02:50:00 okay Mar 17 02:50:59 i need to free up some drive space before i can compile again ... Mar 17 02:51:03 :( Mar 17 02:51:49 actually ... i'll just shutdown mozilla :) Mar 17 03:03:53 use links Mar 17 03:06:24 lol Mar 17 03:06:49 yarssr is set to open mozilla when i click a commit Mar 17 03:07:31 jnc, btw your issue is not related Mar 17 03:07:45 infact i dont know how your missing libgcc Mar 17 03:08:10 its needed to compile glibc and other core libraries Mar 17 03:08:32 no, it's not Mar 17 03:08:45 the lib was libmygcc Mar 17 03:08:55 what is libmygcc? Mar 17 03:09:12 i don't know, but it appeared to be either mysql or gcc Mar 17 03:10:28 searching returns that its probably gcc with a prefix Mar 17 03:10:58 hard to say tho Mar 17 03:11:27 yeah searching libmygcc gives a lot of mysql stuff ... Mar 17 04:12:06 cd .. Mar 17 04:13:41 no thanks Mar 17 04:13:47 ? Mar 17 04:17:26 What is the recomended host gcc version to use for oe builds? Mar 17 04:18:13 3.4.4 I believe Mar 17 04:18:21 3.4.x Mar 17 04:20:28 it depends what your looking for Mar 17 04:20:46 JustinP, you busy for a few min? Mar 17 04:22:41 Danke. Mar 17 04:23:14 can you take a read over this and tell me what you think? Mar 17 04:23:16 http://bugs.treke.net/show_bug.cgi?id=747 Mar 17 04:29:57 ummmm....looks like a bug all right... Mar 17 04:30:00 looks ok to me Mar 17 04:30:33 more curious on your thoughts about introducing -semistable Mar 17 04:31:57 seems useful to me Mar 17 04:32:21 yeah it makes sense to me for the reasons i'll expressed Mar 17 04:32:41 but its not exactly a practice that has been followed in the past Mar 17 04:33:38 i see it as one of the largest issues currently with oe Mar 17 04:33:47 things suddenly get broken Mar 17 04:34:10 even in what are concidered stable branches Mar 17 04:34:37 i know it cannot always be avoided Mar 17 04:35:13 but if we can reduce it and resolve the issues faster it would possibly entice more developers Mar 17 04:38:38 * emte is done wandering through the empty hallways of my brain for five min Mar 17 04:39:50 i am just happy OE builds again with a current repo Mar 17 04:51:19 hey mithro, you busy? Mar 17 05:07:30 emte: use stable Mar 17 05:10:06 jnc, i am Mar 17 05:11:02 oh :/ Mar 17 05:11:02 thats the issue Mar 17 05:23:06 i find that i need applications on my handheld like spreadsheet, text editor, and some sort of password keyring Mar 17 05:29:22 install some then ... Mar 17 05:29:31 they should be in your world feed Mar 17 05:35:27 i'm having trouble filling the keyring thing for GPE Mar 17 05:35:46 there's a number of applications i could use if i was in an OPIE environment Mar 17 06:29:45 good morning Mar 17 06:34:29 hi Mar 17 06:34:44 lack of suspend working on my C3000 keeps me from using it seriously Mar 17 06:34:47 :/ Mar 17 06:34:53 i'm selling it Mar 17 06:53:46 jnc, i am curious if your aware taht opie has far more devs than gpe does Mar 17 06:54:48 i wouldnt mind that device ... if i were rich Mar 17 07:01:21 emte: i'll probably let it go for $350usd to a coworker Mar 17 07:01:31 i paid 500 for it, i think Mar 17 07:03:09 yeah i am (poorly) trying to scratch up enough for some type of sub-notebook Mar 17 07:25:23 erm, sylpheed has permissions issues with bind and stuff on openzaurus Mar 17 07:25:28 i think the issues are with openzaurus Mar 17 07:25:46 normal users for the gpe-image target do not have permission to ping or bind sockets Mar 17 07:25:55 no idea what that is controlled by Mar 17 07:29:10 ofcourse normal users dont Mar 17 07:29:14 why would they? Mar 17 07:29:37 ping is normally setuid root. only root can open raw sockets. thats true on any distro Mar 17 07:29:47 hey kergoth Mar 17 07:29:50 hey Mar 17 07:29:54 i found the problem btw Mar 17 07:30:11 the new binutils has some bad magic Mar 17 07:30:24 ah Mar 17 07:31:01 i was thinking the other day ... Mar 17 07:31:27 how hard do you think it would be to teach bitbake to read from a db? Mar 17 07:35:11 hmm Mar 17 07:35:29 this is taking a while .... .ccache must be a few gb ... Mar 17 07:40:58 morning Mar 17 07:41:40 morn Mar 17 07:51:33 hi all Mar 17 08:07:23 nice, gnumeric built okay Mar 17 08:08:36 but does it run ... Mar 17 08:08:53 not sure Mar 17 08:09:02 anyway sleep time for me Mar 17 08:09:05 o? Mar 17 08:09:16 i'm having problems using my Z, like i said, until the suspend/resume is fixed Mar 17 08:09:34 we will see what happens with my bug tomorrow Mar 17 08:18:27 emte: if you want smoething stable, don't use .dev Mar 17 08:18:35 and good morning all Mar 17 08:18:54 emte: also note that familiar is unmaintained in .dev Mar 17 08:44:01 morning Mar 17 08:44:35 hey hrw|work Mar 17 08:46:30 NOTE: package gpe-image-1.0: completed Mar 17 08:46:48 * koen admires the EABI goodness in deploy/images/c7x0 Mar 17 08:47:55 morning Mar 17 08:48:01 he XorA Mar 17 08:48:09 oooh, EABI angstrom images Mar 17 08:48:31 I did forget to patch the kernel, so it worn't boot Mar 17 08:48:37 but at least it builds Mar 17 08:50:04 heh heh Mar 17 08:53:32 * koen runs bitdoc to generate some docs Mar 17 09:00:53 koen: today i tried a task-bootstrap, i think i make something wrong: http://de.pastebin.ca/46013 Mar 17 09:00:59 koen: any idea ? Mar 17 09:02:46 could it be that OE dosen not like it when using symlinks ? Mar 17 09:02:48 grma: you have a symlink in your path and libtool doesn't like that Mar 17 09:03:54 ahh Mar 17 09:03:57 thanks Mar 17 09:08:40 <_law_> hmm :-( http://pastebin.com/607015 gnu-config-native problem Mar 17 09:09:24 try rebuilding it Mar 17 09:10:25 hail zecke Mar 17 09:11:49 aaargh Mar 17 09:11:49 NameError: global name 'svn_rsh' is not defined Mar 17 09:11:55 * koen stabs bitbake r399 Mar 17 09:14:10 zecke: it seems someone killed the svn fetcher Mar 17 09:14:49 zecke: to reproduce: bitbake -b gnome/gnome-vfs-dbus_svn.bb Mar 17 09:23:50 ~lart perl build Mar 17 09:23:50 * ibot breaks out the Hoover and sucks up perl build Mar 17 09:23:57 its full of hardcoded paths Mar 17 09:27:52 I suppose one thing about gentoo multilib is it exposes all these fsckups Mar 17 09:33:12 damn damn damn Mar 17 09:33:29 ~fishslap zecke Mar 17 09:33:34 * ibot slaps zecke up side the head with a wet fish. Mar 17 09:34:00 koen: don't use a broken bitbake? Mar 17 09:35:02 Good morning Mar 17 09:35:55 morning all Mar 17 09:36:02 bonjour le normand Mar 17 09:36:43 Genesis: hey hey ! one more french here ! Mar 17 09:37:18 one more normand :) Mar 17 09:37:19 alan|home: you are enough? Mar 17 09:37:47 quick someone disinfect the channel :-) Mar 17 09:37:50 alan|home, : i come from caen :) Mar 17 09:38:21 Genesis: indeed, we are quite close... Mar 17 09:38:33 XorA: :p Mar 17 09:39:06 Genesis: first time i see you here... are you a new OE fan ? Mar 17 09:39:14 yes Mar 17 09:39:26 XorA: what did you do to my french friend obergix ?? Mar 17 09:39:44 Genesis: congrats and welcome Mar 17 09:40:03 i block at UART boostrap Mar 17 09:40:36 f*cking proprio tools i've no key Mar 17 09:40:42 UART ? don't know what this is... (^^)' Mar 17 09:40:56 * alan|home not a dev... Mar 17 09:41:00 oki Mar 17 09:41:27 i shall make a e-book plateform with the mx21 Mar 17 09:41:32 woo another perl cross compile bug fixed Mar 17 09:41:44 if someone work with that , tell me i really need help :) Mar 17 09:42:10 koen: sorry, koen, fixed now Mar 17 09:42:41 koen: the code I killed was CVS specific, I didn't expect it was used in the fetcher ata ll Mar 17 09:47:23 haaa...obergix[work] has been released by XorA... Mar 17 09:48:12 eeeep, more of them Mar 17 09:49:14 * XorA needs an amd64 in a pizzabox case to carry to work Mar 17 09:49:20 god, i love XGL... wonder if xegl will work one day on a zaurus... would need a woking 3D first... Mar 17 09:49:50 XorA: me too... :) Mar 17 09:50:57 ~seen proti Mar 17 09:51:15 proti was last seen on IRC in channel #asterisk, 104d 10h 16m 19s ago, saying: 'I want asterisk to register as From: <01233> ? Can't figure the correct parameters ?'. Mar 17 09:53:09 * Bernardo is having problems with kernel 2.4 vs kernel 2.6 in a module .bb Mar 17 09:55:02 anyone can tell me why on_runmake will call "arm-linux-gcc" and make will call "arm-linux-gcc-2.95" (the correct compiler for kernel 2.4.20 on a akita)? Mar 17 09:56:15 did you inherit module or module-base? Mar 17 10:02:05 module Mar 17 10:03:05 I also need ot detect before the build the kernel version, one of the parameters to pass to the makefile is either 4 or 6 depending on kernel evrsion being 2.4.x or 2.6.x Mar 17 10:07:19 use KERNEL_MAJOR_VERSION iirc Mar 17 10:07:26 Bernardo: look t bug #524 Mar 17 10:16:11 * france is back (gone 21:39:07) Mar 17 10:18:28 hrw|work: thanks Mar 17 10:19:35 hrw|work: I don't think bug #524 applies, as I have different build areas for each version Mar 17 10:19:49 Bernardo: you can find how to check 2.4/2.6 there Mar 17 10:20:24 you're right, sorry Mar 17 10:20:30 KERNEL_MAJOR_VERSION is 2.4/2.6 Mar 17 10:20:54 * Bernardo goes for his morning coffee Mar 17 10:36:21 does anyone know about any success with cacao jvm and arm arquitecture? Mar 17 10:48:57 * france is away: Away Mar 17 10:55:30 someone know smallest-possible library which will provide gfx resize functions? Mar 17 10:55:53 hrw|work: scaling of images? Mar 17 10:56:35 epeg? Mar 17 10:56:57 imlib1 Mar 17 10:57:01 thx Mar 17 10:57:23 should I report every error obtained while compiling a full gpe-image with gcc 4.1, or is it crazy? Mar 17 10:57:28 I want to experiment a bit with resizing for cfxga-util mirror function Mar 17 10:57:54 hrw|work: you have a cfxga card now? Mar 17 10:58:03 koen: no - have user with it Mar 17 10:58:12 and 6h timezone difference Mar 17 10:58:23 I noticed IO data EOL'ed it Mar 17 10:58:32 yes Mar 17 10:58:49 last ones was at trisoft and company Mar 17 10:59:16 I wonder if the kdrive bits still work Mar 17 10:59:29 Xfbdev -pcmcia iirc Mar 17 10:59:53 ugh **** ENDING LOGGING AT Fri Mar 17 10:59:56 2006