**** BEGIN LOGGING AT Tue Apr 10 02:59:57 2007 Apr 10 03:21:14 OE Build Configuration: Apr 10 03:21:15 BB_VERSION = "1.6.7" Apr 10 03:21:15 OE_REVISION = "70b9fb5430b130bda41bbfbaff2c613b0b9971df" Apr 10 03:21:15 TARGET_ARCH = "arm" Apr 10 03:21:15 TARGET_OS = "linux-gnueabi" Apr 10 03:21:15 MACHINE = "smdk2440" Apr 10 03:21:17 DISTRO = "angstrom" Apr 10 03:21:19 DISTRO_VERSION = "test-20070410" Apr 10 03:21:21 TARGET_FPU = "soft" Apr 10 03:21:23 OTE: preparing tree for binary locale generation Apr 10 03:21:25 NOTE: generating locale es_NI (UTF-8) Apr 10 03:21:27 NOTE: Task failed: localedef returned an error (command was PATH="/root/work/moko/angstrom/build/tmp/staging/i686-linux/bin/arm-angstrom-linux-gnueabi:/root/work/moko/angstrom/build/tmp/staging/i686-linux/bin:/root/work/moko/angstrom/build/tmp/cross/bin:/root/work/moko/angstrom/bitbake-1.6.0/bin:/root/work/moko/bitbake/bin:/usr/lib/qt-3.3/bin:/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/crossmoko/bi Apr 10 03:21:33 NOTE: package glibc-2.5-r5: task do_package: failed Apr 10 03:21:35 ERROR: TaskFailed event exception, aborting Apr 10 03:21:37 NOTE: package glibc-2.5: failed Apr 10 03:21:39 ERROR: Build of angstrom-x11-image failed Apr 10 03:23:33 LWT001: as everywhere dont flood Apr 10 03:24:36 sorry, Apr 10 07:07:17 morning all Apr 10 07:07:42 Ello Apr 10 07:07:47 hey RP Apr 10 07:10:22 morning' Apr 10 07:11:05 djay-il: "the tipping point" is now in my bookcase :) Apr 10 07:11:34 koen: great :-) I'm now reading his "Blink" Apr 10 07:16:15 hm, how do I set preference for package versions? Apr 10 07:19:00 morning all Apr 10 07:21:40 hi Ifaistos Apr 10 07:23:38 RP: hi ! Seems people are still enjoying Easter :) Apr 10 07:33:21 if I want to have 2 versions of the same package, how do I tell which one to use? Apr 10 07:46:56 has anyone worked with the linux driver that provides support for mmc/sd cards using GPIO pins ? Apr 10 07:54:28 morning Apr 10 07:54:59 Ifaistos: there is SPI by GPIO so you would have to sit an mmc-spi driver on top of that Apr 10 07:56:44 XorA: Thanks. it seems that the wrt guys have an mmc drivers that works directly with gpio Apr 10 08:16:01 * XorA wonders why all his -native tasks have decided to re-run Apr 10 09:01:48 did something change in way stamps are done or something, bitbake is re-running pretty much everything from do_package_write onwards Apr 10 09:17:44 XorA: Shouldn't have done... Apr 10 09:18:17 RP: btw is there a 1.8.1 release of bitbake? Apr 10 09:18:29 XorA: No, 1.8.2 is due any time Apr 10 09:18:41 XorA: Odd numbers are svn versions Apr 10 09:18:53 RP: ah ok, that was confusing me Apr 10 09:19:21 XorA: If we didn't do that, we couldn't tell when people are using svn versions... Apr 10 09:21:25 RP: makes sense Apr 10 10:25:58 rwhitby: I thought most 64 bit issues were solved Apr 10 10:26:49 zecke: people keep reporting problems that don't occur for me on Debian Etch or CentOS 4 (which is a good spread of 32-bit systems), but then report they are running 64 bit amd ... Apr 10 10:27:26 (and remember this is frozen OE version in openmoko now) Apr 10 10:28:39 they are likely to be all fixed in HEAD. Apr 10 10:29:44 well if qemu is mentioned then amd64 is pretty much a no-go Apr 10 10:30:03 I have had it working once, then versions shifted slightly and never again Apr 10 10:32:26 I've never even gotten it to compile.. Too bad, qemu is nice Apr 10 10:32:42 ~prod lrg Apr 10 10:32:50 * ibot zaps lrg with a high voltage cattle prod Apr 10 10:32:56 ouch Apr 10 10:34:11 * rwhitby steers clear of 64 bit for building stuff. Apr 10 10:34:32 my life is too short to deal with more hassles than are absolutely necessary. Apr 10 10:35:21 rwhitby: I've realized that too. I'll get some good vmware image working asap :) Apr 10 10:37:14 I run Debian Etch and CentOS 4, so that I can usually get one of them to build. Apr 10 10:38:18 CM: use a debian chroot instead of vmware Apr 10 10:38:51 CM: or ubuntu chroot (pretty much the same) Apr 10 10:39:01 XorA: Hmm.. Or gentoo ;) Apr 10 10:39:15 CM: nah avoid that unmaintained heap :-D Apr 10 10:39:22 Hehe :P Apr 10 10:39:45 although gentoo does have debootstrap and dchroot ebuilds Apr 10 10:40:12 Aha, thanks. Worth looking into :) Apr 10 10:40:13 you can't reproduce anything on Gentoo, cause every installation is different. Apr 10 10:40:39 Gentoo is the antithesis of reproducable builds. Apr 10 10:40:43 CM: couple of years ago I used a 32bit debian built with debootstrap inside a gentoo 64bit for building Apr 10 10:41:16 yeah, that would be reproducable. Apr 10 10:41:40 I'll give that a try, but first 15 days vacation in the Philippines.. ;) Apr 10 10:41:59 CM: I'm starting 10 days in Queensland from the 17th :-) Apr 10 10:42:12 * XorA has one day in Stranraer :-( Apr 10 10:42:33 * CM cheers for more vacation for everyone Apr 10 10:42:42 stranraer is a hole Apr 10 10:44:34 * CM is away on a meeting Apr 10 10:45:19 hey, whats wrong with Gentoo? Apr 10 10:45:30 * djay-il is using one Apr 10 11:37:30 morning Apr 10 11:52:51 djay-il: I am too, but I totally agree with rwhitby that it's a very inconsistent environment to build in Apr 10 12:03:37 CM: what do you find better? Apr 10 12:04:18 I like gentoo, but I run xubuntu too :) Apr 10 12:04:52 What's wrong in this case is that qemu doesn't work well with 64-bit gentoo Apr 10 12:19:24 hi! Apr 10 12:19:43 ~hail RP for committing kernel-image patch Apr 10 12:19:52 koen|away: is ewi offline? Apr 10 12:19:53 * ibot bows down to RP for committing kernel-image patch and chants, "I'M NOT WORTHY!!" Apr 10 12:20:20 RP: do you want to try an experiment? Apr 10 12:20:37 RP: we could try short lived branches? Apr 10 12:36:47 CM: does qemu work well in 64bit anything? Apr 10 12:53:31 ~lart ipk-index Apr 10 12:53:31 * ibot puts on some milking gloves. "All right, now, ipk-index, this won't hurt a bit...." Apr 10 13:05:45 hi Apr 10 13:06:06 djay-il: qemu builds and works for must stuff here on ubuntu feisty amd-64. Apr 10 13:06:27 djay-il: sadly it gives a core dump when I try to generate binary locales with oe Apr 10 13:16:22 Bernardo: well Apr 10 13:16:39 I have few problems with binary locales as well on Gentoo amd64 Apr 10 13:39:35 03florian 07org.oe.dev * rd44e6bb3... 10/ (1 packages/tasks/task-sdk-sbox.bb): task-sdk-sbox: incude gettext development stuff Apr 10 14:00:20 hi there, I'm trying to do a patch. I include it (SRC_URI += "file://paths.patch;patch=1") but it is not working... I'm getting this error message: NOTE: update-alternatives-dpkg-1.13.22-file://paths.patch has no section, not checking URI Apr 10 14:00:47 does anybody have any idea? Apr 10 14:01:39 vivijim: could you post the whole bb file? Apr 10 14:01:53 vivijim: and full log? e.g. which task does print this message? Apr 10 14:02:17 vivijim: with this information I would bet paths.patch does not exist or you did something funky Apr 10 14:05:31 zecke: I'm including this SRC_URI line at update-alternatives-dpkg.inc Apr 10 14:06:27 at packages/update-alternatives/ I've create the dir "files" and put this paths.patch inside this directory: packages/update-alternatives/files/ Apr 10 14:07:19 vivijim: ah :) Apr 10 14:07:27 red herring :) Apr 10 14:07:33 vivijim: it is not a ERROR bur a NOTE Apr 10 14:07:46 vivijim: and it is from my shasum code that it has no shasum for the paths.patch :) Apr 10 14:07:47 yes, exactly... sorry... Apr 10 14:08:06 vivijim: I'm still making up my mind if it is ever useful to have a shasum of local files :) Apr 10 14:08:31 uhm... Apr 10 14:08:49 vivijim: ignore this output, it comes a lot :) Apr 10 14:09:44 but the patch is not been applied :( Apr 10 14:10:27 vivijim: okay, could you paste the whole output? Apr 10 14:10:28 ~pastebin Apr 10 14:10:30 rumour has it, pastebin is 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 Apr 10 14:10:53 vivijim: a) you don't have yesterday's patch from RP to launch patch resolving in a new screen/xterm Apr 10 14:11:05 vivijim: and use bitbake-1.8 and just don't see the patch resolver Apr 10 14:13:05 there is no other line in the output that refers to my patch :( Apr 10 14:13:15 no I don't have this patch yet... Apr 10 14:13:41 and bitbake-1.8 is not working in my case yet... Apr 10 14:18:41 zecke: uhm since update-alternatives-dpkg has already have a do_patch() function hardcoded implemented the patch is not being applied... removing this function my patch was applied... Apr 10 14:24:25 vivijim: a hardcoded do_patch? Apr 10 14:25:17 that is bad Apr 10 14:27:17 s/do_patch/do_configure_prepend/ Apr 10 14:27:53 zecke: see yourself at packages/update-alternatives/update-alternatives-dpkg.inc Apr 10 14:28:01 there is a do_patch function... Apr 10 14:28:43 vivijim: it will be safe to do what I just posted Apr 10 14:29:14 I believe that it is creating a update-alternatives file from update-alternatives.pl file... without this do_pacth() I got the following error: install: cannot stat `scripts/update-alternatives': No such file or directory Apr 10 14:29:16 vivijim: the normal style for that type of munging is in do_configure_prepend or do_compile_prepend Apr 10 14:29:42 hm, i guess i should ask it instead? :) Apr 10 14:29:43 vivijim: do_patch_prepend then if it needs to be done before your patch Apr 10 14:30:06 XorA: thanks I believe that it will work Apr 10 14:32:03 XorA: did not work :( NOTE: package update-alternatives-dpkg-1.13.22-r0: task do_patch: started Apr 10 14:32:03 ERROR: Error in compiling: /indt/mamona/openembedded/org.openembedded.dev/packages/update-alternatives/update-alternatives-dpkg_1.13.22.bb Apr 10 14:32:03 ERROR: The lines resulting into this error were: Apr 10 14:32:03 ERROR: 2:SyntaxError:' cat ${S}/scripts/update-alternatives.pl | \' Apr 10 14:32:03 ERROR: Printing the environment of the function Apr 10 14:32:04 ERROR: 0001:def do_patch(): Apr 10 14:32:06 ERROR: 0002: cat ${S}/scripts/update-alternatives.pl | \ Apr 10 14:32:08 ERROR: 0003: sed -n -e ' Apr 10 14:32:10 ERROR: 0004: /^\$admindir=.*staging/{ Apr 10 14:32:12 ERROR: 0005: x Apr 10 14:32:14 ERROR: 0006: s/^.*$/$D=$ENV{"D"} || ""\;/; Apr 10 14:33:28 XorA: you can not mix python+bash functions easily Apr 10 14:34:12 XorA: you need to another function Apr 10 14:34:17 python do_patch() { Apr 10 14:34:32 bb.build.exec_function("original_patch_name()", d ) Apr 10 14:34:45 bb.build.exec_function("My weird sed script()", d ) Apr 10 14:34:48 but grep :) Apr 10 14:35:52 do anyone have any idea wtf is this error about? -> http://pastebin.ca/432901 Apr 10 14:36:33 io2: yes, you use a glibc version where my patch is not applied Apr 10 14:36:42 io2: you need to have a readelf binary installed Apr 10 14:36:52 YAY!! someone knows!! Apr 10 14:36:57 io2: generally grep your config.log and don't use explicit language Apr 10 14:37:49 um so.. i need to patch glibc? Apr 10 14:38:07 io2: or install readelf, but look at your config.log to verify my statement Apr 10 14:38:21 oh wait a second :) Apr 10 14:38:22 allright.. just a sec :D Apr 10 14:38:32 I only looked at line 12 Apr 10 14:38:45 zecke: ah yeah I forgot do_patch was python Apr 10 14:39:13 /mnt/sda8/ARM/oe/build/tmp/cross/lib/gcc/arm-angstrom-linux-gnueabi/3.4.4/../../../../arm-angstrom-linux-gnueabi/bin/ld: unrecognised emulation mode: armelf_linux, weird Apr 10 14:39:26 zecke: XorA actually I'm pretty sure that this weird sed script is doing nothing but only creating the update-alternatives file from update-alternatives.pl ... I believe that it should change to a simple cp command Apr 10 14:40:00 indeed, i thought that was weird too.. but that didn't give me much clue about why it hates the linker, unless i am missing something. Apr 10 14:41:26 configure:5495: checking for arm-angstrom-linux-gnueabi-readelf Apr 10 14:41:30 configure:5511: found /mnt/sda8/ARM/oe/build/tmp/cross/bin/arm-angstrom-linux-gnueabi-readelf Apr 10 14:41:42 so i guess it has readelf :/ Apr 10 14:41:48 io2: right Apr 10 14:42:41 so i dunno what to do :( Apr 10 14:48:14 * io2 rolls around Apr 10 14:52:12 greetings, where does the ipkg-collateral package come from? Apr 10 14:52:29 it appears to contain only /etc/ipkg.conf Apr 10 14:52:44 but I don't see a bitbake recipe, or a reference to it in anything ipkg related Apr 10 14:56:57 Hello Apr 10 14:59:49 wow, never mind, there it is under packages/ipkg Apr 10 15:13:42 03florian 07org.oe.dev * rf2f723a4... 10/ (1 packages/tasks/task-sdk-sbox.bb packages/tasks/task-sdk.bb): disapproval of revision '636bac75e27918f87914b16fd745dd92dd014ea4' Apr 10 15:13:49 03florian 07org.oe.dev * r6933e926... 10/ (1 packages/tasks/task-sdk-sbox.bb): task-sdk-sbox: Add some more bits to make autotools work completely. Apr 10 15:15:13 03florian 07org.oe.dev * r2d979990... 10/ (1 MAINTAINERS): Add sbox sdk stuff. Apr 10 15:51:04 hi all Apr 10 15:54:39 * koen|gprs <- bored Apr 10 15:56:51 koen|away: go fix the oe-devel ML hehe Apr 10 15:57:08 CoreDump|afk: I'm not the mailinglist admin Apr 10 15:57:25 I didn't even know we have one ;) Apr 10 15:57:55 rwhitby: btw, the retarded yahoo-group-thingy fubars reply-to as well Apr 10 15:58:12 CoreDump|afk: he's called mickeyl :) Apr 10 15:58:50 ah, well, I guess we can agree that he has other things on his mind mid-term Apr 10 17:06:31 is there a way to add packages to an image other than with DISTRO_EXTRA_RDEPENDS? Apr 10 17:07:04 just because packages in DISTRO_EXTRA_RDEPENDS require -force-depends when removing using ipkg later on Apr 10 17:39:37 cbrake: ping Apr 10 17:40:41 hello all Apr 10 17:41:16 gm Apr 10 17:41:17 03nail 07org.oe.dev * rdc503d93... 10/ (1 packages/uboot-utils/files/fic-gta01/fw_env.config): uboot-utils: Remove package uboot-utils, should've been done when it was merged into uboot-directory Apr 10 17:41:51 NAiL, Thanks! Apr 10 17:42:19 It was a leftover from me, so I should've fixed earlier Apr 10 17:42:26 heh Apr 10 17:42:26 I didn't notice it was still there Apr 10 17:42:31 but wth? Apr 10 17:42:36 I removed more than that file Apr 10 17:42:53 such as mtn :) Apr 10 17:43:51 mtn's recursive remove doesn't work Apr 10 17:43:54 I keep forgetting that Apr 10 17:44:47 03nail 07org.oe.dev * r038512d4... 10/ (9 files in 3 dirs): uboot-utils: Try removing the files one more time Apr 10 17:45:25 that's a bit more like it Apr 10 17:45:36 I keep adding bb files for u-boot, so I feel some responsibilty for reducing the total number :) Apr 10 17:45:37 anyone using bitbake overlays? Apr 10 17:46:03 overlays? or collections? Apr 10 17:46:11 Crofton: whoops, collections Apr 10 17:46:14 http://pastebin.ca/432901 <- anyone knows how to workaround this error? T_T tried everything i could think of Apr 10 17:46:14 :) Apr 10 17:46:22 I do a little Apr 10 17:47:12 My colleague works under Gentoo, and the bitbake collections approach seems to pick up .inc files from others collections. This does not happen on any of my Ubuntu installs. Apr 10 17:48:01 why oe just refuse working for me >< Apr 10 17:48:15 so, bla/packages/bla_2.0.bb picks up foo/packages/bla.inc not bla/packages/bla.inc Apr 10 17:48:38 weird Apr 10 17:48:40 io2: you are in the middle of the steep learning curve Apr 10 17:48:44 I'm no help here Apr 10 17:49:00 and that is a weird message :( Apr 10 17:49:12 likewise, a steep what? Apr 10 17:49:23 you are trying to build bootstrap-image? Apr 10 17:49:29 Crofton, i know.. :/ that's why it's a headache. Apr 10 17:49:39 nope. trying to build binutils xD Apr 10 17:49:46 hmmm Apr 10 17:49:57 it doesn't build ANYTHING. always get stuck at that. Apr 10 17:50:02 try building some know to work machine and distro Apr 10 17:50:10 known to work that is Apr 10 17:50:13 and that error doesn't tell anything except there's a error Apr 10 17:50:18 * io2 is using debian etch Apr 10 17:50:32 my plan is to build and customise angstrom Apr 10 17:50:33 I mean machine and distro in local.conf Apr 10 17:50:39 ah Apr 10 17:50:45 the machine i am using is palmtx Apr 10 17:50:55 and the distro is angstrom-2007.1 Apr 10 17:51:01 try angstrom and omap5912osk Apr 10 17:51:02 io2: The build seems to be confused between EABI and non-EABI for ARM. Apr 10 17:51:11 try machine omap5912osk Apr 10 17:51:19 because that is what I use :) Apr 10 17:51:32 if that builds we can look into differences Apr 10 17:51:34 that machine works for the TX as well? tx has a pxa270 at 312Mhz Apr 10 17:51:44 probably not Apr 10 17:51:55 but we want to see if you can build anything first Apr 10 17:52:02 likewise, and what's a EABI and non-EABI? Apr 10 17:52:07 hm Apr 10 17:52:10 okay, i will try Apr 10 17:52:36 * io2 noticed it confuses about the emulator thing, but what does it has to do with the linker? Apr 10 17:52:49 http://linuxdevices.com/articles/AT5920399313.html Apr 10 17:53:15 the emulator is used to build binary locale stuff Apr 10 17:53:35 at that part? Apr 10 17:53:35 you can add ENABLE_BINARY_LOCALE_GENERATION = "0" Apr 10 17:53:45 to local.conf to prevent that from happening Apr 10 17:53:58 I don't use them so it isn't an issue for me Apr 10 17:54:37 aw. Apr 10 17:54:53 well, having portuguese/italian locale would be nice to me Apr 10 17:54:58 but, working is more important Apr 10 17:55:07 working first :) Apr 10 17:55:08 * io2 waits tmp to be removed.... Apr 10 17:55:25 io2: what problems are you having with qemu-native? Apr 10 17:55:28 mv tmp tmp.del; rm -rf tmp.del & Apr 10 17:55:59 thrash.... Apr 10 17:56:09 JustinP, none that i noticed.. well, all i know is that it seems ot have compiled fine Apr 10 17:56:15 JustinP: See his pastebin (EABI) Apr 10 17:56:35 her xD Apr 10 17:56:38 likewise: her Apr 10 17:56:49 EABI? qemu works with EABI.... Apr 10 17:57:04 Did I say his? Oh I did. Oops :-) Apr 10 17:57:22 we need asexual id terms to go with ugt Apr 10 17:57:26 not your fault tho. not that you could guess :) Apr 10 17:57:57 io2: Well, it was my 'best' bet :-) Apr 10 17:58:01 ah....well I don't know what that error means Apr 10 17:58:05 * io2 starts "bitbake nano" with the omap machine... Apr 10 17:59:13 * Crofton crosses fingers Apr 10 17:59:20 * io2 too.. Apr 10 17:59:26 Now you're the support guy Crofton :-) Apr 10 17:59:31 heh Apr 10 17:59:41 well, at least we can see if the basics are there Apr 10 17:59:55 if it's really the machine. palmtx.conf will need hacking. Apr 10 18:00:02 I need to start getting people to write my MS Thesis in exchange for support Apr 10 18:00:05 Yep, helps to start from a working config and go from there Apr 10 18:00:18 what MS thesis? Apr 10 18:00:22 mine Apr 10 18:00:42 paper so my uni will give me a degree Apr 10 18:00:46 * io2 shrugs.. she should think about her thesis too >< Apr 10 18:01:20 hi likewise! Apr 10 18:01:24 yeah, same here.. except.. that i still have to think of a subject.. Apr 10 18:01:33 HopsNBarley: Hey there HNB! Apr 10 18:01:41 io2: for what course? Apr 10 18:01:48 Laws. Apr 10 18:01:50 I have all that, just need to get words from head to paper Apr 10 18:02:00 io2: GPL? :-) Apr 10 18:02:32 probably i will stick to something related to civil process and procedure Apr 10 18:02:35 nice theme, but perhaps not enough literature for a graduate thesis :/ Apr 10 18:02:35 it'd be a good one for a mestrate tho. Apr 10 18:02:55 there aren't any case that i know in brazilian courts that the GPL was discussed Apr 10 18:02:56 hmm, a lawyer, everyone be quiet :) Apr 10 18:03:18 likewise: I'm back Apr 10 18:03:22 * Crofton 's sister just got a law degree Apr 10 18:03:44 woo! what was her thesis about :D? Apr 10 18:04:16 I'm not sure, she may not have had to do one. Apr 10 18:04:30 io2: Didn't know law student could end up doing cross-compilation builds behind their workstations. Cool. Apr 10 18:04:34 She is working for a firm doin environmental law now Apr 10 18:05:13 likewise: i am an ecletic person. :) Apr 10 18:05:14 http://www.hunton.com/bios/bio.aspx?id=17235&tab=0013 Apr 10 18:06:02 tho, i ended up in a networks classroom from the computer enginnering course for knowing more than my teacher at the "learn how to use a computer, it's important for writting your papers" class. Apr 10 18:06:04 io2: good style. Apr 10 18:06:41 woo, environmental law is on the highs here in brazil, particularly now with so many agreements on ethanol fuel Apr 10 18:06:58 cbrake: Just wondering if you ever had problems with bitbake collections picking up .inc files across collections? Apr 10 18:07:01 likewise, while she was trying to teach me word, i was doing my papers on latex :D Apr 10 18:07:10 cbrake: so, bla/packages/bla_2.0.bb picks up foo/packages/bla.inc not bla/packages/bla.inc Apr 10 18:07:38 * likewise ended up doing his thesis in framemaker after fighting with latex for too many times Apr 10 18:07:50 and it all started with some nice IRC folks convincing me to try out linux Apr 10 18:08:13 that one i don't know, framemaker.. Apr 10 18:08:27 likewise: I don't know about *.inc files. I usually just copy the entire package directory to my custom tree as it is easier to diff that way. Apr 10 18:08:28 io2: commercial, adobe. not linux (yet) Apr 10 18:08:37 likewise: I may have misunderstood the question on the list. Apr 10 18:08:39 ah. Apr 10 18:09:14 cbrake: on the ml? that was another issue, and it wasn't mine (although I did reply) Apr 10 18:09:31 likewise: entire package/, not the *entire* package dir Apr 10 18:09:34 latex is not hard. it's just picky and requires more patience than an average human has. Apr 10 18:10:12 cbrake: I too, cherry-pick packages to copy and diff, and work against in a seperate bb collection. Apr 10 18:10:43 cbrake: It's just that one particular (ixp4xx-kernel) uses an .inc that seems to go wrong on Gentoo systems (but not on any of my Ubuntu systems). Apr 10 18:11:29 likewise: hmm, bad Gentoo system :-) My main OE build system is gentoo Apr 10 18:12:16 cbrake: what MACHINEs do you bake for? Apr 10 18:12:34 likewise: compulab-pxa270, and generic x86 Apr 10 18:13:07 wo. a pxa270 Apr 10 18:13:08 * NAiL kicks CIA-11 Apr 10 18:13:09 ow Apr 10 18:13:17 cbrake: well ok, thanks. I have to dig deeper on why bitbake decides on this. Apr 10 18:13:46 Crofton: So you are the one on the ml that wants the reorganization of u-boot stuff? Apr 10 18:13:56 yeah Apr 10 18:14:08 There was a fair bit of cruft. I recently deleted two tar.bz2 files (u-boot snapshots, 6mb each) Apr 10 18:14:17 I saw that Apr 10 18:14:17 * zecke|away hands -vvv to likewise Apr 10 18:14:27 ah.. Apr 10 18:14:30 got a new error Apr 10 18:14:34 I also contributed the git tag, which is no longer needed Apr 10 18:14:37 at least by me Apr 10 18:14:48 io2, Hopefully one we can solve Apr 10 18:14:54 | checking for suffix of object files... configure: error: cannot compute suffix of object files: cannot compile Apr 10 18:15:06 can you pastebin some more context Apr 10 18:15:14 zecke|away: If bitbake comes across an undefined variable inside a .bb, will it then ignore that .bb? Apr 10 18:15:25 sure.. hold on Apr 10 18:15:36 The less polite would start muttering about our crystal balls being foggy :) Apr 10 18:16:33 likewise: no, undefined in what sense? Apr 10 18:16:44 likewise: UNDEFINED += "new" is like a simple assignment Apr 10 18:17:02 likewise: "${FOO}" will evaluate to ${FOO} iirc Apr 10 18:20:37 Crofton, http://pastebin.ca/433276 <- that's where it died Apr 10 18:22:34 Hi there , In one of our machines we are getting this error: all patches directories are becoming empty ( http://pastebin.ca/433275 ).... We have 4 machines building using the exactly the same conf files but in only one machine we are having this problem... does anybody have an idea about it? Apr 10 18:23:25 vivijim: seems like the same problem I have. Let me check Apr 10 18:24:14 vivijim: Need more output, especially see the fetch and patch stages. Apr 10 18:24:42 vivijim: Do you use bitbake collections? Are the machines all running the same OS? Apr 10 18:25:57 it somehow feels like glib hates me or my system. it's always glibc-intermediate failing at the do_configure task. Apr 10 18:26:22 hmm Apr 10 18:26:30 there is a bad switch Apr 10 18:26:49 this is weird Apr 10 18:28:53 it feels like you are using gcc 3.x Apr 10 18:28:59 why? Apr 10 18:29:00 and just as unexplained as the previous bad linker thing Apr 10 18:29:07 i am Apr 10 18:29:12 angstrom/omap5912osk should build 4.x Apr 10 18:29:38 gcc 4 failed at compiling qemu, so i used 3.4.4 Apr 10 18:29:41 hmm... Apr 10 18:29:46 this is building glibc-intermedaite? Apr 10 18:30:03 think so Apr 10 18:30:10 yeah, it is Apr 10 18:30:10 did you set gcc version in local.conf? Apr 10 18:30:14 likewise: http://rafb.net/p/TxHD8Y15.html Apr 10 18:30:55 vivijim: ping Apr 10 18:31:00 vivijim: sorry Apr 10 18:31:07 i did it at the palmtx.conf iirc Apr 10 18:31:21 me scrollbar was at 99% percent, didn't read your pastebin line yet Apr 10 18:31:24 likewise: we are using the same os, but the system of this problem machine is old and probably have a lot of different version f softwares Apr 10 18:31:25 tho. thinking about it it shouldn't be err'in, since i am using a different machine now. Apr 10 18:31:40 if you look in machine/include/tune-arm926ejs.conf Apr 10 18:31:43 likewise: no problem ;) Apr 10 18:31:51 Crofton: Anyway, what about making u-boot-utils a subpackage of u-boot? Apr 10 18:32:01 you will see there is a different argument format for gcc 3 verus 4 Apr 10 18:32:14 OH Apr 10 18:32:21 I think that is a good idea, I am not 100% certain how to set it up Apr 10 18:32:22 now it makes sense. Apr 10 18:32:46 hm Apr 10 18:32:47 vivijim: note lines 20 and 21 of your paste. Apr 10 18:32:48 wait a sec Apr 10 18:32:59 lemme see what version "gcc" links to Apr 10 18:32:59 basically, most of us work entirely with gcc4 toolchains now Apr 10 18:33:14 ha! bingo! it links to 3.4.4 Apr 10 18:33:20 that might be it Apr 10 18:33:36 likewise: what exactly does it means? Apr 10 18:33:39 * io2 remembers having problems with gcc4 a looong time, probably unset it as default back then Apr 10 18:33:43 the different argument annoyed us for a while :) Apr 10 18:33:58 yeah, a lot of c++ stuff neededfixing for 4 Apr 10 18:34:03 vivijim: It seems as if two variables are concatenated there, that shouldn't be concatenated. Apr 10 18:35:00 lemme do some re-linking here and try again Apr 10 18:35:03 and see if it errs again Apr 10 18:38:18 likewise: but this is happening with all packages.. if we add manually the patch to patches dir this works fine and it breaks on next one... Apr 10 18:38:51 vivijim: Your output shows that the fetch stage fails. Apr 10 18:39:08 I would remove the tabs/spaces in front of the patch names in packages/gnu-config/gnu-config_20050701.bb and then rebuild Apr 10 18:39:18 vivijim: ^^^ Apr 10 18:39:43 vivijim: ...as there seems to be some parsing going wrong. Did you check the bitbake versions on all the machines? Apr 10 18:43:31 likewise: we have a svn that have a pointer entry to svn bitbake and all others files that are necessary to build a package, so we are sure that bitbake and OE are really equals... Apr 10 18:45:38 vivijim: OK. But somehow the fetch stage of packages/gnu-config/gnu-config_20050701.bb goes wrong on the machine output you showed me. Apr 10 18:46:20 vivijim: I would expect that the (type and amount of) whitespace in front of the patch filenames could be an issue. Apr 10 18:47:47 vivijim: suggest you try out this: http://www.pastebin.ca/433315 Apr 10 18:48:25 vivijim: I think the difference in host packages might explain the different behaviour. Apr 10 18:48:41 vivijim: we just have to find out which host package is involved. Apr 10 18:54:26 likewise: I've checked that this NOTE messages like line 20 and 21 happens in all of our machines, but the problem with empty patches dir just happens in one of them... Apr 10 18:54:40 likewise: But I'll try to do that on that machine Apr 10 18:55:12 vivijim: make sure to clean this bitbake with -c clean before rebuilding Apr 10 18:59:37 hm Apr 10 18:59:50 tried with gcc 4, and got the switch complaint again.. Apr 10 19:00:19 err, suffix of obj files complaint Apr 10 19:01:27 likewise: I did that change without success... That note is still there... even cleaning everything before new build... I'm sure that bitbake and OE are not the problem... I believe that the problem is some software in machine like bash instead of dash (just an example) or something like that but I could not figure out what software exactly is making it Apr 10 19:01:51 io2, you have deleted tmp and started over? Apr 10 19:01:57 yep Apr 10 19:02:04 likewise: all machines run ubuntu 6.10 but this one is too old and probably has something installed that are broking the build Apr 10 19:02:25 vivijim: Yes, I agree the problem is with bash or sed, or whatever other tool, but it must be related to parsing or handling some piece of the bitbake work to be done. Apr 10 19:02:49 pastebin the eorror Apr 10 19:02:59 vivijim: Agree with you on all points, maybe we differ a bit in the way to find what the problem is :-) Apr 10 19:03:41 hold on... Apr 10 19:05:43 erg. Apr 10 19:05:57 it's still reading gcc 3.4.4 Apr 10 19:06:13 arm-angstrom-linux-gnueabi-gcc (GCC) 3.4.4 Apr 10 19:06:29 you haven't set it in local.conf? Apr 10 19:07:01 nope... i am sure that i set it on the palmtx.conf, but better doublecheck, just to be sure... Apr 10 19:07:31 yep :) Apr 10 19:10:51 Crofton, i didn't spot anything that specify gcc 3.4.4. http://pastebin.ca/433353 Apr 10 19:13:06 cu, gonna catch a movie at the cinema Apr 10 19:14:28 hm Apr 10 19:14:37 think i found it. Apr 10 19:14:43 the distro conf Apr 10 19:19:46 zecke|away: Sorry, I wasn't around earlier Apr 10 19:25:29 Crofton: Ok, timtimred cleaned away u-boot-linkstationppc and merged it into 1.2.0 Apr 10 19:25:40 slowly going down ;) Apr 10 19:25:50 yay Apr 10 19:25:56 * timtimred wakes up Apr 10 19:26:02 I figured poking at the list would get some progress going :) Apr 10 19:26:07 * timtimred gos back to sleep again Apr 10 19:26:49 im not even on the list. i got poked by proxy Apr 10 19:27:20 :) Apr 10 19:27:37 well, anything to get some of the recipes cleaned up :) Apr 10 19:27:39 thanks! Apr 10 19:28:10 Will you be removing the LABEL thingy? Apr 10 19:28:15 yeah Apr 10 19:28:24 when I am not building though :) Apr 10 19:28:35 I wasn't expecting people to jump in right away :) Apr 10 19:28:59 Thanks for your help Apr 10 19:29:05 np... Apr 10 19:29:33 I'll delete the LABEL version tomorrow AM Apr 10 19:29:35 timtimred: your changes have been pushed, btw Apr 10 19:29:51 ahha :) Apr 10 19:30:13 Better to just push a bunch and see if anything breaks ;) Apr 10 19:30:22 heh Apr 10 19:30:30 I do not want to trip up my build Apr 10 19:30:31 i'll just try and keep ignoring monotone in case someone offers me dev access heh Apr 10 19:30:44 I do not think it would, but not worth the risk :) Apr 10 19:31:15 when/if openlink gets pushed i might need it then Apr 10 19:32:03 NAiL is keen to push it but its not ready ;) Apr 10 19:32:44 it's getting close ;) Apr 10 19:32:52 not that i really want it, i can just hassle NAiL to push stuff :D Apr 10 19:33:41 it is getting close :) Apr 10 19:39:31 And the more people working on it, the better ;) Apr 10 19:39:46 what is it? Apr 10 19:39:53 It's a webinterface Apr 10 19:41:28 for foonas, targeting NAS functionality Apr 10 20:12:58 re Apr 10 20:13:36 io2, any luck? Apr 10 20:21:04 seems like it managed to compile glibc intermediate... :D Apr 10 20:21:09 finnaly.. now it's fetching gcc itself. Apr 10 20:22:07 let's hope it manages it without weird errors.. but for so far, i should say thanks for you people being so patient with me :) Apr 10 20:23:32 XorA|gone, rwhitby: Heh, just what we discussed earlier today: http://www.advogato.org/person/dwmw2/diary.html?start=160 Apr 10 20:50:48 wb zecke|away Apr 10 20:53:27 RP: quick question about serial: Which terminal program /setup are you using to see the boot mesages? I have tried picocom, but I don't get any output until the serial console shows up Apr 10 20:56:28 minicom Apr 10 20:59:19 hvontres|poodle, it's not related to terminal emulator (I'm using picocom). Check your console= kernel parameter Apr 10 20:59:55 should be smth like console=ttyS0,115200 Apr 10 21:00:04 at least it works for me Apr 10 21:05:25 bd2: it is, iirc Apr 10 21:07:26 bd2: does this look right: console=ttyS0,115200n8 Apr 10 21:08:38 try to remove n8, I don't know syntax of that string. but console=ttyS0,115200 works for my 2.6.20 kernel on h5000 device (device does not matter, actually) Apr 10 21:09:15 hvontres|poodle, also.. maybe your boot console not on ttyS0. you can find it out by typing "tty" command when you got shell prompt Apr 10 21:10:08 what you seeing after boot is "agetty" or similar, which runs by init (/etc/inittab) Apr 10 21:40:27 hvontres|poodle: I can send you my config file for minicom if you need Apr 10 22:04:32 03florian 07org.oe.dev * r3f98f312... 10/ (1 packages/meta/meta-sdk-gpe.bb): Add GPE SDK Apr 10 22:04:37 03florian 07org.oe.dev * r4689714f... 10/ (5 files in 2 dirs): task-sdk*: Split up SDKs into multiple different variants. Apr 10 22:04:43 03florian 07org.oe.dev * rcd0b184a... 10/ (4 files in 2 dirs): meta-sdk*: Update to reflect task-sdk changes. Apr 10 22:04:51 03nail 07org.oe.dev * rbd1aedcf... 10/ (14 files in 5 dirs): u-boot: finish moving u-boot utils Apr 10 22:05:00 03nail 07org.oe.dev * r675889c3... 10/ (7 files in 3 dirs): u-boot: Remove u-boot-linkstationppc Apr 10 22:05:07 03nail 07org.oe.dev * r3480600c... 10/ (1 packages/uboot/u-boot_1.2.0.bb): u-boot_1.2.0: Merge in linkstationppc Apr 10 22:05:15 03nail 07org.oe.dev * r88409a86... 10/ (7 files in 2 dirs): u-boot: Add linkstation-patches Apr 10 22:05:22 03nail 07org.oe.dev * r47be8bb4... 10/ (1 packages/uboot/u-boot-linkstationppc): u-boot-linkstationppc: Remove empty dir Apr 10 22:16:09 03crofton 07org.oe.dev * r961e9158... 10/ (1 packages/uboot-utils packages/uboot-utils/files): u-boot-utils : Delete empty directories. Apr 10 22:16:14 03crofton 07org.oe.dev * r89524285... 10/ (4 files in 3 dirs): u-boot_LABEL.2006.06.30.2020.bb : bb file no longer needed. Current release version of u-boot work for the OSK. Apr 10 22:16:38 pierrelux: that would be great :) my email is hvontres AT gmail DOT com Apr 10 22:23:05 hvontres|poodle, report if it changing picocom to minicom will help. I'd be very surprised ;-) Apr 10 22:25:23 bd2: I'll try it and let you know Apr 10 22:33:08 re Apr 10 22:33:56 bd2: I think i know what the problem is... the "quiet" at the end of the cmdline... Apr 10 22:51:12 hvontres|poodle, yup, most probably Apr 10 22:53:27 bd2: I'll try some other kernels when I get home..kinda need the Z here at work still. but at least my serial cable works. Apr 10 22:54:39 RP: It looks like the poodle serial woes are definitly power related. I managed to get into the serial console on a 3.5.4.2-rc2 build of 2.6.17 with my externally powered cable :) Apr 10 22:55:53 RP: I'll experiment with some newer ones when I get home tonight. Apr 10 22:56:15 hvontres|poodle: ok, its probably just a missing gpio setting somewhere then Apr 10 22:57:55 RP: did you see thomas kunzes email on the openzaurus ML? http://www2.informatik.hu-berlin.de/~tkunze/zaurus/patches/collie-r3.patch might have the answer Apr 10 23:10:06 I don't think that will solve the problem Apr 10 23:10:14 hvontres|poodle: I need to look at the list at some point Apr 10 23:11:32 RP: np. he mentioned that patch fixing the colie serial console, so I figured there may be something related in there. Apr 10 23:17:06 hvontres|poodle: Its possible but I don't think it will be enough Apr 10 23:57:41 * hvontres|poodle ponders the possible abuses of IRS Form 6478: "Credit for Alcohol used as Fuel" Apr 11 00:09:55 alcohol is an excellent fuel. works very nicely Apr 11 00:19:08 io2: Yes, but I was wondering how "creative" one could get in the definition of "fuel" :) Apr 11 00:25:13 hm, describing it as what it does? i'd say fuel is anything that's a source for energy generation Apr 11 00:29:46 io2: hehe...so alcohol -> body -> body burns calories -> beer==fuel :) Apr 11 00:32:18 * mwester ponders that a moment Apr 11 00:32:34 * mwester goes off to recalculate his taxes... Apr 11 00:39:11 hvontres|poodle, oh, thought you were talking about ethanol for cars XD Apr 11 00:39:45 io2: well, I think the IRS is too, but I was pondering alternate uses for tax forms :) Apr 11 00:41:00 hvontres|poodle, well, you can always recycle paper? it's fun :D Apr 11 00:41:47 * io2 wish she found the plan/project for building a wind-up generator on the web :( **** ENDING LOGGING AT Wed Apr 11 02:59:57 2007