**** BEGIN LOGGING AT Sun Aug 19 02:59:57 2007 Aug 19 03:01:51 03rwhitby 07org.oe.dev * r464e1df5... 10/ (1 conf/distro/include/slugos.inc): slugos.inc: Added gdk-pixbuf provider stuff (I have no idea why this isn't the default somewhere) Aug 19 04:01:00 03rwhitby 07org.oe.dev * rc09e2ebb... 10/ (1 packages/tasks/task-mokogateway.bb): task-mokogateway: ftdi-eeprom is not required Aug 19 04:26:58 * * OE Bug 2806 has been created by rod(AT)whitby.id.au Aug 19 04:27:00 * * openldap-2.3. 11-r4 fails to build on debian etch - perhaps due to new automake? Aug 19 04:27:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2806 Aug 19 06:01:09 03rwhitby 07org.oe.dev * r0639bf21... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: We no longer support uclibc in SlugOS Aug 19 06:28:34 koen: other than qemu problems, are amd-64 builds working now? Aug 19 06:45:34 hvontres|home: poky has qemu-native working on amd64 Aug 19 06:45:50 morning Aug 19 06:47:18 hrw|gone: thanks. Aug 19 06:48:31 03hrw 07org.oe.dev * r748f0b40... 10/ (1 packages/xorg-xserver/xorg-xserver-common.inc): xorg-xserver: fetch from individual releases dir Aug 19 06:51:32 morning all Aug 19 06:52:15 RP: bitbake now seems to just run rm_work and nothing else. stable 1.8 head Aug 19 06:52:25 we can try to resolve (if possible) the fetch issue of dbus ? Aug 19 06:54:28 RP: I have to explicitly ask for -c build to get something to build. Aug 19 06:54:40 damn !!! the solution is to wipeout the gconf-dbus dir in source/svn !!! probably something was corrupted ... Aug 19 07:23:57 * * OE Bug 2803 has been RESOLVED (FIXED) by Aug 19 07:23:59 * *  xserver-xorg-1.3.0.0-r0-do_fetch Aug 19 07:24:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2803 Aug 19 07:55:04 03rwhitby 07org.oe.dev * rd5497549... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: Sorted out some more building and non-building packages. Promoted mysql. Aug 19 08:14:02 How can I fix lack of lzo compressor in native mkfs.jffs2? Aug 19 08:19:08 RP: http://en.pastebin.ca/663085 Aug 19 08:19:31 bla: mtd-utils_1.0.0+git.bb applies the lzo_1x.patch Aug 19 08:20:00 seems there is lzo support in there? Aug 19 08:20:43 Hm. If you're right there should be. Aug 19 08:21:07 I guess I've broken something. ;p Aug 19 08:21:30 But I've removed mkfs.jffs2 from tmp/ and linked mine native one. It worked somehow. Aug 19 09:54:28 hrm Aug 19 09:54:46 some joker broke net-snmp by forgetting to include a patch Aug 19 10:01:19 NOTE: package net-snmp-5.4.1-r1: task do_patch: started Aug 19 10:01:19 NOTE: Applying patch 'configure-tail.patch' Aug 19 10:01:19 NOTE: Task failed: Error: /oe/org.openembedded.dev/packages/net-snmp/net-snmp-5.4.1/./configure-tail.patch not found. Aug 19 10:10:26 mtn is looking for OE.mtn in a wrong path (I migrated stuff to a diffrent box). Where is that path set for mtn? Aug 19 10:16:33 dcordes: mtn --db=path-to-db/OE.mtn Aug 19 10:33:22 hello guys'n'gals Aug 19 10:35:04 hello Aug 19 10:35:17 thanks Philippe Aug 19 10:49:24 oxo: That'd be me ;( Aug 19 10:51:11 oxo: Pushed out now. Thanks. Aug 19 10:53:36 morning all Aug 19 10:54:01 morning RP Aug 19 10:54:23 RP: can you verify that with rm_work inherited, "bitbake foo" does rm_work on foo instead of building foo? Aug 19 10:54:52 rwhitby: I was just looking at the scrollback :/ Aug 19 11:02:39 rwhitby: The dependency graphs look ok? Aug 19 11:03:06 03lenehan 07org.oe.dev * r86264f83... 10/ (1 packages/net-snmp/net-snmp-5.4.1/configure-tail.patch): net-snmp 5.4.1: Add missing patch. Aug 19 11:03:12 03pH5 07org.oe.dev * r30914cc9... 10/ (1 packages/duma packages/duma/duma_2.5.7.bb): duma: add duma-2.5.7, a red-zone memory allocator Aug 19 11:03:14 rwhitby: Had foo already run up to build? Aug 19 11:05:01 * flo_lap wonders if we can get rid of the merge mail to the commits list Aug 19 11:05:25 03rpurdie * r940 10bitbake/setup.py: setup.py: Update to include bb.ui (thanks pH5) Aug 19 11:09:05 v8jlene: tnx Aug 19 11:12:37 v8jlene: erm..: Aug 19 11:12:38 NOTE: Applying patch 'configure-tail.patch' Aug 19 11:12:38 Segmentation fault (core dumped) Aug 19 11:13:49 rwhitby: I think I've found the glitch ;-) Aug 19 11:15:52 * RP can't decide if its something that should be addressed in bitbake or not... Aug 19 11:23:36 rwhitby: Fix pushed, sorry about the rough ride with this :} Aug 19 11:37:18 RP: where are you pushing to? was this a bitbake push or oe push? Aug 19 11:39:16 hello zecke Aug 19 11:39:27 ho Aug 19 11:43:58 likewise, ping Aug 19 11:44:11 Crofton: pong Aug 19 11:44:40 03rpurdie 07org.oe.dev * rc332c982... 10/ (1 classes/rm_work.bbclass): rm_work.bbclass: Add missing dependency Aug 19 11:44:45 03hrw 07org.oe.dev * r6ffb91c6... 10/ (4 files in 3 dirs): xserver-xorg: apply drmfix.patch to get rid of /usr/include/drm Aug 19 11:44:52 03pH5 07org.oe.dev * r3ee0c672... 10/ (1 packages/scummvm/scummvm.inc): scummvm.inc: improve DESCRIPTION, add HOMEPAGE Aug 19 11:44:57 03pH5 07org.oe.dev * r470abd3e... 10/ (8 files in 4 dirs): Aug 19 11:44:57 scummvm-0.9.1: update packaging, make default, openmoko integration Aug 19 11:44:57 * remove DEFAULT_PREFERENCE="-1" Aug 19 11:44:57 * don't strip executable during install, we do that while packaging Aug 19 11:44:57 * add openmoko-scummvm script for fic-gta01 and scummvm.desktop file. Aug 19 11:45:00 * package the modern theme by default - maybe this should go Aug 19 11:45:02 into a separate scummvm-theme package, RRECOMMENDED by scummvm Aug 19 11:45:23 Crofton: building against libtool 1.5.24 now witthe the OE toolchain, which should reproduce the prob you have, on the very same machine that has alternative toolchains available Aug 19 11:45:31 did you get my local.conf? Aug 19 11:46:00 Crofton|home: hmm, didn't see it, whereto did you send? Aug 19 11:46:18 the email address for you I found on my machione Aug 19 11:47:24 gmail.com? Aug 19 11:48:11 ah yes, got it Aug 19 11:48:54 gmail shows all incoming mails in inbox, independent of filtering and labeling... Aug 19 11:49:19 Crofton|home: tnx Aug 19 11:49:26 thank you Aug 19 11:54:01 likewise: OE, as above :) Aug 19 11:56:03 rp: thanks, rm_work should be in somewhat usable state again (for us to test?) Aug 19 11:56:16 RP: if I set BB_SRCREV_POLICY = "cache", how/when is the cache updated, and can I force that step? Aug 19 12:02:13 likewise: yes Aug 19 12:02:58 pH5: Use BB_SRCREV_POLICY ?= "cache" in local.conf and then run BB_SRCREV_POLICY=clear bitbake foo Aug 19 12:03:08 RP: ah, thanks! Aug 19 12:04:35 pH5: Due to "bugs", that might not even trigger a reparse :} Aug 19 12:05:27 :) Aug 19 12:09:35 pH5: Did you try trunk btw? Aug 19 12:10:41 koen: Could you try bitbake trunk at some point and let me know what needs doing to make it usable again for you ? :) Aug 19 12:10:52 RP: yes I only run trunk Aug 19 12:11:24 the two things I was hit by are 1) python 2.5 dependency due to finally: statement and 2) the rm_work issue Aug 19 12:12:02 pH5: Which finally statement? Aug 19 12:13:22 pH5: Have you been running trunk for a while? It was a bit broken :/ Aug 19 12:14:22 RP: cooker.py, line 746. Aug 19 12:15:21 pH5: try: except: finally: is a 2.5 addition? Aug 19 12:15:46 RP: as far as I am aware, yes. on debian's python2.4 I get a syntax error there. Aug 19 12:16:27 yes Aug 19 12:16:31 hehe Aug 19 12:16:31 finally is a 2.5 addition Aug 19 12:16:51 a good one, if i might add Aug 19 12:16:59 hmm Aug 19 12:17:11 pH5: which rm_work issue? Aug 19 12:17:31 mickeyl: How are we going to handle this in bitbake? Require 1.5 for trunk? Aug 19 12:17:37 2.5? Aug 19 12:17:39 RP: yes. Aug 19 12:18:03 of course, we could rephrase that without finally Aug 19 12:18:06 but 2.5 makes sense Aug 19 12:18:13 so I vote for not doing that Aug 19 12:18:21 zecke: some minutes ago, bitbake just ran rm_work_all, nothing else. Aug 19 12:18:37 mickeyl: Yes, I'm thinking more about what our policy is in general Aug 19 12:18:46 didn't RP kill rm_work_all all together? Aug 19 12:18:58 zecke: Would you agree, python 2.5 only for trunk now? Aug 19 12:19:06 zecke: It came back in a new form Aug 19 12:19:12 RP: trunk will become 1.9 or 2.0 or something? a major version might as well require a major update Aug 19 12:19:13 pH5: That issue should be fixed now Aug 19 12:19:29 RP: judging from the amount of changes, i vote for going 2.0 anyways Aug 19 12:19:41 mickeyl: trunk is 1.9 atm and I'm thinking it will become 2.0 with UI support Aug 19 12:20:55 RP: ACK Aug 19 12:21:06 RP: so let's require python 2.5 Aug 19 12:21:10 for trunk, at least Aug 19 12:21:16 RP: yes, 2.5 is almost a year old Aug 19 12:21:17 and remove all the compat hacks Aug 19 12:21:22 i think we even have 2.3 compat hacks in there Aug 19 12:21:39 now that even OE has 2.5 ;) Aug 19 12:21:46 * mickeyl whistles innocently Aug 19 12:22:15 mickeyl: I guess when we get into UI python bindings, only supporting 2.5 will be an advantage too Aug 19 12:22:30 mickeyl: Feel free to remove the 2.3 and 2.4 compat hacks from trunk ;-) Aug 19 12:22:47 RP: I will. Aug 19 12:23:12 we will be using python3000 by then Aug 19 12:25:45 yes, py3k will rock Aug 19 12:25:50 at least speed-wise Aug 19 12:27:52 will it? Aug 19 12:28:09 my FC6 box is still python 2.4 Aug 19 12:28:11 One night I dreamed I had a python implementation 100x times faster than CPython :) Aug 19 12:29:44 sounds like a sweet dream Aug 19 12:29:54 i would ditch c++ any second then :) Aug 19 12:30:23 yeah, I was on crack though :) Aug 19 12:30:30 heh Aug 19 12:31:19 I wonder if python 2.4 is why my server runs bitbake so slowly... Aug 19 12:36:04 Even slackware has python 2.5 now :) Aug 19 12:37:06 alias python='python -OO' Aug 19 12:38:03 well FC7 is 2.5 Aug 19 12:38:16 , but the box I run OE on is still 2.4 :) Aug 19 12:41:04 mickeyl: I didn't know about that! :) Aug 19 12:51:18 :) Aug 19 12:51:24 it doesn't add a lot, but it's a bit Aug 19 12:51:42 i case you didn't notice, i patched our python in OpenEmbedded to always be python -O Aug 19 12:51:50 that means much less flash usage (no .pyc anymore) Aug 19 12:51:53 and a bit more speed Aug 19 12:51:58 and you still get the docstrings Aug 19 12:52:02 (which you lose in case of -OO) Aug 19 12:52:45 mickeyl: I haven't looked at python 2.5 in OE yet but that sounds cool :) Aug 19 12:53:31 Well, I upgraded python on the slackware box and its broken lots of things. I'd have been better building it myself :-( Aug 19 12:54:04 yea Aug 19 12:54:13 File "tmpHandler(e)", line 74, in tmpHandler Aug 19 12:54:14 AssertionError Aug 19 12:54:20 gotta love such error messages :} Aug 19 13:02:51 03mickeyl 07org.oe.dev * r6f58aae8... 10/ (1 packages/openmoko2/openmoko-contacts2_svn.bb): openmoko2-contacts: add some missing dependencies (AGAIN, who is doing bogus merges lately?) Aug 19 13:02:55 03mickeyl 07org.oe.dev * rb1cb3695... 10/ (14 files in 4 dirs): sip/python-sip/python-pyqt: revamp naming, start to work on a current PyQt that builds against Qt-X11 4.3.0 Aug 19 13:02:59 03mickeyl 07org.oe.dev * r29d2dfa7... 10/ (1 packages/python/python-pyqt_4.3.bb): python-pyqt: set BROKEN and give some hints where to look to fix Aug 19 13:22:35 mickeyl: what is a bogus merge? (just to make sure I'm not to blame...) Aug 19 13:24:09 mickeyl, I get a parse error from python-pyqt .... Aug 19 13:26:29 likewise: sometimes -- if you don't use update / merge enough -- a merge ends up reverting change Aug 19 13:26:30 s Aug 19 13:26:35 Crofton: fixed. please pull again. sorry Aug 19 13:26:44 np Aug 19 13:26:53 likewise: the chance increases the more seldomly you push Aug 19 13:27:09 so merging early and pushing often helps Aug 19 13:27:47 ya Aug 19 13:28:14 ~lart skyhusker for changing starling to use sqlite2 instead of 3 Aug 19 13:28:14 * ibot throws a AN/M-8 smoke grenade at skyhusker for changing starling to use sqlite2 instead of 3 Aug 19 13:34:28 hmm Aug 19 13:34:37 why is ipkg no longer upgrading packages ? Aug 19 13:34:46 root@fic-gta01:/tmp$ ipkg install http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/ipk/armv4t/neod_0.1.0+svnr2740-r0_armv4t.ipk Aug 19 13:34:46 Downloading http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/ipk/armv4t/neod_0.1.0+svnr2740-r0_armv4t.ipk Aug 19 13:34:46 Upgrading neod on root from 0.1.0+svnr2739-r0 to 0.1.0+svnr2740-r0... Aug 19 13:34:46 Configuring neod Aug 19 13:34:49 Collected errors: Aug 19 13:34:51 Multiple packages (neod and neod) providing same name marked HOLD or PREFER. Using latest. Aug 19 13:35:11 we need a new packaging system :/ Aug 19 13:35:32 oh Aug 19 13:35:37 this is not an error Aug 19 13:35:39 it is a warning Aug 19 13:35:43 it does it anyways Aug 19 13:35:51 ~lart ipkg for hardly understandable output Aug 19 13:35:51 * ibot declares ipkg a moron for hardly understandable output Aug 19 14:02:36 03mickeyl 07org.oe.dev * r689420eb... 10/ (1 packages/python/python-pyqt_4.3.bb): python-pyqt: unbreak syntax Aug 19 14:02:40 03pH5 07org.oe.dev * r05d89155... 10/ (1 packages/gsm/libgsmd_svn.bb): libgsmd: package bcm vendor plugin Aug 19 14:14:07 mickeyl: We do need a replacement for ipkg... Aug 19 14:14:24 rpm! Aug 19 14:14:39 alright, only joking Aug 19 14:14:54 msi! Aug 19 14:14:58 Crofton: ;-) Aug 19 14:15:19 rpm spec files contain build instructions, which is way more than we ant Aug 19 14:15:27 want Aug 19 14:23:41 what about .deb then? it is partially in, I tried it because I had thisweird bug with ipkg that noone was able to figure out Aug 19 14:23:50 but the .deb rootfs did not want to build Aug 19 14:23:53 so I gave up on it Aug 19 14:24:59 I tried fixing the ipkg bug myself but well, spent quite some time but also could not figure it out Aug 19 14:26:05 rootfs generation from debs does work in poky Aug 19 14:26:18 ish anyway ;-) Aug 19 14:27:22 also ipkg appears to be pretty unmaintained, I tried contact the dev, but never heard anything back Aug 19 14:27:52 RP: tried that too.. failed, probably because I was not on a debian system, it seems to be using some native dpkg stuff which I had troubles buildin g in OE Aug 19 14:28:23 * RP is pleased to note slimserver is now only using 6% of my servers memory rather than 60%... Aug 19 14:28:44 Jin^eLD: It would be good to track down those issues Aug 19 14:29:46 RP: if someone is interested either switching to .deb or at least keeping a maintained option for .deb stuff in OE I'd surely be willing to help Aug 19 14:30:47 Jin^eLD: The plan is certainly to try and keep it maintained... Aug 19 14:31:20 * RP mutters about 48 hour days Aug 19 14:31:35 well, when I had those issues there seemed to be little interest in a .deb based rootfs Aug 19 14:31:48 it seemed everyone was happy with ipkg Aug 19 14:32:04 I think I spent about a week hacking around and gave up because I had to move on in my project Aug 19 14:34:24 Jin^eLD: There is interest, just nobody has the time to sit down and fix it Aug 19 14:35:18 RP: I could start another attempt when I get back from holidays, I think I got pretty far last time, at least I managed to get the dpkg tools to build Aug 19 14:35:45 but I have no apt/deb/dpkg experience at all, coming from an RPM world ;) so I pretty much got stuck on the rootfs creation process Aug 19 14:37:18 Jin^eLD: If I remember, I'll try and see how the poky apt/dpkg package comapre with OE Aug 19 14:38:42 RP: are you the poky guy? I just dont remember anymore, but if yes - I guess I did talk to you about it then, when I was trying to get this thing going Aug 19 14:38:47 :) Aug 19 14:38:56 Jin^eLD: Yes I am and yes you did Aug 19 14:39:18 oh ok, then it was you who said that it probably works because it is reusing some native stuff that is anyway around on debian Aug 19 14:39:32 deny everything Aug 19 14:39:37 :-) Aug 19 14:39:44 likewise: ;-) Aug 19 14:40:26 hey likewise =) Aug 19 14:40:32 hey Jin^eLD Aug 19 14:40:48 Jin^eLD: still no nda or board in sight.... Aug 19 14:40:57 Jin^eLD: (before you ask :-) ) Aug 19 14:41:46 :) Aug 19 14:41:56 Crofton: hmm, you had glibc errors? Me too now: lt-glib-genmarshal: error while loading shared libraries: libglib-2.0.so.0: cannot open shared object file: No such file or directory Aug 19 14:42:06 really strange that sopmeone who wants to work withthe HW has to go through so much trouble in obtaining it Aug 19 14:42:07 Crofton: or is this libtool hack/workaround related? Aug 19 14:42:18 I'd think they should be interested in people using it.. wlel Aug 19 14:42:36 Jin^eLD: yeah, well they told us 'don't expect support' etc.. Aug 19 14:42:58 Jin^eLD: and we buy too small numbers I guess, a few hundreds maybe Aug 19 14:43:18 I see Aug 19 14:44:19 but it will probably work out, it's just that the $25k is a bit too much to 'eval' on. Aug 19 14:44:53 that is true as well Aug 19 14:56:58 the OE FAQ mentions: "As of this writing, the version of glibc in the .dev branch of OE calls for a version of binutils it really shouldn't use" -- why is that? Aug 19 14:57:13 (http://www.openembedded.org/wiki/OeFaq, #5) Aug 19 14:59:29 likewise, ping Aug 19 15:01:39 pong Aug 19 15:01:44 likewise, I do not remember seeing that problem Aug 19 15:01:51 03freyther 07org.oe.dev * r57751dc9... 10/ (1 classes/tinderclient.bbclass): classes/tinderclient.bbclass: Assert for not None when posting, do not handle MsgNote Aug 19 15:01:57 03freyther 07org.oe.dev * r0745f767... 10/ (1 classes/seppuku.bbclass): Aug 19 15:01:57 classes/seppuku.bbclass: We can't handle NoProvider for, don't handle it properly, more debugging added Aug 19 15:01:57 BitBake 1.8 is firing NoProvider in non fatal cases. We have to honor Aug 19 15:01:57 that for now. Aug 19 15:02:01 03freyther 07org.oe.dev * r7003f77d... 10/ (1 classes/tinderclient.bbclass): classes/tinderclient.bbclass: add the missing 'g' Aug 19 15:02:15 Crofton: hmm, this was with my local.conf and the new libtool* Aug 19 15:02:22 currently running a davinci build Aug 19 15:02:25 what amchine? Aug 19 15:02:29 efika Aug 19 15:02:44 efika fails (for me) building c++ code in libusb Aug 19 15:03:07 you are doing autobuilds? Aug 19 15:03:21 can you build for omap5912osk with libtool-1.5.24? Aug 19 15:03:25 that works for me Aug 19 15:03:37 * Crofton wants to make sure he has no local hacks :) Aug 19 15:08:28 Crofton: If autobuilds is some OE or BB feature, no. But I am running on a build server that I can add builds to. Trying a clean omap build. Aug 19 15:10:12 thanks Aug 19 15:10:23 that is what I mean by autobuilds Aug 19 15:10:57 We should figure out a way to schedule builds for low usage time and rotate configs Aug 19 15:11:09 and try building in alternate tmp dirs Aug 19 15:11:30 then I can do a different build every night Aug 19 15:11:53 as you can see from my local.conf, I am currently hacking at about four -five machines Aug 19 15:26:39 of which I have hardware to test 2 :( Aug 19 15:27:05 has anyone approached the gumstic guys about getting hw to test? Aug 19 15:40:11 Crofton: same error on glib here, this is with OE rev. 5f324e98b61ec50554d929eb59723aaef495ef64 (one day old??), now trying the latest. Aug 19 15:40:39 likewise, I am installing oe on a machine at school Aug 19 15:41:35 my request that it be upgraded to a gig of RAM was translated into get a GIG DIMM Aug 19 15:41:41 so now it has 1.5 G RAM Aug 19 15:42:28 Crofton: I should remember that when I discuss my salary Aug 19 15:42:38 yeah Aug 19 16:00:14 morning Aug 19 16:02:03 03daniel 07org.oe.dev * rd1dc0162... 10/ (3 files in 3 dirs): Aug 19 16:02:03 matchbox-keyboard-inputmethod_svn.bb: Reduce font size for fic-gta01 Aug 19 16:02:03 (so keyboard fits on the screen) Aug 19 16:12:34 chouimat|ibook: good morning Aug 19 16:12:43 chouimat|ibook: how's the weather there? Aug 19 16:14:40 sunny and around 20C Aug 19 16:16:52 Down to 24 bitbake bug reports :) Aug 19 16:17:19 NICE Aug 19 16:17:27 that's pretty damn good Aug 19 16:20:13 hi Aug 19 16:20:23 not really an oe question, but anyways Aug 19 16:20:51 I'm doing an initrd based on busybox Aug 19 16:21:25 and I mount my real (openembedded) root exactly as described in Documentation/initrd.txt Aug 19 16:22:21 mount fails (and after patching busybox to print out errno) it says 2, which is "no such file or directory" according to /usr/include/asm-generic/errno-base.h Aug 19 16:23:00 I'm absolutely sure both the device (/dev/loop0) and the mountpoint (/new-root) exist Aug 19 16:23:12 olegfink: and does /sbin/init exist? Aug 19 16:23:26 ah, mount already fails? Aug 19 16:23:32 yep Aug 19 16:23:40 Crofton: Keep getting this error: "lt-glib-genmarshal: error while loading shared libraries: libglib-2.0.so.0: cannot open shared object file: No such file or directory"... Aug 19 16:24:26 olegfink: could you paste the output? to pastebin.ca for example Aug 19 16:24:52 olegfink: does the mount command exist in your initrd? Aug 19 16:25:05 likewise: output of what? mount? it's one line: mount: mounting /dev/loop0 on /new-root failed: 2 Aug 19 16:25:27 likewise: sure! I've rechecked everything thousands times Aug 19 16:25:50 and it's a bit tricky to paste initrd output... Aug 19 16:25:55 olegfink: well, something basic must be wrong though. Aug 19 16:26:31 I want to know what may trigger mount returning errno=2 Aug 19 16:26:44 both the files I provide it exist for sure Aug 19 16:27:00 olegfink: the doc you mention is from the kernel sources, right? Aug 19 16:27:05 I can show my linuxrc, or the whole initrd if you want Aug 19 16:27:11 likewise: exactly Aug 19 16:28:11 olegfink: loopback devices are builtin your kernel? Aug 19 16:28:25 ? Aug 19 16:28:34 how could device be built in the kernel? Aug 19 16:28:53 olegfink: You use /dev/loop0, it needs a device driver. Aug 19 16:29:00 ah, sure Aug 19 16:29:10 because the userland boots after it Aug 19 16:29:29 (you know, if initrd fails to mount the root kernel does it instead) Aug 19 16:30:15 olegfink: I don't understand how your boot procedure goes. The kernel has a initrd, pls. explain what happens afterwards. Aug 19 16:30:24 * chouimat|ibook hates xorg Aug 19 16:32:10 likewise: http://rafb.net/p/XEr6cz41.html Aug 19 16:34:37 olegfink: I would start adding debugging prints there: an extra mount command right after the losetup command to show the mounts. Aug 19 16:34:55 ls -ald /new-root /dev/loop0 Aug 19 16:35:26 ls -ald /bin/magic Aug 19 16:35:43 do you know if losetup was successful? Aug 19 16:36:19 likewise: yeah, as the kernel with root=/dev/loop0 option boots the userland afterwards Aug 19 16:36:36 likewise: /bin/magic is my busybox applet Aug 19 16:36:46 very silly thing Aug 19 16:38:32 likewise: so you suggest calling 'mount' after my mount command? Aug 19 16:40:00 olegfink: yes, to see what mounts are actually performed. Otherwise, I am out of clues. Aug 19 16:40:56 btw, what's typical number of a read-write cycles a CompactFlash will last? Aug 19 16:42:11 in the thousands Aug 19 16:42:34 so it's not many cycles left Aug 19 16:42:45 (one less after adding that mount call) Aug 19 16:43:04 better to wear out the CF than your NAND flash... Aug 19 16:44:02 likewise: booting... Aug 19 16:44:23 re Aug 19 16:44:56 mount: no /proc/mounts Aug 19 16:44:59 uhm Aug 19 16:45:06 ah Aug 19 16:45:09 simple Aug 19 16:45:19 I haven't compiled that feature for mount Aug 19 16:45:30 (or I don't mount proc in initrd) Aug 19 16:45:47 * chouimat|ibook definitly hates xorg configuration Aug 19 16:46:33 the latter is true Aug 19 16:46:38 likewise: so? Aug 19 16:46:52 olegfink: hmm Aug 19 16:46:54 should I mount proc in initrd? doesn't look like a good idea to me Aug 19 16:47:36 I wonder if you need it to mount the rootfs... Aug 19 16:48:05 every initrd I've seen so far (including one described in initrd.txt) does it Aug 19 16:53:03 olegfink: ok, presumably you need to mount proc first, but I could not find a definitive answer on why Aug 19 16:53:22 olegfink: http://www.denx.de/wiki/DULG/RootFileSystemInAReadOnlyFile Aug 19 16:53:26 Some build path in one of my OE config files seems wrong but I can't locate that config file. In the output of the error log that wrong path is shown: http://www.pastebin.ca/663462 Aug 19 16:53:32 "As a final cleanup we unmount the /proc file system which is not needed any more." Aug 19 16:53:41 hmmmm Aug 19 16:53:45 in fact Aug 19 16:53:48 let me think Aug 19 16:54:42 I have a strange feeling that proc is mounted Aug 19 16:55:16 dcordes: try from a clean shell (or investigate your env variables) Aug 19 16:56:09 likewise: I tried several times. Enviornment script, local.conf and the bitbake.conf files are setup correctly Aug 19 16:56:45 added /bin/sh call to linuxrc, so I can experiment for a while Aug 19 16:58:11 likewise: I have no idea where bitbake is taking that path from Aug 19 16:58:37 RP: NOTE: Handling BitBake files: | (4600/4001) [114 %] Aug 19 16:58:47 could it be that bitbake trunk ignores BBMASK? Aug 19 16:59:04 likewise: no, I really don't have proc mounted Aug 19 16:59:48 likewise: mounted proc by hand, now I have Aug 19 17:00:00 rootfs on / type rootfs (rw) Aug 19 17:00:16 /dev/root.old on / type ext2 (rw) Aug 19 17:00:30 none on /proc type proc (rw) Aug 19 17:02:00 03stefan 07org.oe.dev * r4fe73431... 10/ (1 packages/alsa/alsa-state/fic-gta01/gsmhandset.state): fic-gta01 alsa-statefiles: Merge the tweaking from Daniel Willmann for a better phonecall experience. Aug 19 17:03:14 pH5: It shouldn't :/ Aug 19 17:03:25 btw, if I want to be on the safe side - what bitbake version do I use with current OE? I have not been following the last month but I had the impression that there were some problems with bitbake or something Aug 19 17:03:29 pH5: Do any of your .conf files do BBFILES += ? Aug 19 17:03:53 likewise: any idea where that wrong build path might be set? Aug 19 17:04:42 RP: not that I'm aware of. I have 'BBFILES ?= "${HOME}/src/oe/openembedded/packages/*/*.bb"' in local.conf Aug 19 17:04:50 Jin^eLD: 1.6.6 and 1.8.8 should both work Aug 19 17:05:45 BTW: we have lisppaste7 here now :) Aug 19 17:05:49 lisppaste7: help Aug 19 17:05:49 To use the lisppaste bot, visit http://paste.lisp.org/new/oe and enter your paste. Aug 19 17:05:54 zecke: what'S the main difference between those two? Aug 19 17:06:22 Jin^eLD: 1.8 can parallelize, oh and I think 1.8 is required by OE :) Aug 19 17:06:32 which one would be "better"? Aug 19 17:06:39 1.8 Aug 19 17:06:46 it is faster Aug 19 17:07:13 thanks Aug 19 17:09:12 likewise pasted "hellworld" at http://paste.lisp.org/display/46387 Aug 19 17:10:39 dcordes pasted "config breakage" at http://paste.lisp.org/display/46388 Aug 19 17:10:43 zecke: XML-RPC support? so the autobuilders can paste bugs live :-) Aug 19 17:10:54 nice idea Aug 19 17:11:14 why _paste_ bugs and not submit them to bugtracker? Aug 19 17:11:38 say, to the special bugtracker section "automated build failings reports" Aug 19 17:11:47 olegfink: to first show them and fix them if it's not a dev-error. Aug 19 17:12:37 likewise: it's also possible in bugtracker, but the latter gives you much more possibilities Aug 19 17:12:51 and flexibility Aug 19 17:13:12 olegfink: because we do that already :) Aug 19 17:13:32 olegfink: Well, half of what I see here, are configuration errors (on the host, or in OE itself), not metadata bugs. Aug 19 17:14:26 re Aug 19 17:15:05 wb Aug 19 17:16:39 hm Aug 19 17:16:49 btw any ideas on my trouble? Aug 19 17:18:54 likewise: how can I find out where that path is set? I've been searching for hours Aug 19 17:19:21 olegfink: no, I expected you to debug with a shell? Aug 19 17:20:51 likewise: did it, got a mount table pasted above Aug 19 17:21:51 olegfink: OK, now what if you proceed through your init script step by step Aug 19 17:21:56 manually Aug 19 17:22:11 the next would be mount Aug 19 17:22:19 I can't proceed without mounting Aug 19 17:22:33 the only mount in my script and the one which fails Aug 19 17:22:43 what new can I learn? Aug 19 17:22:48 dcordes: you have moved your workdir and some tools we build(e.g. quilt) hardcode the path to their helper apps... Aug 19 17:23:07 My problem is that kernel has a bad keymap, so in userland I usually switch it with kbd Aug 19 17:23:39 olegfink: ls -ald /new-root /dev/loop0 all look fine? Aug 19 17:24:13 does mount need a helper mount.e2fs or somesuch? Aug 19 17:26:09 uhm, it's busybox mount Aug 19 17:26:19 it's pretty self-contained I _believe_ Aug 19 17:26:48 I don't have ls by the way Aug 19 17:27:17 I can cd to new-root Aug 19 17:27:39 zecke: err how do I do that? Aug 19 17:27:40 loop0 exists Aug 19 17:28:35 dcordes: no idea, your error looks like you did Aug 19 17:29:20 zecke: I tried to symlink stuff to the old path but that doesn't work (too many levels of symbolic links) Aug 19 17:29:25 and "h: /home/erdinger/OE/build/tmp/akita/staging/i686-linux/bin/install-sh: Too many levels of symbolic links" gie a hint as well.. Aug 19 17:31:02 olegfink: I'm pretty clueless then. sure loop0 is a device node? Aug 19 17:31:25 zecke: would rm -rf tmp and rebuild be a solution? Aug 19 17:31:47 likewise: YEAH, and kernel mounts it just fine after initrd is over Aug 19 17:32:24 olegfink: sorry, no new ideas from my side. Aug 19 17:33:03 pH5: It does sound like some kind of masking problem :/ Aug 19 17:34:56 pH5: It is a mask problem... Aug 19 17:35:02 dcordes: I don't know your symlink hell, but it could work Aug 19 17:35:16 RP: you've got it already? I was just going to do some -DD logs Aug 19 17:35:50 zecke: it will take my current machine at least 4 days to accomplish I think Aug 19 17:37:35 pH5: Yes, the total is just not including the masked files yet the count of processed files it. Its just an output error and no internals are affected by it Aug 19 17:37:37 dcordes: alternatively rebuild quilt? Aug 19 17:38:10 RP: ok, that's certainly better than the other way around :) Aug 19 17:38:30 pH5: Yes :) Aug 19 17:38:49 zecke: hmm too late Aug 19 17:39:01 bye Aug 19 17:41:16 03rpurdie * r941 10bitbake/lib/bb/event.py: event.py: masked files are not included in total Aug 19 17:41:33 pH5: Fixed :) Aug 19 17:42:26 RP: thanks! Aug 19 17:44:02 RP: Shouldn't bitbake check on each call if there are new svn revisions that need compiling? Aug 19 17:45:12 likewise, I am doing an omap5912osk with stock revs, it is still runnning Aug 19 17:45:52 Crofton: yes, the glib-2.0 breakage is libtool-version related. Aug 19 17:46:04 interesting Aug 19 17:46:11 for omap5912osk? Aug 19 17:46:12 Crofton: it's now in target userland already, earlier it broke on glib-2.0-native./ Aug 19 17:46:18 checking... Aug 19 17:46:31 MACHINE = "omap5912osk" Aug 19 17:46:38 yep Aug 19 17:47:08 I will run a libtool-1.5.24 run here after this complete .... Aug 19 17:47:18 Crofton: but as you don't get that breakage with libtool 1.5.24, something host-related is probably creeping in Aug 19 17:47:28 Ubuntu 7.04 x86-32 here Aug 19 17:47:30 you are using another tool chain? Aug 19 17:47:38 Fedora 7 and 6 Aug 19 17:48:02 RP: Never mind, I'm stupid :-) Aug 19 17:48:03 Crofton: no, not yet. I will try with a new toolchain in a bit. Need to lookup the way to set the toolchain... Aug 19 17:48:33 Crofton: got sidetracked multiple times (recursive sidetrack problem :-) ) Aug 19 17:48:40 right Aug 19 17:49:11 when koen gets back from the mud bit, I plan to discuss creating a script to do autobuilds Aug 19 17:49:26 something we can run from cron and rotate through combinations Aug 19 17:50:11 this way I can devote each night's cpu time to a different build Aug 19 17:50:25 and we can get a number of people doing it Aug 19 17:52:54 I should remove the libtool-1.5.22 files .. Aug 19 17:53:31 03stefan 07org.oe.dev * r2d845d0c... 10/ (1 packages/alsa/alsa-state.bb): alsa-state.bb: Bump forgotten RP and clean up some trailing whitespaces. Aug 19 17:55:40 Crofton: that autobuild script sounds like a good idea Aug 19 18:00:46 anyone using gstreamer here? Aug 19 18:22:19 03mickeyl 07org.oe.dev * r33d36c3f... 10/ (3 files in 2 dirs): add alsa controller and PCM plugins Aug 19 18:22:24 03mickeyl 07org.oe.dev * r53bc2277... 10/ (1 packages/gstreamer/gst-meta-base_0.10.bb): gst-meta-base: add id3demux plugin Aug 19 18:22:29 03mickeyl 07org.oe.dev * r7a0041ca... 10/ (1 packages/gstreamer/gst-plugin-pulse_0.9.4.bb): add gst-plugin-pulse, GStreamer plugins to use pulseaudio as source and sink Aug 19 18:22:34 03mickeyl 07org.oe.dev * r38796016... 10/ (1 packages/gstreamer/gst-plugin-pulse_0.9.4.bb): gst-plugin-pulse: bypass libtool install (bogus for plugins) Aug 19 18:22:39 03mickeyl 07org.oe.dev * r94f444b2... 10/ (1 conf/distro/openmoko.conf): openmoko.conf: set feed URIs for P1-August-Snapshot Aug 19 18:34:37 anyone doing builds on a system running SELinux? Aug 19 18:36:14 Crofton: that's the first thing I disable on my Fedora installs. It might actually work on Fedora 7, but I've had such poor luck with FC 5 & 6 in just getting the system to run properly, that I've never even got to the point of trying it with OE. Aug 19 18:36:26 heh Aug 19 18:36:31 I have been trying to work it out Aug 19 18:36:54 ... I actually have a spare F7 system that I'll be reloading next week... maybe I should turn it on and try it. Aug 19 18:37:08 the build is continuing, but they have a star that comes up in the toolbar when something happens Aug 19 18:37:18 and a little thing tries to explain what happened Aug 19 18:37:22 yeah Aug 19 18:37:23 try it Aug 19 18:37:42 I'll pastebin some messages when mroe guys are awake Aug 19 18:38:02 ldconfig is trying to fondle some files apparently Aug 19 18:38:25 "fondle"? hehe! Aug 19 18:38:44 ~seen zecke Aug 19 18:38:45 That's a technical term, right? :-) Aug 19 18:38:46 zecke was last seen on IRC in channel #openmoko, 51m 2s ago, saying: 'bbl'. Aug 19 18:38:55 hms Aug 19 18:39:17 http://pastebin.ca/663575 Aug 19 18:39:22 very technical Aug 19 18:39:57 why would ldconfig need to fondle a log file .... Aug 19 18:41:01 That's why I disable selinux. Aug 19 18:41:50 "It is common for user to edit files in their home direcotry or tmp directories..." -- Yeah, ok, I agree. Aug 19 18:41:58 you do not want to know who is fondling what? Aug 19 18:42:05 :) Aug 19 18:42:28 "and then move (mv) them to system directories." -- nope, as a practice I don't move them back. Pointless, methinks... selinux blows it again! Aug 19 18:42:30 I suspect something weird Aug 19 18:42:47 yeah, the verbage does suck Aug 19 19:24:57 hi, maybe OT? Do you know where can I find an arm rootfs image suited for linux-2.4 ? Aug 19 19:27:10 03mickeyl 07org.oe.dev * r8691ff30... 10/ (1 conf/distro/openmoko.conf): openmoko.conf: remove svnnow, fix typo in feed addresses Aug 19 19:27:14 03mickeyl 07org.oe.dev * r14424ba8... 10/ (1 conf/distro/include/openmoko-srcdate-now.inc): remove openmoko-srcdate-now.inc Aug 19 19:36:59 ao2: totally off-topic, we only create images Aug 19 20:44:57 morning Aug 19 20:45:03 hi Aug 19 20:45:10 how are you? where are you? Aug 19 20:45:46 zecke: good, PoznaƄ/Poland still Aug 19 20:46:25 zecke: I am thinking about creating router/AP with alix board as base - to replace wrt54gs (current router/AP) and nslu2 (nfs) with it Aug 19 20:47:15 but with 1Gbit network card + 1Gbit switch instead of 100Mbit network Aug 19 20:47:37 hrw: to prepare the internet connection for your upcoming children? Aug 19 20:47:52 ;D Aug 19 20:47:58 I agree that children shouldn't be born in a world with 100mbit home networks Aug 19 20:48:36 zecke: no - I want to experiment more with alix board and as it has minipci + pci + onboard ethernet it can be intesteing base for such machine Aug 19 20:49:08 * Crofton|home was born in a world without ethernet and it does not seem to have damged him Aug 19 20:50:40 mickeyl: ERROR: No providers of runtime build target python-crypto (for ['/home/hrw/devel/oe/org.openembedded/packages/tasks/task-python-everything.bb']) Aug 19 20:50:40 Crofton|home: hehe, you live near an armby base? that can make things up :) Aug 19 20:50:53 * zecke tries to walk the noble eightfold path Aug 19 20:50:56 hrw: mtpull Aug 19 20:51:11 ok Aug 19 20:51:16 hrw: i think :) Aug 19 20:51:20 if not, please remove it Aug 19 20:51:24 crypto no longer builds atm. Aug 19 20:52:25 27 revs to pull Aug 19 20:53:59 http://en.wikipedia.org/wiki/Green_Bank%2C_West_Virginia Aug 19 20:54:16 1966 through 1971, NRAO is not quite an army base Aug 19 20:54:27 but it was attacked by little rgeen men Aug 19 20:54:50 :) Aug 19 20:54:55 that can make up a lot of thing Aug 19 20:54:56 s Aug 19 20:55:42 The GBT replaced a telescope that fell down, teh weekly world news reported that aliens had destroyed it :) Aug 19 20:55:59 http://www.gb.nrao.edu/fgdocs/300ft/300ft.html Aug 19 20:58:03 lol Aug 19 21:05:18 * hrw found .oe recipe in metadata.. Aug 19 21:05:31 hehe Aug 19 21:06:49 task-python-everything is hardcore build Aug 19 21:06:57 yeah Aug 19 21:07:00 drags in almost everyhting ;) Aug 19 21:07:02 like world Aug 19 21:07:55 mickeyl: if it will break my builddir then I will @#$@%^@^@*$#()%&*@# you Aug 19 21:08:25 as this is not powerfull machine to restart all from beginning ;D Aug 19 21:12:31 hi hrw Aug 19 21:13:34 hrw I whish you all the best for your ibot marriage Aug 19 21:13:38 args Aug 19 21:13:45 hrw I whish you all the best for your marriage Aug 19 21:15:55 thx woglinde Aug 19 21:37:08 03hrw 07org.oe.dev * r6744f035... 10/ (1 packages/tasks/task-python-everything.bb): task-python-everything: removed python-crypto as it does not build (acked by Mickeyl) Aug 19 21:37:12 03hrw 07org.oe.dev * r99210115... 10/ (1 packages/tar/tar_1.18.bb): tar: added 1.18 Aug 19 21:37:17 03hrw 07org.oe.dev * rb05c382a... 10/ (1 packages/tar/tar packages/tar/tar/tar-native_1.13.25.oe): tar: drop VERY OLD recipe which is not used since BitBake was created Aug 19 21:46:23 * rwhitby wakes, and looks for an announcement ... Aug 19 21:46:44 good morning rwhitby Aug 19 21:47:06 rwhitby: what would you like to see announced? Aug 19 21:47:16 oops, wrong channel for the announcement, although mickeyl is here too Aug 19 21:56:00 good nite all Aug 19 22:01:59 mickeyl: When are we planning to bump the OE.dev bitbake version requirement? Aug 19 22:02:33 rwhitby: once Sean wakes up. It's 06:00 in .tw Aug 19 22:02:43 rwhitby: feel free to switch the MokoMakefile now Aug 19 22:03:07 RP: 2 weeks? Aug 19 22:03:14 RP: giving notice tomorrow? Aug 19 22:03:36 mickeyl: ok, we should give notice asap (pending any big 1.8.8 bugs appearing) Aug 19 22:08:40 mickeyl: you build on 32bit machines only? Aug 19 22:09:04 hrw: mostly, yes. occasionally one x86-64. why? Aug 19 22:09:22 mickeyl: does python-everything built on that 64bit one? Aug 19 22:09:36 hrw: i can launch that tomorrow Aug 19 22:09:52 mickeyl: would be nice if you could Aug 19 22:10:29 mickeyl: I do not know how 2.5 but 2.4 python do not like when build != target when it comes to 32/64 bits Aug 19 22:11:37 mickeyl: I fixed some of recipes in poky months ago and most of those fixes are imported into OE but I did not build whole python stuff Aug 19 22:14:13 | checking size of pthread_t... configure: error: cannot run test program while cross compiling Aug 19 22:14:20 glib 1.2 fails for i686 Aug 19 22:17:55 hmm perhaps now it's time to get rid of glib 1.2 and gtk+ 1.2 Aug 19 22:18:00 iirc we have two programs depending on it Aug 19 22:18:55 irdadump is one of them Aug 19 22:19:04 and it is nice debug tool Aug 19 22:19:19 It sounds like a missing site entry Aug 19 22:19:34 it is Aug 19 22:19:47 tomorrow I will bot my i686 laptop and check Aug 19 22:19:53 117 glib_cv_sizeof_system_thread=${glib_cv_sizeof_system_thread=4} Aug 19 22:20:01 does this is correct (rather yes) Aug 19 22:20:42 * hrw -> sleep Aug 19 22:20:47 bye Aug 19 22:20:53 'night hrw Aug 19 22:22:49 03rpurdie * r942 10bitbake/lib/bb/cooker.py: cooker.py: Add support for BB_NICE_LEVEL option (#1627) Aug 19 22:46:02 03rpurdie * r943 10bitbake/lib/bb/ui/knotty.py: knotty.py: Improve BBINCLUDELOGS_LINES handling (#97) Aug 19 22:51:48 03rpurdie * r944 10bitbake/lib/bb/cooker.py: cooker.py: Warn about malformed PREFERRED_PROVIDERS (#1072) Aug 19 23:51:03 03rpurdie * r945 10bitbake/lib/bb/__init__.py: __init__.py: Fix a bug in decodeurl where http://some.where.com/somefile.tgz decoded to host="" (#1530) **** ENDING LOGGING AT Mon Aug 20 02:59:56 2007