**** BEGIN LOGGING AT Tue May 06 02:59:57 2008 May 06 03:06:05 * * OE Bug 2627 has been RESOLVED (FIXED) by May 06 03:06:07 * * altboot seems unable to mount either SD or CF cards on collie May 06 03:06:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2627 May 06 03:18:27 Hmm May 06 03:18:31 is khem not ircing anymore? May 06 07:53:59 bonjour May 06 07:57:06 03koen 07org.oe.dev * r0cf80f75... 10/ (1 packages/linux/linux-omap2_git.bb): linux-omap2 git: move to 2.6.26rc1 May 06 08:32:16 ~ping Crofton May 06 08:32:17 pong Crofton May 06 08:32:43 hi steliosk May 06 08:41:15 woglinde : hi ! May 06 08:51:35 lol May 06 08:51:56 as I see it ok labs took l4 modified it and now sells it May 06 08:55:41 woglinde: OK L4 microkernel sources are not avaiable ? May 06 08:56:27 hm the origial yes, but the sources from ok labs not May 06 08:56:45 shame ... May 06 08:57:22 sorry May 06 08:57:29 I didnt see the skip registrs May 06 08:57:59 s/shame/too bad/ May 06 08:59:30 ah okay May 06 08:59:37 source code is in the download May 06 09:00:56 re May 06 09:01:01 ups May 06 09:01:34 woglinde: :) May 06 09:31:02 jeremy_laine : Did you get the OE running on the Kilauea ? May 06 09:31:17 steliosk: nope May 06 09:31:28 what's the problem ? May 06 09:31:32 steliosk: sorry but have to run, more details later :) May 06 09:31:36 ok May 06 09:32:04 steliosk: can't get the kernel to boot.. for that matter I can't boot the binary image provided on the denx website May 06 09:32:47 jeremy_laine : i am waiting for my board to arrive this week May 06 09:44:53 morning May 06 09:46:55 hey hrw May 06 09:49:23 pxa27x has one nice thing May 06 09:49:33 usb host and device at same time May 06 09:52:30 hi hrw May 06 09:59:15 gdb over usb networking + feed on usb stick == quite comfortable debugging May 06 10:00:36 hrw if the usb is stable enough May 06 10:00:51 not for all archs usb-net works like a charme May 06 10:01:13 woglinde: I know May 06 10:01:22 woglinde: it is pxa27x May 06 10:05:03 03koen 07org.oe.dev * r71a6f5cc... 10/ (1 packages/mono/mono_1.2.6.bb packages/mono/mono_svn.bb): mono: use staging_datadir_native May 06 10:33:44 does someone know can machine be configured to have eth0 available in Xen instance only? May 06 10:38:06 I'd know in virtualbox.org ;-) May 06 10:38:52 steliosk, pong May 06 10:39:29 Crofton : hi May 06 10:39:46 Crofton : are you still working with the gumstix ? May 06 10:40:01 a little May 06 10:40:09 does anyone succeed to compile uicmoc4-native-4.3.3 ? May 06 10:40:11 mostly giving a freind advice with his project May 06 10:40:39 here the compilation process it endless ... May 06 10:40:53 what is the question? May 06 10:41:06 'rcc -name network network.qrc -o .rcc/release-static-emb-x86/qrc_network.cpp' is running for more than 500 minutes :-/ May 06 10:41:11 Crofton : did they switch to OE or still using buildroot ? May 06 10:42:15 basically they use OE May 06 10:42:47 sakoman, made a copy of .dev in October (?) of last year and works from it May 06 10:42:59 but other than that, things are going very well for them May 06 10:43:07 :) May 06 10:44:29 cyrilRomain: It is a long compile, but it works for me May 06 10:44:39 How much power does the machine have? May 06 10:44:46 What distro/machine? May 06 10:45:50 Crofton : i working on a presentation and it looks like gumstix is the smallest machine i can get which supports May 06 10:45:58 OE May 06 10:47:27 Laibsch: Linux kebabordi 2.6.22-14-generic #1 SMP Sun Oct 14 23:05:12 GMT 2007 i686 Intel(R) Pentium(R) Dual CPU E2160 @ 1.80GHz GenuineIntel GNU/Linux May 06 10:47:29 Switching to a package based system helped a lot of people May 06 10:47:52 I'm sure sakoman can answer specific questions, when he wakes up May 06 10:49:04 03Laibsch 07org.oe.dev * r7a849e6b... 10/ (1 packages/tmsnc packages/tmsnc/tmsnc_0.3.2.bb): tmsnc: release 0.3.2. Closes 4217. May 06 10:49:08 03Laibsch 07org.oe.dev * r9f9443b8... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: fix SRC_URI (courtesy jlime project), reorder fields and update HOMEPAGE May 06 10:49:13 03Laibsch 07org.oe.dev * rb2e500e4... 10/ (1 packages/roadmap/roadmap-gtk2_1.1.0.bb): roadmap-gtk2: fix SRC_URI (courtesy jlime project) May 06 10:50:45 cyrilRomain: I meant, what distro/machine are you compiling for? May 06 10:51:50 Laibsch: qemux86 May 06 11:06:10 mhm, I failed at fixing the OE toolchain stuff, in the end just did bitbake -c patch binutils/gcc-cross/etc. and did the configure and build steps by hand, following the CLFS book May 06 11:06:18 generated external toolchain works May 06 11:06:39 so something in the OE scripts is somehow broken and leads to #4203 May 06 11:06:45 but I am not sure how to find out what May 06 11:22:56 Laibsch: did you ever done that fc-cache/update-qtttffontdir stuff? May 06 11:23:21 well, more or less, yes May 06 11:23:31 I think there is one open bug about bdf fonts May 06 11:23:37 Why do you ask? May 06 11:24:00 Or do you mean if I verified that ttf fonts were indeed correctly displayed? May 06 11:24:03 I have to do such thing so wanted to ask before started ;) May 06 11:24:16 It should be all there May 06 11:24:24 So, it should be 0 work May 06 11:24:28 or what is missing? May 06 11:25:05 Laibsch: that part was not merged into poky yet so I will merge May 06 11:26:06 Laibsch: btw - do not remove 'Dusk' theme from bugzilla May 06 11:26:35 OK May 06 11:26:39 Did I? May 06 11:27:52 no, you did not May 06 11:28:00 I just like it more then default one May 06 11:28:05 btw - Undefined subroutine &main::GetVersionTable called at /usr/share/perl5/Bugzilla/Search.pm line 116. May 06 11:29:19 I know May 06 11:29:32 I am working on a bug in Search function May 06 11:29:48 It is fixed in the next release which we cannot install May 06 11:29:59 no package? May 06 11:30:06 no May 06 11:30:10 not even for bugzilla May 06 11:30:20 so, I had to make that concesssion May 06 11:30:30 The upgrade would require more unpackaged stuff May 06 11:30:40 ugly indeed May 06 11:30:41 So, I am trying to make do with the patches from the bug system May 06 11:30:54 Since bugzilla is already unpackaged anyways May 06 11:32:25 hrw: OK, patches in May 06 11:32:29 I hope it is fixed now May 06 11:33:13 Laibsch: works for me May 06 11:33:24 http://bugs.openembedded.net/buglist.cgi?cmdtype=runnamed&namedcmd=LastWeekOpened is only 20 bugs May 06 11:34:00 I miss weekly changelogs from Koen May 06 11:36:20 ~curse mtn log May 06 11:36:22 May you be reincarnated as a Windows XP administrator, mtn log ! May 06 11:38:19 why it does not have 'mtn log --revision 3242384y238927527502702'... May 06 11:38:31 hrw: well, I closed a few of them already May 06 11:38:41 "mtn log --from 23424525235262 --last 1" I would not call intuitive May 06 11:39:00 plus, of course we were down a few days because greg's server went dark suddenly May 06 11:39:23 plus, the DNS issue (which apparently kergoth finally updated) May 06 11:39:43 I'm sure we'll see more May 06 11:42:33 * Genesis is reading [oe] Git versus Hg thread , and wonder if a choice has been taken May 06 11:59:24 nope May 06 11:59:34 RP is still setting up a test server May 06 11:59:53 But I guess in light of the overwhelming tendency it is only a question of time May 06 12:00:06 * Laibsch would like to undelay the process, too May 06 12:02:14 cyrilRomain: uicmo4-native went fine for qemux86 here May 06 12:04:07 woglinde: your uclibc patch seems to get things building again :-) May 06 12:04:27 Laibsch: :) and :( May 06 12:04:46 cbrake hehe, I wonder why it was so long unoticed and the fix is really trival May 06 12:05:00 woglinde: yeah, very strange May 06 12:05:23 woglinde: I'll go head and push for 0.9.29 and _svn unless you know of any reason not to May 06 12:05:28 Laibsch: thanks anyway for the test May 06 12:05:44 crbake okay will do this evening May 06 12:06:03 woglinde: do you mind if I push now? May 06 12:06:27 cbrake no prob go ahead and close the bug please May 06 12:06:33 woglinde: ok May 06 12:08:57 Laibsch : i'm switching to git , i hope i don't do that too fast May 06 12:09:16 what do you mean? May 06 12:09:24 you are switching where and what? May 06 12:09:39 my local developement on oe May 06 12:09:55 for my company May 06 12:13:30 03cbrake 07org.oe.dev * rec52191a... 10/ (4 files in 3 dirs): uclibc_0.9.29/svn: add wcs_upper_buffer patch from Henning Heinold May 06 12:14:14 cbrake hm only for svn? May 06 12:14:35 woglinde: no, I added the patch to both the 0.9.29 and svn recipes May 06 12:14:51 cbrake *g* than better commenting this May 06 12:15:43 woglinde: yeah, I guess that is not obvious May 06 12:20:05 * * OE Bug 4208 has been RESOLVED (FIXED) by cliff.brake(AT)gmail.com May 06 12:20:07 * * uClibc-0.9.29 fails to build May 06 12:20:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4208 May 06 12:20:42 wow, IRC notifications from bugzilla and mtn sure are fast these days ... May 06 12:20:51 yepp May 06 12:28:50 03cbrake 07org.oe.dev * rf9a7308a... 10/ (11 files in 4 dirs): linux-2.6.25: add cm-x270 support May 06 12:44:59 * Laibsch is looking for a hint from you guys on how to fix May 06 12:45:02 !oebug 4212 May 06 12:45:03 * * Bug 4212, Status: UNCONFIRMED, Created: 2008-05-03 04:24 May 06 12:45:04 * * : touchscreen is (again) off by 90° May 06 12:45:05 * * http://bugs.openembedded.net/show_bug.cgi?id=4212 May 06 12:45:31 spitz touchscreen movements are off by 90° May 06 12:45:39 in opie May 06 12:45:48 Laibsch, hi May 06 12:45:53 I believe we had something similar in the past and it was fixed May 06 12:46:04 hi Marex, how are you doing? May 06 12:46:32 Laibsch, well examination period starts soon and I did some upstreaming to handhelds.org recently :) May 06 12:46:48 Laibsch: make it 90% wrong 4 times and all is fixed May 06 12:47:04 XorA: good idea ;-) May 06 12:47:10 How do I do that? May 06 12:47:33 more important: where? May 06 12:47:52 Laibsch, would you mind moving palmtt5 and palmld from linux-hnd to linux-hh in OE? May 06 12:48:06 no, not at all May 06 12:48:14 that'd be great :) May 06 12:48:33 since I have just this poor notebook now, setting up OE is a no go for me :( May 06 12:50:15 Marex: ping CoreDump about it, maybe he can help May 06 12:50:29 the notebook/OE problem May 06 12:51:03 Laibsch, it's not that it's impossible to set it up, it's just too slow machine for that May 06 12:51:12 I understand May 06 12:51:25 thanks :) May 06 12:51:31 You could try icecc as well May 06 12:51:40 you, I know about icecc May 06 12:51:53 s/you/yup/ May 06 12:52:48 but I just need some patch that removes palmtt5 and palmld from linux-hnd.bb and adds it to linux-handhelds.bb May 06 12:53:04 in most recent hh.org cvs, those machines have better support May 06 12:54:01 hmm ... now that I think of it. the patch will have to be a bit more complex, since patching machine files will be needed as well May 06 12:54:47 in -hh... May 06 12:54:50 OK May 06 12:54:58 Let me know when you have something May 06 12:55:05 best via BTS May 06 12:55:33 I wonder when -hh and -hnd guys will learn how to use mainline kernel more May 06 12:55:47 Laibsch, ok :) May 06 12:55:58 hi mickeyl May 06 12:56:04 hrw, mainline isnt ready for palm devices yet, there is lot of stuff missing May 06 12:56:34 hrw: how is "mtn pull|push" doing for you? May 06 12:56:36 Marex: but -hh is ready for them? May 06 12:56:48 It's on and off here May 06 12:56:50 moreover -hnd kernel contains so much crappy code that needs cleanup that it would be best to upstream to -hh and from there to mainline May 06 12:56:52 Laibsch: I pulled at 5 :00 May 06 12:56:56 hi hrw May 06 12:57:02 well, that is long ago May 06 12:57:13 hrw, yea, there are the necessary drivers already ;) May 06 12:57:17 Marex: -hh is stalled at 2.6.21 May 06 12:57:19 It seems to be starting to act up again May 06 12:57:55 hrw, I know :( May 06 12:58:41 Marex: thats why I wonder how moving to 4 releases old tree is good thing May 06 13:00:02 Laibsch: pull does not connect May 06 13:00:32 hrw, because there arent necessary drivers in mainline yet ;) May 06 13:00:54 I can upstream from handhelds.org without too much hassle anytime later May 06 13:01:11 marcan: then someone should extract those drivers, clean, review, push mainline May 06 13:01:43 hrw, hh people can do that :p May 06 13:02:46 Marex: I would not wait rather May 06 13:03:14 hrw, well since they are upstreaming something, I believe that the wm97xx driver will be upstreamed soon too May 06 13:03:29 and after that the palms can be upstreamed too May 06 13:03:49 03koen 07org.oe.dev * rf24ed430... 10/ (3 files in 3 dirs): angstrom: switch to opkg-nogpg since opkg crashed upon encountering .sig files (and signing Packages is bogus anyway, we want signed packages, not feeds) May 06 13:03:52 well at least those from -hh ... those that are still in -hnd are terrible mess and the code is horrible May 06 13:04:29 which wm97xx driver? May 06 13:04:53 the one from wolfson micro for touchscreen and sound May 06 13:05:03 or is that already in mainline ? May 06 13:05:13 you did not even checked??? May 06 13:05:24 umm ... May 06 13:05:42 ;D May 06 13:06:22 well my goal was handhelds.org first and mainline later May 06 13:06:37 I want to kill -hnd kernel tree even though some people dont like the idea May 06 13:07:08 Marex: what about moving to http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog team? May 06 13:07:25 I heard about it May 06 13:07:56 it also has hh branch May 06 13:08:08 I see May 06 13:08:15 that's nice :) May 06 13:12:05 * * OE Bug 4222 has been created by bugs.openembedded.org(AT)rolf.leggewie.biz May 06 13:12:07 * * bugs, mainly in Angstrom and derivatives May 06 13:12:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4222 May 06 13:14:06 ~seen bluelightning May 06 13:14:09 bluelightning was last seen on IRC in channel #opie, 7d 15h 14m 57s ago, saying: 'ok....'. May 06 13:42:44 morning all May 06 13:43:44 hey May 06 13:43:58 hey RP May 06 13:44:30 I've got meta-toolchain working for arm-darwin builds :) May 06 13:44:53 RP: I gave up on it and build an external toolchain by hand :( May 06 13:45:09 after almost a week I could not get any further May 06 13:45:27 Jin^eLD: What was your target again? May 06 13:45:42 RP: arm, eabi, but - uclibc - that's where the problem pops up May 06 13:46:08 I think the problem was you were trying to do this with the stable branch May 06 13:46:14 RP: I tested with .dev May 06 13:46:15 same stuff May 06 13:46:39 http://bugs.openembedded.net/show_bug.cgi?id=4203 May 06 13:46:40 hi rp May 06 13:47:08 the interesting thing is that *sometimes* gcc-cross-sdk will build, but then the resulting toolchain is broken, with the same libc.so.6 issue May 06 13:48:16 the only thing I can confirm is that the sources and patches used in OE are OK - I did bitbake -c patch xxx and then continued to ./configure manually, according to the CLFS book May 06 13:48:27 and that gave me a working external toolchain in the end May 06 13:50:39 Jin^eLD: Can you say whats different in the CLFS approach? May 06 13:50:57 Jin^eLD: Reading that bug report I really don't understand what the problem is :( May 06 13:50:59 RP: I tried to figure that out, but I really got lost in those tons of .inc files, bb files including each other and so on May 06 13:51:28 Trying to look at the gcc recipes in the stable branch will drive you mad... May 06 13:51:30 RP: uhm.. build failure of gcc-cross-sdk with the log I pasted? May 06 13:51:51 RP: exactly... that's what it did :) May 06 13:52:17 RP: if you look at the attachment, there is the compile log May 06 13:52:28 if you say that the bug description is unclear I can rewrite it May 06 13:52:43 Jin^eLD: Sometime if you could retry with the change I pushed to sdk.bbclass, that would be interesting May 06 13:52:56 RP: did that, you changed only one line right? May 06 13:53:11 Added STAGING_DIR_HOST? May 06 13:53:12 yeah May 06 13:53:16 Not the VENDOR one May 06 13:53:41 well, basically, after you commited sdk.bbclass I retried and still had no luck May 06 13:53:52 the change I made yesterday? May 06 13:53:54 yes May 06 13:53:58 ok :( May 06 13:54:10 especially the fact that *sometimes* gcc-cross-sdk does build through May 06 13:54:13 this is scary May 06 13:54:23 imho it should either always fail or always work May 06 13:54:28 you were using parallel make? May 06 13:54:32 yes May 06 13:54:40 ha! ok that could explain why it builds sometimes May 06 13:54:41 that'll be why May 06 13:54:52 but - even if it does build through - the resulting toolchain is broken May 06 13:55:02 so something in the whole process is still not OK May 06 13:55:04 its not just missing a symlink? May 06 13:55:22 I do not think so.. May 06 13:55:39 if I search for libc.so.6 in ${CROSS} and also in staging - there is no such thing there May 06 13:55:57 but still the toolchain that is used by OE internally does build May 06 13:55:59 and does function May 06 13:56:10 hmm. So its finding something from the host? :/ May 06 13:56:20 Where did we get with poky for comparision? May 06 13:56:42 I don't remember where it failed, but it did fail, I pasted the results but that was a couple of days back May 06 13:56:46 I tried pinky I think May 06 13:56:51 but did not get very far May 06 13:57:45 oops wrong window May 06 13:57:47 aem May 06 13:57:48 no not this May 06 13:57:50 the CLFS approach is actually quit esimple May 06 13:57:51 :) May 06 13:58:06 the wrong window landed in the wrong window itself ;) May 06 13:58:25 but if you look at the OE toolchain recipes in stable - really a nightmare May 06 13:58:43 I rewrote them, I know what they're like... May 06 13:58:54 yeah... I really admire your patience May 06 13:59:06 !oebug 3760 May 06 13:59:07 * * Bug 3760, Status: NEW, Created: 2008-02-08 10:10 May 06 13:59:08 * * : integrate bdf fonts into font-update-common May 06 13:59:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3760 May 06 13:59:17 hrw: that is the one I was talking about earlier May 06 13:59:27 nevertheless, its surprising that both, .dev and .stable show the same symttom, allthough the recipes are rewritten and even use sysroot in .dev May 06 13:59:27 Will you possibly take a look at it? May 06 13:59:56 Jin^eLD: I'm afraid I didn't see the poky error but if you can reprouce the failure there I stand a better chance of fixing it May 06 14:00:08 RP: ok, pinky or trunk? May 06 14:00:15 Jin^eLD: use trunk May 06 14:00:24 What is the relation between the A* 2007.12 branch and org.oe.stable? May 06 14:00:48 RP: ok, give me about 15-20 minutes :> May 06 14:01:14 Jin^eLD: I'm running a uclibc meta-toolchain build here too so we'll see what happens May 06 14:02:01 RP: thanks May 06 14:02:17 RP: even if it does build - try to compile a simple app with it May 06 14:02:22 i.e. with the resulting toolchian May 06 14:02:40 you should see the same error as in the bugreport May 06 14:02:41 Jin^eLD: will do May 06 14:03:07 IME if these things compile the rest is much easier May 06 14:03:57 well.. in both cases its related to libc.so.6, in one case it does not compile, in the other case it does compile but nothing can be built May 06 14:04:21 and the whole oe build directory has not a single libc.so.6 reference - and the internal toolchain works May 06 14:04:26 so very, very confusing May 06 14:04:47 It does sound very strange... May 06 14:05:31 it *is* very strange :) May 06 14:11:46 * Laibsch would like to make everyone aware that the bug tracker counts 177 (!) open bugs with patches attached May 06 14:11:50 http://bugs.openembedded.net/buglist.cgi?keywords=patch&resolution=--- May 06 14:12:08 Those are the ones I tagged May 06 14:12:18 What is the relation between the A* 2007.12 branch and org.oe.stable? May 06 14:12:37 Laibsch: I think stable replaces A*2007 May 06 14:12:47 Laibsch: but I'm not 100% sure May 06 14:14:11 it does May 06 14:14:17 it does May 06 14:14:28 4210 looks good May 06 14:14:55 03jeremy_laine 07org.oe.dev * re38c049f... 10/ (3 files in 3 dirs): gpm_1.20.1.bb: fix references to OPEN_MAX (Closes: #3762) May 06 14:15:05 * * OE Bug has been RESOLVED by jeremy.laine(AT)bolloretelecom.eu May 06 14:15:07 * * gpm-1.20.1 fails to build with >=kernel-2.6.23 May 06 14:15:09 * * http://bugs.openembedded.net/show_bug.cgi?id= May 06 14:16:00 sorry to bother you again, CoreDump|afk, but ^^^ May 06 14:16:06 Looks like the number is missing May 06 14:16:19 all of the sudden May 06 14:16:31 !oebug 3762 May 06 14:16:33 * * Bug 3762, Status: RESOLVED (FIXED), Created: 2008-02-10 09:57 May 06 14:16:34 * * c.romain(AT)laposte.net: gpm-1.20.1 fails to build with >=kernel-2.6.23 May 06 14:16:35 * * http://bugs.openembedded.net/show_bug.cgi?id=3762 May 06 14:16:36 zecke: I'll commit 4210 May 06 14:16:44 That one is my fault... May 06 14:18:12 * RP had that patch in a pending queue for poky already May 06 14:21:04 * * OE Bug 4210 has been RESOLVED (FIXED) by rpurdie(AT)rpsys.net May 06 14:21:07 * * Unrecognised cp switch used in autotools.bbclass May 06 14:21:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4210 May 06 14:22:16 03rpurdie 07org.oe.dev * rfaf8b98d... 10/ (1 classes/autotools.bbclass): autotools.bbclass: Remove -t option from cp, there is no longer any need for it (closes #4210) May 06 14:23:24 open-patch-bug count now at 183 May 06 14:24:32 Jin^eLD: I've built the toolchain and it errors when run with crt1.o being missing... May 06 14:25:05 * * OE Bug 4028 has been marked as DUPLICATE of bug 3522 by May 06 14:25:07 * * scummvm-0.9.1-autobuild May 06 14:25:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4028 May 06 14:26:12 RP: that's something new again... May 06 14:26:21 I did not have that error May 06 14:27:12 I wonder why meta-toolchain is so problematic, after all the "internal" toolchain always builds fine, and meta should not be much different, except for some paths May 06 14:27:41 What is the process for committing to .stable, then? ACKs from two devs? May 06 14:27:56 * Laibsch hasn't seen anything in that arena on oe-devel yet May 06 14:29:04 * * OE Bug 2464 has been marked as DUPLICATE of bug 3522 by May 06 14:29:07 * * scummvm doesn't build for qemux86 May 06 14:29:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2464 May 06 14:30:18 Laibsch: policy continues from 2007.1, two acks May 06 14:31:46 well, but I guess that is two OE devs now May 06 14:31:56 not every OE dev is an A* dev May 06 14:32:14 And I guess the discussion should be in oe-devel? May 06 14:32:26 We could also have that in the bug tracker which I think would be better May 06 14:32:40 I'll try out adding a couple of flags May 06 14:32:42 Laibsch: I think it was discussed on the angstrom list May 06 14:32:49 Hehe ;-) May 06 14:32:52 well May 06 14:33:07 Jin^eLD: Its a problem in poky May 06 14:34:09 ERROR: '/home/jin/Poky/poky/meta/packages/uclibc/uclibc_0.9.29.bb' failed May 06 14:34:19 Jin^eLD: With that fixed and the toolchain rebuilt, it builds a hello world fine May 06 14:34:19 you had something different? May 06 14:34:36 Jin^eLD: yes, that worked. Can you pastebin the error? May 06 14:34:45 yep, one moment May 06 14:35:25 hi any news on the planned git repository? May 06 14:35:30 http://pastebin.mozilla.org/423958 May 06 14:36:18 internal compiler error? :> May 06 14:36:46 maybe my local toolchain is borked May 06 14:36:50 gcc version 4.1.2 20070925 (Red Hat 4.1.2-27) May 06 14:37:48 Jin^eLD: Its an ICE in the cross compiler May 06 14:37:59 what is in ICE ? May 06 14:38:18 Jin^eLD: For some reason "/usr/include" in in the compiler commandline as a -I option which is bad May 06 14:38:22 Internal Compiler Error May 06 14:38:29 ah May 06 14:38:36 hm May 06 14:38:36 This is zecke's cross compiling badness patch kicking in May 06 14:38:53 why didn't it hit you? May 06 14:39:16 bp flag created, a request to grant bp results in a cc to oe-devel May 06 14:39:20 * Laibsch will mail the list about it May 06 14:39:21 I've very much like to know why I didn't see it May 06 14:39:26 after some tests May 06 14:39:45 RP: I'm on x86_64 May 06 14:39:50 I can retry on 32bit May 06 14:39:58 Jin^eLD: I was just wondering about i686. I'm on x86_64 May 06 14:40:00 not sure if that will make a difference May 06 14:40:08 hm May 06 14:40:20 so you too.. weird May 06 14:40:21 03tmbinc 07org.oe.dreambox * r296f3346... 10/ (1 packages/dreambox/dreambox-dvb-tools-v3.bb): dreambox-dvb-tools: update to 1.4 May 06 14:40:24 03tmbinc 07org.oe.dreambox * r5043ba81... 10/ (1 conf/distro/opendreambox-1.4.conf): opendreambox-1.4: also use autoconf 2.59 May 06 14:40:28 03tmbinc 07org.oe.dreambox * r3a88f17a... 10/ (6 files in 4 dirs): python-2.5.1: use -fno-strict-aliasing, fix package list May 06 14:40:32 03tmbinc 07org.oe.dreambox * r1a00ff48... 10/ (1 packages/enigma2/enigma2.bb): enigma2: install header files into staging May 06 14:40:55 I sourced the poky-init-build-env script, added POKYLIBC="uclibc" to conf/local.conf and started: bitbake meta-toolchain May 06 14:41:14 Jin^eLD: That should be fine May 06 14:41:30 Jin^eLD: This is also what poky's autobuilder does each night since it tests uclibc... May 06 14:41:55 * Jin^eLD feels like a magnet for errors May 06 14:42:25 Do you know how to make uclibc dump the commands its using? May 06 14:42:35 no May 06 14:42:45 me neither :( May 06 14:43:25 I will retry the build on a different machine May 06 14:43:30 just to see how that goes May 06 14:43:44 I'm using ubuntu FWIW May 06 14:43:47 after all I was on the quest for meta-toolchain, and now got kicked out half way :) May 06 14:43:51 I am on Fedora May 06 14:43:56 F7 on the build server May 06 14:44:02 I will try on older F6 now on my PC May 06 14:44:24 uclibc works for you in OE? May 06 14:44:44 yes May 06 14:45:07 * RP checks for differences poky <-> OE in uclibc May 06 14:45:07 I use angstrom 2007.12 from the stable branch in uclibc mode and everything works nicely May 06 14:45:15 ah, stable branch May 06 14:45:20 not .dev? May 06 14:45:31 I am on product dev, so I picked .stable :P May 06 14:45:48 I can retry .dev too, just to see how that goes May 06 14:47:56 Nothing jumps out as too different poky <-> OE.dev. OE stable is very different though May 06 14:48:37 but I hit the meta-toolchain problem in oe.dev as well May 06 14:50:03 The same ones or different? :) May 06 14:50:12 same :) May 06 14:50:25 Jin^eLD: uclibc from dev just built for me this morning after I applied Henning's patch. May 06 14:50:26 that's my 2nd comment in the bugsubmit, about a test with .dev May 06 14:50:59 cbrake: my problem is with meta-toolchain and uclibc, in both .stable and .dev May 06 14:51:12 either gcc-cross-sdk is failing for me, or it builds through but then the produced toolchain is broken May 06 14:51:20 uclibc as such did build for me May 06 14:51:21 cbrake: What was the patch for? May 06 14:52:12 RP: http://bugs.openembedded.net/show_bug.cgi?id=4208 May 06 14:53:59 cbrake: thanks, Thats doesn't look related to this issue :/ May 06 14:56:39 retrying on FC6 now May 06 15:00:10 Jin^eLD: Add export VERBOSE=1 to uclibc.inc and retry, that should give logs that are useful May 06 15:01:00 ok, one sec May 06 15:01:57 hmm, on FC6 I have a dependency loop :) May 06 15:02:02 what the hell is going on there? heh May 06 15:03:09 aah, have a typo in POKYLIBC :) May 06 15:03:40 the var name was different, so starging the FC6 thing again, and will do the verbose stuff on the build server May 06 15:04:05 * * OE Bug 4223 has been created by bugs.openembedded.org(AT)rolf.leggewie.biz May 06 15:04:07 * * backport be4a9a82ecf6de3dc409f7e421e00f48275630a4 May 06 15:04:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4223 May 06 15:05:17 tmbinc: Isn't steliosk trying to merge the dreambox branch? May 06 15:08:56 RP: this is with the export VERBOSE=1: http://www.deadlock.dhs.org/jin/poky-uclibc.txt May 06 15:10:10 Jin^eLD: ok, the problem is that -I/usr/include entry May 06 15:10:40 Jin^eLD: Can you look at the configure logs and see what KERNEL_HEADERS was set to? May 06 15:11:21 ok, one sec May 06 15:11:47 Linux kernel header location (KERNEL_HEADERS) [/usr/include] (NEW) * May 06 15:11:47 (I'm comparing your logs to mine, hence that guess) May 06 15:11:54 so that's probably the problem? May 06 15:12:07 yes, it should be set to ${STAGING_INCDIR} May 06 15:13:35 I wonder why it is set for you but not for me... May 06 15:13:49 do_configure is doing something very magic with perl May 06 15:13:59 btw, I get a deploop on fc6, so can't build poky trunk there May 06 15:14:02 http://www.deadlock.dhs.org/jin/poky-fc6.txt May 06 15:14:27 python 2.4? May 06 15:14:44 Python 2.4.4 (#1, Oct 23 2006, 13:58:00) May 06 15:14:54 I know who to beat up about that then ;-) May 06 15:15:13 :) May 06 15:15:51 well, as you see - things are not easy for me :) everything fails with various problems hehe May 06 15:16:10 must be my personal magic :> May 06 15:19:07 Jin^eLD: I'll get that python 2.4 issue fixed but in the meantime, this other problem... :) May 06 15:19:13 hehe May 06 15:19:27 Jin^eLD: can you share the .config from the uclibc build? May 06 15:19:51 Jin^eLD: I'm fairly sure that .config is corrupt somehow... May 06 15:20:20 http://www.deadlock.dhs.org/jin/uclibc-poky-config.txt May 06 15:20:35 oh wait, you mean the .config from my failing builds May 06 15:20:39 or from the ICE thing? May 06 15:20:46 i.e. - OE or Poky one? May 06 15:21:03 or in other words - which one of the problems are we talking about now? :) May 06 15:22:34 I mean from the one with the ICE and the bad include path May 06 15:22:43 with poky May 06 15:22:52 ah, ok, thats the one I pasted May 06 15:22:59 posted that is May 06 15:23:04 TARGET_alpha=y ? May 06 15:23:17 lol May 06 15:23:19 that config is insane May 06 15:23:23 yeah, that can't be right May 06 15:23:25 I know why May 06 15:23:39 building for machine which *lack* own config May 06 15:24:30 MACHINE ?= "qemuarm" May 06 15:25:09 RP: what machine were you building for? May 06 15:25:23 03thomas 07org.oe.dev * r2561fb41... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs.inc: bump packagekit version May 06 15:25:23 Jin^eLD: spitz May 06 15:25:28 03thomas 07org.oe.dev * r9a2dacaa... 10/ (1 packages/opkg/opkg.inc packages/opkg/opkg_svn.bb): opkg: update description and version May 06 15:25:31 Jin^eLD: which one did you use? May 06 15:25:57 RP: qemuarm in the poky build May 06 15:26:22 hmm, this sounds ominous... May 06 15:26:33 well, I can take nokia770 May 06 15:26:40 it is also arm926 May 06 15:26:46 and thats what I am building for May 06 15:35:14 XorA: does the wolfonmicro mirror push back to amethyst? May 06 15:35:49 Laibsch: it should do May 06 15:36:04 Laibsch: ask broonie if it got busted sometime May 06 15:36:33 well, because amethyst was misbehaving badly yesterday, I pulled and pulled ffrom there May 06 15:36:52 But this morning the commits hadn't made it to amethyst yet May 06 15:37:03 which is why was asking May 06 15:37:08 .dev branch? May 06 15:37:13 yes May 06 15:37:41 mtn-0.35: connecting to monotone.linuxtogo.org May 06 15:37:41 mtn-0.35: network error: failed to connect: Connection refused May 06 15:37:44 thats why May 06 15:38:23 RP: well, bad luck with Nokia too, I think the uclibc in Poky does not figure out the configuration like oe.dev May 06 15:38:34 I'll try to add one May 06 15:38:48 Laibsch: commits should arrive in a few mins May 06 15:39:15 I pushed them directly this morning and got through May 06 15:39:23 You only sync once a day? May 06 15:39:30 Laibsch: every 5 mins May 06 15:39:40 OK, what I thought May 06 15:39:52 Jin^eLD: Its missing machine.config files May 06 15:40:06 Jin^eLD: You can try copying the spitz directory to qemuarm May 06 15:40:28 * RP 's build has just exploded :/ May 06 15:41:10 Laibsch: mirror should be working again, I guess the ssh tunnel dissapeared sometime May 06 15:42:39 RP: I took one of the arm configs from oe.dev May 06 15:42:43 its building now May 06 15:44:46 Can somebody please log into the BTS and tell me if they have something meaningful under "My Requests"? May 06 15:45:01 IOW, is attachment 9339 shown there? May 06 15:45:21 bye all May 06 15:45:48 When you are logged in, only requests made by you or addressed to you are shown by default. You can change the criteria using the form below. When you are logged out, all pending requests that are not restricted to some group are shown by default. May 06 15:46:15 Laibsch: I see no mention of 9339 May 06 15:46:20 me neither May 06 15:47:08 Jin^eLD: I've fixed my build now and yes, it looks like qemuarm didn't work with uclibc... May 06 15:47:25 I'll at least make it error in future if the machine config doesn't exist May 06 15:47:37 Jin^eLD: Poky should work with python 2.4 again too May 06 15:47:53 let me SVN up on the other machine.. May 06 15:48:18 in the meantime its still building, so I am curious if the meta-toolchain will work out May 06 15:49:24 Jin^eLD, XorA: thanks for testing May 06 15:49:29 np May 06 15:50:01 Can you please check if you can grant (not request) the bp flag for http://bugs.openembedded.net/attachment.cgi?id=9339&action=edit ? May 06 15:50:39 Jin^eLD: Are you a member of the OE grou? May 06 15:50:43 group May 06 15:50:44 no May 06 15:50:49 OK May 06 15:50:52 so I think I will not be able to check that May 06 15:51:05 in the flags I only have "?" in the drop down May 06 15:51:19 Laibsch: I think that works May 06 15:51:20 good May 06 15:51:23 actually one item in the list is empty, the second one is "?" May 06 15:51:43 Jin^eLD: yes, that is intended May 06 15:51:48 k :) May 06 15:51:49 ? is "request" May 06 15:51:58 " " is withdraw request May 06 15:52:08 Jin^eLD: Can you withdraw the request I made? May 06 15:52:17 + is grant May 06 15:52:21 - is deny May 06 15:52:42 uhm May 06 15:52:47 XorA: You have a + to grant the request? May 06 15:52:56 Someone else has made changes to attachment 9339 of bug 4223 at the same time you were trying to. The changes made were: May 06 15:53:02 Laibsch: I did + then - then blank May 06 15:53:06 Laibsch: all seem to do something May 06 15:53:12 OK, nice May 06 15:53:25 Explanation is above May 06 15:53:25 ah, that someone else was XorA then :) May 06 15:53:41 Well, you could attach any kind of semantics to it, I guess May 06 15:53:50 But those seemed kind of natural to me ;-) May 06 15:54:11 oh we have a bugzilla with review flags? May 06 15:54:25 Laibsch: at webkit we have: http://webkit.org/pending-review May 06 15:54:34 Laibsch: and http://webkit.org/pending-commit May 06 15:54:36 Laibsch: hmm, can you make it so same person cant ack twice? May 06 15:55:06 zecke: we are trying that out now May 06 15:55:24 XorA: Yeah, I found that annoying too May 06 15:55:47 But #mozwebtools people think that is the way it SHOULD work May 06 15:55:56 It is called multiplicable May 06 15:56:05 We can have one of two things May 06 15:56:28 a) flags can be raised more than once, despite the problems you mention, I favour this May 06 15:56:44 Laibsch: and My Requests shows properly May 06 15:57:02 Laibsch: just need an all pending requests :-) May 06 15:57:04 well, you have made your own requests now May 06 15:57:28 the "all pending" requests was what I was looking for, too May 06 15:57:42 But you can use a search for that ( and save the search) May 06 15:58:01 Jin^eLD: I've tweaks poky to make that misconfiguration a lot more obvious May 06 15:58:13 XorA: I just sent http://tinyurl.com/5v993n to the list May 06 15:58:47 XorA: I favour a) not least because it is then possible for two people to disagree May 06 15:59:49 If we go b), we'd have to follow this process 1) $user requests bp, 2) $DEV1 ACKs as comment, 3) $DEV2 grants bp flag May 06 16:00:34 b) has the further problem that $DEV1 could be $USERdisguisingASdev and $DEV2 has to catch that May 06 16:01:02 with a) bugzilla takes care of who has permission to do what May 06 16:02:00 zecke: I think, I'll add the review flag May 06 16:02:13 Sort of like May 06 16:02:17 !oebug 2194 May 06 16:02:18 * * Bug 2194, Status: NEW, Created: 2007-05-01 05:03 May 06 16:02:19 * * : meta-bug: bugs waiting to be committed and fixed May 06 16:02:20 * * http://bugs.openembedded.net/show_bug.cgi?id=2194 May 06 16:02:37 Stuff that I did not feel comfortable committing all on my own May 06 16:02:49 The pending-commit thing seems a bit strange May 06 16:02:54 What is that for? May 06 16:03:06 Once I set that flag, I could just commit it as well May 06 16:03:24 RP: k... still building May 06 16:03:40 RP: why is it trying to build sdl when I just wanted meta-toolchain? :> May 06 16:04:24 Jin^eLD: poky ships qemu as part of the toolchain/sdk May 06 16:05:07 Tweak packages/tasks/task-sdk-host.bb if it bothers you ;-) May 06 16:05:19 :) May 06 16:05:38 it started... building gcc 3.4.6 now for whatever reason May 06 16:05:42 well, I'll just let it do its thing May 06 16:05:48 and see what the result is May 06 16:05:51 to build qemu ;-) May 06 16:06:05 heh May 06 16:06:15 that will take a while :) May 06 16:06:20 its horrible to get right, hence I let poky do it rather than trying to explain to people... May 06 16:07:18 hi? how can i remove a runtime dependancy (which is not really necessary, though)? May 06 16:07:28 it's psplash to be precise May 06 16:07:54 it's nowhere in RDEPENDS/RRECOMMENDS, i've check twice and n-ce May 06 16:10:17 grrr uicmoc4 does build here because of an infinite loop in QString::replace in qtopia May 06 16:11:15 how is that possible no one else has this issue, and I always encounter such problem when playing with oe :-/ May 06 16:12:21 * cyrilRomain diffing qtopia qstring.h and qt qtring.h May 06 16:15:53 mrt2k7: it has to be in there somewhere May 06 16:16:01 mrt2k7: check with bitbake -g May 06 16:16:53 oh, nice switch May 06 16:16:58 with which package? May 06 16:17:07 psplash or task-base? May 06 16:17:10 the target causing the problem May 06 16:19:10 okay, what do i do with the dotfile? May 06 16:19:23 look at it May 06 16:19:35 sadly its too big for most graphical viewers... May 06 16:19:58 graphviz?! May 06 16:20:09 or something like that? May 06 16:20:32 it will be too big to work with them :( May 06 16:20:41 i'd like to give it a try May 06 16:20:52 feel free them May 06 16:20:54 then May 06 16:20:55 could be nice for my diploma thesis ;) May 06 16:21:02 (page filling ;)) May 06 16:23:02 "task-base.do_package_write_ipk" -> "psplash.do_package" ? May 06 16:23:42 so something in task-base is triggering a dependency on psplash May 06 16:24:11 erm, yeah, i knew that before :-/ May 06 16:24:19 but it's nowhere configured to be this way May 06 16:25:08 If its referenced in any package task-base generates it will be pulled in due to the renaming in debian.bbclass May 06 16:25:18 hmmm May 06 16:36:32 XorA: on the upside: the flag is prepended with the account name, so the same person ACK'ing twice is usually easy to spot May 06 16:36:43 the dp in front of your vote stands for you May 06 16:36:44 Laibsch: yeah May 06 16:36:49 RP: NOTE: package meta-toolchain-1.0: completed May 06 16:36:50 :> May 06 16:36:51 So, I guess it is OK May 06 16:36:56 Laibsch: pity bugtracker logins are emails though, ah well May 06 16:37:17 yes, certainly not perfect May 06 16:37:21 But workable I guess May 06 16:37:37 And a huge improvement over the cumbersome BACKPORTS.txt May 06 16:39:40 tssssss....i got it May 06 16:40:09 i'm using a local overlay for my stuff, and somehow some parts where taken from the .stable branch May 06 16:40:26 although i have these parts in my local overlay too an the priority is higher May 06 16:40:35 Laibsch: I only know this flag from webkit.org. We have a contribution guideline there May 06 16:40:59 Jin^eLD: yay. But does it work? May 06 16:41:03 no :) May 06 16:41:11 Laibsch: As contributor, you ask for review on a patch. Some of the reviewers say r=- or r=me (+) on the bug report May 06 16:41:19 RP: /usr/local/poky/eabi-uclibc/arm/lib/gcc/arm-poky-linux-uclibcgnueabi/4.2.3/../../../../arm-poky-linux-uclibcgnueabi/bin/ld: crt1.o: No such file: No such file or directory May 06 16:41:35 which seems to be different from the error that I had in OE stable/dev May 06 16:41:37 Jin^eLD: svn up and retry ;-) May 06 16:41:38 Laibsch: other people can land the patches marked as r=+ May 06 16:41:46 Jin^eLD: I've fixed that problem May 06 16:41:57 oh ok :) May 06 16:42:03 Laibsch: the upside is that people like mjs and darin can say yes to patches but don't need to waste the time to merge, update the changelog, commit them May 06 16:42:27 Laibsch: ooops, better remove all my acks, I didnt realise that was a real bug and didnt read the content :-) May 06 16:43:38 XorA: No way back, there ;-) May 06 16:43:42 You're trapped May 06 16:43:46 in a corner May 06 16:43:46 RP: indeed, it seems to work now May 06 16:43:49 back against the wall May 06 16:43:51 ;-) May 06 16:43:58 RP: so.. the big question is - what's different with OE then? :> May 06 16:44:05 zecke: Does the guide have an URL? May 06 16:44:17 Jin^eLD: gcc 4.2.3? May 06 16:44:25 maybe... May 06 16:44:29 Jin^eLD: I don't know to be honest... May 06 16:44:34 Laibsch: even a URI May 06 16:44:42 but then again, .stable and .dev have different gcc versions May 06 16:44:46 and I still had the same problem May 06 16:44:50 Laibsch: http://webkit.org/coding/contributing.html May 06 16:45:04 and I was able to build the toolchain based on gcc 4.1.2 by hand, from OE sources May 06 16:45:40 Jin^eLD: Now you have a known good set of .bb files, finding the problem may be easier May 06 16:45:47 or just use poky ;-) May 06 16:46:03 heh, now I have three things to chose from May 06 16:47:05 RP: thanks for the hints May 06 16:47:26 Jin^eLD: np, its addresses a couple of bugs in poky too which is good May 06 16:47:30 maybe I have more luck figuring out what is missing in OE May 06 16:49:13 hopefully, yes May 06 16:53:05 bonne nuit May 06 16:57:07 RP: btw, why do you use --disable-__cxa_atexit ? May 06 17:03:04 RP: and who is setting the value of BUILDSDK_CFLAGS ? May 06 17:03:11 couldnt find it yet May 06 17:16:12 zecke_: did you ever had an 'infinite compilation loop' building uicmoc4-native-4.3.3 (rcc ... qrc_network.cpp) ? May 06 17:17:29 cyril_: no May 06 17:17:35 cyril_: got a clock skew? May 06 17:18:01 zecke_: no :-/ May 06 17:18:29 zecke_: I tested on two different build machine, same loop when rebuilding uicmoc4-native-4.3.3 May 06 17:18:50 it loops line 633: http://pastebin.ca/1009514 May 06 17:19:43 this is really strange because I wrote a simple main.cpp trying to reproduce the loop, but it built and run fine with both qt4.3.3 and qtopia4.3.3 May 06 17:20:21 cyril_: initName.replace(QRegExp(QLatin1String("[^a-zA-Z0-9_]")), QLatin1String("_")); that is looping? May 06 17:21:08 zecke_: I strongly suspect (I hadded 'fprintf(stderr' in the rcc.cpp file) May 06 17:21:24 but fprintf is not a real debug tool I know May 06 17:21:35 cyril_: use gdb? May 06 17:22:38 zecke_: yes I should, but the executable is rcc, which is built during the uicmoc build May 06 17:23:32 * cyril_ feels too cursed and too lazy May 06 17:25:16 cyril|food: hehe, well you can invoke rcc on the file yourself May 06 17:45:09 XorA|gone: Do you still want to remove that bp+? May 06 17:45:17 Just select that empty thing for it May 06 17:51:05 Laibsch: removed, we can do it right from now on ;-) May 06 17:52:37 great May 06 17:58:54 Laibsch : I am still trying. Have a lot of 2 way merges to perform May 06 17:59:18 How does that work? May 06 17:59:27 with mtn's help or "on foot"? May 06 17:59:37 Where is the list of open tasks? May 06 17:59:46 Can mtn generate that list? May 06 18:20:23 Jin^eLD: Its set from bitbake.conf May 06 18:20:34 Thats a poky only tweak which probably should be removed May 06 18:20:52 and I don't know about the atexit stuff May 06 18:29:13 aah indeed, I missed it in bitbake.conf May 06 18:34:04 zecke_: I invoked rcc myself with gdb ... and there is indeed an infinite loop in qstring::replace, more deeply in QRegExpEngine::parse :-/ May 06 18:34:23 RP: do you use parallel make btw? May 06 18:35:36 Laibsch : well got some help from mtn dev on irc May 06 18:35:42 but the basic idea is this May 06 18:36:26 create a branch of .dreambox named dreamdev ;) and fix the non-content errors May 06 18:36:39 (that is what i ahve done) May 06 18:36:54 cyrilRomain: on which string? May 06 18:37:03 then propagate from .dev and fix 2 way merges that are required May 06 18:37:15 cyrilRomain: but this violates everything I know about finite state automats... May 06 18:37:47 once this is done, push the branch to the main repo so people can test/develop further May 06 18:37:59 once things are stable propagate to .dev May 06 18:40:35 Jin^eLD: yes, heavily May 06 18:40:54 RP: just checking... was thinking if that may be related to the problem May 06 18:42:21 RP: and there are indeed not much differences between .dev and .poky in regard to the meta stuff May 06 18:43:01 meta-toolchain.bb itself is quit edifferent though May 06 18:44:57 The one in OE doesn't take advantage of sysroot yet as far as I remember May 06 18:45:26 I see... but well, the rest is quite similar May 06 18:45:34 so still not sure what the missing part is May 06 18:46:28 Jin^eLD: grep meta-toolchain.bb for libc.so.6 May 06 18:47:21 poky got rid of that mess as part of its cleanup by the looks of things :/ May 06 18:47:29 holy crap.. May 06 18:47:51 what does that do? GROUP ( libc.so.6 libc_nonshared.a ) May 06 18:49:10 RP: well, wait... .dev has the same line in there May 06 18:49:27 Jin^eLD: Its in .dev, its not in poky... May 06 18:49:59 aah, I got all confused heh May 06 18:50:16 I think its historical junk May 06 18:50:29 We did used to have junk like this in gcc too May 06 18:50:50 OK.. that could explain why the generated toolchain looks for libc.so.6 in case gcc-cross-sdk does not fail May 06 18:51:07 ah, no, its in glibc recipes May 06 18:51:38 anyhow, we can get rid of this from meta-toolchain by embrasing the sysroot changes May 06 18:52:30 should the meta-toolchian.bb from poky work in .dev? May 06 18:52:58 Jin^eLD: not without tweaks. I'll try and merge poky's changes in if you give me 5 mins May 06 18:53:05 thanks! May 06 18:56:13 zecke_: the string is _network (at the end of execution of line 632 of my pastebin) May 06 18:57:29 zecke_: it violates also my understanding of finite state automata.. May 06 18:57:59 * cyrilRomain learning gdb May 06 18:58:25 * cyrilRomain too bad to work only on visual studio at work :-/ May 06 18:58:36 cyrilRomain: There is a nice section in the poky manual about using it remotely May 06 18:58:50 Jin^eLD: mtn pull ;-) May 06 18:59:04 :> May 06 18:59:22 03rpurdie 07org.oe.dev * rdcdf37dd... 10/ (1 packages/meta/meta-toolchain.bb): meta-toolchain: Sync with Poky to take full advantage of the sysroot simplifications and add various features. Also unbreaks uclibc toolchains May 06 18:59:35 RP you rock :> May 06 18:59:38 hurray May 06 19:00:36 RP: thanks for the info May 06 19:01:28 Jin^eLD: I'm pretty sure thats it but I'll await your confirmation before celebrating :) May 06 19:01:53 * RP -> back shortly May 06 19:02:00 RP: ok.. and indeed I have to see if I actually come that far, since gcc-cross-sdk was borking out at times May 06 19:19:46 morning May 06 19:19:56 hi hrw May 06 19:20:10 I finally got fedora 9 to build image May 06 19:20:19 :> May 06 19:20:27 re hrw May 06 19:20:41 btw what part of OE is printing stuff like this: NOTE: Couldn't find shared library provider for libc.so.6 May 06 19:24:07 gcc-cross May 06 19:24:21 hrw: what's the trick? I need to upgrade my OE build machine; it's currently running F7 and I'd like to skip F8 if I can make it work with F9 instead. May 06 19:24:37 mwester: first part is patching ipkg-utils-native May 06 19:24:49 second would be fixing fedora kernel to get qemu working May 06 19:25:07 but maybe sysfs tweaking will be enough May 06 19:25:09 That's pretty standard with all RH kernels, of course. :( May 06 19:25:28 RP: I did not get far enough to test the meta-toolchain.bb in oe.dev, it failed again at gcc-cross-sdk: http://www.deadlock.dhs.org/jin/oe-dev-gcc-cross-sdk.txt May 06 19:25:41 sysfs might be able to remove the need to un-patch the RH kernels? That would be good. May 06 19:25:57 RP: and I also find it strange that there are messages like "NOTE: Couldn't find shared library provider for libc.so.6" when I am building for uclibc May 06 19:27:19 mwester: /proc/sys/vm/mmap_min_addr is 65536 - setting it to 4096 fix some systems May 06 19:28:03 * mwester notes to try that, as he's quite tired of unpatching RH kernels with each update. May 06 19:35:25 http://hell.pl/yaal/tmp/Splash.gif May 06 19:35:33 how many colours do you see? May 06 19:36:37 8? May 06 19:36:37 huh? May 06 19:36:42 well, its surely not over 16 May 06 19:37:18 nice gfx? May 06 19:37:52 considering the low amount of colors - yes May 06 19:40:22 http://open.neurostechnology.com/content/neuros-and-texas-instruments-create-new-bounty-program-next-gen-open-internet-television-pla May 06 19:46:05 hrw, do you think it's safe to have 81˚C system temperature ? :) May 06 19:46:34 Marex: boiling some water for tea, are you? May 06 19:47:03 mwester, compiling OE :p May 06 19:47:09 on a laptop May 06 19:47:13 Ouch! May 06 19:47:32 Prop it up so the bottom gets some airflow! May 06 19:47:48 yea, it looks like it will melt soon :) May 06 19:47:51 * Crofton got one of those laptop coolers .... May 06 19:47:57 Marex: on laptop? May 06 19:48:12 hrw, do I have any other option ? May 06 19:48:18 Marex: I had such one and stopped build to make cpu cool a bit May 06 19:49:10 hmm :) May 06 19:49:20 guys... with binutils-cross-sdk: NOTE: Couldn't find shared library provider for libc.so.6 - but not with binutils-cross! any idea why? May 06 19:49:29 something is messed up in the sdk recipes May 06 19:49:41 polyonymous, hi, long time no see :) May 06 19:49:52 someone wants linux journal subscribtion (including t-shirt)? only 30 usd for us citizens May 06 19:49:55 Hey, Marex May 06 19:56:31 RP: if you find some time, I added my findings to #4203, the fix in meta-toolchain does make a difference for the case that gcc-cross-sdk succeeds, but that's still not it; I think there is not much more missing though May 06 19:56:40 and thx again for helping out :> May 06 20:02:40 RP: and this Note thing does not happen in Poky! so the fix must be there :> May 06 20:05:17 bye May 06 20:11:42 Jin|away: The message is from the shlibs code in package.bbclass May 06 20:11:56 Jin|away: Are you mixing glibc and uclibc builds in the same tree? May 06 20:12:17 (Thats supposed to work, I'm just wondering if there is cross contamination) May 06 20:26:31 hello y'all May 06 20:27:06 if I don't have ~/.profile, /etc/profile how do I setup the script for bitbake? May 06 20:34:18 is anyone here familiar with the dm355 from TI ? May 06 20:39:42 btw the error I'm getting with BabelO's image ... 6 sec vid http://www.zippyvideos.com/8717441227574186/pict6743/ May 06 20:40:48 sorry, wrong chat May 06 21:01:42 mwester: fedora9 built binary locales May 06 21:01:55 mwester: with standard kernel May 06 21:02:01 I am dissapointed May 06 21:02:01 ;D May 06 21:03:00 That's excellent!! I'll be upgrading my build system as soon as I can then! May 06 21:03:46 hrw|gone fedora9 what ? May 06 21:04:23 gremlin[it]: first rh distro which did not use broken kernel May 06 21:04:54 and I like fedora kernel response to 'vga=ask' May 06 21:07:03 ohhhh !!!! May 06 21:07:20 i'm using fedora since beginning ;) May 06 21:07:44 I used rh5.1 once May 06 21:07:52 no more rpm crap for me May 06 21:08:05 ahhh no my first redhat was 4.0 ... and for work 5.0 :) May 06 21:08:22 gremlin[it]: that 5.1 was m68k unofficial build ;D May 06 21:09:05 ohhhh cute !!! :) May 06 21:09:37 anyway Debian forever May 06 21:10:22 Jin|away: Can you check some things for me - in that build where gcc-cross fails, did any glibc* task run? (check the stamps directory). Also, is libc.so in staging a text file with libc.so.6 in it? May 06 21:10:48 Jin|away: My best guess is that one of the glibc tasks staged and broke things May 06 21:21:00 hello all! I just downloaded as GettingStarted mentined but seens like my local repo isn't equal the one in http://amethyst.openembedded.net/oe/viewmtn/viewmtn.py/ (I don't have the org.openmoko.dev branch and some others) is that normal? May 06 21:36:17 03crofton 07org.oe.dev * r11ca7ff0... 10/ (1 packages/omniorb/omniorb_4.0.7.bb): omniorb_4.0.7.bb : Fix typo in SRC_URI. May 06 21:37:21 poli: Usually you'd only check out one branch at a time May 06 21:39:43 RP: I understand that. I am just wondering how can I get this particular openmoko branch that I can see in the mtnview of the repo but not in my repo. May 06 22:10:30 poli: You may need to explictly pull it May 06 22:14:17 RP: thanks May 06 22:29:31 03utx 07org.oe.dev * r8dd69611... 10/ (1 packages/udev/udev-118/init packages/udev/udev_118.bb): May 06 22:29:31 udev-118: fixed init failure if /dev/null exist: May 06 22:29:31 * script failed on mknod for some reason, so check /dev/null before calling May 06 22:29:31 mknod May 06 22:29:31 * separate mknod is needed, otherwise ugly errors can appear May 06 22:29:32 * /dev/null not being character device will still cause failure May 06 22:41:33 /dev/null is not a character device? May 06 22:41:58 * mwester wonders if he misunderstands the commit comments. May 06 22:42:10 mwester: In case there's not a real /dev/null there, I'm guessing May 06 22:42:59 ah, "*if* /dev/null is not a character device, then there will be a failure" May 06 22:43:04 :) May 06 22:43:24 Yep :) May 06 22:56:07 * * OE Bug 4218 has been RESOLVED (FIXED) by May 06 22:56:09 * * naim May 06 22:56:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4218 May 06 23:02:50 03Laibsch 07org.oe.dev * r89a03be7... 10/ (1 packages/naim packages/naim/naim_0.11.8.3.1.bb): naim: add 0.11.8.3.1 of this console chat client (courtesy jlime project). Closes 4218. May 06 23:02:55 03Laibsch 07org.oe.dev * r0b294318... 10/ (1 packages/quake/quetoo_0.6.1.bb): quake: fix SRC_URI for quetoo (courtesy jlime project) May 06 23:07:17 can anyone tell me why the helloworld example compiles to helloworld-dev and helloworld-dbg rather than the expected helloworld? May 06 23:19:23 oe May 06 23:33:29 03Laibsch 07org.oe.dev * r8e375838... 10/ (3 files in 3 dirs): scummvm: target check too restrictive. Ease up. Closes bugs 3522, 3594, 3572, 4028 and 2464. May 06 23:33:31 oe May 06 23:39:25 oe May 06 23:42:04 * * OE Bug 3522 has been RESOLVED (FIXED) by May 06 23:42:07 * * scummvm 0.9.1 does not configure with arm-angstrom-linux cross compiler - patch included May 06 23:42:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3522 May 06 23:42:21 oe May 06 23:43:34 lugburtz: please go away May 06 23:44:13 why May 06 23:44:38 either go away, shut up or say something productive May 06 23:44:41 It's that easy May 06 23:45:24 oe May 06 23:48:19 poor Laibsch May 06 23:49:46 oe May 06 23:50:28 oe May 06 23:50:36 indeed May 06 23:50:56 !learn Looooooooong_Cat Longcat is loooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooong May 06 23:50:57 Added "looooooooong_cat" to database May 06 23:51:24 oe May 06 23:52:14 can anyone tell me why the helloworld example compiles to helloworld-dev and helloworld-dbg rather than the expected helloworld? May 06 23:52:29 oe May 06 23:55:13 oe May 07 00:23:30 03Laibsch 07org.oe.dev * r69dbe4f4... 10/ (1 packages/scummvm/files/scummvm-targetcheck.patch): scummvm: send scummvm-targetcheck.patch upstream May 07 00:23:35 03Laibsch 07org.oe.dev * r8fc46c9e... 10/ (1 packages/scummvm/files/scummvm-targetcheck.patch): scummvm: scummvm-targetcheck.patch accepted upstream May 07 00:29:42 hrw|gone, where did you find a F9 machine? May 07 00:46:41 03Laibsch 07org.oe.dev * r6333f882... 10/ (1 packages/frobtads packages/frobtads/frobtads_0.8.bb): frobtads: add version 0.8. Courtesy of jlime project. Closes 4219, May 07 00:51:16 03tmbinc 07org.oe.dreambox * r09bfa2d6... 10/ (1 packages/dreambox/dreambox-secondstage.bb): dreambox-secondstage: use version 59 for dm800 May 07 01:34:51 !forget Looooooooong_Cat May 07 01:34:53 Unable to delete, definition not found **** ENDING LOGGING AT Wed May 07 02:59:56 2008