**** BEGIN LOGGING AT Fri Jun 29 02:59:56 2007 Jun 29 03:36:01 * * OE Bug 2576 has been created by marek.vasut(AT)gmail.com Jun 29 03:36:03 * * Add iwmmxt support to mplayer for palmld Jun 29 03:36:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2576 Jun 29 04:24:59 03rwhitby 07org.oe.dev * rf0a697c2... 10/ (4 files in 3 dirs): ixp4xx-kernel: Updated to 2.6.21.5 for stable version Jun 29 04:25:05 03rwhitby 07org.oe.dev * rd24b671a... 10/ (3 files in 3 dirs): ixp4xx-kernel: Added 2.6.22 defconfig, untested, and not selected by default Jun 29 04:25:10 03rwhitby 07org.oe.dev * r2e2b20a7... 10/ (1 packages/apex/apex_1.5.6.bb packages/apex/apex_1.5.8.bb): apex: updated to 1.5.8 Jun 29 04:25:16 03rwhitby 07org.oe.dev * r4ee29061... 10/ (1 packages/tasks/task-slugos.bb): task-slugos: Added madwifi drivers and tools Jun 29 05:31:14 morning Jun 29 06:28:09 03rwhitby 07org.oe.dev * r6e7ec870... 10/ (1 packages/apex/nslu2/defconfig): apex: Updated defconfig for 1.5.8 Jun 29 06:48:19 good morning Jun 29 06:49:00 * * OE Bug 2577 has been created by jbs(AT)bandos.homelinux.com Jun 29 06:49:02 * * Added iwmmxt support to mplayer for akita Jun 29 06:49:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2577 Jun 29 07:03:49 hi Bernardo Jun 29 07:04:12 Bernardo: you have prism wifi card? Jun 29 07:19:21 hrw: I have a sandisk connect plus Jun 29 07:19:28 I think it is a prism 3 Jun 29 07:20:19 I do know it needs to load firmware, and sometimes has a race condition between loading the hostap* drivers and the ide_cs module Jun 29 07:20:50 we disabled ide_cs for that card Jun 29 07:20:55 but can't recall now if it is prism 2.5 or 3, and have no easy way to check Jun 29 07:21:04 and xora will fight for fixing it Jun 29 07:21:07 its prism3 Jun 29 07:21:09 yep, but it sometimes worked on my older laptop Jun 29 07:21:32 my new laptop has a card express adapter instead of a pcmcia, so I can test it Jun 29 07:21:50 I fixed hostap<>orinoco fight on my zaurus today Jun 29 07:22:19 now need to find out how to make it in OE way Jun 29 07:24:41 pcmcia/cf machines needs full modprobe ;( Jun 29 07:24:46 hi Rod Jun 29 07:25:06 :( Jun 29 07:26:08 well, I have to leave for work now, if you have a patch/something you need me to test, you can dcc it now. My work email is too unreliable (over-agressive spam filter) Jun 29 07:32:07 will handle it with few other guys but thx for help Jun 29 07:32:33 ok, anything I can do Jun 29 07:32:49 bye guys, CUL Jun 29 07:42:25 rwhitby: does slug can have kernel without reflashing? Jun 29 07:45:38 morning Jun 29 07:46:34 hi XorA Jun 29 07:50:38 XorA: you have wcf11 right? Jun 29 07:51:15 12 Jun 29 07:51:39 hostap and orinoco loaded on insert? Jun 29 07:51:56 only hostap Jun 29 07:52:38 you have orinoco in rootfs? Jun 29 07:52:56 * XorA checks Jun 29 07:53:37 no orinoco on my machine Jun 29 07:54:06 damn wombles Jun 29 07:54:17 k Jun 29 07:55:44 hrw: maybe we should organise a "plugfest" at guadec Jun 29 07:56:02 good morning Jun 29 07:57:55 ;) Jun 29 07:58:13 XorA: I do not plan to take any other toys then d400 Jun 29 07:58:23 d400? Jun 29 07:58:49 dell d400 laptop Jun 29 08:14:02 fsck another bomb in london Jun 29 08:24:25 hrw: you forget to remove the if statement around secondary firmware download in the prism-firmware you uploaded Jun 29 08:26:11 Hi Jun 29 08:26:55 hrw: therefore preventing STA download to PRISMIII cards with no flash Jun 29 08:27:50 please fix then Jun 29 08:29:35 * hrw woke up >5 hours ago Jun 29 08:30:46 pushed Jun 29 08:31:29 thx Jun 29 08:40:36 hi mmp Jun 29 08:44:58 03xora 07org.oe.dev * ra72b36a4... 10/ (3 files in 3 dirs): Jun 29 08:44:58 prism-firmware.bb : remove if around secondary firmware download as we always Jun 29 08:44:58 want to do this and logic as presented prevented download on flashless cards. Jun 29 09:20:36 hi, all! Jun 29 09:21:40 do anybody remembers of any method to install all generated module packages to image? (I mean automagically) Jun 29 09:22:43 ipkg install kernel-modules Jun 29 09:23:11 hrw, so it is meta-package? thanks! Jun 29 09:23:18 ipkg status | grep kernel-module- | grep Package | cut -f 2 -d " " | xargs ipkg install :-) Jun 29 09:24:14 XorA, well, I'd like to have it possible at image build time... :) Jun 29 10:35:39 re Jun 29 10:35:45 I would like to do an interview with all people from oe and put them on the wiki Jun 29 10:35:56 gerwinin: 'People behind OE'? Jun 29 10:36:00 Yeah Jun 29 10:36:08 gerwinin: like 'People behind KDE' has Jun 29 10:36:13 hrw: yes Jun 29 10:36:23 gerwinin: I was thinking about it years ago but lacked time Jun 29 10:37:07 hrw: I don't have much time either but as long as we need to extend the wiki a bit Jun 29 10:37:18 gerwinin: I will answer to questions Jun 29 10:37:27 hrw: cool Jun 29 10:37:41 hrw: I will send you a list of questions before next monday Jun 29 10:37:45 ok Jun 29 10:37:51 hrw: than you will be the first one :) Jun 29 10:37:54 ;D Jun 29 10:38:03 tester ;) Jun 29 10:38:38 I was also wondering how things with trolltech where going Jun 29 10:39:29 guinea pig :) Jun 29 10:39:42 Florian: is it time for a new newsletter as well ? Jun 29 10:40:07 I am also looking for possibilities to get some people over to Linuxworld San Fransisco Jun 29 10:41:23 because im to lazy to bugtrack it, wifi-radar.bb is missing a rdepend. it also requires python-pycairo. Jun 29 10:41:33 So in case I succeed I am looking for two volunteers Jun 29 10:42:04 wick900c: and we are lazy to care about such reports Jun 29 10:42:28 heh. Jun 29 10:42:49 wick900c: bugtracker is the only one proper way Jun 29 10:43:15 yea yea, but its a pain. Jun 29 10:43:18 gerwinin: Yes that would be good... I guess that has been neglected since mickeyl started to become busy. Jun 29 11:01:39 florian: hi:) Jun 29 11:02:18 sorry, I went to school and forgot to /away myself:) Jun 29 11:03:04 :-) Jun 29 11:04:29 gerwinin: about linuxworld - if possible I would avoid hh.org sharing Jun 29 11:05:47 heh Jun 29 11:06:11 hh.org is there? Jun 29 11:09:53 read oe ml Jun 29 11:37:04 ok Jun 29 11:37:07 hi koen Jun 29 11:51:10 hrw: why ? Jun 29 11:53:13 because hh.org is on many peoples shitlist these days Jun 29 11:53:14 gerwinin: I replied to your mail... with some reasons. Jun 29 11:53:16 I do not like current hh.org policies Jun 29 11:53:55 they block http links from ltg to hh.org... at least to scap Jun 29 11:54:12 registering trademarks which they do not own, forcing developers to obtain license for own code... Jun 29 11:59:48 one shared booth + half a dozen baseball bats, all solved Jun 29 12:00:18 get george and Florian in booth + popcorn for rest of people? Jun 29 12:00:54 :-) Jun 29 12:04:12 florian is a scary looking man Jun 29 12:05:09 XorA: Nils is the scary looking one ;) Jun 29 12:05:35 florian: dont be offended but you remind me of The Master from Dr Who Jun 29 12:05:45 heh Jun 29 12:05:52 or a better excuse: Jun 29 12:06:19 There is another Florian - one of the MIPS hackers who looks way more scary than I do. Jun 29 12:06:23 florian: do not search for excuses - fight! Jun 29 12:08:18 :-) Jun 29 12:09:15 ibot: botmail for steliosk: in few weeks we can get few magicbox boards - they are waiting for production batch. Jun 29 12:10:07 XorA: http://silicon-verl.de/home/flo/images/themed/flo/img_1223_half.jpg Jun 29 12:10:12 I found a good one ;) Jun 29 12:10:37 thats a swamp donkey if ever I saw one Jun 29 12:11:27 florian: your brother? Jun 29 12:11:35 ;D Jun 29 12:12:05 hrw: no, the MIPS-Florian :-) I met him several times - very nice guy... Jun 29 13:16:00 * * OE Bug 2578 has been created by shrivastava(AT)tesbv.com Jun 29 13:16:02 * * compilation problem in .bb file. Jun 29 13:16:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2578 Jun 29 13:30:00 * * OE Bug 2577 has been RESOLVED (FIXED) by dp(AT)xora.org.uk Jun 29 13:30:02 * *  Added iwmmxt support to mplayer for akita Jun 29 13:30:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2577 Jun 29 13:30:13 * * OE Bug 2576 has been RESOLVED (FIXED) by dp(AT)xora.org.uk Jun 29 13:30:15 * *  Add iwmmxt support to mplayer for palmld Jun 29 13:30:16 * * http://bugs.openembedded.org/show_bug.cgi?id=2576 Jun 29 13:39:13 03xora 07org.oe.dev * r96925818... 10/ (1 packages/mplayer/mplayer_0.0+1.0rc1.bb): mplayer_0.0+1.0rc1.bb : akita and palm can use pxa enhancements Jun 29 13:52:58 movq [r2,#0],%mm6 what cpu does that look like? Jun 29 13:54:41 mmx? Jun 29 13:54:57 yeah it is, thanks for confirmation Jun 29 13:55:10 just googled :) Jun 29 14:04:19 Hi all! Jun 29 14:04:29 Where are all from? Jun 29 14:04:45 Netherlands Jun 29 14:05:00 I am from Russia Jun 29 14:33:02 03tmbinc 07org.oe.dreambox * r0fee1865... 10/ (1 packages/unionfs/unionfs.inc): unionfs: build fix Jun 29 15:20:09 Why does perl-native-5.8.7 always fail? Jun 29 15:20:56 builds for me Jun 29 15:21:59 I thought we are on 5.8.8 Jun 29 15:23:57 Keizer: perl (not perl-native) fails if /bin/sh is linked to dash, but I've not had problems with perl-native. Jun 29 15:24:24 cbrake: iirc it was fixed tooo Jun 29 15:29:18 cbrake: Hey, you're the person who told me about how to make the init file right? Jun 29 15:29:52 hrw: good Jun 29 15:30:03 Keizer: yes, I think that was me :-) Jun 29 15:30:29 cbrake: Does the mtn from OE have the ability to produce that image? Jun 29 15:31:40 Keizer: yes, OE can produce a cpio.gz image. mtn is must the SCM -- does not really have anything to do with OE. Jun 29 15:33:20 Ok Jun 29 15:34:14 I'm about to try again Jun 29 15:45:07 cu all Jun 29 16:16:47 I thought adding -static to the use flags meant that it wouldn't need anything to install, yet the control file lists uclibc and libgcc1 Jun 29 16:16:50 WTF? Jun 29 16:17:13 use flags? Jun 29 16:17:24 -static-libgcc? Jun 29 16:17:29 I didn't know that bitbake had useflags? Jun 29 16:17:39 I need to add -static and -static-libgcc? Jun 29 16:17:50 and then I guess I'd need -static-uclibc too? Jun 29 16:18:00 hehe Jun 29 16:18:02 man gcc Jun 29 16:18:05 and search for static Jun 29 16:18:21 rgr Jun 29 16:18:34 gcc does not care for uclibc or glibc when linking Jun 29 16:23:25 then why does my program have a uclibc dep? Jun 29 16:24:01 Zero_Chaos: tell me. I would check that I have a static uclibc Jun 29 16:24:21 I added -static-libgcc to my ldflags and I still have libgcc dep as well Jun 29 16:24:29 zecke: and I check this how? Jun 29 16:24:45 Zero_Chaos: using GNU ls? Jun 29 16:25:11 Zero_Chaos: did you rebuild your package? Jun 29 16:25:15 yes Jun 29 16:25:32 Zero_Chaos: are you sure you -static and -static-libgcc got passed to your build? Jun 29 16:26:36 um, shit Jun 29 16:26:38 no Jun 29 16:26:40 :-( Jun 29 16:26:45 I wonder why Jun 29 16:27:10 a guess: the Makefile sets LDFLAGS as well? Jun 29 16:27:30 I should be overriding that. Jun 29 16:27:57 LDFLAGS="-static -static-libgcc ${LDFLAGS}" Jun 29 16:28:41 well, the 'Makefile' or whatever this software is using for the build Jun 29 16:31:24 I got it to pass now, and now it ignores the fact I have -lpthread in there and just fails looking for pthread... sweet Jun 29 16:32:09 joys of overriding Makefile variables Jun 29 16:32:33 I switched it to an ldflag and seems to be working better now Jun 29 16:33:19 and it drops it again Jun 29 16:33:20 sweet Jun 29 16:33:34 hehe Jun 29 16:33:36 is there an example of a static reciepe somewhere? Jun 29 16:34:21 Zero_Chaos: dfu-util-native but listen! Jun 29 16:34:33 Zero_Chaos: What is the build-system dropping stuff? Is it only a Makefile? Jun 29 16:34:54 zecke: I'm using OE to build (obviously) and yes, it's only a Makefile Jun 29 16:35:46 Zero_Chaos: for a bad solution you can check netbase, for a better solution see how you can pass variables to make Jun 29 16:36:06 zecke: I thought I knew how to pass variables to make.... Jun 29 16:36:12 Zero_Chaos: normally we pass them as environment, sometimes you have Makefiles that set ANYFLAG directly Jun 29 16:38:10 I have decieded I don't care right now. I apparently don't know how to do this right, so I will work on other things until I learn better Jun 29 16:39:40 hey, you are close, don't give up now Jun 29 16:40:32 close? I set CFLAGS_append and the make line has -lpthread and fails because it can't find pthread, I se LDFLAGS_append and it just plain ignores me Jun 29 16:40:38 * Zero_Chaos doesn't feel very close Jun 29 16:41:22 Zero_Chaos, carry on :) Jun 29 16:41:44 * dcordes trying to motivate 0 for selfish reasons Jun 29 16:42:06 dcordes: you would be wise to tell me to move on, I am working on something other than what you want me working on. Jun 29 16:42:33 Zero_Chaos, move on! :P Jun 29 16:44:29 Zero_Chaos: What are you working at? Jun 29 16:47:25 thesing: I was working on statically building stuff, now I'm back to aircrack-ng-1.0-dev Jun 29 16:50:38 If you want to build static stuff it might be easier to patch the makefile. Jun 29 16:50:57 thesing: yeah, that's not happening right now Jun 29 17:09:31 dcordes: lemme see your .bb again? Jun 29 17:11:39 http://www.pastebin.ca/595747 Jun 29 17:12:11 add src/ before all the files you want to install Jun 29 17:12:16 :-) Jun 29 17:13:23 Zero_Chaos, like "install -m 0755 src/airodump-ng ${D}/${sbindir" ? Jun 29 17:13:41 yeah Jun 29 17:14:35 Zero_Chaos, that made it work for you? Jun 29 17:15:02 dcordes: I didn't try, I am having lunch, just thought of it though Jun 29 17:15:23 Zero_Chaos, I already though about a different dir structure in the 1.0-deb branch being the reason. But than I compared it do the 0.9 and 0.9.1 which were the same Jun 29 17:15:37 but I'll give it a try wait Jun 29 17:20:11 good morning all Jun 29 17:21:57 koen: moin sir Jun 29 17:22:08 koen: hey Jun 29 17:22:11 koen: by chance, did you get my message about viewmtn? Jun 29 17:22:53 koen: I'm installing my Dual Xeon, I'm going to build uclibc+arm+EABI regulary Jun 29 17:27:01 * * OE Bug 2578 has been RESOLVED (INVALID) by Jun 29 17:27:03 * *  compilation problem in .bb file. Jun 29 17:27:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2578 Jun 29 17:29:54 Zero_Chaos: no, and I don't touch the OE website Jun 29 17:32:36 koen: oh, the 404 message was signed by you so I thought you might know who to talk to Jun 29 17:36:09 Zero_Chaos, http://www.pastebin.ca/595781 with src/ appended to filenames Jun 29 17:38:00 dcordes: my QA failed miserably as well, fixing it now Jun 29 17:38:25 Zero_Chaos, sorry but what's QA? Jun 29 17:39:35 Quality Assurance Jun 29 17:39:41 all the errors at the bottom Jun 29 17:41:46 i see Jun 29 17:44:30 Zero_Chaos: ERROR: Architecture did not match (40 to 3) means that the files are x86 binaries Jun 29 17:45:17 also known as "morons that use 'gcc' instead of $(CC) in Makefiles" Jun 29 17:46:48 koen: I am aware of that fact, what surprises me is that my CC="${CC}" doesn't override that Jun 29 17:47:18 nothing helps if the Makefile has 'CC = gcc' hardcoded Jun 29 17:47:23 (maybe make -e) Jun 29 17:48:04 koen: well, if you look at packages/aircrack-ng/aircrack-ng_0.9.bb (or whatever version is on mtn right now) you will notice that it does work and CC=gcc is hardcoded into the makefile Jun 29 17:48:36 Zero_Chaos, yep, aircrack-ng_0.9.bb Jun 29 17:49:11 but that also works with cp aircrack-ng_0.9.bb aircrack-ng_0.9.1.bb Jun 29 17:49:17 already tried Jun 29 17:49:36 koen: so aside from the several other problems, i wonder why aircrack-ng_0.9.bb works to gen arm code but that the 1.0-dev branch doesn't work... Jun 29 17:49:42 dcordes: I know ;-) Jun 29 17:53:43 koen: can I get some words of wisdom? Why does my .bb work for aircrack-ng_0.9 and not for the 1.0-dev? why can I override CC in one and not the other? Jun 29 17:54:26 "crappy makefiles" Jun 29 17:54:26 Zero_Chaos: Difference in do_configure? Jun 29 17:54:39 NAbyss_: there is no configure Jun 29 17:55:06 Zero_Chaos: As koen said then, makefiles.. whatever's generating the makefile isn't capturing $CC Jun 29 17:55:12 there's a strong correlation between '133tn355' of software and the insanity of the buildsystem Jun 29 17:55:15 koen: this is obvious, but two makefiles, very similar, one gets overridden by my code and one doesn't Jun 29 17:55:30 NAbyss_: nothing is generating the makefile, the makefile just is Jun 29 17:55:39 (just look at the THC stuff and weep) Jun 29 17:56:05 Zero_Chaos: diff the makefiles, pay attention to any CC= lines? Jun 29 17:56:22 koen: while I am very entertained as we laugh together about shitty makefiles, this doesn't answer why one works and the other doesn't Jun 29 17:56:33 indeed it doesn't Jun 29 17:56:40 NAbyss_: one is CC=gcc the other is CC=$prefix/gcc Jun 29 17:56:56 NAbyss_: the first one gets overridden, the later does not Jun 29 17:57:11 * Zero_Chaos wonders how koen keeps his job in support. Jun 29 17:57:29 Zero_Chaos: Both invoke make -e? Jun 29 17:57:39 because police officers need strong language before getting the point :) Jun 29 17:57:52 Zero_Chaos: just remove all lines with CC Jun 29 17:57:56 koen: well then, get to the point, I've had enough strong language Jun 29 17:58:13 sed -i -e /^CC/d Makefile Jun 29 17:58:17 koen: I could fix it in aircrack-ng's svn, but the point is why OE treats it differently Jun 29 17:58:49 Zero_Chaos: Check the bb recipes.. one may be invoking make -e, the other sans -e Jun 29 17:58:59 Anyways.. it's 4am.. bed. Jun 29 17:59:04 NAbyss_: they both invoke -e Jun 29 17:59:31 Zero_Chaos: Use make's debug flags then to work out where CC disappears to. Jun 29 17:59:48 * Zero_Chaos doesn't know that much about make... Jun 29 17:59:56 c'mere manual! Jun 29 18:00:25 Zero_Chaos: I'd also attempt to invoke make -e manually with the CC override.. test whether the makefile or recipe is broken Jun 29 18:00:35 03:59 < NAbyss_> Zero_Chaos: Use make's debug flags then to work out where CC disappears to. Jun 29 18:01:01 NAbyss: I was just reading the make manual to learn how to use the debug flags ;-) Jun 29 18:01:34 Zero_Chaos: Good luck :) Jun 29 18:05:54 CC doesn't appear to be properly setting for some reason. how can I confirm that? Jun 29 18:15:02 Wow, it is compiling using $BUILD_CC instead of $CC Jun 29 18:15:05 how wonderful Jun 29 18:16:47 Zero_Chaos, working? Jun 29 18:17:24 dcordes: niet, still working on it Jun 29 18:19:14 wow, uclibc rootfs is throwing alignment traps all over the place! Jun 29 18:19:35 cbrake: EABI one? Jun 29 18:19:41 zecke: yes Jun 29 18:19:48 cbrake: lorn reported that as well, OABI works just fine regarding to woglinde Jun 29 18:20:00 I can't make qemu to boot with my image :) Jun 29 18:20:15 wow GPLv3 is out Jun 29 18:20:34 I also tested uclibc console image on collie and it works fine. Jun 29 18:20:56 thesing: makes sense as collie is probably not eabi? Jun 29 18:21:18 cbrake: right. Jun 29 18:21:41 ~seen anderson Jun 29 18:21:43 anderson was last seen on IRC in channel #htc-linux, 216d 23h 48m 32s ago, saying: 'ok'. Jun 29 18:21:53 ~seen andreson Jun 29 18:21:55 i haven't seen 'andreson', zecke Jun 29 18:22:43 cbrake: what is trapping? busybox? Jun 29 18:23:41 zecke: seems like it Jun 29 18:23:54 http://bugs.uclibc.org/view.php?id=1357 Jun 29 18:25:37 zecke: http://pastebin.ca/595904 Jun 29 18:26:32 cbrake: it is uclibc itself Jun 29 18:26:48 I don't think it is a coincidence that there are only two versions of the ProgramCounter? Jun 29 18:27:12 and psplash-write is not part of busybox Jun 29 18:27:40 for lorn it did boot up to getty, would be interesting to get a core dump, or use addr2line on uclibc Jun 29 18:27:56 zecke: good points Jun 29 18:30:15 zecke: I'll try to get a addr2line with this image later today. Jun 29 18:32:41 zecke: http://bugs.uclibc.org/view.php?id=1357 needs OE's unbreak armv4t patch to gcc Jun 29 18:33:09 hehe Jun 29 18:33:12 koen: write it Jun 29 18:33:28 I want to paste http://www.flickr.com/photos/quimgil/478715643/in/set-72157600160940845/ in there :) Jun 29 18:36:59 lol is fluendo without workers? Jun 29 18:37:58 hmm. i still dont get what 'No buildable providers.. " means Jun 29 18:38:35 in a proper Makefile, what should CC= ? Jun 29 18:38:55 ljp: hmm, what part is unclear? Jun 29 18:39:00 * * OE Bug 2579 has been created by cliff.brake(AT)gmail.com Jun 29 18:39:02 * * alignment faults in uclibc + EABI Jun 29 18:39:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2579 Jun 29 18:40:08 ljp: somewhere in the dependency tree you have an edge to "qtopia-opensource" Jun 29 18:40:30 ljp: now bitbake tries to make sure the thing you point to is getting built Jun 29 18:41:18 ljp: e.g. qtopia-opensource_4.3.0.bb PROVIDES = "qtopia-opensource" by default, but zeckes-openmoko-qtopia-emulator could PROVIDES = "qtopia-opensource" as well Jun 29 18:41:30 ljp: and then BitBake would select one provider from the set of providers Jun 29 18:42:15 ljp: not finding a buildable one can have multiple issues: a) you mis-spelled something b) bitbake knows about qtopia-opensource but a dependency is not buildable c) anything else Jun 29 18:42:35 zecke: I asked one of the blackfin hackers about toolchains and he said "we just use the ADI binaries these days" Jun 29 18:42:50 in a proper Makefile, what should CC= ? Jun 29 18:43:11 a proper makefile has CC as conditional assignment Jun 29 18:43:16 Zero_Chaos: good question. Is there any reason to do an assignment at all? Jun 29 18:43:31 indeed, just leave it out Jun 29 18:43:42 koen: might there be an example I should be looking at? Jun 29 18:43:59 so just drop it? Jun 29 18:44:07 yes Jun 29 18:45:12 ljp: feel free to give me a login and share a screen with me to look into your issue Jun 29 18:45:51 I think I figured out the difference. aircrack-ng-0.9.1 builds with commands like $(CC) and 1.0-dev build like $(MAKE) -C Jun 29 18:46:16 OE builds its own make, right? is it ${MAKE} ? Jun 29 18:46:52 Zero_Chaos: if your question is, should I use oe_runmake to execute make the answer is yes Jun 29 18:47:09 zecke: that was not my question Jun 29 18:47:38 * koen ponders about the usefullness of defining make inside a makefile Jun 29 18:47:57 koen: I agree with you, it's more than slightly rediculous Jun 29 18:48:00 koen: a Makefile that goes into other subdirectories, as in recursive make Jun 29 18:48:11 koen: but that is considered harmful anyway Jun 29 18:48:15 zecke: recurisve make stinks Jun 29 18:48:21 when it calls make, it seems to be getting my host make and not my OE make.... anyone? Jun 29 18:48:29 my spelling stinks as well Jun 29 18:48:50 koen: I'm human I can correct it (turing test passed) Jun 29 18:49:35 how do I make it call OE's make instead of my host make? Jun 29 18:51:33 hmm, OE doesn't seem to have a make... Jun 29 18:53:56 what is packages/make then? Jun 29 18:55:38 koen: okay, well it wasn't in the staging area Jun 29 18:55:42 that I saw Jun 29 18:56:38 automake maybe... Jun 29 19:37:42 Hi all - do you know where is defined, which kernel version is build if I do "bitbake virtual/kernel"? - Unfortunately I can not find it (but would like to change it) ... Jun 29 19:39:21 pgfeller: do a dry run? Jun 29 19:39:54 zecke: what is a "dry run" (sorry for my ignorance) Jun 29 19:40:06 bitbake --help Jun 29 19:41:44 zecke: ah - now I understand :-) ... this I've already done, and I did not like the version - so I tried PREFERED_VERSION in local.conf; but now I have things like this: NOTE: preferred version 2.4.18 of linux-rp not available (for item virtual/kernel) Jun 29 19:43:06 PREFERRED_PROVIDER ? Jun 29 19:44:37 no, only Jun 29 19:44:37 PREFERRED_VERSION_virtual/kernel = "2.4.18" Jun 29 19:44:37 PREFERRED_VERSION_linux = "2.4.18" Jun 29 19:44:37 PREFERRED_VERSION_linux-rp = "2.4.18" Jun 29 19:48:02 RP: sir... about your kernel and the whole not working thing.... Jun 29 19:48:24 zecke: thank you for the hint ... I'll look in packages/linux for a provider & try again Jun 29 19:48:33 RP: it would be really keen if you would commit sharpsl-rc-r1.patch to fix it Jun 29 20:01:13 ~pastebin Jun 29 20:01:14 Zero_Chaos: check the commit logs and bugtracker and you'll notice RP doesn't have anything to do with the sharp remote patch Jun 29 20:01:14 methinks pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/, or http://pastebin.com is usually painfully too slow and unresponsive to use, use one of the other pastebin sites, or dpaste.com is a very nice pastebin as well Jun 29 20:01:41 Zero_Chaos: I suspect you need to pester justinp to add the missing patch, or put the one from the bugtracker in your OE tree Jun 29 20:02:00 koen: linux-rp_2.6.21, I think that would qualify him to fix it. but I can bother JustinP if it was his mistake Jun 29 20:02:33 Zero_Chaos: why don't you check who added that line instead of blindly assuming something? Jun 29 20:02:45 we don't need another goxboxlive... Jun 29 20:03:12 koen: dude, linux-RP, it is RP's kernel. I would be mad if someone messed up my kernel... Jun 29 20:03:13 thx koen you are so nice Jun 29 20:03:39 [22:02] koen: Zero_Chaos: why don't you check who added that line instead of blindly assuming something? Jun 29 20:03:52 koen: yeah, thanks, I get it Jun 29 20:04:09 wow, only after the third time! Jun 29 20:04:15 JustinP: mind commiting your sharpsl-rc-r1.patch instead of just trying to apply it? Jun 29 20:04:20 koen: you missed the point, it's okay Jun 29 20:04:30 tried it again - with the following result: http://paste.debian.net/31673 Jun 29 20:04:49 same if I use bitbake 1.8.4 Jun 29 20:05:05 pgfeller: 1) bitbake 1.6.x is obsolete and won't be supported after tomorrow Jun 29 20:05:27 pgfeller: 2) why do you think there's a 2.4.18 version of linux-rp? Jun 29 20:07:09 koen: probably I completely missunderstood something ... I selected the sharprom distro to build hostap for my collie (need wpa) & saw, that linux-rp 2.6.21 was fetched - so I though I have to force this to 2.4.18 ... - I assume that Jun 29 20:07:09 # Jun 29 20:07:09 PREFERRED_PROVIDERS += "collie-kernel" Jun 29 20:07:33 & PREFERRED_VERSION_linux = "2.4.18" is what I should do? Jun 29 20:07:39 I am searching for cheap virtual hosting(I don't really care about the speed) any golden tips? Jun 29 20:08:09 it's to run mtn or other things that used to run on my home server Jun 29 20:08:14 PREFERRED_PROVIDERS += "collie-kernel" does exactly nothing in the best case and messes up the PREFERRED_PROVIDER glob in the worst case Jun 29 20:09:38 koen: I tried the OE user manual - but unfortunately did not find explanations how to define the kernel to build modules for (as said, I do not plan to do a distro - just hostap-modules :-) ) Jun 29 20:10:09 koen: any hints where I can find info/docu on this? Jun 29 20:13:54 pgfeller: you need openzaurus-sa as kernel. Jun 29 20:14:50 thesing: thanks - so "PREFERRED_PROVIDER_virtual/kernel = "openzaurus-sa" & PREFERRED_VERSION_linux = "2.4.18" - right? Jun 29 20:15:06 look at conf/machine/collie.conf Jun 29 20:15:22 whats the difference between .summary.jffs2 and .jffs2 files? Jun 29 20:15:35 and set MACHINE_KERNEL_VERSION so 2.4 in your local.conf Jun 29 20:15:46 s/so/to/ Jun 29 20:16:52 thesing: thanks - will give it a try :-) Jun 29 20:17:42 pgfeller: and don't set any other PREFERRED_PROVIDERS for kernel in your local.conf Jun 29 20:18:39 thesing: yep ... will check for that. Jun 29 20:18:49 pgfeller: did you read http://www.openembedded.org/wiki/ZaurusKernels ? If not you should to solve a future problem ;) Jun 29 20:19:22 ljp: the summary one mounts faster Jun 29 20:19:34 koen: at the cost of what? Jun 29 20:19:37 thesing: I did, gcc 2.95 installed according to those directions - is there a wiki/faq I could update with the information you gave me? Jun 29 20:19:44 Zero_Chaos: size Jun 29 20:20:07 koen: seems a fair trade to me, espcially with akita/spitz Jun 29 20:20:33 spitz doesn't use NAND by default Jun 29 20:20:53 technically, whats the difference? Jun 29 20:21:11 ljp: the summary ones store a bit of the inode tree Jun 29 20:21:26 ljp: jffs2 by defaults doesn't store it and rebuilds it at mount time Jun 29 20:21:34 hence the excessive ram usage and mount times Jun 29 20:21:53 hmm. ok. thanks Jun 29 20:21:55 logfs ftw Jun 29 20:22:24 koen: why don't we build logfs by default then? or should I add that to my packages? Jun 29 20:22:29 pgfeller: I would add it here: http://www.openembedded.org/usefulconfigurations but I don't know if thats realy the right place. Jun 29 20:22:38 because logfs isn't ready yet Jun 29 20:23:31 http://logfs.org/logfs/ for more info Jun 29 20:23:33 koen :( Jun 29 20:23:37 ~lart logfs for using BUG_ON Jun 29 20:23:37 * ibot beats the living hamstercrap out of logfs for using BUG_ON Jun 29 20:23:48 whats bug_on? Jun 29 20:24:09 thesing: I'll put it there, once I tested it through Jun 29 20:25:52 woglinde: like an ASSERT Jun 29 20:27:01 woglinde: when are you going to arrive? Tieckstrasse Nr. 2., ring the bell with the label "Dr. hendrik" Jun 29 20:27:38 thesing: unfortunately package modutils-cross_2.4.27.bb failes during do_compile ... (this is not yet in the bugtracker for sharpcompatible afaik) Jun 29 20:28:45 pgfeller: what do you want do build eventually? Jun 29 20:29:17 thesing: hostap-modules (need WPA on my collie) Jun 29 20:29:58 pgfeller: I'm not sure if they build with 2.4 kernel. Jun 29 20:30:52 pgfeller: you could try another version of modutils (with PREFERRED_VERSION) or fix the build (better!) ;) Jun 29 20:31:05 thesing: oz 3.5.4 has them for crapix; but the .ipk format does not work with the sharp rom Jun 29 20:32:43 thesing: :-) unfortunately I'll have to go for PREFERRED_VERSION (will try to find out what hrw used for OZ) ...; from the build log I think this is caused by gcc 2.95 - and that's not my leage yet to fix Jun 29 20:33:13 pgfeller: then just repack the ipk. Jun 29 20:33:47 pgfeller: unpack it with 'ar x' and repack it with 'tar czf' Jun 29 20:34:22 thesing: if I do not succeed to build them this night, then I'll do this ... - but like this I'll learn again a little more about OE (little by little :-) Jun 29 20:37:31 thesing: if the older version work (copied configuration from oz 3.5.4), do you think I should update sharprom-compatible distro conf & submit a patch to bugzilla? Jun 29 20:38:58 doesn't hurt to let other people know of the problem and the solution, does it? ;) Jun 29 20:39:45 thesing: yep -> http://www.codeomega.info/index.php?option=com_content&task=view&id=35 ... Jun 29 20:42:33 pgfeller: you don't need the PREFERRED_PROVIDER_virtual/kernel line. MACHINE_KERNEL_VERSION does that for you. Jun 29 20:42:57 thesing: ok Jun 29 20:59:01 lrg: maybe birmingham does have shops where we can buy dark rum :) Jun 29 20:59:37 koen: it does indeed ;) Jun 29 21:00:01 bbl Jun 29 21:01:10 03xora 07org.oe.dev * r6222123e... 10/ (1 packages/aircrack/aircrack-ng_svn.bb): aircrack-ng_svn.bb : add svn of aircrack that will become 1.0 Jun 29 21:01:16 03xora 07org.oe.dev * rf50ef495... 10/ (3 files in 2 dirs): aircrack-ng_0.9.1.bb : version bump + fixups of .bb file Jun 29 21:01:20 03xora 07org.oe.dev * rd0991cf1... 10/ (1 packages/aircrack/files/makefile-ldflags.patch): aircrack_svn.bb : forgotten patch Jun 29 21:06:12 re Jun 29 21:06:21 pgfeller: modutils should just builds Jun 29 21:06:44 and hostap-modules are only for 2.4 kernels. 2.6 kernels use in-kernel hostap modules Jun 29 21:06:46 hrw: was my fault -> bison was missing on the host system Jun 29 21:06:50 ;D Jun 29 21:07:00 pgfeller: so modutils req bison? Jun 29 21:07:01 hrw: he wants to build for sharprom. Jun 29 21:07:44 pgfeller: for shaprom? remember to replace 2.4 kernel with OE generated 2.4 kernel then - otherwise modules will rather not fit Jun 29 21:08:31 hrw: there is no way to build a module which will work with the sharp kernel (I'm a little afraid to replace a kernel I've to confess) Jun 29 21:08:50 hrw: why? the versions are the same. Jun 29 21:09:48 pgfeller: there is no warranty that it will works Jun 29 21:10:03 thesing: our kernel is heavily patched Jun 29 21:10:23 thesing: OE 2.4 crapix kernel has WE18. sharprom kernel has WE12 if not older Jun 29 21:10:39 hrw: I see ... I'll try (I just need hostap) - or then try to replace the kernel Jun 29 21:10:56 I had 2.4-crapix with WE19 even Jun 29 21:11:09 ah ok. thanks. Jun 29 21:11:12 wpa req we18 iirc Jun 29 21:11:49 hrw: do you know which var I did not set if "openzaurus-sa_2.4.18-rmk7-pxa3-embedix20030509.bb" fails with Jun 29 21:11:50 NOTE: make EMBEDIXRELEASE=-3.1 compatible oldconfig Jun 29 21:11:50 make: *** No rule to make target `compatible'. Jun 29 21:11:51 yep. we add WE13-18 into kernel Jun 29 21:12:03 pgfeller: no idea Jun 29 21:12:43 pgfeller: I have no idea from where you got that "make" line Jun 29 21:13:03 2.4-crapix.. argh Jun 29 21:13:07 from ... pxa3-embedix-r23/temp/log.do_configure.23871 Jun 29 21:14:12 pgfeller: you have DISTRO_VERSION = "3.1 compatible" :D Jun 29 21:14:19 pgfeller: remove compatible from it Jun 29 21:14:51 hrw: will do; thanks :-) Jun 29 21:16:29 hrw: it is in org.openembedded.dev/conf/distro/sharprom-compatible.conf even; I'll submit a patch once it works Jun 29 21:16:45 ok Jun 29 21:21:02 03xora 07org.oe.dev * rd291070b... 10/ (1 packages/aircrack/aircrack-ng_0.7.bb): aircrack-ng_0.7.bb : dropped in place of the later alternatives Jun 29 21:21:06 03xora 07org.oe.dev * r9f18c885... 10/ (3 files in 2 dirs): Jun 29 21:21:06 aircrack-ng_1.0-svn.bb : renamed to differentiate from aircrack-ng HEAD also Jun 29 21:21:06 DEFAULT_PREFERENCE = "-1" Jun 29 21:22:03 \me ... it worked :-) Jun 29 21:22:18 * pgfeller ? Jun 29 21:23:27 hrw & thesing: thanks again for your help ... I'l try the .ipks tomorrow (and as promised update wiki & bugtracker) Jun 29 21:24:29 can't mtn update: "mtn: misuse: path _MTN/tmp/1 already exists" Jun 29 21:25:04 dcordes: rm -rf _MTN/tmp;mtn update Jun 29 21:26:38 working again Jun 29 21:39:53 bye Jun 29 21:54:49 03koen 07org.oe.dev * rb29b99ce... 10/ (1 packages/tetex/tetex_3.0.bb): tetex: adjust packaging a bit Jun 29 21:57:04 woglinde_: hi Jun 29 21:57:37 hm crappy connection Jun 29 22:15:21 hi Jun 29 22:15:26 he coredump Jun 29 22:33:26 woglinde: lol, so what did change in your build? Jun 29 22:38:15 dont know Jun 29 22:38:29 uclibc I guess Jun 29 22:40:44 woglinde_: well, do you have the uclibc rev from the previous build around? I thought we use a released version of uclibc? Jun 29 22:40:54 no Jun 29 22:40:54 woglinde_: cbrake filed a bug report in your bt Jun 29 22:40:58 U delete all Jun 29 22:41:00 I Jun 29 22:41:15 maybay its kernel related Jun 29 23:41:01 * * OE Bug 2573 has been RESOLVED (FIXED) by Jun 29 23:41:03 * *  Fix for esmart QA errors Jun 29 23:41:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2573 Jun 29 23:43:03 * * OE Bug 2555 has been REOPENED by r.farina(AT)adelphia.net Jun 29 23:43:05 * *  patch to make kernel driver for Sharp CE-RH2 working Jun 29 23:43:07 * * http://bugs.openembedded.org/show_bug.cgi?id=2555 Jun 30 00:30:04 03Laibsch 07org.oe.dev * r57fa0ccb... 10/ (4 files in 2 dirs): gtkmm: update to 2.10.10. Closes 2544. Jun 30 00:30:16 03Laibsch 07org.oe.dev * rbb1bb675... 10/ (1 packages/efl1/esmart_0.9.0.009.bb): esmart: fix packaging. Closes 2573. Jun 30 01:30:23 03pfalcon 07org.oe.dev * rd876b6a9... 10/ (5 files in 3 dirs): (log message trimmed) Jun 30 01:30:23 udev: Stop booting semi-configured system in case of random miniscule errors. Jun 30 01:30:23 * It makes *completely* no sense to just fail the udev script on first error Jun 30 01:30:23 and continue boot further. Either it should try to ignore them, or halt system Jun 30 01:30:23 boot. Otherwise, system in weird, semi-working state will emerge, without anu Jun 30 01:30:24 notices of such state. Jun 30 01:30:28 * This script for example fails on mount --bind and mount --move, which busybox Jun 30 01:30:30 03pfalcon 07org.oe.dev * rd247b655... 10/ (4 files in 4 dirs): (log message trimmed) Jun 30 01:30:32 udev 092, 097, 100: Stop booting semi-configured system in case of random miniscule errors. Jun 30 01:30:33 * It makes *completely* no sense to just fail the udev script on first error Jun 30 01:30:36 and continue boot further. Either it should try to ignore them, or halt system Jun 30 01:30:38 boot. Otherwise, system in weird, semi-working state will emerge, without anu Jun 30 01:30:39 notices of such state. Jun 30 01:30:42 * This script for example fails on mount --bind and mount --move, which busybox Jun 30 02:29:47 well if anyone is around, how can I make my aircrack-ng preferred version 0.9.99+svn${SRCDATE} ? it doesn't seem to like that as I put it? **** ENDING LOGGING AT Sat Jun 30 02:59:57 2007