**** BEGIN LOGGING AT Sun Jul 01 02:59:56 2007 Jul 01 04:29:55 03mwester 07org.oe.dev * r8fb2c300... 10/ (3 files in 3 dirs): Jul 01 04:29:55 SlugOS: bump kernel up to svn 888 to pick up kexec patches, add kexec-tools Jul 01 04:29:55 to the base rootfs. Jul 01 04:30:01 03mwester 07org.oe.dev * r593f40ff... 10/ (1 packages/linux/ixp4xx-kernel_2.6.21.5.bb): SlugOS kernel: bump svn version to bring in fix for broken kexec patch Jul 01 07:34:56 hello Jul 01 07:43:39 e Jul 01 08:33:59 03freyther 07org.oe.dev * r1e74798b... 10/ (1 packages/orinoco/spectrum-fw.bb): packages/orinoco/spectrum-fw.bb: Building depends on unzip, make sure it is built Jul 01 08:35:18 * koen notices it's the 1st of july Jul 01 08:35:29 * koen bumps BB_MIN_VERSION Jul 01 08:36:33 03koen 07org.oe.dev * rb38c186d... 10/ (1 conf/sanity.conf): sanity.conf: require bitbake 1.8.5, the two month wait is over Jul 01 14:40:17 hey zecke Jul 01 14:40:25 zecke: thanks for tagging 1.8.6 Jul 01 14:40:40 * zecke_ waits for the berlios comment Jul 01 14:41:04 I'm already over that Jul 01 15:55:41 what's that problem with firefox? Is it something with the source? Jul 01 16:01:06 dcordes: it doesnt work is the problem Jul 01 16:03:35 I bet it has something to do with the source, doesn't it? :) Jul 01 16:04:28 polyonymous: I guess, it could be that the source is a pile of stacked turds where one is mis aligned, no-one wants to explore :-) Jul 01 16:04:36 =) Jul 01 16:04:45 which is unfortunate Jul 01 16:05:29 That sometimes happens to sources. Jul 01 16:07:39 well, FF is actively developed Jul 01 16:09:56 The probability of misalignment increases :) Jul 01 16:12:20 hehe Jul 01 16:12:42 anyway, I still have hopes that some nice soul eventually fixes the FF segfault ?= Jul 01 16:12:43 CoreDump|home: define active, lennert patches seem to sit on bugtracker ignored for 6 months Jul 01 16:13:40 yeah, what do I know. My "FireFox" calls himself Iceweasel ;) Jul 01 16:15:06 XorA|gone, active, means that developers are scratching THEIR itches. And they seem to be itchy. Jul 01 16:15:14 Need a bath, probably. Jul 01 16:16:19 damn x86 only project though :-( Jul 01 16:17:08 it means that lennerts EABI port of debian gets ignored Jul 01 16:17:32 because he isn't officially annointed as ARM dude by wookey Jul 01 16:18:03 (which won't happen since wookey's company was hired to said port, but was too incompetent) Jul 01 16:18:13 s/to/to do/ Jul 01 16:18:47 hi koen Jul 01 16:19:18 koen: heh heh intrigue eh Jul 01 16:19:39 * XorA|gone wonders why wookey_ doesnt chime in occasionally Jul 01 16:39:05 * koen wonders why people ask stupid questions in forums when the answer is the *first* link in google for they question Jul 01 16:39:39 and then they say "I searched the forum and didn't find anything" Jul 01 16:39:59 Is a lobotomy or a whack to the head required to join a forum? Jul 01 16:40:27 nah, you can't break it if it ain't fixed. Jul 01 16:40:53 http://www.oesf.org/forums/index.php?showtopic=24299&view=getlastpost Jul 01 16:41:06 hopefully the colours and insane font sizes will get the point across Jul 01 16:41:17 I don't expect so, but one can hope Jul 01 16:43:57 heh, koen at his best :) Jul 01 16:44:54 one wonders why people expect all answer in a forum when the contact page explicitly says otherwise Jul 01 16:46:16 polyonymous, hi ... Ive been pushed to build OE/Angstrom for xscale since some people cant crunch iwmmxt :-/ Jul 01 16:46:27 koen: On form today :) Jul 01 16:46:51 polyonymous, dont you know whether qemu works on amd64 for binary locale generation yet? ? Jul 01 16:47:11 Marex, to the best of my bad knowledge, it doesn't. Jul 01 16:47:35 And you can make me a happier man by making it work :) Jul 01 16:48:12 I was thinking about adding -m32 to flags... Jul 01 16:48:43 polyonymous, bah ... ok, this will be all-nighter, but I dunno whether I can do it Jul 01 16:48:49 dont bet on it at all Jul 01 16:49:20 Marex, I don't. I don't even know what are the problems and how hard the fixes may be. Jul 01 16:49:48 But indeed, provided that host system is prepared for it, doing -m32 may do the trick. Jul 01 16:50:04 I dont like -m32 Jul 01 16:50:15 compiling with iwmmxt should be enabled on a case-by-case basis, like e.g. mplayer Jul 01 16:50:16 all you be 64bit ready Jul 01 16:50:26 koen, I know Jul 01 16:50:43 koen, I just compiled whole OE with iwmmxt to see if it is faster etc, just out of curiosity Jul 01 16:50:50 Marex, of course that's suboptimal... Jul 01 16:50:58 and, is it? :) Jul 01 16:50:59 Marex, polyonymous: I'd advise trying qemu cvs. The qemu in poky is probably worth a shot Jul 01 16:51:17 RP, thanks. I hope Marex tries it before I get to do it and will report :) Jul 01 16:51:18 koen, though, qemu-user cant still handle it, so I had to modify glibc bbclass and supply binary locales Jul 01 16:51:32 * RP doesn't have a 64 bit box but I know qemu cvs got 64 bit fixes recently Jul 01 16:51:32 RP, ok :] Jul 01 16:51:43 ah ... thanks RP Jul 01 16:52:15 RP, I already ported OE qemu patchset to qemu 0.9.0+cvs Jul 01 16:52:16 Marex: iwmmxt is faster for *some* applications Jul 01 16:53:01 Marex: Did you include nptl patches? Jul 01 16:53:24 yes Jul 01 16:53:42 is libbonoboui failing to build for anyone else? Jul 01 16:57:58 any best how many "what does this error mean?" reports we'll get about BB_MIN_VERSION next week? Jul 01 16:58:03 bets* Jul 01 16:58:31 koen, no, I dont care ;-E Jul 01 16:58:52 OT: Any one familiar with GNU R? Jul 01 16:58:57 * Marex makes space for OE on his lappy ... Jul 01 17:01:14 CoreDump|home, Im trying right now Jul 01 17:02:01 Marex: thanks Jul 01 17:02:34 CoreDump|home, it builds some stuff so it might take some time Jul 01 17:04:46 koen, can you check the new qemu personally or I will get another "send it to bugtracker"-like answer if I ask you? Jul 01 17:04:57 what I hate more is people giving bad advice: http://lists.linuxtogo.org/pipermail/angstrom-distro-users/2007-July/000407.html Jul 01 17:05:05 Marex: attach it to the bugtracker Jul 01 17:05:14 unless you want it to get forgotten and lost Jul 01 17:05:21 like I thought :-/ Jul 01 17:05:32 thats what I tried to avoid by asking you to try it ... Jul 01 17:05:45 try what? Jul 01 17:06:00 polyonymous, qemu 0.9.0+cvs Jul 01 17:06:04 for OE Jul 01 17:06:24 why should I go through the trouble of testing it if you are too lazy to attach it to the bugtracker so other people (e.g. not in IRC) can try it as well? Jul 01 17:06:37 Marex, well, you really should attach the patch against oe mtn to bugtracker so that most of us can try it. Jul 01 17:06:59 koen, because you can try and merge it faster :B Jul 01 17:07:18 no Jul 01 17:07:18 Marex, why do you think it will be faster if it's done not via BT? Jul 01 17:07:21 polyonymous, ok, I will revide it once more, update to more recent snapshot and sent to bugtracker Jul 01 17:07:22 not with core parts Jul 01 17:07:45 Marex, after all when you attach it to bugtracker you get more people who can possibly test it. Jul 01 17:08:02 especially with such core parts as qemu Jul 01 17:08:57 Even though you're right, I get this queasy feeling when you can qemu a core part and it doesn't work for me :)) Jul 01 17:09:21 * Marex seriously doubts that some things that were merged were even tested ... Jul 01 17:09:28 but thats different chapter Jul 01 17:10:10 koen, ok ... I will do it this way... Ill revide qemu, try playing with it on amd64 and then Ill put it to bugtracker, ok? Jul 01 17:10:41 * koen doesn't have an amd64 Jul 01 17:10:58 Marex, I'll try to. Jul 01 17:11:05 Marex, do YOU have amd64? Jul 01 17:11:11 * XorA|gone has amd64 Jul 01 17:11:17 polyonymous, yes, I have turion64/x2 based notebook Jul 01 17:11:23 should be enough Jul 01 17:11:26 Marex, so you have tried it? Jul 01 17:11:44 yeah, it failed on amd64 when I have compiling for arm Jul 01 17:11:51 Ill retry now Jul 01 17:12:00 Marex, why would I bother trying it then? :) Jul 01 17:12:30 polyonymous, because it might be potentially merged to OE and should be tested properly before doing so ;] Jul 01 17:13:13 Marex, but if it fails for you, what's the purpose of testing it? I mean, if it's known to fail it's known to fai Jul 01 17:13:15 fail Jul 01 17:13:56 polyonymous, Im getting pretty lost in this ... Jul 01 17:14:05 I meant testing the new version Jul 01 17:14:18 version that wont fail there , but may possibly fail somewhere else Jul 01 17:14:29 somewhere where the previous didnt Jul 01 17:14:42 Marex, ah, that's what I've asked about whether you've tried it. Jul 01 17:15:10 polyonymous, I tried qemu 0.8.2 which is now in OE, it didnt worked Jul 01 17:15:18 So, I'm willing to try version that doesn't fail for you. Jul 01 17:15:30 CoreDump|home, libbonobo blew for me ... Jul 01 17:15:50 Marex: excellent, thanks for trying =) Jul 01 17:16:04 CoreDump|home, but it is libbonobo, not bonoboui Jul 01 17:16:12 err Jul 01 17:16:56 updating libbonoboui to 2.18.0 fixed my compile problem Jul 01 17:17:04 hehe Jul 01 17:17:28 polyonymous, sure Jul 01 17:32:10 koen, is it possible to merge totally untested stuff to OE? Jul 01 17:33:00 koen, and I mean ... totally untested (Im not talking about qemu now) Jul 01 17:33:12 Marex, what for? Jul 01 17:33:24 I just want to know Jul 01 17:33:34 like machine.conf or defconfig Jul 01 17:33:52 Marex, you mean like work in progress? Jul 01 17:34:27 polyonymous, I mean like if someone prepares machine.conf for machine with architecture he never even seen Jul 01 17:34:49 that's possible Jul 01 17:34:51 like eg. I prepared defconfig and machine.conf for some sparcstation Jul 01 17:34:56 (for machine.confs) Jul 01 17:35:00 even though I never even worked with sparc Jul 01 17:35:02 zecke_: ../../../WebCore/loader/FrameLoader.h:45:29: error: ResourceRequest.h: No such file or directory Jul 01 17:35:14 Marex, the important part is that it doesn't break existing stuff, you know. I think it shouldn't. Jul 01 17:35:19 s/I/if I/ Jul 01 17:35:19 And here comes koen :) Jul 01 17:35:30 koen: do you have WebCore/platform/network/curl? Jul 01 17:35:41 polyonymous, yeah, I think it shouldnt too Jul 01 17:36:02 koen: you might need to update your WebKit.pri Jul 01 17:36:06 zecke_: hold on, tinderclient cleaned it Jul 01 17:37:07 koen: you should grab my yesterday's patches, they are not yet reviewed Jul 01 17:37:34 polyonymous, but one certain person plans doing it Jul 01 17:37:49 polyonymous, I have to stop him no matter what :E Jul 01 17:38:25 - $$BASE_DIR/WebCore/platform/network/gdk \ Jul 01 17:38:26 + $$BASE_DIR/WebCore/platform/network/curl \ Jul 01 17:39:00 Marex: btw, OE already has 2 sparc machine configs Jul 01 17:39:00 * * OE Bug 2583 has been created by gfellerpatrik(AT)gmx.net Jul 01 17:39:02 * * modutils-cross_2.4.27.bb; added missing debendency " bison-native" Jul 01 17:39:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2583 Jul 01 17:39:25 koen, that was just an example, I never seen sparc machine in my life Jul 01 17:39:34 maybe on a picture, thats all Jul 01 17:40:03 rwhitby: are you there? Jul 01 17:40:05 * mwester wishes he kept his old Sparcstation Jul 01 17:40:34 Marex: there are 6 sparc systems gathering dust in the bicycle storage here Jul 01 17:40:49 koen: this patch is not needed upstream? Jul 01 17:41:02 koen, Im not your clone nor your brother ;-] Jul 01 17:41:12 and I dont live with you either Jul 01 17:41:13 zecke_: that the diff between OE and upstream Jul 01 17:41:19 ;) Jul 01 17:42:32 zecke_: which patches to you want in OE? Jul 01 17:43:10 Marex: perhaps he's suggesting that you make an offer for purchase of one of them? Jul 01 17:43:17 #15292, #15332, #15336 and #15328 ? Jul 01 17:43:19 koen: well, I will abuse you as remote keyboard. Let us build the GdkLauncher first Jul 01 17:43:40 * RP has a Sparc IPX somewhere but hasn't seen it for a while, probably buried under a load of other old computers... Jul 01 17:43:46 mwester, oh ... Im poor and I will be even more soon Jul 01 17:44:01 koen: I'm in the middle of GNU R... Jul 01 17:44:08 R is pretty neat Jul 01 17:44:15 Im planing to buy new device, probably n311 Jul 01 17:44:27 but it needs gfortran, which OE currently lacks for most archs Jul 01 17:45:13 koen: do you know how I can change the sorting of tables (I have keys "<= 10" and "<= 5" and it thinks "<= 10" < "<= 5)... Jul 01 17:45:26 nope, sorry Jul 01 17:46:21 zecke_: what do you think of using http://www.review-board.org/ for OE? Jul 01 17:46:48 it is using django :) Jul 01 17:47:22 get a Foundation, get servers, install review-board Jul 01 17:50:24 zecke_: http://trac.webkit.org/projects/webkit/changeset/23763 is nice Jul 01 17:51:42 yes Jul 01 17:52:00 a working headerCallback will rock as well. Will hack it tomorrow night Jul 01 17:52:13 and I think I have an answer to my GdkWindow thing Jul 01 17:58:50 03koen 07org.oe.dev * r2a02be85... 10/ (3 files in 2 dirs): webkit: update webkit.pri and webkit.pro Jul 01 18:12:58 Grrr. Suspend is broken on the Zaurus in 2.6.22-rc6 :-( Jul 01 18:13:49 zecke_: armv4t/libwebkitgdk-gdklauncher_0.0+svn20070701-r1_armv4t.ipk Jul 01 18:13:59 koen: lib? Jul 01 18:14:15 koen: how did the libpcre issue go away? Jul 01 18:14:28 yes, it's named after armv4t/libwebkitgdk1_0.0+svn20070701-r1_armv4t.ipk Jul 01 18:15:45 zecke_: iirc your debug qmake magic Jul 01 18:17:22 RP: your xcalibrate stuff looks very interesting and usefull, what do you think about putting it in OE (and pushing it upstream?) Jul 01 18:22:08 koen: how does it work???? :) Jul 01 18:29:19 03koen 07org.oe.dev * ra6fcc49e... 10/ (1 packages/webkit/webkit_svn.bb): webkit: alter do_install Jul 01 18:30:00 * * OE Bug 2584 has been created by  Jul 01 18:30:02 * * qtopia-core 4.2.0 doesn't ship all files Jul 01 18:30:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2584 Jul 01 18:36:13 http://pastebin.ca/598944 - can someone help me with this? problem building qemu Jul 01 18:42:54 kevin do you need qemu? Jul 01 18:51:40 woglinde the build needs it for something or other Jul 01 18:53:52 yes but you can disbale it Jul 01 18:53:59 look at your local.conf Jul 01 18:54:04 you have two ways for it Jul 01 19:11:46 I had great fun today soldering :) http://video.google.nl/videoplay?docid=-2414345125852745646 Jul 01 19:12:35 are there adm5120 hackers around? Jul 01 19:14:27 woglinde what would I do if I disabled it. I could compile it myself and it would probably work, but I don't know where those patches are supposed to come from Jul 01 19:15:24 kevin look ar your local.conf Jul 01 19:15:40 looking.. Jul 01 19:15:59 # ASSUME_PROVIDED += "qemu-native" Jul 01 19:16:06 if I do that, I would need to provide it.. Jul 01 19:16:10 ENABLE_BINARY_LOCALE_GENERATION = "0" or ASSUME_PROVIDED += "qemu-native" Jul 01 19:17:22 if you could point me in the direction of these patches I could install qemu pretty easily.. Jul 01 19:17:51 if you dont really need locale stay with ENABLE_BINARY_LOCALE_GENERATION = "0" Jul 01 19:18:08 than qemu isnt a dependency anymore Jul 01 19:19:06 handling bitbake files.. Jul 01 19:20:22 oh, do you know how I can get a serial console on my pda (ipaq h1940), does it have a port somewhere or do I use usb somehow? Jul 01 19:20:57 sorry dont know Jul 01 19:21:50 oh, one more thing. can I change the machine/distro target without deleting and rebuilding everything or will it cache nicely Jul 01 19:23:48 hey guys, i just finished setting up the OE build environment but im having some problems, OE is up to date, when i try: bitbake nano, i get the error: ERROR NO providers of build target nano (for []) Jul 01 19:24:20 i can see the package in org.openembedded.dev/packages/nano/ Jul 01 19:24:23 hmm, now it's building gcc and family, could take a while :) Jul 01 19:24:51 so im thinking it may be the bitbake path? but ive made sure its correct Jul 01 19:25:41 i'd answer you but I'm here for help too :) Jul 01 19:26:28 creator which image you are building? Jul 01 19:30:00 woglinde : angstrom x11 Jul 01 19:30:03 akita Jul 01 19:30:21 hm Jul 01 19:30:31 could you try a bootstrap-image Jul 01 19:31:14 woglinde : shouldn't bitbake nano work just as well? Jul 01 19:31:28 hm yes Jul 01 19:32:09 woglinde : if i specify the path using -b switch it finds it but then i get errors Jul 01 19:32:37 the errors i get are: Jul 01 19:32:47 thecreator hm you can try bitbake -e and bitbake -g to look at dependencies Jul 01 19:32:58 bb.parse.ParseError: Could not inherit file classes/autotools.bbclass Jul 01 19:33:00 abs traget Jul 01 19:33:12 hm hm Jul 01 19:33:24 maybay your setup is slighty broken Jul 01 19:33:25 this seems to me like PATH errors Jul 01 19:33:30 yes Jul 01 19:33:40 i literally copy and pasted from the guide Jul 01 19:35:27 my BBPATH = /stuff/build:/stuff/org.openembedded.dev Jul 01 19:35:48 my PATH inclides /stuff/bitbake/bin Jul 01 19:36:39 any ideas? Jul 01 19:38:01 and your local.conf? Jul 01 19:38:26 sec Jul 01 19:38:56 and build has a dir conf with the file local.conf? Jul 01 19:41:43 BBFILES = "/stuff/org.openembedded.dev/packages/*/*.bb" Jul 01 19:41:43 DISTRO = "angstrom-2007.1" Jul 01 19:42:12 MACHINE = "akita" Jul 01 19:42:28 and i commented out the last line Jul 01 19:42:52 bitbake bootstrap-image works? Jul 01 19:43:01 sorry asking so many questions Jul 01 19:43:24 woglinde : i appriciate it :) Jul 01 19:43:27 set ill try Jul 01 19:44:25 woglinde : same error Jul 01 19:44:42 hm Jul 01 19:45:04 as I know nano shouldnt be build by default Jul 01 19:45:21 could you do bitbake bootstrap-image -e | less Jul 01 19:45:36 and than search in the less for nano? Jul 01 19:45:47 ill give it a try Jul 01 19:47:02 hm start bitbake parsing or bails it out early? Jul 01 19:48:06 woglinde : first thing it says is: Jul 01 19:48:17 ERROR: no files to build Jul 01 19:48:17 then Jul 01 19:48:34 ERROR: no providers of build target nano (for []) Jul 01 19:48:38 thats all Jul 01 19:48:43 :( Jul 01 19:49:46 hm did you mtn pull? Jul 01 19:49:59 than your setup is wrong Jul 01 19:50:49 woglinde : yeah pull and update, ill try the setup again Jul 01 19:50:59 woglinde : thx for your efforts Jul 01 19:51:46 maybee you misunderstand some phrase Jul 01 19:51:57 better is you take your own paths for installing Jul 01 19:52:05 than its easier to find errors Jul 01 19:52:12 true Jul 01 19:57:17 re koen Jul 01 19:59:05 hey woglinde Jul 01 20:00:54 * woglinde prepares uclibc fix for bluez-utils Jul 01 20:01:30 hi all Jul 01 20:01:34 woglinde: don't forget to send it upstream to bluez Jul 01 20:01:40 no Jul 01 20:01:44 its uclibc error Jul 01 20:01:50 not bluez Jul 01 20:02:05 hm s/error/missing Jul 01 20:17:56 * koen continues to be amazed by how dump people are on zaurus forums Jul 01 20:18:48 you tell them the exact thing to google for so the first 3 hits answer all their questions and they say that "it isn't usefull information" Jul 01 20:18:53 seems a funny day for you Jul 01 20:19:19 maybe they need a forum bot that insert the first 5 hits as plain text into each forum thread as a reply to the opening post Jul 01 20:20:12 morons Jul 01 20:24:33 * mwester observes that often koen's style of response inhibits comprehension Jul 01 20:33:11 woglinde : it seems my environment is being overwritten somewhere Jul 01 20:33:20 when i do: bitbake -e | less Jul 01 20:33:39 i can see BBPATH is /usr/share/bitbake instead of what i set it as Jul 01 20:33:44 also PATH is commented out Jul 01 20:33:53 why would this happen? Jul 01 20:34:14 i made sure to export both PATH and BBPATH before building Jul 01 20:36:16 hm than you have the bitbake from your distro? Jul 01 20:36:22 which distro do you have? Jul 01 20:37:57 slackware 11 Jul 01 20:38:39 if you distro bitbake is 1.8 than stay with it otherwise deinstall it Jul 01 20:39:26 i have version 1.8.7 Jul 01 20:39:55 distro or self installed? Jul 01 20:40:00 self Jul 01 20:40:13 than deinstall the distro Jul 01 20:40:23 its easiest I think Jul 01 20:40:39 ok ill give it as try Jul 01 20:40:54 so you think the bitbake in my OE environment is conflicting with the distro? Jul 01 20:41:54 right Jul 01 20:42:36 ok ill give it a try Jul 01 20:54:24 !google 600 USD in EUR Jul 01 20:54:24 600 U.S. dollars = 445.335115 Euros Jul 01 21:02:10 woglinde : now im getting the error : ERROR: Unable to open conf/bitbake.conf, i have pulled and checkedout again so that should have resolved the problem right? Jul 01 21:02:33 or must i remove org.open.embedded.dev and then pull again from the beginning? Jul 01 21:04:02 bloody slackware Jul 01 21:27:40 creator no Jul 01 21:27:50 only make bitbake right Jul 01 21:29:41 TheCreat0r: do you have BBBPATH set? Jul 01 21:29:45 TheCreat0r: and exported? Jul 01 21:30:07 TheCreat0r: this conf/bitbake.conf can be misleading, this can cope up when an exception is thrown somewhere when parsing conf/bitbake.conf Jul 01 21:30:26 TheCreat0r: use -D (or -DD, or -DDD) to find out what it is actually doing Jul 01 21:35:45 zecke_ : sorry, i was away, yeah, it is misleading Jul 01 21:35:54 ill try with the -D switch Jul 01 21:37:04 zecke_ : with -DDD i still only get cannot open bitbake.conf Jul 01 21:37:37 TheCreat0r: is it not trying to open any other file? Jul 01 21:37:44 TheCreat0r: in that case you should export BBPATH Jul 01 21:37:47 or fix it Jul 01 21:39:29 zecke_ : i just redid the whole thing and it worked, it seems my bitbake distro version was conflicting with the one in my OE dev environment Jul 01 21:39:33 how annoying! Jul 01 21:39:42 thanks for the support guys, appricicate it Jul 01 21:39:46 TheCreat0r: no, it can't conflict Jul 01 21:39:58 zecke_ : it seems to work after i removed it Jul 01 21:40:05 never ever as we setup the PYTHONPATH properly :) Jul 01 21:40:16 * zecke_ is a bitbake dude Jul 01 21:40:19 lol Jul 01 21:40:39 * TheCreat0r bows down to zecke Jul 01 21:40:43 zecke my fault Jul 01 21:40:43 hehe Jul 01 21:40:54 I told him it can conflicted Jul 01 21:40:56 is there a way to track the assigning of memory in a program? Jul 01 21:41:09 vlagrind? Jul 01 21:41:12 valgrind Jul 01 21:41:24 summatusmentis : yeah just print out the address, were talking a c program right? Jul 01 21:41:34 woglinde : dont worry :) Jul 01 21:41:42 same one as the before TheCan Jul 01 21:41:46 TheCreat0r** Jul 01 21:41:51 hehe Jul 01 21:41:56 sorry TheCan :-/ damn tab complete Jul 01 21:42:02 lol i through so Jul 01 21:42:05 :) no problem Jul 01 21:42:18 so you wish to know the address taken up by a string? Jul 01 21:42:26 or something along those lines? Jul 01 21:42:43 TheCreat0r: no, I want to know what function the assigning is happening it Jul 01 21:42:51 in*, I have no idea if this is even possible Jul 01 21:44:01 summatusmentis : say again? sorry i dont quite understand Jul 01 21:44:51 theres a function in the code, that is setting aside memory, and apparently not enough, causing a segfault. I want to know what is doing the setting aside Jul 01 21:45:09 ah Jul 01 21:45:29 ok so look for any malloc function calls Jul 01 21:45:39 you use malloc to set aside memory in the heap Jul 01 21:45:49 so find all the function which malloc memory Jul 01 21:46:31 right... strncpy probably has malloc within it, correct? Jul 01 21:46:31 summatusmentis : is this the same code as before? Jul 01 21:46:36 yeah Jul 01 21:46:41 still on the pastebin? Jul 01 21:46:52 should be, I haven't done anything to it Jul 01 21:47:03 got the address? Jul 01 21:47:14 hold on, I'll check logs Jul 01 21:48:14 http://rafb.net/p/ejlkOD60.html <-- try that Jul 01 21:48:41 summatusmentis : strncpy doesn't need to malloc Jul 01 21:49:14 TheCreat0r: ok, because setting up the dest already has memory set aside? Jul 01 21:50:11 all it does (internally) is create a pointer to a character then loops through the source string up to N (n being the thirs argument) and for each n we plonk that character where the char pointer is pointing to, then increment the char pointer Jul 01 21:50:29 sec ill have another look at the code Jul 01 21:50:34 ok Jul 01 21:53:06 summatusmentis : how are you running this? Jul 01 21:53:53 well, bitbake matchbox-panel, then ./mb-applet-wireless once that binary is copied to the Z Jul 01 21:57:19 TheCreat0r: sorry, I've gotta run. feel free to pm me, or whatever if you've got any suggestions at all Jul 01 21:57:38 summatusmentis : can you paste the gdb output in the paste bin? Jul 01 21:57:59 TheCreat0r: yeah, hold on, then I've gotta run Jul 01 21:58:18 kk Jul 01 21:59:35 TheCreat0r: http://rafb.net/p/RpVoIC85.html again, feel free to pm, or email summatusmentis [at] gmail [dot] com Jul 01 21:59:43 * summatusmentis leaves Jul 01 22:15:19 i take it the first time you build a package you should give it a few hours right Jul 01 22:15:20 ? Jul 01 22:15:59 TheCreat0r: if you start from scratch, yes. it will build a toolchain and specially glibc takes ages to build Jul 01 22:23:12 zecke_ : i set up a dev environment using vmware, i only allocated 8GB for the disk, how many GB's (on average) would you need? i see its filling up quite fast Jul 01 22:23:53 8 gig is enough Jul 01 22:24:21 woglinde : just compiling nano, and i've filled up half half the disk lol Jul 01 22:24:33 yes normal Jul 01 22:24:40 a lot of dependencies Jul 01 22:24:42 woglinde : ah ok Jul 01 22:24:56 so 8GB will be enough Jul 01 22:25:06 * TheCreat0r sighs in relief Jul 01 22:25:13 for angstroem-x11 yes Jul 01 22:25:29 great! Jul 01 22:26:19 TheCreat0r: will, you can easily run out of disk-space Jul 01 22:26:47 TheCreat0r: you can choose between keeping the stuff you build on disk (default) to deleting it after it was built (INHERIT += "rm_work" in your local.conf) Jul 01 22:27:08 lol Jul 01 22:27:25 cant do that, or everytime i build something it will take a year Jul 01 22:27:46 TheCreat0r: let me be more clear Jul 01 22:27:46 zecke, is it possible to migrate the build directory to another disk? Jul 01 22:28:10 TheCreat0r: you will have tmp/work with the build directories of the stuff you build Jul 01 22:28:28 ah Jul 01 22:28:29 TheCreat0r: they are not referened, but e.g. if you want to fix stuff or take a look at the config.log they are around Jul 01 22:28:49 TheCreat0r: with rm_work these are deleted after the package was packaged and the needed stuff copied to other places Jul 01 22:29:01 i seee Jul 01 22:29:07 see* Jul 01 22:29:07 TheCreat0r: and yes basicly conf/bitbake.conf of OE has a set of path's you can freely set Jul 01 22:29:34 zecke_ : great so i can move alot of stuff to another virtual disk Jul 01 22:33:11 TheCreat0r: most of the build systems (e.g. autofoo) don't like symlinks so please try to avoid them Jul 01 22:34:05 zecke_ : ok Jul 01 23:08:23 TheCreat0r: do you have any suggestions as to where to look for this memory (mb-applet-wireless) segfault issue? Jul 01 23:10:37 * summatusmentis leaves again, sorry all Jul 01 23:43:59 koen|away: you rang? Jul 02 00:30:57 hi rwhitby Jul 02 00:34:31 rwhitby: would it make sense to add something along the lines of http://people.openmoko.org/~henryk/makefile-any-image.patch to MokoMakefile, in order to facilitate the building of arbitrary images? (E.g. at university we're probably going to have our own image recipes) Jul 02 01:08:52 TheCreat0r: ping Jul 02 01:09:55 I want to add a task after do_fetch and before do_patch that runs tar xvzf on two tarballs, how can I do that? Jul 02 01:10:09 I know how to add the task, just not how to run tar from a .bb Jul 02 01:11:25 Zero_Chaos: do_install_append() { tar -xzvf /path/tar.gz -C dest } Jul 02 01:12:21 CoreDump|afk: can I append to do_unpack or is that REALLY frowned on? Jul 02 01:12:40 well, you could I assume Jul 02 01:12:54 never saw a custom do_unpack tho ;) Jul 02 01:15:16 CoreDump|afk: inside the distributed file are two other tarballs that need to be unpacked. it's a bit dumb, but I didn't make the release file.... Jul 02 01:24:26 is there anyone around who wants to help a newb troubleshoot gdb output? Jul 02 01:33:03 CoreDump|afk: syntax error, no love Jul 02 01:33:30 try do_configure_append or _prepend then Jul 02 01:34:15 needs to be before patch, so I'll try do_patch_prepend Jul 02 01:35:00 no love Jul 02 01:39:59 any ideas? Jul 02 01:59:56 So, anyone tell me how to run two native commands after do_unpack and before do_fetch? Jul 02 02:04:03 CoreDump|afk: wow, I don't think I had tar-native installed, trying again Jul 02 02:05:25 CoreDump|afk: nope, that didn't fix it Jul 02 02:07:15 Zero_Chaos: I believe gcc_csl-arm-2005q3.bb does what you want Jul 02 02:07:42 yep, pretty much exactly Jul 02 02:07:45 CoreDump|afk: thanks Jul 02 02:07:49 np Jul 02 02:10:03 indeed, and now on to the next part of my nightmare :-) Jul 02 02:20:59 03coredump2 07org.oe.dev * rf598d540... 10/ (1 packages/gnome/libbonoboui_2.18.0.bb): gnome: Update libbonoboui to 2.18.0 to avoid a compile error with 2.14.0 Jul 02 02:21:05 03coredump2 07org.oe.dev * r31bba003... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom-2007.1.conf: Update PREFERRED_VERSION_libgnomeui to 2.18.1, working around a compile problem with 2.16.1 Jul 02 02:28:51 akita/spitz have USB2.0 ports, yes? Jul 02 02:29:06 why in the world do we not have USB2.0 in the kernel? anyone? Jul 02 02:30:13 Zero_Chaos: They have a USB OTG port, which is an extension of USB2.0 technically speaking. spitz has an OHCI controller in the PXA270 though - no need for an EHCI driver Jul 02 02:30:40 OHCI is 1.1 I thought? Jul 02 02:30:41 They only operate at USB 2.0 Full Speed (11Mbps), not High Speed (480Mbps) Jul 02 02:31:32 so ehci_hcd won't work, and really 802.11g cards won't work at full speed either.... Jul 02 02:31:35 right? Jul 02 02:31:41 Correct. Jul 02 02:31:52 that just plain sucks Jul 02 02:31:53 The onboard controller on the xscale only runs at 11Mbps, iirc Jul 02 02:38:00 * * OE Bug 2585 has been created by metalgearsnorkel(AT)gmail.com Jul 02 02:38:02 * * No sound output - SL-C750 Jul 02 02:38:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2585 Jul 02 02:52:00 * * OE Bug 2128 has been RESOLVED (FIXED) by xjqian(AT)gmail.com Jul 02 02:52:02 * *  libc-headers_2.6.15.99 build fail Jul 02 02:52:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2128 **** ENDING LOGGING AT Mon Jul 02 02:59:56 2007