**** BEGIN LOGGING AT Sat Jul 23 23:59:56 2005 Jul 24 01:37:13 foomatic Jul 24 10:40:01 NAiL: ping? Jul 24 11:00:26 Is berlios still down? Jul 24 11:07:44 VoodooZ: No - it's up. At least it was earlier Jul 24 11:09:24 cool. thanks. I tried last night and it was still down. Jul 24 11:09:50 We are talking their svn access right? Jul 24 11:11:08 Yes Jul 24 11:11:22 if it's down, I have put up a current snap to http://www.vanille.de/temp/oe/bitbake-snapshot-050728.tar.bz2 Jul 24 11:11:30 (assuming you talk about bitbake) Jul 24 11:15:24 hey mickeyl! Jul 24 11:15:32 (just like the song!) Jul 24 11:15:44 hey Jul 24 11:15:53 I'm doing an update right now to see if it works. Jul 24 11:16:26 Great! something new: Jul 24 11:16:26 UBRRL = F_CPU / (3277 * 16L) - 1; /* approximate sender's rate of 3276.8 bit/s */ Jul 24 11:16:31 oops. wrong line: Jul 24 11:16:43 svn: Berkeley DB error while opening 'nodes' table for filesystem /svnroot/repos/bitbake/db: Jul 24 11:16:43 Cannot allocate memory Jul 24 11:16:43 svn: bdb: Logging region out of memory; you may need to increase its size Jul 24 11:16:43 make: *** [update-bitbake] Error 1 Jul 24 11:17:39 lucky me. Jul 24 11:18:28 nasty Jul 24 11:19:13 f___! Jul 24 11:19:57 rwhitby has put bitbake in the nslu2-linux.org monotone branch. Jul 24 11:20:17 make update? Jul 24 11:20:27 yep. Jul 24 11:20:45 rwhitby org.nslu2-linux.dev * rc7aaea35... /Makefile: Migrated bitbake from the unstable and three-day-downtime berlios.de SVN to our well-maintained nslu2-linux.org monotone repository Jul 24 11:21:54 My current makefile only fetches oe-symlinks from berlios it seems - bitbake is from monotone Jul 24 11:22:39 And the oe-symlinks update goes ok Jul 24 11:23:49 I'll try mine again. Jul 24 11:23:58 make setup-bitbake Jul 24 11:24:23 should fetch bitbake from monotone Jul 24 11:24:58 yep. I think the first make update got the new makefile so it worked on the subsequent try. Jul 24 11:25:21 Yep - that's how it should work. Jul 24 11:26:40 thanks. Jul 24 11:59:24 ingeba-toandfrom: pong Jul 24 11:59:45 NAiL: Do you have time for helping me commit some stuff? Jul 24 12:00:17 hopefully very soon, yes Jul 24 12:00:25 Ok - ping me whenever Jul 24 12:01:10 just have to make sure my repo is up to date first Jul 24 12:14:27 hmm Jul 24 12:14:38 svn: Berkeley DB error while opening 'nodes' table for filesystem /svnroot/repos/bitbake/db: Jul 24 12:14:54 *sigh* Jul 24 12:15:09 aha Jul 24 12:15:13 make update twice Jul 24 12:15:14 :) Jul 24 12:16:19 ingeba-toandfrom: I asked you to send me stuff yesterday, didn't I? My mailserver appears to have gone berzerk Jul 24 12:16:39 Oh dear, forgot - will do now. Jul 24 12:16:44 ah, ok :) Jul 24 12:17:50 Thought committing procps, groff and man now. Jul 24 12:17:57 I'll mail them. Jul 24 12:36:40 NAiL: Should be on its way to your messed up mail server now. Jul 24 12:41:24 thank you Jul 24 12:41:51 You're most welcome :-) Jul 24 12:42:20 ingeba-toandfrom: pkg_postinst_${PN} would solve the _doc problem, wouldn't it? Jul 24 12:42:52 NAiL: So I believed, but I've tried it and it doesn't help. Jul 24 12:43:01 hmm, ok. Jul 24 12:43:20 I'll take a look at it after some quick food Jul 24 12:43:36 You want me to mail the proftpd bb file? Jul 24 12:43:48 Gonna eat too... way to late. Jul 24 13:07:34 ingeba-toandfrom: ok, looks good by me Jul 24 13:08:03 Cool Jul 24 13:08:10 fixed the MAINTAINTER field in streamripper, missing quotes Jul 24 13:08:13 that's it Jul 24 13:08:17 Ok - will do Jul 24 13:08:31 I did it on my end.. Shall I commit them for you? Jul 24 13:09:00 Ok - at some point I have to learn how, though :-) Jul 24 13:09:17 Have to eat my luxurious dinner now anyway :-) Jul 24 13:09:30 hehe, god appetitt ;) Jul 24 13:09:35 Takker. Jul 24 13:23:47 Should show up in #nslu2-linux in a short while Jul 24 13:35:07 ingeba-eating: what other symlinks did you want added? Jul 24 13:36:48 Hmm ... man-pages for now - also added psmisc, but haven't tested it yet. Jul 24 13:40:06 ok, I'm doing a full build to test your stuff Jul 24 13:40:29 are there any changes to the psmisc stuff, or just the symlink? Jul 24 13:40:35 I can test that later on as well Jul 24 13:41:42 hannesr: you sent me a mail the other day, right? Jul 24 13:41:56 yip, it bounced... Jul 24 13:42:07 should I resend now? Jul 24 13:42:12 yeah, my mailserver went bananas. Fixed now Jul 24 13:42:16 yeah, please do Jul 24 13:42:20 dyoung:) Jul 24 13:43:12 NAiL: resent Jul 24 13:43:18 playingi with new irc client. Jul 24 13:43:34 neat Jul 24 13:43:49 dunno if I like yet. Jul 24 13:44:24 hannesr: btw, why did it bounce? Jul 24 13:45:22 bouncemail from mail.hosted.servetheworld.net said ":Sorry, I wasn't able to establish an SMTP connection." Jul 24 13:45:57 bleh, that should only be delayed, not bounced Jul 24 13:46:26 after that, the mail says "I'm not going to try again; this message has been in the queue too long." Jul 24 13:47:23 weird Jul 24 13:47:26 this was 23 hours after I sent the mail Jul 24 13:47:39 seems a bit quick alright Jul 24 13:48:28 It should keep trying every six hours for two or three days IIRC (is what is recommended) Jul 24 13:48:34 and send a "delayed" message back Jul 24 13:53:49 ingeba-eating: I'll add them to openslug-packages.bb as soon as they all build from a the clean build I'm doing Jul 24 13:54:34 NAiL: THanks Jul 24 13:54:55 ingeba-eating: np Jul 24 14:00:44 hannesr: looked ok by me btw Jul 24 14:01:34 the configure script could be a lot better Jul 24 14:01:37 but I don# Jul 24 14:01:50 ...don't think it's worth the effort Jul 24 14:51:43 [g2]: did you see HowTo/OpenSlugDevInst ? Jul 24 14:52:00 <[g2]> Hey DaKa2 Jul 24 14:52:02 getting close :-) Jul 24 14:52:12 <[g2]> not yet ... looking Jul 24 14:53:20 I fixed cpp-symlinks, so that should go into openslug-native, and the .so-files should get fixed in the package, probably Jul 24 14:54:19 <[g2]> awesome! Jul 24 14:54:23 <[g2]> that looks sweet Jul 24 14:55:09 I havn't looked at the .so-files, whats wrong with them? Jul 24 14:55:48 <[g2]> the .so with the tail ? Jul 24 14:55:56 yes Jul 24 14:56:00 <[g2]> there's a comment in there for an old bug Jul 24 14:56:12 <[g2]> those lines just need to be removed Jul 24 14:56:16 there is a comment in the .so-file? Jul 24 14:56:23 <[g2]> nod Jul 24 14:56:30 who the hell adds comments in an so-file?? Jul 24 14:56:47 <[g2]> it been there for ever Jul 24 14:56:56 <[g2]> 4-6+ months Jul 24 14:57:17 <[g2]> it's a historical thing when the toolchain actually was broken iirc Jul 24 14:58:02 any reason noone removes it from the packages? Jul 24 14:58:47 <[g2]> I don't thing a good one Jul 24 14:59:03 <[g2]> but maybe the #oe folks would know better Jul 24 14:59:18 <[g2]> I'd guess there are few ppl that play with maintaining the toolchains Jul 24 14:59:37 yeah, its not really something you want to mess with Jul 24 15:05:47 <[g2]> DaKa2, so you managed to fix all the update-alternatives ? Jul 24 15:06:14 yep, and only got one small bug into the repo Jul 24 15:06:21 (so far) Jul 24 15:06:21 <[g2]> and is the util-linux shutdown/reboot still borked ? Jul 24 15:06:40 think so, it has lower prio than busybox Jul 24 15:06:48 <[g2]> cool... Jul 24 15:06:50 so we dont get it right now Jul 24 15:06:53 <[g2]> have you actually used it ? Jul 24 15:07:07 * [g2] hugs DaKa2 Jul 24 15:07:08 shutdown or the packages? Jul 24 15:07:12 <[g2]> shutdown Jul 24 15:07:27 from util-linux, no.. or, a week ago Jul 24 15:07:38 kindof fun messages.. Jul 24 15:07:40 but... Jul 24 15:10:32 <[g2]> DaKa2, can't you do the tail fixup as part of the ipkg post install ? Jul 24 15:12:01 I thought of that, rwhitbydid the openslug-native.bb, but I could probably add that Jul 24 15:23:11 [g2]: think I should fix it and check in? Jul 24 15:23:18 in openslug-native.bb? Jul 24 15:27:06 Nail: Discovered that configure in proftpd results in HAVE_LU=1 and HAVE_LLU=0 - this is no in tune with optware and perhaps to blame for the SIGSEGV. Jul 24 15:27:24 hmm Jul 24 15:27:26 Where in OE would I find the setup for the CPU/compiler ? Jul 24 15:27:54 openembedded/site/armeb-linux would probably be the right place to look Jul 24 15:28:09 I recalled the problem from earlier - it is too early to conclude but I just wanted to check. Will test tomorrow. Jul 24 15:28:27 proftpd_HAVE_LU=${proftp_HAVE_LU=X} or something like that Jul 24 15:28:41 In EXTRA_OECONF? Jul 24 15:28:53 no, in site/armeb-linux Jul 24 15:29:02 Ah... Jul 24 15:29:13 I messed a lot with that to get samba to work Jul 24 15:29:34 Can override with compile flags, but it should be automatically correct after configure I think Jul 24 15:30:35 Will do a native test first to make sure it is correct in Optware Jul 24 15:30:36 yeah Jul 24 15:31:03 if you add configure overrides in armeb-linux, configure *should* do stuff correctly Jul 24 15:31:07 (not always though) Jul 24 15:31:35 Nail : I will try that - though it should be a more global setting. Jul 24 15:31:40 Well - NAiL - good night - mail me the failure logs. I'll toss and turn for a few hours contemplating my miserable life. Jul 24 15:31:48 :-) Jul 24 15:31:50 ;) Jul 24 15:31:57 Good nithg Jul 24 15:31:59 night Jul 24 15:43:56 rwhitby: what do you think about fixing the last parts of HowTo/OpenSlugDevInst in the openslug-native postinst? Not really the right solution, but fast Jul 24 15:46:29 good plan for the moment, but slugbug it so we don't forget to fix it properly ... Jul 24 15:46:46 ok Jul 24 16:07:47 <[g2]> DaKa2, it's local to openslug-native Jul 24 16:07:55 <[g2]> that'd be a good checking Jul 24 16:08:15 [g2]: look at HowTo/OpenSlugDevInst again :-) Jul 24 16:08:58 * [g2] pulls again :) Jul 24 16:13:29 <[g2]> DaKa2, can you just ./configure.gnu && make && make test perl after that ? Jul 24 16:13:41 should work Jul 24 16:13:52 <[g2]> does that install the procps ? Jul 24 16:14:10 hm? Jul 24 16:14:59 should it? Jul 24 16:15:45 <[g2]> I dunno if procps build cross Jul 24 16:16:33 <[g2]> iirc it doesnt' Jul 24 16:17:55 <[g2]> I'd think with your latest we could either build perl natively or just package natively after a build Jul 24 16:18:15 <[g2]> s/build perl natively/build perl natively automatically/ Jul 24 16:18:28 I really think it should build perl natively from a bb with bitbake Jul 24 16:18:53 but the bb-file needs to be fixed a bit, or, alot Jul 24 16:19:13 it works, really good Jul 24 16:19:17 but... Jul 24 16:20:45 <[g2]> but ??? Jul 24 16:21:24 well, it .bb doesnt look that good, and probably does something it shouldn't Jul 24 16:21:31 s/it/the/ Jul 24 16:21:53 <[g2]> the perl .bb ? Jul 24 16:22:01 my perl .bb :-) http://david.thg.se/saker/openslug/perl-openslug_5.8.7.bb Jul 24 16:22:05 the one that works Jul 24 16:22:47 <[g2]> well if we've got to extensively change the .bbs to run nativley... then OE is broken in my mind Jul 24 16:22:48 and, as I said, the packaging needs to be discussed Jul 24 16:23:11 <[g2]> nod on the packagaing that's a given Jul 24 16:23:47 we could probably fix their bb, its just that it does so much strange stuff not needed on native builds Jul 24 16:24:24 <[g2]> well... it's interesting that's for sure Jul 24 16:27:26 [g2]: what was it that failed because of the comments in the .so? Jul 24 16:28:10 <[g2]> nothing would link Jul 24 16:28:28 ah, are comments supposed to work in .so-files? Jul 24 16:28:36 <[g2]> npe Jul 24 16:28:38 <[g2]> nope Jul 24 16:28:41 hehe :-) Jul 24 16:28:45 NAiL: how do you, [g2] and DaKa2 want to organise the management of the native feed? Jul 24 16:29:16 its not added to the .so by oe from what I can see, comes from upstream... Jul 24 16:30:17 <[g2]> DaKa2, well I don't know where it comes from either :) Jul 24 16:30:46 its in Makerules Jul 24 16:31:58 Im just testing a small patch to not include that Jul 24 16:33:10 dammit, its in more files.. oh well, 4 files to patch.. Jul 24 16:38:17 but this is strange, its actually in the latest main glibc cvs... Jul 24 16:38:59 [g2]: are you really sure comments are not suppsed to be supported in .so link scripts? Jul 24 16:39:08 supposed* Jul 24 16:39:31 <[g2]> try compiling hello.c with them in there Jul 24 16:39:56 <[g2]> I dunno if that are *supposed* to work but don't ? Jul 24 16:40:22 well.. my build host has that in /usr/lib/libc.so ... Jul 24 16:40:38 debian unstable.. Jul 24 16:40:41 and it works... Jul 24 16:41:13 <[g2]> well try it on the xscale Jul 24 16:41:42 then we should probably find out why it doesn't work on the slug.. Jul 24 16:41:46 linker? Jul 24 16:41:58 <[g2]> loader ? Jul 24 16:44:33 uhm... compiling a simple hello world works with comments in libc.so.. Jul 24 16:45:06 <[g2]> well a couple weeks ago there was only 1 line after the comments Jul 24 16:45:23 <[g2]> I dunno if that output line makes a different Jul 24 16:45:53 hm.. dunno.. Im going to try building perl with this.. Jul 24 16:46:13 <[g2]> sounds good Jul 24 16:55:16 * [g2] hugs his GigE switch Jul 24 16:56:34 .. Jul 24 16:56:34 * DaKa2 hugs his, uhm.. slug? :-) Jul 24 16:57:51 back later Jul 24 17:05:09 well. perl is building.. Jul 24 17:06:53 ..not Jul 24 17:08:10 {standard input}: Assembler messages: Jul 24 17:08:10 {standard input}:11841: Warning: end of file not at end of a line; newline inserted Jul 24 17:08:10 {standard input}:12024: Error: register expected, not '' -- `str r2,[' Jul 24 17:08:10 make: *** [toke.o] Error 1 Jul 24 17:08:52 <[g2]> that sounds familiar.... Jul 24 17:08:57 <[g2]> Assembler Jul 24 17:10:37 well, /usr/bin/as -> armeb-linux-as Jul 24 17:18:53 <[g2]> ipkg install openslug-native 2>&1 | tee openslug-native.log Jul 24 17:21:31 hm.. maybe I should use some swap? :-) Jul 24 17:24:29 <[g2]> DaKa2, http://pastebin.ca/18411 Jul 24 17:24:42 <[g2]> there are a couple errors at the end Jul 24 17:24:49 <[g2]> but overall looks fantastic Jul 24 17:25:16 ah yes, it tried to install wget before busybox was installed Jul 24 17:25:34 try it again and that should work Jul 24 17:25:54 <[g2]> you mean another ipkg insall ? Jul 24 17:25:59 yes Jul 24 17:26:15 do a ipkg upgrade first Jul 24 17:26:40 to get the updated cpio, so tar gets happy Jul 24 17:27:12 and to get the new sysvinit, so util-linux gets happy Jul 24 17:27:52 <[g2]> Downloading http://192.168.123.166/sysvinit_2.86-r16_nslu2.ipk Jul 24 17:27:52 <[g2]> sh: wget: not found Jul 24 17:28:40 well, dammit again Jul 24 17:28:50 ipkg install busybox Jul 24 17:29:13 it has to run the configure script for busybox befor things works for the other packages Jul 24 17:29:18 that is no good.. Jul 24 17:29:55 I prefer the way apt-get does that, ie, install A, configure A, install B, cofigure B, etc Jul 24 17:29:59 <[g2]> Ok... so let's get the current magic chant down Jul 24 17:30:14 <[g2]> I'm rebooting and then I'll init my hd Jul 24 17:30:45 ipkg update && ipkg install busybox && ipkg upgrade && ipkg install openslug-native Jul 24 17:30:58 ipkg install busybox should work for you now Jul 24 17:31:11 and fix the wget symlink Jul 24 17:33:07 well, swap fixed my perl build atleast Jul 24 17:51:48 and paying the bills fixed my problem having money Jul 24 17:52:17 <[g2]> get used to that :) Jul 24 17:54:52 <[g2]> DaKa2, ... ok.. that 99.99% works Jul 24 17:55:08 <[g2]> the ipkg upgrade nuke my ipkg.conf but except for that it seems to work Jul 24 17:55:10 :-) well, its not really that bad, I still have money, actually to much, but still... my spending account went down to ~700 USD for the rest of the month, havn't been that low for 5 years. But who am I to complain, that money could buy several slugs :-) Jul 24 17:55:15 hm.. Jul 24 17:55:31 uhm.. it nukes ipkg.conf? Jul 24 17:55:39 ahh, ipkg doesn't have that as a conffile? Jul 24 17:55:42 <[g2]> appears to Jul 24 17:58:01 hm, well, ipkg really should mark it as a conffile Jul 24 17:59:23 oh.. Jul 24 18:02:41 ipkg-collateral.bb should mark it as a conffile... Jul 24 18:02:49 now why did I put it there... Jul 24 18:03:04 s/I/they/ Jul 24 18:07:19 <[g2]> Ok... perl build as a normal user after install openslug-native is kicked off Jul 24 18:07:46 could you try that with the comments in libc.so and the other one? Jul 24 18:08:17 that is: ipkg install libc6-dev --force-reinstall after the openslug-native Jul 24 18:10:33 <[g2]> my current libc.so is fixed Jul 24 18:12:04 <[g2]> so I'm trying the "other one" first Jul 24 18:12:19 :) Jul 24 18:22:09 [g2]: nod that ipkg.conf should be in conffiles - also note that feeds usually go in seperate files under /etc/ipkg/foo-feed.conf now Jul 24 18:24:50 <[g2]> rwhitby-web, I've seen the /etc/ipkg/foo-feed stuff for awhile now Jul 24 18:25:14 <[g2]> but that's the first time /etc/ipkg.conf got updated Jul 24 18:25:38 nod Jul 24 18:26:02 someone recently did something in ipkg-collateral.bb and bumped the PR Jul 24 18:26:11 then you get a new one on an ipkg upgrade Jul 24 18:26:26 <[g2]> ahh. Jul 24 18:26:56 [g2]: feel free to add ipkg.conf to the CONFFILES in the .bb and push it ... Jul 24 18:27:36 <[g2]> rwhitby-web, I've got to send in my keys first :0 Jul 24 18:27:38 <[g2]> :) Jul 24 18:43:16 <[g2]> DaKa2, oh yeah... I think I forgot about "file" Jul 24 18:43:38 <[g2]> actually.... Jul 24 18:43:56 <[g2]> rm -f libperl.a Jul 24 18:43:56 <[g2]> /usr/bin/ar rcu libperl.a perl.o gv.o toke.o perly.o op.o pad.o regcomp.o dump.o util.o mg.o reentr.o hv.o av.o run.o pp_hot.o sv.o pp.o scope.o pp_ctl.o pp_sys.o doop.o doio.o regexec.o utf8.o taint.o deb.o universal.o xsutils.o globals.o perlio.o perlapi.o numeric.o locale.o pp_pack.o pp_sort.o Jul 24 18:43:57 <[g2]> ar: rcu: No such file or directory Jul 24 18:45:14 uh? Jul 24 18:50:06 <[g2]> something is broke Jul 24 18:50:30 <[g2]> wrong ar Jul 24 18:50:54 <[g2]> which ar Jul 24 18:50:55 <[g2]> /usr/bin/ar Jul 24 18:50:55 <[g2]> tom@fatturbo:~/src/file-4.12$ ls -l /usr/bin/ar Jul 24 18:50:55 <[g2]> lrwxrwxrwx 1 root root 17 Jul 24 20:45 /usr/bin/ar -> ../../bin/busybox Jul 24 18:50:59 DaKa2: http://ewi546.ewi.utwente.nl/OE_qa/cgi-bin//gunzip.cgi?tree=OpenEmbeddedBuild&full-log=1122245521.8967 <- see the busybox postinst error? Jul 24 18:51:46 that's during the openslug image build Jul 24 18:52:27 dammit.. I tested it, and the resulting image works, but I missed the errors.. Jul 24 18:52:54 we have a new build auditing system at http://ewi546.ewi.utwente.nl/OE_qa/cgi-bin//tinder.cgi?tree=OpenEmbeddedBuild&start-time=1122222601&display-hours=48 which picked them up :-) Jul 24 18:53:22 <[g2]> build auditing system --- yummy Jul 24 18:54:48 Ok, this is a fun problem, I need to test if running in image build, or on the device, to do different things with the test binary... fun :-) Jul 24 18:56:22 [g2]: my ar goes the right place, what does cat /usr/lib/ipkg/alternatives/ar say? Jul 24 18:57:32 <[g2]> cat /usr/lib/ipkg/alternatives/ar Jul 24 18:57:32 <[g2]> /usr/bin/ar Jul 24 18:57:32 <[g2]> ../../bin/busybox 50 Jul 24 18:57:46 uhm.. no binutils-symlinks installed? Jul 24 18:57:58 oh.. Jul 24 18:58:00 hehe Jul 24 18:58:07 * DaKa2 updates openslug-native.bb Jul 24 18:58:57 and my perl compiled, so I guess I can remove the .so stuff Jul 24 19:50:36 rwhitby-web: busybox, what do you think about: test -n 2> /dev/null || alias test='busybox test' Jul 24 19:52:03 instead of alias test='busybox test' that doesn't work when building image... Jul 24 19:57:09 DaKa2: I haven't looked at it closely enough to be able to give you an opinion on that. Jul 24 19:58:54 rwhitby-web: ah, well.. it seems to work.. should I bump PR and check that in, or should I check with someone? Jul 24 19:59:51 DaKa2: I think you've earnt the right to just check things in in the area of update-alternatives. But put a detailed comment in the checkin for the #oe folks to read. Jul 24 20:00:20 ok, will do :-) Jul 24 20:11:20 Now I just hope this doesn't break anything... Jul 24 20:11:43 * DaKa2 hides in bed for a few hours, bbl Jul 24 20:11:49 <[g2]> nite Jul 24 20:12:01 <[g2]> I'm about to run the updated version Jul 24 20:12:21 <[g2]> the one with the binutils chagnes Jul 24 20:12:45 ah, its actually just makes sure binutils-symlinks is installed **** ENDING LOGGING AT Sun Jul 24 23:59:56 2005