| 00:17.56 | *** join/#uclibc swishy (n=swishy@121.98.132.55) |
| 00:33.00 | *** join/#uclibc wrobbie (n=rob@203.117.215.163) |
| 01:00.48 | *** join/#uclibc tazmania (n=sandp@118.101.79.155) |
| 01:21.49 | *** part/#uclibc dbtid (i=e37o6u9r@cpe-65-24-173-154.columbus.res.rr.com) |
| 02:12.44 | *** join/#uclibc tsukasa` (n=tsukasa@unaffiliated/tsukasa) |
| 02:28.45 | *** join/#uclibc odesus (n=Omar@201.165.141.180) |
| 02:28.59 | odesus | hi good evening |
| 02:30.16 | *** part/#uclibc odesus (n=Omar@201.165.141.180) |
| 02:30.49 | *** join/#uclibc odesus (n=Omar@201.165.141.180) |
| 02:31.32 | odesus | when I try to mount the root file sytem and copying the busybox structure in the mounted image I got an error space of /mnt partion |
| 02:31.41 | odesus | Any tip for solving this? |
| 02:33.40 | odesus | someone? |
| 02:37.13 | odesus | anybody? |
| 02:37.19 | odesus | no responses? |
| 02:37.54 | odesus | cler |
| 02:38.02 | odesus | mmm :( |
| 02:38.10 | *** part/#uclibc odesus (n=Omar@201.165.141.180) |
| 02:38.31 | *** join/#uclibc hiyuh (n=hiyuh@KD059133117089.ppp.dion.ne.jp) |
| 02:42.04 | *** join/#uclibc ccjoe (n=joeg@75-169-7-104.slkc.qwest.net) |
| 03:22.03 | swishy | why does everyone think there is someone to answer all questions posted in a channel 24x7 |
| 03:22.20 | swishy | mm acutallu |
| 03:22.33 | swishy | s/everyone/a vast quantify of people/g |
| 03:22.36 | swishy | ;p |
| 03:51.16 | *** join/#uclibc quitte_ (n=quitte@f048036232.adsl.alicedsl.de) |
| 04:14.17 | *** join/#uclibc tsukasa (n=tsukasa@unaffiliated/tsukasa) |
| 05:57.49 | *** join/#uclibc doc|schaffen (n=doc@82.115.100.147) |
| 05:58.02 | doc|schaffen | !moin |
| 06:03.19 | solar | swishy: to answer your question.... Sometimes 3hrs.. |
| 06:22.52 | *** join/#uclibc vinilios (n=k88@adsl77-101.ath.forthnet.gr) |
| 06:49.34 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 06:53.49 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 07:15.34 | blindvt` | yawns |
| 07:16.35 | blindvt` | "Can I handle this" is a sweet question |
| 07:19.00 | *** join/#uclibc hw__ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 07:20.59 | blindvt` | the dynamic loader on cris looks suspicious. I think one could do unpleasant things as is |
| 08:05.56 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 08:22.56 | CIA-9 | 03vda * r23679 10busybox/procps/sysctl.c: (log message trimmed) |
| 08:22.57 | CIA-9 | sysctl: fix bug 3894 (by Kryzhanovskyy Maksym) |
| 08:22.57 | CIA-9 | function old new delta |
| 08:29.19 | CIA-9 | 03vda * r23680 10busybox/procps/sysctl.c: |
| 08:29.19 | CIA-9 | sysctl: remove special-cased reporting of EPERM, |
| 08:29.19 | CIA-9 | <PROTECTED> |
| 08:29.38 | CIA-9 | 03aldot * r23681 10uClibc/Rules.mak: - bump version to 0.9.30-rc2 |
| 08:30.15 | CIA-9 | 03aldot 07uClibc_0_9_30_rc2 * r23682 10/: - tag 0.9.30-rc2 |
| 08:33.42 | CIA-9 | 03aldot 07uClibc_0_9_30_rc1 * r23683 10/: - tag r23457 as 0.9.30-rc1 |
| 08:45.55 | CIA-9 | 03vda * r23684 10busybox/procps/sysctl.c: sysctl: remove trivial comments. no code changes |
| 08:46.48 | CIA-9 | 03aldot * r23685 10website/uclibc.org/news.html: - announce 0.9.30-rc2 |
| 08:53.20 | CIA-9 | 03aldot * r23686 10uClibc/Rules.mak: - mark trunk as -rc2-svn |
| 08:54.09 | *** join/#uclibc swishy (n=swishy@60-234-136-97.bitstream.orcon.net.nz) |
| 09:43.37 | CIA-9 | 03vda * r23687 10busybox/procps/sysctl.c: sysctl: fix bug 3894 _for real_. |
| 09:44.38 | CIA-9 | 03vda * r23688 10busybox/procps/sysctl.c: return accidentally deleted "/* vi: set sw=4 ts=4: */" |
| 10:23.36 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 11:49.25 | *** join/#uclibc hw__ (n=hw@p578b3905.dip0.t-ipconnect.de) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc doc|schaffen (n=doc@82.115.100.147) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc wrobbie (n=rob@203.117.215.163) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc dinar_away (n=dinar@212.176.242.38) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc crome (n=crome@lekvar.ezit.hu) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc khem (n=khem@adsl-71-146-28-143.dsl.pltn13.sbcglobal.net) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc CIA-9 (n=CIA@208.69.182.149) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc sega01 (n=sega01@2001:470:8:29:250:4ff:fe20:dbc4) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc Zpenguin (n=Zpenguin@CPE001a70f2faee-CM001bd7ad54fe.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc blindvt` (n=blindvt`@chello213047219050.surfer.at) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc Redhatter (n=vk4msl@gentoo/developer/redhatter) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc sukram (n=markus@ti100710a080-3690.bb.online.no) [NETSPLIT VICTIM] |
| 11:49.25 | *** join/#uclibc Kaloz (i=kaloz@openwrt/developer/Kaloz) |
| 11:49.25 | *** join/#uclibc likewhoa (n=y0@ool-45717734.dyn.optonline.net) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc pepin (n=pepin@208.85.221.34) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc gnomon (n=gnomon@CPE001d60dffa5c-CM000f9f776f96.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc visitors (n=pulstar@198-tar-2.acn.waw.pl) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc mtr (n=michael@v29762.1blu.de) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc vapier (i=UserBah@wh0rd.org) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc torindel (n=torindel@mail.tau.com.pl) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc fabo (i=fabo@ubuntu/member/fabo) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc neli (i=micha@88.159.210.43) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc {Brad} (n=brad@molotov.compsoc.warwick.ac.uk) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc ashes (n=ashes@modemcable123.78-80-70.mc.videotron.ca) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc HcE (i=egtvedt@egtvedt.no) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc Jacmet (n=peko@88.197.163.5) [NETSPLIT VICTIM] |
| 11:49.26 | *** join/#uclibc olistudent (i=psybnc@85.214.116.235) [NETSPLIT VICTIM] |
| 12:02.52 | *** join/#uclibc blindvt` (n=blindvt`@chello213047219050.surfer.at) [NETSPLIT VICTIM] |
| 12:02.52 | *** join/#uclibc sukram (n=markus@ti100710a080-3690.bb.online.no) |
| 12:02.52 | *** join/#uclibc Kaloz (i=kaloz@openwrt/developer/Kaloz) |
| 12:02.52 | *** join/#uclibc pepin (n=pepin@208.85.221.34) [NETSPLIT VICTIM] |
| 12:02.52 | *** join/#uclibc gnomon (n=gnomon@CPE001d60dffa5c-CM000f9f776f96.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 12:02.52 | *** join/#uclibc visitors (n=pulstar@198-tar-2.acn.waw.pl) |
| 12:54.50 | crome | uhmm |
| 12:55.01 | crome | just stumbled into [uclibc-0.9.30_rc1] bash: can't resolve symbol '__uClibc_start_main' |
| 12:55.14 | crome | what about this one? |
| 12:59.12 | blindvt` | crome, first, there is an rc2 and second that is not enough info |
| 12:59.35 | crome | seemingly the last thing happens before this messup is >>> /lib/ld-uClibc.so.0 -> ld-uClibc-0.9.29.so |
| 12:59.45 | crome | after compilation, during postinst |
| 12:59.50 | crome | portage creates that symlink |
| 13:00.53 | blindvt` | crome, perhaps solar would know then |
| 13:01.34 | Jacmet | blindvt`: hi, I have a problem with your busybox r23047 |
| 13:01.54 | Jacmet | blindvt`: the FEATURE_SEAMLESS_ stuff |
| 13:02.18 | Jacmet | blindvt`: we used to have specific settings for that stuff in dpkg / dpkg-deb |
| 13:02.47 | Jacmet | blindvt`: but now FEATURE_SEAMLESS_GZ defaults to n and without it dpkg is pretty useless on "normal" debs |
| 13:03.08 | Jacmet | blindvt`: the question is how to handle this nicely in Kconfig |
| 13:03.45 | Jacmet | blindvt`: you cannot just add a default y if DPKG || DPKG_DEB to FEATURE_SEAMLESS_GZ as that's always visible so the default value doesn't get put to anything |
| 13:04.09 | Jacmet | blindvt`: and a select FEATURE_SEAMLESS_GZ in dpkg / dpkg-deb would mean that you cannot disable it |
| 13:05.25 | Jacmet | blindvt`: we could ofcause reintroduce a dummy CONFIG_FEATURE_DEB_TAR_GZ which would depend on DPKG || DPKG_DEB and select FEATURE_SEAMLESS_GZ, but that also seems kinda silly |
| 13:06.29 | Jacmet | blindvt`: what would you prefer? |
| 13:07.01 | blindvt` | Jacmet, hm. FEATURE_SEAMLESS doesn't sound familiar, ad hoc. Let me have a look at r23047 |
| 13:07.25 | Jacmet | blindvt`: it's for automatically handling gzip/bzip2/lzma stuff |
| 13:08.26 | blindvt` | Jacmet, that was vda's stuff, not mine fwiw |
| 13:10.01 | Jacmet | blindvt`: argh, yes - sorry, I need more coffee |
| 13:10.09 | Jacmet | I'll bring it up on the list then |
| 13:11.56 | blindvt` | Jacmet, a workaround (short of the user configuring it correctly) would be to add a if DPKG\nconfig SEAMLESS_GZ\n\tdef_bool y\nfi, i.e. enable it per default if dpkg is selected, and this avoid a crude select |
| 13:12.32 | blindvt` | Jacmet, please do. Denys might have an idea |
| 13:13.14 | Jacmet | blindvt`: are you sure that will work? if the user already has a .config with a SEAMLESS_GZ symbol? |
| 13:13.37 | Jacmet | the default blah stuff afaik only gets used when there's no definition in the .config |
| 13:13.56 | blindvt` | Jacmet, well you can't really force that same symbol off, of course |
| 13:14.22 | Jacmet | I think the best solution is to either select it from DPKG or reintroduce the CONFIG_FEATURE_DEB_TAR_GZ stuff |
| 13:15.42 | blindvt` | Jacmet, you can "select FORCE_SEAMLESS_GZ" in dpkg to disentangle the selected symbol from a user setting. This would be the prefered thing to do, i guess. |
| 13:16.18 | blindvt` | Jacmet, right. I wasn't aware that the CONFIG_FEATURE_DEB_TAR_GZ was removed. |
| 13:23.18 | blindvt` | in the end we'll just revert r23660 |
| 13:25.20 | blindvt` | it is not a sensible approach to do such a patch without fixing all related paths |
| 13:25.50 | blindvt` | especially not the non-threaded one ;) |
| 13:48.34 | CIA-9 | 03aldot * r23689 10website/uclibc.org/news.html: - fix typo |
| 13:50.26 | CIA-9 | 03vda * r23690 10busybox/procps/sysctl.c: sysctl: sysctl -a was still misbehaving, fix it |
| 14:15.02 | *** join/#uclibc hiyuh (n=hiyuh@KD125054017176.ppp-bb.dion.ne.jp) |
| 14:50.15 | CIA-9 | 03vda * r23691 10busybox/archival/Config.in: build system: dpkg needs .gz handling |
| 14:59.49 | *** join/#uclibc tsukasa (n=tsukasa@unaffiliated/tsukasa) |
| 15:10.17 | *** join/#uclibc wberrier (n=wberrier@12.191.193.40) |
| 15:37.33 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 16:07.41 | *** join/#uclibc quitte (n=quitte@f048036232.adsl.alicedsl.de) |
| 16:09.10 | *** join/#uclibc Lothsahn (n=nospam@8.2.183.242) |
| 16:09.10 | quitte | blindvt`: I used buildroot - and it still fails. |
| 16:11.24 | quitte | blindvt`: this is driving me crazy. i might even try a i386 lenny to build on next |
| 16:12.48 | quitte | blindvt`: also I'd be very glad if you could make a preconfigured buildroot tarball to lower the number of variables some more |
| 16:18.12 | blindvt` | quitte, which buildroot? It worked with mine, as said |
| 16:18.53 | blindvt` | quitte, just clone mine, select linux-2.6.27, latest binutils and gcc |
| 16:20.42 | quitte | ok. i selected 2.6.26. and selecting gcc snapshot didnt work because it wouldn't know what to download |
| 16:23.10 | blindvt` | quitte, checkout trunk, ln -s trunk gcc-4.4.0 && tar -chjf ../dl/gcc-4.4.0.tar.bz2 gcc-4.4.0 |
| 16:23.38 | *** join/#uclibc tsukasa` (n=tsukasa@unaffiliated/tsukasa) |
| 16:23.39 | blindvt` | quitte, alternatively the "4.4 (snapshot)" thing should work for the last snapshot (20081010, IIRC) |
| 16:34.27 | quitte | I was expecting to take that thing everywhere with me and learn c by now :P |
| 16:39.22 | blindvt` | quitte, well, it took me 15 minutes, excluding compile-time :P |
| 16:41.12 | quitte | blindvt`: if it weren't for your support I would have given up. I still feel bad about all the trouble I caused you, though. |
| 16:42.31 | quitte | blindvt`: I hope I can pay back by making scripts that build proper emdebian uclibc toolchains, though. been working on it occassionaly for years now :/ |
| 16:43.42 | *** join/#uclibc bbradley (n=bbradley@78-105-167-16.zone3.bethere.co.uk) |
| 16:44.49 | blindvt` | quitte, np |
| 16:46.19 | blindvt` | quitte, well GyrosGeier and codehelp already have uClibc working, except for the notorious i368_glibc->i386_uClibc case; codehelp applied my patch though, but i didn't have time to see how far such a beast gets by now |
| 16:47.09 | blindvt` | quitte, so if you're not doing awkward stuff like /me, then emdebian w/ uClibc is already in a pretty promising shape :) |
| 16:47.44 | quitte | blindvt`: afaik they just patch the binary toolchain to support uclibc. I'd prefer to do a full bootstrap |
| 16:48.11 | blindvt` | quitte, but i encourage you to give up and send that box to me at any rate ;) |
| 16:48.45 | blindvt` | quitte, yes. As of now only the spec is patched, which is not 100% adequate, at least for my taste |
| 16:48.53 | quitte | blindvt`: hmm. i think i could live without the 620 |
| 16:49.36 | quitte | found it on the wertstoffhof. all it needs is a generic psu |
| 16:49.49 | blindvt` | quitte, unfortunately shipping cost are awful, so thanks, but i doubt that sending it to me makes sense |
| 16:50.28 | quitte | blindvt`: shipping costs to de are quite acceptable for me |
| 16:52.07 | quitte | take it or zumbi gets it |
| 16:53.14 | blindvt` | quitte, he. Let me have a look at what a 620 is first |
| 16:53.27 | quitte | jornada 620 |
| 16:53.39 | blindvt` | quitte, zumbi is busybox with that arm cube, AFAIK |
| 16:53.44 | blindvt` | s/busybox/busy/ |
| 17:01.00 | blindvt` | quitte, send it to zumbi; I don't have time to play with it anyway :-| |
| 17:04.03 | quitte | ok. but then ill have to translate parts of the winCE interface :P |
| 17:38.22 | *** join/#uclibc landley (n=landley@adsl-76-197-33-46.dsl.austtx.sbcglobal.net) |
| 17:38.41 | blindvt` | 'lo |
| 17:40.11 | landley | blindvt: hiya. |
| 17:40.20 | landley | So I say you're maintainer now. I'm going to poke Erik about it. |
| 17:40.23 | landley | You put out releases. |
| 17:41.34 | blindvt` | landley, i'll do the 0.9.30, that's all. I don't care about being maintainer or not |
| 17:41.52 | blindvt` | landley, and i don't like cakes either. Way too sweet ;) |
| 17:42.23 | blindvt` | landley, send it to erik's kids if you really want to send one 8) |
| 17:42.51 | blindvt` | or somebody else who could need a cake. Your pick |
| 17:42.58 | landley | Are you saying you don't want to handle releases after 0.9.30? |
| 17:43.06 | landley | This was a one time thing? |
| 17:43.30 | landley | Because uClibc _needs_ a maintainer. Mike didn't, and Erik's retired. |
| 17:44.09 | blindvt` | landley, one step after the other. I'll keep the thing going, but no promises |
| 17:44.27 | landley | *shrug*. |
| 17:44.37 | landley | I can follow the list and back you up if you get swamped. |
| 17:44.48 | landley | I'm getting back into busybox development too. |
| 17:45.01 | blindvt` | landley, so yes, if you see it that way, i play maintainer *shrug* |
| 17:45.09 | landley | I burned out pretty badly due to overexposure to Bruce Fscking Perens, but I've pretty much gotten over that now. |
| 17:45.12 | landley | Woot. |
| 17:45.16 | blindvt` | landley, yea, i saw it. Very nice to see you back |
| 17:45.33 | landley | (Maintainer isn't really an honor, it's actually a janitorial position.) |
| 17:45.55 | landley | We want to put your name on the little sign on the supplies closet with the mop, just so we know who to blame. :) |
| 17:45.59 | blindvt` | landley, i did _alot_ janitorial stuff recently, so yea |
| 17:46.18 | blindvt` | grins |
| 17:47.07 | blindvt` | landley, so revert 23660 and test out our fine -rc2 |
| 17:47.19 | landley | Revert 23660? |
| 17:47.26 | landley | I'm building a bunch of -rc2 releases now. |
| 17:47.36 | landley | (I saw the message about people wanting uClibc root filesystems.) |
| 17:47.39 | blindvt` | landley, if that IBM guy doesn't fixup his stuff then it will be reverted within a couple of days, fwiw |
| 17:47.52 | landley | I point out I build those, and would be happy to put cross compilers and root filesystem images into the uClibc downloads directory if you like... |
| 17:48.12 | landley | It's got to be better than the old ones from 2005, and I'd be happy to support 'em... |
| 17:48.27 | landley | We also need a page collecting uClibc distros we know about. |
| 17:48.33 | landley | Gentoo embedded, embebian, etc. |
| 17:48.39 | landley | But that's a separate issue. |
| 17:48.47 | blindvt` | landley, well, as said, the thing was to use buildroot. If you provide prebuilt disk-images with your fwl on your site, then cool |
| 17:49.09 | landley | blindvt: I do, but currently only for release versions. |
| 17:49.20 | landley | I can set up a cron job to build the suckers _nightly_, if you like. |
| 17:49.29 | landley | So we can have "most recent release", and we can have "last night's build". |
| 17:50.01 | landley | I've got cross compilers so you can build stuff against uClibc to test it out, and I've got native root filesystems so you can build anything you like under an emulator (or in a chroot) on an actual uClibc system with a native toolchain. |
| 17:50.05 | quitte | was the bruce perens thing about the licensing problems of uclibc? |
| 17:50.13 | landley | It was about the licensing problems of busybox. |
| 17:50.18 | landley | Back when I was maintainer of busybox. |
| 17:50.39 | landley | He drove me away from a project I mantained. It's a long story. |
| 17:50.58 | blindvt` | huh? did somebody remove uClibc versions in mantis? |
| 17:51.15 | landley | Most of which is here: http://lwn.net/Articles/202106/ |
| 17:51.20 | landley | blindvt: mantis? |
| 17:51.23 | quitte | landley: i kind of read into it i think. unless therrre were more such stories. |
| 17:51.37 | quitte | forgot all the details, though |
| 17:51.43 | landley | There are many stories about Bruce. |
| 17:51.45 | blindvt` | landley, hm. no. Odd |
| 17:51.59 | landley | I'm not the only one he's driven away. |
| 17:52.12 | landley | You don't always hear about it because often, after encountering him, they leave the open source community entirely. |
| 17:52.22 | quitte | hehe ok its a new story. sorry about that |
| 17:52.51 | landley | blindvt: uClibc versions in mantis? (Mantis being the bug tracker?) |
| 17:52.55 | landley | Did I miss something? |
| 17:53.26 | blindvt` | landley, yes, mantis being the bug collector, as you say. nm |
| 17:53.54 | landley | The bug tracker used to _be_ buggy, but that was 2-3 years ago. It seems to have been upgraded since... |
| 17:54.29 | landley | blindvt: the thing about buildroot is that it used to be maintained by the same people who did uClibc and busybox. It was all one big project. |
| 17:54.35 | landley | These days, that's not really the case. |
| 17:54.40 | blindvt` | i looked at the wrong project, all is well. |
| 17:54.46 | landley | Ah, ok. |
| 17:55.28 | blindvt` | see you later.. |
| 17:55.30 | blindvt` | & |
| 17:55.31 | landley | g'night. |
| 18:48.03 | quitte | meh! still doesn't work |
| 18:48.54 | blindvt | quitte, what doesn't work |
| 18:50.12 | quitte | it's pretty much the same as it always was |
| 18:50.29 | quitte | except some additional lock ups |
| 18:50.58 | quitte | I removed all traces of the previous toolchain |
| 18:51.45 | blindvt | quitte, the configs that i used for my tree, fwiw http://uclibc.org/~aldot/buildroot/br2-git-sh3el_for_quitte.tar.gz |
| 18:52.12 | quitte | so I'm clueless what you did different from me. except that you are building on i386, i guess |
| 18:53.00 | quitte | blindvt: thanks. ill try those. and then try them in a i386 virtualbox |
| 18:56.34 | blindvt | quitte, busybox is static in that config, so beware |
| 18:57.04 | tsuramoto | quitte, you could compare your config with mine |
| 18:57.06 | blindvt | quitte, yes, i'm on i386 |
| 18:58.00 | blindvt | quitte, modulo bugs this should make no difference of course |
| 19:01.34 | quitte | i know your static busybox was fine. both ls and ifconfig worked. once i get to that point it shouldn't be too hard to figure out what exactly breaks it. |
| 19:04.13 | blindvt | quitte, i'd try to build the exact same config on your normal host first and only then change something, obviously |
| 19:04.14 | quitte | and i'll just believe that you didn't build against glibc just to mess with me ;) |
| 19:04.54 | blindvt | that's heresy! ;) |
| 19:05.02 | quitte | yes. just did make oldconfig&&make with your confs |
| 19:05.45 | blindvt | silentoldconfig. Don't wear out all those little glyphs for no benefit |
| 19:06.15 | quitte | i just put new toner in my monitor |
| 19:06.24 | blindvt | :) |
| 19:11.09 | quitte | hmm. make clean emptied the dl directory :( |
| 19:15.57 | blindvt | quitte, that's why the first thing i do is to set the DL_DIR to "...down". But agree, i'll fix that |
| 19:16.27 | blindvt | (i inherited that, fwiw. Not my notion of fun) |
| 19:17.24 | blindvt | quitte, wait. It's only clean on distclean |
| 19:17.53 | quitte | im pretty sure i didn't do a distclean |
| 19:18.28 | quitte | anyways now one of your local settings bites me:svn: Kann keine ra_local-Verbindung zu einer URL aufbauen |
| 19:19.05 | blindvt | quitte, look at the toplevel Makefile, "DL_DIR". |
| 19:19.22 | blindvt | quitte, if your stuff is uptodate, then disable UPDATES in the .config |
| 19:19.51 | quitte | blindvt: i'll just start from zero if you tell me your git repo address again |
| 19:20.35 | blindvt | quitte, don't know offhand. mompls |
| 19:20.56 | blindvt | quitte, git clone git://repo.or.cz/buildroot.git |
| 19:21.28 | blindvt | should memorize that |
| 19:22.25 | tsuramoto | has countless block corruption in his memory |
| 19:22.58 | tsuramoto | notebook, man's most useful invention |
| 19:22.59 | blindvt | mumbles something about journaling and head-crash |
| 19:24.00 | *** join/#uclibc landley (n=landley@76.197.33.46) |
| 19:24.21 | quitte | sigh. xfs is slow and caused me more trouble than reiser. I'm going back soon! |
| 19:25.11 | landley | Reiserfs is unmaintained, though. |
| 19:25.31 | landley | Its maintainer was convicted of murdering his wife and his company collapsed, that kind of puts a damper on development. |
| 19:25.59 | landley | Plus he'd already been ignoring reiser3 for several years, in favor of a complete rewrite (reiser4) that was unmergeable. |
| 19:26.01 | quitte | i know. would be nice if they gave him a laptop and internet |
| 19:26.04 | landley | Even Novell's given up on it now... |
| 19:26.23 | landley | quitte: Which part of "murdered his wife" are you having trouble with? |
| 19:26.31 | landley | The man is _insane_. |
| 19:26.38 | landley | You want to depend on him for maintaining your filesystem? |
| 19:26.58 | quitte | so what. the filesystem worked for me for years better than any other i tried |
| 19:27.11 | landley | Out of curiosity, what are your criteria |
| 19:27.32 | landley | Define "work better". |
| 19:27.53 | landley | I've found ext3 to work pretty well anywhere you can accept the megabyte overhead for the journal. |
| 19:28.24 | quitte | doesn't overwrite files with zeros if its corrupted. doesn't take forever to delete a directory. no need to fsck every 30 reboots... |
| 19:28.54 | landley | Reiser3 has several known (and unfixable) corner cases where if you interrupt it while it's writing a specific block, the entire filesystem is gone and no fsck yet written can recover your data. |
| 19:29.03 | tsuramoto | <PROTECTED> |
| 19:29.09 | tsuramoto | this is my mdev.conf http://qna.nu/paodqa/ |
| 19:29.11 | landley | Sounds like ext3 to me. |
| 19:29.13 | quitte | i dont want to be able to restore a totally screwed fs. |
| 19:29.40 | tsuramoto | I could symlink them my self, but mdev should take care of it right? |
| 19:29.46 | landley | tsuramoto: there are no stdin/stdout/stderr entries in /sys. |
| 19:29.56 | landley | Have your boot script make those, they're just symlinks. |
| 19:30.08 | quitte | i'll try reiser3 again. i never had a reason to complain about that. |
| 19:30.17 | landley | tsuramoto: mdev just creates the devices listed in /sys. |
| 19:30.38 | landley | It doesn't do things like create a /dev/shm and mount a tmpfs on it. |
| 19:30.40 | landley | That's your job. |
| 19:30.54 | landley | quitte: you realize that filesystem is unmaintained, right? |
| 19:32.10 | quitte | yes. wasn't there a recent effort to revive reiser4? |
| 19:32.46 | landley | quitte: the russians he'd hired tried to find funding to keep the company going, but nobody wanted to give them money. |
| 19:32.54 | tsuramoto | landley thanks |
| 19:33.09 | landley | Keep in mind that reiser4 is the one that A) never got merged, B) needed massive architectural changes before it was mergeable. |
| 19:33.29 | landley | (Changes with Reiser's ego prevented them from making, because he thought he could write a better VFS layer in reiserfs than the linux kernel guys could.) |
| 19:33.53 | quitte | maybe ill try zfs :P |
| 19:34.00 | landley | Apparently, the linux kernel guys didn't see the "genius of his vision", or some such. Neither did his wife, or the people who didn't want to fund his company. |
| 19:34.34 | landley | Eh, zfs is mostly marketing hype as far as I can tell. |
| 19:34.46 | landley | And the sun code isn't licensed in a way that'll ever let it be merged into the linux ekrnel. |
| 19:34.56 | quitte | so ext3 or 4 and just disable the fscking? |
| 19:35.05 | landley | (Intentionally: Sun's trying to use it to differentiate from Linux.) |
| 19:35.21 | landley | Yup. ext3 doesn't _need_ to be fscked. It's journaled. |
| 19:35.29 | landley | tune2fs -c 0 -i 0 |
| 19:36.01 | quitte | growing ext3 online failed for me once, though |
| 19:36.10 | landley | I've never tried it. |
| 19:36.27 | quitte | another reason i prefer reiser. there it always worked for me |
| 19:36.42 | landley | quitte: feel free to keep using the 2.4 kernel as long as you like. |
| 19:37.04 | landley | The maintainer is gone, the filesystem has been abandoned, but if that's not a problem for you, stick with it. |
| 19:37.16 | landley | You can use old kernel versions indefinitely. |
| 19:37.30 | landley | There are still people out there using 2.2 kernels. |
| 19:37.52 | landley | The thing about reiser's failure modes is if you were lucky, you never saw problems. |
| 19:38.00 | quitte | there are plans to remove reiser from the kernel? |
| 19:38.06 | landley | If you were unlucky, the filesystem went away completely and nothing out of it could be recovered. |
| 19:38.12 | landley | it had fewer in-between failure modes. |
| 19:38.37 | landley | quitte: yes, but remember how long it took them to remove devfs? |
| 19:38.53 | quitte | hehe |
| 19:39.04 | landley | They let it bit-rot into clear uselessness over a period of years, then removed it when it finally had zero defenders. |
| 19:39.27 | landley | As long as one person speaks up in its defense, they'll keep it. They won't _fix_ it, they'll just leave it there for the person who complains to try to fix themselves. |
| 19:39.58 | quitte | sigh. so ext it is. thanks for talking some sense into me. well - if it doesn't fail on online growing in a few tests |
| 19:40.53 | landley | I think online growing is only officially supported in ext4, but I'd have to look it up. |
| 19:41.03 | landley | (It should work in ext3, but again, I haven't tried it...) |
| 19:41.29 | landley | Oh, and the thing about growing it is you have to tell it when you originally format it what the maximum size you'll ever grow it to is. |
| 19:41.56 | landley | Otherwise it has to expand existing inode tables, which is possible but a real pain and not something you can do without unmounting it. |
| 19:42.33 | quitte | you're better at giving reasons for reiser than i am ;) |
| 19:42.37 | landley | quitte: do "man mke2fs" and search for -E resize= |
| 19:42.59 | landley | quitte: reiser is a bit like windows. |
| 19:43.49 | landley | It's pretty, it benchmarks well, if you look under the hood it'll make youre brain explode, and if anything nontrivial does go wrong you have to start over from scratch. |
| 19:44.38 | landley | Also, the version everybody uses is unmaintained and the new version is unusable. |
| 19:44.44 | landley | The metaphor holds up quite well, really. :) |
| 19:45.05 | quitte | i dont see how the starting from scratch part is bad. i hate partially broken filesystems. |
| 19:45.26 | landley | quitte: by "start from scratch" I mean "reformat and restore from a backup". |
| 19:46.04 | quitte | yup. thats pretty much what i would do with any other fs once random files are broken |
| 19:47.32 | quitte | i never tried jfs. but i never had the impression anyone did. so not worth giving a try? |
| 19:48.19 | landley | quitte: I haven't heard anything bad about jfs, but it's only really ever been used by IBM's customers. |
| 19:48.36 | landley | It's an existing filesystem IBM had, which they ported from OS/2 to Linux. |
| 19:48.50 | landley | They still actively support it, so it's maintained. |
| 19:49.18 | landley | That said, I'm still not sure why you're avoiding the default filesystem of Linux. |
| 19:49.27 | landley | The most widely debugged of the lot is ext3. |
| 19:49.40 | landley | Something like 90% of all Linux deployments use ext3. |
| 19:50.12 | landley | The "oh no, that one's buggy" argument kind of breaks down if it's the one that's had the most testing by an order of magnitude. |
| 19:54.29 | quitte | well. reiser just worked for me. for years. and just for growing i'd have to plan ahead instead just using the debian installers defaults? i guess it's the best thing i could do. or live with the speed issues of xfs. still it annoys me. can't be that hard not killing anybody. |
| 19:55.47 | landley | quitte: Reiser was crazy for years before he killed his wife. |
| 19:55.59 | landley | Keep in mind he was pushing me to get reiser4 on my crunch time list back in... 2002? |
| 19:57.03 | landley | Yup, 2002: http://lkml.org/lkml/2002/10/27/92 |
| 19:57.22 | landley | So reiser4 was "real soon now, hold up the release for it" _six_years_ ago. |
| 19:57.53 | solar | landley: actually they stopped maintaining jfs |
| 20:00.57 | landley | Did they? |
| 20:01.00 | landley | I haven't been following it. |
| 20:01.34 | solar | it has it's time and place. |
| 20:01.47 | solar | usually on SAN is where it shines |
| 20:02.49 | landley | solar: I'm told that's what xfs gets used for. |
| 20:03.00 | landley | it's great if you need to keep lots of disks spinning in parallel, not so much otherwise. |
| 20:03.24 | landley | ext4 is really "we're adding 48 bit extents, and we've copied the entire directory again to do so". |
| 20:03.25 | landley | Sigh. |
| 20:03.36 | landley | ext2 and ext3 should never have been two separate filesystems. |
| 20:03.56 | landley | If I did much kernel work, I'd try to teach the ext3 driver to do ext2 with a null journal. |
| 20:04.15 | landley | (Technically you can today: use an external journal and put it in a ramdisk.) |
| 20:04.20 | landley | This is HIGHLY INADVISABLE, of course. :) |
| 20:04.30 | landley | And a waste of memory. |
| 20:04.46 | solar | is still hurt over the removal of e2fsprogs from busybox. |
| 20:05.14 | solar | I have to use an old 1.1.3 version to get desired functionality |
| 20:05.16 | landley | solar: the implementation it had was total pants. |
| 20:05.34 | landley | solar: wasn't it still there in 1.2.2? |
| 20:05.54 | landley | I was working on one in toybox, and need to pick that up again. It just got buried with other todo items... |
| 20:06.08 | landley | (Alas, toybox was neither something I was paid to do, nor my top priority hobby project.) |
| 20:06.09 | solar | I think 1.1.3 was the last. I could be wrong. But thats about when it went away afaik. |
| 20:06.27 | landley | Back when I was banging on firmware linux, tinycc, toybox, and kdocs, I git of got spread a bit thin. |
| 20:06.36 | landley | I'm pretty sure it was still there in 1.2.2. |
| 20:06.43 | landley | (That's the last release I put out.) |
| 20:07.05 | landley | I note that of a 1 megabyte busybox executable, e2fsprogs was something like 350k of the total. |
| 20:07.34 | landley | It was over a _third_ of busybox. |
| 20:07.37 | landley | No excuse for that. |
| 20:08.13 | landley | I still havent' quite figured out what I want to do about merging toybox and busybox. |
| 20:08.38 | landley | I still prefer the toybox infrastructure, I might just start wholesale porting the busybox applets I like to toybox. |
| 20:08.51 | landley | I could make pretty darn rapid progress that way. :) |
| 20:09.51 | landley | Denys has good technical judgement, but he doesn't question existing assumptions enough. |
| 20:10.03 | landley | He doesn't go back and do crapectomies on stuff that already works nearly enough. |
| 20:10.21 | tsuramoto | got the /dev/fd thing fixed |
| 20:10.40 | landley | (Not that there's much time to do that when you're maintainer. It's hard to keep up with just reviewing hte patch submissions...) |
| 20:10.41 | landley | tsuramoto: cool. How? |
| 20:10.53 | tsuramoto | how come /var/log/lastlog in not filled with *something* btmp is also empty |
| 20:11.39 | tsuramoto | landley, removed the symlink stuff in mdev.conf, and put ln -s /proc/self/fd /dev/fd etc in an init script |
| 20:12.13 | landley | There was symlink stuff in your mdev.conf? |
| 20:12.20 | tsuramoto | yes |
| 20:12.41 | tsuramoto | I used cross-lfs embedded book as a reference |
| 20:12.55 | tsuramoto | -rw-rw-r-- 1 root utmp 0 Oct 15 20:04 /var/log/lastlog |
| 20:13.08 | landley | Wow, did the -> symlink syntax ever actually get implemented? |
| 20:13.15 | landley | I should look at current busybox mdev... |
| 20:14.29 | tsuramoto | I don't know, I learn as I go |
| 20:14.31 | tsuramoto | http://cross-lfs.org/view/clfs-embedded/x86/bootscripts/mdev.html |
| 20:15.00 | landley | solar: does mdev actually support the -> symlink syntax? |
| 20:15.03 | landley | Or did you patch your mdev? |
| 20:15.53 | tsuramoto | my mdev.conf now looks like this |
| 20:15.55 | tsuramoto | http://qna.nu/p1zcy0/ |
| 20:17.51 | solar | landley: never got it to work. My conf looks like this. http://rafb.net/p/GAg1sQ81.html |
| 20:18.20 | solar | I've yet to see anybody really use mdev.conf anywhere close to it's functionality level cept myself. |
| 20:18.38 | solar | i spent quite a bit of time searching the internet for anybody that might have some decent examples. |
| 20:19.07 | landley | solar: I don't think the -> syntax we discussed ever actually got implemented. |
| 20:19.22 | landley | I know I didn't get around to it. (Maintainership keeps you busy.) |
| 20:19.26 | quitte | solar: so will i find yours once i get to that point? |
| 20:21.13 | solar | quitte: sorry will you find %s ? |
| 20:21.41 | *** join/#uclibc likewise (n=likewise@87.234.195.68) |
| 20:22.28 | tsuramoto | mmm busybox doesn't have lastlog command, assumes /var/log/lastlog can be removed |
| 20:24.23 | tsuramoto | ahh manpage confirms |
| 20:25.47 | quitte | modulo s? |
| 20:36.01 | solar | blindvt: new uclibc-r2 seems to segfault on a few things from the get go. |
| 20:36.59 | solar | right after upgrade from rc1 python/perl segv. It's in that exact spot of threading. |
| 20:41.18 | solar | hrmm strange. And after a downgrade back to the old bins it continues to be broken. Really odd |
| 21:01.19 | landley | solar: there's a broken commit he means to revert. |
| 21:01.57 | landley | solar: http://uclibc.org/lists/uclibc/2008-October/020293.html |
| 21:02.09 | landley | That's the only pending big issue with -rc2 I'm aware of... |
| 21:09.00 | tsuramoto | ahh tftpd isn't being nice |
| 21:09.03 | tsuramoto | tftpd: stdin is not a socket: Socket operation on non-socket |
| 21:09.29 | tsuramoto | lrwxrwxrwx 1 root root 9 Oct 15 20:04 /dev/stdin -> /dev/fd/0 |
| 21:47.55 | *** join/#uclibc gcl (n=gcl@S01060016b61d1226.cg.shawcable.net) |
| 21:53.01 | landley | Is stdin a socket in this case? |
| 21:57.17 | tsuramoto | no a symbolic link |
| 21:57.44 | tsuramoto | /dev/stdin is symbolic link |
| 21:59.38 | tsuramoto | time to sleep here |
| 21:59.42 | tsuramoto | thanks for the help guys |
| 22:17.39 | landley | tsuramoto: you're welcome. |
| 22:17.49 | landley | (What does the symbolic link point to?) |
| 22:18.00 | landley | <PROTECTED> |
| 22:22.17 | *** join/#uclibc irotas (n=irotas@CPE00131042e14d-CM0018c0b37a44.cpe.net.cable.rogers.com) |
| 22:30.20 | *** join/#uclibc pcgeil (n=steffen@p549E62A9.dip.t-dialin.net) |