**** BEGIN LOGGING AT Tue May 18 02:59:57 2010 May 18 04:14:55 build #51 of goldfish is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/goldfish/builds/51 May 18 05:30:11 is "insmod sha1_generic" the correct syntax for loading the sha1 module? May 18 05:53:25 build #46 of pxcab is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/46 May 18 06:24:25 nm, the module seems to be broken. filed a ticket May 18 08:27:58 zombik: please attach the releveant dmsg lines as well May 18 09:14:03 blogic * r21499 /trunk/target/linux/ifxmips/files-2.6.33/arch/mips/ifxmips/danube/Makefile: [ifxmips] oops, missing fix from [21485] May 18 09:27:42 xMff: done. you were right it contained some interesting lines May 18 09:28:11 xMff: looks like not only something with sha1 is amiss May 18 10:21:07 blogic * r21500 /trunk/target/linux/ifxmips/files-2.6.33/arch/mips/ifxmips/danube/Kconfig: [ifxmips] adds proper kconfig to select the new mach types May 18 15:07:35 hi there! has somebody cross compiled an own netfilter extension for openwrt? Do you have a good how-to? May 18 15:09:09 duke83: i never tried it and i don't have a howto, but i think it might be easy to just integrate it as a patch to xtables-addons May 18 15:09:38 then you don't have to come up with your own build scripts for it May 18 15:09:56 nbd. yes this would be really nice May 18 15:11:47 as you probably think.. I am a newbie on openwrt and I just compiled the firmware and loaded it on the device. so the patch I have to put into feeds/packages/net/xtables-addons/patches/ May 18 15:11:47 right? May 18 15:13:06 duke83: are you familiar with the 'quilt' patch management tool? May 18 15:13:20 it makes creating patches a lot easier May 18 15:14:11 nbd: no never heard before ;) May 18 15:14:31 ok, never mind then. yes, the directory you mentioned is the right place for your patches May 18 15:14:41 i just saw your msg in the other channel May 18 15:14:53 you could prepare the patch on your linux pc when you've compiled it with xtables-addons May 18 15:14:54 sorry wrong smiley... actually I have never heard of quilt May 18 15:14:56 and then just copy the finished patch over May 18 15:15:12 nbd: doesnt the gargoyle guy do something similiar to what duke is looking for ? May 18 15:15:30 OutBackDingo: i have no idea what the gargoyle guy does May 18 15:15:39 I have read from this gargoyle guy... but did not find any docs about it May 18 15:15:57 duke83: well its based on a modified patched openwrt May 18 15:16:43 so might just want to peek at his build sequence, it might help, though maybe not May 18 15:16:56 i don't think that's what he needs May 18 15:17:16 it's not about patches to openwrt, it's about patches to a package of openwrt May 18 15:17:23 and that doesn't need a separate build script environment May 18 15:17:48 unless i'm missing something here ... May 18 15:17:56 no you get my point May 18 15:18:31 nbd: just looking at it, he also only patches so yeah your right May 18 15:19:15 duke83: as usual nbd is right :P LOL May 18 15:19:46 what let me puzzled... at the very end of the xtables-addons makefile I can see "$(eval $(call BuildTemplate,tee,TEE,xt_TEE,xt_TEE,46,+kmod-ipt-compat-xtables))" which is actually the same as I want, but I have no clue where the sources of TEE are or how I can put my own sources and let them compile May 18 15:20:34 I mean not the TEE extension is what I want, but my extension is similar to the TEE extension May 18 15:23:37 ok, that makefile uses some weird templating May 18 15:23:53 the first argument is the last part of the package name May 18 15:23:58 then the description May 18 15:24:25 then the names of the extensions May 18 15:24:44 priority is for the module loading order in /lib/modules/ May 18 15:24:49 because we load all modules at boot time May 18 15:24:58 and last one is package depdendencies May 18 15:25:04 you can use that for adding your extension May 18 15:25:10 btw. what kind of extension are you building? May 18 15:25:55 so would it be easier if I change the xtables-addons-1.22.tar.bz2 containing also my extension and putting a similar template instruction at the end of the file? May 18 15:26:39 very ugly but only to get it working for some tests, probably the easiest solution.. what do you think? May 18 15:26:41 no, don't change the tarball May 18 15:26:48 changing patches is easier May 18 15:27:03 and if you want to make it really easy to create patches, you should learn how quilt works May 18 15:27:17 there's a nice pdf here: http://www.google.de/url?sa=t&source=web&ct=res&cd=1&ved=0CBQQFjAA&url=http%3A%2F%2Fwww.suse.de%2F~agruen%2Fquilt.pdf&rct=j&q=agruen+quilt&ei=zbHyS665No6lsAbCuNmFDA&usg=AFQjCNErWjsRGnC2Kd5hTNUsq2bgxXQT-g May 18 15:27:21 oops May 18 15:27:23 i hate google's link rewriting May 18 15:27:45 http://www.suse.de/~agruen/quilt.pdf May 18 15:27:47 here's the right link May 18 15:27:57 ok I will learn it, and make a patch which includes my extension into to the xtables-addons directory, (new folder added, changed kconfig and mconfig) May 18 15:41:10 ping nbd May 18 15:42:08 is it possible to commit to the OpenWRT via the git repository or do I have to use SVN to do commits? May 18 15:43:28 <[florian]> you can use nbd's git repository and commit to the svn openwrt May 18 15:43:45 <[florian]> cshore: there is a script nbd has put online: http://nbd.name/gsc May 18 15:44:10 <[florian]> which does the binding between the git repository and the svn repository May 18 15:44:40 <[florian]> you can use svn+ssh://svn.openwrt as the svn repository to bind to May 18 15:49:59 with the username@svn.openwrt? May 18 15:50:30 with openwrt@svn.openwrt... May 18 15:50:36 not your username May 18 15:50:51 yeah, that's what I meant May 18 15:57:21 nbd: so finally I added a patch (created oldschool using diff) , now how will it be applied? May 18 16:02:22 duke83: if you throw it in patches/ it will be applied automatically May 18 16:02:27 when you rebuild the package May 18 16:05:27 you mean rebuilding by ./scripts/feeds install xtables-addons? May 18 16:05:41 no. make package/xtables-addons/{clean,compile} V=99 May 18 16:05:50 make sure you run make oldconfig first to select your new package after the makefile changes May 18 16:24:34 nbd: gsc gave a message git 'svn' is not a git command....is that normal? May 18 16:24:52 <[florian]> cshore: make sure you have git-svn installed on your machine May 18 16:25:32 [florian]: doh! May 18 17:14:28 nbd: it starts getting compiled :) thanks for your help! but I stumbled again over a problem, I am using some assembler code directly addressing i386 registers. Now I need to change them for the linksys architecture :S May 18 17:18:39 Error: unrecognized opcode `movl %eax,%edx' May 18 17:20:28 why do you need assembler for your module? May 18 18:23:25 for git commits, when I update git (git pull), do I have to resync with svn? May 18 18:24:52 if you git svn rebase, then you can no longer git pull May 18 18:24:58 i suggest sticking to using git svn rebase May 18 18:25:01 that's what i do May 18 18:25:11 i use my public git only for the initial clone May 18 18:25:28 ok May 18 18:25:57 so the command instead of git pull is git svn rebase? May 18 18:26:09 yes May 18 18:26:15 nbd: great, thanks May 18 18:26:29 and git svn dcommit to convert all git commits to svn commits May 18 18:27:44 ok...there a way to see beforehand what will get commited (does it still pop up an edit window, that if not changed cancels the commits)? May 18 18:28:15 well, it will always commit everything that is after the last git-svn commit May 18 18:28:38 git svn rebase always rebases all of your changes on top of the latest git-svn commit May 18 18:28:55 ok, make sense...don't leave un svn commits lying around May 18 20:15:49 jow * r21501 /trunk/package/firewall/files/lib/core.sh: [package] firewall: fix a possible deadlock when the firewall config has syntax errors during restart May 18 20:24:18 <[florian]> rtz2: ping? May 18 20:26:50 [florian]: pong May 18 20:28:16 <[florian]> rtz2: can you test this on your board: http://paste.debian.net/73808/ May 18 20:28:33 <[florian]> rtz2: it works fine for me here with an ic+ ip175c switch May 18 20:29:18 [florian]: I can but soonest tomorrow and maybe not until the weekend May 18 20:29:32 <[florian]> rtz2: sure, there is no hurry May 18 20:29:35 I left the device at home May 18 20:29:49 [florian]: could you mail the patch to me? May 18 20:29:50 <[florian]> rtz2: I just want to make sure that it works with other devices as well May 18 20:29:53 <[florian]> rtz2: sure May 18 20:29:58 ok May 18 20:30:00 <[florian]> rtz2: make sure you enable phylib in your kernel May 18 20:37:18 <[florian]> rtz2: you should have received it by now May 18 21:05:59 [florian]: I did May 18 21:06:04 thanks May 19 00:50:18 jow * r21502 /trunk/package/firewall/ (5 files in 3 dirs): (log message trimmed) May 19 00:50:18 [package] firewall (#7355) May 19 00:50:18 - partially revert r21486, start firewall on init again May 19 00:50:18 - skip iface hotplug events if base fw is not up yet May 19 00:50:19 - get ifname and up state with uci_get_state() in iface setup May 19 00:50:19 since the values gathered by scan_interfaces() may be outdated May 19 00:50:20 when iface coldplugging happens (observed with pptp) May 19 01:20:57 ping epointsystems May 19 01:55:48 jow * r21503 /trunk/package/firewall/files/lib/ (core_init.sh fw.sh): [package] firewall: implement disable_ipv6 uci option **** ENDING LOGGING AT Wed May 19 02:59:57 2010