**** BEGIN LOGGING AT Thu Aug 14 02:59:59 2014 Aug 14 03:07:02 build #606 of kirkwood is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/606 Aug 14 04:04:14 build #242 of mvebu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/242 Aug 14 04:29:52 Is the dnsmasq can resolv the LAN IP? Aug 14 04:30:14 I tried that. but it is not working. Aug 14 06:59:16 jow_lapt1p: netifd: wan6 (1399): /lib/netifd/dhcpv6.script: line 1: fw3: not found Aug 14 06:59:37 i don't have the firewall package installed Aug 14 07:00:20 # CONFIG_PACKAGE_firewall is not set Aug 14 07:03:45 russell--: should be safe to ignore it Aug 14 07:08:28 is it safe to comment out that line? Aug 14 07:09:10 it's crapping up the syslogs ;-) Aug 14 07:14:04 sure, if you don't have the firewall installed anyway Aug 14 07:14:09 it doesn't do anything for you Aug 14 07:14:59 this stuff only effects you if you setup either ds-lite, map-e, map-t or lightweight 4over6 using dhcpv6 Aug 14 07:15:26 and sets the firewall zone of the tunnel to the zone of the parent interface Aug 14 07:15:44 but yeah probably doesn't apply to you in multiple ways ;) Aug 14 07:16:23 this one has native ipv6 (comcast) Aug 14 07:18:48 sure Aug 14 07:19:09 these tunnels are usually used so isps can do ipv6-only and tunnel legacy ipv4 over ipv6 Aug 14 07:19:20 but comcast still has regular dual-stack so its not needed Aug 14 07:23:57 you can do funny things wit these tunnels e.g. share one public ipv4-address across multiple subscribers and each one get's a certain portrange Aug 14 07:25:38 isps like that kind of stuff Aug 14 07:26:25 and of course there is 4 variants of that with 2-3 subvariants for each if you care enough Aug 14 07:26:34 isp's like that as well :P Aug 14 09:03:28 cyrus r42163 packages/net/ntpclient * ntpclient: moved to github Aug 14 10:07:44 <_trine> jow_laptop, the associated messages which are displayed from luci when you click on enable/disable wifi do not make sense Aug 14 10:08:26 <_trine> shall I elaborate further? Aug 14 10:12:08 <_trine> it seems to be ok on my main computer and my 3700 but on my netbook using my nano it does not seem ok Aug 14 10:12:43 <_trine> the disable/enable button always shows the disabled name Aug 14 10:14:38 <_trine> I suppose saying this another way it never looks like it is disabled Aug 14 10:19:48 <_trine> I have isolated a bit more this behaviour only shows iteself when my nano is in client mode and it looks like luci will not let you disable wifi in client mode Aug 14 10:21:38 <_trine> and in fact if you have the wifi disabled before going into client mode it will then not let you enable it Aug 14 10:21:52 <_trine> this is the same on both my 3700 and nano Aug 14 10:24:25 <_trine> however if you go into luci's edit mode you can from there enable or disable the wifi but from the front button you cannot Aug 14 13:17:49 for a package I look after, currently I have two packages, the main package, and a -nossl package, similar to how a few other packages do it. (monit, wget...) Aug 14 13:18:09 this means there's two packages on downloads.openwrt.org, so it's easy to install, but it's also two separate binaries, Aug 14 13:18:34 I've got more options to add now, and I don't want to get into a built binary for every combination, Aug 14 13:18:55 is it reasonable to have the "default" binary be pretty much complete, and just offer a smaller "minimal" one? Aug 14 13:19:08 or just make this config options only, for a single package? Aug 14 13:21:19 similarly, what's the recommended appropach for a library that can use nossl, openssl, or cyassl? one binary package, defaulted to openssl, with config options so it has to be rebuilt for the other forms? Aug 14 13:25:28 karlp: usually there is one build-variant per lib Aug 14 13:25:53 any examples for the various ssl libraries? Aug 14 13:26:11 I know some apps have support for whatever they find, but libwebsockets at least needs different compile flags for them Aug 14 13:26:21 https://github.com/openwrt/packages/blob/master/net/umurmur/Makefile Aug 14 13:26:54 https://dev.openwrt.org/browser/trunk/package/libs/ustream-ssl/Makefile Aug 14 13:26:58 and then a third for "nossl" ? Aug 14 13:27:07 sounds like it could get out of hand quickly? Aug 14 13:27:27 nossl+otherlibrary2+withoutlibrary3 Aug 14 13:27:35 sure Aug 14 13:27:55 i mean it sort of works if it just for ssl with 3 choices Aug 14 13:28:18 if its more complicated usually try to make a minimal, a "useful" and a "full" variant Aug 14 13:29:32 hrm, so a package that depends on this, has to have a depends on (openssl|cya|polar) sort of thing? Aug 14 13:30:20 probably Aug 14 13:30:22 I guess the config menus are nice to only build what you want, but not so good for automated building them all. Aug 14 13:30:39 the thing is it hasn't been a problem as of now Aug 14 13:30:54 or not a pressing one Aug 14 13:30:56 so sub menus should really only be for "uncommon" configs that only someone building from source should ever want to touch then Aug 14 13:31:30 there's a luci "tls support" that offers a choice for them, which is where I got this idea of doing it that way. Aug 14 13:31:58 luci->libraries-tls provider, Aug 14 13:32:56 there is no one-fits-all solution Aug 14 13:33:02 thanks :) Aug 14 13:34:02 think I'll have lunch and have a further thought. Aug 14 13:36:59 so, how does the optional depends work then? Aug 14 13:37:20 like, libwebsockets-example-apps, will depend on _any_ of the libwebsockets, either -openssl, -cyassl or -nossl Aug 14 13:48:24 cyrusff: is it possible to somehow replicate the build of BB rc releases? It seems that mysql modules are broken for telephony packages, but I can't replicate the issue even with the same config. I would love to debug that for final release (current trunk is also ok) Aug 14 13:51:44 here's teh OR depends I'm trying to work out: http://pastebay.net/1482467 Aug 14 14:42:50 nbd: ping Aug 14 14:42:57 nbd: i have such an idea for brcm47xx Aug 14 14:43:26 nbd: i'd like to generate image with initramfs for WNDR4500, even if user/bot selected SquashFS Aug 14 14:43:36 do you think that is acceptable? Aug 14 14:43:41 it's because WNDR4500 uses NAND Aug 14 14:43:49 is it possible to see package download stats, to help me see if it's even worth maintaining a package variant? Aug 14 15:11:10 is there any way to ignore the ssl certificate when doing feeds update -a? Aug 14 15:22:44 i think owrt's compilation can't work on slitaz Aug 14 15:25:32 dlorah_: what ssl cert complain are you getting? Aug 14 15:25:42 (don't use https sources?) Aug 14 15:25:58 just launched feeds update -a Aug 14 15:26:00 wait Aug 14 15:28:37 karlp: http://fpaste.org/125605/03009114/ Aug 14 15:29:49 well, you could probably fix it temporarly by swapping https for http in the feeds.conf, but you know, ssl cert failures are meant tobe treated seriously :) Aug 14 15:30:31 also, what's wrong with your machine that it's failing on those certs? Aug 14 15:32:03 don't know.. it's a slitaz, a very minimal distribution Aug 14 15:34:06 swapped all to http, but still get errors with luci and telephony Aug 14 15:34:11 but tthat doesn't matter Aug 14 15:34:28 i think i will simply copy my owrt folder from a properly distribution Aug 14 15:34:45 yeah, that will help. "shit's broken, let's just work around a symptom" Aug 14 15:36:14 i think it's only for slitaz, didn't get this problem with fedora Aug 14 15:48:15 if i have a makefile for a new package, how would i go about submitting it for inclusion? Aug 14 15:48:39 https://github.com/openwrt/packages/blob/master/CONTRIBUTING.md Aug 14 15:50:00 thanks Aug 14 16:05:17 Hum, I've kinda messed up my commit to the packages feed. Looks like I've lost the Aug 14 16:05:17 merge commits somewhere along the line Aug 14 16:05:22 Does this impose a problem? Aug 14 16:14:15 Naoir: as long as you didn't force commit / rewrote history Aug 14 16:15:43 Nope. I rebased locally before pushing. Guess that's where they got lost Aug 14 16:16:11 ah, yes Aug 14 16:16:12 well, that sounds good then right, why would you want merge commits in the final tree? Aug 14 16:18:13 I kinda thought that's the convention as the log is full of them :D Aug 14 16:20:54 merge pull request maybe, but not the merge upstream branch stuff? Aug 14 16:23:08 Hum looks like you're right Aug 14 16:23:52 Well, I guess everything is okay then. Thanks Aug 14 17:40:52 build #656 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/656 Aug 14 18:02:01 Hauke: hey, do you have access to #openwrt-devel log? to see my messages to nbd when you were offline? Aug 14 18:50:10 zajec: hi Aug 14 18:50:19 I do not have logs for this channel Aug 14 18:50:34 build #262 of imx6 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/262 Aug 14 19:11:39 build #667 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/667 Aug 14 20:19:00 Hauke, zajec: you around? Aug 14 20:19:21 I am around Aug 14 20:19:42 i completed the bisect that zajec suggested Aug 14 20:20:37 ok, i just now got your e-mail Aug 14 20:21:57 early printk is deactivate by default because it causes some problems on some devices Aug 14 20:22:32 DonkeyHotei: could you also remove the call to ssb_pcicore_init(&bus->pcicore); in drivers/ssb/driver_pcicore.c Aug 14 20:22:53 you want me to do both at once? Aug 14 20:23:09 this will deactivate pcmcia support Aug 14 20:23:25 yes you can do both together Aug 14 20:23:57 is that in target/linux/brcm47xx/patches? Aug 14 20:25:12 no it is in mainline kernel, you can directly exit the file in build_dir/target-mipsel_mips32_uClibc-0.9.33.2/linux-brcm47xx_legacy/linux-3.10.49/drivers/ssb/main.c Aug 14 21:19:00 Hauke: now compiling Aug 14 21:21:32 hauke r42164 trunk/target/linux/ brcm47xx/patches-3.10/144-make-reboot-more-reliable.patch brcm47xx/patches-3.14/144-make-reboot-more-reliable.patch * brcm47xx: fix brcm47xx.legacy build Aug 14 21:22:21 hauke r42165 trunk/target/linux/brcm47xx/image/Makefile * brcm47xx: image: update list of Netgear WNR* firmwares Aug 14 21:23:40 hauke r42166 branches/barrier_breaker/target/linux/brcm47xx/patches-3.10/144-make-reboot-more-reliable.patch * BB: brcm47xx: fix brcm47xx.legacy build Aug 14 21:24:18 hauke r42167 branches/barrier_breaker/target/linux/brcm47xx/image/Makefile * BB: brcm47xx: image: update list of Netgear WNR* firmwares Aug 14 23:38:30 build #625 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/625 Aug 15 00:25:43 build #637 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/637 Aug 15 01:51:40 build #702 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/702 Aug 15 01:53:03 build #63 of brcm47xx.legacy is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.legacy/builds/63 Aug 15 02:14:41 build #561 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/561 Aug 15 02:16:16 /awa **** ENDING LOGGING AT Fri Aug 15 03:00:00 2014