**** BEGIN LOGGING AT Wed Apr 09 02:59:58 2014
Apr 09 03:05:50 <builder_> build #566 of cobalt is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/566
Apr 09 04:26:30 <builder_> build #538 of uml is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/538
Apr 09 04:37:32 <builder_> build #40 of sunxi is complete: Failure [failed compile_3]  Build details are at http://buildbot.openwrt.org:8010/builders/sunxi/builds/40
Apr 09 04:54:05 <Guest25729> has anybody tried to package sysdig yet?
Apr 09 05:14:11 <builder_> build #22 of pxa is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/22
Apr 09 06:57:52 <OpenWrt-Commits> cyrus r40432 trunk/package/network/ipv6/odhcp6c/Makefile * odhcp6c: fix recovering from status where all addresses have been lost
Apr 09 07:22:05 <builder_> build #22 of brcm2708 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/22
Apr 09 07:34:02 <builder_> build #22 of omap is complete: Failure [failed shell compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/omap/builds/22
Apr 09 08:24:57 <builder_> build #22 of x86_64 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/22
Apr 09 08:24:58 * russell-- missing something obvious, just trying to add a package locally, not getting picked up by scripts/feeds install -a, despite a makefile in feeds/packages/utils
Apr 09 08:28:35 <blogic> soma: ping
Apr 09 08:42:00 <russell--> scripts/feeds update -i packages is what i was looking for, apparently
Apr 09 10:55:18 <russell--> hmm.  how to tell cmake where to find target's (not host's) kernel?
Apr 09 11:04:02 <jow_laptop> zajec: well, no updates yet
Apr 09 11:47:06 <OpenWrt-Commits> kaloz r40433 trunk/target/linux/ x86_64/Makefile x86_64/patches/011-tune_lzma_options.patch * [x86_64]: upgrade to 3.14
Apr 09 11:55:31 <OpenWrt-Commits> kaloz r40434 trunk/target/linux/x86_64/config-default * [x86_64]: enable native KVM and VMware drivers
Apr 09 11:58:57 <zajec> jow_laptop: is there any place i can read about LuCI2?
Apr 09 11:59:06 <zajec> jow_laptop: LuCI was written in LUA, right?
Apr 09 11:59:12 <jow_laptop> yes
Apr 09 11:59:22 <zajec> jow_laptop: using uci plugin for lua
Apr 09 11:59:26 <zajec> jow_laptop: what about LuCI2?
Apr 09 11:59:30 <zajec> does it still use lua?
Apr 09 11:59:33 <jow_laptop> no
Apr 09 11:59:46 <zajec> what then? ;)
Apr 09 11:59:48 <zajec> ansi c?
Apr 09 11:59:49 <jow_laptop> its almost completely javascript and html
Apr 09 12:00:14 <jow_laptop> actions on the router are triggered via rpc calls or cgi scripts for a few selected tasks like firmware image upload
Apr 09 12:00:24 <jow_laptop> the backend is rpcd, you can find it in trunk already
Apr 09 12:00:43 <zajec> jow_laptop: sec
Apr 09 12:00:53 <zajec> what is "project/luci.git" reporistory? is this old LuCI?
Apr 09 12:01:00 <zajec> old only?
Apr 09 12:01:03 <jow_laptop> rpcd is written in C and utilizes ubus and the native uci api
Apr 09 12:01:09 <zajec> without anything from LuCI2?
Apr 09 12:01:25 <jow_laptop> http://git.openwrt.org/?p=project/luci2/ui.git;a=summary
Apr 09 12:01:43 <zajec> what is that called "luci2/ui"? is this only part of LuCI2?
Apr 09 12:01:48 <jow_laptop> add it as feed, it provides a single package "luci2" which contains all ui stuff
Apr 09 12:01:50 <zajec> do i still need something extra form another repo?
Apr 09 12:01:55 <jow_laptop> no
Apr 09 12:01:58 <zajec> oh, ok
Apr 09 12:02:05 <zajec> that "ui" was a bit misleading
Apr 09 12:02:06 <zajec> thanks
Apr 09 12:02:19 <jow_laptop> well
Apr 09 12:02:32 <jow_laptop> this is because of the organization on nbd.name/gitweb.cgi
Apr 09 12:02:44 <jow_laptop> several new userspace frameworks are called luci2
Apr 09 12:02:54 <jow_laptop> or rather are grouped under the luci2 name
Apr 09 12:03:07 <zajec> ah, i can see now
Apr 09 12:03:49 <jow_laptop> the luci2 parts on nbd.name are supposed to be migrated to git.openwrt.org eventually
Apr 09 12:04:43 <zajec> do i need them for testing LuCI2?
Apr 09 12:04:55 <jow_laptop> no
Apr 09 12:05:12 <jow_laptop> you only need to add the repo I linked above to feeds.conf
Apr 09 12:05:19 <jow_laptop> then enable luci2 in menuconfig
Apr 09 12:05:31 <zajec> ok
Apr 09 12:05:37 <jow_laptop> you might need to manually select uhttpd but in theory the pacakge should depend on anything needed
Apr 09 12:05:44 <jow_laptop> luci2 only needs stuff from trunk/package
Apr 09 12:05:49 <jow_laptop> no things from extra feeds
Apr 09 12:06:15 <zajec> jow_laptop: what's the state of it?
Apr 09 12:06:18 <jow_laptop> uhttpd, uhttpd-mod-ubus, rpcd, rpcd-mod-iwinfo, libiwinfo
Apr 09 12:06:22 <zajec> jow_laptop: do you have some list of missing things?
Apr 09 12:06:37 <jow_laptop> wireless config, network config, firewall config
Apr 09 12:06:50 <jow_laptop> :)
Apr 09 12:07:36 <zajec> damn, luci2.js went huge ;)
Apr 09 12:07:59 <jow_laptop> couldn't find a good js compressor that does not depend on java
Apr 09 12:08:12 <jow_laptop> therefore I did not include js minification yet
Apr 09 12:08:37 <zajec> i mean the code in this file
Apr 09 12:08:45 <zajec> i'm not complaining about compressing ;)
Apr 09 12:09:08 <jow_laptop> yeah, though it will not get much larger
Apr 09 12:09:20 <jow_laptop> it already contains most stuff needed for the entire ui
Apr 09 12:10:32 <zajec> do you want BB to use LuCI2?
Apr 09 12:10:43 <jow_laptop> its unlikely to be ready in time
Apr 09 12:10:48 <zajec> oops :(
Apr 09 12:30:29 <OpenWrt-Commits> cyrus r40435 trunk/package/network/services/odhcpd/Makefile * odhcpd: use correct defaults for interfaces not in UCI
Apr 09 13:49:38 <Devastator> jow_laptop does jQuery produces less code than javascript? I mean, while running
Apr 09 14:04:01 <nighty-_> [   53.659341] cs5535-clockevt: Unable to set up the interrupt.
Apr 09 14:04:02 <nighty-_> [   53.665017] cs5535-clockevt: Unable to set up the MFGPT clock source
Apr 09 14:04:10 <nighty-_> how can that be ?
Apr 09 14:04:19 <nighty-_> (net5501 Soekris)
Apr 09 14:04:29 <nighty-_> running 12.09
Apr 09 14:39:25 <nighty-_> why is 5501 Soekris marked broken in Barrier Breaker ?
Apr 09 14:40:56 <KanjiMonster> likely the kernel was updated by a developer who does not have access to hardware, and "automatically" marked broken until someone confirms it works with the new kernel
Apr 09 15:21:35 <zajec> jow_laptop: why there are two rpcd projects? :| http://git.openwrt.org/?p=project/luci2/ui.git;a=tree;f=luci2/src/rpcd and http://nbd.name/gitweb.cgi?p=luci2/rpcd.git;a=tree
Apr 09 15:22:39 <jow_laptop> they're mirrors of each other
Apr 09 15:22:49 <zajec> but... the source is different
Apr 09 15:23:13 <jow_laptop> ah you mean that
Apr 09 15:23:18 <jow_laptop> its not two rpcd projects
Apr 09 15:23:25 <jow_laptop> luci2 ships its own .so plugin for rpcd
Apr 09 15:23:38 <jow_laptop> which exposes luci2 specific ubus procedures
Apr 09 15:25:06 <karlp> and this is nothing to do with nfs rcpd is it?
Apr 09 15:25:27 <jow_laptop> right
Apr 09 15:25:29 <karlp> is this possibly why nfs server doesn't work on recent builds?
Apr 09 15:25:34 <jow_laptop> no
Apr 09 15:26:06 <karlp> I haven't bisected that down, but I can't for the life of me get nfs-kernel-server working on builds form the last few months
Apr 09 15:27:45 <zajec> jow_laptop: does "ubus" command line also uses RPC? just... locally?
Apr 09 15:27:49 <jow_laptop> yes
Apr 09 15:28:14 <jow_laptop> luci2 uses uhttpd-mod-ubus which is an http-json-rpc to ubus bridge
Apr 09 15:28:25 <jow_laptop> ubus on openwrt itself is locally via unix domain sockets
Apr 09 15:29:01 * zajec starts to understand
Apr 09 15:29:38 <jow_laptop> uhttpd-mod-ubus provides a virtual url /ubus which accepts http post requests that get translated to ubus calls
Apr 09 15:29:48 <jow_laptop> the ubus replies are encoded as json and sent back as http response
Apr 09 15:30:22 <jow_laptop> uhttpd-mod-ubus in turn uses sessioning and access control provided by rpcd, so not any ubus call is simply exposed to any http client
Apr 09 15:31:22 <jow_laptop> as soon as I find more spare time I'm going to write some docs on the process
Apr 09 15:31:43 <zajec> i just wanted to do that
Apr 09 15:31:49 <zajec> but it sounds pretty complex to undesrand that in first place
Apr 09 15:32:27 <zajec> if "ubusd" exposes itself using unix socket (i guess it's called RPC interface), then what for is "rpcd"?
Apr 09 15:32:41 <karlp> the R in rpc I would presume :)
Apr 09 15:32:47 <zajec> why "uhttpd-mod-ubus" uses that magic "rpcd" instead of using unix socket directly?
Apr 09 15:33:01 <jow_laptop> rpcd is a server which simply exposes various ubus objects and procedures
Apr 09 15:33:16 <jow_laptop> ubusd itself is just the message broker, it does not actually expose procedures
Apr 09 15:33:23 <jow_laptop> those are provided by daemons connecting to ubusd
Apr 09 15:33:45 <jow_laptop> e.g. netifd publishes the network.* and wireless.* namespaces in ubus
Apr 09 15:34:04 <jow_laptop> rpcd will provide luci2.*, session.*
Apr 09 15:34:15 <jow_laptop> plus any more stuff added through plugins
Apr 09 15:35:27 <jow_laptop> luci2 -> uhttpd-mod-ubus -> ubusd
Apr 09 15:35:44 <jow_laptop> ubusd gets its data from netifd, rpcd, hostapd, ...
Apr 09 15:37:15 <zajec> does netifd communicate with ubus using unix socket?
Apr 09 15:37:23 <jow_laptop> yes
Apr 09 15:37:55 <nbd> anything that uses ubus connects to ubusd via libubus, which uses unix domain sockets
Apr 09 15:38:09 <nbd> any ubus client can register objects + methods
Apr 09 15:38:14 <nbd> which can be called from other processes
Apr 09 15:41:24 <zajec> what format messages are passed through unix socket (with ubusd)? are the some binary messages or high-level JSONs?
Apr 09 15:41:40 <jow_laptop> binary tlv
Apr 09 15:42:22 <jow_laptop> json is just uses as input/output language
Apr 09 15:42:28 <jow_laptop> could be xml too or anything else
Apr 09 15:42:32 <zajec> ok
Apr 09 15:43:38 <jow_laptop> so think of ubusd as something that simply releys messages between processes
Apr 09 15:43:56 <zajec> yup, got that :)
Apr 09 15:51:05 <zajec> ahh, i get it now
Apr 09 15:51:15 <zajec> i was a bit fouled by "rpcd"
Apr 09 15:51:25 <zajec> i assumed it exposes some extra API over HTTP or sth
Apr 09 15:51:40 <zajec> now I can see it only exposes two namespaces over UBUS
Apr 09 15:51:44 <zajec> that makes sense! :)
Apr 09 15:52:28 <zajec> hey, that's pretty simple after all! ;)
Apr 09 15:54:39 <zajec> jow_laptop: nbd: do you think we could/should describe all of that on http://wiki.openwrt.org/doc/techref/ubus wiki page?
Apr 09 15:54:45 <zajec> or is there some better place for that?
Apr 09 15:57:29 <karlp> so, ubus call log read.... how do you specify the lines parameter (from the shell?)
Apr 09 15:58:05 <karlp> ubus call log read "{'lines':10}" just returns a blank
Apr 09 16:00:19 <zajec> karlp: after "ubus call" you type a path
Apr 09 16:00:26 <zajec> your path "log" doesn't seem to be valid
Apr 09 16:03:17 <karlp> it's listed in ubus -v list
Apr 09 16:03:49 <karlp> ubus call system board works, following the same syntax, I just don't get how to pass parameters
Apr 09 16:03:59 <zajec> oh, ok
Apr 09 16:04:04 <zajec> it just wasn't documented on wiiki page
Apr 09 16:04:06 <zajec> you're right
Apr 09 16:04:50 <karlp> so like "ubus call network.wireless status" works but "ubus call network.wireless list" doesn't, though it's listed in ubus -v list as well
Apr 09 16:05:25 <karlp> because it has a "name" parameter of some sort
Apr 09 16:05:36 <zajec> karlp: your commands works for me
Apr 09 16:05:44 <zajec> ubus call log read "{'lines':1}"
Apr 09 16:05:49 <zajec> this one gives me only 1 message
Apr 09 16:05:57 <karlp> ok, I get nothing at all.
Apr 09 16:06:12 <zajec> :|
Apr 09 16:06:36 <karlp> ubus call log write "{'event':'karl wrote and event'}"
Apr 09 16:06:38 <karlp> that works.
Apr 09 16:06:43 <karlp> I can see it in logread
Apr 09 16:06:51 <karlp> as kern.emerg, which is a little high :)
Apr 09 16:06:57 <nighty-_> KanjiMonster, so that means that someone needs a net5510 soekris and net6501 donnation to better support it ?
Apr 09 16:12:09 <Pteridium> KanjiMonster: jar229 the official SMP tester in seguridadwireless forum said that, for now, the BCM6368 works very fine
Apr 09 16:13:02 <Pteridium> he goes to do more test but he is very confident
Apr 09 16:13:18 <Pteridium> tests*
Apr 09 16:16:39 <builder_> build #22 of adm8668 is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/22
Apr 09 16:30:04 <jow_laptop> zajec: yes, the ubus thing should got the wiki page
Apr 09 16:32:02 <OpenWrt-Commits> blogic r40436 trunk/target/linux/lantiq/base-files/sbin/dsl_notify.sh * lantiq: make dsl_notify.sh executable
Apr 09 16:32:11 <OpenWrt-Commits> blogic r40437 trunk/package/ network/config/ltq-vdsl-app/files/dsl_control kernel/lantiq/ltq-atm/Makefile kernel/lantiq/ltq-ptm/Makefile * lantiq: cleanup vdsl driver bringup
Apr 09 16:37:19 <OpenWrt-Commits> blogic r40438 trunk/package/network/config/ltq-vdsl-app/files/dsl_control * lantiq: vdsl init.d script should auto start
Apr 09 17:08:01 <karlp> does the rpcd+uhttd-mod-x let people use websockets by any chance?
Apr 09 17:17:13 <zajec> karlp: jow_laptop said it's listening on virtual URL for POST
Apr 09 17:17:23 <zajec> karlp: so since it's POST, it's not websocket
Apr 09 17:22:24 <builder_> build #23 of mpc83xx is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/23
Apr 09 17:39:04 <builder_> build #529 of rb532 is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/529
Apr 09 17:39:19 <builder_> build #498 of sibyte is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/498
Apr 09 17:40:17 <builder_> build #529 of ppc44x is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/529
Apr 09 17:54:52 <builder_> build #24 of bcm53xx is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/24
Apr 09 18:04:28 <karlp> zajec: yeah, just wondering/hoping if it could also do websockets :)
Apr 09 18:16:47 <zajec> jow_laptop: you said uhttpd-mod-ubus is a HTTP /bus URL bridge to the ubus
Apr 09 18:17:27 <zajec> jow_laptop: and you said that it's "rpcd" that provides luci2.* and session.*
Apr 09 18:17:52 <zajec> but according to the wiki: http://wiki.openwrt.org/doc/techref/ubus#uhttpd-mod-ubus it's uhttpd-mod-ubus  (not rpcd) that provides session.*
Apr 09 18:17:54 <zajec> so...
Apr 09 18:17:59 <zajec> is the wiki wrong about that?
Apr 09 18:56:30 <danitool> I'm getting this kernel panic http://pastebin.com/HP0iz9W3
Apr 09 18:56:48 <danitool> disabling CONFIG_AUDIT solves the problem
Apr 09 18:56:58 <danitool> any idea about what's going on?
Apr 09 19:01:48 <builder_> build #562 of lantiq is complete: Failure [failed shell compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/562
Apr 09 19:19:53 <nbd> zajec: the wiki hasn't been updated
Apr 09 19:20:04 <nbd> zajec: before it was actively used, session handling was in uhttpd
Apr 09 19:20:32 <nbd> zajec: but it was moved out when jow started using the sessions for luci
Apr 09 19:23:09 <zajec> thanks
Apr 09 19:24:04 <nbd> fyi, making good progress with the cwmp stuff, started to work on the backend
Apr 09 19:24:58 <zajec> nice :)
Apr 09 19:25:21 <zajec> i was looking at TR-098... i don't like it :|
Apr 09 19:25:50 <zajec> many things seem to complex in it
Apr 09 19:26:09 <zajec> like for example specifying encryption
Apr 09 19:26:12 <zajec> or modes
Apr 09 19:26:37 <zajec> and some things are handled in a weird way... like virtual interfaces
Apr 09 19:29:01 <nbd> yeah
Apr 09 19:29:18 <nbd> i will probably have to implement quite a bit of that though
Apr 09 19:29:36 <nbd> i'm sure it'll be manageable
Apr 09 19:29:48 <nbd> and i will be careful to keep all the madness confined within cwmp space
Apr 09 19:29:53 <nbd> so it doesn't leak into the config
Apr 09 19:30:01 <nbd> bbl
Apr 09 19:49:45 <builder_> build #508 of avr32 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/508
Apr 09 21:07:54 <builder_> build #436 of iop32x is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/436
Apr 09 21:42:22 <nighty-_> nbd, how do you manage to compile on your MAC with llvm ?
Apr 09 21:42:33 <nighty-_> well clang
Apr 10 02:23:26 <builder_> build #574 of orion is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/574
**** ENDING LOGGING AT Thu Apr 10 02:59:58 2014