**** BEGIN LOGGING AT Wed Jun 11 02:59:59 2014 Jun 11 08:44:13 cyrus r41099 packages/lang/luabitop/Makefile * luabitop: moved to github Jun 11 08:44:15 cyrus r41100 packages/lang/luaposix/ Makefile patches/100-eglibc-compat.patch * luaposix: moved to github Jun 11 08:44:16 cyrus r41101 packages/net/ (6 files in 4 dirs) * tgt: moved to github Jun 11 08:44:17 cyrus r41102 packages/libs/ libaio/patches/001-arches_from_debian_package_0.3.106-8.patch libaio/patches/002-avr32_support.patch libaio/Makefile * libaio: moved to github Jun 11 08:49:04 swalker: oh, thanks for the heads-up - I didn't notice that there is one already Jun 11 08:59:18 cyrus r41103 packages/admin/monit/ Makefile files/monit.init * monit: moved to github Jun 11 08:59:19 cyrus r41104 packages/admin/htop/Makefile * htop: moved to github Jun 11 08:59:20 cyrus r41105 packages/libs/ libftdi/patches/100-fix-x86_64-build.patch libftdi/Makefile * libftdi: moved to github Jun 11 08:59:21 cyrus r41106 packages/ multimedia/xupnpd/patches/101-root_dir_param.patch multimedia/xupnpd/Makefile multimedia/xupnpd/patches/100-default_config.patch multimedia/xupnpd/files/xupnpd.init * xupnpd: moved to github Jun 11 08:59:22 cyrus r41107 packages/net/ udpxy/files/udpxy.init udpxy/Makefile * udpxy: moved to github Jun 11 08:59:25 cyrus r41108 packages/utils/dmidecode/Makefile * dmidecode: moved to github Jun 11 08:59:28 cyrus r41109 packages/utils/flashrom/Makefile * flashrom: moved to github Jun 11 08:59:30 cyrus r41110 packages/ (6 files in 2 dirs) * pciutils: moved to github Jun 11 08:59:33 cyrus r41111 packages/ utils/screen/patches/110-do_not_add_srcdir_to_include_path.patch utils/screen/files/etc/screenrc utils/screen/patches/100-cross_compile_fix.patch utils/screen/Makefile * screen: moved to github Jun 11 09:56:12 hi all! Jun 11 09:56:39 nbd i have resended the vlandev patch splitted as we talked ;) Jun 11 10:15:13 g10h4ck: yes, i noticed Jun 11 10:15:20 i'll take care of it soon Jun 11 10:15:32 :) Jun 11 10:15:36 thanks Jun 11 10:24:07 g10h4ck: patches look good, pushed to git, will be committed to openwrt now Jun 11 10:24:19 :D Jun 11 10:24:32 so vladev will be available in openwrt trunk ? Jun 11 10:24:40 yes Jun 11 10:24:41 :D Jun 11 10:24:53 nbd r41112 trunk/package/network/config/netifd/Makefile Jun 11 10:24:53 netifd: update to latest version, fixes reload issue (#16754) and adds 802.1ad support Jun 11 10:25:04 thanks for the patches Jun 11 10:25:17 you are welcome :) Jun 11 11:30:39 nbd r41113 trunk/package/kernel/mac80211/patches/ 564-ath9k_cleanup_reset_debug.patch 565-ath9k_restart_after_nfcal_failure.patch Jun 11 11:30:39 ath9k: do a chip reset if noise floor calibration fails, should improve stability Jun 11 12:13:36 cyrus r41114 trunk/package/ network/services/uhttpd/Makefile network/services/uhttpd/files/uhttpd.config * uhttpd: also bind to IPv6 by default Jun 11 12:21:25 cyrus r41115 trunk/package/kernel/linux/modules/usb.mk * kernel: fix kernel builds for chipidea w/ usb-gadget Jun 11 12:32:33 nbd r41116 branches/attitude_adjustment/package/ (108 files in 3 dirs) * AA: mac80211: sync with trunk r41113 Jun 11 12:32:35 nbd r41117 branches/attitude_adjustment/package/ (8 files in 2 dirs) * AA: iw: sync with trunk r40801 Jun 11 13:00:47 blogic r41118 trunk/target/linux/ (8 files) * kernel: backport ubiblock support from 3.15 Jun 11 13:00:52 blogic r41119 trunk/target/linux/ (5 files) * kernel: rootfs auto-mount on ubi Jun 11 13:00:56 blogic r41120 trunk/target/linux/ (13 files) * kernel: backport the ubiblock patches from 3.14->3.10 Jun 11 13:01:00 blogic r41121 trunk/scripts/ubinize-image.sh * scripts: add ubinize-image.sh to generate ubi images Jun 11 13:01:04 blogic r41122 trunk/ (9 files in 9 dirs) * kernel: add a NAND_SUPPORT symbol Jun 11 13:01:08 blogic r41123 trunk/package/libs/libubox/Makefile * libubox: update to latest git Jun 11 13:01:11 blogic r41124 trunk/package/utils/ubi-utils/Makefile * ubi-utils: cleanup the package Jun 11 13:01:15 blogic r41125 trunk/package/base-files/files/lib/upgrade/common.sh * base-files: include more ubi* tools in sysupgrade ramdisk Jun 11 13:01:19 blogic r41126 trunk/package/base-files/files/sbin/sysupgrade * base-files: add an ugly hack for nand to sysupgrade Jun 11 13:01:22 blogic r41127 trunk/include/image.mk * include/image.mk: add a target for generating sysupgrade tar files for nand Jun 11 13:01:27 blogic r41128 trunk/package/system/fstools/Makefile * fstools: the ubi flash tool is obselete now Jun 11 13:01:34 blogic r41129 trunk/package/ system/procd/Makefile system/procd/files/nand.sh * procd: update to latest git HEAD Jun 11 13:01:38 blogic r41130 trunk/target/linux/ar71xx/image/Makefile * ar71xx: remove the explicit ubi.mtd- paramaters and rename ubiroot to ubi Jun 11 13:01:42 blogic r41131 trunk/target/linux/ ar71xx/base-files/lib/upgrade/platform.sh ar71xx/image/Makefile * ar71xx: add sysupgrade support for wndr4300 Jun 11 13:01:45 blogic r41132 trunk/target/linux/ar71xx/nand/config-default * ar71xx: fix up kernel config for the nand subtarget Jun 11 13:01:54 blogic r41133 trunk/target/linux/ar71xx/nand/target.mk * ar71xx: activate the nand sutarget Jun 11 13:01:59 blogic r41134 trunk/target/linux/ (11 files in 8 dirs) * kernel: make sure the new UBIBLOCK symbol does not break builds Jun 11 13:02:07 blogic r41135 trunk/package/base-files/files/lib/upgrade/common.sh * base-files: properly remove stderr from zcat/bzcat Jun 11 13:26:51 nice Jun 11 13:38:11 stintel: you mean the mass join/parting going on? or is there some cnversation I can't see from netsplits? Jun 11 13:38:26 karlp: no the commit spree from blogic :) Jun 11 14:01:09 http://imgur.com/sGxpMWG Jun 11 14:03:04 I know that https://dev.openwrt.org/ticket/16513#comment:2 is not a bug, but what is the recommended way of upgrading from AA to BB while keeping settings? is is simply not supported? Jun 11 14:18:02 karlp: it should work just fine Jun 11 14:25:52 karlp: did you somehow end up having /etc/init.d/boot in conffiles or some other list of to-be-preserved files? Jun 11 14:32:40 no, I haven't honestly tried it yet, I just know that backfire to AA didn't work Jun 11 14:33:00 I just saw that bug in the timeline on dev.openwrt.org Jun 11 14:37:11 jow r41136 packages/net/wshaper * wshaper: moved to github Jun 11 14:42:16 jow r41137 packages/utils/collectd * collectd: moved to github Jun 11 14:46:19 jow r41138 packages/net/vnstat * vnstat: moved to github Jun 11 14:50:18 jow r41139 packages/libs/rrdtool-1.0.x * rrdtool-1.0.x: moved to github Jun 11 15:01:05 jow r41140 packages/utils/coreutils * coreutils: moved to github Jun 11 15:03:38 jow r41141 packages/net/tinyproxy * tinyproxy: moved to github Jun 11 15:04:32 wow openwrt is moving to github, will it fully converted from svn to git at github Jun 11 15:04:44 ausjke: no, packages are moving to github Jun 11 15:04:48 openwrt core stays Jun 11 15:05:32 yeah noticed feeds.conf.default updates, cool Jun 11 15:13:56 isn't rrdtool 1.0.x strictly speaking against the rules of github packages since it is obsoleted by newer versions ? Jun 11 15:15:38 its in active use by collectd and luci-statistics Jun 11 15:15:46 the newer version has four times larger Jun 11 15:15:49 *is Jun 11 15:24:29 nbd: sorry to say my hangs continue. Note I have my network currently configured to have poor signal strength, and my family of 4 stresses it... Jun 11 15:24:41 jg: what openwrt rev? Jun 11 15:24:46 i committed some more fixes Jun 11 15:26:30 nbd: I'll ask dtaht to make it easier to figure out... Jun 11 15:28:05 he should make sure the openwrt base rev stays in /etc/banner Jun 11 15:28:08 not sure this will help, but when I login to it, it says BARRIER BREAKER (3.10.40-6, r40862) Jun 11 15:28:11 yes. Jun 11 15:28:16 ah, ok Jun 11 15:28:18 so that one's too old Jun 11 15:28:42 ok. Jun 11 15:30:06 Ah, I see he's also built a 3.10.41-1 system on 5/31; that one snuck by me. Jun 11 15:30:42 dunno if he announced it. Jun 11 15:36:14 cyrus r41142 trunk/package/system/procd/Makefile * procd: fix overflow in hotplug socket on certain targets Jun 11 15:38:06 nbd, has the sta memleak fix been backported to AA ? Jun 11 15:38:12 yes Jun 11 15:38:18 cool Jun 11 15:38:30 I was too lazy to look thru the patches :) Jun 11 15:40:12 it's a full backport of the trunk package Jun 11 15:40:22 nice :) Jun 11 15:40:38 thanks a lot for that Jun 11 15:40:52 you're welcome Jun 11 15:45:05 jow r41143 trunk/package/network/utils/curl packages/libs/curl * curl: move to core packages Jun 11 16:14:27 if I call luci-reload network, and ucitrack network has "list affects dhcp" why isn't dnsmasq restarted? Jun 11 16:14:40 is luci-reload being clever and not-restarting anything that "hasn't changed" ? Jun 11 16:15:00 I was hoping to use it's dependencies tree in ucitrack to trigger all the changes I knew had happened. Jun 11 16:15:23 ucitrack is obsolete Jun 11 16:15:41 in BB, procd triggers is the way to go Jun 11 16:16:08 and there's a reload_config script that you can call to apply the config Jun 11 16:16:16 for everything that has been converted to procd already Jun 11 16:16:31 yeah, not all of us are thebleeding edge every day :| Jun 11 16:16:44 and there are packages being put in the new packages repo that weren't converted to procd anyway. Jun 11 16:17:06 and there's _some_ documentation on procd init scripts on the wiki, but I haven't seen _any_ yet about how to use these triggers, wonderful as they sound Jun 11 16:17:40 so, is the answer then, "ignore ucitrack, manually work out dependencies and call /etc/init/x,y,z stop,start by hand? Jun 11 16:17:53 are you using AA or some older BB? Jun 11 16:24:35 I'm on AA at present, BB has been churning sufficiently and our hardware is sufficiently old atheros I didn't see any great motivation to move forward yet. Jun 11 16:25:05 backfire -> AA was more pain than I was expecting, so while I've got some prototype new hardware on BB, it's not using any BB only features. Jun 11 16:26:45 http://wiki.openwrt.org/inbox/procd-init-scripts?s[]=procd is all I've managed to scrape together so far, I'd really love to see some more written by someone who knows what's the desired expected behaviour and how you _should_ use the procd features, rather than having to grep the package feeds looking for procd, looking for examples of a package that _might_ do the sort of thing you want to do. Jun 11 16:27:33 now, ok, fair point, ucitrack isn't exactly on the wiki either :) Jun 11 16:28:00 but it's what's been painstakingly learnt over time, Jun 11 16:28:35 so here's how dnsmasq sets itself up to reload after the dhcp config changes: Jun 11 16:28:38 service_triggers() Jun 11 16:28:41 { procd_add_reload_trigger "dhcp" Jun 11 16:28:43 } Jun 11 16:28:54 the interface based triggers have been added recently Jun 11 16:29:01 in the last few days Jun 11 16:29:05 let me find an example for that... Jun 11 16:30:03 ok, interview time :) I'll try and add more to that "procd init script" page, and you answer some questions and show the good examples for me :) Jun 11 16:30:13 sounds good Jun 11 16:30:20 first up, can procd work at all with processes that background themselves? Jun 11 16:30:42 (I _really_ _really_think you authors should write the first round documentation yourselves though, really) Jun 11 16:31:22 i agree Jun 11 16:31:39 it's hard to find the time for it though Jun 11 16:32:02 harder still for people you expect to use it :) Jun 11 16:32:25 procd and processes that background themselves is still untested Jun 11 16:32:30 even just a "procd is how you should handle it, look at package x for an example that is "good"" Jun 11 16:32:35 so far we've made everything stay in foreground Jun 11 16:32:51 for network related stuff, check out r41011 (git commit 363e2707a931e181d686e57e969c07f4ad73bedb) Jun 11 16:32:59 that's going to need some work, lots of processes out there only run in the foreground in some sort of "debug" mode. Jun 11 16:33:06 it's an example that we will apply to dnsmasq as well Jun 11 16:33:24 because when you're dealing with services that need network interface based stuff, technically it doesn't just depend on the config anymore Jun 11 16:33:29 it also depends on runtime state Jun 11 16:33:51 so the way we're dealing with this is whenever an interface is brought up, the trigger issues a reload to the service if it cares about interfaces Jun 11 16:34:10 because some interfaces are brought up with dynamicall assigned devices Jun 11 16:34:13 (e.g. wireless stuff) Jun 11 16:34:54 so will "uci commit" be the trigger? and what's a good example of triggers on the files? Jun 11 16:35:21 there's two parts to this Jun 11 16:35:45 simply using the service_triggers function that i pasted Jun 11 16:36:08 that reloads the init script when the config is changed and 'reload_config' is called Jun 11 16:36:17 so it's not triggered by uci commit Jun 11 16:36:26 reload_config is a very simple script Jun 11 16:36:31 it simply tracks md5sums of config files Jun 11 16:36:43 and when one of them changes, it emits a procd trigger Jun 11 16:36:55 the idea is to keep config state tracking to an absolute minimum Jun 11 16:37:11 since there's no way for it to cover all cases anyway Jun 11 16:37:34 so config triggers come from that script Jun 11 16:37:43 and the second part (network interface related stuff) comes from netifd Jun 11 16:38:14 procd subscribes to the 'network.interface' object Jun 11 16:38:24 on which netifd emits events when interfaces are brought up/down Jun 11 16:39:25 so when you build your stuff based on BB, you only have to make sure that the init scripts tell procd what kind of events they care about, and then you always call reload_config whenever you want to apply the uci changes that you made Jun 11 16:39:30 regardless of whether you committed them or not Jun 11 16:39:41 sometimes it's useful to test uncommitted changes Jun 11 16:39:50 does that make sense? Jun 11 16:41:24 sorry, reading, got interrupted here Jun 11 16:45:27 so you run "reload_config" when you've finished whatever string of edits you like then? Jun 11 16:45:49 yes Jun 11 16:46:13 and "reloads" the init script, that calls what exactly? is there a way to handle that without a stop/start? what hooks does one declare in the init script to get a reload hook? Jun 11 16:49:36 that's handled by some procd magic: Jun 11 16:49:45 so the init script doesn't start/stop services itself Jun 11 16:49:51 it tells procd how it's started Jun 11 16:49:59 and what kinds of things the service cares about Jun 11 16:50:01 config files Jun 11 16:50:03 interfaces Jun 11 16:50:05 command line Jun 11 16:50:22 so when procd receives another request to start a service, it simply compares the new data with the old data Jun 11 16:50:29 to figure out whether something relevant actually changed Jun 11 16:50:47 reload() is basically an alias to start Jun 11 16:50:48 in the init script Jun 11 16:51:11 right, I know, it always was, but you could still provide your own reload if you wished, is that still possible? Jun 11 16:51:15 yes Jun 11 16:51:56 you can overwrite any handler function if you want to Jun 11 16:52:39 ok, so in the past, I had a pid file, and I could override reload_service to send a signal to the service via the pidfile... Jun 11 16:53:40 the parameter to "procd_add_reload_trigger" is always a filename in /etc/config/ ? Jun 11 16:54:19 yes Jun 11 16:54:19 hangong, no there's procd_add_config_trigger too Jun 11 16:54:27 let me check Jun 11 16:54:53 busyboxs/ntpd has a "procd_add_reload_trigger system" which it then validates, that's where the time servers are, Jun 11 16:55:19 procd_add_reload_trigger is a wrapper for procd_add_config_trigger Jun 11 16:55:40 triggers allow you to specify a command that is run when the trigger fires Jun 11 16:55:53 procd_add_reload_trigger passes /etc/init.d/ reload as command Jun 11 16:56:46 so you can watch multiple uci config files if you use open/add_config/close manually, like in netifds script Jun 11 16:56:48 so right now there doesn't appear to be a procd helper function to send a signal to its process Jun 11 16:57:08 in service_triggers you simply add multiple procd_add_reload_trigger lines Jun 11 16:57:24 so netifds is a "bad" example then ;) Jun 11 16:58:16 actually, that is more efficient than what i wrote Jun 11 16:58:25 but even better would be procd_add_reload_trigger taking multiple arguments Jun 11 16:58:42 i think i'll implement that Jun 11 17:02:57 nbd r41144 trunk/include/cmake.mk * include/cmake.mk: fix indentation Jun 11 17:03:00 nbd r41145 trunk/target/linux/ (7 files in 7 dirs) * ramips: fix rt2x00 handling of default packages for subtargets Jun 11 17:07:28 nbd r41146 trunk/package/system/procd/files/procd.sh * procd: allow passing multiple config files to procd_add_reload_trigger Jun 11 17:07:31 nbd r41147 trunk/package/network/config/netifd/files/etc/init.d/network * netifd: use procd_add_reload_trigger Jun 11 17:07:36 karlp: there, better now ;) Jun 11 17:09:33 so, service_reload? that's provided in base-files/files/lib/functions/service.sh, which doesn't seem to know much about procd? how do you provide your own reload implementation at all? Jun 11 17:10:21 you simply define reload() Jun 11 17:10:36 and make it call reload_service first Jun 11 17:12:42 "simply" :) base-files/files/etc/rc.common isn't exactly straight forward... Jun 11 17:12:50 what does reload_service first do then? Jun 11 17:19:52 is the "network" in procd_add_network_trigger "eth0" style, or "lan" style? Jun 11 17:20:18 the only users so far is igmpproxy, whicih uses procd_add_interface_trigger manually Jun 11 17:21:47 and procd_add_interface uses interface.*, my (admittedly not brand new) BB shows ubus has network.interface.* how's that getting resolved? Jun 11 17:22:00 (or is my BB board here just too out of date) Jun 11 17:22:23 r40831? Jun 11 17:22:38 karlp: the netifd init script does this: Jun 11 17:22:39 procd_set_param watch network.interface Jun 11 17:22:50 this causes procd to subscribe to the network.interface object Jun 11 17:22:54 and treat incoming events as triggers Jun 11 17:23:20 netifd events have interface.* as id Jun 11 17:23:37 i mean interface. Jun 11 17:24:47 and I can list those names with? Jun 11 17:25:04 there's no list Jun 11 17:25:15 is it the names that ubus list shows? (ie, lan, wan, loopback) or what ifconfig shows? eth0, lo0? Jun 11 17:25:19 no Jun 11 17:25:33 any object that emits events can choose the id Jun 11 17:25:38 for events Jun 11 17:25:58 I don't have igmpproxy's config here, so igmpproxy uses procd_add_interface trigger directly, but I don't know what the parameter is passed, Jun 11 17:26:40 and procd.sh has both procd_add_interface_trigger, which seems to be lower level, and procd_add_network_trigger, which seems to be higher level, but I can't find examples of what parameters it takes Jun 11 17:27:58 this stuff is still very fresh, i haven't properly reviewed it yet Jun 11 17:28:03 i'll take a look Jun 11 17:29:40 the shell helper functions there are a bit inconsistent Jun 11 17:29:52 i'll sort this out with blogic when he's back Jun 11 17:30:06 ok, moving on then :) there's no helper function for sending a signal, is there a way of grabbing the pid from somewhere and doing it manually? Jun 11 17:30:25 or is that just, "should be available, will be soon(tm)" ? Jun 11 17:31:35 it's part of the data that you get when you run 'ubus call service list' Jun 11 17:31:45 but i want to add the send-signal helper soon Jun 11 17:31:59 because that's quite a bit of json data to crunch for such a simple thing Jun 11 17:34:01 I believe I've captured most of what you've said here now: http://wiki.openwrt.org/inbox/procd-init-scripts but you should review it. Jun 11 17:34:27 anything being packaged in packages2(github) now should not be creating ucitrack entries then right? Jun 11 17:35:29 right Jun 11 17:35:43 what you've written looks good Jun 11 17:35:50 didn't see any obvious errors in there Jun 11 17:35:55 thanks Jun 11 17:40:29 <5EXAA8SU8> build #561 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/561 Jun 11 17:40:48 <5EXAA8SU8> build #592 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/592 Jun 11 17:42:24 <5EXAA8SU8> build #592 of rb532 is complete: Failure [failed shell shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/592 Jun 11 17:45:31 <5EXAA8SU8> build #499 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/499 Jun 11 17:50:17 nbd r41148 trunk/package/system/procd/Makefile * procd: update to the latest version, fixes static linking of upgraded Jun 11 18:28:06 rage incoming: https://github.com/openwrt/packages/pull/10#issuecomment-45777800 Jun 11 18:41:14 <5EXAA8SU8> build #563 of xburst is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/563 Jun 11 18:44:58 <5EXAA8SU8> build #660 of at91 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/660 Jun 11 18:47:06 <5EXAA8SU8> build #544 of kirkwood is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/544 Jun 11 18:59:16 <5EXAA8SU8> build #632 of cobalt is complete: Failure [failed shell compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/632 Jun 11 19:01:12 <5EXAA8SU8> build #638 of orion is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/638 Jun 11 19:03:32 <5EXAA8SU8> build #527 of au1000 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/527 Jun 11 19:06:02 <5EXAA8SU8> build #627 of lantiq is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/627 Jun 11 19:06:19 doesn't look ilke rage anymore, just the (some what expected) surprise factor of feeds.conf.default being updated in a full step, rather than adding both, at least for a little while. Jun 11 19:06:39 <5EXAA8SU8> build #192 of mxs is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mxs/builds/192 Jun 11 19:07:56 <5EXAA8SU8> build #550 of ixp4xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ixp4xx/builds/550 Jun 11 19:10:06 <5EXAA8SU8> build #198 of imx6 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/198 Jun 11 19:12:43 <5EXAA8SU8> build #572 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/572 Jun 11 19:13:25 <5EXAA8SU8> build #603 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/603 Jun 11 19:14:39 <5EXAA8SU8> build #498 of ep93xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/498 Jun 11 20:19:20 <5EXAA8SU8> build #668 of brcm63xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/668 Jun 11 20:20:01 <5EXAA8SU8> build #487 of mcs814x is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mcs814x/builds/487 Jun 11 20:26:17 <5EXAA8SU8> build #87 of pxa is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/87 Jun 11 21:10:51 florian r41149 trunk/feeds.conf.default * feeds.conf.default: restore the old package feed temporarily Jun 11 21:29:00 <5EXAA8SU8> build #88 of x86_64 is complete: Failure [failed compile_1] Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/88 Jun 11 21:45:56 <5EXAA8SU8> build #87 of adm8668 is complete: Failure [failed shell compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/87 Jun 11 21:47:13 <5EXAA8SU8> build #86 of omap is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/omap/builds/86 Jun 11 22:24:55 <5EXAA8SU8> build #88 of mpc83xx is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/88 Jun 11 22:29:09 recursive dependency detected on trunk: http://pastebin.com/n3J9qA4M Jun 11 22:36:38 <5EXAA8SU8> build #87 of brcm2708 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/87 Jun 11 23:05:52 why is CONFIG_FEATURE_drawing-backend_DirectFB=y by default? how can I turn this off as I have no graphic UI at all? Jun 11 23:41:37 luka r41150 trunk/ (21 files in 3 dirs) * kernel: update 3.14 to 3.14.7 Jun 12 02:45:37 <5EXAA8SU8> build #87 of cns3xxx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cns3xxx/builds/87 **** ENDING LOGGING AT Thu Jun 12 02:59:59 2014