**** BEGIN LOGGING AT Tue Sep 16 02:59:58 2014 Sep 16 03:56:30 build #657 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/657 Sep 16 04:58:02 build #669 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/669 Sep 16 07:51:29 build #593 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/593 Sep 16 08:28:37 build #760 of at91 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/760 Sep 16 08:55:23 build #88 of brcm47xx.mips74k is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.mips74k/builds/88 Sep 16 11:56:30 blogic r42570 trunk/package/system/fstools/Makefile * fs-tools: dont use mips16 Sep 16 11:56:51 blogic r42571 trunk/ include/package-ipkg.mk package/Makefile target/imagebuilder/files/Makefile package/base-files/files/lib/functions.sh * postinst trigger: the new postinst trigger broke IB Sep 16 12:27:07 blogic r42572 trunk/package/system/ubox/Makefile * ubox: update to latest git Sep 16 13:31:52 cyrus r42573 trunk/package/libs/ libmnl/patches/100-upstream-fixes.patch libmnl/patches libmnl/Makefile * libmnl: add some upstream fixes Sep 16 14:17:10 has anyone played with routerboard routers? are they any good? Sep 16 14:17:28 oops, probably not the right channel to ask this.. Sep 16 14:53:33 blogic r42574 trunk/target/linux/lantiq/dts/ TDW8970.dts VG3503J.dtsi * lantiq: move the lantiq,switch property into the interface node Sep 16 14:53:38 blogic r42575 trunk/target/linux/lantiq/base-files/etc/init.d/dsl_fs * Lantiq: fix build warning from dsl_fs script Sep 16 15:31:17 gah, tried separating js and htm views in luci, now of course $.getJSON("<%=luci.dispatcher.build_url('home/plugins/list')%>" doesn't work :) Sep 16 15:54:26 build #541 of octeon is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/541 Sep 16 15:56:15 felix: how much memory a scan should take with ath5k ? Sep 16 15:56:28 nbd, how much memory a scan should take with ath5k ? Sep 16 16:46:54 nbd r42576 trunk/tools/Makefile * tools: add missing ldconfig make dependency on the staging area Sep 16 17:02:06 nbd, my scan on ath5k at startup time take 10MB of RAM Sep 16 17:02:19 a simple scan on the command line? Sep 16 17:02:26 no Sep 16 17:02:36 startup scan for WDS STA Sep 16 17:02:57 boot time Sep 16 17:03:09 it crashes the system with OOM sometimes Sep 16 17:03:17 and most times it can not even OOM Sep 16 17:03:36 I tried it on both AA and Trunk Sep 16 17:03:58 and on Abicom Freedom board and soekris 5501 Sep 16 17:04:24 one is an Soc the other is using an ath5k Mini PCI Sep 16 17:04:52 the Abicom Freedom board has 32Mb of RAM Sep 16 17:05:10 I also simulated 32Mb of RAM on the 5501 and also on a 4801 soekris Sep 16 17:05:13 same result Sep 16 17:05:17 crash system dead Sep 16 17:05:24 very hard to debug Sep 16 17:05:39 if using kgdb then system slows down Sep 16 17:05:42 behaviour changes Sep 16 17:06:12 crash do not happen because boot time script release memory before scan takes it up Sep 16 17:06:35 s/script/scripts/ Sep 16 17:06:50 ath5k is software scan Sep 16 17:07:04 this does not happen in ath9k because it is hardware scan Sep 16 17:07:21 so I think something is wrong in software scan code Sep 16 17:07:42 iw dev wlan0 scan Sep 16 17:07:46 is run at boot time Sep 16 17:07:49 and triggers it Sep 16 17:08:35 ath9k does not use hardware scan either Sep 16 17:08:43 oh Sep 16 17:09:35 uhmm I thought ath9k had hardware scan Sep 16 17:11:25 i wonder where the memory it eats ends up Sep 16 17:11:47 no idea Sep 16 17:12:00 I can't even get a log message or a trace Sep 16 17:15:20 it looks like when the scan finishes the memory gets returned Sep 16 17:15:45 I say that because with boards with more memory that is what happens Sep 16 17:16:19 but 10MB sounds like a lot of RAM for scan Sep 16 17:18:29 so on a soekris with mem=32M it crashes Sep 16 17:18:47 the soekris with normal mem (256MB) does not crash Sep 16 18:35:17 blogic r42577 branches/barrier_breaker/package/network/utils/ comgt/files/directip.sh comgt/Makefile comgt/files/directip.gcom comgt/files/directip-stop.gcom * comgt: add directip proto Sep 16 18:35:34 blogic r42578 branches/ barrier_breaker/target/linux/lantiq/base-files/etc/uci-defaults/02_network barrier_breaker/target/linux/lantiq/patches-3.10/0205-lantiq-xrx200-switch.patch barrier_breaker/target/linux/lantiq/dts/VG3503J.dtsi barrier_breaker/target/linux/lantiq/dts/TDW8970.dts * lantiq: fix switch init on vr9 SoC (3.10) Sep 16 18:35:36 blogic r42579 branches/barrier_breaker/package/system/fstools/Makefile * fs-tools: dont use mips16 Sep 16 18:35:38 blogic r42580 branches/barrier_breaker/target/linux/lantiq/base-files/etc/init.d/dsl_fs * Lantiq: fix build warning from dsl_fs script Sep 16 18:35:41 blogic r42581 branches/barrier_breaker/package/system/procd/Makefile * procd: update to latest git HEAD Sep 16 18:35:48 blogic r42582 branches/barrier_breaker/package/system/ubox/Makefile * ubox: update to latest git HEAD Sep 16 21:39:57 hauke r42583 trunk/target/linux/ generic/config-3.14 generic/config-3.13 generic/config-3.10 * kernel: add missing config option Sep 17 00:25:57 build #689 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/689 **** ENDING LOGGING AT Wed Sep 17 02:59:58 2014