**** BEGIN LOGGING AT Wed Mar 12 02:59:58 2014 Mar 12 04:18:39 build #500 of rb532 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/500 Mar 12 04:19:36 build #500 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/500 Mar 12 04:20:50 build #468 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/468 Mar 12 04:55:38 build #407 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/407 Mar 12 07:24:22 build #478 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/478 Mar 12 07:41:46 good morning Mar 12 08:57:39 build #542 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/542 Mar 12 09:58:51 nbd r39882 packages/utils/mc/Makefile * mc: disable the use of libssh2 Mar 12 09:58:52 nbd r39883 packages/net/hsflowd/Makefile * hsflowd: disable mips16 support Mar 12 09:58:53 nbd r39884 packages/libs/lttng-ust/Makefile * lttng-ust: disable mips16 support Mar 12 10:01:15 nbd r39885 trunk/package/libs/openssl/Makefile * openssl: Fix x86_64 build on some 64bit host systems Mar 12 10:01:19 nbd r39886 trunk/package/kernel/mac80211/patches/300-pending_work.patch * ath9k: merge an aggregation related fix Mar 12 10:44:51 nbd r39887 packages/devel/lttng-tools/Makefile * lttng-tools: disable mips16 support Mar 12 10:47:54 nbd r39888 trunk/tools/ Makefile flex/patches flex/patches/100-remove_no_undefined.patch * flex: fix build error due to undefined library symbols Mar 12 11:21:38 nbd r39889 trunk/package/ boot/grub2/Makefile boot/grub2/patches/100-grub_setup_root.patch boot/grub2/patches/200-grub_musl.patch boot/grub2/patches/200-fix-gets-removal.patch * grub2: update to 2.02-beta2, fixes mac os x 10.9 support (and many other things) Mar 12 12:42:41 I am noticing more and more packages have been updated to include this "PKG_USE_MIPS16:=0" line. Is this a mandatory? Mar 12 12:44:54 Mazilo: the upgrade to gcc 4.8 added back mips16 support - if a package breaks due to assembly crap-up when built with mips16 enabled, that line fixes the package Mar 12 12:46:02 nbd: OK and thank you. Mar 12 12:52:59 juhosg r39890 trunk/target/linux/ar71xx/files/arch/mips/ath79/dev-nfc.c * ar71xx: ar934x_nfc: fix memory initialization bug Mar 12 12:53:02 juhosg r39891 trunk/target/linux/ar71xx/files/arch/mips/ath79/mach-rb95x.c * ar71xx: rb95x: use correct SPI flash address Mar 12 13:38:04 kaloz r39892 trunk/target/linux/cns3xxx/files/drivers/usb/ (8 files) * [cns3xxx]: various dwc (OTG) driver fixups Mar 12 13:39:29 kaloz r39893 trunk/target/linux/ixp4xx/patches-3.10/300-avila_support.patch * [ixp4xx]: add GW2375 support Mar 12 13:40:36 kaloz r39894 trunk/target/linux/cns3xxx/files/arch/arm/mach-cns3xxx/laguna.c * [cns3xxx]: Add GW2385 support Mar 12 14:26:11 build #527 of ramips is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/527 Mar 12 16:08:45 My testing with qos-scripts in trunk is that the qos_Default chain doesn't work with IFB (aka inbound wan packets) Mar 12 16:09:30 This is because packets hit IFB before netfilter, and therefore none of the marks are restored Mar 12 16:09:36 Am I in the ballpark? Mar 12 16:10:04 right. the only marking that works on inbound packets is if the connection is marked Mar 12 16:10:40 Whew, that is a relief. Mar 12 16:11:02 There is a patch here, but its not correct if I understand correctly: https://dev.openwrt.org/ticket/9620 Mar 12 16:11:31 because the rules in qos_Default should not always apply to connections Mar 12 16:12:16 am I missing something, or does IFB kind of suck Mar 12 16:13:04 well not suck so much, but seems unfinished perhaps Mar 12 16:13:28 right, the patch is not correct because it breaks normal classifiers in cases where they need more than one packet to match (i.e. content matching) Mar 12 16:13:52 once i find the time for it (maybe in a few months), i plan on rewriting the qos stuff Mar 12 16:14:03 and move more logic and matching to tc instead of netfilter Mar 12 16:15:04 It seems a shame we cant use netfilter for matching anymore, seems like re-inventing the wheel to me Mar 12 16:16:15 well, for per-connection stuff it works Mar 12 16:16:22 when not using it for default rules Mar 12 16:16:48 since marked connections get their connmark restored before qos Mar 12 16:16:51 for ingress Mar 12 16:17:12 yeah, but thats because you wrote a tc module for it right? Mar 12 16:17:18 otherwise that wouldn't work Mar 12 16:17:25 right Mar 12 16:17:36 netprince: yes, ifb is 50% of a replacement to imq Mar 12 16:17:57 but its the favorite of the kernel guys so we need to use it Mar 12 16:18:02 imq handled this at the cost of *extremely* ugly and intrusive hacks in the core network stack Mar 12 16:18:19 that were causing some ugly side effects as well Mar 12 16:18:51 i mainly switched to ifb because the maintenance costs of imq were just becoming too high Mar 12 16:19:46 also imq collides with some other stuff like nfqueue Mar 12 16:19:57 its a bit nasty Mar 12 16:20:16 doesn't change the fact that it supported stuff that was now sweeped under the carpet with the intriduction of ifb Mar 12 16:20:23 right Mar 12 16:20:45 On a positive note, fq_codel seems to work well Mar 12 16:22:14 but i guess in the long run tc-filters and nftables and so on will all use the same bpf-vm for matching Mar 12 16:22:25 if they don't already do this Mar 12 17:38:49 build #408 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/408 Mar 12 17:38:51 build #501 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/501 Mar 12 17:39:00 build #469 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/469 Mar 12 17:39:38 build #501 of rb532 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/501 Mar 12 19:31:58 build #479 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/479 Mar 12 19:33:15 blogic r39895 trunk/ (20 files in 12 dirs) * fstools: add the new fstools package Mar 12 19:33:24 blogic r39896 trunk/target/linux/ramips/patches-3.10/999-clk.patch * ralink: add missing clk_set_rate() Mar 12 19:33:32 blogic r39897 trunk/package/system/ubus/Makefile * ubus: update to latest git HEAD Mar 12 19:33:41 blogic r39898 trunk/package/ system/ubox/files/log.init system/ubox/Makefile * ubox: update to latest git head Mar 12 19:53:36 nbd r39899 trunk/package/Makefile Mar 12 19:53:36 build: fix installation of default-variant packages when the source package has multiple variants (#15237) Mar 12 21:33:19 florian r39900 packages/libs/xmlrpc-c/Makefile * xmlrpc-c: use new repository Mar 12 21:33:21 florian r39901 packages/net/mini_snmpd/ patches/104-memset_fix.patch Makefile * mini_snmpd: fix the size argument to memset (#13764) Mar 12 22:09:23 florian r39902 trunk/target/linux/ generic/patches-3.8/002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch generic/patches-3.8/001-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch generic/patches-3.3/001-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch generic/patches-3.3/002-ARM-7668-1-fix-memset-related-crashes-caused-by Mar 12 22:09:23 OpenWrt-rece.patch * kernel: backport ARM memset fixes for GCC 4.7/4.8 Mar 12 22:09:34 florian r39903 trunk/package/kernel/gpio-button-hotplug/Makefile * gpio-button-hotplug: don't build for 3.3 kernels Mar 12 23:16:50 hauke r39904 trunk/target/linux/generic/config-3.14 * kernel: add missing config options for kernel 3.14 Mar 12 23:26:10 blogic: is mdnsd's repo url (git://nbd.name/mdnsd.git) correct? Mar 12 23:29:05 hauke r39905 trunk/ (62 files in 3 dirs) * brcm47xx: add initial support for kernel 3.14 Mar 12 23:30:05 hauke r39906 trunk/target/linux/generic/patches-3.14/990-gpio_wdt.patch * kernel: fix 990-gpio_wdt.patch patch Mar 12 23:36:24 hauke r39907 trunk/package/kernel/linux/modules/other.mk * kernel: add missing zram option Mar 12 23:50:53 swalker: yes Mar 12 23:50:57 did i push the package ? Mar 12 23:51:41 blogic: yes, with the fstools addition Mar 12 23:51:51 lol Mar 12 23:52:27 swalker: yes, correct but i will mark it broken now Mar 12 23:52:32 the code is not ready yet Mar 12 23:53:08 swalker: i will remove the package Mar 12 23:53:11 ignore it Mar 12 23:53:11 git://git.openwrt.org/project/mdnsd.git is actually available whereas git://nbd.name/mdnsd.git isn't afaict Mar 12 23:53:18 jaja Mar 12 23:53:20 its broken **** ENDING LOGGING AT Thu Mar 13 02:59:58 2014