**** BEGIN LOGGING AT Thu Dec 10 02:59:59 2015 Dec 10 03:02:10 build #169 of brcm63xx is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/169 Dec 10 04:33:02 build #168 of ramips.rt3883 is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.rt3883/builds/168 Dec 10 05:55:19 build #169 of ipq806x is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/ipq806x/builds/169 Dec 10 06:09:48 build #170 of ramips.mt7621 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.mt7621/builds/170 Dec 10 06:32:43 build #167 of realview is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/realview/builds/167 Dec 10 07:42:18 how well supported is the Qualcomm-Atheros QCA9880 ? Dec 10 07:42:34 COMPEX WLE900VX for exampe Dec 10 07:42:43 oops wrong channel maybe Dec 10 07:44:41 build #171 of bcm53xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/171 Dec 10 07:48:50 build #164 of malta is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/malta/builds/164 Dec 10 08:03:15 morning guys Dec 10 08:03:28 anyone knows where Kaloz roams these days? Dec 10 09:54:46 i am looking in the docs how to build packages, for a way to avoid conflicts between packages that cannot be build together. Isnt there a variable like PACKAGE_CONFLICT or something ? Cant seem to find the info Dec 10 10:02:45 no Dec 10 10:04:42 so if we want to make sure that when selecting one package, another will be 'greyed out' or unable to select; how can we do it ? Dec 10 10:05:36 you can use CONFLCITS, that will prevent the package from being <*> if its conflict counterpart is <*> already Dec 10 10:06:31 exactly what i want but have not seen that variable in the wiki. where can i find more info ? Dec 10 10:06:39 grep -r CONFLCITS include/ Dec 10 10:07:49 nothing Dec 10 10:07:56 yeah, typo Dec 10 10:08:05 s/CONFLCITS/CONFLICTS/ Dec 10 10:24:05 how about a variable to load/build packages as modules ? Way easier to specify all modular packages in a Makefile rather than having to go selecting one by one. Is there such thing ? Dec 10 10:27:27 ?? Dec 10 10:27:47 you mean CONFIG_ALL ? Dec 10 10:28:08 also any package can depend on "m", this will force it to be non-builtin Dec 10 10:28:16 but no idea what you plan to do Dec 10 10:37:29 i mean lets say i want to build (m) 50 extra packages only. Dec 10 10:38:28 and want to have them specified in a makefile just like we do for those that are built in (*) +package_name Dec 10 10:41:08 create a dummy package that has them as +depends, then make the dummy package depend on m; that way the =y selection will we forced back to =m and hopefully propagated to the dependencies Dec 10 10:53:04 what is the feedback about hardened openwrt ? does it work or breaks stuff in the builds ? Dec 10 10:54:15 HeXiLeD: it is usually enabled if it doesn't break stuff Dec 10 10:54:25 HeXiLeD: most of the stuff is enabled by default now - you can grep through packages Makefiles to see which packages deactivate it - afaik most of the time its old-er software Dec 10 10:55:08 because fedora and afaik debian now try to get those security patches to the upstream directly Dec 10 11:06:50 hi all! soo, is there a way to tell validate data that an option is required? Dec 10 11:07:20 any obvious mistakes in https://bpaste.net/show/58c672cb0fe6 that I am overlooking ? the .ipk is empty Dec 10 11:07:50 build_dir/target-mips64_octeon_64_glibc-2.21/linux-octeon/linux-4.3/crypto/drbg.ko exists Dec 10 11:09:22 stintel: the KCONFIG is missing the CONFIG_ prefix Dec 10 11:09:44 when using `validate_data 'option:datatype:default'`, if an option is missing and has no default it just gets unset Dec 10 11:09:49 but the validation works Dec 10 11:10:01 KanjiMonster: hah, thanks Dec 10 11:23:42 so, since linux 4.2, DRBG is now the default crypto API RNG, replacing krng Dec 10 11:24:25 and krng is removed, so openwrt with kernel >=4.2 lacks a crypto API RNG (no stdrng in /proc/crypto) Dec 10 11:25:20 the jitterentropy_rng module also doesn't offer the stdrng Dec 10 11:26:25 because of this, ipsec is broken Dec 10 11:26:35 possibly other things requiring stdrng too Dec 10 11:28:56 http://nbd.name/drbg.patch Dec 10 11:29:12 please test this Dec 10 11:29:39 awesome :) Dec 10 11:34:51 nbd: that doesn't work, but let me come up with something Dec 10 11:34:59 based on your suggestion Dec 10 11:50:46 nbd: you forgot to remove the endif, but if you look at https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/crypto/Kconfig#n1586 I think this is a better solution: http://stewie.be.tintel.eu/openwrt_drbg.patch Dec 10 11:51:05 thoughts? compiled + runtime tested on octeon with 4.3 Dec 10 11:55:54 oh and I'm a sucker for alphabetical order Dec 10 11:57:08 that actually makes looking at the crypto.mk file painful sometimes =) Dec 10 12:02:09 so drbg also needs jitterentropy? Dec 10 12:02:12 that's weird Dec 10 12:04:45 http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/crypto/Kconfig -> select CRYPTO_JITTERENTROPY Dec 10 12:04:54 it doesn't need it but it selects it Dec 10 12:05:01 extra info @ http://marc.info/?l=linux-kernel&m=143496272614455&w=2 Dec 10 12:05:54 ok Dec 10 12:06:04 so do we want to follow the kernel behavior ? Dec 10 12:06:11 yeah, i guess so Dec 10 12:06:14 feel free to send that patch Dec 10 12:06:19 I'll send this to the ML as-is Dec 10 12:06:29 with some background info in the commit msg Dec 10 12:08:27 kaloz r47818 trunk/target/linux/ (25 files in 3 dirs) * mvebu: add support for 4.1 and switch over Dec 10 12:11:50 kaloz r47819 trunk/target/linux/ (8 files in 2 dirs) * mvebu: add support for 4.4-rc Dec 10 12:13:28 kaloz r47820 trunk/target/linux/mvebu/base-files/etc/uci-defaults/03_wireless * mvebu: fixup wifi macs for Linksys boards Dec 10 12:14:33 kaloz r47821 trunk/target/linux/ generic/patches-4.1/144-mtd-bcm47xxpart-don-t-fail-because-of-bit-flips.patch generic/patches-4.1/143-mtd-bcm47xxpart-limit-scanned-flash-area-on-BCM47XX-.patch * generic/4.1: refresh patches Dec 10 12:17:36 kaloz r47822 trunk/package/kernel/mac80211/Makefile * mac80211: use the linux-firmware tree from 2015-12-07 Dec 10 12:23:33 glad I finally figured this out Dec 10 12:39:44 nbd r47823 trunk/ include/image.mk target/linux/ramips/image/Makefile * build: fix arch used during dts compilation Dec 10 12:40:03 nbd r47824 trunk/tools/mm-macros/Makefile * tools/mm-macros: update to 0.9.9 Dec 10 12:40:22 nbd r47825 trunk/scripts/download.pl * scripts/download.pl: cleanup GNOME source location definitions Dec 10 12:40:31 nbd r47826 trunk/package/libs/popt/Makefile * popt: remove xgettext prereq check, it is not necessary Dec 10 12:41:19 nbd r47827 trunk/package/kernel/linux/modules/crypto.mk * kernel/modules: fix crypto API RNG for >=4.2 Dec 10 12:44:43 nbd, ping Dec 10 12:44:56 nitroshift: pong Dec 10 12:45:08 Kaloz has introduced kernel 4.4 support for generic Dec 10 12:46:04 in the meantime i managed to build a set of patches for mvebu Dec 10 12:46:17 where can i send them? Dec 10 12:46:42 (apart from openwrt mailing list) Dec 10 12:47:15 i'm running both by wrt1900ac's on 4.4-rc4 Dec 10 12:47:21 nitro he did send sth : https://dev.openwrt.org/changeset/47819 Dec 10 12:47:49 plntyk, thanks, didn't see that commit Dec 10 12:48:42 i've sent him the payches by mail some 2 hours ago Dec 10 12:48:51 i guess he committed them :) Dec 10 12:49:41 ;) Dec 10 12:50:01 thanks :) Dec 10 12:50:42 though he didn;t commit them all Dec 10 12:50:55 i included the patches that enable the buffer manager in mvneta Dec 10 12:51:24 I guess he is still reviewing/testing them Dec 10 12:51:27 as those will be upstreamed soon, i guess there's no need for them Dec 10 12:51:48 Devastator, i'm running with them since rc3 Dec 10 12:51:57 that's about 2 weeks Dec 10 12:52:09 give or take Dec 10 12:53:43 nbd r47828 trunk/package/network/config/netifd/files/sbin/ifup * netifd: ifup-shellscript - fix wrong usage of 'local' Dec 10 12:57:09 build #169 of cobalt is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/169 Dec 10 13:00:04 build #169 of cns21xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cns21xx/builds/169 Dec 10 13:02:52 build #169 of orion is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/169 Dec 10 13:32:56 hmm, tc qdisc crashes with buffer overflow: http://pastebin.com/umPj6C3u Dec 10 13:33:32 DD r47817 on octeon Dec 10 13:33:51 you could try valgrind Dec 10 13:34:07 need to disable sstrip for iproute2 though Dec 10 13:46:25 Hi, I'm using one wireless phy on GL-Inet to have an AP and also simultaneously connect to another network. While it's working fine, I can't restart just one logical interface (wlan or wwan). Whenever I change wwan (supplicant configuration), hostapd does restart too after calling network reload. Any clue how it's implemented? Dec 10 14:15:29 msgctl, not sure, but if the AP that you connect to goes down you're gonna have a bad time Dec 10 14:23:57 otherwise, look in /lib/netifd/ for interface protocol implementations, /sbin/wifi is called whenever wifi settings change, that's probably where the restart gets triggered from Dec 10 14:25:41 wifi configuration is brought up per card, not per individual virtual interface Dec 10 14:25:52 so whenever you make the chanes to the sta, it has to bring down wifi and then bring it back up again Dec 10 14:28:24 nbd: any chance I could patch it in OpenWRT userspace, or the driver just doesn't allow that? Dec 10 14:32:28 s/driver/API Dec 10 14:32:33 ...or would it be too complex? Dec 10 14:32:33 it's not easy to rework the code to allow per-vif reload in a safe way Dec 10 14:40:32 nbd r47829 trunk/package/network/utils/linux-atm/files/br2684ctl * br2684ctl: fix config reload trigger Dec 10 14:41:08 nbd r47830 trunk/package/network/utils/linux-atm/files/br2684ctl * br2684ctl: add atm-bridge disabled option Dec 10 16:12:47 build #174 of ramips is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/174 Dec 10 16:19:10 build #166 of pxa is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/166 Dec 10 17:05:29 Hi. Where are the defaults set for setting up say a new 6in4 interface? Dec 10 17:06:34 I'm of the opinion that "option tos 'inherit'" should really be the default for tunnels that support it. Dec 10 18:49:36 build #135 of ramips.rt305x is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.rt305x/builds/135 Dec 10 18:53:39 build #161 of ppc40x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/161 Dec 10 19:06:49 blogic r47831 trunk/target/linux/ (80 files in 12 dirs) * ralink: bump to the target to v4.3 Dec 10 19:22:21 build #167 of imx6 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/167 Dec 10 20:27:40 build #139 of ramips.mt7620 is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.mt7620/builds/139 Dec 10 20:53:15 kaloz r47832 trunk/target/linux/ (12 files) * mvebu: update and fixup 4.4 patches Dec 10 20:58:33 build #142 of avr32 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/142 Dec 10 20:59:07 build #135 of au1000 is complete: Failure [failed shell_14] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/135 Dec 10 23:26:31 build #146 of xburst is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/146 Dec 11 00:53:19 build #170 of brcm63xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/170 **** ENDING LOGGING AT Fri Dec 11 02:59:58 2015