**** BEGIN LOGGING AT Fri Apr 20 02:59:58 2012 Apr 20 03:35:29 is there a way to chmod a file in base-files/files so, say, it'll be 600 on the squashfs? Apr 20 10:59:00 doh, tried to use git-send-email's "in-reply-to" to make a second version of a patch for the list and patchwork, Apr 20 10:59:09 but it's come out looking like a whole new patch with no context at all. Apr 20 11:03:18 karlp: that's normal patchwork behaviour; you still have to login to patchwork and mark your old patch as superseded Apr 20 11:16:26 ok, done :) Apr 20 11:16:38 though I have a feeling I now have two accounts on patchwork. Apr 20 11:17:11 there's no "I forgot my password" link anywhere, so I don't know whether someone else had my nick, or whether I just didn't have the right password Apr 20 11:52:15 Anyone here good with technical details of mac80211? I'm tracking a bug present in attitude adjustment that is not present in backfire, prevents secure wifi from working at all with this device. Can see differences using wpa_supplicant client with debug options. Apr 20 13:18:41 nbd, others: here's a page with side-by-side comparison of wpa_supplicant in debug mode talking to backfire/attitude adjustment r31296... the client talking to attitude adjustment router shows RX EAPOL before associated and then says "Not associated - Delay processing of received EAPOL frame"... ends up failing and wpa_supplicant loops over and over trying to reconnect [I've only included one iteration] Apr 20 13:19:29 I see behaviour with attitude adjustment is that encrypted wifi does not work at all, at least on SL-R7205. Have tried lots of modes. Non-encrypted works. Apr 20 13:19:49 http://www.aaa.net.au/campbell/wifi.html Apr 20 13:25:03 well, the whole authentication procedure works Apr 20 13:25:14 but shortly after that it gets disassociated for some reason Apr 20 13:27:21 nbd: is uci exposable in perl? Apr 20 13:27:29 ie can i use uci stuff from perl? Apr 20 13:27:43 you'd have to write a module for it Apr 20 13:28:09 ok Apr 20 13:38:28 nbd: thanks for looking... yes it authenticates but then says CTRL-EVENT-DISCONNECTED .. does that imply that openwrt made the disconnection? Apr 20 13:39:11 that only implies that there was a disconnect, it doesn't indicate what triggered it Apr 20 13:40:03 also, the comparison is between two different AP chipsets running different drivers and different versions of the software Apr 20 13:40:08 i'm not sure there's any meaningful data in there Apr 20 13:40:19 also, I have no idea what kind of client chipset you're using, that might be relevant as well Apr 20 13:51:46 nbd: I see this immediate disconnection behaviour with encrypted wifi 100% consistently when connecting from computer with Qualcomm Gobi 2000 wlan hardware, also from Intel 4965 wlan hardware, and also my phone won't connect either. Apr 20 13:52:53 ok Apr 20 13:53:10 did you build the image yourself? Apr 20 13:53:17 nbd: yes Apr 20 13:53:40 edit your .config, set CONFIG_WPA_MSG_MIN_PRIORITY to 0 Apr 20 13:53:47 then rebuild the hostapd package Apr 20 13:53:55 afterwards on the router do this: Apr 20 13:54:00 killall hostapd Apr 20 13:54:07 hostapd -dd /var/run/hostapd-phy0.conf Apr 20 13:54:21 see what debug messages show up when you connect a client Apr 20 13:54:27 if it kicks out a client, it should mention why Apr 20 13:54:31 nbd: ok Apr 20 13:54:57 thanks Apr 20 13:56:03 florian * r31361 /trunk/package/xfsprogs/patches/140-no_po.patch: [package] xfsprogs: disable po files building Apr 20 13:56:08 florian * r31362 /trunk/package/button-hotplug/Makefile: [package] button-hotplug: prevent build on 2.6.30 Apr 20 13:56:11 florian * r31363 /trunk/package/gpio-button-hotplug/Makefile: [package] gpio-button-hotplug: don't build on 2.6.30 Apr 20 13:56:14 florian * r31364 /trunk/target/linux/ps3/config-2.6.30: [ps3] add missing config symbol Apr 20 13:56:17 florian * r31365 /trunk/target/linux/generic/patches-2.6.30/ (3 files): [kernel/2.6.30] remove some annoying warnings and fix snd modules build Apr 20 13:57:05 florian * r31366 /packages/net/openl2tp/Makefile: Apr 20 13:57:05 [package] openl2tp: fix library path Apr 20 13:57:05 When host is x86_64 other than Debian, the Makefile of openl2tp uses /usr/lib64 Apr 20 13:57:05 as SYS_LIBDIR instead of /usr/lib. Apr 20 13:57:05 Aparently setting CROSS_COMPILE is the way the authors intended to handle this Apr 20 13:57:06 situation correctly. Apr 20 13:57:06 Signed-off-by: Daniel Golle Apr 20 13:57:07 florian * r31367 /packages/libs/tiff/ (Makefile patches/ patches/001-CVE-2012-1173.patch): [package] tiff: update to 4.0.1 and include CVE-2012-1173 patch Apr 20 13:57:08 florian * r31368 /packages/libs/freetype/Makefile: Apr 20 13:57:08 [package] freetype: update to 2.4.8 Apr 20 13:57:09 Fixes CVE-2011-3439. Apr 20 13:57:10 florian * r31369 /packages/libs/uclibc++/ (8 files in 2 dirs): Apr 20 13:57:10 [package] uclibc++: update to 0.2.3 Apr 20 13:57:10 Allows us to get rid of some patches Apr 20 13:58:09 claudio * r31370 /trunk/target/linux/at91/image/Makefile: Apr 20 13:58:09 [at91] Ensure that the bin directory exists before copy the files Apr 20 13:58:09 Signed-off-by: Claudio Mignanti Apr 20 14:03:44 nbd: hostapd is a symlink to wpad ... is that ok? Apr 20 14:03:46 claudio * r31371 /trunk/target/linux/at91/ (9 files in 4 dirs): Apr 20 14:03:46 [at91] Update the kernel to 3.3.x series, remove patches that doesn't apply anymore Apr 20 14:03:46 Signed-off-by: Claudio Mignanti Apr 20 14:03:52 claudio * r31372 /trunk/target/linux/at91/ (34 files in 15 dirs): Apr 20 14:03:52 [at91] Reorganize the subtargets, group subtargets per cpu family Apr 20 14:03:52 Signed-off-by: Claudio Mignanti Apr 20 14:03:55 claudio * r31373 /trunk/target/linux/generic/config-3.3: Apr 20 14:03:55 [generic] Add a missing symbol for config-3.3 Apr 20 14:03:55 Signed-off-by: Claudio Mignanti Apr 20 14:04:08 CaneToad: yes, wpad is a multicall binary of hostapd+wpa_supplicant Apr 20 14:04:13 CaneToad: it's built as part of the hostapd package Apr 20 14:04:23 nbd: ok Apr 20 14:14:28 nbd: see http://fpaste.org/Mg1R/ ... note: nl80211: set_key failed; err=-2 No such file or directory) Apr 20 14:15:38 nbd: http://lists.shmoo.com/pipermail/hostap/2010-March/021206.html has some comment on this, search for "No such file or directory" Apr 20 14:25:08 please give me a more complete log Apr 20 14:25:16 not just that short snippet with all the context stripped off Apr 20 14:27:49 nbd: see http://www.aaa.net.au/campbell/hostapd.log Apr 20 14:33:32 hm, odd Apr 20 14:33:38 i can't find any reason for this behavior in the code Apr 20 14:33:50 also, i can't reproduce it myself Apr 20 14:34:00 please send me the output of ./scripts/diffconfig.sh in your build tree Apr 20 14:34:38 nbd: it is vanilla apart from the config, but can do that Apr 20 14:40:02 nbd: http://fpaste.org/LkF6/ Apr 20 14:41:07 ok, looks clean Apr 20 14:42:58 nbd: do you know where in the code it reports the "set_key failed" ? Apr 20 14:43:41 yes, build_dir/linux-ar71xx_generic/compat-wireless-2012-04-17/net/mac80211/cfg.c, function ieee80211_add_key() Apr 20 14:43:45 if the sta lookup fails Apr 20 14:43:51 but i see no reason for the sta lookup to fail Apr 20 14:45:59 nbd: here build_dir/linux-ramips_rt305x/compat-wireless-2012-02-27/net/mac80211/cfg.c Apr 20 14:46:09 different date there Apr 20 14:46:13 yeah Apr 20 14:47:18 nbd: want me to add any debug? Apr 20 14:47:20 maybe try updating to see if the issue goes away Apr 20 14:47:31 i have to leave soon, maybe we can debug this some other time Apr 20 14:47:42 nbd: which time zone you in? Apr 20 14:47:54 or where are you? Apr 20 14:47:58 i'm in germany Apr 20 14:48:15 but at least for the next week, i'll probably only be available in the evening Apr 20 14:48:19 I'm in Australia, tomorrow already here Apr 20 14:48:24 ;) Apr 20 14:49:10 nbd: thanks very much... I might put some debug in and do some tracing, will let you know... :) Apr 20 14:49:50 if updating doesn't work, focus on adding debug code to see if the station entry goes away prematurely or if the lookup fails for some other reason Apr 20 14:51:08 nbd: the shmoo link above says: The problem here is that the key is being configured before association and mac80211/cfg80211 is not exactly ready for it.. I quick workaround would be to re-configure the key after reassociation has been completed. Apr 20 14:52:38 but the stack does throw an event telling the user space that the station entry is there Apr 20 15:36:29 jow * r31374 /trunk/package/base-files/files/etc/ (config/system init.d/sysntpd): (log message trimmed) Apr 20 15:36:29 enable ntpd server for busybox Apr 20 15:36:29 hi Apr 20 15:36:29 Another version, in this one the enable_server option is in the timeserver ntp part of the "system" config file Apr 20 15:36:29 You can patch trunk and bacfire (tested both) Apr 20 15:36:30 You can put busybox ntpd in client mode (if you put server), in client & server Apr 20 15:36:31 (by putting enable_server to 1, ntpd listen to udp 123), and also in server mode Apr 20 15:38:49 jow * r31375 /packages/libs/avahi/files/service-http: (log message trimmed) Apr 20 15:38:49 avahi: do not announce a specific front page Apr 20 15:38:49 Announcing /index.html as frontpage is right when luci or x-wrt is installed Apr 20 15:38:49 as web gui since both applications come with an /index.html which redirect afterwards. Apr 20 15:38:49 However, when using an other front page, e.g. an /index.php then the announcement Apr 20 15:38:50 is wrong and results in a 404 Not Found. Apr 20 15:38:50 So I wonder/suggest whether the default avahi service could not announce Apr 20 15:40:42 jow * r31376 /packages/net/mosquitto/ (Makefile files/ files/mosquitto.init): Apr 20 15:40:43 Provide an init.d script for mosquitto MQTT broker Apr 20 15:40:43 Use the service helper scripts to build a simple init script for the Apr 20 15:40:43 mosquitto broker. To make sure that the standard shipped config file Apr 20 15:40:43 works, add a mosquitto user if not found. Apr 20 15:40:43 Signed-off-by: Karl Palsson Apr 20 15:49:07 jow * r31377 /trunk/package/e2fsprogs/ (Makefile files/e2fsck.sh): [package] e2fsprogs: fix status reaping with fsck piped to logger, based on patch by Lukasz Golec-Biernat Apr 20 15:55:45 juhosg * r31378 /trunk/target/linux/generic/patches-3.3/980-update_arm_machtypes.patch: linux/3.3: update ARM mach-types Apr 20 15:55:48 juhosg * r31379 /trunk/target/linux/cns21xx/ (20 files in 2 dirs): cns21xx: add support for 3.3 Apr 20 15:55:50 juhosg * r31380 /trunk/target/linux/cns21xx/Makefile: cns21xx: switch to 3.3.2 Apr 20 15:55:51 juhosg * r31381 /trunk/target/linux/cns21xx/ (config-3.2 patches-3.2/): cns21xx: remove 3.2 support Apr 20 16:26:20 build #133 of iop32x is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/133 Apr 20 17:21:07 juhosg * r31382 /trunk/target/linux/ar71xx/patches-3.3/610-MIPS-ath79-openwrt-machines.patch: ar71xx: remove duplicated AP121 Kconfig entry Apr 20 17:50:56 jow * r31383 /packages/net/ipset/ (Makefile patches/200-remove-ipv6_addr_copy.patch): [packages] ipset: fix compilations against Linux v3.3 and newer (ipv6_addr_copy() got removed in favor to a direct structure copy) (#11310) Apr 20 18:08:36 jow * r31384 /packages/net/ipset/patches/210-fix-ipv6_skip_exthdr.patch: [packages] ipset: fix another compilation issue against Linux 3.3 and newer (ipv6_skip_exthdr() gained an additional argument) (#11310) **** ENDING LOGGING AT Fri Apr 20 18:23:12 2012 **** BEGIN LOGGING AT Fri Apr 20 18:24:16 2012 **** ENDING LOGGING AT Fri Apr 20 18:25:00 2012 **** BEGIN LOGGING AT Fri Apr 20 18:26:19 2012 Apr 20 18:37:58 build #169 of at91 is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/169 Apr 20 18:43:14 you are a failure :P Apr 20 21:34:03 juhosg * r31385 /trunk/target/linux/ppc40x/config-3.3: ppc40x: sync 3.3 config Apr 20 21:34:05 juhosg * r31386 /trunk/target/linux/ppc40x/Makefile: ppc40x: switch to 3.3.2 Apr 20 21:34:06 juhosg * r31387 /trunk/target/linux/ppc40x/ (config-default patches/): ppc40x: nuke 3.2 support Apr 20 22:23:28 jogo * r31388 /packages/net/znc/Makefile: packages: ZNC: update to 0.206 Apr 21 00:07:22 build #157 of pxcab is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/157 Apr 21 01:20:01 build #153 of ppc40x is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/153 Apr 21 02:47:58 jow * r31389 /packages/net/openswan/ (3 files in 2 dirs): [packages] openswan: update to v2.6.38, fixes compilation against Linux v3.3.x (#11316) **** ENDING LOGGING AT Sat Apr 21 02:59:59 2012