**** BEGIN LOGGING AT Tue Nov 13 02:59:57 2012 Nov 13 08:47:17 blogic, ping if i may and if you take care of ramips like dir 615 h1.. Nov 13 08:47:47 i dont have that unit Nov 13 08:48:48 i do, serial on it right here, i applied dangole's two RF patches but there's another problem that doesnt manifest so often, now and then the router hangs Nov 13 08:49:13 the only thing i modified is the /etc/config/network which on the 2-3 month old trunk wouldnt work for this device Nov 13 08:49:50 fucking hate google search results Nov 13 08:50:00 oops .. wrong window Nov 13 08:50:02 why? Nov 13 08:50:26 http://www.google.com/url?sa=t&rct=j&q=rtl8366%20pdf&source=web&cd=1&cad=rja&ved=0CC8QFjAA&url=http%3A%2F%2Frealtek.info%2Fpdf%2Frtl8366_8369_datasheet_1-1.pdf&ei=kwmiUMatLorFswbh2YGQDA&usg=AFQjCNFAi2A3ueI_bzrS4ZRTyHHMdX1QRw Nov 13 08:50:29 this is it - http://pastebin.com/yrJL4cJ4 Nov 13 08:50:46 wat Nov 13 08:50:48 realtek pdf? Nov 13 08:50:52 its refuses to give you the resulting link and eavesdrops on you and what you clicked Nov 13 08:50:56 dont be evil google Nov 13 08:51:07 oh yeah Nov 13 08:51:10 it's been doing that a while Nov 13 08:51:18 I keep meaning to set up a greasemonkey script Nov 13 08:51:27 i did Nov 13 08:51:34 oh, nice Nov 13 08:51:35 and then upgraded google and now it fails Nov 13 08:51:42 lame Nov 13 08:51:46 link to script? Nov 13 08:52:32 no idea Nov 13 08:54:07 blogic: what device has that realtek chip? Nov 13 08:54:23 netgear dgn3500 Nov 13 08:54:39 oh cool the modem/router combos Nov 13 08:54:41 blogic r34180 trunk/target/linux/ lantiq/files-3.3/arch/mips/lantiq/xway/mach-netgear.c lantiq/ar9/config-default * [lantiq] fixes ar9 / dgn3500 support Nov 13 08:57:47 eigma: ^ Nov 13 08:57:48 that one Nov 13 08:57:54 cool Nov 13 08:57:55 w/e Nov 13 08:57:58 hehe Nov 13 08:58:03 no drivers for modem eh? Nov 13 08:58:11 is it a PCI device or..? Nov 13 08:58:12 its all working Nov 13 08:59:25 blogic: google isn't being evil, it's just helping you to stay sharp! Nov 13 09:00:01 lolwut Nov 13 09:00:24 eigma: the unit is available worldwider really off amazon Nov 13 09:00:51 so what does the modem look like? Nov 13 09:00:56 is it an eth interface? Nov 13 09:01:00 or maybe a ppp interface? Nov 13 09:01:08 nm Nov 13 09:01:11 ppp Nov 13 09:01:14 eth just above physical probably Nov 13 09:01:20 nas0 Nov 13 09:01:21 then ppp on top Nov 13 09:01:26 its a proper atm device Nov 13 09:01:46 window Nov 13 09:01:48 fuck Nov 13 09:01:51 awesome Nov 13 09:02:08 can you sniff atm frames Nov 13 09:02:10 with like wireshark Nov 13 09:03:03 tcpdum Nov 13 09:03:05 tcpdump Nov 13 09:03:10 nice Nov 13 09:03:19 anything other than your ip packets on the line? :> Nov 13 09:03:28 also can you inject atm frames? Nov 13 09:03:56 zioproto r34181 packages/net/ olsrd/patches/100-olsrd-fix-stack-corruption-in-net_output.patch olsrd/Makefile * [packages] olsrd: version bump to 0.6.4 Nov 13 09:04:25 do you have a pedal under your desk that you push with your foot, and each time OpenWrt-Commits prints a commit? Nov 13 09:06:48 ? Nov 13 09:06:51 that last was not me Nov 13 09:39:27 <[florian]> eigma: there is this irc bot hooked on the post-commit hook of the server Nov 13 12:23:18 * blogic feels like hitting bastel basti Nov 13 12:23:25 dude you got issues Nov 13 12:36:31 Hi guys/galls (if any) Nov 13 12:36:57 one quite ugly bug bit me Nov 13 12:40:06 I had issues with attitude adjustment trunks on tplink wr1043nd with Atheros AR9103, it looks like some ath9k bug that caused wifi to just stop working, you would see ssid beacons but couldn't associate to network Nov 13 12:40:34 valentt: we will push rc1 that fixes a few ath9k bugs Nov 13 12:40:43 please retest in 2-3 days with the new bianries Nov 13 12:41:09 now with R34158 after few hours wifi works but it is slooow, under 1mbit! Nov 13 12:41:14 blogic, thanks, will do. Nov 13 13:09:48 when will dev.openwrt.org be up? I'm interested in details in few bugs Nov 13 13:10:40 whenever Kaloz feels like restarting tracd... Nov 13 13:10:46 we cannot do anything about it Nov 13 13:15:19 blogic r34182 branches/attitude_adjustment/target/linux/lantiq/vr9/target.mk * [lantiq] vr9 i not really usable, lets just ignore it Nov 13 13:15:21 blogic r34183 branches/attitude_adjustment/target/linux/ramips/files/drivers/net/ethernet/ramips/ramips_main.c * [ramips] backport ethernet fix Nov 13 13:15:23 blogic r34184 branches/attitude_adjustment/target/ linux/lantiq/files/arch/mips/lantiq/xway/mach-netgear.c linux/lantiq/ar9/target.mk linux/lantiq/ar9/profiles/002-netgear.mk linux/lantiq/ar9/config-default * [lantiq] fixes dgn3500 Nov 13 13:15:25 blogic r34185 branches/attitude_adjustment/ include/version.mk package/base-files/image-config.in * beta2 -> rc1 Nov 13 13:42:01 Kaloz, could you setup some watchdog script to restart tracd automatically? Nov 13 13:42:19 or is it more complicated than I think? ;) Nov 13 13:42:42 i am sure there is such a script Nov 13 13:42:45 but thanks for the idea Nov 13 13:43:40 valentt: afaik monit is already monitoring tracd, but currently it looks like there is something preventing trac from coming back up Nov 13 13:44:30 that happens if one rely on pythons. first they overindulge, then they get lazy and stop moving ;) Nov 13 13:45:03 lol Nov 13 13:46:48 watch out when they get hungry again ;) Nov 13 14:42:46 ping dape Nov 13 14:42:58 hey dangole Nov 13 14:43:35 want me to plug the router and give you ssh Nov 13 14:48:24 juhosg r34186 trunk/target/linux/ ar71xx/patches-3.3/408-mtd-redboot_partition_scan.patch ar71xx/patches-3.6/408-mtd-redboot_partition_scan.patch * ar71xx: refresh patches Nov 13 14:48:26 juhosg r34187 trunk/target/linux/ ar71xx/patches-3.3/612-MIPS-ath79-TL-WA7510N-v1-support.patch ar71xx/patches-3.6/612-MIPS-ath79-TL-WA7510N-v1-support.patch ar71xx/files/arch/mips/ath79/mach-tl-wa901nd.c * ar71xx: add support for TP-Link TL-WA7510N Nov 13 14:48:27 juhosg r34188 trunk/target/linux/ ar71xx/base-files/lib/ar71xx.sh ar71xx/base-files/lib/upgrade/platform.sh * ar71xx: user space support for the TL-WA7510N Nov 13 14:48:28 juhosg r34189 trunk/tools/firmware-utils/src/mktplinkfw.c * firmware-utils/mktplinkfw: add support for TL-WA7510N Nov 13 14:48:29 juhosg r34190 trunk/target/linux/ ar71xx/generic/profiles/tp-link.mk ar71xx/image/Makefile * ar71xx: add profile and build image for the TL-WA7510N board Nov 13 14:55:41 dape: that'd be nice Nov 13 14:55:58 i still got to compare what went down the drain when i polished serge' Nov 13 14:56:00 's patches Nov 13 14:57:28 aww, snap, i forgot the a/c adapter at work goddamit Nov 13 14:59:41 damn, and nothing similar at 5v 1,2a Nov 13 15:00:30 dangole apologies, i'm gonna get back to you tomorrow Nov 13 16:03:37 dape: no worries Nov 13 16:04:03 btw: you did test the v3 version of the patches, i guess...? Nov 13 16:04:47 because in the version before the if-clause selecting which rf_vals to use was inverted by mistake Nov 13 17:09:03 what about: https://dev.openwrt.org/ticket/12459 Nov 13 17:09:08 ? Nov 13 17:20:41 florian r34191 trunk/target/linux/ malta/patches-3.6/index.html?p=ralf%2Flinux.git;a=patch;h=4ade440f81a3 malta/patches-3.6/001-MIPS-fix-CBUS-UART-irq.patch malta/patches-3.6 * [malta] fix CBUS UART registration Nov 13 17:20:45 florian r34192 trunk/target/linux/malta/Makefile * [malta] switch to kernel 3.6.6 Nov 13 17:40:10 florian r34193 trunk/target/linux/ malta/patches-3.6/index.html?p=ralf%2Flinux.git;a=patch;h=4ade440f81a3 malta/patches-3.6/002-MIPS-Malta-fix-build-failure.patch * [malta] fix badly named patch file Nov 13 19:12:18 blogic: please commit this: http://paste.openwrt.org/22176 Nov 13 19:28:04 the diag.sh stuff is very dependent on dgn3500 leds Nov 13 19:31:39 and why remove xway_cmdline ? Nov 13 19:31:53 because you want to pass the ethaddr ? Nov 13 19:57:07 blogic: yes Nov 13 19:57:31 setting the xway_cmdline in uboot works much easier Nov 13 19:57:49 erm no Nov 13 20:00:00 as for being dependent on the leds, that's why i have the test for their presence Nov 13 20:00:51 another issue i have not yet looked into is that nas0 has no mac addr Nov 13 20:01:02 that is normal Nov 13 20:01:27 it should be eth+1 Nov 13 20:02:01 why ? Nov 13 20:02:29 because zeros reported to the dslam gateway is bad Nov 13 20:02:47 the stock fw uses eth+1 Nov 13 20:03:55 one more thing: i cannot find where the reset and gpio buttons are exposed in /sys, but there is a third button for wifi on/off Nov 13 20:04:30 /sys/class/gpio Nov 13 20:05:07 gotta find the gpio number for the third button Nov 13 20:05:12 please do Nov 13 20:05:52 in the meantime, i think the diag.sh is very sane, with the change to mach-nergear.c Nov 13 20:06:35 and if you have a better idea about ethaddr, i'd like to hear it Nov 13 20:06:57 it sees the mtd's just fine now, btw Nov 13 20:10:30 here is what i typed in uboot: Nov 13 20:11:02 setenv net_owrt tftp 80500000 openwrt-lantiq-ar9-DGN3500B-uImage\;setenv bootargs \$(bootargs) machtype=DGN3500B console=ttyLTQ1,\$(baudrate) ethaddr=\$(ethaddr) rootfstype=squashfs,jffs2\;bootm;saveenv Nov 13 20:11:22 and to boot, "run net_owrt" Nov 13 20:14:05 blogic: https://dev.openwrt.org/ticket/12459 - you know what's wrong? Nov 13 20:15:21 DonkeyHotei: do something like $(call Image/BuildKernel/Template,DGN3500B),$(uboot_cmdline)), and than add uboot_cmdline to makefile just like the one for sx76x Nov 13 20:15:34 i still don't see the buttons exposed in /sys/class/gpio Nov 13 20:15:58 Delboy: can't, because it has no access to uboot env vars Nov 13 20:17:27 when you remove that - from the beginning of the line it uses that + what you set in uboot Nov 13 20:18:57 Linux blap6 3.2.0-3-amd64 #1 SMP Mon Jul 23 02:45:17 UTC 2012 x86_64 GNU/Linux Nov 13 20:19:33 oops Nov 13 20:20:02 written_tabletpc: the deps are not set in the makefile Nov 13 20:21:18 blogic: DEPENDS:=+libXi +libstdcpp +libfreetype Nov 13 20:21:21 is not enough? Nov 13 20:24:56 Delboy: uh, no, it replaces what uboot passes Nov 13 20:25:22 written_tabletpc: Package fltk2 is missing dependencies for the following libraries: Nov 13 20:25:25 libXft.so.2 Nov 13 20:25:27 libpthread.so.0 Nov 13 20:25:29 add xft and pthread Nov 13 20:25:39 ok, will try Nov 13 20:26:58 - DEPENDS:=+libXi +libstdcpp +libfreetype Nov 13 20:26:59 + DEPENDS:=+libXi +libXft +libstdcpp +libpthread +libfreetype Nov 13 20:27:04 correct, blogic ? Nov 13 20:27:34 pthread i think Nov 13 20:27:38 not libpthread Nov 13 20:27:46 ok Nov 13 20:28:11 ah no Nov 13 20:28:12 CONFIG_PACKAGE_libpthread=y Nov 13 20:28:16 libpthread it is Nov 13 20:28:31 where to put this in the fltk2/Makefile? Nov 13 20:28:40 yes Nov 13 20:28:55 ah, ok Nov 13 20:32:22 blogic: https://dev.openwrt.org/ticket/12459#comment:1 - this patch fixes it Nov 13 20:35:36 Delboy: any other bright ideas? Nov 13 20:43:17 spotting another error with xkbdata: Nov 13 20:43:17 checking for xkbcomp... not_found Nov 13 20:43:18 configure: error: xkbcomp is required to install the xkb data files Nov 13 20:43:40 it's in the dependencies of xkbdata Nov 13 20:44:48 xkbcomp got compiled Nov 13 20:45:05 well it did not get installed properly then Nov 13 20:45:13 check the config.og of xkbdata Nov 13 20:45:21 will do Nov 13 20:45:24 inside the build directory there will be a config.log Nov 13 20:45:31 in there is the real error Nov 13 20:45:37 it might be something totally different Nov 13 20:46:42 the error is at the end of the config.log or somewhere in the log file? Nov 13 20:48:14 hmm Nov 13 20:48:22 configure:1719: error: xkbcomp is required to install the xkb data files Nov 13 20:49:43 xkbdata's config.log is only 8k in size Nov 13 20:51:13 DonkeyHotei: try $(call Image/BuildKernel/Template,DGN3500B),$(sx76x_cmdline)), Nov 13 20:57:40 juhosg r34194 trunk/target/linux/generic/patches-3.6/062-mips_decompressor_build_fix.patch * generic: refresh 3.6 patches Nov 13 20:57:41 juhosg r34195 trunk/target/linux/ ar71xx/generic/profiles/pcs.mk ar71xx/image/Makefile * ar71xx: add profiles and build images for old PowerCloud System devices Nov 13 20:57:43 juhosg r34196 trunk/target/linux/ ar71xx/patches-3.3/613-MIPS-ath79-ubnt-xm-add-unifi-outdoor.patch ar71xx/patches-3.6/613-MIPS-ath79-ubnt-xm-add-unifi-outdoor.patch * ar71xx: add support for the Ubiquiti UniFi AP-Outdoor device Nov 13 20:57:44 juhosg r34197 trunk/target/linux/ ar71xx/base-files/lib/ar71xx.sh ar71xx/base-files/lib/upgrade/platform.sh ar71xx/base-files/etc/uci-defaults/network * ar71xx: add user-space support for the Ubiquiti UniFi AP-Outdoor board Nov 13 20:57:45 juhosg r34198 trunk/target/linux/ ar71xx/generic/profiles/ubnt.mk ar71xx/image/Makefile * ar71xx: add profile and build image for the Ubiquiti UniFi AP-Outdoor board Nov 13 20:59:28 any idea? Nov 13 20:59:31 Delboy: sx76x_cmdline is identical to xway_cmdline, so that would effectively revert it back to the way it is in svn, which i propose changing to what is in the pastebin Nov 13 21:00:47 DonkeyHotei: xway_cmdline=-console=..... sx76x_cmdline=console= that "-" after cmdline= is removed Nov 13 21:02:33 did not notice that Nov 13 21:04:02 i will try it Nov 13 21:09:55 Delboy: that actually worked, thanx! Nov 13 21:12:49 blogic: i still don't see the buttons exposed in /sys/class/gpio Nov 13 21:15:09 the buttons are not there Nov 13 21:15:22 buttons trigger a hotplug event Nov 13 21:15:30 /etc/hotplug.d/buttons/ Nov 13 21:15:35 oh. Nov 13 21:15:44 then how do i probe? Nov 13 21:17:14 is gpio-keys-polled loaded ? Nov 13 21:17:17 lsmod | grep gpio Nov 13 21:19:07 nope Nov 13 21:20:00 make an image with that included Nov 13 21:20:02 btw, setting the wps button to be a reboot will be damn inconvenient, considering how easy it is to brush against it on the device Nov 13 21:20:15 the buttons point at random events Nov 13 21:22:38 blogic: ? Nov 13 21:23:16 ? Nov 13 21:27:59 blogic: ok, added that module in, now what? Nov 13 21:28:28 now the buttons should work Nov 13 21:29:10 they don't seem to trigger the events in /etc/hotplug.d/button/10-generic.sh Nov 13 21:34:18 blogic: ideas? Nov 13 21:34:45 blogic r34199 feeds/xorg/lib/fltk2/Makefile * [packages] xorg/libs/fltk2 build is broken Nov 13 21:34:56 written_tabletpc: ^^ Nov 13 21:36:12 Hello, I'm trying to modify the script "/etc/hotplug.d/iface/00-netstate" in order to launch wpa_supplicant and then dhcpd. Right now I added [ "$DEVICE" = "$(uci -q get network.wan.ifname)" ] && { touch /tmp/it_works } but it doesnt work. How can I make this work? --> http://ideone.com/JBm0wq Nov 13 21:38:33 apocn: [ "$INTERFACE" = "wan" ] ? Nov 13 21:39:14 and launching dhcpd from 00-netstate seems very broken Nov 13 21:41:05 yes... I agree Nov 13 21:41:35 the thing is that I have proto="dhcp", but don't know how to force wpa_supplicant to start before sending the first DHCP-Discover Nov 13 21:41:58 even better... I would like the DHCP to be launched only if wpa_supplicant is successful. (802.1X) Nov 13 21:44:28 something like: iface eth0.2 inet dhcp \ pre-up wpa_supplicant -i eth0.2 .... \ post-down killall -q wpa_supplicant Nov 13 21:47:12 blogic, is this possible? Nov 13 21:50:34 blogic: what could be wrong with the button polling? Nov 13 21:51:03 apocn: right now there are no pre-up facilities Nov 13 21:51:50 apocn: edit /lib/netifd/proto/dhcp.sh, adjust proto_dhcp_setup() Nov 13 21:52:13 only execute proto_run_command() after the wpa-supp worked Nov 13 21:53:04 great! Nov 13 21:58:04 if wpa-supp fails to authenticate you can use proto_notify_error "$config" AUTH_FAILED to notify netifd about a permanent error, so it won't try to re-execute udhcpc over and over Nov 13 22:05:33 jow_laptop, thanks for mentioning it... I was thinking about creating a temporary file (like a flag) to avoid this. Nov 13 22:15:49 blogic: ? Nov 13 22:25:42 thanks, blogic. what about the other problem? Nov 14 00:10:15 * russell-- trying to decipher wgt634u flash organization ... looking at the stock fw partitioning, when i cited 0xa0000, it sounded wrong. cfe is only 384k, but the stock fw had a "config" partition in the 128k after that. finding a wide variation in the content of that section in the stock firmware'd devices we initially received. first 384k is identical everywhere. Nov 14 00:30:55 where is hauke when you want him? Nov 14 00:37:57 right now, probably sleeping :P Nov 14 01:06:32 sleep-- **** ENDING LOGGING AT Wed Nov 14 02:59:59 2012