**** BEGIN LOGGING AT Tue Jan 24 02:59:57 2012
Jan 24 03:12:58 <mirko> Chocks: i remember this error
Jan 24 03:13:54 <mirko> either it is due to naming scheem errors (-2.12 instead of -2.14) or a symlink-issue
Jan 24 05:34:03 <philipp64|laptop> woo-hoo!  all of my platform-driver-x86 submissions were accepted upstream for alix2.c, net5501.c, and geos.c...
Jan 24 05:34:32 <philipp64|laptop> should be out in 3.4...  now if only getting them into OpenWRT was as easy. :-P
Jan 24 08:21:27 <builder___> build #95 of ep93xx is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/95
Jan 24 08:27:20 <acoul> gmorning
Jan 24 08:28:19 <acoul> looking for a gentoo VPS, too good to be true: http://www.greatnet.de/cms/front_content.php?client=1&lang=1&idcat=5&idart=12
Jan 24 09:42:04 <builder___> build #127 of ubicom32 is complete: Failure [failed compile_1]  Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/127
Jan 24 09:48:54 <CIA-84> juhosg * r29877 /trunk/target/imagebuilder/files/Makefile:
Jan 24 09:48:54 <CIA-84> ImageBuilder: respect the PROFILE parameter
Jan 24 09:48:54 <CIA-84> Based on a patch from #10766
Jan 24 09:48:55 <CIA-84> juhosg * r29878 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: prebuild all loaders in Image/Prepare stage
Jan 24 09:48:56 <CIA-84> juhosg * r29879 /trunk/target/linux/ar71xx/image/Makefile:
Jan 24 09:48:56 <CIA-84> ar71xx: create interim kernel files in a separate dir
Jan 24 09:48:56 <CIA-84> This reduces the size of the ImageBuilder tarball a lot.
Jan 24 09:48:58 <CIA-84> juhosg * r29880 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: nuke fs_all variable in image/Makefile
Jan 24 09:49:26 <builder___> build #129 of at91 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/129
Jan 24 11:32:58 <CIA-84> blogic * r29881 /trunk/package/ltq-dsl-app/ (5 files in 2 dirs): (log message trimmed)
Jan 24 11:32:58 <CIA-84> [lantiq] Initial DSL framework
Jan 24 11:32:58 <CIA-84> - Removed ifx_cpe_control_init.sh (the old init script)
Jan 24 11:32:58 <CIA-84> - Created dsl_control (/etc/init.d)
Jan 24 11:32:58 <CIA-84> - tidied up start and stop
Jan 24 11:32:59 <CIA-84> - stop calls dsl_notify before killing daemon
Jan 24 11:32:59 <CIA-84> - has status and lucistat options to provide detail
Jan 24 11:50:19 <CIA-84> juhosg * r29882 /trunk/target/linux/ramips/ (5 files in 3 dirs):
Jan 24 11:50:19 <CIA-84> ramips: add kernel support for ALL0256N
Jan 24 11:50:19 <CIA-84> Signed-off-by: Daniel Golle <dgolle@allnet.de>
Jan 24 11:50:21 <CIA-84> juhosg * r29883 /trunk/target/linux/ramips/base-files/ (6 files in 5 dirs):
Jan 24 11:50:21 <CIA-84> ramips: add user-space support for ALL0256N
Jan 24 11:50:21 <CIA-84> This adds uci-defaults and sysupgrade support for the ALL0256N.
Jan 24 11:50:21 <CIA-84> Signed-off-by: Daniel Golle <dgolle@allnet.de>
Jan 24 11:50:22 <CIA-84> juhosg * r29884 /trunk/target/linux/ramips/image/Makefile:
Jan 24 11:50:22 <CIA-84> ramips: generate sysupgrade image for the ALL0256N
Jan 24 11:50:22 <CIA-84> Signed-off-by: Daniel Golle <dgolle@allnet.de>
Jan 24 12:45:50 <CIA-84> swalker * r29885 /packages/libs/openldap/ (Makefile patches/020-autofs-schema.patch): [packages] openldap: remove unused ldif directory (#10712, thanks flyn), add missing libuuid dependency, remove symlink copies
Jan 24 14:16:36 <CIA-84> jow * r29886 /packages/dstat/ (. Makefile):
Jan 24 14:16:36 <CIA-84> New package: dstat
Jan 24 14:16:36 <CIA-84> Dstat is a versatile replacement for vmstat, iostat, netstat and ifstat
Jan 24 14:16:36 <CIA-84> written in python and with a very good plugin system.
Jan 24 14:16:36 <CIA-84> Signed-off-by: Roberto Riggio <roberto.riggio@create-net.org>
Jan 24 14:19:08 <jow_laptop> gah not again
Jan 24 14:20:45 <nbd> ?
Jan 24 14:20:48 <CIA-84> jow * r29887 /packages/packages/ (. net/ net/aria2/ net/aria2/Makefile):
Jan 24 14:20:49 <CIA-84> new package - aria2
Jan 24 14:20:49 <CIA-84> aria2 is a lightweight multi-protocol & multi-source download utility
Jan 24 14:20:49 <CIA-84> operated in command-line. It supports HTTP/HTTPS, FTP, BitTorrent
Jan 24 14:20:49 <CIA-84> and Metalink. aria2 has built-in JSON-RPC and XML-RPC interface.
Jan 24 14:20:49 <CIA-84> Signed-off-by: Cezary Jackiewicz <cezary.jackiewicz@gmail.com>
Jan 24 14:20:51 <CIA-84> jow * r29888 /packages/ (dstat/Makefile utils/dstat/ utils/dstat/Makefile): [packages] dstat: move to the correct directory
Jan 24 14:21:05 <nbd> ah
Jan 24 14:21:55 <nbd> i made it a habit of mine to run git log --name-status before git-svn dcommit
Jan 24 14:22:21 <nbd> reduces the number of accidents that make it to svn
Jan 24 14:22:40 <jow_laptop> yeah
Jan 24 14:23:41 <jow_laptop> gr and aria2 is wrong too
Jan 24 14:26:07 <CIA-84> jow * r29889 /packages/ (net/aria2/ net/aria2/Makefile packages/net/aria2/Makefile): [packages] aria2: move to the correct directory
Jan 24 14:33:07 <CIA-84> jow * r29890 /packages/ (dstat/ packages/): [packages] remove empty directories
Jan 24 15:26:30 * russell-- continues to have trouble with ath5k on ar2315 hardware
Jan 24 15:26:52 <loswillios> russell--: fonera?
Jan 24 15:27:12 <russell--> accton mr3201a and meraki mini both show it
Jan 24 15:27:29 <russell--> i haven't tried a fonera
Jan 24 15:28:09 <loswillios> hm ok. might try my fonera later
Jan 24 15:28:23 <russell--> seems to be something introduced with compat-wireless 2011-12-01
Jan 24 15:28:25 <loswillios> ath5k works well with my wgt634u
Jan 24 15:28:39 <russell--> yeah, i haven't seen it on a wgt
Jan 24 15:28:54 <russell--> or on the net4521's i just put together
Jan 24 15:29:47 <russell--> it might take a day or so to manifest itself, and then i see gain calibration errors and wireless stops working until i reboot
Jan 24 15:34:33 <nbd> russell--: i'm pretty sure that was introduced by a patch series from nick kossifidis
Jan 24 15:35:17 <russell--> yeah, i pinged him, but he's stopped replying to my emails ;-(
Jan 24 15:35:29 <nbd> i'm looking for the patch series
Jan 24 15:35:40 <nbd> maybe you can try to revert these patches one by one to find out which one caused the regression
Jan 24 15:36:15 <russell--> i could, but it would be quite slow, since it takes a while to show up.  still, might be necessary.
Jan 24 15:36:17 <nbd> if you can figure that out, i can help you with tracking down the bug
Jan 24 15:36:42 <nbd> can't find it in gmane
Jan 24 15:36:47 <nbd> i'll export them from my git tree for you
Jan 24 15:37:10 <russell--> nick suggested fastchanswitch=1 and debug=0x20, which i did, also changed channels.
Jan 24 15:38:23 <nbd> http://nbd.name/ath5k-patches.tar.gz
Jan 24 15:40:14 <nbd> i'd recommend doing a pseudo-bisect, i.e. first reverting half of the patches, then depending on the result either revert some more, or add some back
Jan 24 15:41:24 <russell--> right, thanks!
Jan 24 15:46:53 <russell--> hmm. i may be able to speed things up by doing a bunch of devices at a time
Jan 24 15:50:06 <Chocks> yes
Jan 24 15:50:26 <russell--> i happen to have a box of 60 of these ;-)
Jan 24 15:52:10 <Chocks> nbd: FWIW (nothing), I don't care for git
Jan 24 16:40:11 <CIA-84> swalker * r29891 /packages/net/umurmur/Makefile: [packages] umurmur: fix library linkage (#10664, #10287, thanks martin)
Jan 24 17:02:14 <builder___> build #134 of brcm63xx is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/134
Jan 24 17:02:14 <builder___> build #130 of s3c24xx is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/130
Jan 24 17:02:16 <builder___> build #127 of brcm47xx is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/127
Jan 24 17:02:18 <builder___> build #125 of atheros is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/125
Jan 24 18:09:37 <CIA-84> juhosg * r29892 /branches/backfire/target/imagebuilder/files/Makefile: backfire: ImageBuilder: respect the PROFILE parameter (backport of r29877)
Jan 24 18:11:36 <jow_laptop> oh this is going to make a lot of people happy
Jan 24 18:13:54 * Chocks does a dance
Jan 24 18:14:06 <Chocks> oh, not using backfire, no matter ;-)
Jan 24 18:14:26 <jow_laptop> trunk has been merged already
Jan 24 18:14:32 <jow_laptop> yesterday or so
Jan 24 18:14:45 * Chocks examines precise change
Jan 24 18:15:17 <jow_laptop> ah no this morning. r29877
Jan 24 18:16:11 <Chocks> right. I must admit to forgetting the issue, only that it's something I complained about
Jan 24 18:16:33 <jow_laptop> well its got excessive altely
Jan 24 18:16:38 <jow_laptop> the image generation part I mean
Jan 24 18:16:57 <Chocks> for sure
Jan 24 18:42:08 <CIA-84> juhosg * r29893 /trunk/target/linux/ar71xx/files-3.2/arch/mips/include/asm/mach-ath79/ar913x_flash.h: ar71xx: remove an unused header file
Jan 24 18:42:10 <CIA-84> juhosg * r29894 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: pass mtdlayout to Planex boards via command line
Jan 24 18:42:10 <CIA-84> juhosg * r29895 /trunk/target/linux/ar71xx/ (4 files in 2 dirs): ar71xx: remove hardcoded MTD layout from planex files
Jan 24 18:42:13 <CIA-84> juhosg * r29896 /trunk/target/linux/ar71xx/base-files/etc/uci-defaults/leds: ar71xx: preset WLAN LED config on MZK-W300NH
Jan 24 18:42:13 <CIA-84> juhosg * r29897 /trunk/target/linux/ar71xx/ (2 files in 2 dirs): ar71xx: fix the AP/Router LED definitons on MZK-W300NH
Jan 24 18:52:51 <jow_laptop> swalker: btw, do you have patchwork access?
Jan 24 19:10:02 <CIA-84> jow * r29898 /packages/mail/ssmtp/Makefile: [packages] ssmtp: add shebang (#10851)
Jan 24 19:46:21 <luka12345> nbd: ping - about netifd
Jan 24 20:13:45 <CIA-84> juhosg * r29899 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: rename Sysupgrade/{KR,RK} to Sysupgrade/{KR,RK}uImage
Jan 24 20:13:46 <CIA-84> juhosg * r29900 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: use the lzma loader on the Planex boards
Jan 24 20:53:07 <nbd> luka12345: pong
Jan 24 20:55:50 <CIA-84> jow * r29901 /trunk/package/hostapd/ (Makefile files/wpa_supplicant.sh): [package] hostapd: introduce a client_cert option which is passed on to wpa_supplicant (#8876)
Jan 24 20:56:27 <CIA-84> jow * r29902 /branches/backfire/package/hostapd/ (Makefile files/wpa_supplicant.sh): [backfire] merge r29901
Jan 24 21:06:12 <luka12345> nbd: the same confing i showed you yesterday in netifd is not working...
Jan 24 21:06:20 <nbd> ok, any details on what is not working?
Jan 24 21:06:33 <luka12345> yes :)
Jan 24 21:06:43 <luka12345> none of the interfaces has ip
Jan 24 21:06:46 <luka12345> not even lo
Jan 24 21:07:11 <nbd> ok, and netifd died?
Jan 24 21:08:07 <luka12345> i did not noticed it
Jan 24 21:08:19 <luka12345> there is no netifd in process list
Jan 24 21:08:35 <nbd> is there a core file in /tmp?
Jan 24 21:08:36 <luka12345> i'm hooked up via serial to the board now
Jan 24 21:08:50 <luka12345> yes there are several
Jan 24 21:09:00 <luka12345> should i remove them and fire up netifd again?
Jan 24 21:09:03 <nbd> no
Jan 24 21:09:15 <nbd> configure networking manually and send me those files
Jan 24 21:19:39 <luka12345> nbd: i'm having troubles with file transfer
Jan 24 21:20:12 <luka12345> i configured ip with 'ifconfig eth0 192.168.1.20'
Jan 24 21:20:38 <luka12345> i can ping that address localy from router, but i can not ping it from outside
Jan 24 21:20:52 <nbd> tried /etc/init.d/firewall stop?
Jan 24 21:20:55 <luka12345> i run tcpdump on the router and nothing gets to that interface
Jan 24 21:21:57 <luka12345> no change after stoping firewall
Jan 24 21:23:01 <luka12345> any other ideas?
Jan 24 21:23:47 <nbd> vconfig set_name_type DEV_PLUS_VID_NO_PAD
Jan 24 21:23:52 <nbd> vconfig add eth0 1
Jan 24 21:23:54 <nbd> ifconfig eth0.1 up
Jan 24 21:23:58 <nbd> then set the ip address on eth0.1
Jan 24 21:24:08 <nbd> and remove it from eth0
Jan 24 21:30:38 <luka12345> ok, i got them
Jan 24 21:30:54 <luka12345> should i send them to nbd@openwrt ?
Jan 24 21:31:02 <nbd> nbd@nbd.name
Jan 24 21:32:37 <luka12345> i have sent them...
Jan 24 21:32:50 <nbd> got them, thx
Jan 24 21:32:52 <nbd> i'll take a look
Jan 24 21:36:01 <nbd> ok, figured out how to trigger the issue
Jan 24 21:36:06 <nbd> will have a fix in a few minutes
Jan 24 21:37:55 <luka12345> ok
Jan 24 21:45:37 <CIA-84> nbd * r29903 /trunk/package/netifd/Makefile: netifd: update to latest, fixes a crash when the interface protocol was not specified
Jan 24 21:45:52 <nbd> luka12345: there, fixed
Jan 24 21:46:12 <btsimonh> blogic: I have nand...  now will work out why/how and let you know.
Jan 24 21:46:48 <luka12345> nbd: ok, i'm building it
Jan 24 21:47:14 <luka12345> nbd: can you look at this : http://dev.phrozen.org/gitweb/?p=owsip.git;a=blob;f=owsip.c;h=d87d8ba3f67ea64dbae019be85ba5a4d4cc2907d;hb=HEAD
Jan 24 21:47:22 <luka12345> lines 126 - 147
Jan 24 21:47:45 <luka12345> that is the way i get ip of interface (and wait until i get it)
Jan 24 21:48:34 <luka12345> do you know a better way of doing it and/or would you accept a patch of that functionality to libubox?
Jan 24 21:48:55 <nbd> i'm not sure libubox should cover networking related stuff
Jan 24 21:49:02 <nbd> i don't have a better way right now
Jan 24 21:49:09 <luka12345> ok
Jan 24 21:49:23 <jow_laptop> luka12345: I suggest to look into the radvd sources. It uses netlink to listen on interface change events
Jan 24 21:49:50 <jow_laptop> it should be possible to use a subscriber model to get notified
Jan 24 21:50:37 <luka12345> jow_laptop: ok, i will
Jan 24 21:51:20 <nbd> luka12345: btw. why does it need to wait for an ip?
Jan 24 21:52:29 <luka12345> because i want to bind owsip on specific interface
Jan 24 21:52:37 <luka12345> and i can not do it without ip
Jan 24 21:52:40 <nbd> why?
Jan 24 21:53:34 <luka12345> because pjsip is not hadling rtp trafic unless i bind it to right ip
Jan 24 21:53:44 <luka12345> im using the same stuff in freecwmp
Jan 24 21:53:58 <luka12345> there i'm also waiting until i get ip
Jan 24 21:54:36 <luka12345> because in tr069 specificiation you should do specific timeouts if you can not contact acs
Jan 24 21:55:00 <luka12345> and i start failure countion only when i have ip...
Jan 24 21:56:27 <nbd> it seems to me that the source ip for rtp traffic needs to be dynamic based on the local ip of the sip connection anyway
Jan 24 21:56:41 <nbd> why not just bind globally, and select the ip for rtp based on the local ip of the incoming connection?
Jan 24 21:57:01 <nbd> as for tr069, you can probably use the error code to figure out if the connection failed because you have no route, or if it's some other issue
Jan 24 22:00:09 <luka12345> for that i should look at pjsip library and see why if failed... for now it was easier to bind it to right interface/ip
Jan 24 22:00:40 <luka12345> and for tr069 you are right, but there is another thing
Jan 24 22:01:20 <luka12345> i need to tell acs on which ip it should contact me back, and for that i need to wait for interface to get an ip
Jan 24 22:01:37 <nbd> why? if you establish a connection to the acs, you can get the local ip from that sockjet
Jan 24 22:02:00 <luka12345> but i can not construct inform message without ip
Jan 24 22:02:11 <luka12345> it's the first message i need to send
Jan 24 22:02:13 <nbd> you can connect the socket
Jan 24 22:02:28 <nbd> that should be enough to get the local ip
Jan 24 22:02:51 <luka12345> now i lost you
Jan 24 22:02:59 <jow_laptop> getsockname()
Jan 24 22:03:12 <jow_laptop> on the fd you use to send()
Jan 24 22:03:23 <jow_laptop> use its result in the inform message
Jan 24 22:03:39 <nbd> is the connection to the acs udp?
Jan 24 22:03:57 <luka12345> no, http
Jan 24 22:04:00 <nbd> ok, even easier
Jan 24 22:04:00 <luka12345> tcp
Jan 24 22:04:09 <nbd> well, the same actually
Jan 24 22:04:28 <nbd> as soon as you've established the connection, you can use getsockname
Jan 24 22:04:41 <luka12345> ok, now i see where you are going
Jan 24 22:04:42 <nbd> at least i think it's getsockname
Jan 24 22:04:44 <luka12345> but
Jan 24 22:04:58 <luka12345> i have divided freecwmp into two parts
Jan 24 22:05:14 <luka12345> into core c stuff that handles communication with acs
Jan 24 22:05:35 <luka12345> and into script part that handles integration with openwrt and other stuff
Jan 24 22:05:59 <luka12345> the parameter where i need to get ip is like all other parameters handled in scripts
Jan 24 22:06:16 <nbd> yeah, but you already need to pass that info from the core to the scripts, right?
Jan 24 22:06:35 <luka12345> yes, you are right
Jan 24 22:07:21 <luka12345> i mean no
Jan 24 22:07:53 <luka12345> well i'm waiting in core until i know that interface has ip, then program goes on
Jan 24 22:08:03 <nbd> ok, you can do the same with a connect() attempt
Jan 24 22:08:22 <nbd> so you wait for the connect to succeed before continuing with the rest
Jan 24 22:08:23 <luka12345> and when i need ip parameter i go to scripts again
Jan 24 22:08:39 <nbd> which will make the program more robust, imho
Jan 24 22:08:49 <nbd> because then you also test if you have a working connection to the ACS
Jan 24 22:08:53 <nbd> instead of just testing for an ip address
Jan 24 22:09:03 <nbd> and the tool doesn't need to be aware of what outgoing interface to use anymore
Jan 24 22:09:15 <nbd> so it becomes more flexible wrt. having multiple wan connections
Jan 24 22:09:17 <nbd> without making it more complex
Jan 24 22:09:21 <jow_laptop> as nbd said, just implement some  while (connect(...) < 0) { log_error(); sleep(3); }
Jan 24 22:09:28 <jow_laptop> .. logic
Jan 24 22:10:03 <jow_laptop> as soon as you get a connect you can figure out the effective local ip from getsockname()
Jan 24 22:12:45 <luka12345> i admit that you aproach makes sence
Jan 24 22:13:23 <luka12345> but i would need to rewrite it and i'm lazy to do that now
Jan 24 22:13:51 <nbd> i think it'd be worth it
Jan 24 22:14:05 <luka12345> but i will try and see if it makes it better next time will touch the c code
Jan 24 22:22:45 <luka12345> netifd now is working
Jan 24 22:23:02 <luka12345> i will plug it in dsl now
Jan 24 22:23:35 <jow_laptop> luka12345: another problem I see with the hardcoded local ip is that it will change eventually without your service or the router restarting
Jan 24 22:24:11 <luka12345> in owsip i will add periodical checks
Jan 24 22:25:04 <luka12345> and in freecwmp it binds to the interface, and when if ip changes i get it from scripts when i need it
Jan 24 22:25:16 <luka12345> so it should not be a problem
Jan 24 22:27:17 <jow_laptop> ok
Jan 24 22:29:57 <luka12345> now i'm connected via netifd, anything else i should test?
Jan 24 22:35:26 <luka12345> jow_laptop: i'm reading about netlink
Jan 24 22:36:08 <luka12345> so in owsip instead periodical checks if ip has changed i could use netlink to "notify" me if there were ip changes, right?
Jan 24 22:36:16 <jow_laptop> yes
Jan 24 22:36:34 <luka12345> that makes more sence then
Jan 24 22:37:01 <Chocks> not to mention sense
Jan 24 22:37:16 <luka12345> well, i suck at spelling
Jan 24 23:44:16 <EqUaTe> hm.. anyone know a decent source of pcb pin headers in the UK that does 2x7's? got a trendnet device here that i want to stick a jtag header on.. (partly because i can..)
Jan 24 23:45:34 <EqUaTe> closest i can find is a 2x8 at maplin.. might work, but might be too big.
Jan 24 23:46:18 <EqUaTe> i suppose i could just solder the pins right in..
Jan 24 23:46:53 <Chocks> ebay is most likely
Jan 24 23:47:07 <Chocks> find ones you can cut if need be
Jan 24 23:47:28 <EqUaTe> pretty sure the maplin one could be cut.. i do have a dremel :P
Jan 24 23:47:36 <Chocks> enjoy
Jan 24 23:47:36 <EqUaTe> http://www.maplin.co.uk/idc-pcb-header-plugs-32803
Jan 24 23:47:56 <EqUaTe> dremels ftw! :P
Jan 24 23:48:28 <EqUaTe> 'course, i'll probably just throw in serial and leave it at that for now.
Jan 24 23:53:22 <EqUaTe> depending on how much i drink tomorrow i may do that when i get home.. we shall see.
Jan 25 01:38:47 <swalker> jow_laptop: I have an account for it
**** ENDING LOGGING AT Wed Jan 25 02:59:56 2012