**** BEGIN LOGGING AT Fri Oct 22 02:59:58 2010
Oct 22 04:11:35 <solca> nbd: does ath9k always advertise 11b rates
Oct 22 04:12:04 <solca> I'm trying to limit to 1M with hostapd.conf supported_rate=10 but clients do connect to 11M
Oct 22 04:12:24 <solca> s/supported_rate/supported_rates
Oct 22 04:14:59 <solca> well, it seems stations can connect to 1M 2M 5.5M and 11M rates when supported_rates=10 and basic_rates=10 in hostapd.conf
Oct 22 06:26:40 <builder> build #12 of mpc52xx is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/12
Oct 22 07:42:55 <builder> build #10 of xburst is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/10
Oct 22 07:56:46 <acoul> hello
Oct 22 09:40:45 <CIA-39> acoul * r23582 /packages/utils/smartmontools/ (Makefile patches/): utils/smartmontools: update package to version 5.40 (closes #8116)
Oct 22 12:01:34 <CIA-39> acoul * r23583 /packages/net/iperf/Makefile: net/iperf: fix a typo
Oct 22 13:38:25 <nunojpg> wireless on wrt160nl appears not to be working
Oct 22 13:39:16 <nunojpg> I've just compiled the current trunk and using two computers one found the network for a few seconds and the other never found it
Oct 22 14:03:02 <CIA-39> acoul * r23584 /packages/utils/smartmontools/Makefile: utils/smartmontools: fix a cross compile issue (thank you mbm)
Oct 22 14:13:47 <builder> build #14 of at91 is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/at91/builds/14
Oct 22 14:20:00 <builder> build #11 of ubicom32 is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/11
Oct 22 14:46:49 <KanjiMonster> xMff: regarding the kernel configs - shouldn't essentially CONFIG_*=m never be in any config-*? because either there's a kmod-package for it, then that should declare this config symbol, or there isn't a kmod, and then it isn't used/included anywhere, right?
Oct 22 14:50:06 <larsc> KanjiMonster: correct
Oct 22 14:51:13 <KanjiMonster> "grep -r --include="config*" '=m' target/linux" gets quite a lot of hits ;)
Oct 22 16:05:24 <nunojpg> I can confirm my previous report, wireless on the wrt160nl only works for around 60s after boot
Oct 22 16:05:31 <nunojpg> then the AP disappears
Oct 22 16:05:44 <nunojpg> I've just done a clean compile, and tested on 2 units
Oct 22 16:51:20 <CIA-39> jow * r23585 /packages/lang/php5/Makefile: (log message trimmed)
Oct 22 16:51:20 <CIA-39> [PATCH] php5: Php's sqlite3 module needs to be explicite linked to -lpthread with uClibc.
Oct 22 16:51:20 <CIA-39> Also add the required package dependencies. See also ticket #7237 for
Oct 22 16:51:20 <CIA-39> reference.
Oct 22 16:51:20 <CIA-39> A simple test case is the following php script:
Oct 22 16:51:21 <CIA-39> <?php
Oct 22 16:51:21 <CIA-39> $db = new SQLite3("test.db");
Oct 22 16:51:55 <xMff> ping OutBackDingo
Oct 22 16:52:06 <OutBackDingo> xMff: pong
Oct 22 16:52:15 <xMff> can you look into the Wing package update?
Oct 22 16:52:29 <OutBackDingo> xMff: what about it
Oct 22 16:52:40 <xMff> new version on openwrt-devel
Oct 22 16:52:46 <xMff> the author splitted host and target etc.
Oct 22 16:52:53 <xMff> should lead to less build issues in the long run
Oct 22 16:53:20 <OutBackDingo> yeah i emailed him patch doesnt apply cleanly, needs to be refreshed against the tree
Oct 22 16:53:27 <xMff> okay
Oct 22 16:53:42 <OutBackDingo> im on it though, once its fixed and tested ill commit
Oct 22 16:53:56 <OutBackDingo> xMff: you need a working version ? i have one
Oct 22 16:54:11 <xMff> no, no time for mesh stuff atm :)
Oct 22 16:54:17 <OutBackDingo> lol ok
Oct 22 17:09:24 <nunojpg> jow_laptop: did you notice my wireless broken report
Oct 22 17:23:39 <xMff> nunojpg: yes but not enough details
Oct 22 17:26:56 <nunojpg> xMff: can you ask me for details?
Oct 22 17:27:00 <nunojpg> I'm not sure what to give...
Oct 22 17:27:14 <xMff> is hostapd running after wifi disappeared?
Oct 22 17:27:22 <xMff> do you see any errors in dmesg?
Oct 22 17:27:33 <xMff> can you reestablish it with /sbin/wifi ?
Oct 22 17:27:47 <nunojpg> not with "wifi"
Oct 22 17:27:51 <nunojpg> hostapd with ps?
Oct 22 17:27:59 <xMff> any error messages with "killall -9 hostapd; hostapd -dd /var/run/hostapd-phy0.conf" ?
Oct 22 17:38:20 <nunojpg> it looks to disappear after the first sta disconnects, then only a "wifi" will restore the AP.  hostapd is still running when this happens
Oct 22 17:38:40 <xMff> ok. then do  killall -9 hostapd; hostapd -dd /var/run/hostapd-phy0.conf
Oct 22 17:38:44 <xMff> and make it disappear
Oct 22 17:39:04 <xMff> you'll see a lot of debug output
Oct 22 17:48:47 <builder> build #11 of x86 is complete: Failure [failed shell shell_13]  Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/11
Oct 22 18:17:06 <nunojpg> xMff: http://pastebin.com/EjCbE1vy
Oct 22 18:18:07 <nunojpg> this is a little random, but after "killall -9 hostapd; hostapd -dd /var/run/hostapd-phy0.conf" I never saw it happening again
Oct 22 18:18:26 <nunojpg> but I tried another router and I confirmed it happens
Oct 22 18:19:04 <xMff> I need it to happen with hostapd in debug mode :)
Oct 22 18:28:08 <nunojpg> xMff: cant it be something related to boot?
Oct 22 18:28:18 <xMff> unlikely
Oct 22 18:28:28 <nunojpg> possibly I should change the hostapd call on boot so it is in debug mode?
Oct 22 18:28:53 <nunojpg> forget, it just "happened"!
Oct 22 18:29:42 <nunojpg> but the console also dropped..(via another computer)
Oct 22 18:35:59 <nunojpg> I guess I messed up with the test, can I put hostapd on debug from boot?
Oct 22 18:36:03 <nunojpg> where is it started?
Oct 22 18:37:20 <xMff>  /lib/wifi/hostapd.sh
Oct 22 18:42:10 <Memphiz> hey nbd ... is i saw you committed some patches for ar71xx plattform ... and i just wanted to remember you for the "switch off gpio mode on init of flash" thingy - which still prevents the kernel from kexecing (you wanted to put that thing at the right place ... you remember?)
Oct 22 19:02:15 <nunojpg> xMff: it's called on hostapd.sh, but the one I see on boot is being started in mac80211.sh
Oct 22 19:02:39 <xMff> mac80211.sh starts procedures defined in hostapd.sh
Oct 22 19:02:48 <nunojpg> I added -dd, but forwading the output to a file doesn't work
Oct 22 19:03:20 <xMff> did you use stderr?
Oct 22 19:03:30 <nunojpg> owh...
Oct 22 19:03:55 <xMff> and use /tmp
Oct 22 19:04:06 <xMff> the file will grow really fast so it might crash your router at some point
Oct 22 19:04:12 <xMff> or just fill up /tmp
Oct 22 19:04:57 <nunojpg> yes, that I was using
Oct 22 19:05:10 <nunojpg> not because of your explanation, but because I love my flash
Oct 22 19:20:33 <nunojpg> &> /tmp/foo?
Oct 22 19:20:42 <nunojpg> foo is empty
Oct 22 19:20:49 <nunojpg> &>> /tmp/foo either
Oct 22 19:21:13 <xMff> 2>/tmp/foo
Oct 22 19:27:01 <nunojpg> mine is still empty...
Oct 22 19:29:29 <nunojpg> http://pastebin.com/wt3VC0rd
Oct 22 19:32:11 <xMff> remove the -B
Oct 22 19:32:17 <xMff> and put an & in the end
Oct 22 19:32:52 <xMff> hostapd -dd -P /var/run/wifi-$phy.pid /var/run/hostapd-$phy.conf  2>/tmp/logzinho & true || {
Oct 22 19:49:14 <CIA-39> zandbelt * r23586 /packages/net/asterisk-1.6.x/Makefile: [packages] asterisk-1.6.x: add func_channel and func_blacklist
Oct 22 21:09:01 <Nicklas`> Hi
Oct 22 21:09:04 <Nicklas`> a little question
Oct 22 21:09:36 <Nicklas`> For module kernel loading, I use the AUTOLOAD macro
Oct 22 21:10:02 <Nicklas`> AUTOLOAD:=$(call AutoLoad,<order>,<module1>...<moduleN>)
Oct 22 21:10:11 <Nicklas`> How can I set module parameters ?
Oct 22 21:13:50 <xMff> you can't
Oct 22 21:14:15 <Nicklas`> ?
Oct 22 21:14:26 <Nicklas`> it's boring...
Oct 22 21:14:54 <xMff> kmod definitions are sahred among different targets, adding a param there would make it work on some and break it on others
Oct 22 21:15:12 <Nicklas`> of course...
Oct 22 21:15:15 <xMff> if the param is the same everywhere, patch the module
Oct 22 21:15:25 <Nicklas`> but modules.mk can be written into the target directory
Oct 22 21:15:30 <xMff> if the kmod is specific to your target, patch it too
Oct 22 21:15:37 <Nicklas`> target/linux/<board>/modules.mk
Oct 22 21:15:48 <Nicklas`> and for my board, I could use params...
Oct 22 21:16:21 <xMff> override package/kmod-whatever/install then
Oct 22 21:16:36 <xMff> and write the autoload file yourself
Oct 22 21:44:30 <CIA-39> mirko * r23587 /packages/utils/vim/Makefile: [vim] enable multibyte feature in vim-full package
Oct 22 21:45:20 <danitool> [florian]: is there any known performance problem in bcm63xx boards?
Oct 22 22:23:20 <nunojpg> xMff: "hostapd -dd -P /var/run/wifi-$phy.pid /var/run/hostapd-$phy.conf  2>/tmp/logzinho & true || {"
Oct 22 22:23:23 <nunojpg> doesn't work either
Oct 22 22:23:48 <nunojpg> /tmp/logzinho is still, even after killing the process
**** ENDING LOGGING AT Sat Oct 23 02:59:57 2010