**** BEGIN LOGGING AT Mon Jan 11 02:59:57 2010
Jan 11 04:44:48 <CIA-36> lars * r19098 /trunk/target/linux/xburst/ (124 files in 40 dirs): Merge xburst target.
Jan 11 05:58:01 <CIA-36> nbd * r19099 /trunk/package/mac80211/patches/ (590-endian_fix.patch 600-monitor_select_queue_fix.patch): mac80211: fix some monitor mode queue selection bugs
Jan 11 08:02:12 <CIA-36> juhosg * r19100 /trunk/target/linux/ar71xx/ (12 files in 7 dirs): ar71xx: add support for the D-Link DIR-600 rev. A1 board
Jan 11 09:04:54 <[florian]> swalker: ping ?
Jan 11 09:16:44 <CIA-36> kaloz * r19101 /trunk/target/linux/ (omap35xx/Makefile xburst/Makefile): targets using 2.6.32 should use the latest release
Jan 11 09:40:30 <CIA-36> kaloz * r19102 /trunk/package/madwifi/patches/453-procps.patch: refresh patch
Jan 11 11:02:56 <CIA-36> lars * r19103 /trunk/tools/ (4 files in 3 dirs): Add libuuid hosttool package. It's needed in order to build mkfs.ubifs
Jan 11 11:04:31 <CIA-36> lars * r19104 /trunk/tools/mtd-utils/patches/101-ubifs-optional_lzo.patch: mtd-utils: Disable lzo for mkfs.ubifs
Jan 11 11:18:28 <CIA-36> lars * r19105 /trunk/tools/libuuid/Makefile: Add libuuid Makefile. Oops...
Jan 11 13:00:42 <tmonjalo> Hi all
Jan 11 13:01:37 <tmonjalo> Is there anybody who has already have a problem with madwifi while configuring a WEP AP and connecting to another WEP AP at the same time ?
Jan 11 13:01:51 <tmonjalo> I've just open this bug: https://dev.openwrt.org/ticket/6487
Jan 11 13:03:06 <tmonjalo> sorry for my english, s/.../who has ever seen/
Jan 11 13:23:10 <danage> tmonjalo: must be same channel afaik, questions like this are more likely to be answered over in #openwrt
Jan 11 13:23:59 <jow_laptop> tmonjalo: have you tried to use different wep key slots?
Jan 11 13:24:34 <jow_laptop> e.g.: option key '1' + option key1 '...' in the AP vif and option key '2'
Jan 11 13:24:45 <jow_laptop> + option key2 '...' in the STA vif
Jan 11 13:28:57 <tmonjalo> jow_laptop: the problem is on the AP side: we cannot connect to it if STA is connected with WEP
Jan 11 13:29:40 <jow_laptop> "If you set the same WEP key for AP and STA, it is possible to connect to the AP."
Jan 11 13:29:56 <tmonjalo> jow_laptop: strange, I know
Jan 11 13:30:06 <jow_laptop> therefor I asked you whether using different key slots for ap and sta fixes your issue
Jan 11 13:30:26 <jow_laptop> what is your current wireless configuration?
Jan 11 13:30:51 <danage> i think you must run the ap and the sta on the same wireless channel
Jan 11 13:31:03 <tmonjalo> danage: of course
Jan 11 13:31:32 <danage> perhaps best to pastebin your config files
Jan 11 13:32:30 <nbd> tmonjalo: jow's suggestion is correct, imho. the problem is in the conflict between the key index on the sta and the key index on the ap
Jan 11 13:32:39 <nbd> wep keys are 4 static keys per card
Jan 11 13:32:47 <nbd> thus cannot be fully per-vif
Jan 11 13:33:36 <nbd> why use wep at all?
Jan 11 13:33:43 <nbd> i mean it's not like it provides any security...
Jan 11 13:33:45 <jow_laptop> legacy requirements most likely
Jan 11 13:33:47 <tmonjalo> nbd, jow_laptop: are you suggested that the wep keys are shared between virtual interfaces ?
Jan 11 13:33:50 <nbd> can be broken in a minute
Jan 11 13:33:56 <nbd> tmonjalo: yes
Jan 11 13:33:58 <danage> wep is like CSS for dvd
Jan 11 13:34:30 <tmonjalo> I'm working for someone who wants to launch a new router
Jan 11 13:34:43 <tmonjalo> wep must be working
Jan 11 13:34:55 <tmonjalo> we need to test all configurations
Jan 11 13:35:06 <jow_laptop> tmonjalo: keys are shared between virtual interfaces but you have four available slots, by default slot 1 is used for both ap and sta, you should change your configuration to use slot 1 for ap and slot 2 for sta - this way you can have two different wep kes
Jan 11 13:35:33 <tmonjalo> thx for the tip
Jan 11 13:35:51 <nbd> jow_laptop: we could change the default key index to be (vif index % 4) + 1
Jan 11 13:36:03 <tmonjalo> I will test in one or two hours and brb
Jan 11 13:36:09 <nbd> in madwifi.sh
Jan 11 13:36:24 <jow_laptop> nbd: yeah, but why bother? :)
Jan 11 13:36:46 <nbd> to reduce the number of questions we get about this issue
Jan 11 13:36:50 <nbd> in case more people stumble upon this
Jan 11 13:36:58 <jow_laptop> if we do it for madwifi then we should do the same for broadcom and mac80211, shouldn't we?
Jan 11 13:37:17 <nbd> yeah, probably
Jan 11 13:37:19 <jow_laptop> esp. since wep is the only available crypto for wds on broadcom
Jan 11 13:37:34 <jow_laptop> smoking break brb
Jan 11 13:37:46 <nbd> wpa on broadcom wds is possible, we just need to fix it
Jan 11 13:38:05 <nbd> iirc it partially depends on the crypto settings of the base interface as well
Jan 11 13:38:21 <nbd> so if you do wds with wpa, the ap interface needs to be wpa-encrypted as well on both sides
Jan 11 13:44:42 <Hauke> nbd: Should your patch to mac80211 fixes my problem with wifi on brcm47xx?
Jan 11 13:50:45 <jow_laptop> nbd: hmm okay, I guess that there are more issues, I've seen that the MAC of the WDS vif is not incremented, this could be the reason why WPA on WDS fails
Jan 11 14:19:35 <nbd> Hauke: hopefully yes
Jan 11 14:19:48 <nbd> jow_laptop: the mac should not be incremented for wds
Jan 11 14:20:48 <jow_laptop> nbd: hmm but we had to do that for sta/ap vifs in order to fix wpa
Jan 11 14:21:20 <jow_laptop> maybe a different issue
Jan 11 14:37:59 <nbd> jow_laptop: wds uses the bssid of the ap vif
Jan 11 14:38:17 <nbd> and it's supposed to do that
Jan 11 14:38:26 <jow_laptop> allright
Jan 11 14:40:31 <blogic> nbd: does sta+ap work reliable on trunk madwifi ?
Jan 11 14:43:41 <nbd> it should. it's been a while since i've tested it, though
Jan 11 14:44:52 <acoul> @blogic: it does
Jan 11 14:45:15 <blogic> acoul: thx
Jan 11 14:45:26 <acoul> np
Jan 11 14:45:53 <_trine> trunk does not seem to be compiling today
Jan 11 14:46:14 <nbd> details?
Jan 11 14:46:59 <_trine> http://pastebin.com/m77aa0b8c
Jan 11 14:47:29 <nbd> odd
Jan 11 14:47:43 <nbd> did you use -j?
Jan 11 14:47:55 <_trine> yes I always do
Jan 11 14:48:05 <_trine> and I did a distclean
Jan 11 14:48:11 <nbd> please don't use it when showing errors
Jan 11 14:48:24 <nbd> it hides error messages
Jan 11 14:48:26 <_trine> I will try again without it
Jan 11 14:50:43 <_trine> it seems to be getting further without -j
Jan 11 14:51:27 <_trine> which is strange as I always compile with -j and never normally have any problem
Jan 11 14:52:30 <_trine> failed again
Jan 11 14:53:32 <_trine> http://pastebin.com/m755b54ae
Jan 11 15:11:25 <nbd> _trine: try updating again, looks like your bug was fixed (or supposed to be fixed) earlier today, in r19104
Jan 11 15:11:51 <nbd> the reason why it seems to get further without -j is that -j changes the order in which subdirectories are built
Jan 11 15:12:13 <jklee> for the WEP-WEP issue, I've tried to use this configuration http://pastebin.com/d64eeaaf3 without any success though
Jan 11 15:12:28 <jklee> the configuration is correct, right ?
Jan 11 15:12:45 <nbd> yes
Jan 11 15:12:56 <nbd> maybe we should simply consider this not supported then
Jan 11 15:13:11 <jow_laptop> "s:ppppp"
Jan 11 15:13:16 <jow_laptop> this didn't work for me
Jan 11 15:13:22 <jow_laptop> I had to hex encode it manually
Jan 11 15:13:32 <nbd> odd
Jan 11 15:13:36 <nbd> the s: syntax always worked for me
Jan 11 15:15:36 <jklee> allright, thanks, I'll also try with two keys of the same kind (HEX&HEX)
Jan 11 15:15:52 <_trine> nbd, I an at revision 19105
Jan 11 15:16:05 <_trine> nbd, I am at revision 19105
Jan 11 15:16:41 <_trine> and it does not compile with just using make without -j
Jan 11 15:19:31 <jow_laptop> _trine: this is caused by mkfs.ubi
Jan 11 15:19:41 <jow_laptop> it needs liblzo-dev and libuuid-dev on the host system
Jan 11 15:21:23 <_trine> since when ,,,, It all compiled yesterday
Jan 11 15:21:24 <jow_laptop> on Debian it is:  liblzo2-dev and  uuid-dev
Jan 11 15:21:29 <jow_laptop> since today
Jan 11 15:21:34 <_trine> ok
Jan 11 15:22:22 <_trine> I will try again with them
Jan 11 15:26:09 <_trine> it is compiling ok it seems now
Jan 11 15:26:21 <_trine> ty
Jan 11 15:40:48 <CIA-36> florian * r19106 /trunk/tools/mtd-utils/Makefile:
Jan 11 15:40:48 <CIA-36> [tools] do not build nor install mkfs.ubifs until we have it fixed
Jan 11 15:40:48 <CIA-36> (#6483)
Jan 11 15:52:32 <alesan> Kaloz, hi
Jan 11 16:15:54 <CIA-36> lars * r19107 /trunk/tools/mtd-utils/patches/101-ubifs-optional_lzo.patch: mtd-utils: Don't include lzo headers aswell...
Jan 11 17:18:15 <CIA-36> florian * r19108 /packages/net/ntpd/ (Makefile patches/100-ntpd_nmea.patch): [package] update ntpd to 4.2.6 (#6457)
Jan 11 19:48:36 <CIA-36> lars * r19109 /trunk/target/linux/xburst/config-2.6.32: [xburst] Disable CONFIG_MODVERSIONS
Jan 11 21:34:19 <netprince> is there a reasonable alternative for nslookup?  it doesn't seem to respect the [host] parameter... I would like to poll a specific dns server.
Jan 11 21:34:51 <xMff> dig / host
Jan 11 21:35:13 <xMff> should be in the bind-tools package
Jan 11 21:35:16 <netprince> cool, are they reasonably sized?
Jan 11 21:35:28 <xMff> yes as far as I remeber they're not big
Jan 11 21:35:35 <netprince> nice, thanks!
Jan 12 02:33:34 <stintel> xMff: I created a ticket for the bug with WEP encrypted client mode with mac80211 drivers, and attached a patch. https://dev.openwrt.org/ticket/6491
**** ENDING LOGGING AT Tue Jan 12 02:59:56 2010