**** BEGIN LOGGING AT Sat Dec 05 02:59:57 2009 Dec 05 03:48:54 anyone know if git provides a way of getting the latest commit similar to svn info or hg id without having to git clone the repository? Dec 05 03:56:53 swalker: git log Dec 05 03:58:57 commit a1ca4142c765bb354a68b9aa67298a4c6130dcd6 Dec 05 03:58:57 Author: nbd Dec 05 03:58:57 Date: Fri Dec 4 22:18:36 2009 +0000 Dec 05 03:58:57 oops, forgot one small change in the last commit Dec 05 03:58:57 git-svn-id: svn://svn.openwrt.org/openwrt/trunk@18640 3c298f89-4303-0410-b956-a3cf2f4a3e73 Dec 05 04:15:04 doesn't work without first cloning afaict Dec 05 04:41:16 perl build is dying on GDBM_File.c Dec 05 04:41:34 somewhere between r18567 and r18640 Dec 05 04:41:53 (on a wgt634u compile) Dec 05 05:02:57 swalker: whoops didnt see that part of thwe question, hrmmm Dec 05 05:16:18 hmm. no gdbm.h in staging_dir/target-mipsel_uClibc-0.9.30.1/usr/include Dec 05 05:31:08 weird, it shows for mips and x86 (in older staging trees) Dec 05 05:34:54 dingo: git ls-remote works although my local git is missing the 1.6.5.3 ls-remote http bugfix Dec 05 05:45:58 rebuilding gdbm manually installed it in staging Dec 05 05:46:24 missing dependency Dec 05 06:04:52 or something Dec 05 06:17:38 [florian]: hmm, sometimes next week, I may have a working driver for the realtek switch, maybe even with vlan support Dec 05 07:53:59 xMff: your /etc/init.d/netconfig fix solved my problem Dec 05 09:59:35 here's a crappy patch that fixes the syslog-ng3 build for wgt634u, proper fix probably looks a little different: http://pastebin.ca/1702883 Dec 05 10:03:32 gzip build failure on: getopt_int.h:22: error: conflicting types for '_getopt_internal' Dec 05 12:39:30 jow * r18657 /trunk/target/linux/brcm-2.4/base-files/etc/init.d/netconfig: [brcm-2.4] fix typo in switch config migration script Dec 05 12:54:32 florian * r18658 /trunk/package/ (hostap-driver/Makefile kernel/modules/wireless.mk): [package] allow building hostap-driver and its dependencies on 2.6.32 (#6305) Dec 05 13:52:21 florian * r18659 /packages/libs/curl/ (3 files in 2 dirs): [package] revert r18554, until we fix curl compilation on all platforms Dec 05 13:52:23 florian * r18660 /packages/libs/libtorrent/Makefile: [package] update libtorrent to 0.12.6 (#6263) Dec 05 13:52:26 florian * r18661 /packages/net/rtorrent/Makefile: [package] update rtorrent to 0.8.6 (#6263) Dec 05 14:21:16 mmmh. is luci broken in trunk somehow? Dec 05 14:21:29 roh: what are the symptoms? Dec 05 14:30:38 every second page is white and now i get zero sized replys Dec 05 14:31:07 roh: you can workaround this: /etc/init.d/lucid stop; /etc/init.d/lucid disable; /etc/init.d/httpd restart Dec 05 14:31:09 also.. is ath5k useful for 5ghz station mode on a wl500gp ? Dec 05 14:32:22 aah.. yeah.. that helped Dec 05 14:33:06 the daemon has some issues since the lates lua updates Dec 05 14:34:43 how do i configure 5ghz wifi? Dec 05 14:36:43 hmm, should be enough to just sepcify a 5ghz channel Dec 05 14:52:59 xMff you know if wl500g can have more than 2 vlans on the ethernet? Dec 05 14:55:05 its eth0.0 on the 4 'lan' ports and eth0.1 on 'wan' .. i'd like to have one of the lan in a extra vlan, not shared with the others Dec 05 14:55:19 sure Dec 05 14:55:31 just make an additional vlan and take a port ouf of lan Dec 05 14:55:41 you can have up to 16 vlans Dec 05 14:56:03 well.. i find weird not seeing the default config at all Dec 05 14:56:17 no switch config? Dec 05 14:57:44 hm.. the hw-switch? Dec 05 14:58:08 i mean.. the linux part is easy... what tool configs the switch chiip? Dec 05 14:58:17 depends Dec 05 14:58:30 on brcm-2.4 you need to poke /proc/switch/ Dec 05 14:58:35 2.6 Dec 05 14:58:43 on all other arches it is "swconfig" if supported Dec 05 14:59:09 swconfig is used to configure the hardware Dec 05 14:59:23 hm.. does it have a list? Dec 05 14:59:31 in order to actually use the hardware vlans you need to define vlan interfaces with vconfig Dec 05 14:59:47 roh: it uses a small netlink based kernel api for configuring switches Dec 05 15:00:21 and we have various kernel drivers and a small core that implement it Dec 05 15:01:10 that reminds me... i really should submit that to netdev@ ;) Dec 05 15:04:10 hm. Dec 05 15:04:37 well.. i am a bit puzzled.. the switch is on the 5 external ports and on a 6th which goes to mii on eth0 on the soc Dec 05 15:04:45 where does eth1 connect to? Dec 05 15:05:19 maybe it's unconnected Dec 05 15:05:39 hm... weird... nobody ever asked that ;) Dec 05 15:08:27 jow * r18662 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: [package] mac80211: fix wpa handling in sta mode Dec 05 15:09:22 jow * r18663 /trunk/package/wpa_supplicant/files/wpa_supplicant.sh: [package] wpa_supplicant: relax parsing of encryption mode, drop support for uppercase modes Dec 05 15:10:47 how do i config tagged/untagged ? Dec 05 15:11:33 putting a 't' after a port in the config marks it as tagged Dec 05 15:11:37 in a vlan port list Dec 05 15:11:48 jow * r18664 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: [package] mac80211: fix a shell syntax error Dec 05 15:15:09 nbd there is a star for vlan0 and prt 5 and no star for vlan1 Dec 05 15:15:27 star means assigning the primary vlan for a port Dec 05 15:15:35 network.eth0_0.ports=1 2 3 4 5* and network.eth0_1.ports=0 5 Dec 05 15:15:39 some switches need this to be set on the cpu port for some reason Dec 05 15:15:50 weird. Dec 05 15:16:03 because one is tagged and one untagged then Dec 05 15:16:19 oh, wait Dec 05 15:16:23 that is broadcom, right? Dec 05 15:16:37 ack Dec 05 15:16:47 5325 Dec 05 15:16:52 wl500gP Dec 05 15:16:55 v1 Dec 05 15:16:58 the driver defaults to tagged for the cpu port Dec 05 15:17:03 legacy stuff Dec 05 15:17:04 ;) Dec 05 15:17:23 yes. makes sense.. but why not for the wan port too? Dec 05 15:17:41 looks like lan comes in tagged and wan untagged by default Dec 05 15:17:42 why not what for the wan port? Dec 05 15:17:49 if * is equvalent to t Dec 05 15:17:52 it's not Dec 05 15:18:04 * assigns the primary vlan id Dec 05 15:18:08 t marks a port as tagged Dec 05 15:18:15 well... thats what i pulled from the default config Dec 05 15:18:21 eh.. so its untagged?! Dec 05 15:18:31 i just said that the cpu port defaults to being tagged ;) Dec 05 15:18:33 how does it pull wan and lan from each other then? Dec 05 15:19:07 so it's equivalent to writing 1 2 3 4 5t* and 0 5t Dec 05 15:19:21 nbd ok.. but why dont i see that in uci then? Dec 05 15:19:32 as i said... legacy stuff Dec 05 15:19:48 do i need a star on my port 4 vlan id 2 then too? Dec 05 15:19:51 the broadcom driver does it implicitly and we haven't made it explicit yet Dec 05 15:19:56 other platforms are more clear Dec 05 15:20:06 i think only the cpu port needs a * Dec 05 15:20:46 nbd isnt the cpu port 5 for all vlans defined? Dec 05 15:21:14 sure, cpu port is always 5 on these things Dec 05 15:21:43 and why is there no * for wan then? Dec 05 15:22:18 because each port can only have one primary vlan Dec 05 15:56:56 [florian]: ping Dec 05 16:01:13 hhm hm.. ~. Dec 05 16:01:29 somehow the vlanconfig/switchconfig doesnt happen Dec 05 16:01:54 can it be that the initscript just uses whats hardcoded in it by model? Dec 05 16:02:52 roh: it's not supported with all types of switches Dec 05 16:03:52 rtz2 bc5325 in a wl500gP V1 Dec 05 16:04:57 roh: sorry, no idea where exactly Dec 05 16:05:27 roh: the inner workings of this vlan stuff are a mystery to me :/ Dec 05 17:12:53 nbd * r18665 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: mac80211: set the country code through iw reg, setting it only through hostapd apparently does not work Dec 05 17:12:58 nbd * r18666 /trunk/package/mac80211/ (17 files in 2 dirs): mac80211: update to compat-wireless 2009-12-05 and fix a critical bug in ath9k tx status reporting Dec 05 17:45:29 fnord23!42 Dec 05 23:12:29 roh: perhaps change your password? Dec 06 01:56:54 I just got a little aggressive on the forums :( Dec 06 02:02:58 ? Dec 06 02:06:08 https://forum.openwrt.org/viewtopic.php?id=6256 Dec 06 02:06:13 Maybe I have a rage problem? Dec 06 02:19:15 kupesoft: I wouldn't call this aggressive Dec 06 02:19:21 and he asked for it Dec 06 02:20:11 :) **** ENDING LOGGING AT Sun Dec 06 02:59:58 2009