**** BEGIN LOGGING AT Mon Aug 10 02:59:57 2009
Aug 10 04:12:15 <russell_> is there an arm922t target in openwrt already?
Aug 10 05:44:39 <dkostousov> Hi. I have built my own theme for LuCI. What I should do to add it as a package to OpenWRT firmware?
Aug 10 05:46:22 <loswillios> dkostousov: talk to xMff
Aug 10 05:56:42 <Bartman007> russell_: s3c24xx is arm920, orion uses a feroceon core which is arm926, kirkwood uses a feroceon/xscale derived core, but I think it's arm11
Aug 10 06:13:12 <frogonwheels> hey Bartman007 did you get to look at the apache changes?  any thoughts?
Aug 10 06:43:15 <Hopscotch> good morning
Aug 10 09:54:27 <dkostousov>  I have some problem with OpenWRT trunk and DLink DIR320. I configure WAN interface but ping goes to one address from the same network.
Aug 10 09:54:27 <dkostousov>  I flush iptables with iptables -F and iptables -t nat -F but nothing changes.
Aug 10 09:54:27 <dkostousov>  Ping to the neighbour host work fine. Ping to the GW fail. Ping to the any other host fails too.
Aug 10 09:55:50 <dkostousov> I can ssh to the WAN from the neighbour host after flushing iptables.
Aug 10 09:56:21 <dkostousov> But I can't connect via ssh from any another host.
Aug 10 10:05:27 <dkostousov> A tcpdump helps me to determinate that all packet goes from openwrt successfully. And answer packets successfully goes to openwrt from remote. But openwrt drop it
Aug 10 10:06:09 <frogonwheels> dkostousov: you might have more luck in #openwrt
Aug 10 10:06:41 <frogonwheels> dkostousov: I've gotta dash now, but I suggest you draw a mud-map of your config and pastebin it
Aug 10 10:08:13 <dkostousov> ifconfig result http://openwrt.pastebin.com/m5748db9e
Aug 10 10:08:28 <frogonwheels> I've read what you've sent a bunch of times, and still can't work out which machine is where, &c
Aug 10 10:09:29 <frogonwheels> you say 'configure WAN' interface - is it dhcp? what?  have you got modules required for iptables log target?
Aug 10 10:09:47 <frogonwheels> anyway, it's really a #openwrt discussion - I'm off for some dinner.
Aug 10 10:10:02 <dkostousov> /etc/config/network http://openwrt.pastebin.com/m5ecd4e11
Aug 10 10:12:01 <dkostousov> frogonwheels: ok
Aug 10 11:23:04 <frogonwheels> does anybody know if it's somehow possible that sizeof(void *)  is actually 4  in the gcc cross compiler??
Aug 10 11:23:19 <frogonwheels> erm.. for 64bit x86 that is
Aug 10 11:30:56 <frogonwheels> nope, it's definitely 8.  I just cross-compiled a 64bit test program  main() {printf("%d",sizeof(void*)}   copied it tothe target 64bit machine, and it printed 8
Aug 10 11:32:04 <frogonwheels> .. using essentially the same command-line from the config.log
Aug 10 11:32:23 <frogonwheels> http://pastebin.ca/1523521
Aug 10 11:32:40 <frogonwheels> http://pastey.net/120422
Aug 10 11:32:44 <frogonwheels> sorry the second one.
Aug 10 11:37:43 <frogonwheels> AH. This could do it:     checking size of void *... (cached) 4
Aug 10 11:41:25 <frogonwheels> where does ./configure cache stuff?
Aug 10 11:41:30 <frogonwheels> (That was after a clean,compile)
Aug 10 11:44:59 <dkostousov> Are here any openwrt kernel developers?
Aug 10 12:22:46 <frogonwheels> aaaaaaaaahhhh..   include/site/x86_64-linux  has ac_cv_sizeof_void_p=4  !!!!!!!!
Aug 10 12:23:04 <frogonwheels> that's a top-level define.
Aug 10 12:23:26 <frogonwheels> I mean.. trunk/include/site/  ..
Aug 10 12:25:04 <frogonwheels> hazzah!
Aug 10 15:18:01 <mun24> I changed lan subnet from 192.168.1.0 to 192.168.4.0, and then it stopped assigning any ip addresses, I reverted back to 192.168.1.0, still not working
Aug 10 15:18:05 <mun24> I am using trunk
Aug 10 15:18:58 <xMff> check logread
Aug 10 15:22:16 <mun24> logread doesn't give any error
Aug 10 15:38:48 <xMff> restart dnsmasq and check logread again
Aug 10 16:00:37 <dkostousov_h> I have some problem with WAN interface on dlink dir-320 (openwrt trunk, kamikaze 8.09.1). It silently drops any incoming packets.
Aug 10 16:01:07 <dkostousov_h> How I can turn on debugging and try to find problem?
Aug 10 16:49:52 <dkostousov_h> where I can get channel logs?
Aug 10 18:26:20 <CIA-36> hauke * r17208 /trunk/package/ (kernel/modules/fs.mk mac80211/Makefile):
Aug 10 18:26:20 <CIA-36> Correct bool expressions in dependencies.
Aug 10 18:26:20 <CIA-36> This fixes #5627
Aug 10 20:11:22 <kupesoft> Kaloz: You bumped opkg in r17152 and I've been getting a weird non-fatal error since, I emailed -devel, did you get a chance to take a look?
Aug 10 20:11:56 <Dogge_> lol
Aug 10 20:17:16 <[florian]> Dogge_: usual sarcacism?
Aug 10 20:18:11 <xMff> kupesoft: are you sure it's the update causing the problem or the switch from arch to board in pkg names?
Aug 10 20:18:45 <kupesoft> 99% sure it's the update, since I got the error using a local repo
Aug 10 20:18:56 <kupesoft> I've tracked it down, basically /etc/opkg/opkg.conf gets parsed twiced every time opkg starts up
Aug 10 20:19:02 <kupesoft> I just don't know why or how to stop it :P
Aug 10 20:19:15 <kupesoft> /etc/opkg/opkg.conf [sic]
Aug 10 20:19:31 <kupesoft> xMff: tried it?
Aug 10 20:19:48 <xMff> kupesoft: nope but got similar reports from other users
Aug 10 20:20:06 <xMff> actually the error message appears twice in the source as well
Aug 10 20:20:32 <kupesoft> xMff: yeah twice in the source, but for basically the same error. One is parsing gzip'd sources and one is plantext, if I recall correctly
Aug 10 20:20:42 <xMff> yeah, just looked
Aug 10 20:22:35 <kupesoft> opkg_conf_parse_file() in libopkg/opkg_conf.c gets called twice. That's as far as I traced it.
Aug 10 20:23:37 <xMff> do you have an old /etc/opkg.conf lying around?
Aug 10 20:24:23 <xMff> hm wait
Aug 10 20:24:43 <kupesoft> Nope
Aug 10 20:24:50 <kupesoft> Everyone's on a clean build
Aug 10 20:24:50 <xMff>  "%s/etc/opkg/*.conf", args->offline_root
Aug 10 20:25:01 <xMff> passed to glob later on
Aug 10 20:25:32 <xMff> then iterated over the results
Aug 10 20:26:43 <xMff> maybe it thinks / is the offline root and wants to read the config from there as well?
Aug 10 20:26:57 <kupesoft> peut etre
Aug 10 20:28:25 <kupesoft> I don't have a recent build running to poke around with settings
Aug 10 20:28:57 <xMff> me neither
Aug 10 20:31:47 <kupesoft> I'll make an x86 one with debugging for qemu
Aug 10 20:31:52 <kupesoft> And get to the bottom of it
Aug 10 20:31:54 <kupesoft> :)
**** ENDING LOGGING AT Mon Aug 10 21:05:15 2009

**** BEGIN LOGGING AT Mon Aug 10 23:46:19 2009
Aug 10 23:48:12 <CIA-36> jow * r17217 /trunk/package/base-files/ (3 files in 3 dirs): [package] base-files: move ipv6 module loading from setup_interface() to addif() in the hotplug call, this ensures that ipv6 is loaded before any interfaces or aliases with ip6addr option are configured (#5356)
Aug 10 23:53:29 <frogonwheels> xMff:  no probs.  I'll have a small collection related to 64bit.  This one is pretty obvious (and I've tested its validity), but some of the others I'm not sure I've quite got the right conditions.
Aug 10 23:54:46 <CIA-36> jow * r17218 /branches/8.09/package/base-files/ (3 files in 3 dirs): [8.09] merge fix from r17217, bump patch level
Aug 11 00:02:09 <xMff> frogonwheels: can you paste the patch somewhere?
Aug 11 00:03:14 <frogonwheels> xMff: I will later.  There's no real rush - as there's no 64bit x86 targets yet :)
Aug 11 00:03:47 <frogonwheels> xMff:  Actually, at the moment the 64bit has been done as a sub-target of x86, and I'm wondering if it should be a new target.
Aug 11 00:04:32 <xMff> hm other arches have big and low endian subtargets, why not a 32 and 64bit subtarget then, but I'm not in the position to decide this
Aug 11 00:10:39 <frogonwheels> that sounds ok then.
Aug 11 00:10:50 <frogonwheels> I mean, it's working fine, so I'm happy with that.
Aug 11 00:12:28 <frogonwheels> xMff: did you see my apache changes come through on the list?
Aug 11 00:12:38 <xMff> subject?
Aug 11 00:13:06 <frogonwheels> [Patch] Switch to dynamic-linked apache with module packages
Aug 11 00:13:39 <frogonwheels> still no movement on that.  anyway I gg and take kids to school, go visit the building site & stuff
Aug 11 00:15:41 <xMff> will take a look
Aug 11 00:19:46 <kupesoft> anyone wanna apply the apf package, should I UCI-ify startup?
Aug 11 00:43:19 <xMff> "Apple Filing Protocol" ?
Aug 11 01:24:08 <mankash> how to include tcpdump
Aug 11 01:25:41 <xMff> ./scripts/feeds -d y install tcpdump
Aug 11 01:31:04 <mankash> xMff: Now I am at home, any other way to see why openwrt is unable to assign ip address to lan clients
Aug 11 01:31:32 <xMff> do you see incoming dhcp requests in logread?
Aug 11 01:31:37 <mankash> nope
Aug 11 01:32:23 <mankash> but when I do ifconfig eth0 down then I can see on winxp  disconnection icon in system tray
Aug 11 01:32:27 <xMff> what's the output of echo $(sed -nep /proc/$(pidof dnsmasq)/cmdline)  ?
Aug 11 01:32:42 <mankash> and when do ifconfig etho up it start trying to get ip address
Aug 11 01:35:34 <mankash> /usr/sbin/dnsmasq -K -D -y -Z -b -E -s lan -S /lan/ --log-dhcp
Aug 11 01:35:47 <mankash> I addedd --log-dhcp
Aug 11 01:36:00 <xMff> it did not configured a lan address range...
Aug 11 01:37:01 <mankash> then how to fix it
Aug 11 01:37:24 <xMff> how does your /etc/config/dhcp looks like?
Aug 11 01:37:47 <xMff> and your /etc/config/network
Aug 11 01:38:06 <mankash> i havent toucj dhcp file
Aug 11 01:38:53 <xMff> try adding option 'force' '1' to the lan section in /etc/config/dhcp
Aug 11 01:39:05 <mankash> ok
Aug 11 01:41:36 <mankash> no result
Aug 11 01:41:45 <mankash> do I need to restart dnsmasq
Aug 11 01:41:50 <mankash> I restarted network
Aug 11 01:41:58 <xMff> yes you need to restart dnsmasq
Aug 11 01:42:03 <mankash> ok
Aug 11 01:44:40 <mankash> openwrt does not support vlan in KS8995 switch, can I remove option bridge and then try
Aug 11 01:45:39 <xMff> do you have two eth interfaces?
Aug 11 01:45:46 <mankash> yes
Aug 11 01:46:02 <mankash> eth0 and eth1 and wiif
Aug 11 01:49:20 <xMff> the dnsmasq command line still does not contain a range?
Aug 11 01:50:23 <xMff> should look like this: /usr/sbin/dnsmasq -K -D -N -y -Z -b -E -s lan -S /lan/ -l /tmp/dhcp.leases -r /tmp/resolv.conf.auto --dhcp-range=lan,192.168.1.100,192.168.1.250,255.255.255.0,12h
Aug 11 01:51:15 <mankash> ok thx
Aug 11 01:51:20 <xMff> seems like everything after your "--log-dhcp" is discarded
Aug 11 01:51:36 <xMff> for some reason, where did you specify it?
Aug 11 01:51:56 <mankash> when I restarted it
Aug 11 01:52:08 <xMff> ?
Aug 11 01:52:19 <xMff> so you executed dnsmasq manually ?
Aug 11 01:52:21 <mankash> dnsmasq
Aug 11 01:52:37 <mankash> yes whenit wasn't working
Aug 11 01:53:05 <xMff> well, the command line was not complete
Aug 11 01:53:16 <mankash> strange
Aug 11 01:54:02 <xMff> the init script usually handles this, it expects an interface called "lan" in the network config
Aug 11 01:54:19 <xMff> sometimes it refuses to do dhcp on the lan if it thinks there's another dhcp server
Aug 11 01:54:42 <xMff> but option force 1 should take care of that
Aug 11 01:54:42 <mankash> there is no dhcp server
Aug 11 01:55:27 <mankash> let me create the RAM image, may be it is good
Aug 11 01:57:15 <mankash> any other suggestion
Aug 11 01:59:23 <xMff> what's the current dnsmasq command line?
Aug 11 01:59:48 <xMff> have you tried to disable the firewall?
Aug 11 01:59:55 <mankash> nope
Aug 11 02:03:46 <mankash> one thing I have noticed that dnmasq is run under nobody whereas other process run under root
Aug 11 02:04:04 <mankash> may be dnsmasq is unable to read the file
Aug 11 02:04:07 <xMff> this is normal
Aug 11 02:04:15 <mankash> ok
Aug 11 02:04:40 <mankash> can I specify the address range manually and then see if it works
Aug 11 02:04:53 <xMff> /usr/sbin/dnsmasq -K -D -N -y -Z -b -E -s lan -S /lan/ -l /tmp/dhcp.leases -r /tmp/resolv.conf.auto --dhcp-range=lan,192.168.1.100,192.168.1.250,255.255.255.0,12h
Aug 11 02:05:22 <mankash> ok
Aug 11 02:05:45 <xMff> oh wait
Aug 11 02:05:47 <xMff> typo
Aug 11 02:06:03 <xMff> no, it's okay
Aug 11 02:06:11 <mankash> ok
Aug 11 02:14:17 <CIA-36> nbd * r17219 /trunk/package/madwifi/patches/ (5 files): madwifi: fix compile error on kernels without net_device api compatibility
Aug 11 02:14:29 <CIA-36> nbd * r17220 /trunk/package/madwifi/patches/442-ibss_rx_filter.patch: fix software rx filter for ibss mode (based on a patch by harald schiöberg)
Aug 11 02:14:35 <CIA-36> nbd * r17221 /trunk/package/madwifi/patches/414-txpower.patch: madwifi: fix compiler warning
Aug 11 02:18:54 <mankash> xMff: no result
Aug 11 02:20:01 <xMff> what's in logread?
Aug 11 02:20:24 <mankash> nothing
Aug 11 02:21:06 <xMff> not even startup messages from dnsmasq?
Aug 11 02:24:49 <mankash> I see one error Jan  1 00:38:21 OpenWrt daemon.err dnsmasq-dhcp[3193]: failed to read /etc/ethers: No such file or directory
Aug 11 02:26:14 <mankash> http://openwrt.pastebin.com/m49106477
Aug 11 02:26:35 <xMff> Jan  1 00:38:21 OpenWrt daemon.info dnsmasq-dhcp[3193]: DHCP, IP range 192.168.1.100 -- 192.168.1.250, lease time 12h
Aug 11 02:27:01 <mankash> you can check the pastebin
Aug 11 02:27:17 <xMff> I checked the pastebin
Aug 11 02:27:33 <xMff> your dnsmasq has now a range for dhcp
Aug 11 02:27:39 <mankash> yeah
**** ENDING LOGGING AT Tue Aug 11 02:59:57 2009