**** BEGIN LOGGING AT Fri Feb 08 02:59:56 2008
Feb 08 03:48:23 <nbd> thepeople: the more i review this crappy codebase, the more memory allocation bugs i find
Feb 08 03:48:37 <nbd> thepeople: might have a version soon that is a bit more robust against this stuff
Feb 08 03:51:05 <nbd> thepeople: want to test a preliminary patch?
Feb 08 03:51:47 <thepeople> nbd: sure
Feb 08 03:52:02 <nbd> ok, let me just check if it compiles ;)
Feb 08 03:52:30 <nbd> thepeople: http://nbd.name/331-memory_alloc.patch
Feb 08 03:52:52 <nbd> compiles, *theoretically* shouldn't break anything :)
Feb 08 03:54:12 <thepeople> ok, I will test its theoretical nonbrokeness
Feb 08 03:54:18 <nbd> :)
Feb 08 03:54:19 <nbd> thanks
Feb 08 03:59:19 <thepeople> nbd: on the meraki's can I just update madwifi or do I need to reflash?
Feb 08 03:59:30 <nbd> you can update madwifi
Feb 08 03:59:36 <thepeople> ok
Feb 08 04:19:07 <nbd> thepeople: does it start without oopsing? :)
Feb 08 04:21:17 <thepeople> my box is still compiling, otherwise I have to restart my laptop back into linux
Feb 08 04:21:55 <nbd> ah
Feb 08 04:46:55 <nbd> thepeople: found a problem with my patch
Feb 08 04:46:58 <nbd> thepeople: please reload it
Feb 08 04:47:09 <nbd> there was a potential hidden double-free
Feb 08 04:47:15 <nbd> which theoretically shouldn't happen
Feb 08 04:47:19 <nbd> but is checked for now
Feb 08 04:49:18 <thepeople> ok
Feb 08 04:57:52 <thepeople> nbd: can you send me the link for that patch again
Feb 08 04:58:02 <nbd> http://nbd.name/331-memory_alloc.patch
Feb 08 04:58:07 * thepeople is tired of waiting for my box to compile it
Feb 08 05:06:11 <thepeople> it installed and runs, my clients are connected
Feb 08 05:07:23 <nbd> ok
Feb 08 05:07:30 <nbd> let's see how iperf is doing ;)
Feb 08 05:10:42 <thepeople> unfortunately I can't tonight, the client is on a shed and I would wake the people up. I will iperf it in the morning, I have a couple of camera viewing windows open to see if it will die that way
Feb 08 05:10:59 <nbd> ok, i think i'll just commit the patch
Feb 08 05:11:12 <nbd> it was ack'd by another madwifi dev, so it should be ok
Feb 08 05:11:28 <thepeople> ok
Feb 08 05:11:56 <thepeople> I need to put a linux box behind it
Feb 08 05:11:59 <thepeople> :-)
Feb 08 05:13:04 <CIA-27> nbd * r10416 /trunk/package/madwifi/patches/ (315-power_drop_PR1695.patch 330-beaconcal.patch): madwifi: add ap mode specific improvements to the calibration process. reported to produce a noticeable increase in rx/tx strength
Feb 08 05:13:07 <CIA-27> nbd * r10417 /trunk/package/madwifi/patches/331-memory_alloc.patch: add a fix for memleaks caused by failed memory allocations
Feb 08 05:13:40 <thepeople> hmm, your first patch description look interesting
Feb 08 05:14:27 <nbd> yeah, and there's more to come for sta mode as well
Feb 08 05:14:41 <nbd> did lots of low level debugging on this one lately
Feb 08 05:14:46 <nbd> until i finally figured it out
Feb 08 05:16:12 <thepeople> hopefully it fixes some performance issues
Feb 08 05:16:24 <nbd> i think it will
Feb 08 05:18:36 <thepeople> I will be testing of the wireless this weekend, so I will let you know how things shape up for me
Feb 08 05:29:07 <thepeople> nbd: I can create another error http://pastebin.ca/895946 this is the first card with no traffic on it and me flooding the network on the second card
Feb 08 05:29:21 <thepeople> odd that it would create errors on the first card
Feb 08 05:29:53 <nbd> does it lock up the interface?
Feb 08 05:31:38 <thepeople> looks like it might have, it deauthed my client on that interface
Feb 08 05:31:42 <thepeople> if I stop flooding now it came back
Feb 08 05:32:31 <nbd> ok
Feb 08 05:32:40 <thepeople> the ap shows that the client is connected but I can not ping it
Feb 08 05:32:42 <nbd> temporary lockup is much better than permanent lockup
Feb 08 05:33:57 <thepeople> ok it finally came all the way back and responds to a ping
Feb 08 05:34:37 <thepeople> still odd that the first card would get affected by flooding of the second
Feb 08 05:35:27 <nbd> yeah
Feb 08 05:35:32 <nbd> well, not really
Feb 08 05:35:43 <nbd> the flooding of the first one eats up lots of memory
Feb 08 05:35:56 <nbd> which would be needed to keep the other one working
Feb 08 05:36:12 <thepeople> ok, yea I could see that
Feb 08 05:37:53 <thepeople> well I will let the network be with normal loads on it and check to see if it is still alive come morning
Feb 08 14:57:18 <CIA-27> florian * r10418 /packages/libs/libnfnetlink/Makefile: Cleanup makefile
Feb 08 14:58:08 <CIA-27> florian * r10419 /packages/ipv6/babel/Makefile: Upgrade to babel 0.8
Feb 08 15:00:35 <CIA-27> florian * r10420 /packages/libs/libnetfilter-queue/Makefile: Cleanup makefile
Feb 08 17:17:48 <CIA-27> matteo * r10421 /trunk/target/linux/ar7/ (4 files in 3 dirs): [AR7] Update kernel to 2.6.24
Feb 08 17:27:05 <CIA-27> matteo * r10422 /trunk/target/linux/ar7/config-2.6.24: [AR7] build cpmac driver
Feb 08 17:33:43 <thepeople> nbd: so far with your patch I have not had any errors
Feb 08 17:33:54 <thepeople> under regular load
Feb 08 17:34:25 <nbd> great
Feb 08 17:34:27 <thepeople> normally I would have by now
Feb 08 18:11:20 <witchdoc> re
Feb 08 18:18:15 <CIA-27> matteo * r10423 /trunk/package/ar7-atm/patches/100-compile_fix.patch: fix AR7 DSL with 2.6.24
Feb 08 19:41:17 <CIA-27> matteo * r10424 /trunk/target/linux/ar7/patches-2.6.24/140-cpmac_fix.patch: cpmac: reapply [9664]
Feb 08 20:15:27 <CIA-27> matteo * r10425 /trunk/package/acx/patches/003-2.6.24-compat.diff: backport acx100 2.6.24 fixes from SVN
Feb 08 20:39:05 <pedrosa> Hello
Feb 08 20:39:09 <pedrosa> anyone is working in b43 wifi script?
Feb 08 20:39:11 <wrt> hi
Feb 08 20:39:35 <thepeople> pedrosa: I am not aware of anyone working on one
Feb 08 20:41:27 <pedrosa> i'v coded small script to configure this board
Feb 08 20:41:42 <pedrosa> but only in open mode
Feb 08 20:43:59 <thepeople> you could submit it as a patch to get it started https://dev.openwrt.org/wiki/SubmittingPatches
Feb 08 20:44:43 <pedrosa> ok
Feb 08 20:49:08 <pedrosa> tnks
Feb 08 22:24:40 <nabcore> Can any devs with svn write access apply the git patches from Linus' tree to the AR7's cpmac? https://dev.openwrt.org/ticket/3124
Feb 08 22:25:55 <thepeople> nabcore: you will to talk with matteo
Feb 08 22:26:07 <nabcore> ok
**** ENDING LOGGING AT Sat Feb 09 02:59:57 2008