**** BEGIN LOGGING AT Fri May 14 02:59:56 2010
May 14 04:25:47 <builder> build #41 of x86 is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/41
May 14 06:16:08 <builder> build #43 of xburst is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/43
May 14 07:00:32 <Mr_Smoke> mo'in
May 14 08:38:01 <CIA-33> juhosg * r21443 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-nbg460n.c:
May 14 08:38:01 <CIA-33> ar71xx: fix nbg460n mtd partitions
May 14 08:38:01 <CIA-33> This fixes the NBG460N/550N/550NH mtd partitions.
May 14 08:38:01 <CIA-33> Signed-off-by: Michael Kurz <michi.kurz@googlemail.com>
May 14 08:38:42 <CIA-33> juhosg * r21444 /trunk/target/linux/ar71xx/base-files/sbin/wget2nand: (log message trimmed)
May 14 08:38:42 <CIA-33> ar71xx: Make wget2nand fail if copying the kernel fails and use the correct file extension for the rootfs image
May 14 08:38:42 <CIA-33> wget2nand didnt check the exit status after copying the kernel, if the
May 14 08:38:42 <CIA-33> copying failed for some reason ( for example not enougs space on the
May 14 08:38:42 <CIA-33> kernel partition) it simply continued extracting the rootfs.
May 14 08:38:42 <CIA-33> I also changed the filename, which wget2nand trys to download (
May 14 08:38:43 <CIA-33> .tar.gz instead of .tgz ).
May 14 08:46:13 <acoul> hello
May 14 09:12:51 <rtz2> [florian]: ping
May 14 09:20:36 <CIA-33> juhosg * r21445 /trunk/tools/firmware-utils/src/mkzynfw.c:
May 14 09:20:36 <CIA-33> firmware-utils/mkzynfw: don't use next_offset twice (closes #7273)
May 14 09:20:36 <CIA-33> The test if a file can fit in an image considers the current offset
May 14 09:20:36 <CIA-33> twice. So a image that would fit is too big.
May 14 09:20:36 <CIA-33> Thanks-to: Michael Kurz <michi.kurz@googlemail.com>
May 14 09:20:38 <CIA-33> juhosg * r21446 /trunk/tools/firmware-utils/src/ (mkzynfw.c zynos.h):
May 14 09:20:38 <CIA-33> firmware-utils/mkzynfw: add support for the NBG460N board
May 14 09:20:38 <CIA-33> Adding NBG460N/550N/550NH boards to mkzynfw.
May 14 09:20:39 <CIA-33> Signed-off-by: Michael Kurz <michi.kurz@googlemail.com>
May 14 09:25:58 <wolfspraul> I get a segfault in the scripts/config/mconf binary, how can I turn on debug symbols?
May 14 09:27:21 <Mr_Smoke> Is there a man page for wlc somewhere ?
May 14 09:36:03 <CIA-33> zandbelt * r21447 /packages/net/asterisk-1.6.x/Makefile: [packages] asterisk-1.6.x: add a bunch of modules through templating
May 14 09:47:42 <xMff> Mr_Smoke: no
May 14 09:49:33 <Mr_Smoke> Ok
May 14 09:50:03 <Mr_Smoke> I think this broadcom.sh issue all boils down to the way the script counts the interfaces
May 14 09:50:18 <Mr_Smoke> There is a new switch/case in the script, I think it's part of the problem
May 14 09:56:32 <Mr_Smoke> xMff: would you have a couple of minutes to spare ? regarding issue 7335
May 14 09:56:48 <xMff> sure
May 14 09:56:57 <Mr_Smoke> Thanks
May 14 09:57:14 <Mr_Smoke> I sent an email on openwrt-devel yesterday about it
May 14 09:57:36 <Mr_Smoke> I'm willing to help, but I am probably not wlc-fluent yet, so I'm not sure where to start
May 14 09:58:27 <Mr_Smoke> As you can see form the ticket, c.shore suggested a patch that (I think) causes mssid to be set to 1 when there's an AP + WDS (otherwise it is not set and WDS alone doesn't work)
May 14 09:58:51 <Mr_Smoke> However, with his patch, I can't seem to get the WDS up
May 14 09:59:18 <xMff> well wlc does nothing more than exposing various driver ioctls
May 14 09:59:33 <xMff> basically a reverse engineered wl
May 14 09:59:39 <Mr_Smoke> Yeah, so it's basically the way broadcom.sh translates the config files into wlc input
May 14 09:59:52 <xMff> the right order of ioctls invokations is documented nowhere
May 14 10:00:04 <Mr_Smoke> Gees, thank you BroadCom :/
May 14 10:00:16 <xMff> it is all knowledge gathered by examining vendor firmwares
May 14 10:00:52 <xMff> ... and trial and error
May 14 10:01:12 <Mr_Smoke> Indeed.
May 14 10:01:18 <xMff> you can trace the commands the 7.09 kamikaze does and then compare them against 10.03
May 14 10:01:33 <Mr_Smoke> I think jow made me do that yes
May 14 10:01:46 <Mr_Smoke> I'll probably have to do it again
May 14 10:02:47 <rtz2> Mr_Smoke: xMff=jow
May 14 10:02:54 <Mr_Smoke> oh
May 14 10:02:57 <Mr_Smoke> Haha
May 14 10:03:08 <Mr_Smoke> even jow.laptop ?
May 14 10:04:08 <xMff> yea
May 14 10:04:57 <Mr_Smoke> Well, nice to meet you again :P
May 14 10:05:23 <Mr_Smoke> I'll flash a third unit with 7.09
May 14 10:10:23 <Mr_Smoke> It's like the WRT54GLs are breeding in here ...
May 14 10:12:32 <rtz2> acoul: ping
May 14 10:16:11 <rtz2> Hauke: ping
May 14 10:17:54 <Hauke> rtz2: pong
May 14 10:17:59 <acoul> pong
May 14 10:19:05 <rtz2> acoul, Hauke: this actually concerns both of you
May 14 10:19:25 <rtz2> it's about the crash in b43 with the brcm4716 patches
May 14 10:19:44 <rtz2> I have no clue, where exactly this comes from
May 14 10:20:52 <rtz2> so, if you none of you has the time to track it down, could I ask for serial access to a WL-500GP ?
May 14 10:21:19 <Hauke> rtz2: I have looked into the code but only for ~5 minutes ;-)
May 14 10:21:59 <Hauke> rtz2: I could track it down in the evening
May 14 10:22:25 <rtz2> Hauke: basically, the problem is somewhere in the b43_probe function, not b43info
May 14 10:22:49 <rtz2> Hauke: key word is "somewhere", this is a pretty big think (with all subfunctions)
May 14 10:23:01 <Hauke> rtz2: It would be nice to split the patches into one part changing the ssb code without adding the new fetaures and an other patch adding the new features
May 14 10:23:37 <Hauke> the b43_probe function calls some ssb functions
May 14 10:24:13 <rtz2> Hauke: I will see, what I can do, but it may take 1-2 days
May 14 10:24:58 <rtz2> Hauke: as for the ssb functions, it calls a number of them, and I don't understand how this could happen
May 14 10:27:15 <rtz2> Hauke: as for the no new features part, do you mean no functionality change or no new functionality?
May 14 10:40:07 <Mr_Smoke> xMff: I think I understand now. In fact, even in Kamikaze 7.09, one *cannot* define a WDS on its own (ie without an AP alongside it), and this is still true in 10.03
May 14 10:40:19 <Mr_Smoke> Maybe there should be a reference to this fact somewhere
May 14 10:40:40 <Mr_Smoke> I admit that WDS without AP might be a little strange, but who knows.
May 14 10:48:17 <rtz2> Hauke: sorry, I got disconnected, did you write anything?
May 14 10:50:54 <Hauke>  rtz2: I have not write anything
May 14 12:41:52 <CIA-33> craigc * r21448 /packages/net/multiwan/ (Makefile files/etc/config/multiwan files/usr/bin/multiwan):
May 14 12:41:53 <CIA-33> [packages] multiwan:
May 14 12:41:53 <CIA-33> - updated revision
May 14 12:41:53 <CIA-33> - overhauled probability matrix for netfilter route selection
May 14 12:41:53 <CIA-33> - added configuration options for specifying method of traffic distribution per default route/rule/failover (iproute/netfilter)
May 14 14:05:30 <_trine> I have just flashed the newest trunk r21446 and now there does not appear to be ipv6 connectivity although I can ping ipv6 address's from withing the router
May 14 15:14:57 <Mr_Smoke> _trine: what do you mean by "no ipv6 connectivity" ?
May 14 15:18:40 <_trine> Mr_Smoke, hello
May 14 15:19:06 <_trine> just that I cant connect to ipv6.google
May 14 15:19:22 <_trine> but I can ping it from my router
May 14 15:20:07 <_trine> I have reflashed now to an older version and am in the process of setting it up
May 14 15:20:19 <Mr_Smoke> Sounds like (p)MTU trouble to me
May 14 15:20:29 <Mr_Smoke> If you ping but can't get the contents ...
May 14 15:21:29 <Mr_Smoke> _trine: What kind of IPv6 do you have ? native, 6rd, tunnel ... ?
May 14 15:21:36 <_trine> tunnel
May 14 15:21:48 <_trine> It all was working b4 I flashed
May 14 15:22:16 <Mr_Smoke> from what to what ?
May 14 15:27:58 <__trine> Mr_Smoke, I am flashed back now
May 14 15:28:05 <__trine> I can type better
May 14 15:28:23 <__trine> the newest trunk is broken badly for ipv6
May 14 15:28:53 <__trine> this is r20754 and this works
May 14 15:30:29 <__trine> on the newest trunk I cannot connect to any ipv6 website nor can the router connect to ipv6.chat.freenode.net
May 14 15:30:35 <Mr_Smoke> Hm interesting
May 14 15:31:06 <__trine> but i can ping from within the router ipv6 ip address
May 14 15:31:16 <Mr_Smoke> you can ping the outside ?
May 14 15:31:22 <__trine> no
May 14 15:31:27 <Mr_Smoke> Oh
May 14 15:31:40 <Mr_Smoke> Is your v6 route set ?
May 14 15:31:51 <__trine> what do you mean
May 14 15:32:22 <__trine> do you mean is everyting looking ok in ifconfig
May 14 15:32:26 <__trine> it is
May 14 15:33:09 <__trine> I think r21176 worked ok too
May 14 15:33:13 <__trine> from memory
May 14 15:33:39 <__trine> we had this problem a couple of weeks ago and it was resolved
May 14 15:34:08 <__trine> but it appears to have raised it's ugly head again    :))))
May 14 15:34:36 <Mr_Smoke> what does "route -A inet6" say about your default route ?
May 14 15:36:59 <__trine> http://dpaste.com/194559/
May 14 15:37:37 <__trine> but this older one is working
May 14 15:37:42 <Mr_Smoke> Can you ping 2001:470:1f08:bf6::1 ?
May 14 15:39:09 <__trine> yes
May 14 15:39:28 <__trine> but as i said this one is working a reflashed back to an older one
May 14 15:39:30 <Mr_Smoke> Oh but you're using the working version yeah
May 14 15:39:31 <Mr_Smoke> ok
May 14 15:39:58 <__trine> if you want to run some tests I can reflash back to the not working one
May 14 15:40:14 <Mr_Smoke> Well I won't be able to test much from the outside
May 14 15:40:19 <Mr_Smoke> Things to consider :
May 14 15:40:37 <Mr_Smoke> LL ping, ping the other end of the tunnel, ping with larger packets (depending on your MTU)
May 14 15:41:13 <__trine> you can have root access if you want I have nothing I care too much about
May 14 15:41:35 <__trine> I reflash frequently anyway
May 14 15:41:42 <Mr_Smoke> I can try, if you want
May 14 15:42:21 <__trine> ok let me flash back,,, this router odes not know if it's coming or going today
May 14 15:42:36 <__trine> it will be getting dizzy
May 14 15:53:18 <_trine> its flashing now
May 14 15:57:03 <_trine> I hope its not bricked again
May 14 15:57:19 <_trine> its taking a long time to come back up
May 14 15:57:26 <_trine> ah its back
May 14 15:57:33 <Mr_Smoke> Hehe
May 14 15:57:36 <Mr_Smoke> Those long minutes
May 14 15:57:52 <Mr_Smoke> It's even worse when it's a dedicated servers with plenty of VMs on it :)
May 14 15:58:37 <_trine> its like waiting for a kettle to boil
May 14 15:59:09 <_trine> I bricked it this morning
May 14 16:01:10 <Mr_Smoke> You've got jtag ?
May 14 16:04:25 <_trine> well i have but I used my serial
May 14 16:04:54 <Mr_Smoke> Yeah, same here
May 14 16:19:34 <__trine> well I have no idea why but I have reflashed the router with trunk and now ipv6 is working again
May 14 16:19:40 <Mr_Smoke> Hahaha
May 14 16:19:49 <Mr_Smoke> Linux does that to you :)
May 14 16:19:51 <Mr_Smoke> At times :)
May 14 16:20:00 <__trine> the gremlins must be on overtime
May 14 16:20:08 <Mr_Smoke> Sounds like it
May 14 16:20:24 <Mr_Smoke> I think my 10.03 router is almost ready for production now
May 14 16:20:31 <__trine> the first time I flashed today it bricked
May 14 16:20:44 <__trine> so it's one of those days I guess
May 14 16:28:00 <epointsystem> cshore: Any new thoughts on auto-mounting hotplugged unpartitioned block devices?
May 14 17:46:59 <rtz2> what the heck should I do with this guy: https://lists.openwrt.org/pipermail/openwrt-devel/2010-May/007127.html ?
May 14 17:47:27 <jow_laptop> whats wrong?
May 14 17:49:32 <rtz2> jow_laptop: well, it's basically the same brcm4716 patch I'm trying to get into openwrt since weeks
May 14 17:49:39 <rtz2> minus a bunch of bugfixes
May 14 17:49:57 <rtz2> jow_laptop: besides that, it's a ugly #ifdef mess
May 14 17:50:25 <rtz2> so I need a friendly way to say no
May 14 17:50:33 <rtz2> +#ifndef CONFIG_WNR3500L
May 14 17:50:33 <rtz2>  void plat_irq_dispatch(void)
May 14 17:50:33 <rtz2> +#else
May 14 17:50:33 <rtz2> +void asmlinkage plat_irq_dispatch(void)
May 14 17:50:33 <rtz2> +#endif
May 14 17:52:13 <rtz2> it's not the first time he's come up with it, there were already a mail 2 weeks or so ago with a patch for the 8.09 kernel
May 14 17:52:57 <rtz2> with the same problems, which I already pointed out to him
May 14 17:52:58 <rtz2> hmmmmm
May 14 17:54:28 <jow_laptop> well explain him kindly that the patch is not properly integrated
May 14 17:54:42 <jow_laptop> and that it will be used as a starting point to add real support
May 14 17:57:51 <rtz2> hmmm
May 14 17:58:35 <rtz2> diplomacy is not exactly an area where I'm good at
May 14 17:59:14 <rtz2> hmmm
May 14 19:15:12 <rtz2> nbd: ping
May 14 20:46:25 <epointsystem> cshore: ping
May 14 20:49:50 <nbd> rtz2: pong
May 14 20:59:54 <rtz2> nbd: sorry, I have to ask again
May 14 21:00:09 <rtz2> nbd: did you have time to review the rest of the mtd patch?
May 14 21:08:34 <cshore> ping epointsystem
May 14 21:09:18 <epointsystem> Hi cshore. Any new thoughts on auto-mounting hotplugged unpartitioned block devices?
May 14 21:09:56 <cshore> epointsystem: nothing new, just the blkid idea....I'd use /sys but that 2.6 specific
May 14 21:11:00 <epointsystem> By the way, is the proprietary broadcom driver the sole reason to keep 2.4 around?
May 14 21:11:11 <cshore> epointsystem: AFAIK
May 14 21:12:45 <epointsystem> And what is the current status of the open source effort to replace it?
May 14 21:13:30 <cshore> epointsystem: not sure...I think there is progress being made but I don't know how long before it will be ready to replace the proprietary driver in regular use
May 14 21:22:49 <epointsystem> cshore: are you willing to put together the patch for checking blkdev in blkid output?
May 14 21:35:58 <rtz2> epointsystem: the support for older broadcom wireless chips is reasonable good
May 14 21:36:07 <rtz2> epointsystem: n support is still kinda wip
May 14 21:39:17 <xl0> Is it a good idea to build openwrt with make -jN? Looks like there are some problems with it, but can't tell, how severe.
May 14 21:41:30 <rtz2> xl0: in the core system, not very severe
May 14 21:41:43 <rtz2> xl0: but I wouldn't try it for rarely used packages
May 14 21:42:02 <xl0> Well, I can't even build the toolchain.
May 14 21:42:29 <xl0> http://dumpz.org/19448/
May 14 21:42:44 <xl0> Or am I missing something?
May 14 21:44:46 <rtz2> xl0: yes
May 14 21:44:49 <rtz2> xl0: V=99
May 14 21:47:36 <xl0> rtz2: Problem is, it worked fine the next time..
May 14 21:47:56 <cshore> epointsystem: it's not high on my priority list at all, but if you want to submit a patch I'll look at it much sooner
May 14 21:48:09 <xl0> And I did not get this error after git clean -dfx. :/
May 14 21:48:14 <rtz2> xl0: that's more or less expected
May 14 21:48:33 <rtz2> xl0: there are dependencies missing
May 14 21:48:59 <xl0> So, no -j before someone fixes it. :/ Thanks.
May 14 21:52:03 <nbd> rtz2: nope, not yet
May 14 21:52:55 <rtz2> nbd: ok
May 14 21:53:11 <rtz2> nbd: but I have to ask from time to time :)
May 14 21:53:18 <nbd> sure
May 14 21:53:25 <nbd> no problem
May 14 22:59:19 <cshore> ping nbd
May 14 23:05:05 <cshore> ping epointsystems
May 14 23:11:22 <cshore> epointsystems: if you see this, remind me in a few days and I'll see if I can find sometime sooner
May 14 23:16:37 <Weedy> fffffffffffffffffffffffffff compat-wireless
May 14 23:16:40 * Weedy rages
May 14 23:17:20 * cshore mocks Weedy's pain; then cries as he experiences it
May 14 23:18:02 <Weedy> i can't find a date that works with .32
May 15 01:15:06 <OutBackDingo> Weedy: whats the target?
**** ENDING LOGGING AT Sat May 15 02:59:56 2010