**** BEGIN LOGGING AT Thu Feb 24 02:59:57 2011
Feb 24 04:04:18 <builder> build #80 of pxcab is complete: Failure [failed compile_3]  Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/80
Feb 24 06:11:32 <Epsylon3> www.plugcomputer.org
Feb 24 06:44:57 <PsyMan> hmm
Feb 24 06:44:59 <PsyMan> great
Feb 24 06:45:21 <PsyMan> any config i try fails
Feb 24 06:45:45 <PsyMan> flashed router, rebooted, rooter starts, configs self
Feb 24 06:46:06 <PsyMan> yet it never asigns any adressed
Feb 24 06:46:23 <PsyMan> and even manually cinfiguring ips does not work
Feb 24 07:42:08 <CIA-84> blogic * r25694 /trunk/package/uboot-lantiq/ (33 files in 5 dirs):
Feb 24 07:42:08 <CIA-84> [uboot-lantiq]
Feb 24 07:42:08 <CIA-84> * adds stage1 lzma
Feb 24 07:42:08 <CIA-84> * new boards
Feb 24 07:42:08 <CIA-84> * fixes settings for PSC ram
Feb 24 07:42:09 <CIA-84> * lost of cleanups
Feb 24 07:51:25 <philipp64|laptop> jow_laptop: ping
Feb 24 08:09:41 <builder> build #80 of ppc44x is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/80
Feb 24 08:13:36 <CIA-84> kaloz * r25695 /trunk/target/linux/avr32/image/u-boot/Makefile: [avr32]: include u-boot version in filename
Feb 24 08:15:12 <CIA-84> kaloz * r25696 /trunk/toolchain/gcc/patches/4.3.5/930-avr32_support.patch: [toolchain/gcc/4.3.5]: sync avr32 patch with the latest ones from atmel
Feb 24 08:16:01 <CIA-84> kaloz * r25697 /trunk/target/linux/avr32/Makefile: [avr32]: upgrade to 2.6.37.1
Feb 24 12:57:20 <CIA-84> kaloz * r25698 /trunk/toolchain/uClibc/ (16 files in 2 dirs): [toolchain/uClibc]: upgrade to latest git, refresh patches
Feb 24 13:01:38 <jdoe123> hi, can anyone help me with jtag adapter selection?
Feb 24 13:07:06 <blogic> jdoe123: sure
Feb 24 13:07:50 <blogic> jdoe123: what do you need ?
Feb 24 13:07:57 <blogic> something to work with openocd ?
Feb 24 13:08:05 <blogic> http://www.amontec.com/jtagkey-tiny.shtml
Feb 24 13:08:09 <blogic> this is a good unit
Feb 24 13:08:24 <blogic> and openocd ships a working config for it
Feb 24 13:08:32 <jdoe123> well i dont know much about hardware i reacently started using serial
Feb 24 13:09:03 <jdoe123> what i would like is to get a usb jtag adapter cable
Feb 24 13:09:11 <jdoe123> primary for this device: http://wiki.dns323.info/hardware:jtag
Feb 24 13:09:32 <jdoe123> it would be great if i could use that jtag adapter for other projects as well
Feb 24 13:09:53 <jdoe123> in the page above soldering is the part that scares me :((
Feb 24 13:10:05 <blogic> ok, let me have a look
Feb 24 13:10:35 <blogic> ok the wiki actually mentions the unit i just posted a link to
Feb 24 13:10:37 <blogic> :)
Feb 24 13:10:39 <blogic> 2: The “Amontec JTAGKey Tiny”
Feb 24 13:11:03 <blogic> and those solder pictures are for a parallel port wiggler
Feb 24 13:11:23 <blogic> so in theory the amontec should work 1:1 without an extra circuit
Feb 24 13:11:59 <blogic> yes, get the amontec and it will "just work"
Feb 24 13:12:02 <_trine> and if its wrt54g type router you dont need an adapter to jtag it
Feb 24 13:12:46 <jdoe123> what about this: Solder 47 KOhm resistors @ R202 R201 R200 Connect JTAG pins just like the Wiggler port, and also bridge R199 (EN pin to VCC) to enable JTAG. Refer to the top half of the pdf as attached in the previous section. Pinout of the jtagkey tiny is available  here
Feb 24 13:13:23 <jdoe123> that's why i didn't buy that jtag already
Feb 24 13:15:36 <blogic> yes
Feb 24 13:15:41 <blogic> you need to solder them as it seems
Feb 24 13:16:07 <blogic> i would first try without
Feb 24 13:16:44 <jdoe123> could i brich something - jtag/dns323/usb
Feb 24 13:16:54 <jdoe123> *brick
Feb 24 13:18:28 <blogic> well
Feb 24 13:18:37 <blogic> if oyu do you have all the tools to debrick :)
Feb 24 13:19:20 <jdoe123> yeah
Feb 24 13:19:54 <jdoe123> i wrote patch for openwrt to enable support for this device but i did not get any feedback
Feb 24 13:20:16 <jdoe123> i have documented it here: http://dns323.lukaperkov.net/
Feb 24 13:20:41 <jdoe123> now i would like to replace the stock uboot with my own...
Feb 24 13:21:21 <jdoe123> thats why i'm looking for jtag for dummies solution...
**** BEGIN LOGGING AT Thu Feb 24 17:26:17 2011
Feb 24 17:28:08 <PsyMan> hmm
Feb 24 17:28:18 <PsyMan> [florian] around?
Feb 24 17:33:10 <PsyMan> i'll put this here anyway
Feb 24 17:34:06 <PsyMan> rev 25675 does not seem to fix that dg834g issue
Feb 24 17:35:01 <PsyMan> did clean builds, with or without webui, with or without wireless, nothing seemed to work
Feb 24 17:35:54 <[florian]> PsyMan: what does that mean nothing seems to work?
Feb 24 17:36:08 <PsyMan> tried stripped ar7 build just with support for that marval switch and adsl... not even that worked
Feb 24 17:36:16 <PsyMan> i mean that it compiles fine
Feb 24 17:36:23 <PsyMan> flashed router
Feb 24 17:36:34 <PsyMan> modem configured itself on boot
Feb 24 17:36:53 <[florian]> do you have a serial console attached to your dg834g ?
Feb 24 17:36:53 <PsyMan> yet i can't access it using ethernet
Feb 24 17:36:58 <PsyMan> if i had
Feb 24 17:37:02 <PsyMan> i'd post the log
Feb 24 17:37:12 <PsyMan> or better, find the issue and apply a patch
Feb 24 17:37:13 <PsyMan> :p
Feb 24 17:38:44 <[florian]> well without some useful debug I can't simply provide you with a patch
Feb 24 17:38:51 <PsyMan> i know
Feb 24 17:39:05 <PsyMan> didn't you have this router?
Feb 24 17:39:07 <[florian]> is there any way you can plug in a serial console?
Feb 24 17:39:11 <PsyMan> hmm
Feb 24 17:39:13 <[florian]> I don't
Feb 24 17:39:22 <[florian]> otherwise you would certainly not have had that kind of stuff
Feb 24 17:39:41 <[florian]> but anyway, there's been something like 6 versions of the dg834g on the market now
Feb 24 17:39:44 <[florian]> all different
Feb 24 17:39:55 <PsyMan> no soldering iron here, no serial ports on desktop nor laptop, not even the parts to jtag it
Feb 24 17:39:59 <PsyMan> sry :|
Feb 24 17:40:19 <PsyMan> ahh
Feb 24 17:40:23 <PsyMan> here is the catch
Feb 24 17:40:35 <PsyMan> v1 v2 and v3 are more or less the same
Feb 24 17:41:23 <PsyMan> v2 eliminated some design flaws, v3 introduced new ones to decrease build cost
Feb 24 17:41:36 <PsyMan> then it moved away from ti :p
Feb 24 17:48:22 <builder> build #73 of uml is complete: Failure [failed compile_10]  Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/73
Feb 24 17:54:24 <xMff> ping cshore
Feb 24 17:54:34 <xMff> and/or ping [florian]
Feb 24 18:01:46 <cshore> pong xMff
Feb 24 18:02:47 <xMff> do you recognize this error? http://pastebin.j-zero.de/lclvk1j4t
Feb 24 18:02:47 <xMff> apparently it happens in backfire, could it be caused by an outdated imagetag.h ?
Feb 24 18:04:22 <cshore> well there's definitely something mismatched because 'struct bcm_tag' has not member 'flashRootLength'
Feb 24 18:04:36 <cshore> is this on a fresh checkout or an update?
Feb 24 18:05:50 <xMff> don't know, let me ask
Feb 24 18:07:01 <xMff> r25679
Feb 24 18:07:10 <xMff> so quite recent
Feb 24 18:07:54 <cshore> odd, I'll have to check the commits.  I haven't changed anything that recently (it would have been a couple of months ago for me)
Feb 24 18:08:52 <xMff> hm
Feb 24 18:09:01 <xMff> his error messages do not match the sources
Feb 24 18:10:10 <xMff> now he said its trunk...
Feb 24 18:11:01 <xMff> cshore: it was you actually :)
Feb 24 18:11:18 <xMff> https://dev.openwrt.org/changeset/24838
Feb 24 18:11:55 <cshore> ah, yes, a couple of months ago, but if it's trunk it works here
Feb 24 18:12:05 <xMff> I assume his kernel tree is old
Feb 24 18:12:16 <xMff> however, see the last hunk at the link above
Feb 24 18:12:29 <xMff> there you use both flashRootLength and rootLength, this is okay?
Feb 24 18:12:45 <cshore> yes, they're different
Feb 24 18:12:49 <xMff> I see
Feb 24 18:13:38 <cshore> rootLength is the 'real' length, which gets used after the tag gets rewritten by the crc fix
Feb 24 18:14:22 <cshore> flashRootLength gets changed to 0 by the crc fix, so that jffs2 filesystem don't fail to boot on stupid routers
Feb 24 18:14:33 <cshore> *doesn't
Feb 24 18:14:48 <cshore> or for that matter jffs2 part of a squashfs
Feb 24 18:15:35 <cshore> anyway he probably needs to do a fresh checkout
Feb 24 18:16:23 <xMff> probably yes
Feb 24 18:18:02 <cshore> so I saw the patch for doing stuff after a fstab hotplug event ... I'm hoping to get a chance to add it soon; I like it
Feb 24 18:18:30 <cshore> *review I mean; like the concept
Feb 24 18:19:32 <cshore> but have to check what it does with a rootfs during preinit
Feb 24 18:20:36 <xMff> I do not like it that much
Feb 24 18:20:44 <cshore> oh?
Feb 24 18:21:04 <cshore> I want to the option to plain (non-overlay) mount too
Feb 24 18:22:19 <xMff> I mean the exec-after-mount is kind of arbritary
Feb 24 18:22:58 <xMff> I'd say an /etc/hotplug.d/mount/ would make way more sense
Feb 24 18:23:36 <cshore> hmmm....right, just do a flag that indicates the mount happened and then they can drop in /etc/hotplug.d/block
Feb 24 18:24:07 <cshore> or do you mean do a new directory?
Feb 24 18:24:33 <cshore> just like interfaces vs net?
Feb 24 18:25:03 <xMff> correct, new event class
Feb 24 18:25:09 <xMff> mount & umount
Feb 24 18:25:24 <xMff> I mean subsystem mount and action mount/umount
Feb 24 18:25:31 <xMff> maybe even swapon and swapoff
Feb 24 18:25:37 <xMff> and probably fsck too
Feb 24 18:26:13 <cshore> xMff: ok, would those be launch via the block script or is it a matter of creating some special hotplug thing?
Feb 24 18:26:47 <cshore> I don't know that much about the hotplug layout, I just modified the block files
Feb 24 18:26:51 <xMff> env -i ACTION=mount DEVICE=/dev/foo DIRECTORY=/foo/bar /sbin/hotplug-call mount
Feb 24 18:27:13 <xMff> env -i ACTION=umount DEVICE=/dev/foo DIRECTORY=/foo/bar /sbin/hotplug-call mount
Feb 24 18:27:21 <cshore> ah, ok
Feb 24 18:27:23 <xMff> etc. in the appropraite palces within the mount scripts
Feb 24 18:27:32 <xMff> right before mount ... and umount ...
Feb 24 18:27:57 <xMff> hotplug-call will block while the scripts run (unless scripts decide to detach themselves)
Feb 24 18:29:01 <xMff> well them ount action after mount and the umount action before umount
Feb 24 18:29:36 <cshore> yeah, that makes sense
Feb 24 18:29:45 <xMff> one could then still ship a standard hotplug handler which implements the uci options proposed on the ml
Feb 24 18:30:07 <xMff> like /etc/hotplug.d/mount/01-run-services
Feb 24 18:30:55 <xMff> I don't know whether there maybe already are standard uevents for mount and umount
Feb 24 18:31:09 <xMff> so that you do not even have to call hotplug-call
Feb 24 18:31:44 <cshore> I wouldn't be surprised
Feb 24 18:32:11 <xMff> http://alinux.tv/Kernel-2.6.34/filesystems/gfs2-uevents.txt
Feb 24 18:32:23 <xMff> I suppose other filesystems provide those as well
Feb 24 18:36:52 <xMff> hm nope, just checked
Feb 24 18:37:14 <xMff> there are uevents for appearing and disappearing blockdevs or partitions, but none for actual mount and umount
Feb 24 18:37:27 <cshore> ok, good to know
Feb 24 18:37:35 <xMff> at least not with ext2/3/4
Feb 24 18:41:23 <tripolar> ping nbd
Feb 24 18:41:53 <tripolar> xMff: do you have write access to busybox?
Feb 24 18:41:56 <cshore> for the non-overlay extroot, which about an is_overlay parameter which defaults to true, and if false mounts the rootfs doing a straign pivot_root, instead of fopivot?
Feb 24 18:42:38 <xMff> hm
Feb 24 18:43:34 <xMff> not sure tbh. In retrospect it would've maybe been better to just handle target/ and target /overlay specially
Feb 24 18:45:45 <cshore> xMff: maybe we should deprecate is_rootfs (it will still work for now, but if the target is / or /overlay we act accordingly in any case)?
Feb 24 18:46:08 <xMff> yes, that would be way more natural imo
Feb 24 18:46:17 <xMff> or "intuitive"
Feb 24 18:47:38 <xMff> oh another thing, we should maybe add an "/etc/init.d/fstab enable-overlay" action
Feb 24 18:47:42 <xMff> to clear out the md5sums
Feb 24 18:47:48 <cshore> ok
Feb 24 18:48:21 <cshore> I was wondering how to deal with that
Feb 24 18:49:31 <cshore> I try to work on those two things in the coming week (probably not till this Sunday, but we'll see)
Feb 24 18:51:29 <cshore> Any ETA on the new init stuff?
Feb 24 18:52:16 <cshore> I want to try to make using non-extroot external storage more seamless
Feb 24 19:04:18 <builder> build #91 of s3c24xx is complete: Failure [failed shell_4 compile_6]  Build details are at http://tksite.gotdns.org:8010/builders/s3c24xx/builds/91
Feb 24 19:08:55 <CIA-84> acoul * r25703 /trunk/target/linux/generic/patches-2.6.38/218-mini_fo_fix_2_6_38.patch: linux/generic: fix r25699 take II
Feb 24 19:13:11 <CIA-84> acoul * r25704 /trunk/target/linux/generic/patches-2.6.38/067-block2mtd_2.6.38_fix.patch: linux/generic: fix block2mtd for 2.6.38
Feb 24 19:13:47 <KanjiMonster> acoul: you probably need to update mini_fo's mini_fo_d_compare signature, too
Feb 24 20:13:46 <philipp64|laptop> xMff: does the openswan or ipsec-tools support roadwarrior type VPN configs out of the box?
Feb 24 20:13:58 <xMff> probably not
Feb 24 20:14:16 <philipp64|laptop> we could work together on the scripting to make that happen if you're interested.
Feb 24 20:14:29 <tripolar> hi
Feb 24 20:14:46 <philipp64|laptop> I had some of it working but couldn't get cert-based authentication to work without using xauth.
Feb 24 20:14:49 <tripolar> how can i find out if my /dev/watchdog  is implemented by hard or software?
Feb 24 20:14:54 <tripolar> its BCM47xx Watchdog Timer enabled
Feb 24 20:14:58 <tripolar> in dmesg
Feb 24 20:15:01 <xMff> then its hardware
Feb 24 20:15:18 <tripolar> xMff: so when i kill the watchdog the device should reboot?
Feb 24 20:15:53 <xMff> philipp64|laptop: feel free to script away, I have no experience with it, imo ipsec is overengineered committee software :)
Feb 24 20:19:42 <tripolar> xMff: could please subbmit my busybox patch?
Feb 24 20:19:52 <tripolar> im using it by now without a problem
Feb 24 20:20:06 <tripolar> xMff: btv killing watchdog worked ;)
Feb 24 20:20:19 <xMff> you said that about openssl as well, yet a number of packages broke afterwards
Feb 24 20:20:47 <tripolar> btv=btw
Feb 24 20:21:08 <tripolar> xMff but that has nothing to do with openssl it a package problem
Feb 24 20:21:12 <tripolar> i cant test every package
Feb 24 20:21:16 <tripolar> on every arch
Feb 24 20:21:59 <tripolar> xMff so we should stop updating everything because updating could brake something ;)
Feb 24 20:22:04 <xMff> it is a problem in so far as 1.0.0 broke the api, which was a bit more than the usual bump
Feb 24 20:22:47 <xMff> anyway, point is that I won't commit the busybox update now
Feb 24 20:23:04 <xMff> I do not maintain it anyway, and I want to give it some tries first
Feb 24 20:23:45 <tripolar> xMff: yeah but test it - was a lot of work - like the openssl update - because i ported the ocf stuff - and afterwards i read the README :(
Feb 24 20:24:43 <tripolar> the most time cosuming work on busybox is the config updates ...
Feb 24 20:25:26 <xMff> exactly, which leads me to another point
Feb 24 20:25:36 <xMff> why did you disable base64 with this update?
Feb 24 20:25:56 <tripolar> was it enabled in the default config
Feb 24 20:26:31 <xMff> no idea, you wrot you disabled it, therfore I assumed it was
Feb 24 20:27:08 <tripolar> yeah i think it was therefor
Feb 24 20:27:30 <tripolar> i disabled it
Feb 24 20:29:45 <tripolar> xMff the only thing i wasnt shure about was hunk 1 in 610-syslog-remote-retry-connection.patch
Feb 24 20:30:09 <tripolar> but because hank 2 was in i didnt apply it
Feb 24 20:30:48 <cshore> 610 can removed...I had meant to anyway
Feb 24 20:31:23 <tripolar> like 110 <2.6.37
Feb 24 20:31:35 <tripolar> @cshore
Feb 24 20:31:50 <tripolar> in generic kernel patches
Feb 24 20:31:58 <cshore> ?
Feb 24 20:32:14 <tripolar> it breaks some firewall scripts i use
Feb 24 20:32:28 <tripolar> without it, they work
Feb 24 20:32:59 <tripolar> i didnt dig deeper
Feb 24 20:33:28 <cshore> er, and that has to do with busybox 110 how?
Feb 24 20:33:36 <cshore> sorry 610
Feb 24 20:33:49 <xMff> you mean the netfilterm atch speedup?
Feb 24 20:33:59 <tripolar> who is the maintainer of busybox?
Feb 24 20:34:29 <tripolar> i should read the makefile ;)
Feb 24 20:34:32 <tripolar> ping nico
Feb 24 20:34:36 <tripolar> ;)
Feb 24 20:34:44 <tripolar> shit he isnt here
Feb 24 20:35:55 <tripolar> xMff: do you have it a compiled busybox 1.18.3 by now?
Feb 24 20:39:11 <xMff> no, I am busy with other stuff right now
Feb 24 20:45:25 <tripolar> cshore could you also please test it - when more pleople test it - maybe some problem that may arrise can be fix befor it is in trunk
Feb 25 00:25:13 <PsyMan> Wipster there?
Feb 25 00:25:27 <Wipster> PsyMan, just
Feb 25 00:25:47 <PsyMan> ?
Feb 25 00:25:57 <PsyMan> are ya busy?
Feb 25 00:26:28 <Wipster> was just shutting down, whats up?
Feb 25 00:26:33 <PsyMan> ahh
Feb 25 00:26:47 <PsyMan> well, i tried to compile the latest rev
Feb 25 00:27:23 <PsyMan> but it seems to have the same switch issue it had before your patch
Feb 25 00:27:44 <PsyMan> and i wonder if i missed anything :\
Feb 25 00:27:56 <Wipster> can you send me a dmesg?
Feb 25 00:28:05 <PsyMan> not really
Feb 25 00:28:08 <PsyMan> no serial here
Feb 25 00:29:52 <PsyMan> all i can tell is that the ethernet ports are dead, no ip assignment
Feb 25 00:30:37 <Wipster> ok well check that your svn rev is >= 25675 or that you are using the .37 kernel
Feb 25 00:30:45 <PsyMan> did
Feb 25 00:30:50 <PsyMan> :p
Feb 25 00:31:40 <PsyMan> actually, the "latest rev" was actually 25675
Feb 25 00:31:51 <PsyMan> back when i built the thing
Feb 25 00:32:57 <PsyMan> i tried with the marvell switch selected from menu config, without it too
Feb 25 00:33:23 <PsyMan> even took all non basic packages out just to make sure
Feb 25 00:34:49 <PsyMan> nothing seemed to work
Feb 25 00:34:58 <Wipster> its built in as default, well a serial output would be most helpfull it works on my v3 which is essentially the same hardware
Feb 25 00:36:17 <PsyMan> you sure you didn't mess with anything, edited related src, etc?
Feb 25 00:37:15 <Wipster> sorry?
Feb 25 00:38:54 <PsyMan> ?
Feb 25 00:39:15 <PsyMan> well, it is supposed to work
Feb 25 00:39:52 <PsyMan> so i wonder if there was something missing there, or if if some patch on trunk breaks the switch
Feb 25 00:41:57 <Wipster> ok well if you can get a serial output to me and are willing to test a patch or two we can get it fixed up
Feb 25 00:42:11 <PsyMan> lemme see
Feb 25 00:43:42 <PsyMan> serial cable and extra stuff to fixup voltages... nope, serial ports on pc... nope, some nokia cable as a better choice... nope, how about soldering iron...? nope :|
Feb 25 00:45:02 <PsyMan> so yeah, this is not an option
Feb 25 00:45:05 <PsyMan> :|
Feb 25 00:49:34 <PsyMan> hmm
Feb 25 00:49:42 <PsyMan> clean build finished
Feb 25 00:50:24 <PsyMan> no errors, kernel 2.6.32, with some patches taken out
Feb 25 00:50:51 <PsyMan> lemme test that and see how it goes
Feb 25 00:52:21 <PsyMan> so bbl
Feb 25 01:23:10 <builder> build #74 of x86 is complete: Failure [failed compile_6]  Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/74
Feb 25 01:54:07 <neutronscott> OT question. last time i read email on my server i used pine. whats the cool MUA nowadays? :)
Feb 25 01:56:39 <neutronscott> alpine? :D
Feb 25 02:04:09 <cshore> neturonscott: for text mutt I guess (though I don't think it's as featureful as pine)
Feb 25 02:04:40 <cshore> neutronscott: most people use a GUI now
Feb 25 02:05:42 <cshore> in which case I'd say claws-mail - light, fast, featureful, and apparently has good keyboard shortcuts (though I never use them)
Feb 25 02:12:57 <neutronscott> GUI? like Windows XP?
Feb 25 02:13:57 <neutronscott> heh. only want to read local mail on a server. don't really want pop3/imap or X11 dependancies
Feb 25 02:14:38 <neutronscott> maybe i'll just use a lightweight webmai that does mbox
Feb 25 02:38:02 <cshore> ah, in that case mutt or webmail
Feb 25 02:41:17 <neutronscott> seems all these lil webmails want imap
Feb 25 02:41:23 <neutronscott> i'll stick with 'mail'
Feb 25 02:41:39 <neutronscott> which is mailx..
Feb 25 02:43:25 <cshore> why not mutt?
Feb 25 02:43:32 <cshore> ncurses is all it needs
**** ENDING LOGGING AT Fri Feb 25 02:59:57 2011