**** BEGIN LOGGING AT Mon Jan 24 02:59:58 2011
Jan 24 03:00:01 <philipp64|laptop> nbd: ok, it links now. sent out a patch.
Jan 24 03:24:50 <philipp64|laptop> umm.... sent out the correct patch the 2nd time.
Jan 24 04:26:07 <russell--> hmm.  maybe nothing wrong with sysupdate -n afterall.  i just flashed the nanostation5 from tftp and ended up with the same 204k /overlay afterwards.
Jan 24 04:28:13 <russell--> wtf?
Jan 24 04:47:59 <russell--> so the weird thing is that the 'find -type f' files in /overlay only add up to 5644 bytes
Jan 24 05:17:10 <russell--> for about a dozen files, most of which don't diff from /rom versions
Jan 24 05:22:42 <neutronscott> maybe they've been touch'ed?
Jan 24 05:35:05 <russell--> not by me
Jan 24 06:23:00 <russell--> maybe df is just lying to me
Jan 24 06:36:35 <russell--> here is a paste of the jffs2dump -b -v -c output.  There isn't anything 204k in there!  http://paste.openwrt.org/f8558c58
Jan 24 06:41:26 <CIA-84> kaloz * r25080 /trunk/ (Config.in include/kernel-defaults.mk): 2.6.38 supports xz ramdisks, too
Jan 24 06:51:28 <xMff> russell--: the internal journal might also be counted there
Jan 24 07:01:07 * russell-- confesses a certain ignorance of how jffs2 works
Jan 24 07:02:03 <russell--> although, once upon a time i read the paper describing it
Jan 24 07:05:17 <xMff> there is one block, like the superblock on ext* afaik
Jan 24 07:05:26 <xMff> which holds fs metadata
Jan 24 07:05:45 <xMff> depending on your flash that should be 64-128 missing k already
Jan 24 07:06:39 <russell--> i think the erase block size is 64k on this
Jan 24 07:07:59 <xMff> then there's a "sto file" which should take another few kb
Jan 24 07:12:53 <russell--> that dump shows an aweful lot of empty space
Jan 24 07:15:56 <russell--> i can't finger the culprit, but somethings seems fish, is all
Jan 24 07:16:02 <russell--> fishy*
Jan 24 07:17:48 <builder> build #67 of orion is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/67
Jan 24 07:54:09 <WTGPhoben> Seeing switch problem on wrt160nl. anyone out there that can help?  Had a working config in august, just updated backfire build and now it doesn't work with same config.
Jan 24 07:54:32 <xMff> does your config use vlan 0 ?
Jan 24 07:55:15 <WTGPhoben> xMff: iwconfig gives correct output (eth0.0 and eth0.1) and swconfig seems to give correct output
Jan 24 07:55:25 <WTGPhoben> but eth0.1 is dead
Jan 24 07:55:30 <xMff> change it to 1 and 2 instead of 0 and 1
Jan 24 07:55:40 <WTGPhoben> tcpdumping it shows it's not reading packets at all
Jan 24 07:56:01 <WTGPhoben> will try. now...
Jan 24 08:00:08 <WTGPhoben> xMff: that makes a mess.  swconfig registers vlan2  (pvid: 2) but the one that's supposed to be 1 is labeled pvid: 0
Jan 24 08:00:21 <WTGPhoben> ifconfig shows eth- and eth 0.0
Jan 24 08:00:25 <WTGPhoben> err eth0
Jan 24 08:00:34 <WTGPhoben> and eth 0.0
Jan 24 08:02:52 <russell--> okay, i can confirm that what i dumbly though was a persistent /overlay is really just a weirdly created one.  with sysupgrade -n, /overlay is not persistent across the upgrade, as expected.
Jan 24 08:07:38 <WTGPhoben> xMff: if I change to vlans to 2 and 3, i get eth0.0 and eth0.1 again
Jan 24 08:08:10 <WTGPhoben> scratch that
Jan 24 08:08:17 <WTGPhoben> wrong place in scrollback
Jan 24 08:08:35 <WTGPhoben> i get 0.2 and 0.3
Jan 24 08:10:54 <WTGPhoben> xMff: same behavior there, but now swconfig is not registering it correctly
Jan 24 08:12:10 <WTGPhoben> it shows vlan0 as 0,5; vlan2 as 2,5 and vlan3 as 1234
Jan 24 08:12:16 <WTGPhoben> ideas?
Jan 24 08:19:45 * philipp64|laptop yawns and wanders off to bed.
Jan 24 08:35:48 <jow_laptop> you should not have a vlan 0
Jan 24 08:35:55 <jow_laptop> 0 is invalid on most drivers
Jan 24 09:01:48 <WTGPhoben> xMff: posted ticket, fyi  https://dev.openwrt.org/ticket/8719
Jan 24 09:05:45 <jow_laptop> you still have vlan 0 in there
Jan 24 09:06:59 <jow_laptop> 4* is invalid syntax
Jan 24 09:07:06 <jow_laptop> and none of your vlans contains a cpu port
Jan 24 09:07:50 <jow_laptop> try this default: https://dev.openwrt.org/browser/trunk/target/linux/ar71xx/base-files/etc/defconfig/wrt160nl/network
Jan 24 09:07:59 <jow_laptop> then base your modifications on that
Jan 24 09:08:03 <jow_laptop> and don't use vlan 0
Jan 24 09:13:48 <Acinonyx> gmorning
Jan 24 09:15:25 <WTGPhoben> <jow_laptop>: what is invalid about 4*?  Why is vlan0 not valid?
Jan 24 09:15:48 <WTGPhoben> I thought 4 was cpu port on linksys
Jan 24 09:16:59 <jow_laptop> yeah ok port 4
Jan 24 09:17:15 <jow_laptop> most drivers reject vlan 0 / treat it special
Jan 24 09:17:29 <WTGPhoben> I tried vlan 1, and same result
Jan 24 09:17:36 <WTGPhoben> err 1,2
Jan 24 09:18:00 <jow_laptop> the network config you attached is incomplete
Jan 24 09:18:42 <WTGPhoben> I can post the rest, one sec...
Jan 24 09:21:55 <WTGPhoben> <jow_laptop>: updated, https://dev.openwrt.org/ticket/8719#comment:1
Jan 24 09:22:39 <jow_laptop> eth0.0 cannot work as the cpu port is not tagged
Jan 24 09:22:47 <jow_laptop> besides the fact that 0 is invalid
Jan 24 09:23:51 <WTGPhoben> so change 0,1 to 1,2 and then explain the tagging part?  should both 4s have t?
Jan 24 09:24:02 <jow_laptop> yes
Jan 24 09:24:07 <WTGPhoben> and remove *?
Jan 24 09:24:10 <jow_laptop> yes
Jan 24 09:24:17 <WTGPhoben> ok, trying now. tnx
Jan 24 09:30:53 <WTGPhoben> jow_laptop: that worked, thanks!
Jan 24 09:31:02 <WTGPhoben> will update defect...
Jan 24 09:39:33 <CIA-84> kaloz * r25081 /trunk/target/linux/generic/patches-2.6.38/ (8 files): [target/generic]: refresh 2.6.38 patches with -rc2
Jan 24 09:40:40 <CIA-84> kaloz * r25082 /trunk/toolchain/uClibc/ (4 files in 2 dirs): [toolchain/uClibc]: update to latest git
Jan 24 09:56:35 <loswillios> I get this when trying to build gcc-4.5.2 for x86: configure: error: Building GCC requires GMP 4.2+, MPFR 2.3.1+ and MPC 0.8.0+.
Jan 24 10:49:38 <russell--> is there a reason that atheros and x86 are back on linux-2.6.32.x?  i've built 2.6.37 recently and it seems fine (with a few kernel config fixups).
Jan 24 11:01:56 <CIA-84> kaloz * r25083 /trunk/target/linux/ (9 files in 9 dirs): platform specific patches don't belong here
**** BEGIN LOGGING AT Mon Jan 24 11:22:50 2011
Jan 24 11:50:13 <CIA-84> cshore * r25084 /packages/net/miniupnpd/Makefile: [packages] miniupnpd: Bump version to 1.5
Jan 24 12:11:42 <CIA-84> kaloz * r25085 /trunk/target/linux/ (51 files in 14 dirs): platform specific patches don't belong here - take #2
Jan 24 12:15:37 <neutronscott> ehh where'd they move to
Jan 24 12:17:14 <CIA-84> kaloz * r25086 /trunk/target/linux/generic/patches-2.6.38/013-gcc4_inline_fix.patch: remove mips gcc4 inline patch - fixed upstream
Jan 24 12:17:18 <neutronscott> oh i see
Jan 24 12:18:19 <CIA-84> kaloz * r25087 /trunk/target/linux/generic/patches-2.6.38/207-powerpc_asm_segment_h.patch: this missing header patch is way too old for any sane software to depend on it
Jan 24 12:33:41 <CIA-84> kaloz * r25088 /trunk/target/linux/generic/ (3 files in 2 dirs): [target/generic/2.6.38]: replace the mips_boot patch with one that makes BOOT_RAW selectable, and enable it in the generic config
Jan 24 12:36:36 <CIA-84> kaloz * r25089 /trunk/target/linux/generic/patches-2.6.38/ (021-mips_image_cmdline_hack.patch 025-mips_disable_fpu.patch): [target/generic/2.6.38]: refresh patches
Jan 24 12:52:08 <CIA-84> kaloz * r25090 /trunk/target/linux/orion/ (3 files in 3 dirs): [target/orion]: upgrade to 2.6.37
Jan 24 12:53:02 <CIA-84> kaloz * r25091 /trunk/target/linux/orion/config-default: [target/orion]: add missing symbol for 2.6.38
Jan 24 13:42:35 <CIA-84> tripolar * r25092 /packages/net/openssh/Makefile: [openssh] update openssh to 5.7p1
Jan 24 14:05:55 <mazilo> I just did an 'svn up' + 'make package/symlinks' and the package/openssl/Makefile is still not updated, yet.
Jan 24 14:06:40 <mazilo> Oops, I meant net/openssh.
Jan 24 14:08:29 <loswillios> mazilo: ./scripts/feeds update
Jan 24 14:09:44 <mazilo> loswillios: Got it. Thanks.
Jan 24 17:57:45 <Epsylon3> hmm, trunk has switched to 2.6.37 ?
Jan 24 18:00:59 <blogic> apparently yes, otherwise you would not ask
Jan 24 18:05:55 <Epsylon3> yea... i hope that will compile :)
Jan 24 18:52:40 <Acinonyx> philip prondeville
Jan 24 18:54:36 <CIA-84> nbd * r25093 /trunk/package/mac80211/ (17 files in 2 dirs): mac80211: update to wireless-testing 2011-01-24
Jan 24 18:54:40 <CIA-84> nbd * r25094 /trunk/package/hostapd/ (16 files in 2 dirs): hostapd: update to 20110117
Jan 24 19:11:14 <Epsylon3> !link
Jan 24 19:12:13 <philipp64|laptop> Acinonyx: close but no cigar.
Jan 24 19:33:09 <philipp64|laptop> nbd: do the patches look right?
Jan 24 19:36:00 <nbd> sorry, but my todo list is too long for me to be able to regularly follow your patches
Jan 24 19:37:25 <nbd> my primary concern is ath9k, and the new to-be-written infrastructure ubus, netifd, configd and anything that's in the way of me completing that
Jan 24 19:37:30 <nbd> don't have much time for other stuff
Jan 24 19:38:06 <nbd> so please stop poking me directly about your patches and find somebody else that has more time for review and design discussion
Jan 24 19:43:19 <danage> philipp64|laptop: he means that in a nice, sincere way :)
Jan 24 20:00:44 <philipp64|laptop> well, he did ask me to make sure that l2tpv3tun still built, so I took that to mean he had a special interest in it...
Jan 24 20:02:04 <EqUaTe> philipp64|laptop: well, if it builds and works, then the patches are likely right.. :)
Jan 24 20:06:36 <Epsylon3> i dunno what it builds, but it builds :)
Jan 24 20:07:01 <Epsylon3> it build and rebuild all except the needed thing :p
Jan 24 20:07:27 <Epsylon3> (my build box)
Jan 24 20:10:48 <EqUaTe> Epsylon3: what are you trying to build?
Jan 24 20:17:22 <nbd> philipp64|laptop: my only special interest in l2tpv3tun was that i identified it as the only package directly depending on libnl
Jan 24 20:17:34 <nbd> and i wanted to make sure your upgrade wouldn't cause regressions ;)
Jan 24 20:17:56 <nbd> (everything else had been converted to libnl-tiny already)
Jan 24 20:19:48 <nbd> and yes, danage was right. i didn't mean to come across as rude, i simply get distracted when too many people are repeatedly poking me about their stuff ;)
Jan 24 20:20:31 <danage> like me!
Jan 24 20:20:37 <russell--> nbd: scratch my problem with sysupgrade off your list, btw
Jan 24 20:23:15 <Acinonyx> https://dev.openwrt.org/wiki/SubmittingPatches
Jan 24 20:27:01 <Epsylon3> submit on trac
Jan 24 20:27:20 <Epsylon3> it s how i found the basics for wnr854t
Jan 24 20:27:27 <Epsylon3> leds
Jan 24 20:32:35 <Epsylon3> i will do that now...
Jan 24 20:32:44 <Epsylon3> no more patches, only solutions (full files)
Jan 24 20:32:52 <Acinonyx> lol
Jan 24 20:34:05 <Epsylon3> patch to create patches that patch seems to be ridiculous :)
Jan 24 20:35:16 <Epsylon3> i will make a little link on wiki to the device...
Jan 24 20:35:34 <Epsylon3> if people want it... they download and put in the target dir..
Jan 24 20:36:32 <danage> Epsylon3: sounds like patchwork, cause it will be revision specific?
Jan 24 20:36:49 <Epsylon3> no
Jan 24 20:36:54 <Epsylon3> that's patches :p
Jan 24 20:37:36 <Epsylon3> but sendfiles diff files to patch patch files is long to do and cant be reviewed
Jan 24 20:37:48 <Epsylon3> sending files
Jan 24 20:37:51 <Epsylon3> arfgh
Jan 24 20:38:03 <Epsylon3> my hands are tired
Jan 24 20:42:48 <Epsylon3> CONFIG_LEGACY_PTY_COUNT=256
Jan 24 20:43:02 <Epsylon3> hmm not too much ?
Jan 24 20:44:47 <Epsylon3> - /dev/ptya1 to /dev/ptyzf :p
Jan 24 20:45:22 <Epsylon3> and same for tty
Jan 24 20:51:33 <philipp64|laptop> as I remember, swconfig also links to -lnl and not -lnl-tiny
Jan 24 20:52:05 <Epsylon3> maybe a bad "depends"
Jan 24 20:52:06 <philipp64|laptop> from my point of view, what's frustrating is that the "peer review" process is opaque and not very predictable.
Jan 24 20:52:21 <philipp64|laptop> it would be nice if someone submitting a patch new what to expect.
Jan 24 20:53:28 <Epsylon3> ...xt_connmark.ko' specified more than once
Jan 24 20:53:44 <Epsylon3> hmm no it's a warning
Jan 24 20:54:57 <Epsylon3> ok, almost done.. wrn854t seems compatible with 2.6.37
Jan 24 20:55:15 <Epsylon3> lets go... compile all again... erf
Jan 24 20:56:04 <Epsylon3> there is a new orion device...
Jan 24 20:56:28 <Epsylon3> Freecom DataTank Gateway
Jan 24 20:56:35 <Epsylon3> (dt2)
Jan 24 20:57:02 <Epsylon3> with reset button irq implementation :)
Jan 24 21:22:21 <Epsylon3> coffee time...
Jan 24 21:22:52 <Epsylon3> hope to finish this reset button support this evening... tomorrow i get a WNR3500L
Jan 24 22:01:41 <Epsylon3> ok... finally... image ready after ... 4 hours ?:p
Jan 24 22:07:38 <Epsylon3> Linux version 2.6.37 (root@epsybox) (gcc version 4.3.3 (GCC) ) #1 Mon Jan 24 22:58:22 CET 2011
Jan 24 22:07:45 <Epsylon3> tada... :)
Jan 24 22:16:51 <Epsylon3> so... this reset button :)
Jan 24 22:18:02 <Epsylon3> hmmm no
Jan 24 22:18:03 <Epsylon3> :/
Jan 24 22:20:30 <Epsylon3> bad irq... 33 instead of 36
Jan 24 22:22:16 <Epsylon3> hmm no, 36 is for the wifi card
Jan 24 22:23:37 <Epsylon3> oh.. its maybe because i use gpio-keys too...
Jan 24 22:25:20 <_trine> xMff, ping
Jan 24 22:35:55 <Epsylon3> wnr854t
Jan 24 22:35:58 <Epsylon3> arf
Jan 24 22:36:18 <Epsylon3> EqUaTe : ... its what i'm building
Jan 24 22:37:10 <Epsylon3> hmm, if the irq doesnt works without gpio-keys... ill try hotplug script
Jan 24 22:38:30 <soma> hi. does anyone know if regdomain 0x00 works with all atheros cards or does it (still) depend on the value in the cards eeprom?
Jan 24 22:39:11 <soma> (talking about madwifi)
Jan 24 22:39:44 <Epsylon3> http://www.mobilnews.cz/blog/?p=36 : Atheros EEPROM checksum fix
Jan 24 22:40:34 <Epsylon3> soma : i don't know... in fact ive no problem (i think) with that
Jan 24 22:40:34 <soma> thx, but thats not what i'm looking for
Jan 24 22:40:48 <Epsylon3> i bough the card in US
Jan 24 22:40:51 <Epsylon3> and im in france
Jan 24 22:40:55 <Epsylon3> t
Jan 24 22:41:22 <Epsylon3> ath 5413
Jan 24 22:41:23 <soma> yes, my cm-9s are also from us and do work with 0x00, but i remember in the past i had to change the eeprom
Jan 24 22:41:45 <Epsylon3> its why i paste this link :)
Jan 24 22:41:59 <Epsylon3> to change eeprom if needed
Jan 24 22:42:16 <soma> and i'm currently working on a wizard and wonder if it makes more sense to set 0x00 or the real local regdomain
Jan 24 22:42:37 <Epsylon3> 0 is default ?
Jan 24 22:42:51 <Epsylon3> maybe all channels
Jan 24 22:43:03 <soma> 0 allows all country codes
Jan 24 22:43:41 <soma> but i'm unsure if this is accepted for all cards
Jan 24 22:43:44 <Epsylon3> you mean all channels ? 1-14 ?
Jan 24 22:44:21 <soma> no, all country codes. the regdomain says what ccs you are allowed to set
Jan 24 22:44:39 <Epsylon3> oh
Jan 24 22:45:07 <Epsylon3> how can i check that ?
Jan 24 22:45:11 <Epsylon3> i use madwifi too
Jan 24 22:45:35 <soma> cat /proc/sys/dev/wifi0/regdomain
Jan 24 22:45:46 <Epsylon3> 245
Jan 24 22:45:51 <Epsylon3> france is 250
Jan 24 22:46:07 <Epsylon3> seen that on mailing list i think
Jan 24 22:47:38 <soma> that doesn't look right, it should be something with 0x..
Jan 24 22:50:24 <soma> i wonder if i should try to set it at all, probably its to use whats in the eeprom and leave the rest to the user.
Jan 24 22:51:02 <soma> *best
Jan 24 22:53:01 <Epsylon3> yea.. i've set 'FR'
Jan 24 22:53:06 <Epsylon3> in wireless
Jan 24 22:53:18 <Epsylon3> but look like i need to set '250'
Jan 24 22:53:31 <soma> yes
Jan 24 22:54:02 <soma> i have a patch for that (accept also the two-char-codes), but no response until now :(
Jan 24 22:54:03 <Epsylon3> in fact its country key
Jan 24 22:54:06 <Epsylon3> for regdomain ?
Jan 24 22:54:13 <soma> no, for country
Jan 24 22:54:19 <Epsylon3> ok :)
Jan 24 22:54:45 <Epsylon3> wifi0: ath_getchannels: Unable to collect channel list from HAL; regdomain likely 245 country code 250
Jan 24 22:55:10 <soma> http://madwifi-project.org/wiki/UserDocs/CountryCode
Jan 24 22:55:22 <Epsylon3> so 'FR' is the right setting
Jan 24 22:55:24 <soma> 0x32 would be france, but try to set it to 0x00
Jan 24 22:55:27 <Epsylon3> no warning with that
Jan 24 22:55:50 <Epsylon3> hmm oki
Jan 24 22:56:21 <Epsylon3> wifi0: ath_getchannels: Unable to collect channel list from HAL; regdomain likely 245 country code 50
Jan 24 22:56:25 <Epsylon3> (for 0x32)
Jan 24 22:56:35 <soma> no, got me wrong again ;)
Jan 24 22:56:45 <Epsylon3> 0 is ok
Jan 24 22:56:45 <soma> just follow the link i posted
Jan 24 22:57:30 <Epsylon3> dev.wifi0.regdomain = 245
Jan 24 22:57:37 <Epsylon3> dev.wifi0.countrycode = 0
Jan 24 22:59:34 <philipp64|laptop> oh, deja vu!  I thought I was on #linux-wireless for a moment!  :-)
Jan 24 23:00:09 <nbd> they don't usually talk about madwifi on #linux-wireless ;)
Jan 24 23:00:23 <Epsylon3> :)
Jan 24 23:00:29 <Epsylon3> wifi0: ath_getchannels: Reverting to defaults
Jan 24 23:00:42 <soma> ah, nbd :) is regdomain setting independent from the eeprom meanwhile?
Jan 24 23:00:43 <Epsylon3> in fact yea, with a String code, its like 0
Jan 24 23:00:59 <nbd> soma: if you override the regdomain in the config, it'll ignore the eeprom
Jan 24 23:01:02 <Epsylon3> so its a module option
Jan 24 23:01:09 <Epsylon3> oh
Jan 24 23:01:14 <soma> thanks, thats what i wanted to know
Jan 24 23:01:23 <Epsylon3> maybe my eeprom is empty
Jan 24 23:01:40 <nbd> if your eeprom was empty, the driver would refuse to load
Jan 24 23:01:53 <Epsylon3> yea :p i ve seen that
Jan 24 23:01:59 <Epsylon3> i've 2 cards
Jan 24 23:01:59 <nbd> the hardware doesn't work without calibration data ;)
Jan 24 23:02:06 <Epsylon3> and one is broken (eeprom)
Jan 24 23:02:22 <nbd> did it break all by itself?
Jan 24 23:02:23 <Epsylon3> its possible to repair directly from the router ?
Jan 24 23:02:29 <nbd> or did it get some help?
Jan 24 23:02:33 <Epsylon3> yea... but not on this router
Jan 24 23:02:36 <Epsylon3> in my laptop
Jan 24 23:02:52 <nbd> hm
Jan 24 23:02:55 <Epsylon3> maybe... i played with different drivers
Jan 24 23:03:04 <Epsylon3> could be the reason
Jan 24 23:03:06 <nbd> if the calibration data is gone, you should get a new card
Jan 24 23:03:09 <nbd> unless you happen to have a backup
Jan 24 23:03:18 <Epsylon3> yea i ve
Jan 24 23:03:26 <Epsylon3> i've 2 cards
Jan 24 23:03:33 <nbd> calibration data is unique for every card
Jan 24 23:03:40 <Epsylon3> hmm
Jan 24 23:03:47 <nbd> every card is calibrated individually
Jan 24 23:04:04 <Epsylon3> http://www.mobilnews.cz/blog/?p=36
Jan 24 23:04:10 <Epsylon3> found that
Jan 24 23:04:14 <soma> epsilon: try regdomain=0x00 and country=250. if that doesn't work you could try to set regdomain with ath_info. if the eeprom is really borked listen to nbd :)
Jan 24 23:04:40 <Epsylon3> k, but its not this card which is broken
Jan 24 23:05:17 <_trine> I have a small problem with luci,,   if I have CONFIG_ATH_USER_REGD=y luci limits me to 15mW
Jan 24 23:05:28 <nbd> sometimes cloning the data from one card to another works
Jan 24 23:05:30 <_trine> on my nanostation M2
Jan 24 23:05:31 <Epsylon3> yea, no warning soma
Jan 24 23:05:43 <nbd> but you'll never get back the signal quality again
Jan 24 23:05:43 <Epsylon3> cat /proc/sys/dev/wifi0/regdomain
Jan 24 23:05:44 <Epsylon3> 0
Jan 24 23:05:51 <nbd> and sometimes the transmit power gets completely messed up
Jan 24 23:05:55 <Epsylon3> cat /proc/sys/dev/wifi0/countrycode : 250
Jan 24 23:06:34 <Epsylon3> its better than nothing
Jan 24 23:06:41 <Epsylon3> i'll try that...
Jan 24 23:06:51 <Epsylon3> is there a tool to backup/restore eeprom ?
Jan 24 23:06:58 <Epsylon3> on openwrt
Jan 24 23:07:20 <nbd> what kind of router do you have?
Jan 24 23:07:34 <Epsylon3> wnr854t
Jan 24 23:07:37 <Epsylon3> arm
Jan 24 23:07:39 <Epsylon3> orion
Jan 24 23:07:46 <nbd> hm
Jan 24 23:07:51 <Epsylon3> on trunk
Jan 24 23:07:52 <nbd> it's probably better to try it on the laptop instead
Jan 24 23:08:03 <Epsylon3> self compiled (2.6.37)
Jan 24 23:08:08 <nbd> i don't know if anybody has made the tools working on arm or mips yet
Jan 24 23:08:09 <philipp64|laptop> nbd: no, but crda is a frequent source of torment. :-)
Jan 24 23:08:14 <nbd> at least on mips they're know to be broken
Jan 24 23:08:28 <Epsylon3> i've source code
Jan 24 23:08:28 <nbd> philipp64|laptop: on openwrt it's more manageable than on regular linux distros ;)
Jan 24 23:08:38 <Epsylon3> http://www.mobilnews.cz/blog/files/ar5k_plnt.c
Jan 24 23:08:45 <Epsylon3> http://www.mobilnews.cz/blog/files/idchanger_plnt.c
Jan 24 23:09:44 <Epsylon3> its for 5211
Jan 24 23:11:02 <Epsylon3> ok... sysupgrade... trying to resolve the reset irq problem...
Jan 24 23:13:15 <philipp64|laptop> Openssh 5.7 is out!  Woohoo!  Now doing ssh-based backups and file transfers will no longer hose my VoIP sessions!!!
Jan 24 23:14:05 <nbd> what changed to make it not hose your voip sessions?
Jan 24 23:15:03 <Epsylon3> hmm no
Jan 24 23:15:04 <Epsylon3> setting trigger mode 8 for irq 33 failed (0xc0062ef0)
Jan 24 23:15:04 <Epsylon3> WNR854T: Reset Button IRQ 33 not available
Jan 24 23:15:13 <Epsylon3> what is this "mode 8"
Jan 24 23:15:28 <Epsylon3> i used TRIGGER_LOW
Jan 24 23:18:24 <philipp64|laptop> nbd: it now supports RFC-2474 traffic marking... it was until now using RFC-791 TOS markings if you can believe it.
Jan 24 23:19:01 <Epsylon3> i get that from "wifi up" : ioctl[IEEE80211_IOCTL_DELKEY]: Invalid argument
Jan 24 23:19:07 <Epsylon3> in 2.6.36 too
Jan 24 23:19:24 <Epsylon3> two times
Jan 24 23:19:34 <philipp64|laptop> https://bugzilla.mindrot.org/attachment.cgi?id=1949&action=diff
Jan 24 23:20:46 <nbd> ah, interesting
Jan 24 23:23:33 <Epsylon3> will try IRQF_TRIGGER_FALLING
Jan 24 23:29:22 <Epsylon3> another tool here : http://www.mobilnews.cz/honza/ath-eeprom-tool
Jan 24 23:31:17 <Epsylon3> :) found one for windows (x86/x64) : http://rghost.net/2341199
Jan 24 23:34:40 <_trine> this is weird why should CONFIG_ATH_USER_REGD=y limit the RF power to no more then 15mW
Jan 24 23:39:30 <_trine> going to bed gnite
Jan 24 23:41:30 <Epsylon3> maybe for some low power devices
Jan 24 23:42:01 <Epsylon3> ohh
Jan 24 23:42:19 <Epsylon3> maybe not authorized by your country
Jan 24 23:42:26 <Epsylon3> to use more power...
Jan 24 23:42:34 <Epsylon3> dunno how much in france...
Jan 24 23:43:53 <Epsylon3> ohh... i ve now new channels available :)
Jan 24 23:44:13 <Epsylon3> 36 40 42 48 50 52
Jan 24 23:44:25 <Epsylon3> 56 58 60 64
Jan 24 23:44:50 <Epsylon3> 50mW max
Jan 24 23:45:48 <Epsylon3> sysupgrade...
Jan 24 23:52:28 <Epsylon3> nope
Jan 24 23:52:31 <Epsylon3> WNR854T: Reset Button IRQ 33 not available
Jan 24 23:52:39 <Epsylon3> but no message on trigger method
Jan 24 23:57:03 <Epsylon3> mdio ...
Jan 24 23:57:10 <Epsylon3> what is this bus
Jan 24 23:57:22 <Epsylon3> i've 4 devices detected
Jan 24 23:58:11 <Epsylon3> ok, ethernet :p
Jan 25 00:03:10 <OutBackDingo> wow, you guys are scary
Jan 25 00:15:04 <Epsylon3> hmm i'll keep the gpio-keys settings...
Jan 25 00:24:02 <Epsylon3> :) https://sourceforge.net/projects/wifi-box/
Jan 25 00:24:21 <Epsylon3> was on this project... in 2004
Jan 25 00:25:59 <Epsylon3> lol, 125 000 downloads
Jan 25 00:26:15 <CIA-84> nbd * r25095 /trunk/target/linux/generic/ (7 files in 7 dirs):
Jan 25 00:26:15 <CIA-84> kernel: when bridging, do not forward EAP frames to other ports, only deliver them locally
Jan 25 00:26:15 <CIA-84> fixes WPA authentication issues with multiples APs that are connected to each other via bridges
Jan 25 00:26:15 <Epsylon3> of my release
Jan 25 00:29:14 <philipp64|laptop> nbd: was in a meeting... what's interesting?  the QoS stuff?
Jan 25 00:31:44 <philipp64|laptop> nbd: r25095 -- nice.
Jan 25 00:36:22 <Epsylon3> the git is updated in real time ?
Jan 25 00:36:38 <Epsylon3> i'm on r25091
Jan 25 00:36:40 <nbd> every few minutes
Jan 25 00:36:45 <nbd> cron job
Jan 25 00:36:49 <Epsylon3> ok
Jan 25 00:37:26 <Epsylon3> svn://svn.openwrt.org/openwrt/trunk@25091
Jan 25 00:37:32 <Epsylon3> weird
Jan 25 00:37:50 <KanjiMonster> nbd: ++      if (skb->protocol == htons(ETH_P_PAE)) { <- shouldn't this be ntohs?
Jan 25 00:38:12 <nbd> no
Jan 25 00:38:22 <nbd> skb->protocol is in network byte order
Jan 25 00:38:33 <nbd> ETH_P_PAE is in host byte order
Jan 25 00:38:37 <KanjiMonster> err, right
Jan 25 00:38:47 <KanjiMonster> got myself confused
Jan 25 00:38:47 <Epsylon3> oh ok... a merge was needed
Jan 25 00:38:57 <KanjiMonster> too late ;)
Jan 25 00:40:41 <builder> build #55 of xburst is complete: Failure [failed shell_13]  Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/55
Jan 25 01:25:45 <Epsylon3> samba 3 always broken ?
Jan 25 01:25:52 <Epsylon3> still
Jan 25 02:57:44 <builder> build #70 of ubicom32 is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/70
**** ENDING LOGGING AT Tue Jan 25 02:59:57 2011