**** BEGIN LOGGING AT Thu Jan 07 02:59:56 2010
Jan 07 07:21:01 <glp> phaidros: ping
Jan 07 08:12:40 <phaidros> glp: pong
Jan 07 10:07:16 <danage> Bartman007: http://en.wikipedia.org/wiki/In_re_Bilski
Jan 07 12:00:00 <cshore> ping [florian]
Jan 07 12:26:16 <[Fate]> hi. in a DEPENDS line, how do I check for a kernel symbol to be selected as 'Y'es, not 'M'odular?
Jan 07 12:27:24 <cshore> if the package is 'Y'es it dependencies should automatically be 'Y'es as well
Jan 07 12:28:39 <[Fate]> they are, but that was not my question ;)
Jan 07 12:28:40 <[florian]> cshore: pong
Jan 07 12:30:48 <cshore> have you looked at the -A for dnsmasq?
Jan 07 12:32:49 <cshore> also for CTRL-C not working in brcm63xx to enter failsafe, could you look at ticket #5516 (the latest patch)....I'd like to be able to enter failsafe without patching every build
Jan 07 12:33:46 <[florian]> allright
Jan 07 12:37:51 <CIA-36> jow * r19061 /trunk/package/opkg/ (8 files in 2 dirs): [package] opkg: update to r513, move lock-file to /var/lock and drop cleanup patch - went upstream
Jan 07 12:44:13 <cshore> [florian] sorry to pounce on you...it's just hard to get in touch when I'm normally asleep when you're around
Jan 07 13:07:13 <nunojpg> some button clicks go unotice if they are too short
Jan 07 13:07:38 <nunojpg> is that something to do with threshold or _BUTTONS_POLL_INTERVAL?
Jan 07 13:07:56 <jow_laptop> poll interval sounds reasonable
Jan 07 13:08:17 <nunojpg> #define WRT160NL_BUTTONS_POLL_INTERVAL	20
Jan 07 13:08:25 <nunojpg> this means mili seconds?
Jan 07 13:09:05 <jow_laptop> maybe, I don't know
Jan 07 13:09:11 <nunojpg> :)
Jan 07 13:10:32 <nunojpg> my doubt is about threshold...maybe it's for debouce, but that hardly is done in software:S
Jan 07 13:10:34 <puchu> hi
Jan 07 13:10:45 <puchu> how can i decode kernel oops on mipsel?
Jan 07 13:10:55 <puchu> i compiled ksymoops but it doesnt seem to work
Jan 07 13:11:22 <jow_laptop> compile your kernel with debugging symbols
Jan 07 13:11:41 <jow_laptop> you can also manually decode the oops by comparing the memory addresses with your System.map in build_dir
Jan 07 13:11:43 <puchu> where do i enable this?
Jan 07 13:12:33 <puchu> CONFIG_DEBUG_BUGVERBOSE this option?
Jan 07 13:12:43 <puchu> ahh CONFIG_DEBUG_KERNEL
Jan 07 13:12:54 <jow_laptop> Global build settings -> Compile the kernel with symbol table information
Jan 07 13:13:42 <puchu> thanks
Jan 07 13:26:03 <Lalloso> how can i check if my architecture is little endian or big endian?
Jan 07 13:28:51 <cshore> #define LITTLE_ENDIAN 0
Jan 07 13:28:51 <cshore> #define BIG_ENDIAN    1
Jan 07 13:28:51 <cshore> int machineEndianness()
Jan 07 13:28:51 <cshore> {
Jan 07 13:28:51 <cshore>    int i = 1;
Jan 07 13:28:52 <cshore>    char *p = (char *) &i;
Jan 07 13:28:54 <cshore>    if (p[0] == 1) // Lowest address contains the least significant byte
Jan 07 13:28:56 <cshore>       return LITTLE_ENDIAN;
Jan 07 13:28:58 <cshore>    else
Jan 07 13:29:00 <cshore>       return BIG_ENDIAN;
Jan 07 13:29:02 <cshore> }
Jan 07 13:29:06 <cshore> from wikipedia
Jan 07 13:29:08 <cshore> compile and run
Jan 07 13:29:42 <cshore> with suitable wrapper code
Jan 07 13:30:17 <cshore> or
Jan 07 13:30:17 <cshore> #include <stdio.h>
Jan 07 13:30:17 <cshore> typedef union
Jan 07 13:30:17 <cshore> {
Jan 07 13:30:17 <cshore>         int i;
Jan 07 13:30:18 <cshore>         char c[4];
Jan 07 13:30:20 <cshore> }u;
Jan 07 13:30:22 <cshore> int main()
Jan 07 13:30:24 <cshore> {
Jan 07 13:30:26 <cshore>         u temp;
Jan 07 13:30:28 <cshore>         temp.i = 0x12345678;
Jan 07 13:30:30 <cshore>         printf("%x\n", temp.i);
Jan 07 13:30:32 <cshore>         printf("%x %x %x %x\n", temp.c[0], temp.c[1], temp.c[2], temp.c[3]);
Jan 07 13:30:34 <cshore> }
Jan 07 13:30:36 <cshore> If you get the output of the second line as "78 56 34 12", the machine is little endian.
Jan 07 13:30:38 <cshore> Instead if the output is "12 34 56 78", it's a big endian machine.
Jan 07 13:30:40 <cshore> __________________
Jan 07 13:30:44 <cshore> google is your friend
Jan 07 13:35:22 <jow_laptop> here's a simple way to test using lua: lua -e 'print(string.byte(string.dump(function() end), 7) == 0 and "big endian" or "low endian")'
Jan 07 13:35:59 <cshore> cool
Jan 07 13:37:41 <mb__> From the commandline:
Jan 07 13:37:42 <mb__> mb@homer:~$ file /bin/cp
Jan 07 13:37:44 <mb__> /bin/cp: ELF 32-bit MSB executable, PowerPC or cisco 4500, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.8, with unknown capability 0x41000000 = 0xf676e75, stripped
Jan 07 13:48:05 <Lalloso> ah thx lol
Jan 07 13:48:16 <Lalloso> guro hackers here
Jan 07 13:48:35 <Lalloso> next hint will be
Jan 07 13:48:38 <Lalloso> echo little endian :-D
Jan 07 13:48:52 <loswillios> heh
Jan 07 13:49:12 <Lalloso> but where do you get the endianess from the output of file command?
Jan 07 13:49:32 <Lalloso> is little the L in elf?
Jan 07 13:50:11 <cshore> MSB = most-significant bit machine, LSB = least-signifcant-bit-machine
Jan 07 13:50:33 <cshore> aka bigendian/littlendian
Jan 07 13:51:22 <cshore> well really executable but if you're running it, it's what your machine handles
Jan 07 13:51:50 <danage> i liked the one with the 50 lines of code
Jan 07 13:53:33 <cshore> heh...it's probably what I would have used had I need to do it....it depends on what you know
Jan 07 13:57:50 <cshore> I'm still not used to being able ask people actually know something a question...I've spent over two decades figuring out things from docs and code (and textbooks)
Jan 07 14:09:27 <cshore> is there an open source replace for the b43 wireless driver yet?
Jan 07 14:10:11 <loswillios> b43 is opensource
Jan 07 14:11:20 <cshore> oh, I thought it required proprietary stuff....or is only the firmware that is proprietary....for some reason I thought there was a binary-only driver in use for bcm wireless...or is that how it used to be?
Jan 07 14:11:48 <loswillios> that's how it used to be and it is still the case on 2.4
Jan 07 14:11:56 <cshore> ah, ok
Jan 07 14:12:07 <loswillios> there's also an opensource firmware for b43 available
Jan 07 14:12:26 <cshore> is it available on OpenWRT?
Jan 07 14:12:48 <mb__> yes, it doesn't support AP mode, however
Jan 07 14:12:49 <loswillios> i think i have seen it, yes
Jan 07 14:14:42 <cshore> what chipsets is there AP access for? (both opensource and closed source)
Jan 07 14:15:15 <cshore> hostapd is fairly limited that way on Debian, IIRC
Jan 07 14:15:26 <loswillios> http://linuxwireless.org/en/users/Drivers
Jan 07 14:23:54 <cshore> ah, I found the hostapd docs there...it seems I was looking at old docs that were before nl80211
Jan 07 14:25:31 <stintel>  
Jan 07 15:00:39 <nunojpg> what's the path of a button click event up to hotplug?
Jan 07 15:01:05 <nunojpg> can I check the button state without hotplug only with a kernel module?
Jan 07 15:53:21 <nunojpg> do I have to bride someone to get an account? :S
Jan 07 16:00:49 <loswillios> bride or bribe :p
Jan 07 16:05:59 <nunojpg> bribe looks easier, after checking google :D
Jan 07 16:06:07 <CIA-36> mb * r19062 /trunk/target/linux/brcm47xx/ (27 files in 2 dirs): bcm47xx: Add basic 2.6.32 support.
Jan 07 16:08:30 <loswillios> ;-)
Jan 07 16:22:51 <Lalloso> guys
Jan 07 16:23:02 <Lalloso> how do I check if i'm on a floating point or a fixed point device?
Jan 07 16:24:11 <Lalloso> well it's this board
Jan 07 16:24:12 <Lalloso> http://www.ubnt.com/products/rspro.php
Jan 07 16:29:35 <puchu> hi
Jan 07 16:29:36 <puchu>  is there some way to define options for modules that get loaded at boot time?
Jan 07 16:31:50 <jow_laptop>  /etc/modules/
Jan 07 16:34:59 <puchu> jow_laptop yeah but how can i define an option? $modulename $moduleoption=1 ? there are only files in /etc/modules.d/ that contain no option (on my sys) ...
Jan 07 16:35:32 <jow_laptop> just add the options after the module name
Jan 07 16:36:31 <stintel> hrrm, r18662 seems to have broken client mode + WEP
Jan 07 16:36:41 <jow_laptop> arch/driver ?
Jan 07 16:36:43 <puchu> jow_laptop: thanks
Jan 07 16:37:29 <stintel> jow_laptop: ar71xx/ath9k
Jan 07 16:38:32 <stintel> the variable keystring is never set, so the iw dev wlan0 connect ssid key $keystring is never executed
Jan 07 16:40:51 <stintel> or, if I set keymgmt then the keystring variable is set but then the iw dev wlan0 connect ... isn't exeecuted because keymgmt is not zero
Jan 07 16:44:58 <Lalloso> my atheros chip has mips32 24k
Jan 07 16:45:08 <Lalloso> i guess it's not 24kf but plain 24k
Jan 07 16:45:16 <Lalloso> therefore without floating point support
Jan 07 16:46:24 <jow_laptop> stintel: how does your /e/c/wireless looks like?
Jan 07 16:46:52 <stintel> jow_laptop: http://pastebin.com/m18a9786b
Jan 07 16:47:10 <stintel> jow_laptop: below is output of "wifi" with -x added to the shebang line
Jan 07 16:47:45 <jow_laptop> stintel: the wep configuration is slightly different... "key" is supposed to be "1" and "key1" should contain the actual secret
Jan 07 16:47:54 <stintel> i tried also to use key0 or key1, put d:0:0123456789, etc, but I ended up hacking mac80211.sh
Jan 07 16:48:28 <stintel> possible that it has changed recently?
Jan 07 16:48:56 <stintel> coz I normally just use Luci, and that worked before
Jan 07 16:49:22 <jow_laptop> well yes, LuCI implements it wrong and the wpa-like mode worked for some time but it was never the right way
Jan 07 16:50:14 <jow_laptop> I noticed that only recently as well while examining the scripts to write the wireless documentation
Jan 07 16:51:20 <jow_laptop> personally I think the wep key schema is awkward, but then I never understood the concept of the four different key slots
Jan 07 16:53:43 <jow_laptop> maybe the scripts should be changed to accept just "key" as well
Jan 07 16:56:31 <stintel> jow_laptop: you meant something like this? http://pastebin.com/m4c55be9b
Jan 07 16:56:43 <stintel> or I didnt understand you correctly I'm afraid
Jan 07 16:56:54 <jow_laptop> yes, I meant this
Jan 07 16:57:00 <stintel> still fails somehow
Jan 07 16:57:12 <jow_laptop> okay, then we have an actual bug :)
Jan 07 16:57:45 <stintel> ahh lame, I would prefer to debug this issue from home, but I also like to fix it before leaving here :P
Jan 07 17:21:24 <[Fate]> nbd: ping
Jan 07 19:15:07 <CIA-36> nbd * r19063 /trunk/package/mac80211/ (23 files in 2 dirs): mac80211: update to compat-wireless 2010-01-07 (unoffical snapshot, official ones are not working yet)
Jan 07 19:15:31 <CIA-36> nbd * r19064 /trunk/package/mac80211/patches/550-mac80211_action_frame_fix.patch: mac80211: fix spurious client reconnects triggered by rejected action frames being interpreted as chan switch announcements
Jan 07 19:59:08 <[Fate]> is there a way to check for a kernel configuration symbol from within a DEPENDS line?
Jan 07 21:59:53 <CIA-36> nbd * r19065 /trunk/package/mac80211/Makefile: mac80211: export a few ath header files
Jan 07 21:59:58 <CIA-36> nbd * r19066 /trunk/package/carl9170/ (5 files in 2 dirs): add the carl9170 wifi driver for ar9170 devices
Jan 07 22:22:12 <CIA-36> nbd * r19067 /trunk/package/carl9170/ (Makefile patches/120-fix_missing_include.patch): carl9170: fix debugfs support
Jan 07 22:51:10 <CIA-36> nbd * r19068 /trunk/package/carl9170/patches/130-ht_default.patch: carl9170: enable ht support by default
Jan 08 00:03:10 <CIA-36> jow * r19069 /trunk/package/dnsmasq/ (Makefile files/dnsmasq.init):
Jan 08 00:03:10 <CIA-36> [PATCH] dnsmasq: Fix setting the network-id for config host sections
Jan 08 00:03:10 <CIA-36> This renames name to networkid where applicable, and re-adds the name option for hosts.
Jan 08 00:03:10 <CIA-36> Signed-off-by: Stefan Bethke <stb@lassitu.de>
Jan 08 00:08:44 <CIA-36> jow * r19070 /packages/libs/liboping/ (Makefile patches/100-fd_error.patch): [packages] liboping: update to v1.3.4 (#6462)
Jan 08 00:09:15 <CIA-36> jow * r19071 /packages/libs/liboping/patches/100-fd_error.patch: [packages] liboping: remove now empty 100-fd_error.patch
Jan 08 00:12:02 <CIA-36> jow * r19072 /packages/net/httping/ (3 files in 2 dirs): [packages] httping: update to v1.4.0 (#6461)
Jan 08 00:14:13 <CIA-36> jow * r19073 /packages/net/transmission/Makefile: [packages] transmission: update to v1.77 (#6465)
Jan 08 00:15:08 <CIA-36> jow * r19074 /packages/net/transmission/patches/ (. 001-debian_dont_build_libevent.patch): [packages] transmission: add missing patch
**** ENDING LOGGING AT Fri Jan 08 02:59:57 2010