**** BEGIN LOGGING AT Sun Aug 09 02:59:57 2009 Aug 09 07:25:09 is there a way of showing what variables are defined in the makefiles? Aug 09 07:30:40 .. trying to work out what defines to use to conditionally add options etc for x86_64.. using TARGET_ARCH will hopefully work. Aug 09 07:58:12 http://www.flickr.com/photos/lumachrome/3802191079/sizes/l/in/set-72157621984992162/ Aug 09 07:59:03 omg.. what's that for? Aug 09 07:59:09 heh Aug 09 07:59:49 that is a skypilot SkyExtender, which i am currently dismantling and trying to figure out of there are salvagable parts Aug 09 08:00:09 from the defunct metrofi network in portland oregon Aug 09 08:00:27 ah. cool. Aug 09 08:01:15 has a peplink manga-winti board inside, arm9 cpu Aug 09 08:01:37 biggest problem i have right now is finding the serial port on the peplink board Aug 09 08:02:02 http://www.tdb.com/~russell/photos/2009-08-09/P1030259.html Aug 09 08:02:19 i've probed around with an oscilloscope and haven't found it so far Aug 09 08:12:06 hey russel? what kind of antenna is this? is it horizontally polarized omni? Aug 09 08:14:54 Acinonyx: the skypilot has two radios Aug 09 08:15:12 the horns sticking down are 7.5 dBi 2.4GHz omnis Aug 09 08:15:29 under the radome are 8 45-degree 5.8GHz sectors Aug 09 08:15:39 with a special antenna selector board Aug 09 08:16:05 the omnis are normal vertically polarized jobbies Aug 09 08:20:01 in fact, i accidentally opened one of them up and it looks like one of these, except there are three helical sections: http://bargainhopping.com/cart/images/7DBI.jpg Aug 09 08:23:14 hmm.. for some reason, the lzo configure has decided my 64 bit void * is 4 bytes... even though size_t seems to be 8 - this is then failing my confugre. Aug 09 08:23:24 (x86_64 this is) Aug 09 08:24:51 http://pastey.net/120422 Aug 09 08:27:38 russell_: why so many antennas? do you know if all the sector antennas work simultaneously or a single one is selected each time? Aug 09 08:27:53 single at a time Aug 09 08:28:05 the sectors are for mesh backhaul Aug 09 08:28:25 they give them higher gain in a particular direction and don't pollute rf in other directions Aug 09 08:29:23 i think each sector is 18 dBi Aug 09 08:35:22 i see.. Aug 09 10:08:09 florian * r17188 /trunk/target/linux/brcm63xx/files/drivers/mtd/maps/bcm963xx-flash.c: [brcm63xx] fix assignment instead of comparison in the mtd driver (#5649) Aug 09 10:18:32 florian * r17189 /trunk/target/linux/brcm63xx/files/drivers/spi/bcm63xx_spi.c: [brcm63xx] remove useless hard reset in the SPI controller driver Aug 09 10:19:48 florian * r17190 /trunk/ (3 files in 3 dirs): [brcm63xx] add support for the Pirelli AG226G board (#5337) Aug 09 10:25:30 florian * r17191 /packages/net/vsftpd/ (4 files in 2 dirs): [package] update vsftpd to 2.1.2 (#5647) Aug 09 10:33:57 florian * r17192 /packages/net/vpnc/ (4 files in 2 dirs): [package] update vpnc to 0.5.3 (#5648) Aug 09 10:44:53 florian * r17193 /packages/net/stunnel/ (5 files in 3 dirs): [package] update stunnel to 4.27 (#5650) Aug 09 10:49:02 florian * r17194 /packages/net/tinc/Makefile: [package] update tinc to 1.0.9, fix missing dependency against kmod-tun (#5651) Aug 09 10:51:29 florian * r17195 /packages/net/imspector/Makefile: [package] fix missing libpthread dependency in imspector makefile (#5654) Aug 09 10:58:49 florian * r17196 /packages/ipv6/miredo/Makefile: [package] update miredo to 1.1.7 (#5632) Aug 09 11:03:09 florian * r17197 /packages/ipv6/miredo/ (Makefile files/miredo-server.init files/miredo.init): [package] fix miredo init scripts and bump release number (#5632) Aug 09 11:11:13 florian * r17198 /trunk/target/sdk/Makefile: [sdk] do not exclude the share directory for bison and other programs to work (#5629) Aug 09 11:12:59 florian * r17199 /branches/8.09/target/sdk/Makefile: merge r17198 to 8.09 Aug 09 12:04:55 <_trine> can someone have a look at why since yesterday webif won't compile anymore in trunk? Aug 09 12:05:43 <_trine> make[3] -C feeds/xwrt/webif compile it gets to here then fails Aug 09 12:06:13 <_trine> it compiles OK if I don't include webif Aug 09 12:08:23 hard to tell w/o V=99 output Aug 09 12:08:33 maybe some required libaray/package does not compile? Aug 09 12:09:47 <_trine> ok I'll do V=99 Aug 09 12:10:17 <_trine> it compiled ok yesterday Aug 09 12:10:43 then it was changed in the meantime ;) Aug 09 12:11:09 <_trine> and this morning it compiles ok with luci and without any gui also Aug 09 12:11:34 <_trine> my guess is,, something in the universe has changed :) Aug 09 12:15:54 <_trine> http://pastebin.com/m7ea98fa5 Aug 09 12:18:41 this is not enough info, do you build with -j ? Aug 09 12:18:48 <_trine> yes Aug 09 12:18:58 run again without -j and V=99 Aug 09 12:19:03 <_trine> ok Aug 09 12:19:15 moment Aug 09 12:19:38 you can speed it up: make package/feeds/xwrt/webif/{clean,compile} V=99 Aug 09 12:20:03 <_trine> I had already started the make Aug 09 12:20:06 ok Aug 09 12:20:24 <_trine> it does not take long Aug 09 12:22:16 <_trine> same error Aug 09 12:26:02 <_trine> This is the whole thing Aug 09 12:26:05 <_trine> http://pastebin.com/d7bc97411 Aug 09 12:26:21 xMff ping Aug 09 12:27:00 xMff http://openwrt.pastebin.com/d71f162ef Aug 09 12:27:47 * russell_ finds the serial console! Aug 09 12:28:46 _trine: this is no xwrt problem, looks like the build-system pulls in your host /tmp somehow Aug 09 12:29:42 russell_ : which one was the serial console? Aug 09 12:29:49 J4? Aug 09 12:30:15 _trine: what does ls -l /home/john/trunk/staging_dir/target-mipsel_uClibc-0.9.30.1/root-brcm47xx/var tell? Aug 09 12:30:23 s/console/port/ Aug 09 12:30:29 <_trine> xMff, it compiled this time Aug 09 12:30:42 <_trine> when I used sudo make -j2 Aug 09 12:30:56 Dogge_: is it gone in trunk? Aug 09 12:32:24 xMff yes Aug 09 12:32:33 yeah just seen it Aug 09 12:32:44 can't help much there beside removing the related app Aug 09 12:33:10 nbd * r17200 /trunk/target/linux/generic-2.6/ (2 files in 2 dirs): revert some accidentally committed test/debug changes from the module symbol export patch - should fix #5655 Aug 09 12:33:55 xMff it's ok to remove the app, since i'm preparing to write an luci app for mjpg-streamer, since there is now uci config for mjpg-streamer :) Aug 09 12:34:46 allright. thing is, I'm working on the very basics (better wireless and network config) right now so I don't have much time to care for the multimedia packages atm Aug 09 12:36:04 Acinonyx: yes, j4 pins 5 and 6 Aug 09 12:39:06 it's got 32 meg of ram and 16 meg of flash on two chips Aug 09 12:40:06 unusual pinout... Aug 09 12:41:14 http://www.personaltelco.net/~russell/peplink-sbc-winti.bootlog Aug 09 12:43:05 the bootloader is kind of funny, it loads a ramdisk image into a ramdisk and runs from that... ext2 in a ramdisk. Aug 09 12:44:10 <[florian]> russell_: there is a linux 2.6 micrel ks8695 port, did you try it? Aug 09 12:44:25 where do i find it? Aug 09 12:47:57 <[florian]> oh it is in the vanilla sources Aug 09 12:48:30 yeah, i think i did see that, but i'm not sure how to turn that into something i can try Aug 09 12:49:02 <[florian]> you should determine how to rebuild an image, then see what kind of files you assemble to make it a flashable firmware image Aug 09 12:49:30 yeah, i've just started to look at the peplink sdk Aug 09 12:50:44 http://downloads.sourceforge.net/project/pemare/MANGA%20SDK/2.4.1/manga-sdk-2.4.1.tar.gz Aug 09 12:51:28 the kernel and rdisk.gz are flashed separately iirc Aug 09 12:51:57 using some weird binary tool called ethflash or something Aug 09 12:52:05 not sure if it's tftp or what Aug 09 12:54:16 s/ethflash/ethloader/ Aug 09 12:56:08 juhosg * r17201 /trunk/target/linux/ar71xx/files/drivers/mtd/wrt160nl_part.c: [ar71xx] WRT160NL: more mtd parser cleanup Aug 09 13:52:04 nbd trunk is still broken Aug 09 13:52:19 or again Aug 09 14:01:46 details? Aug 09 14:02:23 just rebuilding target/linux to see if it's the same error as yesterday Aug 09 14:05:16 nbd yeah, same error as yesterday Aug 09 14:06:26 nbd http://openwrt.pastebin.com/d2375082c Aug 09 14:12:26 juhosg * r17202 /trunk/target/linux/ar71xx/base-files/lib/upgrade/platform.sh: [ar71xx] enable sysupgrade on the WRT160Nl Aug 09 14:13:28 juhosg * r17203 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-wrt160nl.c: [ar71xx] WRT160NL: add default trigger for the power led Aug 09 14:14:08 whow Aug 09 14:14:16 someone is having a wrt160nl spree :) Aug 09 14:22:56 any idea, nbd? Aug 09 14:23:09 Hopscotch: no, but i got one friday, so i can work on that :) Aug 09 14:23:16 Hopscotch: btw, thank you for the patches Aug 09 14:23:23 Dogge_: testing Aug 09 14:23:29 nbd ok Aug 09 14:23:43 juhosg: no problem - just for my own benefit :) Aug 09 14:24:18 juhosg: i suspect a problem with the ehci though Aug 09 14:25:06 Hopscotch: what kind of problem? Aug 09 14:25:15 wwhen i tried to get dvb-t working i saw lots of oopses/segfaults/etc Aug 09 14:25:39 further investigation showed, that the ehci seems to write beyond the end of the request buffers Aug 09 14:26:10 i doesn't show in the storage code Aug 09 14:27:21 i made all the buffers in the dvb-t driver bugger to cope with the garbage Aug 09 14:27:33 after that the oopses/segfaults/etc disappeared Aug 09 14:28:48 weird, can you try it with 2.6.30.4? Aug 09 14:29:52 noot today, but i will Aug 09 14:30:02 ok Aug 09 14:58:02 Dogge_: btw. before cleaning your kernel tree, did you update to r17200? Aug 09 15:09:06 nbd: guide me some programming so that I can be helpful to you guys :) Aug 09 15:13:04 nbd: why don't you open up some smal online academy so that people can learn more about openwrt and some openwrt get some fund, I don't know how openwrt get funds now Aug 09 15:14:45 You have very good and many years experience, spread your knowledge and get some fund for openwrt Aug 09 15:16:55 nbd seems to be working now. no idea what was wrong Aug 09 15:31:23 mankash: i like doing stuff more than i like talking about it Aug 09 16:07:30 nbd: even I am like that, I agree with you one should do what one like, np Aug 09 16:08:06 yeah, anyone is to his own :) Aug 09 16:20:49 Is there any other way to see txpower other then using iwconfig. iwconfig always shows 0, Aug 09 16:21:13 This is atheros based card, clients gets associated fine Aug 09 18:06:55 how to trouble shoot if wifi clients are unable to associate Aug 09 18:08:38 one more question, when router is started it detect wan and lan ports, but in my case wan does't starts untill KS8995 is actiavted and detection happens before that, how to fix that Aug 09 20:51:05 nbd * r17204 /trunk/target/linux/generic-2.6/patches-2.6.28/027-module_exports.patch: arm: when linking the kernel, move the section discards after the other sections to fix discarding unused symbol exports Aug 09 22:45:13 acinonyx * r17205 /packages/utils/dbus/patches/01-dbus-nopie-fix.patch: [packages] dbus: Fix whitespaces in no pie patch Aug 09 22:51:05 acinonyx * r17206 /trunk/package/kernel/modules/wireless.mk: [package] kmod-net-hermes-pci, kmod-net-hermes-plx: Fix kmod-net-hermes dependency Aug 09 22:54:28 acinonyx * r17207 /packages/libs/net-snmp/Makefile: [packages] net-snmp: Enable ipv6 support if defined globally Aug 09 22:59:45 g'night all Aug 09 23:18:36 blah my patches fall silent Aug 09 23:54:30 kupesoft: I'm afraid that can happen. I've had some patches go through quickly, but others just seem to fall on deaf ears.. or 'busy' ears more likely. Aug 09 23:57:40 kupesoft: I've got a big patch for apache Makefile that's been kicking round for about a year. every now and again I try to get it through.. Aug 09 23:57:47 lets hope this last one is the ticket. Aug 09 23:59:15 oh, stop complaining :P Aug 09 23:59:25 it makes us look bad :( Aug 10 00:01:51 :) Aug 10 01:46:18 Anybody got clues on why this happens? This is on a x86 64bit target (I created one) - see that void* (void_p) is down as 4bytes long: http://pastey.net/120422 Aug 10 01:46:31 .. whereas size_t is 8 bytes as expected. Aug 10 01:53:05 - this is compiling LZO library required by OpenVPN. Aug 10 02:01:33 can you paste "src/miniacc.h" ? Aug 10 02:09:43 It should be the same on your system. Aug 10 02:10:00 it just checks that the types agree.. which they don't seem to. Aug 10 02:12:34 http://pastebin.ca/1523521 Aug 10 02:28:59 xMff: ac_cv_sizeof_void_p=4 Aug 10 02:56:52 xMff: are you there? **** ENDING LOGGING AT Mon Aug 10 02:59:57 2009