**** BEGIN LOGGING AT Mon Oct 10 02:59:57 2011 Oct 10 06:32:13 acoul * r28398 /packages/net/mtr/Makefile: net/mtr: update to version 0.81 Oct 10 09:27:14 florian * r28399 /trunk/package/kernel/modules/i2c.mk: [package] kernel: fix i2c section after r28334 Oct 10 11:22:16 nico * r28400 /packages/libs/libdlna/patches/300-ffmpeg_compat.patch: [packages] libdlna: fix build after ffmpeg update (closes: #10002) Oct 10 11:46:11 nico * r28401 /branches/packages_10.03.1/libs/tokyocabinet/patches/200-uclibc_0.9.30_compat.patch: Oct 10 11:46:11 [backfire/packages] tokyocabinet: remove uClibc compat patch Oct 10 11:46:11 We have the needed math support in source:branches/backfire/toolchain/uClibc/patches-0.9.30.1/150-c99_ldbl_math.patch Oct 10 14:01:23 Is anyone here working to port OpenWRT to a PogoPlug Pro (http://www.google.com/products/catalog?q=PogoPlug+Pro&um=1&ie=UTF-8&tbm=shop&cid=8167861145955058396&sa=X&ei=3PmSTs_GBMr10gHZrOnnBA&ved=0CG8Q8wIwBA)? Oct 10 14:24:29 nico * r28402 /packages/lang/distribute/Makefile: [packages] distribute: install host modules in hostpython search path Oct 10 14:25:48 nico * r28403 /packages/lang/mako/Makefile: [packages] mako: remove @BROKEN, it is building againg after [28402] Oct 10 15:08:52 nbd * r28404 /trunk/package/ubus/Makefile: ubus: update to latest version, includes a small bugfix for object call replies Oct 10 15:17:10 blogic * r28405 /trunk/target/linux/lantiq/ (127 files in 19 dirs): Oct 10 15:17:10 [lantiq] Oct 10 15:17:10 * update patches to 3.0 Oct 10 15:17:10 * add basic vr9 support Oct 10 15:17:10 * backport 3.1 fixes Oct 10 15:17:10 * backport 3.2 queue (falcon) Oct 10 15:17:39 blogic * r28406 /trunk/package/ (30 files in 3 dirs): Oct 10 15:17:39 [ltq-dsl] Oct 10 15:17:39 * add support for ar9 Oct 10 15:17:39 * add support for ase Oct 10 15:17:39 (vr9 support is still a todo) Oct 10 15:44:04 build #92 of s3c24xx is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/92 Oct 10 16:44:39 ping [florian] Oct 10 16:44:58 <[florian]> danage: pong Oct 10 16:45:07 hi there. did you get my message? Oct 10 16:45:09 <[florian]> danage: your device does not have any ethernet device registered, that's why it's broken for you Oct 10 16:45:28 <[florian]> danage: you can edit arch/mips/bcm63xx/boards/board_bcm963xx.c Oct 10 16:45:56 ok - did i maybe flash the wrong image? i used the -gw2 one (which corresponds to the board info in the boot loader) Oct 10 16:46:07 <[florian]> nope, that's theg good image Oct 10 16:46:16 <[florian]> it just lacks the missing knobs to register the ethernet drive Oct 10 16:46:18 <[florian]> +r Oct 10 16:46:23 alright, let's do it Oct 10 16:46:37 <[florian]> danage: so what you are missing as as simple as : Oct 10 16:46:39 * danage pulling up a console Oct 10 16:46:43 <[florian]> .has_enet0 = 1, Oct 10 16:46:46 <[florian]> .enet0 = { Oct 10 16:46:51 <[florian]> .force_speed_100 = 1, Oct 10 16:46:57 <[florian]> .force_duplex_full = 1, Oct 10 16:46:59 <[florian]> }, Oct 10 16:47:11 <[florian]> there is a switch connected to the BCM6345 on your device, right? Oct 10 16:47:48 i don't think so. it has one ethernet port only. well, plus maybe the connection to the DSL Oct 10 16:48:07 <[florian]> allright, let's try it that way anyway Oct 10 16:48:14 <[florian]> this should be working even though there is a PHY Oct 10 16:48:22 <[florian]> if not, it's relatively easy to change later on Oct 10 16:48:37 ok. i'm going to make the necessary changes and recompile. i am at work right now but i will try and report back later (when i'm at home) Oct 10 16:48:50 cd .. Oct 10 16:48:55 <[florian]> no problem Oct 10 16:49:02 fail Oct 10 17:00:22 [florian]: i see some .has_pci with other devices. should i add that too, for the wlan? Oct 10 17:15:47 mazilo * r28407 /packages/net/freeswitch/ (6 files in 5 dirs): Update to the latest GIT commit and added the new application mod_http_cache to download/play audio files from a webserver Oct 10 17:22:47 If I'm getting messages like, "note: someone does not honour COPTS correctly, passed 0 times" is that likely to be a problem in the owrt package makefile, or the original package makefile? Oct 10 17:22:54 (I can fix both) Oct 10 17:23:11 it seems to all work ok, but it seems like something that should be fixed Oct 10 17:24:09 karlp: it simply says that the openwrt TARGET_CFLAGS do not make it to the gcc commands Oct 10 17:24:31 so it could be either the owrt file or the package file or both Oct 10 17:24:47 you just need to make sure that the cflags get passed on properly Oct 10 19:34:53 [florian]: no workie. same issues. i think i might have changed the wrong file. which board_bcm963xx.c is the right one? Oct 10 19:35:18 <[florian]> build_dir/ ... Oct 10 19:35:23 ./build_dir/linux-brcm47xx/linux-3.0.3/arch/mips/bcm63xx/boards/board_bcm963xx.c Oct 10 19:35:37 ok. my bad. brb. Oct 10 19:36:04 <[florian]> np Oct 10 19:38:41 gotta ask again: or rather build_dir/linux-brcm63xx/linux-2.6.39.4/arch/mips/bcm63xx/boards? Oct 10 19:40:04 here's a crashdump that just turned up, btw: http://openwrt.pastebin.ca/2088849 Oct 10 19:44:03 <[florian]> 8MB of ram is very little Oct 10 19:44:09 <[florian]> you might consider disabling wifi entirely Oct 10 19:49:45 trunk/build_dir/linux-brcm63xx/linux-2.6.39.4/arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h:63:21: error: array type has incomplete element type Oct 10 19:50:18 <[florian]> can you show me the file? Oct 10 19:54:28 [florian]: http://openwrt.pastebin.ca/2088858 and http://openwrt.pastebin.ca/2088855 Oct 10 19:56:43 <[florian]> humm weird, let me build this too Oct 10 19:57:17 maybe i'm in the wrong directories. the file is in several Oct 10 19:57:51 <[florian]> no, this definitively looks like the right one Oct 10 19:58:47 <[florian]> I fail to reproduce this Oct 10 19:59:22 <[florian]> danage: can you start with a fresh make target/linux/clean && make target/linux/prepare ? Oct 10 19:59:39 sure Oct 10 21:04:01 [florian]: it did something to the eth, but still has the same problem. http://openwrt.pastebin.ca/2088882 Oct 10 21:05:03 [florian]: low memory does not seem to be of the issue. i invoked the "free" command a couple of times on the console. it's really slow, but freemem seemed finwe Oct 10 21:06:20 if it's really slow then the kernel is trying really hard to find some last remaining bits of free memory Oct 10 21:06:28 and atomic allocations for network driver buffers are much more likely to fail Oct 10 21:08:01 hi nbd. it reports 600k of free mem, non? Oct 10 21:08:20 that's very low Oct 10 21:08:50 * danage omitting bill gates quote Oct 10 21:09:01 i'll try without the wifi then Oct 11 01:25:47 tripolar * r28408 /packages/net/ntpd/Makefile: [packages]: update ntpd to 4.2.6p4 **** ENDING LOGGING AT Tue Oct 11 02:59:57 2011