**** BEGIN LOGGING AT Fri Apr 06 03:00:00 2007 Apr 06 04:51:36 w Apr 06 07:40:52 VFS: Cannot open root device "mtdblock1" or unknown-block(31,1) Apr 06 09:36:20 <[florian]> armijn: apparently this did not work out of the box with the c54apra, right ? Apr 06 09:37:42 from what he said, it appeared to be needing to be repartitioned Apr 06 09:38:44 <[florian]> humm Apr 06 09:39:41 we will definitely need to use an lzma loader for rdc-2.6 after all, but i've worked around that for the moment Apr 06 09:40:15 [florian]: indeed Apr 06 09:40:38 [florian]: but I don't know if it was because of one device, or if it is default behaviour Apr 06 09:40:38 atm, i'm getting the error above on boot Apr 06 09:41:27 armijn: i would assume it to be default until demonstrated otherwise Apr 06 09:41:45 sn9_: yeah, that's why I will try with a few more devices today Apr 06 09:41:59 I got a bunch of them :) Apr 06 09:46:47 <[florian]> armijn: this is pretty strange, because I just ran adam2flash.pl on the device, and it worked Apr 06 09:50:11 <[florian]> sn9_: any flash map update to test ? Apr 06 09:50:16 yes Apr 06 09:50:41 the flash mapping seems to work now on 2.6 Apr 06 09:50:53 <[florian]> ok, big.patch Apr 06 09:50:55 <[florian]> ? Apr 06 09:50:56 except for that error halting the boot Apr 06 09:51:07 wait, lemme reupload Apr 06 09:51:17 [florian]: that's why I will try with another device as well Apr 06 09:53:24 [florian]: ok, download Apr 06 09:54:22 <[florian]> sn9_: much cleaner :) Apr 06 09:54:43 yes Apr 06 09:56:29 <[florian]> I will test it in few minutes Apr 06 09:56:47 it will not boot Apr 06 09:59:48 <[florian]> size limitation ? Apr 06 09:59:51 no Apr 06 09:59:59 i worked around that Apr 06 11:07:57 roh: ping Apr 06 14:27:51 nbd * r6872 /trunk/target/linux/au1000-2.6/patches/012-mtx1_cmdline.patch: meshcube: override the boot loader's (wrong) kernel command line Apr 06 16:07:37 nbd * r6873 /trunk/package/base-files/files/etc/init.d/usb: use rc.common for usb init script Apr 06 16:28:04 could somebody please revert changesets 6849 and 6850? they cause a boatload of breakage Apr 06 16:28:22 http://www.ploetzli.ch/forumtst/kmod-bluetooth-bnep.diff Apr 06 16:29:29 loswillios ? eh.. yep? Apr 06 16:29:47 nbd * r6874 /trunk/package/kernel/modules/other.mk: add bnep to the bluetooth kmod package (patch by Henryk) Apr 06 16:35:38 <[florian]> sn9_: be more specific Apr 06 16:35:59 the resulting command lines are invalid Apr 06 16:36:31 <[florian]> looks like it is working for most people Apr 06 16:36:44 i find that hard to believe Apr 06 16:36:55 sn9_: what packages in peticular are being broekn by it. Apr 06 16:37:04 paste us some error messages.... Apr 06 16:37:13 <[florian]> sn9_: if it was the case we would have seen tickets about that Apr 06 16:37:16 err, pastebin* Apr 06 16:38:36 http://www.pastebin.ca/427090 Apr 06 16:38:47 all packages are affected Apr 06 16:39:30 it was correct before Apr 06 16:40:32 i have never seen any system where + can terminate -exec, only \; Apr 06 16:42:14 + has been in use for ages Apr 06 16:43:47 not only have i never seen it before, but i tried it on a couple of systems, and got stuff like in that paste Apr 06 16:44:23 <[florian]> it has been working wherever I tested it : macosx, gentoo, fedora Apr 06 16:44:39 Works for me also Apr 06 16:44:48 i tested on osx and slack -- will also test on ubuntu Apr 06 16:46:28 anything wrong with "\; 2>/dev/null" instead? Apr 06 16:59:58 nbd * r6875 /trunk/package/qos-scripts/ (9 files in 7 dirs): update qos-scripts to v1.1 - rewritten hfsc rate calculation Apr 06 17:00:13 a possible reason there are no tickets on the matter is that one would not easily realize what causes the problem Apr 06 17:13:59 http://www.opengroup.org/onlinepubs/009695399/utilities/find.html - perfectly legitimate (in freebsd find also) Apr 06 17:17:19 according to that page, the intended behavior of + is aggregation, which was the very problem with xargs in the first place Apr 06 17:40:10 I have never had a problem with either xargs or {} + (but I have just tested and ; is noticably slower) Apr 06 17:40:53 true, but when too many args are aggregated into one command line, it is rejected Apr 06 17:41:30 i was having that problem with xargs, and so were others Apr 06 17:42:00 what shell ? Apr 06 17:42:05 bash Apr 06 17:43:06 I don't use bash only zsh Apr 06 17:43:33 zsh is certainly less popular Apr 06 17:44:34 true Apr 06 17:46:09 "\; 2>/dev/null" seems like the least trouble for all concerned Apr 06 17:46:24 sn9_: you have broken xargs Apr 06 17:46:58 ejka: and find, apparently. but others were having the xargs issue, too Apr 06 17:48:10 ISTR find in older findutils didn't handle -exec ... + correctly Apr 06 17:48:25 right Apr 06 17:51:11 * ejka wonders what system is it, if it has both find and xargs broken Apr 06 17:51:58 xargs is also part of findutils, so the problems are inseparable Apr 06 17:52:49 I've never seen xargs not respecting cmdline length restrictions Apr 06 17:53:17 i was apparently not the only one who had that issue Apr 06 19:39:23 roh: query? Apr 06 20:34:22 nbd: i get this after /etc/init.d/qos restart: "RED: failed to calculate EWMA constant." (2x) Apr 06 21:51:56 interesting Apr 06 21:52:03 I installed on more c54apra machines Apr 06 21:52:09 at least, on one Apr 06 21:52:15 turned out it had another mtd layout Apr 06 23:09:08 nbd * r6876 /trunk/Makefile: use ~/.openwrt/defconfig as default config if it exists Apr 06 23:10:38 nbd: that is potentially dangerous Apr 06 23:10:42 I would advise against it Apr 06 23:11:12 it could cause some hefty debug sessions for people Apr 06 23:11:24 only to find out they have a ~/.openwrt/defconfig they forgot about Apr 06 23:11:49 as in, it would happen to me :P Apr 06 23:15:40 nbd * r6877 /trunk/ (Config.in scripts/download.pl): add a menuconfig option for specifying a local download mirror Apr 07 00:20:49 nbd * r6878 /trunk/package/qos-scripts/files/usr/lib/qos/tcrules.awk: qos-scripts: fix RED burst calculation **** ENDING LOGGING AT Sat Apr 07 02:59:57 2007