**** BEGIN LOGGING AT Wed Jun 10 02:59:57 2009 Jun 10 04:08:07 what does "mini_fo_open: invalid file state detected." mean? Jun 10 04:08:23 i see three of those messages on shutting down Jun 10 04:08:30 (on the serial console) Jun 10 07:49:52 running configure fragment for nptl/sysdeps/pthread Jun 10 07:49:52 "configure: error: compiler support for __thread is required" when compiling trunk with glibc ? Jun 10 07:50:01 Any good ideas ? Jun 10 08:24:15 florian * r16396 /trunk/ (5 files in 4 dirs): Jun 10 08:24:15 The attached patch adds building of firmware Jun 10 08:24:15 images for more routers based on the brcm63xx chipset. Jun 10 08:24:15 Signed-Off By: Daniel Dickinson Jun 10 10:31:36 [florian]: i will be gone from tomorrow morning-sunday Jun 10 11:58:38 <[florian]> danage: ok Jun 10 11:58:59 np? Jun 10 12:00:11 <[florian]> danage: no problems at all Jun 10 12:00:36 ok cool. will i miss all the action though? Jun 10 12:01:56 <[florian]> danage: not sure, I will be working on bcm6345 for sure, but I do not expect gettings much results Jun 10 12:02:37 :( Jun 10 12:02:48 <[florian]> danage: can you send me your System.map by mail please ? Jun 10 12:03:07 sure, florian@openw....? Jun 10 12:03:36 i can't do it until i get home though, i have the build env on my desktop Jun 10 12:04:34 <[florian]> yup Jun 10 12:04:45 <[florian]> no problems Jun 10 12:10:38 !seen paul Jun 10 12:23:57 hcg * r16397 /trunk/target/linux/generic-2.6/patches-2.6.30/ (023-mips_delay_gcc4.4.0.patch 024-mips_delay.patch): Remove deprecated 023-mips_delay_gcc4.4.0.patch and add new delay fix for MIPS Jun 10 12:30:50 acinonyx * r16398 /packages/net/quagga/patches/140-holdtimer-set.patch: [packages] quagga: Set BGP Hold Timer to a large value on OpenSent state (RFC1771, Par.8) Jun 10 12:32:41 acinonyx * r16399 /packages/net/quagga/patches/ (7 files): [packages] quagga: refresh patches Jun 10 13:43:50 florian * r16400 /trunk/target/linux/brcm63xx/image/Makefile: [brcm63xx] use build system REVISION variable (#5317) Jun 10 13:59:30 florian * r16401 /packages/admin/monit/ (Makefile patches/001-no_sslv2.patch): [package] update monit to 5.0.3 (#5284) Jun 10 14:15:01 [florian]: email sent. is the patch up there relevant? Jun 10 14:19:30 <[florian]> danage: got the System.map thanks ! Jun 10 14:19:47 <[florian]> danage: seems definitively related to the uart though I do not know how to fix it up yt Jun 10 14:20:03 excuse my nosiness, but system.map is unique on my system? Jun 10 14:21:05 <[florian]> it is relatively unique for a given build Jun 10 14:21:22 <[florian]> whenever you compile something in or out in the kernel the System.map will change Jun 10 14:25:04 interesting. well i wish i would be capable of providing any more help than this Jun 10 14:37:38 nbd * r16402 /trunk/package/wprobe/ (9 files in 6 dirs): wprobe: move measurement task to the kernel, add some configurability (work in progress) Jun 10 14:37:47 nbd * r16403 /trunk/include/target.mk: enable ucitrigger by default Jun 10 15:18:27 hcg * r16404 /trunk/target/linux/generic-2.6/patches-2.6.30/024-mips_delay.patch: Cleaned up whitespace in patch 024-mips_delay.patch Jun 10 15:44:45 russell_: ping Jun 10 19:17:55 Hello xMff Jun 10 19:20:30 xMff if I want compile lua script, with build root... I can do that in .postinst file or It will compile with lua of my system? Jun 10 19:30:38 fofware: you should do that during host build ... with the host lua Jun 10 19:30:48 the standard lua bytecode is incompatible with openwrt lua Jun 10 19:31:39 host lua = lua compiled by openwrt for the host, not on the host isntalled lua Jun 10 19:35:58 xMff, I did this question because maybe one year ago... before luci start, I did try it and when I did compile it with luac in make script, the script do not work in router because was for other platform, after i see in luci they have that like option in menuconfig Jun 10 19:36:46 I have not experience in linux, so some times I can't understand the scripts Jun 10 19:37:51 I did looking luci makes but I don't understand very well how they do :o) Jun 10 19:41:28 I belive it doesn't work for luci either right now, it just calls luac -s blah.lua Jun 10 19:42:02 if PATH is not overrideen, the OS will use standard lua then and you end up with incompatible bytecode Jun 10 19:42:19 just inside makefile like when i do a cp Jun 10 21:06:33 xMff: pong Jun 10 21:52:09 acinonyx * r16405 /trunk/package/base-files/files/lib/upgrade/common.sh: [package] base-files: format hexdump output instead of using awk for sysupgrade Jun 10 21:55:45 florian * r16406 /trunk/target/linux/rdc/ (files/ files/drivers/mtd/maps/rdc3210.c files-2.6.28/): [rdc] move files-2.6.28 to files Jun 10 22:22:50 acinonyx * r16407 /trunk/package/base-files/files/ (lib/upgrade/common.sh sbin/sysupgrade): [package] base-files: Detect and decompressed gzipped images automatically when flashing with sysupgrade Jun 10 22:34:16 acinonyx * r16408 /trunk/package/base-files/files/lib/upgrade/common.sh: [package] base-files: Add support for bzipped images in sysupgrade Jun 10 23:55:21 jow * r16409 /trunk/feeds.conf.default: [feeds] switch to LuCI 0.9.x branch Jun 11 01:02:13 Bartman007: wrt the $(TOPDIR)/files business, it would be handy if there was a way to remove existing files from the image, maybe something like the META files in the mini_fo overlay thing (e.g. what you see in /jffs). Jun 11 01:07:51 I see a skipped image/preconfig step every time the image is baked, maybe one could hook a shell script in there Jun 11 01:14:12 * russell_ 's initial "real" deployment of the nocatauth'd wgt failed btw. not sure why (no messages in evidence) but the gateway process died, was restarted, and died again. back to the drawing board. Jun 11 01:15:58 ah btw, is there some generic way to recover from broken nvram on the wgt634u? I've seen a forum thread were some users attempted to do it w/o much success so far Jun 11 01:16:11 it worked for me Jun 11 01:16:20 just get to CFE with a serial console Jun 11 01:16:25 k Jun 11 01:16:36 I do not own such a device so I can't really comment Jun 11 01:16:36 and setenv -p the stuff in there, quoting anything with spaces Jun 11 01:17:00 STARTUP and kernel_args or something were the two needing quoting Jun 11 01:17:13 then reboot Jun 11 01:18:25 what i did is fashion a new mtdblock4 from a different device and hexl-mode fixup the macaddrs and then from the running system, mtd write nvram Jun 11 01:19:23 but it was booting well enough even without that mtd write once i did the setenv -p stuff Jun 11 01:19:37 ok Jun 11 01:20:17 i think the people who were reporting that not working for them either weren't doing the quoting or had other problems Jun 11 01:21:19 ah, looks like they solved it in the meanwhile Jun 11 02:09:15 jow * r16410 /trunk/package/fuse/Makefile: [package] fuse: update to v2.7.4 (#5186) Jun 11 02:09:55 jow * r16411 /trunk/package/fuse24/ (8 files in 2 dirs): [package] fuse24: introduce fuse package for brcm-2.4 (#5186) Jun 11 02:10:43 jow * r16412 /trunk/target/linux/brcm47xx/patches-2.6.28/ (2 files): [brcm47xx] add patches to workaround dcache realted fuse problems (#5186) Jun 11 02:12:55 jow * r16413 /packages/utils/ntfs-3g/Makefile: [packages] update ntfs-3g to version 2009.4.4 (#5186) Jun 11 02:17:15 jow * r16414 /packages/net/sshfs/Makefile: [packages] sshfs: update to v2.2 (#5319) **** ENDING LOGGING AT Thu Jun 11 02:59:57 2009