**** BEGIN LOGGING AT Thu May 07 02:59:57 2009 May 07 03:04:02 [florian]: send you an email on dlink2590/cavium boot, memory map mismatch May 07 03:06:06 jow * r15652 /trunk/package/kernel/modules/netfilter.mk: [package] kernel: prepare kconfig for imq update May 07 03:06:57 jow * r15653 /trunk/package/iptables/patches/1.4.3.2/ (005-imq.patch 005-imq1.patch 009-table-alignment.patch): [package] iptables: update imq userspace part May 07 03:07:47 jow * r15654 /trunk/target/linux/generic-2.6/ (5 files): [kernel] prepare default configuration for kernel side imq update May 07 03:09:06 jow * r15655 /trunk/target/linux/generic-2.6/ (22 files in 5 dirs): [kernel] update imq patches for 2.6.26, 2.6.27, 2.6.28, 2.6.29 and 2.6.30 May 07 03:16:37 jow * r15656 /trunk/include/netfilter.mk: [include] adept netfilter.mk to updated imq May 07 04:53:48 xMff: /etc/function.sh come from openwrt or is part of x-wrt May 07 08:14:25 ping aoz May 07 08:14:28 ping glp May 07 08:35:14 florian * r15657 /trunk/package/kernel/modules/other.mk: [kernel] add missing backslash, thanks juhosg May 07 10:10:08 nbd * r15658 /trunk/package/madwifi/patches/ (420-diversity_fix.patch 421-channel_handling.patch): madwifi: improve antenna diversity handling May 07 10:34:04 florian * r15659 /branches/8.09/package/ (6 files in 4 dirs): merge r14882 to 8.09 (#5703) May 07 10:59:28 florian * r15660 /packages/net/lighttpd/ (Makefile patches/500-configure_cross.patch): [package] update lighttpd to 1.4.22 (#5072) May 07 11:01:51 [florian]: any updates on the driver? was busy the past few days May 07 11:02:14 hello to all May 07 11:03:06 I have been having problems with the madwifi driver and an AR5212, I am not using openwrt but building my own initramfs images May 07 11:03:18 and I have seen that you apply a set of patches to the madwifi driver May 07 11:03:40 <[florian]> mordy: not right now, but the good news is I will brought back with me my vortex board ;) May 07 11:03:53 oh :D May 07 11:04:03 btw does this thing have any sensor chip on it? May 07 11:04:05 if I understand correctly, you use SVN version 3314 as base for patching, but I am not able to build 3314+patches against kernel 2.6.25.17 (which is the version you use in kamikaze) May 07 11:04:34 I suppose that I am using the wrong version of some component, could you provide some insight, please? May 07 11:05:37 <[florian]> mordy: not that I remember of May 07 11:05:45 hmm.. :( May 07 11:05:50 i run irssi on it though :P May 07 11:06:53 i wonder if my wifi card is drawing too much power - it's the R52 from routerboard May 07 11:06:57 florian * r15661 /trunk/target/linux/au1000/ (13 files in 5 dirs): [au1000] update to 2.6.28 May 07 11:07:09 and i am using the ath5k driver for the AP May 07 11:07:45 mordy: does ath5k work in AP mode for you? I tried yesterday but was not able of making it work May 07 11:08:30 moebius_: yes but only wiht the latest wireless-testing git kernel May 07 11:08:40 and of course you need to modify the source and add a line enabling AP mode May 07 11:09:21 mordy: I have seen that small patch in some blog, I will try with the Git sources, as I also have R52 cards in my APs -- and madwifi is driving me... "mad" May 07 11:09:24 and also you'd want to comment out the "beacon queue (something) didn't stop?" message (it's somewhere in ath5k.c) because that will spam you May 07 11:09:44 well i fouond it easier to simply modify it by hand:P May 07 11:09:56 mmh, so is it safe to have that commented out? May 07 11:09:58 madwifi would make my kernel panic May 07 11:10:09 moebius_: well it's just an error message, it's not a function of the card May 07 11:10:19 hehe, okay May 07 11:10:25 but if your hardware is weak it may not like being spammed with that annoying messagwe May 07 11:10:40 madwifi is making all the stations in our company suffer periodic reconnects every ~10minutes May 07 11:10:48 lol May 07 11:10:58 ath5k isn't perfect either May 07 11:11:07 it takes quite a nice amount of time to associate using 11g and WPA May 07 11:11:25 in the earlier kernels (rc1) it was a matter of minutes May 07 11:11:39 now it's usually a bit under a minute. association is faster when either not using WPA or not using 11g May 07 11:11:44 it is okay for me if then there are o further reconnects aftwerwards May 07 11:11:53 yeah i've never been diisconnected May 07 11:11:59 unless of course the machine itself reboots :( May 07 11:12:13 whgich is why i need the wdt working and why i'm nagging [florian] about it : May 07 11:12:16 :P May 07 11:12:31 well, I have alix3 boards, which are pretty stable by themselves May 07 11:13:02 my problem is that I tested several combinations of madwifi/hostapd/kernel versions without too much luch regarding reconnnects May 07 11:13:04 yeah this one isn't as stable, but it's moroe powerful than the alix, which is why i got it May 07 11:13:09 mmaybe it's just a lemon, who knows May 07 11:13:15 :) May 07 11:13:55 i'm using the most 'l33test versions of everything wireless here :P May 07 11:14:15 hehe, that's nice (I also like testing), but even if I get ath5k working, I would like to know whether it is possible to have a decent madwifi build May 07 11:14:30 well madwifi is being depreciated anyway May 07 11:15:13 I know, but openwrt is more stable than the builds I made, that's why I keep trying to build the thing with the patches used in kamikaze May 07 11:18:56 lol it just rebooted again :( May 07 11:19:45 i need a proper testing environment. it may even be the PSU itself failing May 07 11:20:48 but i've placed it under the heaviest stress possible sometimes and it never rebooted May 07 11:20:55 but sometimes it will reboot randomly May 07 11:21:14 looks hard to debug May 07 11:21:32 <[florian]> ram problem causing an unhandled exception ? May 07 11:21:39 no, RAM is fine May 07 11:21:44 did a few passes with memtest86 May 07 11:21:55 did a few passes with memtest86 May 07 11:21:56 did a few passes with memtest86 May 07 11:21:57 did a few passes with memtest86 May 07 11:21:58 did a few passes with memtest86 May 07 11:21:58 did a few passes with memtest86 May 07 11:22:01 did a few passes with memtest86 May 07 11:22:04 did a few passes with memtest86 May 07 11:22:04 <[florian]> ok I got it :) May 07 11:22:04 if memtest86 says so, it should be ok then May 07 11:22:06 dammit May 07 11:22:09 sory about that May 07 11:22:11 wrong window May 07 11:22:25 i was testing whether it would make a noise if i entered the same command recursively May 07 11:24:46 no... false alarm May 07 11:25:07 there was a noise bieng made, but i guess it's from the serial port on my host machine rather than the ebox May 07 11:25:12 sorry about the spam May 07 11:25:13 lol May 07 11:25:35 and since i interface with it via serial... May 07 11:25:58 i've set txpower to 15dbm and i'll see how that goes May 07 11:26:14 i don't know how accurate iwconfig is withthat May 07 11:27:24 <[florian]> really depends on what the driver reports May 07 11:27:32 <[florian]> I guess it is ok in your case May 07 11:27:47 <[florian]> 15 dBm does not seem delirious as the default output power on an ath5k card May 07 11:27:59 hmm... i don't think it works May 07 11:28:13 i am monitoring my client to see whether the signal level will change when i change the txpower in iwconfig May 07 11:28:18 but it's not budging May 07 11:28:33 right now the AP is at "1dbm" but i still get a nice signal May 07 11:32:40 nbd * r15662 /trunk/package/madwifi/patches/123-ccmp_checks.patch: make madwifi buildable without CONFIG_CRYPTO May 07 11:32:45 nbd * r15663 /trunk/package/madwifi/patches/ (5 files): madwifi: improve mode selection, do not scan channels which do not fit the desired mode May 07 11:32:48 nbd * r15664 /trunk/package/madwifi/patches/422-confchange_reset.patch: madwifi: fix run-time changing of iwpriv settings (thx, brainslayer) May 07 11:33:24 mordy: have you tried hostapd+ath5k with a bridge? May 07 11:33:31 (I am building wireless-testing right now) May 07 11:34:29 moebius_: what do you mean bridge? i mean i have the wired and wirelss interfaced bridged May 07 11:34:38 otherwise i have no intertubez May 07 11:34:58 I plan to du exaclty that thing: br0 = wlan0+eth0 May 07 11:35:14 is the "bridge=br0" line in hostapd.conf ? May 07 11:35:29 <[florian]> nope it would be called br-lan May 07 11:35:43 <[florian]> oh unless you are not doing this with openwrt ? May 07 11:35:44 (it says in the comments that must be set for madwifi, but I dunno about nl80211-based setups) May 07 11:35:52 I am building my own images ;) May 07 11:36:11 moebius_: well specifying the bridge in hostapd.conf doesn't make a bridge for you AFAIK May 07 11:36:17 it just lets it work together with the bridge May 07 11:36:18 because I am serving only the kernel and a minimal initramfs over PXE May 07 11:36:28 tyou need to make the bridge before hostapd is run May 07 11:36:36 mordy: I know, I have also a script which uses brctl to bring up the bridge May 07 11:36:48 ifup? :P May 07 11:37:08 brctl addbr br0 ; brctl addif br0 wlan0 ; brctl addif br0 eth0 May 07 11:37:11 :P May 07 11:38:00 nope, i think txpower is broken for ath5k, at least for my card :( May 07 11:45:08 hmm.. it IS the wireless in this case May 07 11:45:20 i see that the message before reboot is always wireless-related May 07 11:47:21 http://pastebin.com/f23504b20 May 07 12:13:12 I made a quick script to auto-apply madwifi patches, haha May 07 12:13:46 for each patch; it applies the patch, builds the sources, and if that fails, reverts the patch before continuing with the next patch >:-) May 07 12:14:08 hello all May 07 12:14:08 now that's *evil* May 07 12:17:19 that's evil indeed May 07 12:17:26 but hey, that's what scripts are for May 07 12:17:57 florian * r15665 /trunk/target/linux/ar7/ (11 files in 3 dirs): [ar7] update to 2.6.27 (#4122) May 07 12:17:58 of course, I have never imagined that I would write so much shell-crap... but now I work as sysadmin May 07 12:18:01 xDD May 07 12:18:07 and *that's* evil, too May 07 12:18:11 XP May 07 12:20:28 change your nick to evilius_ then May 07 12:20:28 :P May 07 12:20:38 haha May 07 12:21:03 florian * r15666 /trunk/ (4 files in 4 dirs): [kernel] remove 2.6.26 since there are no remaining candidates for it May 07 12:31:33 florian * r15667 /trunk/package/acx/patches/004-add_request_info.patch: [package] fix acx compilation on 2.6.27 May 07 12:48:55 jow * r15668 /branches/8.09/package/pptp/files/pptp.sh: Fix typo in r15659 May 07 13:18:46 jow * r15669 /trunk/package/util-linux-ng/Makefile: May 07 13:18:46 This patch adds a new package "flock" that allows to manage file locks May 07 13:18:46 from shell scripts. It provides more functionality than the busybox May 07 13:18:46 "lock" command, i.e. locking existing file descriptors and non-blocking May 07 13:18:46 operation. May 07 13:20:49 nbd * r15670 /trunk/toolchain/uClibc/patches-0.9.30.1/150-c99_ldbl_math.patch: May 07 13:20:49 uclibc 0.9.30.1: add c99 long double math fix from gentoo (thx, ralph) May 07 13:20:49 fixes compile error on powerpc May 07 13:21:59 jow * r15671 /packages/lang/php5/Makefile: [PATCH] Support for SPL (Standard PHP Library) May 07 13:27:12 jow * r15672 /trunk/package/base-files/files/ (lib/upgrade/common.sh sbin/sysupgrade): [PATCH] Add option in sysupgrade to flash gzipped images May 07 13:33:09 jow * r15673 /packages/net/olsrd/files/olsrd.init: [PATCH] [olsrd] Start olsrd after starting quagga May 07 13:35:15 jow * r15674 /packages/net/quagga/files/quagga: May 07 13:35:15 [Quagga] Fix typo when executing watchquagga May 07 13:35:15 Signed-off-by: Vasilis Tsiligiannis May 07 13:38:30 jow * r15675 /packages/utils/pciutils/Makefile: May 07 13:38:30 [pcituils] Fix post-installation error May 07 13:38:30 This patch fixes an issue where installation of pciutils fails if pci IDs May 07 13:38:30 cannot be downloaded (e.g. when there isn't any internet access) May 07 13:38:30 Signed-off-by: Vasilis Tsiligiannis May 07 13:40:32 jow * r15676 /packages/net/quagga/patches/130-netlink_ignore_wireless.patch: May 07 13:40:32 [Quagga] Ignore wireless netlink messages May 07 13:40:32 This patch fixes a bug where zebra terminates randomly with SIGABRT. The May 07 13:40:32 problem originates from zebra trying to parse wireless netlink messages. The May 07 13:40:32 following patch makes zebra ignore these messages. May 07 13:40:33 Signed-off-by: Acinonyx May 07 13:41:26 jow * r15677 /packages/net/olsrd/Makefile: May 07 13:41:26 [olsrd] Add quagga plugin package May 07 13:41:26 Signed-off-by: Vasilis Tsiligiannis May 07 13:42:46 jow * r15678 /packages/net/quagga/ (4 files in 3 dirs): May 07 13:42:46 Add default configuration for Quagga daemons May 07 13:42:46 This patch adds default configs for quagga daemons. It adds password zebra May 07 13:42:46 and restricts vty access to everyone except localhost. If these confs do not May 07 13:42:46 exist the daemons do not start when running quagga init script. May 07 13:45:04 jow * r15679 /packages/utils/pciutils/patches/ (. 100-remove-no-timestamping.patch): May 07 13:45:04 Remove --no-timestamping wget option from update-pciids script May 07 13:45:04 This bug is introduced after upgrading pciutils package and prevents successful May 07 13:45:04 installation of the package. This patch fixes update-pciids script not to use May 07 13:45:04 --no-timestamping wget switch, which is not supported by busybox. May 07 13:45:07 Signed-off-by: Vasilis Tsiligiannis May 07 13:47:05 jow * r15680 /trunk/package/hostap-driver/Makefile: May 07 13:47:05 [hostapd] Fix installation of hostap pci and plx driver May 07 13:47:05 This fixes an issue where installation of hostap pci and plx driver fails May 07 13:47:05 because both were trying to overwrite hostap.sh file. May 07 13:47:05 Signed-off-by: Vasilis Tsiligiannis May 07 13:48:24 jow * r15681 /trunk/toolchain/uClibc/Makefile: May 07 13:48:24 Fix Image Builder building - added missing /bin dir to toolchain May 07 13:48:24 Signed-off-by: Vasilis Tsiligiannis May 07 13:50:11 jow * r15682 /trunk/target/linux/x86/image/Makefile: May 07 13:50:11 Fix building of iso filesystem May 07 13:50:11 This fixes two errors when building iso filesystem: May 07 13:50:11 a) dir root.grub/boot doesn't get created. May 07 13:50:11 b) the creation of .image file by gen_image.sh which apparently is not needed May 07 13:50:12 when building an iso. May 07 13:50:14 Signed-off-by: Vasilis Tsiligiannis May 07 13:51:50 jow * r15683 /trunk/package/base-files/files/sbin/sysupgrade: (log message trimmed) May 07 13:51:50 [sysupgrade] Don't try to pivot to new ramfs if already running on one May 07 13:51:50 This patch allows the sysupgrade script to continue with the upgrade if May 07 13:51:50 openwrt is already running on a ramfs. This allows the use of the sysupgrade May 07 13:51:50 script as an _installer_ when running from an iso image. A user could boot the May 07 13:51:53 system from a bootable cd and invoke sysupgrade -n to write an image to May 07 13:51:55 the hard disk or CF. May 07 13:53:28 sorry for the spam... May 07 13:54:22 ;) May 07 14:05:44 jow * r15684 /trunk/ (8 files in 2 dirs): May 07 14:05:44 Add support for pc speaker May 07 14:05:44 PC speaker could be very useful on platforms which implement it. One could May 07 14:05:44 easily use the speaker to produce a sound where its pitch or duration varies May 07 14:05:44 according to signal strength or SNR of a wireless link. May 07 14:05:45 Signed-off-by: Vasilis Tsiligiannis May 07 14:07:05 jow * r15685 /trunk/target/linux/x86/Makefile: May 07 14:07:05 Add pcmcia support on x86 target May 07 14:07:05 This is useful for some notebooks which still exist and have this hardware. May 07 14:07:05 Signed-off-by: Vasilis Tsiligiannis May 07 14:15:01 jow * r15686 /packages/net/redir/ (9 files in 2 dirs): May 07 14:15:01 New package redir May 07 14:15:01 This patch adds a new package "redir" that allows to redirect incoming May 07 14:15:01 TCP connections to a specific destination in a proxy-like way. May 07 14:26:36 jow * r15687 /packages/utils/ink/ (. Makefile): May 07 14:26:36 New package: ink 0.5.0 May 07 14:26:36 this adds a small command line utility for reading ink levels from May 07 14:26:36 printers via libinklevel. May 07 14:39:00 jow * r15688 /packages/net/djbdns/dnsmasq.patch: May 07 14:39:00 New package: djbdns May 07 14:39:00 This adds D.J. Bernstein's djbdns package to OpenWRT. Documentation is in the package's included README file. May 07 14:39:00 Signed-off-by: Johnny Halfmoon May 07 14:39:58 hmm, that one was borked May 07 14:54:26 jow * r15689 /packages/net/djbdns/dnsmasq.patch: [packages] djbdns: remove junk from last commit May 07 15:07:22 florian * r15690 /trunk/package/acx-mac80211/patches/ (3 files): [package] make acx-mac80211 compile against latest compat-wireless, still does not work though ;) May 07 15:07:49 florian * r15691 /trunk/target/linux/ar7/Makefile: [ar7] include kmod-acx in default images May 07 15:10:02 jow * r15692 /packages/libs/libinklevel/ (5 files in 2 dirs): [packages] Update libinklevel to 0.8.0rc2, fix usb device location May 07 15:25:17 florian * r15693 /trunk/target/linux/brcm63xx/Makefile: [brcm63xx] also supports pcmcia May 07 18:30:19 nbd * r15694 /trunk/package/madwifi/patches/421-channel_handling.patch: madwifi: fix setting outdoor mode May 07 19:13:46 juhosg * r15695 /trunk/package/kernel/modules/usb.mk: [package] kernel/modules: add ar71xx specific KCONFIG options for the OHCI/EHCI drivers May 07 19:19:16 kaloz * r15696 /trunk/target/linux/generic-2.6/ (5 files): add missing RTC_DEBUG symbol May 07 19:23:06 kaloz * r15697 /trunk/target/linux/orion/patches/030-mv88e6095_support.patch: backport mv88e6095 support for orion - this also fixes the non-working switch-port on some routers May 07 20:21:22 florian * r15698 /trunk/package/base-files/Makefile: [package] create the lib64 symbolic links for 64-bits supported architectures (#5075) May 07 20:29:35 florian * r15699 /packages/net/hiawatha/Makefile: [package] update hiawatha to 6.13 (#5074) May 07 20:45:14 florian * r15700 /packages/net/transmission/Makefile: [package] update transmission to 1.60 (#5065) May 07 20:57:15 florian * r15701 /packages/libs/polarssl/Makefile: [package] update polarssl to 0.11.0 (#5048) May 07 21:00:37 nbd * r15702 /trunk/package/madwifi/patches/ (15 files): madwifi: improve autochannel, nuke some dead code, fix scanlist building from channel mode May 07 21:04:49 florian * r15703 /packages/libs/cyassl/ (Makefile patches/): [package] update cyassl to 1.0.2 (#5049) May 07 21:10:34 nbd * r15704 /trunk/package/madwifi/patches/411-autochannel_multi.patch: madwifi: improve autochannel decision with lots of cards on one board May 07 21:32:04 hello guys May 07 21:34:18 I'm looking for mysql server in menuconfig, but I didn't find it, Is mysql server available for openwrt? May 07 21:35:03 fofware: scripts/feeds --help May 07 21:36:22 Have you looked in package svn rather than the feeds? (I don't know if it exists, but if it does and it's not the in the feed it could be in the package svn) May 07 21:38:59 thanks, but i don't use feeds to upgrade my build system only svn May 07 21:40:33 thanks, I will find it in directories May 07 21:41:45 Be aware package svn contains packages that aren't ready for use yet May 07 21:41:58 (As well as ones that are) May 07 21:42:23 ah!!! ok, thanks very much May 07 21:43:19 the dev.openwrt.org page for svn access tells you the svn for the packages as well as the one for openwrt base system, just in case you didn't know May 07 21:47:07 yes, thanks , I'm not use feeds, because I keep a one clean copy of openwr and build a directories to each platform that I want, but they have symbolics links to propers dirs, so I save download time and spaces in my system May 07 21:51:11 and when i did that feeds was not pressent in openwrt or i did have not notice about it May 07 21:51:14 :o) May 07 21:52:21 hmmm... not... mysql-server is not present only mysql-client May 07 22:04:51 does anyone know how to add led support to a device. my internet light does not seem to have a dev hook it /sys/class/leds May 07 22:05:27 the only led listed is "status" which controls the power light May 07 22:05:45 I would like to control the dsl and internet light. its ar7 device btw May 07 22:14:55 florian: someone mentioned that I should talk to you about getting leds to work May 07 22:21:41 florian's gone to bed May 07 23:50:21 nbd * r15705 /trunk/include/toplevel.mk: override some variables to make sure that perforce does not mess with our build process **** ENDING LOGGING AT Fri May 08 02:59:57 2009