**** BEGIN LOGGING AT Thu Apr 15 02:59:56 2010 Apr 15 03:20:30 nico * r20869 /trunk/package/broadcom-diag/src/diag.c: package/broadcom-diag: fix USB led polarity on Asus WL-HDD (closes: #7055) Apr 15 03:24:40 nico * r20870 /branches/backfire/package/broadcom-diag/src/diag.c: [backfire] backport r20869 (closes: #7055) Apr 15 03:27:57 nico * r20871 /packages/utils/alsa-utils/Makefile: [packages] alsa-utils: add arecord binary (closes: #7155) Apr 15 04:57:19 build #57 of avr32 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/avr32/builds/57 Apr 15 05:00:02 nico * r20872 /trunk/toolchain/kernel-headers/Makefile: toolchain/kernel-headers: restore support for patches, needed for ubicom32 Apr 15 05:59:12 nico * r20873 /trunk/toolchain/uClibc/ (15 files in 3 dirs): toolchain/uClibc: add preliminary support for 0.9.31 (needs testing) Apr 15 06:01:47 nico * r20874 /trunk/package/base-files/Makefile: package/base-files: use a single $(CP) instead of two to copy shared library files Apr 15 06:23:02 build #58 of ar71xx is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ar71xx/builds/58 Apr 15 06:25:43 build #62 of atheros is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/atheros/builds/62 Apr 15 06:28:41 build #63 of ppc44x is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/63 Apr 15 07:24:03 nico * r20875 /trunk/package/block-mount/Makefile: package/block-mount: mark /etc/config/fstab as conffiles (closes: #7130) Apr 15 07:27:38 nico * r20876 /branches/backfire/package/block-mount/Makefile: [backfire] backport r20875 Apr 15 08:09:50 nico * r20877 /packages/net/pptpd/Makefile: (log message trimmed) Apr 15 08:09:50 [packages] pptpd: various fixes Apr 15 08:09:50 * enable bcrelay and package its binary (closes: #7150) Apr 15 08:09:50 * mark conffiles Apr 15 08:09:50 * add md5sum Apr 15 08:09:50 * use PKG_INSTALL Apr 15 08:09:50 * use $(CP) instead of $(INSTALL_BIN) for files installed by upstream 'make install' Apr 15 08:22:41 nico * r20878 /trunk/package/base-files/Makefile: package/base-files: fix typo in r20874 (thanks to rhk for spotting it) Apr 15 08:58:46 gmorning Apr 15 11:00:31 cshore, ping Apr 15 11:15:40 mirko: ping Apr 15 11:16:45 KanjiMonster, pong Apr 15 11:17:21 https://dev.openwrt.org/changeset/20860 <- doesn't increasing the start of aiccu to 51 do exactly the opposite? (radvd has 50) Apr 15 11:17:49 KanjiMonster, yes, i swapped the daemons within the commit-message, however the commit is correct Apr 15 11:18:17 wasn't intented ;) Apr 15 11:19:24 ah, I see Apr 15 11:21:27 my aiccu/radvd installation is too old to check myself which one needs to be started first (it might have changed in current versions) Apr 15 11:23:14 mirko * r20879 /packages/Xorg/lib/qt4/Makefile: Apr 15 11:23:14 package support for optional qt4-plugins and Apr 15 11:23:14 - thanks a lot to Dave Smith Apr 15 11:23:35 KanjiMonster, the problem is kernel 2.4-related Apr 15 11:23:56 here routes are set in a different manner and they interfere Apr 15 11:23:58 ah, I see, that's why probably I never saw it Apr 15 11:25:17 the only problem I have with radvd is that I have to restart it manually after a network restart or else ipv6 forwarding is broken, but I am using an oooold trunk (2.6.25 - didn't have the time to update yet, and too much custom stuff) Apr 15 11:25:39 hehe, as usual Apr 15 11:25:55 cshore, http://pastebin.com/a0yN12cN - can you reproduce this? Apr 15 14:58:19 build #53 of brcm_2_4 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/brcm_2_4/builds/53 Apr 15 15:03:38 cshore, ping :) Apr 15 15:19:45 [florian], i got advised to ask you about gettign open wrt on the netgear dg834g V4 but i need to do it wirewireless if posible cos i burnt out all the ethernet ports on it by accident and havent had chance to resolder some new ones on or buy new ones for it Apr 15 15:19:48 is it posible Apr 15 15:20:51 tefx: what you really need is a serial connection Apr 15 15:27:16 ahh cant i tftp over wlan Apr 15 15:27:37 cos i think i have fouidn a way to forcfully start tftp in the acctual booted system Apr 15 15:28:36 tefx: well, if you have a telnet/ssh connection to it, it should work Apr 15 15:28:41 tefx: tftp won't work Apr 15 15:28:52 tefx: and i'm not sure, if this thing is even supported Apr 15 15:29:59 someoen said to talk to [florian] he supposedly has got it workign or sumamt Apr 15 15:30:34 tefx: do you have access to the vendor web interface? or telnet access? Apr 15 15:30:54 i have ssh telnet and wbeinterface Apr 15 15:31:02 i alrdy flashed it with the dgteam firmware Apr 15 15:31:10 to unlock full ssh accses Apr 15 15:31:59 tefx: well, it could work then, but you would need to build a custom image to enable wireless by default Apr 15 15:32:09 tefx: and stuff may very well be broken Apr 15 15:32:41 tefx: I really wouln't try it without wired network or better serial, even if it is supported Apr 15 15:34:56 nico * r20880 /packages/net/unbound/ (7 files in 2 dirs): [packages] add unbound (closes: #7022) Apr 15 15:38:46 <[florian]> tefx: which chip is this router based on? Apr 15 15:38:51 hum coudl eb difficult as i dotn have cash to repalace the hternets and solderign as for solderign a seirial prot on me and a solderign iron = danerous Apr 15 15:39:06 erm cant renmebr 2 secs ill find out Apr 15 15:41:53 [florian], i deffinatley know its broadcom based cpu and wlan Apr 15 15:42:57 <[florian]> tefx: they are quite some broadcom chips out there;) Apr 15 15:43:51 yeah i knwo im no were near the router to look in ti atm Apr 15 15:44:01 google isnt been my friend either Apr 15 15:44:29 BCM6348 V0.7 Apr 15 15:44:52 <[florian]> good, then you should be able to build an image for it, pray for it to have a recognized board id, and enable wireless Apr 15 15:44:57 <[florian]> like rtz2 said, having serial console is a mut Apr 15 15:45:01 <[florian]> s/mut/must/ Apr 15 15:48:30 [florian], can i get away withotu a serial console ??? Apr 15 15:48:45 as logn as i build it right first time Apr 15 15:50:19 tefx: I would guess the chance of everything working is < 10% Apr 15 15:50:43 as logn as i get atleast ssh wlan work stright outr Apr 15 15:50:47 if it dosent work iui can recover Apr 15 15:51:18 <[florian]> tefx: you will loose your ui Apr 15 15:51:25 heh dotn need ui Apr 15 15:52:21 i can replace that with the opensource uis for open-wrt after i get ti running Apr 15 15:54:10 tefx: wireless is disabled by default Apr 15 15:54:23 tefx: I doubt you will manage to enable it on first try Apr 15 15:54:30 yeah hense a recompiel with wirless enabled Apr 15 15:57:32 tefx: well, go ahead, give it a try Apr 15 15:57:50 tefx: but don't come here and complain, if it blows up in your face Apr 15 15:59:42 heh i know but as it woudl eb ncie if someoen can help me out throguht he process so i gott a better chance of makign ti fail safe Apr 15 16:01:33 tefx: best way is to modify the /sbin/wifi script to ouput and config file with wifi enabled Apr 15 16:02:09 tefx: but this won't help you, if you board id is not supported Apr 15 16:02:48 tefx: or if the flashmap doesn't work Apr 15 16:03:05 tefx: or if you run into problems with the image checksum Apr 15 16:36:32 mirko: ping Apr 15 16:36:40 cshore: ping Apr 15 16:36:44 cshore, pong Apr 15 16:36:51 xMff: pong Apr 15 16:37:10 mirko: yes, I noticed it last night Apr 15 16:37:12 cshore: did you had a chance to test the uhttpd changes? Apr 15 16:37:20 xMff: not yet, no Apr 15 16:37:27 kk Apr 15 16:37:36 xMff: I slept a long time last night Apr 15 16:38:13 cshore, would be nice to get it fixed or at least built again Apr 15 16:38:37 mirko: mod_iax is the problem...I'll look at it today sometime Apr 15 16:38:49 cshore, thanks Apr 15 16:40:15 mirko: do you happen to have the part of the log where mod_iax is attempted to be built (rather than what you have, which is the packaging)? Apr 15 16:40:33 cshore, yeah, wait Apr 15 16:40:55 need a minute Apr 15 16:41:28 mirko: sure, thanks Apr 15 16:46:26 acoul * r20881 /trunk/target/linux/generic-2.6/patches-2.6.34/006-squashfs_add_lzma.patch: 2.6.34: fix kzalloc error Apr 15 16:47:56 cshore, mind, i've made some local modifications - however - i don't think they interfere with this actual component Apr 15 17:12:29 mirko: no, that happened to me last night too Apr 15 17:49:16 build #56 of brcm63xx is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/56 Apr 15 18:10:15 just booted an uclibc 0.9.31 image on my rs pro, didn't notice anything not working yet (but only browsed around luci for now) Apr 15 18:10:21 cshore, http://nanl.de/foo/log_freeswitch Apr 15 18:11:21 it will break as soon as you turn your back on it ;) Apr 15 18:11:30 *g* Apr 15 18:11:52 hmm...looks like it never actually even tries to build mod_iax, which is why it can't package it Apr 15 18:33:02 <{Nico}> cshore: i will give it a try too Apr 15 18:33:42 {Nico}: ok, you might want to wait until I do the fix though....I'm pretty sure it's something I did to thepackaging Apr 15 18:56:53 hmmm....odd, the build worked in x86 and ar71xx Apr 15 18:58:46 <{Nico}> cshore: propably a configure test that failed somewhere Apr 15 18:59:28 well in mirko's build it looks like it doesn't even try to build Apr 15 18:59:45 there's not mention if iax except the packaging Apr 15 19:12:51 {Nico} if you still want to do a build to take a look the configure output, that could be useful Apr 15 19:16:19 build #54 of brcm47xx is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/54 Apr 15 19:25:44 <{Nico}> cshore: will do... after dinner :) Apr 15 19:27:15 aw man, dinner more important the FS? :-P Apr 15 19:28:13 juhosg * r20882 /trunk/toolchain/uClibc/patches-0.9.31/901-MIPS-restore-INLINE_SYSCALL-macro.patch: toolchain/uClubc-0.9.31: fix syscalls on MIPS Apr 15 19:46:39 jow * r20883 /trunk/package/uhttpd/ (8 files in 3 dirs): Apr 15 19:46:39 [package] uhttpd: Apr 15 19:46:39 - make network timeout configurable, increase default to 30 seconds (#7067) Apr 15 19:46:39 - follow symlinks in docroot and add option to disable that Apr 15 19:46:39 - fix mimetype detection for files with combined extensions (.tar.gz, ...) Apr 15 19:48:36 build #52 of sibyte is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/sibyte/builds/52 Apr 15 19:53:31 jow * r20884 /branches/backfire/package/uhttpd/ (8 files in 3 dirs): [backfire] merge r20883 Apr 15 20:08:58 jow * r20885 /trunk/package/uhttpd/src/uhttpd-utils.c: [package] uhttpd: fix bug in path canonization introduced by r20883 Apr 15 20:09:36 jow * r20886 /branches/backfire/package/uhttpd/src/uhttpd-utils.c: [backfire] merge r20885 Apr 15 20:28:24 http://pastebin.com/FkCmh8nb Apr 15 20:28:33 I'm getting this "dumps" on every boot Apr 15 20:28:51 could someone tell me what is that? Apr 15 20:37:23 nbd * r20887 /trunk/package/mac80211/patches/ (7 files): ath9k: fix crash issues caused by the ar9300 support patch Apr 15 20:50:55 acoul * r20888 /trunk/target/linux/brcm47xx/ (25 files in 2 dirs): brcm47xx: add preliminary 2.6.34 support Apr 15 20:52:45 nbd: thanks for the fix Apr 15 21:03:23 installing kmod-ipt-conntrack-extra produces that output Apr 15 21:03:24 http://pastebin.com/FRCfL9ta Apr 15 21:05:21 self-built image? Apr 15 21:11:08 acoul * r20889 /packages/net/ (11 files in 4 dirs): fix mini_snmpd on ipv4. Create a separate mini_snmpd_ipv6 package (closes #6578 & #7094) Apr 15 21:15:28 nbd: yes, selft-built image Apr 15 21:15:50 I didn't understood your explanation, but I'm building it with the package selected Apr 15 21:16:11 anyway that is not normal behaviour? Apr 15 21:16:41 well, it's known that kmod-* package selections affect core data structures or other kmod-* packages Apr 15 21:16:45 which could lead to crashes Apr 15 21:16:57 unfortunately we don't have a good approach for dealing with this yet Apr 15 21:17:23 we could run a crc32 over the selected options and add that to the kernel version :) Apr 15 21:17:30 as hex string Apr 15 21:17:53 too many false positives, imho Apr 15 21:17:54 isn't there already some module magic embedded in the kernel? Apr 15 21:18:03 maybe we can come up with a kernel patch that deals with the core changes Apr 15 21:18:18 by removing ifdefs Apr 15 21:20:53 acoul * r20890 /packages/libs/uclibc++/Makefile: fix TARGET_CFLAGS (closes #7172) Apr 15 21:21:25 I really don't understand why there is not many kernel hackers suicides Apr 15 21:21:35 they should feel depressed for never achieving perfection Apr 15 21:21:48 linux hackers are typically not about perfection Apr 15 21:21:56 but about getting stuff done without messing it up Apr 15 21:22:10 also this issue is typically not a problem for regular kernel useres Apr 15 21:22:31 it's a problem with configurability vs. binary distribution Apr 15 22:14:17 does trac have anything to deal with the idiot that continually posts comments such as the one on https://dev.openwrt.org/ticket/7173? Apr 15 22:17:06 lol Apr 15 22:17:23 NO FAX, NO PITCH! Apr 15 22:22:54 iSteve: ping Apr 15 22:23:06 well... kind of pong, but I'm about to hit the hay Apr 15 22:23:28 ok, quickly then... "recent" like how recent for a hotplug2 version bump? Apr 15 22:23:35 currently we have 201 in openwrt trunk. Apr 15 22:23:47 nbd: no dump when compiled. Then I do opkg remove *, opkg update, opkg install *, and I start getting the dump again on boot Apr 15 22:23:52 that sounds right Apr 15 22:24:05 sounds like what you know already? Apr 15 22:24:18 iSteve: (not for you) Apr 15 22:24:26 iSteve: if we were to bump, what version would you have us bump to? Apr 15 22:24:43 r201 is the best version to be Apr 15 22:24:56 then I'm confused by your email. Apr 15 22:24:57 perhaps you only have obsolete rules file version Apr 15 22:25:13 Well, your rules syntax suggests the rules file was written for _way_ older version of hotplug2 Apr 15 22:25:25 That's what led me to the conclusion that you are actually running an older version Apr 15 22:26:37 nunojpg: hm Apr 15 22:26:46 nunojpg: can you try installing the package from your locally built ones instead? Apr 15 22:27:01 It is not unlikely that your problems stem from using older rules syntax -- while I did my best to keep them as compatible as possible, it still was quite a leap (a completely new parser), so some things changed. For example, what struck me rightaway, was the semicolon you have in your "exec" Apr 15 22:27:33 The semicolon is obsolete, it should not be used, unless you actually want a semicolon in the argv of the application Apr 15 22:30:32 I put the semi-colon because that's what http://isteve.bofh.cz/~isteve/hotplug2/hotplug2.rules.doc.txt said. Apr 15 22:30:34 (which might actually be the source of your problems... if not, I'd rather continue tomorrow) Apr 15 22:30:43 awww crap Apr 15 22:31:17 nbd: no dump when my local built package is reinstalled Apr 15 22:31:29 philipp64: That's my fault... since I haven't made a proper hotplug2 1.0 release, I haven't updated the on-site docs. However, since everybody uses current svn anyway, I'll update them. Apr 15 22:31:40 philipp64: The docs in the sourcecode tree are authoritative Apr 15 22:31:47 ok. Apr 15 22:31:53 philipp64: My fault entirely, thanks for telling me Apr 15 22:32:02 nunojpg: ok, then it's probably just a config mismatch somewhere else Apr 15 22:32:18 sure. btw: what's the best way to express a rule that gets executed unconditionally? Apr 15 22:32:29 ACTION is set Apr 15 22:32:57 ok, was using DEVPATH. Apr 15 22:33:06 (without a set action, the uevent would be absolutely malformed) Apr 15 22:33:27 can't think of a case when "DEVPATH is set" wouldn't work equally well Apr 15 22:33:39 but "ACTION is set" should be a sure hit Apr 15 22:33:48 ok. Apr 15 22:33:52 nbd: thanks Apr 15 22:34:06 let me just check, to make sure Apr 15 22:34:40 yeah Apr 15 22:58:07 nico * r20891 /packages/utils/lcd4linux/ (Config.in Makefile): [packages] lcd4linux: rework drivers/plugins configuration and dependencies (closes: #5932) Apr 15 23:31:20 nbd * r20892 /trunk/target/linux/ar71xx/files/drivers/net/ag71xx/ (ag71xx_ar8216.c ag71xx_main.c): Apr 15 23:31:20 ar71xx: fix up alignment handling Apr 15 23:31:20 - on ar724x, rx buffers can be aligned with an offset of 2, which keeps the ip header aligned Apr 15 23:31:20 - alignment offset is only added if the ar8216 workaround is not active and the phy driver does not advertise its own packet alignment Apr 15 23:31:20 - ar71xx and ar91xx can not handle rx alignment offsets, however taking a hit on unaligned exceptions seems to have less overhead than re-aligning the data for large packets Apr 15 23:31:20 - use memmove to re-align small packets, if necessary Apr 15 23:31:21 tested on ar9132, ar7240 and ar7242 based devices without ar8216 headers Apr 16 00:08:54 nico * r20893 /trunk/package/kernel/modules/netsupport.mk: package/kernel: add missing modules autoloading for kmod-ipsec, kmod-ipsec4 & kmod-ipsec6 (partially closes: #7164) Apr 16 00:09:17 {Nico}: how did you build go? Apr 16 00:11:22 *your Apr 16 00:13:50 <{Nico}> cshore: not started yet Apr 16 00:20:25 ok, np Apr 16 00:27:50 I'm looking at extending the luci-fw and uci_flrewall packages to support more of iptable's features as well as implement more hardening and threat management techniques, and create more clear iptables configurations. I would love to contribute them to OpenWrt. Should I look at submitting patches for them or should I rather look at creating new packages for OpenWrt to support this separately? Apr 16 00:31:31 greezybacon: I think the best would be to implement as separate packages Apr 16 00:31:52 greezybacon: maybe even a separate firewall package which is a superset of the current one Apr 16 00:32:13 yeah, the problem is in the uci_firewall Apr 16 00:32:21 I can create like a luci-fw-adv package Apr 16 00:32:28 to probably stack Apr 16 00:32:39 but uci_firewall has to understand the rules I would create Apr 16 00:32:56 {Nico}: mod_iax no longer exists Apr 16 00:33:12 or you create a separate package which supports all of thre Apr 16 00:33:25 ... current rules plus additinal ones Apr 16 00:35:51 good idea Apr 16 00:36:40 this also has a higher chance to being adopted I think Apr 16 00:37:21 so you think like a module set to replace the luci-fw and uci-firewall packages? Apr 16 00:37:41 cshore * r20894 /packages/net/freeswitch/Makefile: packages/net/freeswitch: mod_iax no longers exists in 1.0.6 Apr 16 00:37:59 replace when installed, not in development Apr 16 00:38:14 greezybacon: yep Apr 16 00:38:56 sweet Apr 16 00:40:07 cshore * r20895 /packages/net/freeswitch/Makefile: packages: net/freeswitch: Makefile cleanup. Removed duplicate var from site rules and old snapshot text Apr 16 00:42:45 hmmm....if a init script backgrounds the process it starts (via &) will the stdout/stderr output still go the logger? Apr 16 00:43:50 I think yes Apr 16 00:45:07 cshore * r20896 /packages/net/freeswitch/files/freeswitch.init: packages: net/freeswitch: restore logging to syslog instead of file on /tmp/freeswitch at nico's request Apr 16 01:14:33 cshore: ping Apr 16 01:18:55 thepeople: pong Apr 16 01:19:39 cshore: add the maintainer line to freeswitch please Apr 16 01:19:46 doh, yep, sorry Apr 16 01:21:22 thepeople: a reminder: does that go in define Package/freeswitch or in at the head of the Makefile? Apr 16 01:21:38 in the define Apr 16 01:22:20 thepeople: any word on the devs thoughts about my becoming a dev? Apr 16 01:22:44 no objections yet Apr 16 01:22:51 cshore * r20897 /packages/net/freeswitch/Makefile: package: net/freeswitch: Add myself as maintainer Apr 16 01:28:44 <{Nico}> cshore: did you get some feedback from lalloso and/or trine regarding FS ? Apr 16 01:29:07 not yet, only that they were interested in it looking at it Apr 16 01:32:47 {Nico}: did you see that mod_iax is no longer available in the FS Apr 16 01:34:16 nico * r20898 /packages/net/siproxd/Makefile: [packages] siproxd: explicitly set libosip path for configure (closes: #5791), use PKG_INSTALL Apr 16 01:34:26 <{Nico}> cshore: yeah, saw your commits Apr 16 01:36:11 the funny thing is I had successful builds a couple of times with iax selected, which is weird Apr 16 01:36:21 because it's not in the source tree anymore Apr 16 02:31:54 nico * r20899 /packages/Xorg/lib/libast/Makefile: [packages] libast: add dependency on libpcre (#7137), add conditionnal dependencies on libX11 & libXt (needed for X11/Intrinsic.h) when drawing backend is X11, use PKG_INSTALL, cleanup Apr 16 02:33:07 <{Nico}> cshore: it's not downloaded from FS site ? Apr 16 02:55:32 nico * r20900 /trunk/toolchain/gcc/patches/3.4.6/701-pr16276-fix.patch: toolchain/gcc: add a 3.4.6 patch fixing a g++ bug where it generates local references to linkonce (see http://gcc.gnu.org/PR16276, closes: #7014) **** ENDING LOGGING AT Fri Apr 16 03:00:02 2010