**** BEGIN LOGGING AT Fri Feb 12 02:59:56 2010 Feb 12 03:45:28 build #9 of brcm63xx is complete: Failure [failed compile_6 compile_15] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/9 Feb 12 03:46:19 build #6 of orion is complete: Failure [failed compile_6 compile_15] Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/6 Feb 12 03:53:05 build #6 of au1000 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/au1000/builds/6 Feb 12 03:57:16 build #6 of rdc is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/rdc/builds/6 Feb 12 09:39:53 kaloz * r19598 /trunk/target/linux/generic-2.6/patches-2.6.32/850-if_no_generic_time.patch: remove 850-if_no_generic_time.patch -- breaks compilation and 2.6.32.8 already includes a fix for this.. Feb 12 14:24:12 kerneis * r19599 /packages/net/polipo/Makefile: (log message trimmed) Feb 12 14:24:12 [packages] polipo: update to 1.0.4.1 Feb 12 14:24:12 31 January 2010: Polipo 1.0.4.1: Feb 12 14:24:12 Cherry-picked fixes from 1.0.5 Feb 12 14:24:12 * Fixed an integer overflow that may lead to a crash Feb 12 14:24:12 (http://secunia.com/advisories/37607/). Discovered by Jeremy Brown. Feb 12 14:24:12 (CVE-2009-4413) Feb 12 14:25:33 xMff: taking a look at your patch... I'm not certain where stop_interface_dhcp is called from? If its only called from ifdown, then I like it... Feb 12 14:31:02 netprince: correct, /sbin/ifdown scans for a procedure stop_interface_$proto and calls it if it exists Feb 12 14:31:37 netprince: see https://dev.openwrt.org/changeset/19597 Feb 12 14:32:13 this revert only kicks in on ifdown foo or /etc/init.d/network restart but not on dhcp renews or releases Feb 12 14:33:15 I did some testing in a vm yesterday and couldn't reproduce the bug submitters issue in the first place but I assume it is related to swconfig somehow Feb 12 14:33:27 ... which is not present on x86/vbox Feb 12 14:51:22 xMff: looks good to me :) Feb 12 14:55:48 rhk * r19600 /packages/libs/pcre/Makefile: [packages] pcre: update to 8.01 Feb 12 15:00:35 rhk * r19601 /packages/libs/libinklevel/Makefile: [packages] libinklevel: fix parallel build Feb 12 15:12:15 rhk * r19602 /packages/libs/clearsilver/Makefile: [packages] clearsilver: fix parallel build, don't install static libraries Feb 12 15:13:22 thepeople * r19603 /packages/libs/libol/Makefile: fix parallel build, from #6577 Feb 12 15:16:11 thepeople * r19604 /packages/utils/ntfs-3g/Makefile: fix parallel build, from #6577 Feb 12 15:24:19 thepeople * r19605 /packages/admin/syslog-ng/Makefile: [syslog-ng] fix parallel build thanks Raphael HUCK Feb 12 15:25:51 thepeople * r19606 /packages/net/tcpdump/Makefile: [tcpdump] fix parallel build thanks Raphael HUCK Feb 12 15:38:59 thepeople * r19607 /packages/utils/powertop/ (5 files in 2 dirs): add powertop, thanks Raphael HUCK Feb 12 16:26:12 rtz: ping Feb 12 16:29:30 codertux: hello Feb 12 16:30:14 rtz: hi, sorry for my absence... I had exams until today... what did you want to talk to me about the other day? Feb 12 16:31:35 codertux: I had also much stuff to do, so no problem Feb 12 16:31:50 codertux: I could try to compile the dd-wrt kernel, if you are still interested Feb 12 16:32:03 rtz: be my guest... Feb 12 16:32:19 rtz: unfortunately, I'm still at uni waiting for the results for my last exam... Feb 12 16:32:56 and I don't know what's my home IP right now, had some networking problems last night... Feb 12 16:33:04 codertux: sorry, but I don't have a mips toolchain and my hdd is kinda full Feb 12 16:33:17 codertux: so it will have to wait for a few hours Feb 12 16:33:19 but I'll be back home in ~2-3 hours hopefully... Feb 12 16:33:24 no problem Feb 12 16:33:44 ok... so I'll ping you again when I get home, ok? Feb 12 16:35:13 ok Feb 12 16:35:25 codertux: but I may be offline for some time Feb 12 16:46:08 xMff: ping Feb 12 16:47:03 pong rtz Feb 12 16:47:32 jow_laptop: do you know by any chance which package is missing for luci in 8.09.2? Feb 12 16:47:39 ? Feb 12 16:47:59 jow_laptop: well, luci doesn't seem to work Feb 12 16:48:09 details Feb 12 16:48:09 jow_laptop: it produces a 404 error Feb 12 16:48:26 jow_laptop: or does this only happen on first login? Feb 12 16:48:27 in the official release images? Feb 12 16:48:32 jow_laptop: yes Feb 12 16:48:43 there were several people complaining about it Feb 12 16:48:57 jow_laptop with the wrt54g image Feb 12 16:49:12 well, maybe its just a cached index.html leading to nowhere Feb 12 16:49:22 hmm Feb 12 16:49:50 or a theme is missing Feb 12 16:50:07 last time I checked the 8.09.2 was complete Feb 12 16:50:30 there is somebody with the problem in #openwrt, let me ask him to refresh his caches Feb 12 16:51:02 jow_laptop can you add a new luci collection with luci-admin-full and luci-theme-openwrtlight ? Feb 12 16:51:34 rtz: ask him what the exact url is when he gets the 404 Feb 12 16:52:14 jow_laptop: will do once he responds Feb 12 16:53:46 crap, he's gone Feb 12 16:54:28 and he's back Feb 12 16:57:48 xMff: yes there still comes the text "Not Found". i get redirected to /luci and get this error. Feb 12 16:57:48 rtz: luci-static is present. Feb 12 18:57:03 jow * r19608 /trunk/package/base-files/files/etc/hotplug2-common.rules: [package] base-files: add /dev/hvcX to hotplug common rules Feb 12 18:57:49 jow * r19609 /trunk/target/linux/generic-2.6/config-2.6.31: [generic-2.6] add some missing symbols Feb 12 19:01:11 jow * r19610 /trunk/target/linux/x86/ (7 files in 4 dirs): [x86] add Xen DomU subtarget Feb 12 19:01:47 jow * r19611 /trunk/package/kernel/modules/xen.mk: [package] kernel: add xen kmods Feb 12 19:04:14 jow * r19612 /trunk/target/linux/x86/image/ (Config.in Makefile menu.lst): [x86] make console port editable through menuconfig, gzip ext2 images, add suitable defaults for Xen Feb 12 19:30:42 build #6 of s3c24xx is complete: Failure [failed shell_6 compile_11] Build details are at http://tksite.gotdns.org:8010/builders/s3c24xx/builds/6 Feb 12 19:33:12 lars * r19613 /feeds/efl/python-evas/Makefile: Feb 12 19:33:12 [packages] python-evas: Don't fail if python-package.mk is not installed yet. Feb 12 19:33:12 Fixes #6676 Feb 12 20:30:11 lars * r19614 /packages/libs/gettext/Makefile: [packages] gettext: Install m4 files. Fixes #6591 Feb 12 20:37:30 Looks like 19395 is not compatible with brcm 2.4 Feb 12 20:39:03 netprince: compile error? Feb 12 20:39:20 yeah, I can get it for you... Feb 12 20:39:41 would be nice Feb 12 20:41:11 http://openwrt.pastebin.com/m7f1d6964 Feb 12 20:42:53 my bad :( Feb 12 20:44:27 if think some ifdefing will fix it Feb 12 20:44:52 that patch I attached should probably only be applied if kernel version > 2.6.something Feb 12 20:45:06 >=2.6.32 iirc Feb 12 20:50:00 netprince: can you try this? http://openwrt.pastebin.com/m536725e7 Feb 12 20:51:44 yeah, will try Feb 12 21:00:50 hmm, it seems that upstream also just replaced it rather than using ifdef, let me check if anyone reported issues about that upstream Feb 12 21:01:34 maybe they just dropped 2.4 support ;) Feb 12 21:03:23 hehe Feb 12 21:04:03 xMff: it seems I am having a brain fart on building just one package?? make feeds/packages/utils/strace-compile V=99 (is that correct?) Feb 12 21:04:22 make package/strace/{clean,compile} V=99 Feb 12 21:05:24 thanks Feb 12 21:13:45 hmm, it didn't seem to help Feb 12 21:14:39 I checked the net.c has the version check now, so patch was applied Feb 12 21:19:12 hm Feb 12 21:20:05 make target/linux/install ? (if you want it in the image?) Feb 12 21:20:37 hmm, net.c shouldn't be edited, but configure.ac and configure (afaik) Feb 12 21:21:28 stintel: yeah, trying another approach - hold on Feb 12 21:21:28 Just tried building without 000-linux_socket.patch, seems to build ok Feb 12 21:22:02 yeah but that'll break it again on 2.6.32, so the #if approach is needed ;-) Feb 12 21:22:31 yeah, I was just making sure we were looking in the right place Feb 12 21:23:34 it has issues with struct sockaddr_nl which seems to be provided by linux/netlink.h Feb 12 21:25:23 http://openwrt.pastebin.com/m7d4b0450 Feb 12 21:31:59 that one gives a slightly different error message, http://openwrt.pastebin.com/m27f23407 Feb 12 21:32:59 can you post /home/psycadmin/trunk/build_dir/target-mipsel_uClibc-0.9.30.1/strace-4.5.19/config.h ? Feb 12 21:35:50 sure, its 479 lines, is there an easy way to pastebin that, or just copy/paste? Feb 12 21:36:28 grep ^# /... Feb 12 21:39:25 http://openwrt.pastebin.com/m744c371b Feb 12 21:41:17 rtz: ping Feb 12 21:42:02 netprince: ah as I suspected Feb 12 21:42:14 it fails at detecting netlink.h Feb 12 21:44:41 http://openwrt.pastebin.com/m39276778 Feb 12 21:44:51 and revert the earlier patch-patch Feb 12 21:48:06 that fixed it, nice Feb 12 21:48:38 I've got to go now, ttyl Feb 12 21:48:56 thanks for the help Feb 12 21:49:03 yw - thanks for testing Feb 12 21:49:23 jow * r19615 /packages/utils/strace/Makefile: [packages] strace: fix compilation against Linux 2.4 Feb 12 21:58:34 cshore: I'd like to create a new preinit addon to mount xenfs on boot (for the xen subtarget), where should I place that? Mounting is only possible after module loading Feb 12 22:00:34 xMff: I would add module loading code in 50_xxx ... just as sec I'll show you what I'm doing with USB Feb 12 22:06:44 xMff: http://openwrt.pastebin.ca/1794806 Feb 12 22:06:55 xMff: it's two files Feb 12 22:07:20 xMff: you don't need all that, but it gives you and idea Feb 12 22:09:06 thanks Feb 12 22:11:30 is there some reason that dd shouldn't write to /dev/hda on an alix? Feb 12 22:12:13 Alix board has /dev/sda Feb 12 22:12:29 not according to /sys/block Feb 12 22:12:47 cshore: and to which hook array? preinit_mount_root as well? Feb 12 22:13:06 yes to preinit_mount_root Feb 12 22:35:09 cshore: http://openwrt.pastebin.com/mddd6d29 ? Feb 12 22:36:15 xMff: I think that should work Feb 12 22:37:20 jow * r19616 /trunk/target/linux/x86/xen_domu/base-files/lib/ (. preinit/ preinit/45_mount_xenfs): [x86] add preinit xenfs mount handler Feb 12 23:12:11 cshore: ping Feb 12 23:15:45 pong thepeople Feb 12 23:15:50 ping xMff Feb 12 23:16:18 cshore: your trac access is active, you should be able to close tickets Feb 12 23:16:31 thepeople: ok, thanks Feb 12 23:23:37 cshore: pong Feb 12 23:24:05 xMff: that cat for the insmod will break if there is more than one modules loaded by ??-xenfs Feb 12 23:24:20 e.g. mbcache and xenfs Feb 12 23:24:21 will not happen Feb 12 23:24:30 or maybes yes... Feb 12 23:24:33 hm Feb 12 23:24:42 there should be an insmod helper Feb 12 23:24:45 that would the sed line is for Feb 12 23:24:52 would=what Feb 12 23:25:22 it sticks insmod in front of each module line and then executes each line Feb 12 23:25:45 sed 's/^\([^#]\)/insmod \1/' ??\-$module | ash 2>&- || : Feb 12 23:25:51 do_insmod() { Feb 12 23:25:51 while read modline; do insmod $modline; done < /etc/modules.d/??-$1 Feb 12 23:25:51 } Feb 12 23:27:41 do_insmod() { [ -f /etc/modules.d/??-$1 ] && while read modline; do insmod $modline; done < /etc/modules.d/??-$1; } Feb 12 23:28:01 ah ^# ... hmm Feb 12 23:39:04 * russell_ figures it out, needed to wait longer for the dd to sync, all better now Feb 12 23:41:20 how about do_insmod() { [ -f /etc/modules.d/??-$1] && sed 's/^\([^#]\)/insmod \1/' ??\-$modules | ash 2>&- || : ; } Feb 12 23:41:35 whoops Feb 12 23:41:50 do_insmod() { [ -f /etc/modules.d/??-$1] && sed 's/^\([^#]\)/insmod \1/' ??\-$1 | ash 2>&- || : ; } Feb 12 23:43:30 hmmm... maybe need more braces, or if then instead of [] && Feb 12 23:43:52 I wanted to avoid sed Feb 12 23:43:58 ah Feb 12 23:44:01 why? Feb 12 23:44:08 extra fork, whatever Feb 12 23:48:20 cshore: what about this: http://openwrt.pastebin.com/md8a4c46 Feb 12 23:51:50 but I do not know whether this construct (piping to sh) would work on the hush shell Feb 12 23:52:54 does load_modules in /etc/functions.sh work in hush? That's where I cribbed the sed line from Feb 12 23:53:53 ah Feb 12 23:57:02 which variation to pick specific module(s) Feb 12 23:58:01 just noticed, the line you posted is almost exactly the same Feb 12 23:58:26 but I'd replace "ash" with "sh" Feb 12 23:58:43 ok Feb 12 23:58:47 because there are at least two different shells in use on openwrt Feb 12 23:59:19 probably you should update functions.sh then Feb 12 23:59:30 yes Feb 13 00:00:02 is the advantage of hush that it's smaller? Feb 13 00:00:18 it works on non-mmu systems Feb 13 00:00:42 which is required for the ubicom target at least Feb 13 00:37:14 jow * r19617 /trunk/target/linux/ (9 files in 5 dirs): [x86] prepare kernel 2.6.32 support Feb 13 00:45:36 lars * r19618 /trunk/target/linux/generic-2.6/config-2.6.32: [kernel] Add some missing debug symbols to the generic 2.6.32 config. Feb 13 00:53:23 thepeople * r19619 /trunk/target/linux/brcm47xx/config-2.6.32: sync kernel config, closes #6675 Feb 13 01:05:32 thepeople * r19620 /trunk/target/linux/brcm47xx/config-2.6.32: fix mistake, thanks xMff Feb 13 01:07:46 thepeople * r19621 /trunk/scripts/download.pl: add geodns addresses for the kernel, closes #6574 Feb 13 01:11:26 thepeople * r19622 /packages/net/ethtool/Makefile: Feb 13 01:11:27 this patch fixes ethtool parallel build. Feb 13 01:11:27 -Raphael Feb 13 01:25:24 build #11 of ppc40x is complete: Failure [failed shell_6] Build details are at http://tksite.gotdns.org:8010/builders/ppc40x/builds/11 Feb 13 01:56:50 jow * r19623 /trunk/target/imagebuilder/files/Makefile: [imagebuilder] fix custom package selection **** ENDING LOGGING AT Sat Feb 13 02:59:58 2010