**** BEGIN LOGGING AT Sun Feb 28 02:59:57 2010 Feb 28 04:16:38 build #18 of mpc52xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/18 Feb 28 05:52:24 updated openwrt/upstream, http://pastehtml.com/view/5tle51a.html Feb 28 06:08:13 builder: commands Feb 28 06:08:16 buildbot commands: commands, dance, destroy, excited, force, hello, help, join, last, leave, list, notify, source, status, stop, version, watch Feb 28 06:08:30 builder: status brcm63xx Feb 28 06:08:30 brcm63xx: building('./upload_files.sh xwrt ...') [ETA 5h39m35s] Feb 28 08:22:48 build #18 of ixp4xx is complete: Failure [failed compile_6 shell_19] Build details are at http://tksite.gotdns.org:8010/builders/ixp4xx/builds/18 Feb 28 09:41:02 juhosg * r19888 /trunk/target/linux/adm5120/Makefile: adm5120: switch to 2.6.32 Feb 28 09:41:05 juhosg * r19889 /trunk/target/linux/adm5120/ (16 files in 3 dirs): adm5120: remove 2.6.31 related stuff Feb 28 10:16:49 hauke * r19890 /trunk/ (19 files in 19 dirs): kernel: update to kernel 2.6.32.9 Feb 28 10:16:55 is juhosg appearing here? Feb 28 10:21:06 hauke * r19891 /trunk/target/linux/brcm47xx/ (3 files in 3 dirs): [brcm47xx] remove unneeded patch Feb 28 10:50:07 <_trine> cshore, ping Feb 28 10:50:17 Trine: pong Feb 28 10:51:09 <_trine> morning,, I have a problem with wifi when I pivot the root and I was wondering if it is also there when using your method Feb 28 10:51:25 no idea...what's the problem? Feb 28 10:51:31 <_trine> well Feb 28 10:52:02 <_trine> I have basically 2 different roots ,, one on the routers flash and one on the pendrive Feb 28 10:52:12 right Feb 28 10:52:21 <_trine> and from that I have 2 lots of configs Feb 28 10:52:40 yep Feb 28 10:53:11 <_trine> I don't want wifi to be on when I first boot my router so I make sure wifi is set to disabled on the first root Feb 28 10:53:30 ok Feb 28 10:54:07 <_trine> but i have noticed that when I have configured in this way and then subsequently pivot my root Feb 28 10:54:27 <_trine> I cannot then bring up the 5GHz radio Feb 28 10:54:36 <_trine> using wifi up Feb 28 10:54:48 <_trine> I will show you what I get Feb 28 10:55:00 is the pivot_root in preinit, or is it later than that? Feb 28 10:55:51 mine is in preinit, so the only configs that matter are the ones of the active root (i.e. if you have the key drive and boot, then the only config that matters is the key drive config) Feb 28 10:56:40 Trine, however, it could also be a bug with the wifi code Feb 28 10:56:45 <_trine> http://pastebin.com/nYgLt8p8 Feb 28 10:57:00 <_trine> cshore, I am not a linux expert Feb 28 10:57:14 <_trine> just a beginner really Feb 28 10:57:44 <_trine> it all works perfectly if I don't disable wifi in the first root Feb 28 10:58:06 ah, then I would suspect you will be fine with my rootfs on key drive Feb 28 10:58:22 <_trine> I think I need to use yours Feb 28 10:58:45 <_trine> but I will wait for you to finish it first Feb 28 10:58:50 oki Feb 28 10:59:18 <_trine> I just thought I would mention this in case it was also a problem on your version too Feb 28 11:09:30 florian * r19892 /trunk/target/linux/generic-2.6/config-2.6.32: [kernel] add missing libertas symbols Feb 28 11:09:36 florian * r19893 /trunk/target/linux/malta/ (. Makefile README config-default image/ image/Makefile): [malta] add a MIPS Malta target to be used with qemu Feb 28 11:09:41 florian * r19894 /trunk/target/linux/brcm63xx/image/Makefile: [brcm63xx] build images for RG100A (bcm6358-based boards) (#6755) Feb 28 11:09:48 florian * r19895 /trunk/target/linux/brcm47xx/patches-2.6.32/180-ssb_extif_interrupt.patch: Feb 28 11:09:48 [brcm47xx] explicitely enable shared interrupt 2 for any core that did not a dedicated IRQ Feb 28 11:09:48 Explicitely enable shared interrupt 2 for any core that didn't get a dedicated IRQ Feb 28 11:09:48 anymore (fallthrough case) and for EXTIF cores to make gpio interrupts work. Feb 28 11:09:49 Also remove a bogus comment. Feb 28 11:09:49 Signed-off-by: Jochen Friedrich Feb 28 11:09:51 florian * r19896 /trunk/package/broadcom-diag/src/diag.c: [diag] add missing hunks to recognize the Microsoft MN-700 device (#6749) Feb 28 11:26:19 <_trine> cshore, it's interesting that this problem only affects the 5GHz radio and not the 2.4GHz one Feb 28 11:29:50 build #24 of brcm63xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/24 Feb 28 12:26:42 florian * r19897 /packages/net/umurmur/ (3 files in 2 dirs): [package] prevent umurmur server crashed on client join, remove useless patch, bump relase number (#6745) Feb 28 12:26:44 florian * r19898 /packages/net/privoxy/ (Makefile patches/102-remove_rej_files.patch): [package] update privoxy to 3.0.16, move to correct section, remove now useless patch (#6747) Feb 28 12:26:46 florian * r19899 /packages/net/tor/Makefile: [package] update tor to 0.2.1.24 (#6748) Feb 28 12:26:50 florian * r19900 /packages/net/irssi/Makefile: [package] switch irssi to using the build variants infrastructure (#6726) Feb 28 12:26:52 florian * r19901 /packages/mail/msmtp/Makefile: [package] switch mstmp to using build variants (#6723) Feb 28 12:26:55 florian * r19902 /packages/net/memcached/ (. Makefile files/ files/memcached files/memcached.init): [package] add memcached (#6746) Feb 28 12:57:59 florian * r19903 /trunk/package/switch/src/ (switch-adm.c switch-core.c switch-robo.c): [package] switch: do not include linux/autoconf.h (#6513) Feb 28 12:58:04 florian * r19904 /trunk/package/busybox/patches/460-truncated_ping_results.patch: [package] fix busybox's ping compilation when the fancy feature is turned off (#6489) Feb 28 14:05:12 xMff, cshore: I kinda made some progress with the problem from yesterday Feb 28 14:05:23 cool, found something out? Feb 28 14:06:13 xMff: not sure what the exact problem is, but the crc32 algorithm produces a different result on mips then on x86 Feb 28 14:06:29 ah, so it is some endian / rollover stuff after all Feb 28 14:07:26 xMff: I get 0x9a45e350 on mips and on x86 Feb 28 14:07:27 0x708e9077 Feb 28 14:07:57 x86 or x86_64? Feb 28 14:08:04 x86 Feb 28 14:08:21 hmmm...so it's the same byte-size Feb 28 14:10:37 rtz2: can you upload you test code and an example image somewhere? Feb 28 14:11:27 rtz2: and have you tried to run that code on some big endian machine? Feb 28 14:11:49 just to rule out endian issues Feb 28 14:14:11 and the bootloader produces the a 3rd result ... Feb 28 14:14:13 one moment Feb 28 14:18:56 ok, maybe I have messed up somewhere, but I have no clue where Feb 28 14:19:19 xMff: you can reproduce it with any file and the standard trx tool in openwrt Feb 28 14:19:43 well I need a file where the uboot checksum is known for Feb 28 14:20:01 xMff: it produces completly different results on my x86 and on my amazon (mips) board Feb 28 14:20:15 and where this checksum differs from what is calculated by trx.c Feb 28 14:20:25 xMff: not only swapped bytes Feb 28 14:20:31 understood Feb 28 14:22:37 xMff: let me upload the file, but it can take a few minutes, my upstream is really slow Feb 28 14:22:58 xMff: can you open .rar archives? Feb 28 14:23:12 of course Feb 28 14:37:02 crap, I think I messed up somewhere Feb 28 14:56:08 and my upload failed :/ Feb 28 14:56:12 argh Feb 28 15:00:01 I have to go for an hour or so Feb 28 15:20:03 matteo * r19905 /trunk/ (111 files in 3 dirs): kernel: add 2.6.33 patches Feb 28 15:30:09 matteo * r19906 /trunk/target/linux/atheros/ (13 files in 2 dirs): atheros: 2.6.33 patches Feb 28 15:43:34 hi Feb 28 15:43:43 ping [florian] Feb 28 16:31:06 matteo * r19907 /packages/net/ptpd/ (. Makefile files/ files/ptpd.init): add package PTPd Feb 28 16:33:16 matteo * r19908 /trunk/target/linux/ar71xx/ (5 files in 2 dirs): ar71xx: 2.6.33 patches Feb 28 16:34:20 nbd: ping Feb 28 16:40:08 jow * r19909 /trunk/ (2 files in 2 dirs): Feb 28 16:40:08 [package] switch: Feb 28 16:40:08 eliminate the asterisk from the switch port configuration and set it Feb 28 16:40:08 automatically on the first vlan, this aligns the broadcom vlan config Feb 28 16:40:08 with the swconfig format Feb 28 16:40:35 rtz2: pong Feb 28 16:44:41 jow * r19910 /trunk/target/linux/brcm-2.4/base-files/etc/init.d/netconfig: [brcm-2.4] remove asterisk from default vlan ports in netconfig Feb 28 16:50:56 nbd: did you have time to look into the wert160nl boot problem? Feb 28 16:51:03 no Feb 28 16:51:29 nbd: hardpat told me you talked with him about it Feb 28 16:51:34 yes Feb 28 16:52:46 nbd: I tried to find out, what's happening, but I have no f* clue, what the problem is :/ Feb 28 16:55:50 well, maybe I will get lucky Feb 28 16:58:18 jow * r19911 /packages/net/updatedd/Makefile: [packages] updatedd: mark uci as config file (#6766) Feb 28 17:06:55 nbd * r19912 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: mac80211: fix interface startup order for AP+STA (patch by yann lopez) Feb 28 17:19:36 jow * r19913 /packages/net/memcached/ (4 files in 2 dirs): [packages] memcached: remove submenu, fix copyright year, add memory option to default config Feb 28 17:32:20 tripolar * r19914 /packages/net/ctorrent/Makefile: Feb 28 17:32:20 switch ctorrent to using the build variants infrastructure Feb 28 17:32:20 * ctorrent is linked to openssl Feb 28 17:32:20 * ctorrent-nossl uses builtin SHA-1 Feb 28 17:54:31 anyone have this problem on ar71xx/trunk/r19912: http://openwrt.pastebin.com/V4JrRXxr ? Feb 28 18:43:44 stintel: does iptables work afterwards? Feb 28 18:45:54 xMff: I noticed that I also have it on my home router, and I don't seem to have any problems there Feb 28 18:46:08 xMff: probably some module that needs to be loaded before ip_tables.ko? Feb 28 18:47:18 well iiptables tries to load kmods from time to time, until recently it tried to call modprobe for this which isn't present in openwrt, so I patched it to use insmod - and since module load attempts now work, insmod will throw a lot of unresolved symbols into dmesg since it does not support dependency resolving Feb 28 18:47:21 oh well Feb 28 18:47:34 seems like we have to disable the modprobe shit entirely Feb 28 18:48:31 ic Feb 28 18:48:33 it is not more than white noise Feb 28 18:50:30 the actual reason why I changed that was that iptables segfaulted on avr32 when it tried to call the not existing modprobe Feb 28 18:50:50 bad error handling apprently Feb 28 18:59:29 hmmm, 19912 broke client mode again Feb 28 19:00:42 yeah, i see why Feb 28 19:00:44 fixing that Feb 28 19:00:50 and from just looking at the code it also breaks adhoc Feb 28 19:01:04 it assigns bssid to ssid (or is it the other way around) Feb 28 19:01:45 the patch author probably wanted to get ssid but that's already done a bit above Feb 28 19:02:07 nbd * r19915 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: mac80211: fix client mode Feb 28 19:02:25 should both be fixed now Feb 28 19:03:05 nbd: check https://dev.openwrt.org/browser/trunk/package/mac80211/files/lib/wifi/mac80211.sh?rev=19915#L359 for adhoc Feb 28 19:03:29 ? Feb 28 19:03:47 should be: config_get ssid "$vif" ssid Feb 28 19:03:47 no ? Feb 28 19:03:56 oh Feb 28 19:03:57 right Feb 28 19:04:03 missed that Feb 28 19:04:12 easy to miss ;-) Feb 28 19:05:35 nbd * r19916 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: mac80211: fix essid setting for ad-hoc (thx, stintel) Feb 28 19:07:16 got two identical routers to play with atm, rspro + sr71a Feb 28 19:07:31 currently doing simple ap<->client test Feb 28 19:07:37 well, not yet :-) Feb 28 19:12:48 jow * r19917 /trunk/package/iptables/ (3 files in 2 dirs): [package] iptables: disable kernel module autoloading Feb 28 19:16:24 I wonder if 802.11w will work between the 2, it doesn't work if I try to enable it in wpa_supplicant on my laptop Feb 28 19:17:16 noticed in the hostapd/wpa_supplicant changelogs that it was updated to to use IEEE Std 802.11w-2009 Feb 28 19:18:30 I'm going to send some patches to allow to enable it in /etc/config/wireless to openwrt-devel soon, should allow easier testing by others Feb 28 19:18:48 ok Feb 28 19:20:53 nbd: I have a question regarding adding ar8316 support to owrt. which would be preferred way? a) as changes to the ar8216 driver? b) a separate ar8316 driver? or c) same as a, but renamed to e.g. ar8x16 Feb 28 19:24:05 changes to ar8216 first, we can rename it later Feb 28 19:24:21 ok Feb 28 19:25:54 nbd: can confirm that client mode works again, fwiw Feb 28 19:26:10 xMff: the unresolved symbols are gone from the log as well Feb 28 19:51:59 stintel: great Feb 28 19:57:38 anyone with a recent brcm-2.4 install around? Feb 28 19:59:03 I can get one of mine back within an hour if needed :-) Feb 28 19:59:29 well somebody claimed that the iptables mark match is broken Feb 28 19:59:44 required by qos-scripts Feb 28 19:59:50 and I suspect human error Feb 28 19:59:58 I was planning to get a wrt54gl back anyway, might as well do it tonight Feb 28 20:01:37 but first, dinner! :_) Feb 28 20:19:30 oh, forgive me for not signing off patches I sent to openwrt-devel Feb 28 20:23:35 no problem Feb 28 20:33:32 Hi all Feb 28 20:33:45 xMff: I have 19769, it appears to be working fine Feb 28 20:33:55 what was the problem Feb 28 20:33:57 ? Feb 28 20:34:10 hold on Feb 28 20:34:42 Just to let you know I am going to be away from Mar 2 to 5, and will probably be busy tomorrow and the rest of today, though I will try to be online for the /overlay changes to block-extroot Feb 28 20:36:20 netprince_: https://dev.openwrt.org/ticket/6769 Feb 28 20:37:19 blogic: ping Feb 28 20:38:22 si Feb 28 20:39:38 blogic: I played around with adding amazon support to ifxmips Feb 28 20:40:12 blogic: and there is some ASC offset (not sure what exactly this is) Feb 28 20:40:16 https://dev.openwrt.org/browser/trunk/target/linux/amazon/files/include/asm-mips/amazon/amazon.h#L631 Feb 28 20:40:30 https://dev.openwrt.org/browser/trunk/target/linux/ifxmips/files/arch/mips/include/asm/mach-ifxmips/ifxmips.h#L40 Feb 28 20:41:01 blogic: would it be possible to use the amazon offset on the danube? Feb 28 20:41:21 blogic: it's used in prom.c Feb 28 20:41:51 xMff: I've not tried 8.09.2, but I'm not seeing that in trunk. Its possible 8.09.2 suffered from the multiport issue we recently resolved. But there appears to be more wrong than just that... Feb 28 20:42:19 rtz2: dont think so Feb 28 20:42:36 netprince_: yeah, he probably hadn't rebooted after installing the qos stuff Feb 28 20:42:59 it does look like the result of some missing insmods Feb 28 20:43:57 blogic: ok, I could either switch it on compile time, or I could autodetect it during bootup, if there is some difference of the cpus Feb 28 20:44:49 xMff: is it me or is he missing ipt_mark in his lsmod? Feb 28 20:45:00 it is missing Feb 28 20:45:05 blogic: i'm not sure, if there is something in the cpuinfo_mips Feb 28 20:45:24 you can detect which cpu you are on Feb 28 20:45:24 and he can not install the userspace part without pulling the kmod along, therfore I think it is just a missing reboot Feb 28 20:45:37 or insmod Feb 28 20:45:59 blogic: what should I use, processor_id? Feb 28 20:46:26 blogic: http://lxr.linux.no/linux+*/arch/mips/include/asm/cpu-info.h#L50 Feb 28 20:48:06 is there an easy way to split changes into a few patches rather than in just one patch? Feb 28 20:48:45 stintel: ususally, you can edit the patches manually without any problems Feb 28 20:49:12 stintel: or you can svn diff only certain directories Feb 28 20:49:26 ah but it's on the same file :-) Feb 28 20:49:43 stintel: well, then, edit the patch file Feb 28 20:49:51 ok Feb 28 20:50:01 stintel: as long as the parts are different chunks Feb 28 20:50:42 florian * r19918 /trunk/package/kernel/modules/usb.mk: [package] fix kmod-nls-base dependency with usb-core (#6753) Feb 28 20:50:52 florian * r19919 /trunk/target/linux/brcm63xx/ (24 files in 2 dirs): [brcm63xx] add support for 2.6.32, dropped the SPI patch for now Feb 28 20:50:59 florian * r19920 /trunk/target/linux/generic-2.6/ (19 files in 2 dirs): [kernel] misc 2.6.33 compilation and config symbols fixes Feb 28 20:52:32 [florian]: ping Feb 28 21:09:50 <[florian]> rtz2: pong Feb 28 21:15:05 [florian]: I don't want to be too annoying, but I have to ask again after the rdc patch, but I have to decide, which router to use as a development platform and which one for my network at home Feb 28 21:15:47 [florian]: and I can't change this too easily and there were still the problems with platform.c and I would also like to get sysupgrade working ... Feb 28 21:15:53 <[florian]> rtz2: allright, well I cannot test your rdc patch tonigh but I will commit it in a reworked version in a couple of days Feb 28 21:16:25 <[florian]> rtz2: the rdc platform does not provide good runtime detection, unless we can find a way to do so, I am not keen on doing nultime kernel for multiple boards Feb 28 21:16:48 <[florian]> rtz2: one possibily would be to use a second stage loader which modifies the x86 setup_data structure to pass board specific information Feb 28 21:17:52 <[florian]> rtz2: if your device can be identified by the rdc cpu id stuff that's even better Feb 28 21:20:03 [florian]: well, the current way of checking the flash for magic headers in certain places is not ideal, but I don't see, why this shouldn't work Feb 28 21:20:11 and it seems reasonable easy Feb 28 21:20:36 <[florian]> it is "easy" but it does not work reliably because rdc designs have flash chips ranging from 8MB to 2MB Feb 28 21:20:39 messing around with a second stage loader will be way harder Feb 28 21:21:01 <[florian]> please check if your model can be identified by the rdc cpu id Feb 28 21:21:29 [florian]: will do, but what's the problem with the current flash-size detection code? Feb 28 21:21:55 <[florian]> rtz2: one should check that it can ma 8MB, if not 4Mb, if not 2MB, and then check for the magic Feb 28 21:22:13 <[florian]> s/ma/map/ Feb 28 21:22:37 [florian]: that's how it is currently done Feb 28 21:22:55 [florian]: by cpu id, you mean the cpuid instruction? Feb 28 21:23:16 <[florian]> rtz2: by this: target/linux/rdc/patches-2.6.32/010-rdc_cpu_ident.patch Feb 28 21:23:53 <[florian]> rtz2: and by reading a magic on the flash it is fine Feb 28 21:23:56 <[florian]> I got to leave now Feb 28 21:24:36 florian * r19921 /trunk/target/linux/brcm63xx/ (23 files in 2 dirs): [brcm63xx] add support for 2.6.33 Feb 28 21:24:42 <[florian]> rtz2: you can mail me if you prefer, because I can answer emails in the day Feb 28 21:25:06 [florian]: ok Feb 28 21:26:43 [florian]: it prints R3210X/Edimax Feb 28 21:26:59 [florian]: not sure, if this is also used for other devices beside the sitecom Feb 28 21:27:22 blogic: you still here? Feb 28 21:42:47 si Feb 28 21:50:12 blogic: first a general question Feb 28 21:50:49 blogic: looking through the code, it seems to be pretty similiar, but most register offsets are different Feb 28 21:51:19 blogic: you still think it's a good idea to integrate amazon into ifxmips? Feb 28 21:52:14 dunno Feb 28 21:52:21 that is why i have not done so yet Feb 28 21:52:30 alternatively you clean up amazon Feb 28 21:53:40 hmmm Feb 28 21:54:47 blogic: https://dev.openwrt.org/browser/trunk/target/linux/ifxmips/files/arch/mips/pci/pci-ifxmips.c#L132 Feb 28 21:55:05 blogic: do you know offhand, what those register writes do exactly? Feb 28 22:07:48 blogic: do you know offhand, what those register writes do exactly? Feb 28 22:07:58 blogic: sorry, lost my internet connection Feb 28 22:21:50 no Feb 28 22:22:40 i think it is the mappings for the individual bars Feb 28 22:22:49 yes Feb 28 22:22:59 its the mappings for the different pci slots Feb 28 22:23:08 internally usb is connected using pci imho Feb 28 22:23:15 or sort of Feb 28 22:31:48 blogic: maybe the best idea would be to lump them together but making the choice at compile time Feb 28 22:32:14 blogic: and having different header files with offsets for both Feb 28 22:32:42 yes Feb 28 22:32:47 that is the best way imho Feb 28 22:32:58 #ifdef or a struct Feb 28 22:33:08 struct is cleaner but takes a bit longer to hack Feb 28 22:34:23 blogic: well, not sure how useful a struct would be for #define constants Feb 28 22:36:32 ah ok Feb 28 22:36:42 it is not the regions but th emapping within the regions Feb 28 22:36:47 yes #ifedef Feb 28 23:06:53 build #22 of brcm_2_4 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/brcm_2_4/builds/22 Feb 28 23:09:33 blogic: do you have datasheets for both the amazon and the danube? Feb 28 23:11:47 si Mar 01 00:38:05 build #18 of rdc is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/rdc/builds/18 Mar 01 00:59:15 ping xMff: just curious, why did you have to remove iptables kernel module autoloading? Mar 01 00:59:46 we have no modprobe so I patched iptables to use insmod Mar 01 01:00:19 but since insmod does not support dependency resolving it just resulted in dozens of unresolved symbol errors in dmesg each time it tried to request a kmod Mar 01 01:00:48 ahh, ok thanks Mar 01 01:01:06 while iptables still works fine it will scare most users and produce lots of invalid bug reports, so I disabled it entirely now Mar 01 01:01:55 understood Mar 01 01:07:58 hmm.. I'm getting "configure: error: compiler support for __thread is required" compiling glibc .. any clues? Mar 01 01:08:18 I build it successfully yesterday . not sure what changed :| Mar 01 01:08:52 hmm.. was using a different gcc v. though. Mar 01 01:09:07 yes, google indicates that this might be a gcc4 issue Mar 01 01:09:11 (was having problems with kernel not booting - think it might have been enabling locale support.. Mar 01 01:09:21 thanks xMff I'll go back to 4.2 then. Mar 01 01:09:22 if you have the option to use 3.x, try that Mar 01 01:09:31 or 4.2 Mar 01 01:09:49 I went back to 4.1 due to kernel crashing on boot.. :| Mar 01 01:10:21 at least the kernel doesn't crash now- I've tried it with the current kernel build, even though I have no _init :) Mar 01 01:10:30 .. at least it boots. Mar 01 01:10:47 ( I was getting a gazillion kernel strack traces scrolling past) Mar 01 01:12:59 nico * r19922 /trunk/package/switch/files/switch.sh: switch: suppress boot error message on 2.4 Mar 01 01:36:38 build #18 of pxcab is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/18 Mar 01 02:40:59 xMff: ok. trying with 3.x now. v. weird - it compiled b4 Mar 01 02:41:10 xMff: I have no idea what's different :| **** ENDING LOGGING AT Mon Mar 01 02:59:57 2010