**** BEGIN LOGGING AT Sat Feb 26 02:59:57 2011 Feb 26 05:48:11 there's a bug in the menuconfig with the luci-app-oslr Feb 26 05:48:33 some visualiser option magically appears but you can't truly enable it Feb 26 06:18:42 hmm Feb 26 06:18:46 damn Feb 26 06:19:01 maybe i should le luci out Feb 26 06:22:25 last time i checked, 4mb flash seemed enough Feb 26 06:22:28 damn >_< Feb 26 06:23:34 nbd * r25724 /trunk/target/linux/ixp4xx/config-2.6.37: ixp4xx: recreate the 2.6.37 config from the 2.6.32 one to clean up several faulty overrides Feb 26 06:23:37 nbd * r25725 /trunk/target/linux/ixp4xx/patches-2.6.37/020-gateworks_i2c_pld.patch: ixp4xx: fix the i2c pld driver for 2.6.37 - i2c uses a rt_mutex instead of a mutex now Feb 26 06:23:43 nbd * r25726 /trunk/target/linux/ixp4xx/ (config-2.6.36 patches-2.6.36/): ixp4xx: remove 2.6.36 support Feb 26 06:23:47 nbd * r25727 /trunk/target/linux/ixp4xx/patches-2.6.37/ (16 files): ixp4xx: clean up patches, fold the 2.6.37 update patch into the relevant board support patches Feb 26 07:20:06 build #16 of lantiq is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/lantiq/builds/16 Feb 26 07:34:16 build #81 of pxcab is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/81 Feb 26 08:33:24 hmm Feb 26 09:47:22 dammit i don't know how much free space should i leave for jffs2 xD Feb 26 09:47:53 nor how to keep the thing usable while having enough free space xD Feb 26 11:11:32 looks like i'll have to get rid of some client mode packages :\ Feb 26 11:54:53 made it Feb 26 11:54:58 should have hmm Feb 26 11:55:03 6 blocks free Feb 26 11:55:36 3 reserved, 2 for config, 1 for deletion Feb 26 11:55:50 (just 1 for config is too risky) Feb 26 12:22:54 gmorning Feb 26 13:00:23 what is Luci2 intended for? Feb 26 13:00:43 misterious and obscure options.. Feb 26 13:12:28 hey guys im trying to build an image for my Siemens Sl2-141 device, wich has a RTA1025WB board, i know this device is not supported officially but i thought i give it a try. Now to my problem, the router wants some special tags in the image file to know that this image is "valid". How can i create my own rule to build an image? I just need to change a few strings in the imagetags... Feb 26 13:14:38 danitool: right now its a parent project under which various system level libraries and interprocess communication techniques are developed Feb 26 14:40:27 Syburg: see target/ if I flash with mtd, should I use the factory or sysupdate image binary? Feb 26 14:47:55 thank you KanjiMonster Feb 26 14:48:58 Olipro: sysupgrade Feb 26 14:49:30 ok, so I definitely did that right Feb 26 14:49:36 it broke my router :( Feb 26 14:49:52 it was my own build mind you Feb 26 14:50:02 still, I'd expect the kernel to at least be able to get me to failsafe Feb 26 14:51:26 florian * r25728 /packages/utils/bc/Makefile: Feb 26 14:51:26 [package] enable readline support in bc Feb 26 14:51:26 Signed-off-by: Xiangfu Liu Feb 26 14:51:39 florian * r25729 /packages/libs/libxslt/Makefile: Feb 26 14:51:39 [package] libxslt: add InstallDev for libexslt Feb 26 14:51:39 The libxslt package does not InstallDev the exslt dev files when Feb 26 14:51:39 libexslt package is selected. Feb 26 14:51:40 Signed-off-by: Andrew Peebles Feb 26 14:54:32 KanjiMonster: if I make a ram image, how do I actually test it? Feb 26 14:56:02 Olipro: normally you need serial access for that; many bootloaders support loading an (elf) binary into ram through tftp and executing it Feb 26 14:56:20 arse Feb 26 14:56:34 I guess I'll just have to wait for my SPI flashing board to arrive Feb 26 14:56:45 still, in the meantime, I would like to figure out why my image killed my router Feb 26 14:56:57 [florian] there? Feb 26 14:57:03 <[florian]> PsyMan: yes sir Feb 26 14:57:14 Olipro: well, you can also flash it, but then no settings will survive a reboot (assuming it works ;) Feb 26 14:57:35 nice Feb 26 14:57:51 KanjiMonster: I backed up the flash in its entirety before flashing it Feb 26 14:57:58 wipster found a bug on one of the patches Feb 26 14:58:05 sent a patch here Feb 26 14:58:46 http://pastebin.com/sjiLshw6 <-- the patch for the 972 patch Feb 26 14:59:05 2.6.37 kernel Feb 26 14:59:22 basically, I turn it on, the power and switch lights come on Feb 26 14:59:22 <[florian]> PsyMan: that's a patch Wipster sent you right? Feb 26 14:59:24 but the sys led never starts flashing Feb 26 14:59:26 KanjiMonster: I specified "firmware" when flashing with mtd, that's correct yet? Feb 26 14:59:27 yeah Feb 26 14:59:39 <[florian]> PsyMan: I have not committed it yet, but does it help getting your device towork? Feb 26 14:59:45 it did Feb 26 15:00:08 Olipro_: depends on the target. on broadcom targets, its linux Feb 26 15:00:16 it's a WR1043ND Feb 26 15:00:18 <[florian]> PsyMan: awesome! Feb 26 15:00:20 Atheros Feb 26 15:00:35 <[florian]> PsyMan: so your device with a marvell switch not works fine? Feb 26 15:00:44 <[florian]> PsyMan: have you updated acx-mac80211? Feb 26 15:00:44 he forgot that sagem and orion had different mappings Feb 26 15:00:49 yeah Feb 26 15:00:55 I'm thinking I might have a pop at making a modded uBoot to give some basic network access Feb 26 15:01:00 but if i set it to ap mode it freezes the router Feb 26 15:01:15 Olipro_: then firmware should be right Feb 26 15:01:35 <[florian]> PsyMan: ok, you might want to contact the acx-mac80211 maintainer about that Feb 26 15:01:54 ahh, so you haven't tested wifi? Feb 26 15:02:06 damn i was hoping for a workaround xD Feb 26 15:02:08 KanjiMonster: then the key question is, why has it killed my router :P Feb 26 15:02:26 oh another question Feb 26 15:02:31 <[florian]> PsyMan: since I updated it, no Feb 26 15:02:37 Olipro: you probably won't find this out without serial ;) Feb 26 15:02:47 I fancied as much Feb 26 15:02:59 I've ordered a TTL to USB cable from Hong Kong Feb 26 15:03:04 last time i flashed, jffs2 had only 3 blocks available Feb 26 15:03:09 cost under a fiver with postage too Feb 26 15:03:13 router worked but no settings were saved Feb 26 15:03:25 <[florian]> PsyMan: yeah, that's way too small Feb 26 15:03:28 what's a safe number of 64k blocks to work with? Feb 26 15:03:40 I had a weird issue with my jffs2 while it was working Feb 26 15:03:44 5 is minimum afaik, but safe? Feb 26 15:03:47 the router could boot from extroot just fine Feb 26 15:03:57 but if I turned extroot off, it failed and I could only get into failsafe Feb 26 15:04:00 <[florian]> PsyMan: depends on the erase block size, but I do not think that 5 is safe either Feb 26 15:04:09 64k Feb 26 15:04:09 <[florian]> PsyMan: how much flash do you have on your device? Feb 26 15:04:16 4mb Feb 26 15:04:18 :| Feb 26 15:04:36 <[florian]> PsyMan: did you put many additionnal software? Feb 26 15:04:51 <[florian]> PsyMan: you might want to consider using squashfs+jffs2 to save some space on the "static" software Feb 26 15:05:01 <[florian]> PsyMan: software that you do not want to uninstall later Feb 26 15:05:05 upnp, ipv6 tunneling, ip6to4, radvd Feb 26 15:05:16 yeah, stick whatever you *know* you'll need into SquashFS Feb 26 15:05:18 luci with packages related to the above Feb 26 15:05:42 i use only squash cause well Feb 26 15:05:48 i'm gonna need all of that Feb 26 15:05:53 xD Feb 26 15:06:00 <[florian]> oh, so rootfs_data only has 5 blocks available? Feb 26 15:06:10 yeah Feb 26 15:06:26 now it should be to 6 or 7 Feb 26 15:07:00 not planning to install packages on that jffs partition Feb 26 15:07:37 even checked the options to strip unnecessary stuff from kernel and packages Feb 26 15:07:38 <[florian]> considering the space remaining, that would be hard :) Feb 26 15:08:11 <[florian]> if I recall right on ar7, you should be able to save some space by reducing the kernel partition to fit the kernel size exactly Feb 26 15:08:34 done Feb 26 15:09:27 so, while having all files on squash, how much is it used by openwrt? Feb 26 15:10:09 only edits will be stored on jffs of course, no new packages or updates Feb 26 15:11:31 lemme guess... you have no idea do you? xD Feb 26 15:12:10 <[florian]> sorry, I was afk Feb 26 15:12:19 <[florian]> I can't give you the exact space needed Feb 26 15:12:47 <[florian]> you are going to need some space to store configuration files you modify etc .. Feb 26 15:13:01 <[florian]> but for sure that should fit within an erase block Feb 26 15:13:32 ahh Feb 26 15:13:36 just one? Feb 26 15:13:51 will not become 2 in the future? :p Feb 26 15:14:14 <[florian]> apart from configuration files, that's basically what you are going to modify Feb 26 15:14:18 if that's the case, 6 will be enough, since deletion needs 3+1 free Feb 26 15:15:40 and creation 3+2 iirc Feb 26 15:16:25 hmm Feb 26 15:17:25 also, how exactly is the zone info helpful? Feb 26 15:17:43 <[florian]> no idea Feb 26 15:18:15 can you set time and time saving settings without it? Feb 26 15:18:21 florian * r25730 /trunk/target/linux/ar7/patches-2.6.37/972-cpmac_fixup.patch: [ar7] fix MII register ioremap on when high cpmac is available, patch from Wipster Feb 26 15:18:29 (cause i have no idea what it does either) Feb 26 15:18:45 <[florian]> PsyMan: you could set time without storage, that's just about synchronizing to a remote server Feb 26 15:18:51 <[florian]> but time savings needs the zoneinfo Feb 26 15:19:14 hmm Feb 26 15:19:30 <[florian]> but you could do this in the squashfs partition in ro manner Feb 26 15:19:33 specific to region? or the core will do? Feb 26 15:20:03 <[florian]> you basically need to set your timezone and there you go Feb 26 15:20:10 <[florian]> system date gets updated correctly Feb 26 15:20:21 ehm Feb 26 15:20:26 that's the issue Feb 26 15:20:34 i don't know if the core one will do Feb 26 15:20:44 <[florian]> what do you mean by core? Feb 26 15:20:49 or if i need the one for my region too Feb 26 15:21:00 <[florian]> oh yes you need the info for your region Feb 26 15:21:09 ahh Feb 26 15:21:52 <[florian]> that's what is in /usr/share/zoneinfo/* on usual Linux systems Feb 26 15:22:10 could anyone explain what this patch https://dev.openwrt.org/browser/trunk/target/linux/generic-2.6/patches-2.6.32/202-mips_mem_functions_performance.patch?rev=20151 is for? Feb 26 15:22:10 kernel already has these functions, and I don't believe in kdev dumbness Feb 26 15:22:14 ahh Feb 26 15:22:42 so i'm more or less having a problem Feb 26 15:22:45 great :\ Feb 26 15:23:19 <[florian]> PsyMan: strip down what you need in your squashfs image Feb 26 15:23:28 ? Feb 26 15:23:34 what do you mean? Feb 26 15:23:44 <[florian]> PsyMan: basically install the zone for your region, configure /etc from the squashfs image and you won't need post-flashing configuration Feb 26 15:23:56 <[florian]> that saves you some space on the flash Feb 26 15:24:02 hmm Feb 26 15:24:15 <[florian]> theMIROn: by default, such functions are not using gcc's builtin optimizations, thus the patch Feb 26 15:24:25 basically i could just take out whatever i don't need from every part of the code Feb 26 15:24:25 <[florian]> theMIROn: but if you want some metrics, ask nbd, he committed that Feb 26 15:24:28 then recompile Feb 26 15:24:29 xD Feb 26 15:24:42 <[florian]> PsyMan: in fact yes, but mklibs is there for you in that case ;) Feb 26 15:25:07 by default these funcs are not to be optimised, they're coded in asm already Feb 26 15:25:39 thanks for all the info btw :) Feb 26 15:25:40 <[florian]> theMIROn: except that they are written for the most compatible ISA Feb 26 15:25:44 <[florian]> PsyMan: you are welcome Feb 26 15:26:01 only issue now is that acx driver freezing the router Feb 26 15:26:08 ISA? what's it? Feb 26 15:26:09 you never tried ap mode there? Feb 26 15:26:09 <[florian]> theMIROn: so if you compile your kernel with a different ISA, say mips32r2, you would not have these optimized for mips32r2 Feb 26 15:26:16 <[florian]> ISA = instruction set architecture Feb 26 15:26:20 ic Feb 26 15:26:25 <[florian]> PsyMan: I did, but without encryption Feb 26 15:26:33 same here Feb 26 15:26:42 but it just freezes the router Feb 26 15:26:47 i have mips32r1 and mips32r2 kernels, feel no prerf degradation ever Feb 26 15:27:00 I'd better contact with nbd about Feb 26 15:27:17 maybe you used the old acx branch (that one without wpa2 support)? Feb 26 15:27:50 <[florian]> or maybe we just do not have the same tnetw chipset Feb 26 15:28:16 <[florian]> theMIROn: have you measured it? Feb 26 15:28:40 I'm going to right now Feb 26 15:29:07 which router are you on/ Feb 26 15:29:08 ? Feb 26 15:29:41 <[florian]> PsyMan: I am not using an openwrt router, I use my ISP router Feb 26 15:29:49 HAHA Feb 26 15:30:01 <[florian]> long story short, if I don't, I do not benefit from all features, and I need them Feb 26 15:30:13 <[florian]> so from time to time, I switch, but never long enough Feb 26 15:30:19 :p Feb 26 15:31:18 TNETW1130GVF <-- that's the wireless card here Feb 26 15:33:24 theMIROn: hi Feb 26 15:33:41 hi, nbd Feb 26 15:33:43 here's what this patch is for: Feb 26 15:34:06 if gcc uses its builtins for these functions, it does not just see a function call to it Feb 26 15:34:14 but it also has more knowledge about what the functions do Feb 26 15:34:30 so it can make more assumptions about the surrounding code Feb 26 15:35:00 in some cases that did produce some slight performance improvements in my tests Feb 26 15:35:35 i guess it should depend from gcc version and arch Feb 26 15:35:54 could you please note what were exactly the test env? Feb 26 15:36:04 gcc 4.3.3+cs, ar71xx Feb 26 15:36:18 cpu utilization during lan->wireless transfers Feb 26 15:37:00 why not just strighn str* random loop tests? Feb 26 15:37:10 *stright Feb 26 15:37:27 did you ever try this with bcm target? Feb 26 15:37:42 no, i didn't try this with the bcm target, but i don't think this patch can make things worse Feb 26 15:37:50 it simply allows gcc to test whether optimizations can improve things Feb 26 15:37:54 atheros cpu may lack of asm equivalents Feb 26 15:38:06 and if the optimizations don't help, it'll automatically fall back to the kernel implementation of these functions Feb 26 15:38:07 hmm Feb 26 15:38:15 no same chipset i suppose Feb 26 15:38:16 :\ Feb 26 15:38:54 so during my test i used oprofile Feb 26 15:39:00 ic, just remember realtek cpus without mul/div funcs... Feb 26 15:39:01 and i found memcmp to be visible on the list Feb 26 15:39:13 after my changes, cpu utilization triggered by memcmp was quite a bit lower Feb 26 15:39:43 my guess is the compiler was able to inline some constant-length memory compares Feb 26 15:39:55 got it, so I need to profile utilization with my boards Feb 26 15:40:05 <[florian]> PsyMan: no, I need to check Feb 26 15:40:19 ahh Feb 26 15:40:32 maybe i should fill a bug report? :\ Feb 26 15:40:48 <[florian]> you should contact the acx-mac80211 maintainer as I told you Feb 26 15:41:09 <[florian]> he also has an ar7 device, and he tests AP mode in general Feb 26 15:41:12 not only does it hang when applying the change, it ends up disabled on reboot Feb 26 15:41:55 infrastructure mode works though... which is supposed to be based on the same theory xD Feb 26 15:50:20 nbd * r25731 /trunk/include/netfilter.mk: netfilter: add missing modules for v6 conntrack (patch from #8940) Feb 26 15:56:17 hmm Feb 26 15:56:47 no bug tracking on the acx git? that's not really good :\ Feb 26 16:20:23 nbd, hi, I've updated my trunk today and I get an error about compiling ubus Feb 26 16:20:43 show me Feb 26 16:22:05 libubus.c:628:3: error: case value '9' not in enumerated type 'const enum blobmsg_type' Feb 26 16:24:06 nbd, My target is an ar71xx board Feb 26 16:24:12 ahh Feb 26 16:24:13 ok Feb 26 16:29:16 nbd * r25732 /trunk/package/ubus/Makefile: ubus: update to the latest version to fix a compile error with gcc 4.5 Feb 26 16:29:20 tornado: there, fixed ;) Feb 26 16:29:52 nbd, oh thanks :) Feb 26 16:40:31 build #81 of ppc44x is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/81 Feb 26 16:43:12 hmm Feb 26 16:43:17 actually Feb 26 16:43:34 changing ANY wifi settings hangs the system Feb 26 16:43:50 it's just that the rest work after the reboot xD Feb 26 16:44:29 that can't be an acx bug xD Feb 26 16:59:58 the kernel shows a similar enum warning Feb 26 17:00:07 net/ipv4/netfilter/nf_nat_standalone.c:117:2: warning: case value '4' not in enumerated type 'enum ip_conntrack_info' Feb 26 17:00:36 still using .32.29 though Feb 26 17:00:50 I'm getting this error now after ubus fixed (it occurs while compiling xfsprogs): FATAL ERROR: msgmerge does not seem to be installed Feb 26 17:01:14 tornado: I'll take a look Feb 26 17:01:34 build #92 of s3c24xx is complete: Failure [failed compile_10] Build details are at http://tksite.gotdns.org:8010/builders/s3c24xx/builds/92 Feb 26 17:02:32 xMff, thanks Feb 26 17:05:09 xMff, I forgot to write the next line: xfsprogs cannot be built without a working gettext installation Feb 26 17:05:22 its a lie :) Feb 26 17:05:41 should rather be "autofail is too stupid to work without gettext" Feb 26 17:06:06 I still waiting for my build to finish unless I can tackle xfsprogs Feb 26 17:06:15 xMff, :D Feb 26 17:06:15 would be grateful if you could paste the buildlog in the meanwhile Feb 26 17:09:45 ok just a question. is gettext any package that I should install on my distro? Feb 26 17:10:19 no Feb 26 17:10:37 well you can and get away with it but the build should not depend on it in the first place Feb 26 17:10:59 therefore it'd be useful to see the current log Feb 26 17:11:18 I have gettext on my system and cannot reproduce it most likely Feb 26 17:14:17 xMff, http://pastebin.com/E1BavMmy Feb 26 17:15:22 ah ok. I guess I'll add some more fake macros Feb 26 17:15:32 happy pills for autoconf Feb 26 17:15:51 but I can solve it by installing gettext yes? Feb 26 17:15:57 wait Feb 26 17:16:42 yes, you can Feb 26 17:16:51 ok thanks Feb 26 17:17:30 or ln -s /bin/true staging_dir/host/bin/msgmerge Feb 26 17:18:42 xMff, thanks a lot. it fixed Feb 26 17:40:29 jow * r25733 /trunk/package/xfsprogs/Makefile: [package] xfsprogs: explicitely disable gettext and 64bit library support Feb 26 18:04:08 nbd, I have a problem with ath9k. some times the ap starts to show this error: Feb 26 18:04:10 ath: Could not stop RX, we could be confusing the DMA engine when we start RX up Feb 26 18:04:24 ath: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x42000020 Feb 26 18:05:10 and I can't ping ap from station Feb 26 18:16:30 I can reproduce this error when have two systems in ap mode and I put one of them in managed mode by iw command to use scanning and after that I change the mode of this system to sta (in wireless config file) and use wifi command Feb 26 18:18:09 nbd, I hadn't this error before compat-wireless-2011-02-07 Feb 26 18:21:29 nbd, sorry the station shows that error message Feb 26 18:48:46 jow * r25734 /branches/backfire/ (include/host-build.mk include/package.mk rules.mk): [backfire] backport host build hook support Feb 26 19:28:28 xMff: why is it that FPU emulation is set to off in kernel config but there's an enabled FPU option in the OpenWRT config? (AR71xx) Feb 26 19:29:05 softmpu vs. fpu emulator Feb 26 19:29:21 *softfpu Feb 26 19:31:15 I see Feb 26 19:31:35 and using the kernel FPU emulator is slower? or it actually breaks stuff? Feb 26 19:32:17 also, doesn't having softFPU mean that packages using it will be larger than if they simply relied on FPU emulation from the kernel since each package will have the same code built into it Feb 26 19:32:59 slightly, yes. but its faster Feb 26 19:33:12 since gcc can generate optimized code right away Feb 26 19:35:35 Olipro: fpu emulation is much slower, since every fpu instruction results in an mips exception whith a context change to kernel, then the emulation of the instruction, then back to the userspace program, while softfpu just computes it in userspace directly (without using any fpu instructions) Feb 26 19:41:59 tornado: the DMA failed to stop thing is not new Feb 26 19:42:20 it has been there for a long time now and it's usually not fatal Feb 26 19:44:32 nbd, but when I've done that test before a lot but I hadn't seen that error. and when it happens I can't ping the AP properly. I get one pong with about 2 seconds and one request timeout and it repeats contintiniously Feb 26 19:46:04 nbd, and I found that it happens just after what I said for scanning Feb 26 19:56:23 is there any possibility to clear the arp cache under openwrt? the net-tools-arp doesn't seem to support any commandline options :( Feb 26 20:05:02 none that I am aware of Feb 26 20:05:10 k ... thx Feb 26 20:05:37 have a nice weekend dudes ... Feb 26 20:12:29 jow * r25735 /branches/backfire/tools/mm-macros/: [backfire] tools: backport mm-macros Feb 26 20:12:47 jow * r25736 /branches/backfire/tools/xorg-macros/: [backfire] tools: backport xorg-macros Feb 26 20:13:09 jow * r25737 /branches/backfire/tools/missing-macros/: [backfire] tools: backport missing-macros Feb 26 20:13:33 jow * r25738 /branches/backfire/tools/autoconf/Makefile: [backfire] tools: update autoconf to v2.68 Feb 26 20:13:56 jow * r25739 /branches/backfire/tools/automake/ (2 files in 2 dirs): [backfire] tools: update automake to v1.11.1 Feb 26 20:14:25 jow * r25740 /branches/backfire/tools/m4/Makefile: [backfire] tools: update m4 to v1.4.15 Feb 26 20:15:18 jow * r25741 /branches/backfire/ (3 files in 3 dirs): [backfire] backport libtool/libltdl split from trunk Feb 26 20:16:16 jow * r25742 /branches/backfire/tools/Makefile: [backfire] tools: update Makefile after r25735-r25741 Feb 26 20:17:00 jow * r25743 /branches/backfire/package/opkg/ (6 files in 3 dirs): [backfire] package: sync opkg with trunk Feb 26 20:17:17 jow * r25744 /branches/backfire/package/iptables/Makefile: [backfire] package: fix iptables after libtool changes Feb 26 20:17:37 jow * r25745 /branches/backfire/include/autotools.mk: [backfire] include: sync autotools.mk from trunk Feb 26 21:11:05 build #65 of mpc52xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/65 Feb 26 23:52:38 mb * r25746 /trunk/target/linux/omap24xx/ (17 files in 2 dirs): omap24xx: Add 2.6.38 files Feb 27 00:08:17 mb * r25747 /trunk/target/linux/omap24xx/patches-2.6.38/810-mmc-fixes.patch: omap24xx: Update MMC fixes Feb 27 01:54:28 build #86 of ppc40x is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ppc40x/builds/86 Feb 27 02:19:23 build #74 of uml is complete: Failure [failed compile_10] Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/74 Feb 27 02:37:40 build #89 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/89 **** ENDING LOGGING AT Sun Feb 27 02:59:57 2011