**** BEGIN LOGGING AT Sun Sep 25 02:59:57 2011
Sep 25 05:18:38 <builder_> build #82 of s3c24xx is complete: Exception [exception compile_7 shell_14 compile_12]  Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/82
Sep 25 05:18:38 <builder_> build #77 of cobalt is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/77
Sep 25 05:18:41 <builder_> build #61 of au1000 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/61
Sep 25 05:18:43 <builder_> build #54 of rdc is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/rdc/builds/54
Sep 25 06:25:28 <builder_> build #83 of s3c24xx is complete: Exception [exception compile shell_14 compile_12]  Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/83
Sep 25 06:25:29 <builder_> build #81 of brcm47xx is complete: Exception [exception download shell_14 compile_12]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/81
Sep 25 06:25:32 <builder_> build #83 of at91 is complete: Exception [exception shell_2 shell_14 compile_12]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/83
Sep 25 06:25:34 <builder_> build #81 of ubicom32 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/81
Sep 25 07:49:16 <builder_> build #79 of atheros is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/79
Sep 25 07:49:17 <builder_> build #78 of lantiq is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/78
Sep 25 07:49:20 <builder_> build #75 of ps3 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ps3/builds/75
Sep 25 07:49:22 <builder_> build #76 of ramips is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/76
Sep 25 09:54:47 <builder_> build #71 of ppc44x is complete: Exception [exception shell_2 shell_14 compile_12]  Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/71
Sep 25 09:54:48 <builder_> build #76 of pxcab is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/76
Sep 25 09:54:51 <builder_> build #71 of ppc40x is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/71
Sep 25 09:54:53 <builder_> build #74 of rb532 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/74
Sep 25 10:01:47 <builder_> build #71 of x86 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/71
Sep 25 10:01:48 <builder_> build #68 of ar7 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ar7/builds/68
Sep 25 10:01:51 <builder_> build #66 of ixp4xx is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/ixp4xx/builds/66
Sep 25 10:01:53 <builder_> build #64 of kirkwood is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/64
Sep 25 10:55:07 <dottedmag> What do directories staging_dir/target-$arch/{host,packages,root-$shortarch} do?
Sep 25 10:55:39 <dottedmag> Shouldn't at least staging_dir/$target-$arch/host contentsbe installed to staging_dir/host?
Sep 25 13:27:58 <swalker> updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/index.html
Sep 25 14:16:06 <CIA-73> jow * r28298 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: [package] mac80211: rework mac address generation for virtual interfaces; set locally administered bit on generated macs and change the last two bytes isntead of the first one
Sep 25 14:16:41 <CIA-73> jow * r28299 /branches/backfire/package/mac80211/files/lib/wifi/mac80211.sh: [backfire] merge r28298
Sep 25 15:54:11 <DelphiWorld> hello
Sep 25 17:06:30 <EqUaTe> would someone poke whoever is in charge of the oidentd package and get them to apply http://patchwork.openwrt.org/patch/1325/ ?
Sep 25 18:36:08 <CIA-73> nbd * r28300 /trunk/package/mac80211/patches/300-pending_work.patch: ath9k: merge a pending fix for powersave
Sep 25 18:36:24 <CIA-73> nbd * r28301 /trunk/package/mac80211/patches/ (3 files): ath9k: fix some tx aggregation issues - reduces the number of tx dma stop failures
Sep 25 18:37:22 <jow_laptop> EqUaTe: do you know something about the origin of these patches?
Sep 25 18:53:19 <EqUaTe> jow_laptop: the oidentd one?
Sep 25 18:53:42 <EqUaTe> ones*
Sep 25 18:53:51 <jow_laptop> yep
Sep 25 18:55:00 <EqUaTe> yes. give me a sec.
Sep 25 18:58:17 <EqUaTe> jow_laptop: some are from here: http://sa.me.uk/pub/patches/oidentd/ , the extra bit(s) are from myself and KanjiMonster.
Sep 25 19:03:50 <jow_laptop> EqUaTe: ok
Sep 25 19:52:01 <CIA-73> jow * r28302 /packages/net/samba3/ (Makefile files/samba.init): [packages] samba3: don't source /etc/functions.sh in the initscript, rc.local takes care of that in cross-compile firendly way
Sep 25 19:53:06 <CIA-73> jow * r28303 /branches/packages_10.03.1/net/samba3/ (Makefile files/samba.init): [packages_10.03.1] r28302
Sep 25 23:37:46 <philipp64|laptop> it's impressive that www.kernel.org is still down. I guess linux-3.1 won't be out any time soon, as most development will grind to a halt until then.
Sep 25 23:53:39 <jow_laptop> well the lkml is online and linux moved his repo to github
Sep 25 23:53:51 <jow_laptop> so I think it has not such a big impact
Sep 26 00:02:13 <philipp64|laptop> that's reassuring.
Sep 26 00:02:26 <philipp64|laptop> git is down too.
Sep 26 00:04:25 <larsc> 3.1 will probably released in the first week of october, iirc
Sep 26 00:06:46 <philipp64|laptop> up to rc7... wow.
Sep 26 00:07:17 <philipp64|laptop> I thought the point of the whole 3.x renumbering was to make it easier to crank out versions more regularly.
Sep 26 00:07:32 <philipp64|laptop> or did I misunderstand?
Sep 26 00:07:57 <larsc> maybe
Sep 26 00:08:09 <philipp64|laptop> I'm looking at the branch torvalds/linux if that wasn't clear.
Sep 26 00:09:52 <philipp64|laptop> odd that https://github.com/torvalds/linux/tree/master/arch/x86/platform doesnt yet contain geode.
Sep 26 00:10:26 <philipp64|laptop> I thought Andrew's stuff had already bubbled up.
Sep 26 00:10:54 <philipp64|laptop> it was in linux-next right after 3.0 was released.
Sep 26 00:12:08 <philipp64|laptop> maybe it got backed out as they reconstruct the merges...
Sep 26 00:12:41 <philipp64|laptop> were any specific exploits found in the tree?
Sep 26 00:26:48 <builder_> build #54 of mpc52xx is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/54
Sep 26 00:40:03 <builder_> build #82 of ubicom32 is complete: Failure [failed compile_3]  Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/82
Sep 26 00:53:19 <jow_laptop> philipp64|laptop: according to the press, no
Sep 26 01:00:10 <builder_> build #84 of at91 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/84
Sep 26 01:10:40 <jow_laptop> philipp64|laptop: zoneinfo works for me after fixing the default -> Default thing
Sep 26 01:11:06 <jow_laptop> philipp64|laptop: ./scripts/feeds install zoneinfo-core, afterwards it appears in menuconfig  LuCI -> ZoneInfo
Sep 26 01:15:27 <philipp64|laptop> hm... thanks. I did a "feeds install zoneinfo" which silently did nothing... and I took that to mean success.
Sep 26 01:16:03 <philipp64|laptop> does it need a Build/Configure definition, or can it be absent one?
Sep 26 01:18:13 <jow_laptop> nope, its due to the way the feed install works
Sep 26 01:18:37 <jow_laptop> you need to install one of the packages provided by a Makefile, that usually corresponds to the foldername but not in this case
Sep 26 01:18:52 <jow_laptop> simply do  ./scripts/feeds install zoneinfo-core and it will pull the Makefile
Sep 26 01:19:26 <jow_laptop> and also update the luci feed while you're at it
Sep 26 01:19:31 <jow_laptop> I just updated it to 2011i
Sep 26 01:20:06 <jow_laptop> it = zoneinfo
Sep 26 01:20:10 <philipp64|laptop> yeah, saw that. thanks. :-)
Sep 26 01:21:11 <philipp64|laptop> would it make sense to make zoneinfo-core a dependency of eglibc?
Sep 26 01:21:23 <jow_laptop> hm, I think I'll relocate it into the packages feed, it does not really belong to luci
Sep 26 01:21:52 <philipp64|laptop> I don't know anyone who enjoys their logs not being in localtime. :-)
Sep 26 01:22:51 <philipp64|laptop> oh, and that raises the question... what scripting changes are required to support the timezone with zoneinfo?
Sep 26 01:23:21 <philipp64|laptop> uClibc uses /etc/TZ, but zoneinfo uses /etc/localtime if I remember...
Sep 26 01:23:33 <jow_laptop> I think none
Sep 26 01:23:55 <jow_laptop> zoneinfo is a binary database format which (e)glibc understands natively
Sep 26 01:24:16 <jow_laptop> /etc/localtime simply contains a name like "Europe/Berlin"
Sep 26 01:27:45 <philipp64|laptop> on my fedora system, it's not just the last line of the file but the entire file with the binary structure at the top.
Sep 26 01:28:17 <philipp64|laptop> but my question was really, do we need to take the value from /etc/config/system and do an "echo $xxx > /etc/localtime" ?
Sep 26 01:29:27 <jow_laptop> the "timezone" variable used there is unsuitable for that
Sep 26 01:29:37 <jow_laptop> luci additionally sets a variable "zonename" which can be used
Sep 26 01:31:04 <philipp64|laptop> ah. so to make use of zoneinfo, you need to have luci, or be prepared to set it manually?
Sep 26 01:31:29 <philipp64|laptop> (and by manually, I mean doing a copy of the time at the command line...)
Sep 26 01:31:50 <jow_laptop> so there needs to be something like  zone="$(uci get system.@system[0].zonename)"; [ -n "$zone" && -f "/usr/share/zoneinfo/$zone" ] && ln -s "/usr/share/zoneinfo/$zone" /etc/localtime
Sep 26 01:32:45 <jow_laptop> and actually /etc/localtime should point to /tmp/localtime which then is the actual symlink into /usr/share... (flash wearout)
Sep 26 01:33:10 <philipp64|laptop> that will work, yeah...
Sep 26 01:33:21 <philipp64|laptop> I just did: ln -s /usr/share/zoneinfo/Los_Angeles /etc/localtime
Sep 26 01:33:25 <philipp64|laptop> and it did the trick.
Sep 26 01:35:54 <philipp64|laptop> du -k on /usr/share/zoneinfo says 937K, if that matters.
Sep 26 01:36:49 <jow_laptop> http://luci.subsignal.org/~jow/base-files-link-zoneinfo.patch
Sep 26 01:36:51 <philipp64|laptop> I suspect the squashfs size is significantly less, and looking in bin/x86/ seems to confirm this.
Sep 26 01:38:30 <philipp64|laptop> make -j1 package/{base-files,linux}/install V=99 to reimage?
Sep 26 01:39:20 <jow_laptop> just base-files
Sep 26 01:39:39 <philipp64|laptop> that builds the squashfs as well?
Sep 26 01:40:10 <jow_laptop> make package/base-files/clean world V=99
Sep 26 01:40:27 <jow_laptop> make package/base-files/clean target/linux/install V=99
Sep 26 01:40:31 <jow_laptop> that one should work too
Sep 26 01:40:34 <jow_laptop> and is faster
Sep 26 01:40:49 <philipp64|laptop> errr... target/linux/install yes, sorry.
Sep 26 01:42:44 <jow_laptop> k target/linux/install does not trigger compile, so
Sep 26 01:42:53 <jow_laptop> make package/base-files/{clean,compile} target/linux/install V=99
Sep 26 01:43:15 <jow_laptop> make package/base-files/{clean,compile,install} target/linux/install V=99
Sep 26 01:43:25 <jow_laptop> phew
Sep 26 01:48:50 <philipp64|laptop> ok, will re-run it.
Sep 26 01:49:51 <philipp64|laptop> that picks up files/ as well, right?
Sep 26 01:50:13 <jow_laptop> yep
Sep 26 01:51:05 <philipp64|laptop> was a decision made on using more than one processor for mksquashfs?
Sep 26 01:51:30 <jow_laptop> wasn't even ware that question was raised
Sep 26 01:51:39 <philipp64|laptop> I'm not aware of any instability that running it on more than one processor...
Sep 26 01:51:47 <jow_laptop> one could probably just pass -j through
Sep 26 01:51:52 <philipp64|laptop> someone submitted a patch for it, I think... let me look.
Sep 26 01:52:09 <philipp64|laptop> detecting -j in makefiles is a lot harder than one would think, as I found out.
Sep 26 01:52:46 <philipp64|laptop> I tried to force all wget's to not be chatty in parallel make because it makes the logs look unreadable, but couldn't figure out how.
Sep 26 01:53:12 <philipp64|laptop> not in a clean way, anyway.... had to do a ps and get the process group then grep for other make instances...
Sep 26 01:54:37 <philipp64|laptop> hmmm. Eugene San, 8/29... getting the patchwork link...
Sep 26 01:55:42 <philipp64|laptop> http://patchwork.midlink.org/patch/1368/
Sep 26 01:56:49 <jow_laptop> I don't have CONFIG_PKG_BUILD_JOBS in my .config
Sep 26 01:57:13 <philipp64|laptop> yeah, I was just going to mention that I hadn't seen that variable before...
Sep 26 01:58:33 <philipp64|laptop> well, there's $(shell $((`grep -c '^processor' /proc/cpuinfo`+1)) )
Sep 26 01:58:42 <philipp64|laptop> but that would break on Darwin.
Sep 26 02:01:57 <philipp64|laptop> I like your patch except... maybe we should make /etc/localtime be a permanent part of the r/o filesystem?
Sep 26 02:02:05 <CIA-73> jow * r28304 /packages/utils/zoneinfo/ (. Makefile): [packages] utils: add zoneinfo package (moved from LuCI feed)
Sep 26 02:02:10 <philipp64|laptop> create the link at build time, in other words?
Sep 26 02:04:07 <CIA-73> jow * r28305 /trunk/include/image.mk: [include] image.mk: pass through $(CONFIG_PKG_BUILD_JOBS) to the mksquashfs -processors option
Sep 26 02:07:09 <jow_laptop> philipp64|laptop: yes it could go into base-files directly eventually
Sep 26 02:10:51 <jow_laptop> philipp64|laptop: I updated the patch, same location
Sep 26 02:11:06 <philipp64|laptop> looking...
Sep 26 02:11:12 <jow_laptop> its part of the glibc specific install section which is also called by eglibc
Sep 26 02:11:26 <jow_laptop> I mean the symlink generation
Sep 26 02:12:50 <philipp64|laptop> on my 3 year old 4-core, the parallel squashfs just screams on a 53MB squash-size filesystem.
Sep 26 02:13:42 <philipp64|laptop> unhappiness
Sep 26 02:13:43 <philipp64|laptop> ln: failed to create symbolic link `/home/philipp/openwrt2/build_dir/linux-x86_geos/base-files/ipkg-x86/libc/etc/localtime': No such file or directory
Sep 26 02:15:17 <philipp64|laptop> looks like we're missing an mkdir .../etc
Sep 26 02:17:38 <jow_laptop> move it after the glibc-files copy
Sep 26 02:19:09 <philipp64|laptop> I put it after the second done in the macro and that seems to be working...
Sep 26 02:21:02 <philipp64|laptop> what happens when log_type isn't circular?
Sep 26 02:21:23 <jow_laptop> it will grow
Sep 26 02:25:06 <philipp64|laptop> ah. so we'd need some sort of script to rotate the logs?
Sep 26 02:25:34 <philipp64|laptop> anyway, with the 2 lines you added after the 'for' loops, it works beautifully.
Sep 26 02:29:37 <CIA-73> jow * r28306 /trunk/package/base-files/ (Makefile files/etc/init.d/boot): [package] base-files: use "zonename" system option to symlink /etc/localtime via /tmp/localtime to /usr/share/zoneinfo for glibc and eglibc
Sep 26 02:30:35 <philipp64|laptop> and thanks again.
Sep 26 02:33:02 <jow_laptop> np. bbl (in about a week)
Sep 26 02:43:05 <philipp64|laptop> happy travels.
**** ENDING LOGGING AT Mon Sep 26 02:59:56 2011