**** BEGIN LOGGING AT Mon May 04 02:59:59 2015 May 04 06:55:41 build #380 of adm8668 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/380 May 04 06:58:25 build #390 of mpc83xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/390 May 04 07:09:20 build #795 of ep93xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/795 May 04 08:40:46 build #130 of ramips.mt7621 is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.mt7621/builds/130 May 04 08:50:43 jow r45602 trunk/package/network/ services/openvpn/patches/120-polarssl-disable-record-splitting.patch services/openvpn/Makefile * openvpn: disable CBC record splitting in PolarSSL/mbedTLS (#19101) May 04 10:15:24 build #934 of ppc40x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/934 May 04 12:28:00 hrm, I'm trying to rebuild a backfire image (I know, I know) and it can't build the packages feed at all. May 04 12:28:38 I've udpated the urls for luci and packages, but I get this odd message installing the packages feed https://pastee.org/wxm9w May 04 12:28:55 this should still be buildable right? May 04 12:29:12 karlp: thats just the latest gnu bullshit May 04 12:29:31 gnu grep now treats files with utf8 in them as binary, leading to the message you see May 04 12:29:40 you can work around it with export LC_ALL=C iirc May 04 12:29:42 urgh :( May 04 12:30:15 the feeds.conf.default url for luci doesn't work in backfire anymore, May 04 12:30:54 http://paste.fedoraproject.org/218186/07426441/ May 04 12:31:21 moment May 04 12:31:40 ah yes, use the github mirror May 04 12:32:41 the xwrt url will fail when google code shuts down too, though I don't personally care much about that one May 04 12:33:09 might be a good idea to clone it while it is still there May 04 12:34:53 huh, the redboot fconfig tool isn't in backfire either, May 04 12:40:52 no idea if anyone ses xwrt at all these days May 04 12:41:17 uses* May 04 12:45:33 weird, redboot sends arps, my laptop replies, but ping fails, I May 04 12:45:37 stupid old busted crap May 04 13:03:26 build #753 of mpc52xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/753 May 04 13:23:27 kaloz r45603 trunk/tools/ sed/patches sed/patches/001-musl_host_fixup.patch * tools/sed: fix compile on musl hosts May 04 13:23:45 kaloz r45604 trunk/tools/autoconf/patches/002-musl_host_fixup.patch * tools/autoconf: fix compile on musl hosts May 04 13:24:03 kaloz r45605 trunk/tools/pkg-config/patches/002-musl_host_fixup.patch * tools/pkg-config: fix compile on musl hosts May 04 16:05:35 russell--: hey, i'm back home May 04 16:05:44 russell--: so you can catch me on IRC again May 04 16:05:51 russell--: goot luck with WGT634U May 04 16:26:26 anyone know _why_ this was done? the commit comment just says the same as the patch, so no context there at all :( http://git.openwrt.org/?p=openwrt.git;a=commitdiff;h=babe7a96eab00e5d8e9748786ef608e656514328 May 04 16:27:22 karlp: a few commits earlier: cmake.mk: enable -Bsymbolic-functions by default for linking shared libs, slightly improves performance and compression May 04 16:40:49 so no LD_PRELOAD for projects that happened to have been built with cmake, but still normal for non-cmake projects? May 04 16:54:07 karlp: LD_PRELOAD still works, just overriding library-internal functions doesn't May 04 16:55:00 karlp: do you have a use case in mind that fails with symbolic-functions? May 04 16:55:16 no, not at all, it just seems a rather blanket change May 04 16:56:41 but at the same time, not everything, jsut cmake, anda few other libs here and there. May 04 16:57:13 for cmake, i can control library ldflags separately from executable ldflags May 04 16:57:28 how does ld_preload stay working, when there's a default impelemtnation of something in a library, and you're meant to link in your own? May 04 16:57:34 for other things, sometimes it might call ld directly, in which case -Wl,... would not work May 04 16:58:08 symbolic-functions does not affect resolving cross-library symbol references May 04 16:58:33 so if libA has symbol X, and libB uses it, you can override it for libB's use of it May 04 16:58:43 e.g. for libc functions and stuff May 04 16:59:00 which i think is the more common case of LD_PRELOAD use May 04 16:59:04 but you can't get libA to use one provided elsewhere with this, which you could before? May 04 16:59:19 right May 04 16:59:24 I know this is esoteric, it just seems like it woudl pop up unusual bugs for some people May 04 16:59:59 it's the sort of thing I would have liked to have seen at least _announced_ to -devel, May 04 17:00:25 not asking for any discussion of it, but just a "hey, this has changed, if you're doing weird things, you might want to have noticed this commit in amongst all the other hundreds of commits" May 04 17:00:44 some numbers on the performance and compressions changes you got would have been nice May 04 17:00:54 the other debates on line about this have all been rather handwavy May 04 17:01:10 0.5-1% improvement in compressed size May 04 17:01:27 runtime performance is harder to quantify May 04 17:01:43 btw. i found quite a few libs that already use this upstream May 04 17:02:01 i think json-c does, polarssl does May 04 17:02:17 it's a common optimization, i'm just making it a bit more generic May 04 19:37:44 rmilecki: /me is away from hardware atm, but some of my initial difficulty involved applying patches that depended on other ones (particularly with respect to KSEG, iirc) within that rev change May 04 20:50:37 build #877 of avr32 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/877 May 04 21:12:52 russell--: that's why I prepared a nice patchset for you :) May 04 21:13:09 russell--: it splits that first bad rev into 5 smaller ones May 04 21:13:23 russell--: you can apply only one of my patch and it'll compile May 04 21:13:29 then you can apply 2nd one May 04 23:07:41 build #944 of cobalt is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/944 May 04 23:10:30 build #395 of cns21xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cns21xx/builds/395 May 04 23:13:13 build #941 of orion is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/941 May 05 00:49:05 build #394 of pxa is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/394 May 05 01:18:41 build #396 of x86_64 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/396 **** ENDING LOGGING AT Tue May 05 02:59:59 2015