**** BEGIN LOGGING AT Sat Jun 20 02:59:58 2015 Jun 20 03:11:19 is there a way to build just one pkg with -g for gdb? Jun 20 05:13:38 build #35 of ep93xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/35 Jun 20 05:35:39 build #36 of lantiq is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/36 Jun 20 06:23:06 build #35 of mpc52xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/35 Jun 20 06:47:16 plntyk: your squid fix causes a bus error at runtime Jun 20 06:47:44 Program received signal SIGBUS, Bus error. Jun 20 06:47:44 0x0829d461 in Ipc::MemMapSlot::MemMapSlot (this=0xb7b3244c) at MemMap.cc:288 Jun 20 06:51:53 you mean the change with squid_cv_gnu_atomics=yes on x86 ? Jun 20 06:52:05 the one you pastebinned Jun 20 06:52:21 can u link it - i might have forgotten what else was there Jun 20 06:52:50 [Sun 2015-06-14 09:55:12 PM PDT] see http://paste.debian.net/229538/ Jun 20 06:54:21 an if you always deactivat ethe gnu_atomics thingy ? Jun 20 06:55:05 when building for 486 that's a linker error Jun 20 07:27:20 build #35 of octeon is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/35 Jun 20 07:34:31 can you pastebin the linker error ? Jun 20 07:35:27 x86 seems to be broken - at least gnutls would not compile due to stack prot default atm Jun 20 07:35:39 another rerun of compiling toolchain Jun 20 08:24:01 DonkeyHotei: disabling hardened build flags + stack protector and squid compile fine with musl a makefile that has http://paste.debian.net/244068/ as diff to github/packages compiles Jun 20 08:24:55 config.log/ logs/compile.txt are >1mb too big for pastebin Jun 20 08:26:22 the code in that squid source (memmap.cc) hasnt been touched in ages - chances are its not some source error but some other fail Jun 20 08:28:04 you could check / explicitly activate async io - all other configure options or patches look normal (i looked at fedora squid changes, alpine linux (uses musl libc) and buildroot) Jun 20 08:39:23 build #36 of mpc85xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mpc85xx/builds/36 Jun 20 09:24:09 build #37 of lantiq.xrx200 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq.xrx200/builds/37 Jun 20 09:58:12 build #36 of rb532 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/36 Jun 20 10:08:22 plntyk: not being touched in a while doesn't mean bug free, might be that the bug is just benign and wasn't cought yet. When gcc started to warn about foo *bar; ... mem(..., sizeof(foo)); it cought quite a lot of instances Jun 20 10:41:08 <_trine> http://paste.debian.net/244284/ there is a fault with the GCC compiler Version (gcc 4.8.x with Linaro enhancements) ---> Jun 20 11:02:18 DonkeyHotei: make package/busybox/{clean,compile} V=99 CONFIG_DEBUG=y Jun 20 11:02:22 from the http://wiki.openwrt.org/doc/devel/gdb page Jun 20 11:03:11 I'd add it to the http://wiki.openwrt.org/doc/howtobuild/single.package page too, but I don't have my login saved on this compuer Jun 20 11:18:03 _trine: try disabling mips16 for the package Jun 20 11:18:42 <_trine> KanjiMonster, I will but I can't right now as I am recompiling with 4.9 Jun 20 11:19:56 <_trine> the package compiles ok providing you dont have it slected in makemenu which is that case stops the package being made Jun 20 11:20:26 <_trine> *selected Jun 20 11:20:45 <_trine> ggr typo's galore Jun 20 11:21:10 <_trine> the package compiles OK providing you don't have it selected in makemenu which in that case stops the package being made Jun 20 11:22:14 <_trine> so if I do make package/reaver/{clean,compile} it completes without moaning Jun 20 11:22:35 <_trine> but of course there is no package made because I don't have it selected Jun 20 11:22:54 <_trine> when it is selected then I see the error Jun 20 11:22:59 does it actually compile at all then? Jun 20 11:23:08 doesn't it stkip it as it's not selected? Jun 20 11:23:12 <_trine> yes it compiles ok Jun 20 11:23:33 <_trine> but the package cannot be made Jun 20 11:24:46 <_trine> this is not the normal reaver by the way Jun 20 13:35:03 build #35 of adm5120 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/35 Jun 20 13:44:01 nbd, can you update musl to include http://git.musl-libc.org/cgit/musl/commit/?id=ce3688eca920aa77549323f84e21f33522397115 ? It fixes the segfault mentioned in https://github.com/openwrt/packages/pull/1390#issuecomment-113306435 Jun 20 13:53:28 <_trine> KanjiMonster, your suggestion worked thank you Jun 20 15:15:27 build #35 of ar71xx.mikrotik is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ar71xx.mikrotik/builds/35 Jun 20 15:39:25 <_trine> macchanger isnt compiling http://paste.debian.net/245211/ Jun 20 15:46:50 _trine: add #include Jun 20 15:47:01 <_trine> ok thanks plntyk Jun 20 17:37:25 cyrus r46070 trunk/include/ package.mk hardening.mk * buildroot: move hardening options into separate file Jun 20 17:37:27 cyrus r46071 trunk/toolchain/ gcc/final/Makefile gcc/patches/4.8-linaro/030-always-link-ssp-noshared.patch * gcc: musl comes with its own SSP, so use it instead of libssp Jun 20 17:37:30 cyrus r46072 trunk/include/package-defaults.mk * package-defaults: remove libssp for musl (redundant) Jun 20 17:37:31 cyrus r46073 trunk/toolchain/musl/common.mk * musl: apply hardening options (e.g. SSP) also to libc itself Jun 20 17:37:35 cyrus r46074 trunk/package/libs/libpcap/Makefile * libpcap: fixup libtool Jun 20 18:36:57 cyrus r46075 trunk/toolchain/musl/patches/ 001-git-2015-06-16.patch 001-git-2015-06-20.patch * musl: update musl to 2015-06-20 Jun 20 18:36:59 cyrus r46076 trunk/package/libs/uclibc++/Makefile * uclibc++: link against libssp_nonshared instead of libssp Jun 20 18:42:26 plntyk: sorry, i'm not sure what you're suggesting Jun 20 18:43:57 [Sun 2015-06-14 09:26:10 PM PDT] undefined reference to `__sync_fetch_and_add_8' Jun 20 18:44:16 that is the linker error when atomics are enabled Jun 20 18:54:23 DonkeyHotei: well always disable them Jun 20 18:56:43 disabling them means that bus error Jun 20 19:00:34 adding some more compile options from alpine linux does not work ? Jun 20 19:04:36 what compile options? you linked only to buildroot, and they don't use others Jun 20 19:14:45 and alpine appears to use none at all Jun 20 19:14:55 plntyk: ^ Jun 20 19:27:32 here is a simple fix for the mklibs build problems when using musl: https://gist.github.com/glycoknob/0c82c17e48c19e06d3ee not sure if it's doing any stripping Jun 20 19:27:52 should I send this to the list or is there already someone working on fixing this? Jun 20 19:52:08 build #37 of cobalt is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/37 Jun 20 19:52:20 build #37 of pxa is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/37 Jun 20 19:52:24 build #37 of ppc40x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/37 Jun 20 19:52:43 build #37 of cns21xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/cns21xx/builds/37 Jun 20 19:53:10 build #37 of orion is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/37 Jun 20 19:53:39 build #36 of avr32 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/36 Jun 20 20:58:20 build #36 of netlogic is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/netlogic/builds/36 Jun 20 21:33:22 build #37 of x86.64 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86.64/builds/37 Jun 20 22:00:28 build #36 of x86 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/36 Jun 20 22:02:01 build #36 of x86.kvm_guest is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86.kvm_guest/builds/36 **** BEGIN LOGGING AT Sat Jun 20 23:28:00 2015 Jun 21 00:01:23 build #36 of x86.xen_domu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/x86.xen_domu/builds/36 Jun 21 00:04:06 build #36 of mpc83xx is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/36 **** ENDING LOGGING AT Sun Jun 21 02:59:58 2015