Date: Fri, 05 Apr 2013 13:13:12 +0200 From: =?ISO-8859-1?Q?Ren=E9_Ladan?= <rene@freebsd.org> To: Adrian Chadd <adrian@freebsd.org> Cc: arm@freebsd.org, Ian Lepore <ian@freebsd.org> Subject: Re: signal 11 after kernel update r247742 -> r248706 Message-ID: <515EB1C8.3050708@freebsd.org> In-Reply-To: <515D9239.40606@freebsd.org> References: <51545932.9050901@freebsd.org> <1364484652.36972.81.camel@revolution.hippie.lan> <515466B5.4020403@freebsd.org> <5155BC1C.2000104@freebsd.org> <CAJ-VmokCuu1B9hT4XxhyEmegT7X5N31tFBtvz2fuHQY3k=akwg@mail.gmail.com> <515D7E0A.6010504@freebsd.org> <515D9239.40606@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 04-04-2013 16:46, René Ladan wrote: > On 04-04-2013 15:20, René Ladan wrote: >> On 29-03-2013 18:58, Adrian Chadd wrote: >>> On 29 March 2013 09:06, René Ladan <rene@freebsd.org >>> <mailto:rene@freebsd.org>> wrote: >>> >>> >>> >> I'm at r248509 on my rpi and everything is good; maybe that'll help >>> >> narrow the binary search. >>> > Ok, I'll start looking from there. >>> >>> The 'skeleton' unmapped IO commit was r248508, so that one is OK. >>> There seem to be some unmapped IO changes between r248509 and r248706, >>> (e.g. r248510-r248512, r248514-r248522, r248550, r248568-r248569, >>> r248596) but I think it is more related to something closer to >>> userland? >>> >>> But the other commits between r248508 and r248706 look OK to me >>> (including r248534 (jilles, SOCK_CLOEXEC, SOCK_NONLBOCK, >>> MSG_CMSG_CLOEXEC)) >>> >>> >>> Are you able to continue bisecting those changes? >>> >>> Thanks for your help so far! >>> >> Checkpoint: r248509 boots fine (including init), but running e.g. python >> built under r247742 dumps core during startup. >> kdump output: ftp://rene-ladan.nl/pub/freebsd/python2.7.ktrace >> > Maybe it has to do with the fix applied to libsupc++ in r248624 to get > this program working: > > % cat testcpp.cc > #include <iostream> > > using namespace std; > > int main(void) { > try { > throw "throwme"; > } > catch (const char* msg) { > cout << msg << endl; > } > > return 0; > } > > With an older libsupc++ , one gets: > % /home/pi/testcpp > /usr/lib/libsupc++.so.1: Undefined symbol "_Unwind_RaiseException" > % > I just upgraded to r248706 again and it boots fine. The difference with the earlier upgrade is that I now cross-install kernel and world so that the architecture is compatible with the cross-build architecture. I used a native 'make installkernel' / 'make installworld' upgrade earlier which required some ad-hoc install hacks. testcpp runs fine again with the libsupc++ fix: % /home/pi/testcpp throwme % python still dumps core, maybe it needs to be forcibly rebuilt? I'll upgrade to todays CURRENT to see how that behaves. René
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?515EB1C8.3050708>