valgrind illegal instruction

[Amit Chaudhuri]

...well it does warn about experiment support for OS/X (versions) at the head...

Google has some examples of unhandled syscall:443 which appear to
relate to lack of translation for particular opcodes on particular
architectures.  So maybe this is a problem with the Valgrind support
for OSX.

Is worth re-running with the 2 suggestions the logs contain:

1) dSYM directory is missing; consider using --dsymutil=yes
2) Rerun with --leak-check=full to see details of leaked memory

?

A

On Tue, Jul 8, 2014 at 7:45 PM, Robert C. Helling <helling at atdotde.de> wrote:
>
> On 07 Jul 2014, at 21:55, Thiago Macieira <thiago at macieira.org> wrote:
>
> Hi,
>
>> Valgrind also prints the bytes near the instruction pointer that caused the
>> SIGILL, so we should be able to decode manually. And with the fault address,
>> you should be able to find the instruction in the output of objdump -Cdr.
>
> I think the most likely explanation is that valgrind does not like to run with Mac OSC Mavericks and its versions of clang and lldb. But see for yourself, here is its output when running subsurface:
>
> http://euve10195.vserver.de/~robert/valgrind.run
>
> Best
> Robert
>
> --
> .oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oO
> Robert C. Helling     Elite Master Course Theoretical and Mathematical Physics
>                       Scientific Coordinator
>                       Ludwig Maximilians Universitaet Muenchen, Dept. Physik
> print "Just another   Phone: +49 89 2180-4523  Theresienstr. 39, rm. B339
>     stupid .sig\n";   http://www.atdotde.de
>