unwinding fallback for mips-irix6 n32

Rainer Orth ro@CeBiTec.Uni-Bielefeld.DE
Fri Jun 17 13:57:00 GMT 2011

Hi Olivier,

> Rainer Orth wrote:
>> I've finally gotten around to this.  Apart from some comment and code
>> cleanups along the lines of the sol2-unwind.h files, I had to minimally
>> adapt the N32 multithreaded code sequence for IRIX 6.5.30 that I'm
>> running here.  While I was at it, I added N64 support which proved to be
>> almost trivial.  You'll probably have to adapt this for the version of
>> IRIX 6.5 you're running, or we could simply skip the single varying
>> insn.
>  Either way is fine with me. There's a micro stronger confidence
>  in exact matches, but this could lead to spurious propagation failures
>  on other variants of the OS where a third version of that insn
>  could show up while still part of a valid context.

agreed: if the number variations we observe remain in the 3-5 range, we
can continue with matching all of them.  I've installed the patch in the
meantime, so feel free to update it with the N64 variants you see on
IRIX != 6.5.30 (might even vary in patches, though).

>> 64-bit stack_check2.adb remains broken, though.  It SEGVs in memcpy, but
>> the stack is corrupted, so I cannot say yet what's going on.

Still can't, but am currently looking into the remaining libjava failures.

>> I'll run a full bootstrap over the weekend.  It seems that the remaining
>> libjava failures are unrelated.
>  OK ...

They were, but libjava on both IRIX and Tru64 UNIX didn't make use of
MD_FALLBACK_FRAME_STATE_FOR yet.  I've just fixed that, which allowed me
to get rid of another testsuite failure there.


Rainer Orth, Center for Biotechnology, Bielefeld University

More information about the Gcc-patches mailing list