On Tue, Feb 12, 2013 at 09:17:45PM +0100, Matthias Killian wrote:
> While getting bleeding eyes reading ghostscript code and getting
> closer to the problem

Ouch! that bad?

> i'm really curious why it doesn't happen on amd64 and wether it
> only happens on 32bit archs

I wonder if this recent post on bugs.ghostscript.com could be
related: http://bugs.ghostscript.com/show_bug.cgi?id=693612
"Indeterministic seg fault with 32bit build":

With the 32 bit build of Ghostscript the following command seg faults:

  bin/gs -o test.pgm -dMaxBitmap=400000000 -sDEVICE=pgmraw -r300 ./Bug692368.pdf

This appears to be an indeterminism, looking through the logs the segfault has 
come and gone with various commits.

Bisecting shows the last time the seg fault didn't happen was with the commit 
before:

commit 2740bef445c51680d831ec40753436291f01760a
Author: Alex Cherepanov <[email protected]>
Date:   Thu Aug 23 02:10:20 2012 -0400

    Bug 693268: reinstate big references.
    
    To make big references compatible with 8-byte alignment on 32-bit Windows
    force ref structure to take 16 bytes by adding a dummy uint64_t member
    to the union.

Just a thought.....
MARTIN

Reply via email to