From: Andrew Walrond <andrew@...>
Date: 2005-02-02T06:31:06+09:00
Subject: Re: [BUG] Segmentation fault

On Tuesday 01 February 2005 21:13, Bertram Scharpf wrote:
>
> Sample code to reproduce it?
>

Hi Bertram,

Contrary to (it seems) popular belief, not all bugs can be reduced to a 
testcase, small or otherwise.

There is a large class of bugs which are difficult or impossible to reproduce 
with a testcase. Memory or stack scribbling, that sort of stuff.

This is likely one of those. Same code, same dataset, 1000 runs, one segfault.

And no, I can't run off another 1000 to catch the bug. Each runs takes 7hrs,  
even using a cluster of dual opteron machines. I expect Rite to be out before 
then ;)

Andrew