Re: [Lynx-dev] Re: reproducible crash

[Thorsten Glaser]

Thomas Dickey dixit:

>I'd just reduce that to -Werror for a first try, to eliminate compiler
>optimization as a cause.  Aside from some comments regarding gcc 4.0,
>I haven't noticed any recent complains about gcc's optimizer.

This is pretty bad. I cannot reproduce the problem on
 gcc version 3.4.4 20050314 (prerelease) (Debian 3.4.3-13)

Okay, what "we" have is
 gcc version 3.4.5 20050614 (propolice; MirOS 0832)

Another difference is the stack protector...

Tadaa! [compiler line] -fno-stack-protector -fno-unit-at-a-time works.

Then I know to whom I should report this bug. I still wonder why
it is only caused by lynx (at least I don't have ever seen a
problem with -f*no-*unit-at-a-time, only the other way round).

//mirabile
-- 
I believe no one can invent an algorithm. One just happens to hit upon it
when God enlightens him. Or only God invents algorithms, we merely copy them.
If you don't believe in God, just consider God as Nature if you won't deny
existence.              -- Coywolf Qi Hunt