Re: memory protection

[Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>]

Martin-Eric Racine <q-funk@citenet.net> writes:

|>> Kristoffer Lawson <setok@fishpool.com> writes:
|>> 
|>> |> On 14 Nov 1997, Andreas Schwab wrote:
|>> |>> Broken software deserves to be broken.  Any program that crashes due to MP
|>> |>> is simply broken, because it accesses memory it doesn't own.  There is no
|>> |>> excuse for this.
|>> 
|>> |> Yes, but it does so happen that lots of software like this crash the whole
|>> |> system instead of just being killed by the system. This is not a sign of 
|>> |> a working memory protection system..
|>> 
|>> The problem could be that when a non-MiNT-aware program is killed it
|>> typically won't release any resources like changed vectors or the like.
|>> Especially the AES would mostly be affected.  Remember that MiNT is not a
|>> secure system.  Use Linux instead.

|> Except Thing IS supposed to be MiNT-friendly. N.AES too. 
|> Thing cannot even be started by N.AES correctly.

Then why does it access memory it doesn't own?  That's still a bug.  If
it's not bug-free (which will of course never happen for any non-trivial
program) it should catch SIGSEGV and properly clean up in the handler.

-- 
Andreas Schwab                                      "And now for something
schwab@issan.informatik.uni-dortmund.de              completely different"
schwab@gnu.org