[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [MiNT] XaAES / GEM memory issues



> -----Original Message-----
> From:	Konrad M. Kokoszkiewicz [SMTP:draco@obta.uw.edu.pl]
> Sent:	Friday, January 12, 2001 3:22 PM
> To:	MiNT mailing list
> Subject:	Re: [MiNT] XaAES / GEM memory issues
> 
> The call only would allow to verify pointers. While the main evil thing is
> that an AES with traditional design needs the memory protection to be
> hacked and overridden, which in turn creates all sort of crap problems
> with stability, signals, shutdown, exception handling and so on. That's
> why the F_OS_SPECIAL facility _will_ be removed some day. You apparently
> 
I believe that F_OS_SPECIAL can be improved, but it would require some
modifications to the AES (or adding a TSR that steals the AES trap). If I
understand you, F_OS_SPECIAL currently allows the AES to read and write
*anywhere*, as if memory protection was disabled. What if it only added the
AES-applications' memoryspace to the memoryspace of the AES? appl_init and
appl_exit could trigger this (either directly from the AES, or by a TSR),
adding and removing apps to the AES memoryspace as needed. This way the AES
would still be allowed to f**k with AES applications, but everything else
would be protected, thus allowing traditional AES-designs to still work with
MP but with increased stability and security.

> And btw about removals: we have just exterminated the critical error
> handler. In 1.16 changing this vector will have no effect. I think AES
> developers may be interested in this information.
> 
Doesn't STiNG use this vector? If so, cutting off STiNG won't increase the
popularity of MiNT, but rather support the myth that MiNT is for the elite.

What problems does this solve btw?

Jo Even Skarstein

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

This email with attachments is solely for the use of the individual or 
entity to whom it is addressed. Please also be aware that 
Vital Insurance/DnB Group cannot accept any payment orders or other 
legally binding correspondance with customers as a part of an email. 

This email message has been virus checked by the virus programs used 
in the Vital Insurance/DnB Group.

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *