From: amylaar@meolyon.hanse.de (Joern Rennecke) Subject: Re: Process protection and Debugging. Date: Sun, 28 Mar 1993 00:15:03 GMT
s_titz@ira.uka.de (Olaf Titz) writes:
>In article <1993Mar18.144500.2669@infograph.com> darin@infograph.com (Darin Wayrynen) writes:
>[protecting addressNULL from being read]
>> But it also breaks code, because not everyone assumes that you
>> can't pass null pointers to string functions for example. On
>> Sun's that causes a core dump, but on HP's/PC's passing nulls
>> to strcat, strcpy, etc (as second operands) is perfectly valid.
>Depends on the library. A library function that checks every pointer
>for NULL before accessing it will cause no problem even if NULL is
>read-protected.
Hmm, I think it's less resource consuming to let the page fault occur
in the rare cases where the software passes NULL pointers, and
have the option to emulate reading 0 - or maybe an error string - if
the offending code was one of said libc functions.
Joern Rennecke