Todd Vierling writes:

> On 30 Sep 1998, Chris G. Demetriou wrote:
> 
> : I just read the solaris manual page.
> : 
> : It specifically says "the kernel uses this information."  I don't see
> : any place to indicate that the kernel just throws it away, or actually
> : even that the kernel may ignore the advice, but it's unreasonable to
> : expect that the advice will be followed unconditionally.  However,
> : there is the bald statement that the kernel _does_ use the
> : information, rather than just throw it away.
> 
> Then why would it be so bad to pretend to accept the data?
> 
> Not to mention it's rather rude that we define the call, document the call,
> have defines for the call in <sys/mman.h>, and die with EOPNOTSUPP when
> given valid memory addresses.  The least we could do is pretend to be happy
> with the information, as long as the memory range is valid.

It should probably do sanity checking on the arguments and check that
they are ok before doign nothign with them. :)  EOPNOTSUPP should
indicate that the arguments were ok but the advisory was ignored.

Then it's not so rude.  (Or ruder, depending on how you look at it :)

Cheers,

Dave