Memory-mapped gettimeofday(2)?

[Jeff Van Epps]

It seems to me that Sun has been emphasizing memory mapping as a way to
solve problems lately, so I got to wondering whether certain system calls
had been or could be implemented as accesses in a page of memory which
every process can read but only the kernel can write.  So that something
like gettimeofday() would be only a fetch from memory vs.  the normal
trap/syscall overhead (and possibly blocking??).  I confess to not having
any numbers, so perhaps someone else can tell me that this is a non-issue?

Just a dreamy break from my more mundane Sun questions ;-)

    Jeff Van Epps          jeffv at bisco.kodak.com
                           rochester!kodak!bisco!jeffv