syscall messaging interface API

[David Leimbach]

On Wednesday, July 23, 2003, at 5:40PM, Peter da Silva wrote:

David Leimbach wrote:
 Didn't the L4 folks find a way to make system calls on Pentiums 
without using
software interrupts?  Isn't this like 10x faster?
The nice thing about the way Matt's designing the API is that you
can change the mechanism for the system call and you only need to
change the messageport send routines. So this kind of thing is more
an implementation detail than a basic design decision.
Sure... that's why I brought it up.... Matt's design seems to shield us 
from the
backend which could use any optimal/reliable path.