I think a (partial) bitmask is the simpler interface.  I'd think the most
common case is setting up access for just one device, i.e. a small
contiguous range of io ports, so supplying a starting port number and a
word or less of all-ones is very easy for the client side.  And bottom
line, that's the Linux ioperm interface and we're best off having every
actual program in question just use that interface and nothing
Hurd-specific (at least for now), so we can change it all later.

As to the question of new msg ids, this is a generic thing.  In practice,
we are quite sure nothing we have is using the old interfaces at all.  But
there is still a right way to do these things.  The message IDs must change
if the message format or its meaning has changed.  So, e.g. we could make
i386_io_port_add work on a task_t instead of or as well as a thread_t
without changing the message ID, but we should not change the type or
meaning of its parameters wihtout changing the message ID.


_______________________________________________
Bug-hurd mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-hurd

Reply via email to