Re: vk_l4 -- CVS Setup

[Farid Hajji]

> Network transparent RPC definitely seems like a userspace candidate to
> me. The RPC you are describing would be built as messages to some sort
> of RPC server (L4 server) making use of the L4 ipc calls to communicate
> the message to the server. All concepts of networking and such would be
> strictly outside of the L4 kernel.
Most of it would be done outside L4 via a proxy L4 task that would forward
messages to remove proxies. I see no reason to do this inside L4. And you're
right: care must be take of network transparency issues like endianness etc.
All this (say: marshalling/unmarshalling the message) belongs into the realm
of user-space [the proxy task].

The question is wether GUIDs (globally unique IDs) are needed or not.
A related question is wether L4 should provide a redirect mechanism,
so that non-local GUIDs are redirected to the TID of the proxy task
for further processing.

Actually, GUIDs seem like a bad idea anyway, because L4 would waste cycles
on _every_ IPC, just to discriminate between remote and local IDs. If it
provided a different syscall, the user-thread would need to discriminate
itself (then again: why provide GUIDs at all, just contact the proxy directy
via regular local IPC).

So, GUIDs are unlikely to be implemented in L4 in the forseeable future.
IPC transparency will have to be dealt with in user-space, probably through
some kind of general purpose library that would provide:
  * macros for local IPCs,
  * discrimination code, including redirecting to proxy IDs,
  * marshalling/unmarshalling messages (protocol for the proxy),
and of course proxy tasks themselves, including the inter-proxy
protocol ;-).

> -- Ian

-Farid.

-- 
Farid Hajji -- Unix Systems and Network Admin | Phone: +49-2131-67-555
Broicherdorfstr. 83, D-41564 Kaarst, Germany  | address@hidden
- - - - - - - - - - - - - - - - - - - - - - - + - - - - - - - - - - - -
One OS To Rule Them All And In The Darkness Bind Them... --Bill Gates.