One basic question:
What is the actual physical storage place for message queues and named
pipes? Do they exist as part of disk or memory always, implementation
dependent?

In article <1134926135.917040.210370@g47g2000cwa.googlegroups.com>,
 qazmlp1209@rediffmail.com wrote:

> One basic question:
> What is the actual physical storage place for message queues and named
> pipes? Do they exist as part of disk or memory always, implementation
> dependent?

It's implementation-dependent, but most implementations use kernel 
memory for them, since they don't need any permanent storage.  They 
might end up temporarily in the swap partition as a result of paging.

-- 
Barry Margolin, barmar@alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***
*** PLEASE don't copy me on replies, I'll read them in the group ***

qazmlp1209@rediffmail.com wrote:
# One basic question:
# What is the actual physical storage place for message queues and named
# pipes? Do they exist as part of disk or memory always, implementation
# dependent?

There is an i-node and directory entry for a named pipe, but no disk
contents you can access. IPC entities don't even make it as far as a
directory. The pages holding the data may be paged to disk, but normal
users cannot open the swap areas; that's handled by virtual memory
magic for you.

Pay no attention to the kernel behind the curtain.

--
SM Ryan http://www.rawbw.com/~wyrmwif/
Don't say anything. Especially you.

qazmlp1209@rediffmail.com wrote:
> One basic question:
> What is the actual physical storage place for message queues and named
> pipes?
Usually kernel-only accessible memory.

  Do they exist as part of disk or memory always, implementation
> dependent?

Implementation dependant.