On Thursday 22 May 2003 6:32 pm, Eugene Crosser wrote:

> I think that it would be ideal to have a function that you can pass a
> pointer to a MIME message in memory (possibly mmaped file), which would

If a file is small enough to fit in a mmaped file then writing to a disc file 
won't result in a physical write since it'll sit in the disc buffer cache 
system, the file being removed before a physical write taked place.

If a file is so big that it won't fit in the cache and will cause a physical 
write then by definition it's too big to fit in ram in a mmaped file so it'll 
cause paging/swapping increasing the disc activity that you tried to stop.

What would work better is feeding to a pipe instead of the current mechanism 
of building a copy in memory. The new clamdscan achitecture makes that more 
difficult to implement than with the old clamscan system which is why I 
haven't moved over to that scheme, not because clamdscan is worse than 
clamscan, it's just the API into the mbox code which doesn't allow the 
flexiblity to implement such a solution

However I am still thinking about it to see if I can come up with a solution.

-Nigel

-- 
Nigel Horne. Arranger, Composer, Typesetter.
NJH Music, Barnsley, UK.  ICQ#20252325
[EMAIL PROTECTED] http://www.bandsman.co.uk


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


Reply via email to