Re: designing fast, streaming disk i/o with mmap: help wanted

From: Matthias Urlichs (smurf@noris.net)
Date: Fri Apr 07 2000 - 07:35:57 EST


Hi,

Matti Aarnio:
> Read-ahead (there is no write ahead)

There cannot be write-ahead (the data either hasn't been written yet, or
it's still too "hot"), but there's the possibility of write-behind.

That is, if you have look at page N, which needs to be written, check
N-1 and N-2 and ... to see if they are dirty too. If so, write them.

> [ a thread to do the read-ahead ]
> using aio_*() might be usefull too.
>
Under Linux, that's the same thing... right?

-- 
Matthias Urlichs  |  noris network GmbH   |   smurf@noris.de  |  ICQ: 20193661
The quote was selected randomly. Really.    |      http://www.noris.de/~smurf/
-- 
Be valiant, but not too venturous.
Let thy attire be comely, but not costly.
               --John Lyly

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 21:00:18 EST