Re: [PATCH] Recent VM fiasco - fixed

From: Paul Barton-Davis (pbd@Op.Net)
Date: Fri May 12 2000 - 08:28:59 EST

>From: "Jim Nance" <>
>Date: Thu, 11 May 2000 22:51:21 -0400

>On Wed, May 10, 2000 at 10:10:13PM -0700, Linus Torvalds wrote:
>> Because it's a classic "walk a large array in order" test, which means
>> that the worst possible order to page things out in is LRU.
>> So toreally speed up mmap002, the best approach is to try to be as non-LRU
>> as possible, which is obviously the wrong thing to do in real life. So in
>> that sense optimizing mmap002 is a _bad_ thing.
>I agree with you. However, some "real" programs do make sequential passes
>through large blocks of memory. It would be nice if we could detect that
>a program was doing this and keep it from flushing out other processes
>pages. I realize that now is probably not a good time to work on this.

Lets take my audio recorder app (again). It has 24+ 1MB buffers that
it scans sequentially all the time. It also mlocks them, which may
affect this in important ways, I don't know. This behaviour, along
with the concomittant heavy disk i/o, was supported just fine in
2.2.10 and in 2.3.51.

I would say its a fine time to work on it. I just wish I knew more
about the VM system, otherwise I would be trying to help.


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

This archive was generated by hypermail 2b29 : Mon May 15 2000 - 21:00:20 EST