Re: [PATCH] 2.2.17pre7 VM enhancement Re: I/O performance on

[Rik van Riel]

On Mon, 26 Jun 2000, Andrea Arcangeli wrote:
> On Sun, 25 Jun 2000, Marcelo Tosatti wrote:
> 
> >With the attached patch, try_to_free_buffers() wakes up kflushd if the
> >percentage of dirty memory is high, independant of sync_page_buffers
> 
> It's wrong to do that there. kflushd is just wakenup by the guy
> that generated the dirty data.

Actually that's not the case. You're forgetting that
kswapd _frees_ memory...

Say we have 11MB of buffer memory and 2MB of dirty
buffers. Now the system frees 7MB of clean buffers and
suddenly we have 4MB of buffers with 2MB dirty.

In that situation we *definately* want kflushd to be
woken up so it can try to sync out the other dirty buffers
before we want to free that memory.

regards,

Rik
--
The Internet is not a network of computers. It is a network
of people. That is its real strength.

Wanna talk about the kernel?  irc.openprojects.net / #kernelnewbies
http://www.conectiva.com/		http://www.surriel.com/