Re: [PATCH][5/?] count writeback pages in nr_scanned

[Andrew Morton <akpm@osdl.org>]

Rik van Riel <riel@redhat.com> wrote:
>
> On Thu, 6 Jan 2005, Nick Piggin wrote:
> 
> > I think what Andrea is worried about is that blk_congestion_wait is 
> > fairly vague, and can be a source of instability in the scanning 
> > implementation.
> 
> The recent OOM kill problem has been happening:
> 1) with cache pressure on lowmem only, due to a block device write
> 2) with no block congestion at all
> 3) with pretty much all pageable lowmme pages in writeback state

You must have a wild number of requests configured in the queue.  Is this
CFQ?

I've done testing with "all of memory under writeback" before and it went
OK.  It's certainly a design objective to handle this well.  But that
testing was before we broke it.

> It appears the VM has trouble dealing with the situation where
> there is no block congestion to wait on...

It's misnamed.  We don't "wait for the queue to come out of congestion". 
We simply throttle until a write completes, or, rarely, timeout.

The bug which you fixed would cause the VM to scan itself to death without
throttling.  Did the fix fix things?