xfs
[Top] [All Lists]

Re: pagebuf_prepare_write doesn't kmap the page

To: Steve Lord <lord@xxxxxxx>
Subject: Re: pagebuf_prepare_write doesn't kmap the page
From: Marcelo Tosatti <marcelo@xxxxxxxxxxxxxxxx>
Date: Wed, 31 Jan 2001 13:13:49 -0200 (BRST)
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <200101311643.f0VGhgR26778@xxxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx

Does your benchmarks make something hit swap? 

On Wed, 31 Jan 2001, Steve Lord wrote:

> 
> To follow up on my own message - running some micro benchmarks, every single
> individual operation got faster, but dbench got slower.....
> 
> Ho Hum.
> 
> Steve
> 
> 
> > > 
> > > On Wed, 31 Jan 2001, Steve Lord wrote:
> > > 
> > > > 
> > > > OK, I see what is going on here - and I think this could actually
> > > > break regular XFS as well - if you use the loop device on a highmem
> > > > machine. I will fix it in my 2.4.1 merge which should be out in cvs
> > > > in a day or two. 2.4.1 is running with XFS, but I am not happy with
> > > > what the kernel changes have done to performance.
> > > 
> > > Steve, 
> > > 
> > > Could you please describe what performance problems are going on? 
> > > 
> > 
> > 
> > It is difficult to nail down the cause yet, but dbench has taken a
> > significant hit (40%). Single file (bonnie) is comparable with 2.4.0, the
> > dbench issue may be related to the throttling being placed on the
> > ll_rw_block interface by Jens' changes. The problem with working out
> > a bottleneck in dbench is it is such a random pile of stuff that
> > you really cannot point a finger at any one thing. I need to go
> > find some more deterministic measurements.
> > 
> > The performance drop off is not limited to xfs though ext2 appears
> > to suffer as well.
> > 
> > Steve
> > 
> 
> 


<Prev in Thread] Current Thread [Next in Thread>