xfs
[Top] [All Lists]

Re: [patch 2/2] xfs: use scalable vmap API

To: Nick Piggin <npiggin@xxxxxxx>
Subject: Re: [patch 2/2] xfs: use scalable vmap API
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Mon, 25 Jan 2010 07:39:44 -0500
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx, linux-mm@xxxxxxxxx
In-reply-to: <20100125123746.GA24406@laptop>
References: <20081021082542.GA6974@xxxxxxxxxxxxx> <20081021082735.GB6974@xxxxxxxxxxxxx> <20081021120932.GB13348@xxxxxxxxxxxxx> <20081022093018.GD4359@xxxxxxxxxxxxx> <20100119121505.GA9428@xxxxxxxxxxxxx> <20100125075445.GD19664@laptop> <20100125081750.GA20012@xxxxxxxxxxxxx> <20100125083309.GF19664@laptop> <20100125123746.GA24406@laptop>
User-agent: Mutt/1.5.19 (2009-01-05)
On Mon, Jan 25, 2010 at 11:37:46PM +1100, Nick Piggin wrote:
> On Mon, Jan 25, 2010 at 07:33:09PM +1100, Nick Piggin wrote:
> > > Any easy way to get them?  Sorry, not uptodate on your new vmalloc
> > > implementation anymore.
> > 
> > Let me try writing a few (tested) patches here first that I can send you.
> 
> Well is it easy to reproduce the vmap failure? Here is a better tested
> patch if you can try it. It fixes a couple of bugs and does some purging
> of fragmented blocks.
> 
> If it does not help, can you tell me how many CPUs in your system?

The simplest one to reproduce it is a 1 cpu kvm virtual machine.  Will
give your patch a try ASAP - while it's easy to reproduce it takes some
time as I appears only when doing a second xfstests run after a first
finished fine, which makes it look like a leak to me.

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