From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
To: Andi Kleen <ak@linux.intel.com>
Subject: Re: [PATCH 6/8] Add __GFP_OTHER_NODE flag
Cc: kosaki.motohiro@jp.fujitsu.com, Andi Kleen <andi@firstfloor.org>,
        akpm@linux-foundation.org, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org, aarcange@redhat.com
In-Reply-To: <20110307163334.GB13384@alboin.amr.corp.intel.com>
References: <20110307173042.8A04.A69D9226@jp.fujitsu.com> <20110307163334.GB13384@alboin.amr.corp.intel.com>
Message-Id: <20110308091843.8A95.A69D9226@jp.fujitsu.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Date: Tue,  8 Mar 2011 09:19:06 +0900 (JST)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 822
Lines: 25

> > Yes, less intrusive. But are you using current NUMA stastics on
> > practical system?
> 
> Yes I do. I know users use it too.
> 
> We unfortunately still have enough NUMA locality problems in the kernel
> so that overflowing nodes, causing fallbacks for process memory etc. are not uncommon. 
> If you get that then numastat is very useful to track down what happens.
> 
> In an ideal world with perfect NUMA balancing it wouldn't be needed,
> but we're far from that.
> 
> Also the numactl test suite depends on them.

If so, I have no objection of cource. :)

Thanks.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/