Re: -mmX 4G patches feedback [numbers: how much performance impact]

From: "Martin J. Bligh" <mbligh@aracnet.com>
To: Andrea Arcangeli <andrea@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>, Eric Whiting <ewhiting@amis.com>,
	akpm@osdl.org, linux-kernel@vger.kernel.org
Subject: Re: -mmX 4G patches feedback [numbers: how much performance impact]
Date: Wed, 07 Apr 2004 16:34:51 -0700	[thread overview]
Message-ID: <33900000.1081380891@flay> (raw)
In-Reply-To: <20040407231806.GV26888@dualathlon.random>

--On Thursday, April 08, 2004 01:18:06 +0200 Andrea Arcangeli <andrea@suse.de> wrote:

> On Wed, Apr 07, 2004 at 04:21:44PM -0700, Martin J. Bligh wrote:
>> Speaking of which, pte_highmem is stinking expensive itself. There's
>> probably a large class of workloads that'd work with out pte_highmem
>> if we had 4/4 split (or shared pagetables. Grrr ;-))
> 
> hey, I can add a sysctl in 5 minutes to disable pte_highmem at runtime,
> why do you think it's expensive, it should be not, it's all atomic kmaps
> only doing invlpg. The few workloads trashing on the ptes manipulation
> needs pte_highmem anyways. If I thought it was expensive for any common
> load the sysctl would be already there.

I measured it - IIRC it was 5-10% on kernel compile ... and that was on a
high ratio NUMA which it should have made *better* (as with highmem, the
PTEs can be allocated node locally). I'll try to dig up the old profiles.

M.