[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20161115185015.GA5854@lst.de>
Date: Tue, 15 Nov 2016 19:50:15 +0100
From: Christoph Hellwig <hch@....de>
To: Dan Williams <dan.j.williams@...el.com>
Cc: Christoph Hellwig <hch@....de>,
Dave Hansen <dave.hansen@...el.com>,
"linux-nvdimm@...ts.01.org" <linux-nvdimm@...ts.01.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Linux MM <linux-mm@...ck.org>,
Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH] mm: add ZONE_DEVICE statistics to smaps
Hi Dan,
On Mon, Nov 14, 2016 at 07:14:22PM -0800, Dan Williams wrote:
> Wanted to get your opinion on this given your earlier concerns about
> the VM_DAX flag.
>
> This instead lets an application know how much of a vma is backed by
> ZONE_DEVICE pages, but does not make any indications about the vma
> having DAX semantics or not. I.e. it is possible that 'device' and
> 'device_huge' are non-zero *and* vma_is_dax() is false. So, it is
> purely accounting the composition of the present pages in the vma.
>
> Another option is to have something like 'shared_thp' just to account
> for file backed huge pages that dax can map. However if ZONE_DEVICE
> is leaking into other use cases I think it makes sense to have it be a
> first class-citizen with respect to accounting alongside
> 'anonymous_thp'.
This counter sounds fine to me, it's a debug tool and not an obvious
abuse candidate like VM_DAX. But I'll defer to the VM folks for a real
review.
Powered by blists - more mailing lists