From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Rik van Riel <riel@redhat.com>
Cc: paulmck@linux.vnet.ibm.com, linux-kernel@vger.kernel.org,
linux-mm@kvack.org, jglisse@redhat.com, mgorman@suse.de,
aarcange@redhat.com, airlied@redhat.com,
aneesh.kumar@linux.vnet.ibm.com,
Cameron Buschardt <cabuschardt@nvidia.com>,
Mark Hairgrove <mhairgrove@nvidia.com>,
Geoffrey Gerfin <ggerfin@nvidia.com>,
John McKenna <jmckenna@nvidia.com>,
akpm@linux-foundation.org
Subject: Re: Interacting with coherent memory on external devices
Date: Sat, 25 Apr 2015 13:32:39 +1000 [thread overview]
Message-ID: <1429932759.16571.31.camel@kernel.crashing.org> (raw)
In-Reply-To: <553AFCC1.5070502@redhat.com>
On Fri, 2015-04-24 at 22:32 -0400, Rik van Riel wrote:
> > The result would be that the kernel would allocate only
> migratable
> > pages within the CCAD device's memory, and even then only if
> > memory was otherwise exhausted.
>
> Does it make sense to allocate the device's page tables in memory
> belonging to the device?
>
> Is this a necessary thing with some devices? Jerome's HMM comes
> to mind...
In our case, the device's MMU shares the host page tables (which is why
we can't use HMM, ie we can't have a page with different permissions on
CPU vs. device which HMM does).
However the device has a pretty fast path to system memory, the best
thing we can do is pin the workload to the same chip the device is
connected to so those page tables arent' too far away.
Cheers,
Ben.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2015-04-25 3:32 UTC|newest]
Thread overview: 97+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-21 21:44 Interacting with coherent memory on external devices Paul E. McKenney
2015-04-21 23:46 ` Jerome Glisse
2015-04-22 0:36 ` Benjamin Herrenschmidt
2015-04-22 12:42 ` Paul E. McKenney
2015-04-21 23:49 ` Christoph Lameter
2015-04-22 0:05 ` Jerome Glisse
2015-04-22 0:50 ` Christoph Lameter
2015-04-22 1:01 ` Benjamin Herrenschmidt
2015-04-22 13:35 ` Paul E. McKenney
2015-04-22 13:18 ` Paul E. McKenney
2015-04-22 16:16 ` Christoph Lameter
2015-04-22 17:07 ` Jerome Glisse
2015-04-22 18:17 ` Christoph Lameter
2015-04-22 18:52 ` Paul E. McKenney
2015-04-23 14:12 ` Christoph Lameter
2015-04-23 19:24 ` Paul E. McKenney
2015-04-24 14:01 ` Christoph Lameter
2015-04-24 14:13 ` Paul E. McKenney
2015-04-24 15:53 ` Rik van Riel
2015-04-23 2:36 ` Benjamin Herrenschmidt
2015-04-23 14:10 ` Christoph Lameter
2015-04-23 15:42 ` Jerome Glisse
2015-04-24 14:04 ` Christoph Lameter
2015-04-23 22:29 ` Benjamin Herrenschmidt
2015-04-23 2:30 ` Benjamin Herrenschmidt
2015-04-23 14:25 ` Christoph Lameter
2015-04-23 15:25 ` Austin S Hemmelgarn
2015-04-23 19:33 ` Paul E. McKenney
2015-04-24 14:12 ` Christoph Lameter
2015-04-24 14:57 ` Paul E. McKenney
2015-04-24 15:09 ` Jerome Glisse
2015-04-25 11:20 ` Paul E. McKenney
2015-04-24 15:52 ` Christoph Lameter
2015-04-23 22:37 ` Benjamin Herrenschmidt
2015-04-24 14:09 ` Christoph Lameter
2015-04-23 16:04 ` Rik van Riel
2015-04-22 0:42 ` Benjamin Herrenschmidt
2015-04-22 0:57 ` Paul E. McKenney
2015-04-22 1:04 ` Benjamin Herrenschmidt
2015-04-22 15:25 ` Christoph Lameter
2015-04-22 16:31 ` Jerome Glisse
2015-04-22 17:14 ` Christoph Lameter
2015-04-22 19:07 ` Jerome Glisse
2015-04-23 2:34 ` Benjamin Herrenschmidt
2015-04-23 14:38 ` Christoph Lameter
2015-04-23 16:11 ` Jerome Glisse
2015-04-24 14:29 ` Christoph Lameter
2015-04-24 15:08 ` Jerome Glisse
2015-04-24 16:03 ` Christoph Lameter
2015-04-24 16:43 ` Jerome Glisse
2015-04-24 16:58 ` Christoph Lameter
2015-04-24 17:19 ` Jerome Glisse
2015-04-24 18:56 ` Christoph Lameter
2015-04-24 19:29 ` Jerome Glisse
2015-04-24 20:00 ` Christoph Lameter
2015-04-24 20:32 ` Jerome Glisse
2015-04-25 11:46 ` Paul E. McKenney
2015-04-27 15:08 ` Christoph Lameter
2015-04-27 15:47 ` Jerome Glisse
2015-04-27 16:17 ` Christoph Lameter
2015-04-27 16:29 ` Rik van Riel
2015-04-27 16:48 ` Christoph Lameter
2015-04-27 23:54 ` Benjamin Herrenschmidt
2015-05-13 14:10 ` Vlastimil Babka
2015-05-13 23:38 ` Benjamin Herrenschmidt
2015-05-14 7:39 ` Vlastimil Babka
2015-05-14 7:51 ` Benjamin Herrenschmidt
2015-05-28 18:18 ` Paul E. McKenney
2015-04-27 16:43 ` Jerome Glisse
2015-04-27 16:51 ` Christoph Lameter
2015-04-27 17:21 ` Jerome Glisse
2015-04-27 19:26 ` Christoph Lameter
2015-04-27 19:35 ` Rik van Riel
2015-04-27 20:52 ` Jerome Glisse
2015-04-28 14:18 ` Christoph Lameter
2015-04-28 17:20 ` Jerome Glisse
2015-04-27 16:15 ` Paul E. McKenney
2015-04-27 16:31 ` Christoph Lameter
2015-04-24 23:45 ` Benjamin Herrenschmidt
2015-04-23 18:52 ` Paul E. McKenney
2015-04-24 14:30 ` Christoph Lameter
2015-04-24 14:54 ` Paul E. McKenney
2015-04-24 15:49 ` Christoph Lameter
2015-04-24 16:06 ` Rik van Riel
2015-04-25 11:49 ` Paul E. McKenney
2015-04-24 16:00 ` Jerome Glisse
2015-04-24 16:08 ` Rik van Riel
2015-04-23 17:28 ` Rik van Riel
2015-04-23 2:27 ` Benjamin Herrenschmidt
2015-04-23 14:20 ` Christoph Lameter
2015-04-23 16:22 ` Jerome Glisse
2015-04-24 18:41 ` Oded Gabbay
2015-04-23 19:00 ` Paul E. McKenney
2015-04-22 15:20 ` Christoph Lameter
2015-04-25 2:32 ` Rik van Riel
2015-04-25 3:32 ` Benjamin Herrenschmidt [this message]
2015-04-25 11:55 ` Paul E. McKenney
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1429932759.16571.31.camel@kernel.crashing.org \
--to=benh@kernel.crashing.org \
--cc=aarcange@redhat.com \
--cc=airlied@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=aneesh.kumar@linux.vnet.ibm.com \
--cc=cabuschardt@nvidia.com \
--cc=ggerfin@nvidia.com \
--cc=jglisse@redhat.com \
--cc=jmckenna@nvidia.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=mhairgrove@nvidia.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=riel@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).