All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andi Kleen <ak@linux.intel.com>
To: David Rientjes <rientjes@google.com>
Cc: Andi Kleen <andi@firstfloor.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	Andrea Arcangeli <aarcange@redhat.com>,
	lwoodman@redhat.com
Subject: Re: [PATCH 6/8] Add __GFP_OTHER_NODE flag
Date: Tue, 22 Feb 2011 14:52:11 -0800	[thread overview]
Message-ID: <4D643E1B.7050507@linux.intel.com> (raw)
In-Reply-To: <alpine.DEB.2.00.1102221402150.5929@chino.kir.corp.google.com>


> You could make the same argument for anything using kmalloc_node() since
> preferred_zone may very well not be on the allocating cpu's node.

You're right. It is not always, that is why I defined a new flag. In the 
cases where the flag
is passed it is.



>   So you
> either define NUMA_LOCAL to account for when a cpu allocates memory local
> to itself (as it's name implies) or you define it to account for when
> memory comes from the preferred_zone's node as determined by the zonelist.

That's already numa_hit as you say.

I just don't think "local to some random kernel daemon that means 
nothing to the user"
is a useful definition for local_hit.

When I defined the counter I intended it to be local to the user 
process. It always was like
that too, just THP changed the rules.

-Andi



WARNING: multiple messages have this Message-ID (diff)
From: Andi Kleen <ak@linux.intel.com>
To: David Rientjes <rientjes@google.com>
Cc: Andi Kleen <andi@firstfloor.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	Andrea Arcangeli <aarcange@redhat.com>,
	lwoodman@redhat.com
Subject: Re: [PATCH 6/8] Add __GFP_OTHER_NODE flag
Date: Tue, 22 Feb 2011 14:52:11 -0800	[thread overview]
Message-ID: <4D643E1B.7050507@linux.intel.com> (raw)
In-Reply-To: <alpine.DEB.2.00.1102221402150.5929@chino.kir.corp.google.com>


> You could make the same argument for anything using kmalloc_node() since
> preferred_zone may very well not be on the allocating cpu's node.

You're right. It is not always, that is why I defined a new flag. In the 
cases where the flag
is passed it is.



>   So you
> either define NUMA_LOCAL to account for when a cpu allocates memory local
> to itself (as it's name implies) or you define it to account for when
> memory comes from the preferred_zone's node as determined by the zonelist.

That's already numa_hit as you say.

I just don't think "local to some random kernel daemon that means 
nothing to the user"
is a useful definition for local_hit.

When I defined the counter I intended it to be local to the user 
process. It always was like
that too, just THP changed the rules.

-Andi


--
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/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  reply	other threads:[~2011-02-22 22:52 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-02-21 19:07 Fix NUMA problems in transparent hugepages and KSM Andi Kleen
2011-02-21 19:07 ` Andi Kleen
2011-02-21 19:07 ` [PATCH 1/8] Fix interleaving for transparent hugepages Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-22 15:34   ` Christoph Lameter
2011-02-22 15:34     ` Christoph Lameter
2011-02-22 16:29     ` Andrea Arcangeli
2011-02-22 16:29       ` Andrea Arcangeli
2011-02-22 17:11     ` Andi Kleen
2011-02-22 17:11       ` Andi Kleen
2011-02-21 19:07 ` [PATCH 2/8] Change alloc_pages_vma to pass down the policy node for local policy Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-22 15:42   ` Christoph Lameter
2011-02-22 15:42     ` Christoph Lameter
2011-02-21 19:07 ` [PATCH 3/8] Preserve local node for KSM copies Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-22 15:47   ` Christoph Lameter
2011-02-22 15:47     ` Christoph Lameter
2011-02-22 16:20     ` Andrea Arcangeli
2011-02-22 16:20       ` Andrea Arcangeli
2011-02-22 17:15       ` Andi Kleen
2011-02-22 17:15         ` Andi Kleen
2011-02-22 18:24     ` Andi Kleen
2011-02-22 18:24       ` Andi Kleen
2011-02-21 19:07 ` [PATCH 4/8] Preserve original node for transparent huge page copies Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-21 19:07 ` [PATCH 5/8] Use correct numa policy node for transparent hugepages Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-21 19:07 ` [PATCH 6/8] Add __GFP_OTHER_NODE flag Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-22 21:42   ` David Rientjes
2011-02-22 21:42     ` David Rientjes
2011-02-22 21:47     ` Andi Kleen
2011-02-22 21:47       ` Andi Kleen
2011-02-22 22:08       ` David Rientjes
2011-02-22 22:08         ` David Rientjes
2011-02-22 22:52         ` Andi Kleen [this message]
2011-02-22 22:52           ` Andi Kleen
2011-02-21 19:07 ` [PATCH 7/8] Use GFP_OTHER_NODE for transparent huge pages Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-21 19:07 ` [PATCH 8/8] Add VM counters for transparent hugepages Andi Kleen
2011-02-21 19:07   ` Andi Kleen
2011-02-22 16:36   ` Dave Hansen
2011-02-22 16:36     ` Dave Hansen
2011-02-22 16:43     ` Andrea Arcangeli
2011-02-22 16:43       ` Andrea Arcangeli
2011-02-22 18:02       ` Andi Kleen
2011-02-22 18:02         ` Andi Kleen
  -- strict thread matches above, loose matches on Subject: below --
2011-02-23  1:51 Fix NUMA problems in transparent hugepages v2 Andi Kleen
2011-02-23  1:52 ` [PATCH 6/8] Add __GFP_OTHER_NODE flag Andi Kleen
2011-02-23  1:52   ` Andi Kleen
2011-03-03  0:45 Fix NUMA problems in transparent hugepages and KSM Andi Kleen
2011-03-03  0:45 ` [PATCH 6/8] Add __GFP_OTHER_NODE flag Andi Kleen
2011-03-03  0:45   ` Andi Kleen
2011-03-03  2:33   ` KAMEZAWA Hiroyuki
2011-03-03  2:33     ` KAMEZAWA Hiroyuki
2011-03-03 19:59 Fix NUMA problems in transparent hugepages and KSM Andi Kleen
2011-03-03 19:59 ` [PATCH 6/8] Add __GFP_OTHER_NODE flag Andi Kleen
2011-03-03 19:59   ` Andi Kleen
2011-03-07  8:34   ` KOSAKI Motohiro
2011-03-07  8:34     ` KOSAKI Motohiro
2011-03-07 16:33     ` Andi Kleen
2011-03-07 16:33       ` Andi Kleen
2011-03-08  0:19       ` KOSAKI Motohiro
2011-03-08  0:19         ` KOSAKI Motohiro

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=4D643E1B.7050507@linux.intel.com \
    --to=ak@linux.intel.com \
    --cc=aarcange@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lwoodman@redhat.com \
    --cc=rientjes@google.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.