[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.DEB.2.10.1502251701340.12985@chino.kir.corp.google.com>
Date: Wed, 25 Feb 2015 17:04:56 -0800 (PST)
From: David Rientjes <rientjes@...gle.com>
To: Christoph Lameter <cl@...ux.com>
cc: Andrew Morton <akpm@...ux-foundation.org>,
Vlastimil Babka <vbabka@...e.cz>,
Pekka Enberg <penberg@...nel.org>,
Joonsoo Kim <iamjoonsoo.kim@....com>,
Johannes Weiner <hannes@...xchg.org>,
Mel Gorman <mgorman@...e.de>,
Pravin Shelar <pshelar@...ira.com>,
Jarno Rajahalme <jrajahalme@...ira.com>,
Greg Thelen <gthelen@...gle.com>, linux-kernel@...r.kernel.org,
linux-mm@...ck.org, netdev@...r.kernel.org, dev@...nvswitch.org
Subject: Re: [patch 1/2] mm: remove GFP_THISNODE
On Wed, 25 Feb 2015, Christoph Lameter wrote:
> On Wed, 25 Feb 2015, David Rientjes wrote:
>
> > NOTE: this is not about __GFP_THISNODE, this is only about GFP_THISNODE.
>
> Well but then its not removing it. You are replacing it with an inline
> function.
>
Removing GFP_THISNODE, not __GFP_THISNODE. GFP_THISNODE, as the commit
message says, is a special collection of flags that means "never try
reclaim" and people confuse it for __GFP_THISNODE.
There are legitimate usecases where we want __GFP_THISNODE, in other words
restricting the allocation to only a specific node, and try reclaim but
not warn in failure or retry. The most notable example is in the followup
patch for thp, both for page faults and khugepaged, where we want to
target the local node but silently fallback to small pages instead.
This removes the special "no reclaim" behavior of __GFP_THISNODE |
__GFP_NORETRY | __GFP_NOWARN and relies on clearing __GFP_WAIT instead.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists