From mboxrd@z Thu Jan  1 00:00:00 1970
From: Alasdair G Kergon <agk@redhat.com>
Subject: Re: [PATCH] Don't mlock guardpage if the stack is growing up
Date: Tue, 10 May 2011 23:57:00 +0100
Message-ID: <20110510225659.GE9253@agk-dp.fab.redhat.com>
References: <alpine.DEB.2.00.1105082045250.15552@artax.karlin.mff.cuni.cz> <BANLkTikEpaRZK3X4YReeQAwgcv3HAGUwSw@mail.gmail.com> <alpine.DEB.2.00.1105091246550.32299@artax.karlin.mff.cuni.cz> <4DC7D37F.9040308@redhat.com> <20110509224511.GC15227@parisc-linux.org> <BANLkTikz-sGiy2NdzQMPzCWF64cVF-DZ7w@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
To: Linus Torvalds <torvalds@linux-foundation.org>,
	Matthew Wilcox <matthew@wil.cx>,
	Zdenek Kabelac <zkabelac@redhat.com>,
	Mikulas Patocka <mikulas@artax.karlin.mff.cuni.cz>,
	linux-kerne
Return-path: <linux-parisc-owner@vger.kernel.org>
In-Reply-To: <BANLkTikz-sGiy2NdzQMPzCWF64cVF-DZ7w@mail.gmail.com>
List-ID: <linux-parisc.vger.kernel.org>
List-Id: linux-parisc.vger.kernel.org

On Mon, May 09, 2011 at 03:56:11PM -0700, Linus Torvalds wrote:
> So there are sane semantics for the concept, and it would be easy to
> do in the kernel. Whether it's worth doing, I dunno.
 
At this point we have a workaround that seems to work for us.  I find it
ugly and it has needed some tweaking already but we can cope.

If others find similar problems to ours and start replicating the logic
we're using, then that would be the time to give serious thought to a
clean 'sparse' extension.  (Maybe marking individual mappings as MAP_SPARSE
and adding a MCL_NO_SPARSE option to ignore them sounds most promising to me.)

(What other software packages make use of mlockall() and under what
circumstances?)

Alasdair