[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1305060353.2437.26.camel@edumazet-laptop>
Date: Tue, 10 May 2011 22:45:53 +0200
From: Eric Dumazet <eric.dumazet@...il.com>
To: Christoph Lameter <cl@...ux.com>
Cc: Vegard Nossum <vegardno@....uio.no>,
Pekka Enberg <penberg@...helsinki.fi>,
casteyde.christian@...e.fr,
Andrew Morton <akpm@...ux-foundation.org>,
netdev@...r.kernel.org, bugzilla-daemon@...zilla.kernel.org,
bugme-daemon@...zilla.kernel.org
Subject: Re: [Bugme-new] [Bug 33502] New: Caught 64-bit read from
uninitialized memory in __alloc_skb
Le mardi 10 mai 2011 à 15:33 -0500, Christoph Lameter a écrit :
> On Tue, 10 May 2011, Eric Dumazet wrote:
>
> > Le mardi 10 mai 2011 à 14:38 -0500, Christoph Lameter a écrit :
> >
> > > Optimizing? You think about this as concurrency issue between multiple
> > > cpus. That is fundamentally wrong. This is dealing with access to per cpu
> > > data and the concurrency issues are only with code running on the *same*
> > > cpu.
> > >
> >
> > If you enable irqs, then this object can be allocated by _this_ cpu and
> > given to another one.
>
> That will cause an incrementing of the tid.
>
> > Another cpu can free the page, forcing you to call a very expensive
> > function, that might give obsolete result as soon it returns.
>
> No the other cpu cannot free the page since the page is pinned by
> the current cpu (see PageFrozen()).
>
What happens then ? Other cpu calls kfree() on last nonfreed object for
this slab, and yet the page stay frozen ? How this page is going to be
freed at all ?
> > Maybe I am just tired tonight, this seems very obvious, I must miss
> > something.
>
> Yeah you are way off thinking about cpu to cpu concurrency issues that do
> not apply here.
I fail to understand how current cpu can assert page ownership, if IRQs
are enabled, this seems obvious it cannot.
--
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