[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <137549.1285723160@localhost>
Date: Tue, 28 Sep 2010 21:19:20 -0400
From: Valdis.Kletnieks@...edu
To: "N.P.S. N.P.S." <napohybelskurwysynom2010@...il.com>
Cc: akpm@...ux-foundation.org, mm-commits@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: mmotm 2010-09-28-16-13 uploaded
On Wed, 29 Sep 2010 02:57:59 +0200, "N.P.S. N.P.S." said:
> 2010/9/29 <Valdis.Kletnieks@...edu>:
> > On Wed, 29 Sep 2010 02:27:55 +0200, "N.P.S. N.P.S." said:
> >> Hi,
> >>
> >> 2010/9/29 =A0<akpm@...ux-foundation.org>:
> >> > The mm-of-the-moment snapshot 2010-09-28-16-13 has been uploaded to
> >>
> >> I wonder why are these two functions in include/linux/fs.h
> >> static inline void lock_flocks()
> >> {
> >> }
> >>
> >> static inline void unlock_flocks()
> >> {
> >> }
> >
> > Those were added as part of the kill-the-BKL project. =A0They used to be
> > done as #defines to lock_kernel and unlock_kernel, but became full-fledge=
> d
> > functions in this release. =A0Unfortunately, if you build with CONFIG_FIL=
> E_LOCKING=3Dn
> > you need to provide a dummy stub, which is what was intended here.
> >
> > Does the following (untested) fix the problem for you?
>
> Yes, it does. Thanks!
And right after that I spotted that Randy Dunlap posted an almost identical
patch (with better changelog) earlier this afternoon...
Content of type "application/pgp-signature" skipped
Powered by blists - more mailing lists