Re: [PATCH 1/2] mm,fs: introduce helpers around i_mmap_mutex

From: Davidlohr Bueso
Date: Thu Oct 03 2013 - 21:28:32 EST


On Thu, 2013-10-03 at 14:00 -0700, Davidlohr Bueso wrote:
> On Thu, 2013-10-03 at 13:58 -0700, Andrew Morton wrote:
> > On Wed, 2 Oct 2013 13:17:45 -0700 Davidlohr Bueso <davidlohr@xxxxxx> wrote:
> >
> > > Various parts of the kernel acquire and release this mutex,
> > > so add i_mmap_lock_write() and immap_unlock_write() helper
> > > functions that will encapsulate this logic. The next patch
> > > will make use of these.
> > >
> > > ...
> > >
> > > --- a/include/linux/fs.h
> > > +++ b/include/linux/fs.h
> > > @@ -478,6 +478,16 @@ struct block_device {
> > >
> > > int mapping_tagged(struct address_space *mapping, int tag);
> > >
> > > +static inline void i_mmap_lock_write(struct address_space *mapping)
> > > +{
> > > + mutex_lock(&mapping->i_mmap_mutex);
> > > +}
> >
> > I don't understand the thinking behind the "_write". There is no
> > "_read" and all callsites use "_write", so why not call it
> > i_mmap_lock()?
> >
> > I *assume* the answer is "so we can later convert some sites to a new
> > i_mmap_lock_read()". If so, the changelog should have discussed this.
> > If not, still confused.
> >
>
> Yes, that's exactly right. I'll resend with an updated changelog.

So here's an updated changelog, I left it generic enough for us not to
rely on the lock type characteristics, since if/when changed, it can
remain a sleeping lock (rwsem) or not, for rwlock_t.

8<-----------------------------------------
From: Davidlohr Bueso <davidlohr@xxxxxx>
Subject: [PATCH 1/2] mm,fs: introduce helpers around i_mmap_mutex

Various parts of the kernel acquire and release this mutex,
so add i_mmap_lock_write() and immap_unlock_write() helper
functions that will encapsulate this logic. The next patch
will make use of these.

Note that since this lock is currently a mutex, only introduce
write related functions. In the future, the lock type can be
changed and reading semantics can be added such that the lock
can be shared when allowed.

Signed-off-by: Davidlohr Bueso <davidlohr@xxxxxx>
---
include/linux/fs.h | 10 ++++++++++
1 file changed, 10 insertions(+)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 3f40547..b32e64f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -478,6 +478,16 @@ struct block_device {

int mapping_tagged(struct address_space *mapping, int tag);

+static inline void i_mmap_lock_write(struct address_space *mapping)
+{
+ mutex_lock(&mapping->i_mmap_mutex);
+}
+
+static inline void i_mmap_unlock_write(struct address_space *mapping)
+{
+ mutex_unlock(&mapping->i_mmap_mutex);
+}
+
/*
* Might pages of this file be mapped into userspace?
*/
--
1.8.1.4



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/