Re: memory barrier question

[Paul E. McKenney]

On Sun, Sep 19, 2010 at 10:15:51PM +0200, Miklos Szeredi wrote:
> On Sun, 19 Sep 2010, Paul E. McKenney wrote:
> > Give it a few years.  There are reportedly already other compilers that do
> > this, which is not too surprising given that the perception of insanity
> > is limited to lockless parallel code.  If you have single-threaded code,
> > such as code and data under a lock (where the data is never accessed
> > without holding that lock), then this sort of optimization is pretty safe.
> > I still don't like it, but the compiler guys would argue that this is
> > because I am one of those insane parallel-programming guys.
> > 
> > Furthermore, there are other ways to get into trouble.  If the code
> > continued as follows:
> > 
> >    LOAD inode = next.dentry->inode
> >    if (inode != NULL)
> >    	LOAD inode->f_op
> > 	do_something_using_lots_of_registers();
> > 	LOAD inode->some_other_field
> > 
> > and if the code expected ->f_op and ->some_other_field to be from the
> > same inode structure, severe disappointment could ensue.  This is because
> > the compiler is within its rights to reload from next.dentry->inode,
> > especially given register pressure.  In fact, the compiler would be within
> > its rights to reload from next.dentry->inode in the "LOAD inode->f_op"
> > statement.  And it might well get NULL from such a reload.
> 
> Except the VFS doesn't allow that.  dentry->inode can go from NULL to
> non-NULL anytime but will only go from non-NULL to NULL when there are
> no possible external references to the dentry.
> 
> The compiler and the CPU cannot move the "LOAD inode->some_field"
> before the "LOAD dentry->inode", because of the conditional, right?

Other than Alpha, the CPU cannot.  The standard -does- permit the
compiler to guess the value of the pointer, thus effectively moving the
load prior to the conditional.  At present, as far as I know, gcc does
not actually do this.

Again, please put at least an ACCESS_ONCE() in.  Trivial to do now,
possibly saving much pain and headache later on.

							Thanx, Paul
--
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/