Re: [PATCH 1/2] list.h: add debug version of list_empty

["Kirill A. Shutemov" <kirill@shutemov.name>]

On Mon, Mar 21, 2011 at 04:52:06PM -0700, Andrew Morton wrote:
> On Tue, 15 Mar 2011 15:08:42 +0200
> Phil Carmody <ext-phil.2.carmody@nokia.com> wrote:
> 
> > Heed the notice in list_del: "Note: list_empty() on entry does not
> > return true after this, the entry is in an undefined state.", and
> > check for precisely that condition.
> > 
> > There are currently a few instances in the code of this sequence:
> >     if(!list_empty(pnode))
> >         list_del(pnode);
> > which seems to be useless or dangerous if intended to protect from
> > repeated del's. And given that I've seen an oops pointing to a
> > dereference of poison in such a list_empty, I'm veering towards
> > dangerous. This patch would make such errors obvious.
> > 
> > Nothing is changed in the non-DEBUG_LIST build.
> > 
> > ...
> >
> > +
> > +/**
> > + * list_empty - tests whether a list is empty
> > + * @head: the list to test.
> > + */
> > +int list_empty(const struct list_head *head)
> > +{
> > +	if ((head->prev == LIST_POISON2) || (head->prev == LIST_POISON1))
> > +		WARN(1, "list_empty performed on a node "
> > +		     "at %p removed from a list.\n", head);
> > +	else
> > +		WARN((head->prev == head) != (head->next == head),
> > +		     "list_empty corruption. %p<-%p->%p is half-empty.\n",
> > +		     head->prev, head, head->next);
> > +
> > +	return head->next == head;
> > +}
> > +EXPORT_SYMBOL(list_empty);
> 
> The second warning here is triggering maybe a hundred times from all
> over the place just when booting the kernel.
> 
> Here's the first two:
> 
> 
> [   64.295941] WARNING: at lib/list_debug.c:89 list_empty+0x79/0x85()
> [   64.296129] list_empty corruption. ffff880255bcb788<-ffff880255bcb788->ffff88024c3a3c20 is half-empty.

It looks like a race between __list_del() and list_empty().

-- 
 Kirill A. Shutemov