Re: SLUB defrag pull request?

[Miklos Szeredi]

On Mon, 13 Oct 2008, Nick Piggin wrote:
> In many cases, yes it seems to. And some of the approaches even if
> they work now seem like they *might* cause problematic constraints
> in the design... Have Al and Christoph reviewed the dentry and inode
> patches?

This d_invalidate() looks suspicious to me:

+/*
+ * Slab has dropped all the locks. Get rid of the refcount obtained
+ * earlier and also free the object.
+ */
+static void kick_dentries(struct kmem_cache *s,
+                               int nr, void **v, void *private)
+{
+       struct dentry *dentry;
+       int i;
+
+       /*
+        * First invalidate the dentries without holding the dcache lock
+        */
+       for (i = 0; i < nr; i++) {
+               dentry = v[i];
+
+               if (dentry)
+                       d_invalidate(dentry);
+       }

I think it's wrong to unhash dentries while they are possibly still
being used.  You can do the shrink_dcache_parent() here, but should
leave the unhashing to be done by prune_one_dentry(), after it's been
checked that there are no other users of the dentry.

Miklos
--
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/