On Fri, 21 Jan 2011 15:44:20 -0600 [email protected] wrote: > From: Shirish Pargaonkar <[email protected]> > > A new type of key, cifs_idmap_key_type, is used. > > Keys are instantiated and searched using credential of the root by > overriding and restoring the credentials of the caller requesting the key. > Once a SID is mapped, it is stored along with the mapped id, in one > of the rb trees and key is released. > > > Signed-off-by: Shirish Pargaonkar <[email protected]> > --- > fs/cifs/cifsacl.c | 94 > +++++++++++++++++++++++++++++++++++++++++++++++++++ > fs/cifs/cifsacl.h | 5 +++ > fs/cifs/cifsfs.c | 6 +++ > fs/cifs/cifsproto.h | 2 + > fs/cifs/connect.c | 12 ++++++ > 5 files changed, 119 insertions(+), 0 deletions(-) >
[...] > index 18d3c77..ac6f536 100644 > --- a/fs/cifs/connect.c > +++ b/fs/cifs/connect.c > @@ -3099,6 +3099,18 @@ cifs_umount(struct super_block *sb, struct > cifs_sb_info *cifs_sb) > } > spin_unlock(&cifs_sb->tlink_tree_lock); > > + root = &cifs_sb->uidtree; > + spin_lock(&cifs_sb->siduidlock); > + while ((node = rb_first(root))) > + rb_erase(node, root); > + spin_unlock(&cifs_sb->siduidlock); > + > + root = &cifs_sb->gidtree; > + spin_lock(&cifs_sb->sidgidlock); > + while ((node = rb_first(root))) > + rb_erase(node, root); > + spin_unlock(&cifs_sb->sidgidlock); > + > tmp = cifs_sb->prepath; > cifs_sb->prepathlen = 0; > cifs_sb->prepath = NULL; This cache grows without bound until an unmount occurs? I don't think that's a good idea. It would be easy for someone malicious to DoS the box by simply putting a bunch of files owned by unique SIDs in a directory and then doing a "ls -l" in that dir. Perhaps you should register a shrinker for this cache to allow the VM to prune LRU entries whenever memory gets tight? See "struct shrinker" in include/linux/mm.h. -- Jeff Layton <[email protected]> -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
