From: Al Viro Date: Wed, 14 Sep 2011 17:55:41 +0000 (+0100) Subject: restore pinning the victim dentry in vfs_rmdir()/vfs_rename_dir() X-Git-Tag: v3.0.5~120 X-Git-Url: https://git.kernelconcepts.de/?a=commitdiff_plain;h=97abc52eb26f2ae8d033100c4c024e6138bdfd60;p=karo-tx-linux.git restore pinning the victim dentry in vfs_rmdir()/vfs_rename_dir() commit 1d2ef5901483004d74947bbf78d5146c24038fe7 upstream. We used to get the victim pinned by dentry_unhash() prior to commit 64252c75a219 ("vfs: remove dget() from dentry_unhash()") and ->rmdir() and ->rename() instances relied on that; most of them don't care, but ones that used d_delete() themselves do. As the result, we are getting rmdir() oopses on NFS now. Just grab the reference before locking the victim and drop it explicitly after unlocking, same as vfs_rename_other() does. Signed-off-by: Al Viro Tested-by: Simon Kirby Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- diff --git a/fs/namei.c b/fs/namei.c index 14ab8d3f2f0c..b456c7a2e6b4 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2582,6 +2582,7 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry) if (!dir->i_op->rmdir) return -EPERM; + dget(dentry); mutex_lock(&dentry->d_inode->i_mutex); error = -EBUSY; @@ -2602,6 +2603,7 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry) out: mutex_unlock(&dentry->d_inode->i_mutex); + dput(dentry); if (!error) d_delete(dentry); return error; @@ -3005,6 +3007,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, if (error) return error; + dget(new_dentry); if (target) mutex_lock(&target->i_mutex); @@ -3025,6 +3028,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, out: if (target) mutex_unlock(&target->i_mutex); + dput(new_dentry); if (!error) if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) d_move(old_dentry,new_dentry);