ovl: during copy up, switch to mounter's creds early
Now, we have the notion that copy up of a file is done with the creds
of mounter of overlay filesystem (as opposed to task). Right now before
we switch creds, we do some vfs_getattr() operations in the context of
task and that itself can fail. We should do that getattr() using the
creds of mounter instead.
So this patch switches to mounter's creds early during copy up process so
that even vfs_getattr() is done with mounter's creds.
Do not call revert_creds() unless we have already called
ovl_override_creds(). [Reported by Arnd Bergmann]
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c
index abadbc3..796d06f 100644
--- a/fs/overlayfs/copy_up.c
+++ b/fs/overlayfs/copy_up.c
@@ -348,7 +348,6 @@
struct path parentpath;
struct dentry *upperdir;
struct dentry *upperdentry;
- const struct cred *old_cred;
char *link = NULL;
if (WARN_ON(!workdir))
@@ -369,8 +368,6 @@
return PTR_ERR(link);
}
- old_cred = ovl_override_creds(dentry->d_sb);
-
err = -EIO;
if (lock_rename(workdir, upperdir) != NULL) {
pr_err("overlayfs: failed to lock workdir+upperdir\n");
@@ -391,7 +388,6 @@
}
out_unlock:
unlock_rename(workdir, upperdir);
- revert_creds(old_cred);
if (link)
free_page((unsigned long) link);
@@ -401,9 +397,9 @@
int ovl_copy_up(struct dentry *dentry)
{
- int err;
+ int err = 0;
+ const struct cred *old_cred = ovl_override_creds(dentry->d_sb);
- err = 0;
while (!err) {
struct dentry *next;
struct dentry *parent;
@@ -435,6 +431,7 @@
dput(parent);
dput(next);
}
+ revert_creds(old_cred);
return err;
}