diff options
| author | Al Viro <viro@zeniv.linux.org.uk> | 2016-04-11 01:50:48 +0300 |
|---|---|---|
| committer | Greg Kroah-Hartman <gregkh@linuxfoundation.org> | 2016-10-22 13:26:55 +0300 |
| commit | be90a09bd915e417feae76d9a72c2a37dc5cf1b7 (patch) | |
| tree | 4489674dd10b0a83665f169959fbf715b4da0b81 /crypto | |
| parent | 5c16520bdcd41740abcadb58133637faecf713c9 (diff) | |
| download | linux-be90a09bd915e417feae76d9a72c2a37dc5cf1b7.tar.xz | |
reiserfs: switch to generic_{get,set,remove}xattr()
commit 79a628d14ec7ee9adfdc3ce04343d5ff7ec20c18 upstream.
reiserfs_xattr_[sg]et() will fail with -EOPNOTSUPP for V1 inodes anyway,
and all reiserfs instances of ->[sg]et() call it and so does ->set_acl().
Checks for name length in the instances had been bogus; they should've
been "bugger off if it's _exactly_ the prefix" (as generic would
do on its own) and not "bugger off if it's shorter than the prefix" -
that can't happen.
xattr_full_name() is needed to adjust for the fact that generic instances
will skip the prefix in the name passed to ->[gs]et(); reiserfs homegrown
analogues didn't.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[jeffm: Backported to v4.4: adjust context]
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'crypto')
0 files changed, 0 insertions, 0 deletions