JFS: Remove redundant xattr permission checking

The vfs handles most permissions for setting and retrieving xattrs.
This patch removes a redundant and wrong check so that it won't override
the correct behavior which is being fixed in the vfs.

Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>

diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c
index 4c7985e..b753ba2 100644 (file)
--- a/fs/jfs/xattr.c
+++ b/fs/jfs/xattr.c
@@ -756,6 +756,11 @@ static int can_set_system_xattr(struct inode *inode, const char *name,
        return -EOPNOTSUPP;
 }
 
        return -EOPNOTSUPP;
 }
 

+/*
+ * Most of the permission checking is done by xattr_permission in the vfs.
+ * The local file system is responsible for handling the system.* namespace.
+ * We also need to verify that this is a namespace that we recognize.
+ */

 static int can_set_xattr(struct inode *inode, const char *name,
                         const void *value, size_t value_len)
 {
 static int can_set_xattr(struct inode *inode, const char *name,
                         const void *value, size_t value_len)
 {


@@ -771,10 +776,6 @@ static int can_set_xattr(struct inode *inode, const char *name,
            strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN))
                return -EOPNOTSUPP;
 
            strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN))
                return -EOPNOTSUPP;
 

-       if (!S_ISREG(inode->i_mode) &&
-           (!S_ISDIR(inode->i_mode) || inode->i_mode &S_ISVTX))
-               return -EPERM;
-

        return 0;
 }
 
        return 0;
 }