Hello,
Coverity has drawn my attention to a strange code in generic_readlink().
We do there:
const char *link = inode->i_link;
if (!link) {
link = inode->i_op->follow_link(dentry, &cookie);
if (IS_ERR(link))
return PTR_ERR(link);
}
res = readlink_copy(buffer, buflen, link);
if (inode->i_op->put_link)
inode->i_op->put_link(inode, cookie);
However when inode->i_link is set and the inode also has put_link method,
we would pass uninitialized cookie to put_link.
Now if I understand the code right, each inode is supposed to use either
i_link or cookies so this shouldn't really happen but IMHO it's
unnecessarily fragile. I'd prefer to call put_link only if follow_link was
called in generic_readlink(). Thoughts?
Honza
--
Jan Kara <jack@xxxxxxx>
SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
