On Wed, Apr 18, 2012 at 09:17:26AM +0100, David Laight wrote: > > > > > Why not have: > > > > diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c > > index d510353..f9f77a7 100644 > > --- a/net/unix/af_unix.c > > +++ b/net/unix/af_unix.c > > @@ -216,6 +216,9 @@ static int unix_mkname(struct sockaddr_un > > *sunaddr, int len, unsigned *hashp) > > */ > > ((char *)sunaddr)[len] = 0; > > len = strlen(sunaddr->sun_path)+1+sizeof(short); > > + /* No null terminator was found in the path. */ > > + if (len > sizeof(*sunaddr)) > > + return -EINVAL; > > return len; > > That could generate a kernel page fault! > (Depending on what follows (or rather doesn't follow!) sun_path.) > You'd need to use memchr() not strlen(). > > David > Hi, David. What follows is a 0 byte, because it's set that way in the line before strlen. Note that len is tested for sizeof(*sunaddr), and there is a huge comment about that extra byte that was omitted. The whole function is at net/unix/af_unix.c:203. Regards, Cascardo. -- To unsubscribe from this list: send the line "unsubscribe linux-api" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
