On Sun, Mar 30, 2003 at 11:58:09PM +0900, YOSHIFUJI Hideaki wrote:
> > And, patch does not seem optimal. I'd take a look at very soon.
>
> Here's our patch based on our fix in August, 2001.
> Question: should we use spin_lock_bh() instead of spin_lock()?
Because everywhere else in the file {read,write}_lock_bh() is used
instead of {read,write}_lock(), so I'm assuming that _bh is required
but I really don't know why.
Anyway I have some critics over your patch:
- locking inside ipv6_add_addr() is simpler and more linear but
semantically wrong because you're unable to tell the user why his
"ip addr add" failed. E.g. you answer ENOBUFS instead of EEXIST.
- your ipv6_chk_same_addr() does a useless check for (dev != NULL)
> +static
> +int ipv6_chk_same_addr(const struct in6_addr *addr, struct net_device
*dev)
> +{
> + struct inet6_ifaddr * ifp;
> + u8 hash = ipv6_addr_hash(addr);
> +
> + read_lock_bh(&addrconf_hash_lock);
> + for(ifp = inet6_addr_lst[hash]; ifp; ifp=ifp->lst_next) {
> + if (ipv6_addr_cmp(&ifp->addr, addr) == 0) {
> + if (dev != NULL && ifp->idev->dev == dev)
> break;
> }
your never "break" if dev == NULL, so you could return 0 before
even acquiring the lock.
Regards,
Simone
--
Simone Piunno -- http://members.ferrara.linux.it/pioppo
.------- Adde parvum parvo magnus acervus erit -------.
Ferrara Linux Users Group - http://www.ferrara.linux.it
Deep Space 6, IPv6 on Linux - http://www.deepspace6.net
GNU Mailman, Mailing List Manager - http://www.list.org
`-------------------------------------------------------'
|
