|
Hi Martin,
Thanks a lot for this fix.
I am awfully sorry for leaving that stupid locking bugs there.
I will change things for LBLC code, will release ipvs-1.0.0-2.2.17 soon.
Thanks,
Wensong
On Wed, 15 Nov 2000, Martin Hamilton wrote:
> Hi folks - I wasn't able to get the LBLC code to compile (as a module,
> FWIW) against a 2.2.17 kernel patched with ReiserFS 3.5.27 on a RedHat
> 6.2 box. Making the changes below lets it compile, but as a newbie
> kernel hacker I'm not sure whether this is horribly broken :-)
>
> Cheers,
>
> Martin
>
>
> --- linux-2.2.17+ipvs-1.0.0beta1+reiserfs-3.5.27/net/ipv4/ip_vs_lblc.c.FCS
> Wed Nov 15 11:53:09 2000
> +++ linux-2.2.17+ipvs-1.0.0beta1+reiserfs-3.5.27/net/ipv4/ip_vs_lblc.c Wed
> Nov
> 15 11:53:33 2000
> @@ -134,10 +134,10 @@
> */
> hash = ip_vs_lblc_hashkey(en->addr);
>
> - write_lock(tbl->lock);
> + write_lock(&tbl->lock);
> list_add(&en->list, &tbl->bucket[hash]);
> atomic_inc(&tbl->entries);
> - write_unlock(tbl->lock);
> + write_unlock(&tbl->lock);
>
> return 1;
> }
> @@ -182,19 +182,19 @@
>
> hash = ip_vs_lblc_hashkey(addr);
>
> - read_lock(tbl->lock);
> + read_lock(&tbl->lock);
>
> l = &tbl->bucket[hash];
> for (e=l->next; e!=l; e=e->next) {
> en = list_entry(e, struct ip_vs_lblc_entry, list);
> if (en->addr == addr) {
> /* HIT */
> - read_unlock(tbl->lock);
> + read_unlock(&tbl->lock);
> return en;
> }
> }
>
> - read_unlock(tbl->lock);
> + read_unlock(&tbl->lock);
>
> return NULL;
> }
>
>
>
>
|
