On Wed, 2016-11-09 at 15:23 -0500, J. Bruce Fields wrote:
On Wed, Nov 09, 2016 at 02:47:24PM -0500, J. Bruce Fields wrote:
For now I wish we could just like to continue assuming the workqueue
processes only one item at a time. Do we have that now, or do we need
to switch to (looking at workqueue.h...) alloc_ordered workqueue()?
Oh, wait, I missed the _ordered_ in:
#define create_singlethread_workqueue(name) \
alloc_ordered_workqueue("%s", __WQ_LEGACY | WQ_MEM_RECLAIM, name)
So our existing create_singlethread_workqueue is fine--but we probably
don't need those flags:
--b.
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 211dc2aed8e1..eb78109d666c 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -1061,7 +1061,7 @@ static const struct rpc_call_ops nfsd4_cb_ops = {
int nfsd4_create_callback_queue(void)
{
- callback_wq = create_singlethread_workqueue("nfsd4_callbacks");
+ callback_wq = alloc_ordered_workqueue("nfsd4_callbacks", 0);
if (!callback_wq)
return -ENOMEM;
return 0;
That looks good to me. Eventually we could better serialize the handling
of the callback code and move to an unordered workqueue, but for now I
think Bruce is right that we have to keep it.
Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
