On 2019/8/6 下午8:04, Jason Gunthorpe wrote:

On Mon, Aug 05, 2019 at 12:20:45PM +0800, Jason Wang wrote:

quoted

On 2019/8/2 下午8:46, Jason Gunthorpe wrote:

quoted

On Fri, Aug 02, 2019 at 05:40:07PM +0800, Jason Wang wrote:

quoted

quoted

This must be a proper barrier, like a spinlock, mutex, or
synchronize_rcu.

I start with synchronize_rcu() but both you and Michael raise some
concern.

I've also idly wondered if calling synchronize_rcu() under the various
mm locks is a deadlock situation.

Maybe, that's why I suggest to use vhost_work_flush() which is much
lightweight can can achieve the same function. It can guarantee all previous
work has been processed after vhost_work_flush() return.

If things are already running in a work, then yes, you can piggyback
on the existing spinlocks inside the workqueue and be Ok

However, if that work is doing any copy_from_user, then the flush
becomes dependent on swap and it won't work again...

Yes it do copy_from_user(), so we can't do this.

quoted

quoted

quoted

1) spinlock: add lots of overhead on datapath, this leads 0 performance
improvement.

I think the topic here is correctness not performance improvement>

quoted

But the whole series is to speed up vhost.

So? Starting with a whole bunch of crazy, possibly broken, locking and
claiming a performance win is not reasonable.

Yes, I admit this patch is tricky, I'm not going to push this. Will post 
a V3.

quoted

Spinlock is correct but make the whole series meaningless consider it won't
bring any performance improvement.

You can't invent a faster spinlock by opencoding some wild
scheme. There is nothing special about the usage here, it needs a
blocking lock, plain and simple.

Jason

Will post V3. Let's see if you are happy with that version.

Thanks


_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization