Re: [PATCH 0/2] Postcopy migration and vhost-user errors

[Peter Xu]

On Tue, Jul 23, 2024 at 10:33:58AM +0530, Prasad Pandit wrote:
> On Sun, 21 Jul 2024 at 01:11, Michael S. Tsirkin <mst@redhat.com> wrote:
> > > > > On Wed, Jul 17, 2024 at 04:55:52AM -0400, Michael S. Tsirkin wrote:
> > > > > > I just want to understand how we managed to have two threads
> > > > > > talking in parallel. BQL is normally enough, which path
> > > > > > manages to invoke vhost-user with BQL not taken?
> > > > > > Just check BQL taken on each vhost user invocation and
> > > > > > you will figure it out.
> > > > >
> > > > OK, so postcopy_ram_ things run without the BQL?
> > >
> > > There are a lot of postcopy_ram_* functions, I didn't check all of them 
> > > but
> > > I think it's true in this case.  Thanks.
> > >
> > So pls document this in the commit log.
> 
> * ie. IIUC, if we take BQL in postcop_ram_* functions, we don't need
> this 'vhost_user_request_reply_lock patch'? I'll check the
> postcopy_ram_* functions to see what's happening there.

Go ahead, Prasad.  But just to mention postcopy stuff doesn't take BQL may
be because there's some good reason. So if you're not confident on some
changes, you can share a patch before testing in an involved environment.

Personally, I'd avoid using BQL as long as it can be replaced by a finer
grained lock.  Otherwise the lock semantics can be more unclear, and we'll
never get rid of BQL.  IOW, I'd prefer a smaller lock in general if
possible to avoid BQL, but we can see how the patch look like when it's
ready.

Thanks,

-- 
Peter Xu