Re: [SMB3][PATCHes] parallelizing decryption of large read responses

[ronnie sahlberg <ronniesahlberg@gmail.com>]

Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>


We have now a decently large number of new mount options so we need a
patch to the manpage.
That said, we should also make sure that we try to set reasonable
values by default,
or even longer term, remove the options with heuristics.

(Very very few people read the manpage or ever use any of these mount options
so our default should be "as close to optimal as possible" and using a
mount option
should be the rare exception where our heuristics just went wrong.)

On Tue, Sep 10, 2019 at 12:21 AM Steve French <smfrench@gmail.com> wrote:
>
> Had a minor typo in patch 2 - fixed in attached
>
> On Sun, Sep 8, 2019 at 11:31 PM Steve French <smfrench@gmail.com> wrote:
> >
> > I am seeing very good performance benefit from offload of decryption
> > of encrypted SMB3 read responses to a pool of worker threads
> > (optionally).  See attached patches.
> >
> > I plan to add another patch to only offload when number of requests in
> > flight is > 1 (since there is no point offloading and doing a thread
> > switch if no other responses would overlap in the cifsd thread reading
> > from the socket).
> >
> > --
> > Thanks,
> >
> > Steve
>
>
>
> --
> Thanks,
>
> Steve