[linux-cifs-client] use non-zero vcnumbers patch

[Steve French]

On Wed, Feb 11, 2009 at 6:06 AM, Leonardo Chiquitto <
leonardo.lists at gmail.com> wrote:

> >>> +     if (max_vcs < 2)
> >>> +             max_vcs = 0xFFFF;
> >>> +
> >>
> >> A comment about the above situation seems warranted since it's basically
> >> a workaround for buggy servers.
> >
> > Yes - Windows set(s) max vcs to 1 when it means unlimited so I agree,
> > this should be commented.
> >
> >> So what happens at reconnect? Shouldn't we "reset" all the vcnums?
> >> Shouldn't the first session setup on the reconnected socket use vcnum
> >> 0? I seem to recall reading that vcnum of 0 has some significance for
> >> the server...
> >>
> >> Other than that, the patch looks reasonable to me.
> > I forgot to add apart that resets vcnumbers, I had originally intended
> > to do that, but I am not certain that it is necessary, although
> > without it, we increase the possiblility of vc number collisions.   I
> > thought Windows begins with vcnumber 0, but don't remember (that is
> > probably safest, to follow Windows behavior here, although it is not
> > always easy to get certain Windows clients to send more than one vc
> > from the same client - as you could e.g. with terminal server/
> >
> > I also remember the old situation that Chris H. mentions about vc
> > numbers and NATs and the nasty behavior change it required.
>
> Hi Steve,
>
>  Do you have plans to push this patch upstream in the near future?
>
>
Yes - I plan to push upstream, but wanted to fix the reconnection case
(so on reconnect we would send vc number 0, on the first smb session, and
non-zero unique vc numbers on the subsequent smb sessions).


-- 
Thanks,

Steve
-------------- next part --------------
HTML attachment scrubbed and removed