Re: [ 105/175] CIFS: Respect negotiated MaxMpxCount

From: Jeff Layton
Date: Sun Apr 01 2012 - 09:55:07 EST


On Sun, 1 Apr 2012 02:26:58 -0500
Steve French <smfrench@xxxxxxxxx> wrote:

> On Sat, Mar 31, 2012 at 7:18 PM, Jeff Layton <jlayton@xxxxxxxxx> wrote:
> > On Fri, 30 Mar 2012 12:50:10 -0700
> > Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> >
> >> 3.3-stable review patch. ÂIf anyone has any objections, please let me know.
> >>
> >
> > I'm not sure it's a good idea to put this patch into any stable release
> > just yet. I think it's a correct and necessary fix, but it has some
> > potential to cause regressions too (or uncover other preexisting
> > problems with this code). Might it be best to wait until we have some
> > more experience with this before we push it into stable?
>
> Delaying it a little is probably ok, but weigh that against the
> frequency/severity
> of the problem that Pavel's fix addresses: ie sequential writes to some
> versions of Windows 7 and Windows Vista can fail partway
> through large file copy (which also causes those servers
> to become unresponsive to subsequent connection requests
> as well) so the fix does address a hot problem.
>
>
>

I'm not disputing whether this patch is correct. It's clearly a bug
that cifs.ko has never respected this value, and it's quite problematic
with certain servers.

The problem is that we don't have any confirmation that:

a) this fixes any of the problems that we think it will

b) this doesn't introduce any regressions

The request slot allocation code is quite fiddly and fragile, and I
think the potential for "b" is somewhat high. I'd feel more comfortable
if we waited until we have more experience with this patch before
merging it into stable.

--
Jeff Layton <jlayton@xxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/