> -----Original Message----- > From: linux-cifs-owner@xxxxxxxxxxxxxxx <linux-cifs-owner@xxxxxxxxxxxxxxx> On > Behalf Of Robin P. Blanchard > Sent: Thursday, August 16, 2018 4:51 PM > To: Steve French <smfrench@xxxxxxxxx> > Cc: linux-cifs@xxxxxxxxxxxxxxx > Subject: Re: regression in CIFS(?) between 4.17.14 and 4.18.0 > > Correct. > > Interplay between DFS referrals and vers=2.[01] seems to be where we're at... Curious. What OS and SMB dialect are these DFS servers running? The DFS referral protocol is pretty much identical across dialects, so it must be something in the new mount/connection plumbing in the client that's going flooey. Tom. > On Thu, Aug 16, 2018 at 3:46 PM Steve French <smfrench@xxxxxxxxx> wrote: > > > > and to clarify - DFS referral to Windows 2016 works with 3.0 or later, > > but reboots with 2.0 or 2.1? > > On Thu, Aug 16, 2018 at 3:42 PM Robin P. Blanchard > > <robin.blanchard@xxxxxxxxx> wrote: > > > > > > Summary of regression between 4.17.14 and 4.18.0 > > > > > > pam_mount is/was red herring > > > > > > vers=2.0 and vers=2.1 trigger spontaneous reboots **using DFS pathing** > to > > > - Microsoft Windows Server 2008 R2 Enterprise > > > - Microsoft Windows Server 2016 Datacenter > > > > > > mounting directly to one of the underlying DFS member servers does NOT > > > trigger spontaneous reboot. > > > > > > > > > On Thu, Aug 16, 2018 at 3:21 PM Steve French <smfrench@xxxxxxxxx> > wrote: > > > > > > > > On Thu, Aug 16, 2018 at 3:14 PM Robin P. Blanchard > > > > <robin.blanchard@xxxxxxxxx> wrote: > > > > > > > > > > On Thu, Aug 16, 2018 at 2:49 PM Tom Talpey <ttalpey@xxxxxxxxxxxxx> > wrote: > > > > > > > > > > > > > -----Original Message----- > > > > > > > From: linux-cifs-owner@xxxxxxxxxxxxxxx <linux-cifs- > owner@xxxxxxxxxxxxxxx> On > > > > > > > Behalf Of Robin P. Blanchard > > > > > > > Sent: Thursday, August 16, 2018 3:34 PM > > > > > > > To: Steve French <smfrench@xxxxxxxxx> > > > > > > > Cc: linux-cifs@xxxxxxxxxxxxxxx > > > > > > > Subject: Re: regression in CIFS(?) between 4.17.14 and 4.18.0 > > > > > > > > > > > > > > Ok. "Good" news. > > > > > > > > > > > > > > The issue is specific to vers=2.0. > > > > > > > > > > > > Your Windows Server 2008R2 target supports SMB2.1, is there some > reason you > > > > > > are not using the 2.1 dialect? It is *much* preferred to the baselevel > 2.0, though of > > > > > > course any 3.x is better still. > > > > > > > > > > > > Tom. > > > > > > > > > > Sure. This, however, doesn't explain the regression. > > > > > > > > > > FWIW - just rested: vers=2.1 also triggers spontaneous reboot > > > > > > > > So barring other differences in the server (unexpected return code > > > > that Windows 2008R2 > > > > returns that Samba or Windows 2012 or Windows 2016 or Mac servers or > NetApp > > > > wouldn't return) ... best theory is that only the older two dialects > > > > SMB2.0 and SMB2.1 broke for your scenario, but only if mounted via > pam_mount > > > > not if mounted explicitly from bash? > > > > > > > > Is that correct? > > > > > > > > If anyone else is able to repro this let us know ASAP, and of course any > data > > > > on location of oops would be huge help ... perhaps build kernel with > different > > > > kconfig options for debugging > > > > > > > > -- > > > > Thanks, > > > > > > > > Steve > > > > > > > > -- > > Thanks, > > > > Steve
