Date: Tue, 29 Aug 2017 13:02:11 -0300
From: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
To: Tejun Heo <tj@kernel.org>
Cc: David Ahern <dsahern@gmail.com>, Christoph Hellwig <hch@lst.de>,
        linux-ide@vger.kernel.org, LKML <linux-kernel@vger.kernel.org>,
        Robert Elliott <elliott@hpe.com>
Subject: Re: boot failure with 4.13.0-rc6 due to ATA errors
Message-ID: <20170829160211.lfyo7jmvhdatay72@khazad-dum.debian.net>
References: <3117ae58-d432-101e-3f0b-68d72fdee28b@gmail.com>
 <20170828195916.GA491396@devbig577.frc2.facebook.com>
 <37033f6d-2ad0-7817-10b5-cbd7ff565624@gmail.com>
 <20170828212225.GB491396@devbig577.frc2.facebook.com>
 <20170829124206.GA26738@lst.de>
 <a5caf398-3ae3-4585-23d2-b2afbbc28d7d@gmail.com>
 <20170829153037.GN491396@devbig577.frc2.facebook.com>
 <20170829155102.vgbzyh7fhjrurdhl@khazad-dum.debian.net>
 <20170829155505.GR491396@devbig577.frc2.facebook.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20170829155505.GR491396@devbig577.frc2.facebook.com>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1491
Lines: 33

On Tue, 29 Aug 2017, Tejun Heo wrote:
> On Tue, Aug 29, 2017 at 12:51:02PM -0300, Henrique de Moraes Holschuh wrote:
> > > > > ATA-8 and later mirrors the TRUSTED COMPUTING SUPPORTED bit in word 48 of
> > > > > the IDENTIFY DEVICE data.  Check this before issuing a READ LOG PAGE
> > > > > command to avoid issues with buggy devices.  The only downside is that
> > > > > we can't support Security Send / Receive for a device with an older
> > > > > revision due to the conflicting use of this field in earlier
> > > > > specifications.
> > > 
> > > Christoph, I'm gonna revert the horkage patch and apply this one.  If
> > > you can think of a better way to do this, please let me know.
> > 
> > The one thing that comes to mind would be an additional patch to allow
> > people with ATA-7 to bypass the identify device data, and rely just on
> > the read log page check, based on a kernel command line parameter.
> > 
> > If we get enough sucessful reports to make it worth it, an whitelist
> > could be added...
> 
> If the ones we miss are the ones based on old revisions, does it
> matter?  If we have to, we can just whitelist those devices and I

Well, only if the opal functionality would actually be worth something
on those old revisions I guess.  It is certainly possible that it would
be better off disabled.

> don't expect there to be many.

Indeed, but we will not get much on the way of people testing this if we
don't give them a kernel parameter ;-)

-- 
  Henrique Holschuh