linux-kernel - Re: [PATCH] nvme-pci: Prevent mmio reads if pci channel offline

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190222215938.GG10237@localhost.localdomain>
Date:   Fri, 22 Feb 2019 14:59:38 -0700
From:   Keith Busch <keith.busch@...el.com>
To:     Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     "Derrick, Jonathan" <jonathan.derrick@...el.com>,
        "linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>,
        Jens Axboe <axboe@...com>, Christoph Hellwig <hch@....de>,
        Sagi Grimberg <sagi@...mberg.me>,
        Alex Gagniuc <alex_gagniuc@...lteam.com>,
        Linux List Kernel Mailing <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] nvme-pci: Prevent mmio reads if pci channel offline

On Fri, Feb 22, 2019 at 01:28:42PM -0800, Linus Torvalds wrote:
> On Thu, Feb 21, 2019 at 5:07 PM Jon Derrick <jonathan.derrick@...el.com> wrote:
> >
> > Some platforms don't seem to easily tolerate non-posted mmio reads on
> > lost (hot removed) devices. This has been noted in previous
> > modifications to other layers where an mmio read to a lost device could
> > cause an undesired firmware intervention [1][2].
> 
> This is broken, and whatever platform that requires this is broken.
> 
> This has absolutely nothing to do with nvme, and should not be handled
> by a driver.
> 
> The platform code should be fixed.

This is, of course, the correct answer. We just find platform firmware
uncooperative, so we see these attempts to avoid them.

I really don't like this driver piecemeal approach if we're going to
quirk around these platforms, though. I'd rather see the invalidated
address ranges remapped to a fault handler fixup exception once and be
done with it.

Or we can say you don't get to use this feature if you bought that
hardware.