linux-kernel - Re: [PATCH v2 1/1] vfio/type1: Respect IOMMU reserved regions in vfio_test_domain

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] [day] [month] [year] [list]

Date:   Mon, 09 Jan 2023 10:00:33 +0100
From:   Niklas Schnelle <schnelle@...ux.ibm.com>
To:     Alex Williamson <alex.williamson@...hat.com>
Cc:     Cornelia Huck <cohuck@...hat.com>, Jason Gunthorpe <jgg@...pe.ca>,
        kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-s390@...r.kernel.org,
        Matthew Rosato <mjrosato@...ux.ibm.com>,
        Pierre Morel <pmorel@...ux.ibm.com>,
        Christian Bornträger 
        <borntraeger@...ux.ibm.com>
Subject: Re: [PATCH v2 1/1] vfio/type1: Respect IOMMU reserved regions in
 vfio_test_domain_fgsp()

On Fri, 2023-01-06 at 10:24 -0700, Alex Williamson wrote:
> On Wed,  4 Jan 2023 16:42:02 +0100
> Niklas Schnelle <schnelle@...ux.ibm.com> wrote:
> 
> > Since commit cbf7827bc5dc ("iommu/s390: Fix potential s390_domain
> > aperture shrinking") the s390 IOMMU driver uses reserved regions for the
> 
> Are you asking for this in v6.2?  Seems like the above was introduced
> in v6.2 and I can't tell if this is sufficiently prevalent that we need
> a fix in the same release.

If possible yes I'd hope for this to go into v6.2 so we don't break ISM
pass-through. Support for ISM pass-through has only been available
since v6.0 where Matt added the interpretation support but it is one of
the most useful pass-through uses at the moment since the ISM device
uses long lived DMA mappings and as such is pretty much unaffected by
the performance impact of our virtualized IOMMUs. Together with SMD-D
this allows high performance communication of VMs with other VMs or
LPARs. Now we don't have this problem in distribution kernels that have
Matt's patches as backports but lack my newer IOMMU changes and few if
any of our customers run upstream but still of course I'd prefer not to
have known broken upstream releases.

Thanks,
Niklas