Re: [PATCH V3 0/9] vfio virtual address update

From: Cornelia Huck <cohuck@redhat.com>
To: Alex Williamson <alex.williamson@redhat.com>
Cc: Steven Sistare <steven.sistare@oracle.com>,
	kvm@vger.kernel.org, Kirti Wankhede <kwankhede@nvidia.com>
Subject: Re: [PATCH V3 0/9] vfio virtual address update
Date: Tue, 2 Feb 2021 08:54:37 +0100	[thread overview]
Message-ID: <20210202085437.51ad8423.cohuck@redhat.com> (raw)
In-Reply-To: <20210201132318.36795807@omen.home.shazbot.org>

On Mon, 1 Feb 2021 13:23:18 -0700
Alex Williamson <alex.williamson@redhat.com> wrote:

> On Sat, 30 Jan 2021 11:54:03 -0500
> Steven Sistare <steven.sistare@oracle.com> wrote:
> 
> > On 1/29/2021 4:55 PM, Alex Williamson wrote:  
> > > On Fri, 29 Jan 2021 08:54:03 -0800
> > > Steve Sistare <steven.sistare@oracle.com> wrote:
> > >     
> > >> Add interfaces that allow the underlying memory object of an iova range
> > >> to be mapped to a new virtual address in the host process:
> > >>
> > >>   - VFIO_DMA_UNMAP_FLAG_VADDR for VFIO_IOMMU_UNMAP_DMA
> > >>   - VFIO_DMA_MAP_FLAG_VADDR flag for VFIO_IOMMU_MAP_DMA
> > >>   - VFIO_UPDATE_VADDR for VFIO_CHECK_EXTENSION
> > >>   - VFIO_DMA_UNMAP_FLAG_ALL for VFIO_IOMMU_UNMAP_DMA
> > >>   - VFIO_UNMAP_ALL for VFIO_CHECK_EXTENSION
> > >>
> > >> Unmap-vaddr invalidates the host virtual address in an iova range and blocks
> > >> vfio translation of host virtual addresses, but DMA to already-mapped pages
> > >> continues.  Map-vaddr updates the base VA and resumes translation.  The
> > >> implementation supports iommu type1 and mediated devices.  Unmap-all allows
> > >> all ranges to be unmapped or invalidated in a single ioctl, which simplifies
> > >> userland code.
> > >>
> > >> This functionality is necessary for live update, in which a host process
> > >> such as qemu exec's an updated version of itself, while preserving its
> > >> guest and vfio devices.  The process blocks vfio VA translation, exec's
> > >> its new self, mmap's the memory object(s) underlying vfio object, updates
> > >> the VA, and unblocks translation.  For a working example that uses these
> > >> new interfaces, see the QEMU patch series "[PATCH V2] Live Update" at
> > >> https://lore.kernel.org/qemu-devel/1609861330-129855-1-git-send-email-steven.sistare@oracle.com
> > >>
> > >> Patches 1-3 define and implement the flag to unmap all ranges.
> > >> Patches 4-6 define and implement the flags to update vaddr.
> > >> Patches 7-9 add blocking to complete the implementation.    
> > > 
> > > Hi Steve,
> > > 
> > > It looks pretty good to me, but I have some nit-picky comments that
> > > I'll follow-up with on the individual patches.  However, I've made the
> > > changes I suggest in a branch that you can find here:
> > > 
> > > git://github.com/awilliam/linux-vfio.git vaddr-v3
> > > 
> > > If the changes look ok, just send me an ack, I don't want to attribute
> > > something to you that you don't approve of.  Thanks,    
> > 
> > All changes look good, thanks!  
> > Do you need anything more from me on this patch series?  
> 
> Here's a new branch:
> 
> git://github.com/awilliam/linux-vfio.git vaddr-v3.1
> 
> Extent of the changes are s/may not/cannot/ on patches 1 & 4 and
> addition of Connie's R-b for all (rebased to rc6).  If there are any
> final comments, speak now.  Thanks,
> 
> Alex

LGTM.