Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 06 Sep 2016 14:29:40 -0700
From:      John Baldwin <jhb@freebsd.org>
To:        src-committers@freebsd.org
Cc:        svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   Re: svn commit: r305502 - in head: share/man/man9 sys/amd64/vmm/io sys/dev/pci
Message-ID:  <1571206.j1ezrSnmG7@ralph.baldwin.cx>
In-Reply-To: <201609062115.u86LFZaU030916@repo.freebsd.org>
References:  <201609062115.u86LFZaU030916@repo.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday, September 06, 2016 09:15:35 PM John Baldwin wrote:
> Author: jhb
> Date: Tue Sep  6 21:15:35 2016
> New Revision: 305502
> URL: https://svnweb.freebsd.org/changeset/base/305502
> 
> Log:
>   Reset PCI pass through devices via PCI-e FLR during VM start and end.
>   
>   Add routines to trigger a function level reset (FLR) of a PCI-express
>   device via the PCI-express device control register.  This also includes
>   support routines to wait for pending transactions to complete as well
>   as calculating the maximum completion timeout permitted by a device.
>   
>   Change the ppt(4) driver to reset pass through devices before attaching
>   to a VM during startup and before detaching from a VM during shutdown.

With the recent fixes, it should (in theory) be safe to do something like:

kldload vmm
iovctl -C -f /path/to/iovctl.conf
devctl set driver ppt <location>
vmrun.sh -p <location> <other args>

And have it work correctly, including if you hotplug or create new VFs 
after an initial VM is running/has run, or you delete a VF, etc.  FLR
should (in theory) make VMs with pass through devices a bit more robust
if they do not cleanly shutdown the device hardware.

I also have some thoughts (still kicking around in my head) about adding
a 'devctl reset' command.  In particular the trickier case is what to do
for attached devices (I think we suspend / resume devices around reset
perhaps).  Also, there are different ways you can trigger a PCI reset
that we should eventually support.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1571206.j1ezrSnmG7>