qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Alex Williamson <alex.williamson@redhat.com>
Cc: chen.fan.fnst@cn.fujitsu.com, izumi.taku@jp.fujitsu.com,
	Cao jin <caoj.fnst@cn.fujitsu.com>,
	qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [patch v5 07/12] pci: add a pci_function_is_valid callback to check function if valid
Date: Sun, 27 Mar 2016 15:19:10 +0300	[thread overview]
Message-ID: <20160327121910.GA26438@redhat.com> (raw)
In-Reply-To: <20160324165433.4806cba1@t450s.home>

On Thu, Mar 24, 2016 at 04:54:33PM -0600, Alex Williamson wrote:
> On Wed, 23 Mar 2016 18:12:02 +0800
> Cao jin <caoj.fnst@cn.fujitsu.com> wrote:
> 
> > From: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
> > 
> > PCI hotplug requires that function 0 is added last to close the
> > slot.  Since vfio supporting AER, we require that the VM bus
> > contains the same set of devices as the host bus to support AER,
> > we can perform an AER validation test whenever a function 0 in
> > the VM is hot-added.
> > 
> > Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
> > ---
> >  hw/pci/pci.c         | 32 ++++++++++++++++++++++++++++++++
> >  include/hw/pci/pci.h |  1 +
> >  2 files changed, 33 insertions(+)
> 
> 
> This would of course need Michael's ack.
> 
> Michael, we can't do this in vfio-pci code because we can't guarantee
> that function 0 is a vfio-pci device.  That would allow users to bypass
> our configuration requirements by putting any random non-vfio-pci
> device at function 0 of a hot-added multifunction device.

I got that. What I don't understand is how is the non hotplug
variant handled, given that is_valid_func is not called
in that case.

Also, why does it scan all devices on bus, not all functions
of the device?

> > 
> > diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> > index e67664d..2a5291a 100644
> > --- a/hw/pci/pci.c
> > +++ b/hw/pci/pci.c
> > @@ -1840,6 +1840,22 @@ PCIDevice *pci_find_device(PCIBus *bus, int bus_num, uint8_t devfn)
> >      return bus->devices[devfn];
> >  }
> >  
> > +static void pci_function_is_valid(PCIBus *bus, PCIDevice *d, void *opaque)
> > +{
> > +    PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(d);
> > +    Error **errp = opaque;
> > +
> > +    if (*errp) {
> > +        return;
> > +    }
> > +
> > +    if (!pc->is_valid_func) {
> > +        return;
> > +    }
> > +
> > +    pc->is_valid_func(d, errp);
> > +}
> > +
> >  static void pci_qdev_realize(DeviceState *qdev, Error **errp)
> >  {
> >      PCIDevice *pci_dev = (PCIDevice *)qdev;
> > @@ -1882,6 +1898,22 @@ static void pci_qdev_realize(DeviceState *qdev, Error **errp)
> >          pci_qdev_unrealize(DEVICE(pci_dev), NULL);
> >          return;
> >      }
> > +
> > +    /*
> > +     *  If the function number is 0, indicate the closure of the slot.
> > +     *  then we get the chance to check all functions on same device
> > +     *  if valid.
> > +     */
> > +    if (DEVICE(pci_dev)->hotplugged &&
> > +        pci_get_function_0(pci_dev) == pci_dev) {
> > +        pci_for_each_device(bus, pci_bus_num(bus),
> > +                            pci_function_is_valid, &local_err);
> > +        if (local_err) {
> > +            error_propagate(errp, local_err);
> > +            pci_qdev_unrealize(DEVICE(pci_dev), NULL);
> > +            return;
> > +        }
> > +    }
> >  }
> >  
> >  static void pci_default_realize(PCIDevice *dev, Error **errp)
> > diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> > index 0be07c8..4a2f7d4 100644
> > --- a/include/hw/pci/pci.h
> > +++ b/include/hw/pci/pci.h
> > @@ -190,6 +190,7 @@ typedef struct PCIDeviceClass {
> >  
> >      void (*realize)(PCIDevice *dev, Error **errp);
> >      int (*init)(PCIDevice *dev);/* TODO convert to realize() and remove */
> > +    void (*is_valid_func)(PCIDevice *dev, Error **errp);
> >      PCIUnregisterFunc *exit;
> >      PCIConfigReadFunc *config_read;
> >      PCIConfigWriteFunc *config_write;

  reply	other threads:[~2016-03-27 12:19 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-23 10:11 [Qemu-devel] [patch v5 00/12] vfio-pci: pass the aer error to guest, part2 Cao jin
2016-03-23 10:11 ` [Qemu-devel] [patch v5 01/12] vfio: extract vfio_get_hot_reset_info as a single function Cao jin
2016-03-23 10:11 ` [Qemu-devel] [patch v5 02/12] vfio: squeeze out vfio_pci_do_hot_reset for support bus reset Cao jin
2016-03-23 10:11 ` [Qemu-devel] [patch v5 03/12] vfio: add pcie extended capability support Cao jin
2016-03-23 10:11 ` [Qemu-devel] [patch v5 04/12] vfio: add aer support for vfio device Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 05/12] vfio: refine function vfio_pci_host_match Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 06/12] vfio: add check host bus reset is support or not Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 07/12] pci: add a pci_function_is_valid callback to check function if valid Cao jin
2016-03-24 22:54   ` Alex Williamson
2016-03-27 12:19     ` Michael S. Tsirkin [this message]
2016-03-31  7:23       ` Chen Fan
2016-03-27 12:52   ` Michael S. Tsirkin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 08/12] vfio: add check aer functionality for hotplug device Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 09/12] vfio: vote the function 0 to do host bus reset when aer occurred Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 10/12] vfio-pci: pass the aer error to guest Cao jin
2016-03-23 10:12 ` [Qemu-devel] [patch v5 11/12] vfio: device may stuck in D3 when doing aer recovery Cao jin
2016-03-24 22:54   ` Alex Williamson
2016-03-25  1:38     ` Chen Fan
2016-03-25  2:22       ` Alex Williamson
2016-03-31  6:55         ` Chen Fan
2016-03-31 15:44           ` Alex Williamson
2016-04-01  1:40             ` Chen Fan
2016-04-01  1:53               ` Chen Fan
2016-03-23 10:12 ` [Qemu-devel] [patch v5 12/12] vfio: add 'aer' property to expose aercap Cao jin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160327121910.GA26438@redhat.com \
    --to=mst@redhat.com \
    --cc=alex.williamson@redhat.com \
    --cc=caoj.fnst@cn.fujitsu.com \
    --cc=chen.fan.fnst@cn.fujitsu.com \
    --cc=izumi.taku@jp.fujitsu.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).