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 v2 06/11] pci: add a is_valid_func callback to check device if complete
Date: Wed, 9 Mar 2016 19:14:49 +0200 [thread overview]
Message-ID: <20160309171449.GA8144@redhat.com> (raw)
In-Reply-To: <20160309095031.7635e897@t450s.home>
On Wed, Mar 09, 2016 at 09:50:31AM -0700, Alex Williamson wrote:
> On Wed, 9 Mar 2016 18:22:24 +0200
> "Michael S. Tsirkin" <mst@redhat.com> wrote:
>
> > On Mon, Mar 07, 2016 at 11:22:59AM +0800, Cao jin wrote:
> > > From: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
> > >
> > > Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
> >
> > So if you create a mess, you discover it when
> > you later add function 0.
> > Why not call this when function is added?
> > O(N^2) on # of functions, but that # is up to 256 so maybe
> > that is not too bad.
>
> Because the configuration isn't valid until the slot is closed. Take a
> dual port NIC example again, after we add the first function, the
> configuration is invalid because we have an AER indicated device that
> can't do a bus reset because the second function, which may be in a
> separate IOMMU group, hasn't been added yet. Therefore we can only
> check the configuration when the slot is complete. Thanks,
>
> Alex
I see. The name mislead me. So what you want to do is validate a
multi-function device, not the bus.
> > > ---
> > > hw/pci/pci.c | 39 +++++++++++++++++++++++++++++++++++++++
> > > include/hw/pci/pci.h | 1 +
> > > 2 files changed, 40 insertions(+)
> > >
> > > diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> > > index d940f79..72650c5 100644
> > > --- a/hw/pci/pci.c
> > > +++ b/hw/pci/pci.c
> > > @@ -1836,6 +1836,31 @@ PCIDevice *pci_find_device(PCIBus *bus, int bus_num, uint8_t devfn)
> > > return bus->devices[devfn];
> > > }
> > >
> > > +static void pci_bus_check_device(PCIDevice *pdev, Error **errp)
Is not it true that what you are really after is validating
functions of the given device?
Pls rename this pci_check_valid_functions or something
like this, and change it to only scan functions of the device,
not all devices on the bus.
> > > +{
> > > + PCIBus *bus = pdev->bus;
> > > + PCIDeviceClass *pc;
> > > + int i;
> > > + Error *local_err = NULL;
> > > +
> > > + for (i = 0; i < ARRAY_SIZE(bus->devices); ++i) {
> > > + if (!bus->devices[i]) {
> > > + continue;
> > > + }
> > > +
> > > + pc = PCI_DEVICE_GET_CLASS(bus->devices[i]);
> > > + if (!pc->is_valid_func) {
> > > + continue;
> > > + }
> > > +
> > > + pc->is_valid_func(bus->devices[i], &local_err);
> > > + if (local_err) {
> > > + error_propagate(errp, local_err);
> > > + return;
> > > + }
> > > + }
> > > +}
> > > +
> > > static void pci_qdev_realize(DeviceState *qdev, Error **errp)
> > > {
> > > PCIDevice *pci_dev = (PCIDevice *)qdev;
> > > @@ -1878,6 +1903,20 @@ static void pci_qdev_realize(DeviceState *qdev, Error **errp)
> > > pci_qdev_unrealize(DEVICE(pci_dev), NULL);
> > > return;
> > > }
> > > +
> > > + /*
> > > + * If the function is func 0, indicate the closure of the slot.
> > > + * then we get the chance to check all functions on same device
> > > + * if valid.
> > > + */
> > > + if (pci_get_function_0(pci_dev) == pci_dev) {
> > > + pci_bus_check_device(pci_dev, &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 dedf277..4e56256 100644
> > > --- a/include/hw/pci/pci.h
> > > +++ b/include/hw/pci/pci.h
> > > @@ -191,6 +191,7 @@ 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;
> > > --
> > > 1.9.3
> > >
> > >
next prev parent reply other threads:[~2016-03-09 17:14 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-07 3:22 [Qemu-devel] [PATCH v2 Resend 00/11] vfio-pci: pass the aer error to guest, part2 Cao jin
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 01/11] vfio: extract vfio_get_hot_reset_info as a single function Cao jin
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 02/11] vfio: squeeze out vfio_pci_do_hot_reset for support bus reset Cao jin
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 03/11] vfio: add pcie extended capability support Cao jin
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 04/11] vfio: add aer support for vfio device Cao jin
2016-03-08 22:55 ` Alex Williamson
2016-03-09 1:21 ` Chen Fan
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 05/11] vfio: add check host bus reset is support or not Cao jin
2016-03-08 22:55 ` Alex Williamson
2016-03-09 1:26 ` Chen Fan
2016-03-09 16:47 ` Michael S. Tsirkin
2016-03-09 16:59 ` Alex Williamson
2016-03-09 17:21 ` Michael S. Tsirkin
2016-03-07 3:22 ` [Qemu-devel] [PATCH v2 06/11] pci: add a is_valid_func callback to check device if complete Cao jin
2016-03-09 16:22 ` Michael S. Tsirkin
2016-03-09 16:50 ` Alex Williamson
2016-03-09 17:14 ` Michael S. Tsirkin [this message]
2016-03-10 2:00 ` Chen Fan
2016-03-07 3:23 ` [Qemu-devel] [PATCH v2 07/11] vfio: add check aer functionality for hotplug device Cao jin
2016-03-07 3:23 ` [Qemu-devel] [PATCH v2 08/11] pci: introduce pci bus pre reset Cao jin
2016-03-07 3:23 ` [Qemu-devel] [PATCH v2 09/11] vfio: vote a device to do host bus reset Cao jin
2016-03-09 10:07 ` Michael S. Tsirkin
2016-03-09 16:37 ` Alex Williamson
2016-03-10 6:15 ` Chen Fan
2016-03-10 14:16 ` Michael S. Tsirkin
2016-03-09 16:39 ` Michael S. Tsirkin
2016-03-09 17:09 ` Alex Williamson
2016-03-09 17:31 ` Michael S. Tsirkin
2016-03-07 3:23 ` [Qemu-devel] [PATCH v2 10/11] vfio-pci: pass the aer error to guest Cao jin
2016-03-07 3:23 ` [Qemu-devel] [PATCH v2 11/11] vfio: add 'aer' property to expose aercap Cao jin
-- strict thread matches above, loose matches on Subject: below --
2016-02-19 10:42 [Qemu-devel] [PATCH v2 00/11] vfio-pci: pass the aer error to guest, part2 Cao jin
2016-02-19 10:42 ` [Qemu-devel] [PATCH v2 06/11] pci: add a is_valid_func callback to check device if complete 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=20160309171449.GA8144@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).