From: Mark McLoughlin <markmc@redhat.com>
To: "Han, Weidong" <weidong.han@intel.com>
Cc: "'Avi Kivity'" <avi@redhat.com>,
"kvm@vger.kernel.org" <kvm@vger.kernel.org>
Subject: Re: [PATCH 4/4] kvm: qemu: Assign irq in init_assigned_device
Date: Wed, 10 Dec 2008 10:20:14 +0000 [thread overview]
Message-ID: <1228904414.5384.40.camel@blaa> (raw)
In-Reply-To: <715D42877B251141A38726ABF5CABF2C018BF9ACE2@pdsmsx503.ccr.corp.intel.com>
On Wed, 2008-12-10 at 17:45 +0800, Han, Weidong wrote:
> assign_dev_update_irq may not be invoked when hot add a device, so
> need to assign irq after assign device in init_assigned_device.
Makes sense, but ...
> Signed-off-by: Weidong Han <weidong.han@intel.com>
> ---
> qemu/hw/device-assignment.c | 99 ++++++++++++++++++++++++++++--------------
> 1 files changed, 66 insertions(+), 33 deletions(-)
>
> diff --git a/qemu/hw/device-assignment.c b/qemu/hw/device-assignment.c
> index 20618a4..557beb8 100644
> --- a/qemu/hw/device-assignment.c
> +++ b/qemu/hw/device-assignment.c
> @@ -490,6 +490,65 @@ static uint32_t calc_assigned_dev_id(uint8_t bus, uint8_t devfn)
> return (uint32_t)bus << 8 | (uint32_t)devfn;
> }
>
...
> +static int assign_irq(AssignedDevInfo *adev)
> +{
> + struct kvm_assigned_irq assigned_irq_data;
> + AssignedDevice *dev = adev->assigned_dev;
> + int irq, r = 0;
> +
> + irq = pci_map_irq(&dev->dev, dev->intpin);
> + irq = piix_get_irq(irq);
> +
> +#ifdef TARGET_IA64
> + irq = ipf_map_irq(&dev->dev, irq);
> +#endif
If you added here:
if (irq == dev->girq)
return;
then ...
> + memset(&assigned_irq_data, 0, sizeof(assigned_irq_data));
> + assigned_irq_data.assigned_dev_id =
> + calc_assigned_dev_id(dev->h_busnr, dev->h_devfn);
> + assigned_irq_data.guest_irq = irq;
> + assigned_irq_data.host_irq = dev->real_device.irq;
> + r = kvm_assign_irq(kvm_context, &assigned_irq_data);
> + if (r < 0) {
> + fprintf(stderr, "Failed to assign irq for \"%s\": %s\n",
> + adev->name, strerror(-r));
> + fprintf(stderr, "Perhaps you are assigning a device "
> + "that shares an IRQ with another device?\n");
> + return r;
> + }
> +
> + dev->girq = irq;
> + return r;
> +}
> +
> /* The pci config space got updated. Check if irq numbers have changed
> * for our devices
> */
> @@ -511,20 +570,8 @@ void assigned_dev_update_irq()
> #endif
>
> if (irq != assigned_dev->girq) {
> - struct kvm_assigned_irq assigned_irq_data;
> -
> - memset(&assigned_irq_data, 0, sizeof(assigned_irq_data));
> - assigned_irq_data.assigned_dev_id =
> - calc_assigned_dev_id(assigned_dev->h_busnr,
> - (uint8_t) assigned_dev->h_devfn);
> - assigned_irq_data.guest_irq = irq;
> - assigned_irq_data.host_irq = assigned_dev->real_device.irq;
> - r = kvm_assign_irq(kvm_context, &assigned_irq_data);
> + r = assign_irq(adev);
> if (r < 0) {
> - fprintf(stderr, "Failed to assign irq for \"%s\": %s\n",
> - adev->name, strerror(-r));
> - fprintf(stderr, "Perhaps you are assigning a device "
> - "that shares an IRQ with another device?\n");
> free_assigned_device(adev);
> adev = next;
> continue;
.... you end up with:
void assigned_dev_update_irq(PCIDevice *d)
{
AssignedDevInfo *adev;
adev = LIST_FIRST(&adev_head);
while (adev) {
AssignedDevInfo *next = LIST_NEXT(adev, next);
int r;
r = assign_irq(adev);
if (r < 0)
free_assigned_device(adev);
adev = next;
}
which is much nicer.
> @@ -579,27 +626,13 @@ struct PCIDevice *init_assigned_device(AssignedDevInfo *adev, PCIBus *bus)
> dev->h_busnr = adev->bus;
> dev->h_devfn = PCI_DEVFN(adev->dev, adev->func);
>
> - memset(&assigned_dev_data, 0, sizeof(assigned_dev_data));
> - assigned_dev_data.assigned_dev_id =
> - calc_assigned_dev_id(dev->h_busnr, (uint32_t)dev->h_devfn);
> - assigned_dev_data.busnr = dev->h_busnr;
> - assigned_dev_data.devfn = dev->h_devfn;
> -
> -#ifdef KVM_CAP_IOMMU
> - /* We always enable the IOMMU if present
> - * (or when not disabled on the command line)
> - */
> - r = kvm_check_extension(kvm_context, KVM_CAP_IOMMU);
> - if (r && !adev->disable_iommu)
> - assigned_dev_data.flags |= KVM_DEV_ASSIGN_ENABLE_IOMMU;
> -#endif
> + r = assign_device(adev);
> + if (r < 0)
> + return NULL;
>
> - r = kvm_assign_pci_device(kvm_context, &assigned_dev_data);
> - if (r < 0) {
> - fprintf(stderr, "Failed to assign device \"%s\" : %s\n",
> - adev->name, strerror(-r));
> - return NULL;
> - }
> + r = assign_irq(adev);
> + if (r < 0)
> + return NULL;
Hmm, why are we doing no cleanup if these fail?
Cheers,
Mark.
next prev parent reply other threads:[~2008-12-10 10:21 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-12-10 9:45 [PATCH 4/4] kvm: qemu: Assign irq in init_assigned_device Han, Weidong
2008-12-10 10:20 ` Mark McLoughlin [this message]
2008-12-10 12:18 ` Han, Weidong
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=1228904414.5384.40.camel@blaa \
--to=markmc@redhat.com \
--cc=avi@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=weidong.han@intel.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.