public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Alex Williamson <alex.williamson@redhat.com>
To: kvm@vger.kernel.org
Cc: chrisw@redhat.com, alex.williamson@redhat.com
Subject: [PATCH] device-assignment: be more selective in interrupt disabling
Date: Thu, 17 Jun 2010 11:51:55 -0600	[thread overview]
Message-ID: <20100617174948.4792.54699.stgit@localhost.localdomain> (raw)

An 82576 physical function assigned to a Windows 7 guest currently
doesn't work because the driver seems to gratuitiously disable
MSI and MSIX interrupts.  When it does this, we blindly deassign
the current interrupt setup, leaving the device with no interrupts.
Instead let's only deassign the irq if we were previously using
MSI/MSIX or if we're going to start using MSI/MSIX.

Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
---

Note, this doesn't fix 82576 virtual functions on win7.  I can't find a
driver that even tries to claim the vf.  Please send a pointer if one exists.

 hw/device-assignment.c |   34 +++++++++++++++++++++++++---------
 1 files changed, 25 insertions(+), 9 deletions(-)

diff --git a/hw/device-assignment.c b/hw/device-assignment.c
index ba02157..85cd414 100644
--- a/hw/device-assignment.c
+++ b/hw/device-assignment.c
@@ -1031,13 +1031,20 @@ static void assigned_dev_update_msi(PCIDevice *pci_dev, unsigned int ctrl_pos)
         calc_assigned_dev_id(assigned_dev->h_segnr, assigned_dev->h_busnr,
                 (uint8_t)assigned_dev->h_devfn);
 
-    if (assigned_dev->irq_requested_type) {
-	    assigned_irq_data.flags = assigned_dev->irq_requested_type;
-	    free_dev_irq_entries(assigned_dev);
-	    r = kvm_deassign_irq(kvm_context, &assigned_irq_data);
-	    /* -ENXIO means no assigned irq */
-	    if (r && r != -ENXIO)
-		    perror("assigned_dev_update_msi: deassign irq");
+    /* Some guests gratuitously disable MSI even if they're not using it,
+     * try to catch this by only deassigning irqs if the guest is using
+     * MSI or intends to start. */
+    if ((assigned_dev->irq_requested_type & KVM_DEV_IRQ_GUEST_MSI) ||
+        (ctrl_byte & PCI_MSI_FLAGS_ENABLE)) {
+
+        assigned_irq_data.flags = assigned_dev->irq_requested_type;
+        free_dev_irq_entries(assigned_dev);
+        r = kvm_deassign_irq(kvm_context, &assigned_irq_data);
+        /* -ENXIO means no assigned irq */
+        if (r && r != -ENXIO)
+            perror("assigned_dev_update_msi: deassign irq");
+
+        assigned_irq_data.flags = 0;
     }
 
     if (ctrl_byte & PCI_MSI_FLAGS_ENABLE) {
@@ -1188,17 +1195,26 @@ static void assigned_dev_update_msix(PCIDevice *pci_dev, unsigned int ctrl_pos)
             calc_assigned_dev_id(assigned_dev->h_segnr, assigned_dev->h_busnr,
                     (uint8_t)assigned_dev->h_devfn);
 
-    if (assigned_dev->irq_requested_type) {
+    /* Some guests gratuitously disable MSIX even if they're not using it,
+     * try to catch this by only deassigning irqs if the guest is using
+     * MSIX or intends to start. */
+    if ((assigned_dev->irq_requested_type & KVM_DEV_IRQ_GUEST_MSIX) ||
+        (*ctrl_word & PCI_MSIX_ENABLE)) {
+
         assigned_irq_data.flags = assigned_dev->irq_requested_type;
         free_dev_irq_entries(assigned_dev);
         r = kvm_deassign_irq(kvm_context, &assigned_irq_data);
         /* -ENXIO means no assigned irq */
         if (r && r != -ENXIO)
             perror("assigned_dev_update_msix: deassign irq");
+
+        assigned_irq_data.flags = 0;
     }
-    assigned_irq_data.flags = KVM_DEV_IRQ_HOST_MSIX | KVM_DEV_IRQ_GUEST_MSIX;
 
     if (*ctrl_word & PCI_MSIX_ENABLE) {
+        assigned_irq_data.flags = KVM_DEV_IRQ_HOST_MSIX |
+                                  KVM_DEV_IRQ_GUEST_MSIX;
+
         if (assigned_dev_update_msix_mmio(pci_dev) < 0) {
             perror("assigned_dev_update_msix_mmio");
             return;


             reply	other threads:[~2010-06-17 17:51 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-06-17 17:51 Alex Williamson [this message]
2010-06-17 19:07 ` [PATCH] device-assignment: be more selective in interrupt disabling Chris Wright
2010-06-23 13:05 ` Avi Kivity

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=20100617174948.4792.54699.stgit@localhost.localdomain \
    --to=alex.williamson@redhat.com \
    --cc=chrisw@redhat.com \
    --cc=kvm@vger.kernel.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