On 07/21/2016 08:43 PM, Greg Kurz wrote:
From: Greg Kurz <gk...@linux.vnet.ibm.com>
Without presuming if we got there because of a user mistake or some
more subtle bug in the tooling, it really does not make sense to
implement a non-functional device.
Signed-off-by: Greg Kurz <gk...@linux.vnet.ibm.com>
Signed-off-by: Greg Kurz <gr...@kaod.org>
---
v3: - rebased on top of:
https://lists.gnu.org/archive/html/qemu-devel/2016-07/msg04744.html
- use virtio_pci_legacy/modern helpers
- rephrased error message to be shorter and use the on/off logic
Marcel, this still results in > 80 char line in the code but I'd rather not
split it to ease grepping, nor shorten the message even more to keep it
meaningful.
---
hw/virtio/virtio-pci.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index 755f9218b77d..1f5f00a50a0b 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -1842,6 +1842,11 @@ static void virtio_pci_dc_realize(DeviceState *qdev,
Error **errp)
VirtIOPCIProxy *proxy = VIRTIO_PCI(qdev);
PCIDevice *pci_dev = &proxy->pci_dev;
+ if (!(virtio_pci_modern(proxy) || virtio_pci_legacy(proxy))) {
Hi Greg,
Thanks for rebasing it.
A minor thing, disable-legacy is now auto/on/off.
If the user sets [disable-legacy=auto, disable-modern=on]
will not pass this test, but is possible that later on
will be enabled:
See virtio_pci_realize:
if (proxy->disable_legacy == ON_OFF_AUTO_AUTO) {
proxy->disable_legacy = pcie_port ? ON_OFF_AUTO_ON : ON_OFF_AUTO_OFF;
}
On the the other hand, if the user sets disable-modern=on, is reasonable to ask
him to set disable-legacy to off.
Reviewed-by: Marcel Apfelbaum <mar...@redhat.com>
Thanks,
Marcel
+ error_setg(errp, "device cannot work when both disable-modern and
disable-legacy are set to on.");
+ return;
+ }
+
if (!(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_PCIE) &&
virtio_pci_modern(proxy)) {
pci_dev->cap_present |= QEMU_PCI_CAP_EXPRESS;
