On 4/9/25 01:17, Aneesh Kumar K.V wrote:
Dan Williams <dan.j.williams@xxxxxxxxx> writes:
...
+/**
+ * pci_tsm_bind() - Bind @pdev as a TDI for @kvm
+ * @pdev: PCI device function to bind
+ * @kvm: Private memory attach context
+ * @tdi_id: Identifier (virtual BDF) for the TDI as referenced by the TSM and DSM
+ *
+ * Returns 0 on success, or a negative error code on failure.
+ *
+ * Context: Caller is responsible for constraining the bind lifetime to the
+ * registered state of the device. For example, pci_tsm_bind() /
+ * pci_tsm_unbind() limited to the VFIO driver bound state of the device.
+ */
+int pci_tsm_bind(struct pci_dev *pdev, struct kvm *kvm, u32 tdi_id)
+{
+ const struct pci_tsm_ops *ops;
+ struct pci_tsm_pf0 *tsm_pf0;
+ struct pci_tdi *tdi;
+
+ if (!kvm)
+ return -EINVAL;
+
+ guard(rwsem_read)(&pci_tsm_rwsem);
+
+ if (!pdev->tsm)
+ return -EINVAL;
+
+ ops = pdev->tsm->ops;
+
+ if (!is_link_tsm(ops->owner))
+ return -ENXIO;
+
+ tsm_pf0 = to_pci_tsm_pf0(pdev->tsm);
+ guard(mutex)(&tsm_pf0->lock);
+
+ /* Resolve races to bind a TDI */
+ if (pdev->tsm->tdi) {
+ if (pdev->tsm->tdi->kvm == kvm)
+ return 0;
+ else
+ return -EBUSY;
+ }
+
+ tdi = ops->bind(pdev, kvm, tdi_id);
+ if (IS_ERR(tdi))
+ return PTR_ERR(tdi);
+
+ pdev->tsm->tdi = tdi;
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(pci_tsm_bind);
+
Are we missing assigning pdev and kvm in the above function?
modified drivers/pci/tsm.c
@@ -356,6 +356,8 @@ int pci_tsm_bind(struct pci_dev *pdev, struct kvm *kvm, u32 tdi_id)
if (IS_ERR(tdi))
return PTR_ERR(tdi);
+ tdi->pdev = pdev;
This signals that this pdev backref is not exactly needed :)
+ tdi->kvm = kvm;
iommufd_vdevice holds the reference and AMD TSM only needs kvm* during bind() so it can go unnoticed but it is needed, yeah. Thanks,
pdev->tsm->tdi = tdi;
return 0;
-aneesh
--
Alexey
