[pve-devel] [PATCH docs v7 3/4] added vIOMMU documentation

[Markus Frank]

Signed-off-by: Markus Frank <m.frank at proxmox.com>
---
 qm-pci-passthrough.adoc | 53 +++++++++++++++++++++++++++++++++++++++++
 qm.adoc                 |  1 +
 2 files changed, 54 insertions(+)

diff --git a/qm-pci-passthrough.adoc b/qm-pci-passthrough.adoc
index b90a0b9..7f7d4a7 100644
--- a/qm-pci-passthrough.adoc
+++ b/qm-pci-passthrough.adoc
@@ -408,6 +408,59 @@ properly used with HA and hardware changes are detected and non root users
 can configure them. See xref:resource_mapping[Resource Mapping]
 for details on that.
 
+[[qm_pci_viommu]]
+vIOMMU (emulated IOMMU)
+~~~~~~~~~~~~~~~~~~~~~~~
+
+Using the vIOMMU option allows you to to passthrough PCI devices to layer-2
+VMs in layer-1 VMs via
+https://pve.proxmox.com/wiki/Nested_Virtualization[nested virtualization].
+There are currently two vIOMMU implementations available: Intel and VirtIO.
+
+Host requirement:
+
+* Add `intel_iommu=on` or `amd_iommu=on` depending on your CPU to your kernel
+command line.
+
+Nested virtualization VM requirement:
+
+* Enable *kvm* and use the *host* cpu type to use nested virtualization.
+
+Intel vIOMMU
+^^^^^^^^^^^^
+
+Intel vIOMMU specific VM requirements:
+
+* Whether you are using an Intel or AMD CPU on your host, it is important to set
+`intel_iommu=on` in the VMs kernel parameters.
+
+* To use Intel vIOMMU you need to set *q35* as the machine type.
+
+If all requirements are met, you can add `viommu=intel` to the machine parameter
+in the configuration of the VM that should be able to passthrough PCI devices.
+
+----
+# qm set VMID -machine q35,viommu=intel
+----
+
+https://wiki.qemu.org/Features/VT-d[QEMU documentation for VT-d]
+
+VirtIO vIOMMU
+^^^^^^^^^^^^^
+
+This vIOMMU implementation is more recent and does not have as many limitations
+as Intel vIOMMU but is currently less used in production and less documentated.
+
+With VirtIO vIOMMU there is *no* need to set any kernel parameters.
+It is also *not* necessary to use q35 as the machine type, but it is advisable
+if you want to use PCIe.
+
+----
+# qm set VMID -machine q35,viommu=virtio
+----
+
+https://web.archive.org/web/20230804075844/https://michael2012z.medium.com/virtio-iommu-789369049443[Blog-Post by Michael Zhao explaining virtio-iommu]
+
 ifdef::wiki[]
 
 See Also
diff --git a/qm.adoc b/qm.adoc
index b3c3034..2e5b109 100644
--- a/qm.adoc
+++ b/qm.adoc
@@ -145,6 +145,7 @@ default https://en.wikipedia.org/wiki/Intel_440FX[Intel 440FX] or the
 https://ark.intel.com/content/www/us/en/ark/products/31918/intel-82q35-graphics-and-memory-controller.html[Q35]
 chipset, which also provides a virtual PCIe bus, and thus may be desired if
 one wants to pass through PCIe hardware.
+Additionally, you can select a xref:qm_pci_viommu[vIOMMU] implementation.
 
 [[qm_hard_disk]]
 Hard Disk
-- 
2.39.2