[pve-devel] [PATCH pve-installer v9] update the PCI(e) docs
Friedrich Weber
f.weber at proxmox.com
Wed Oct 4 16:53:46 CEST 2023
Ping due to user interest [1]. The patch still applies (though the repo
in the subject should be `pve-docs`), and AFAICT the feedback for v8 has
been addressed.
[1] https://forum.proxmox.com/threads/134202/#post-592682
On 20/07/2023 11:32, Noel Ullreich wrote:
> A little update to the PCI(e) docs. The PCI wiki article has been
> reworked as well, in line with changes from this patch.
>
> Along some minor grammar fixes added:
> * how to check if kernelmodules are being loaded
> * how to check which drivers to blacklist
> * how to add softdeps for module loading
> * where to find kernel params
>
> Signed-off-by: Noel Ullreich <n.ullreich at proxmox.com>
> ---
> changes from v1:
> * fixed spelling mistakes
> * reduced code snippets of how to check iommu groupings to one
> * moved where to find kernel params to kernel cmdline section
> * removed wrong info on display output. will add correct info to
> Examples-Wiki
> * changed module names to variable-names, so that people can't
> blindly copy-paste.
> * restructured commit message ;)
>
> changes from v2:
> * while moving where to find the kernel params to the kernel
> cmdline section, I forgot to remove it from the pci(e) section
> * fixed typo in the link to the kernel param section
>
> changes from v3:
> * Some restructuring of the layout as well as moving parts of the
> PCI examples wiki to the docs here. This should lead to well-
> structured, concise docs that are independent from the PCI wiki.
> * found some more minor grammar errors
> * found a spelling mistake in qm.adoc
>
> changes from v4:
> * formatted the git message wrong again :/
>
> changes from v5:
> * fixed links to wiki
> * moved where to find kernel params to end of its chapter
> * the `vfio_virqfd` does not need to be loaded anymore with kernel 6.2
>
> changes from v6: the public wiki was updated -> fixed the links
>
> changes from v7:
> Forum user Leesteken noted that for Intel cpus IOMMU is not automatically
> activated anymore.
> https://forum.proxmox.com/threads/fix-pci-passthrough-documentation.122521/post-566926
> Thanks Leesteken :)
>
> changes from v8:
> Amended Dominiks notes:
> * added a note to remove `vfio_virqfd` with pve 7 eol
> * fixed formatting of a note-block
>
> qm-pci-passthrough.adoc | 165 +++++++++++++++++++++++++++++++---------
> qm.adoc | 2 +-
> system-booting.adoc | 8 ++
> 3 files changed, 137 insertions(+), 38 deletions(-)
>
> diff --git a/qm-pci-passthrough.adoc b/qm-pci-passthrough.adoc
> index b90a0b9..693deb7 100644
> --- a/qm-pci-passthrough.adoc
> +++ b/qm-pci-passthrough.adoc
> @@ -13,19 +13,27 @@ features (e.g., offloading).
> But, if you pass through a device to a virtual machine, you cannot use that
> device anymore on the host or in any other VM.
>
> +Note that, while PCI passthrough is available for i440fx and q35 machines, PCIe
> +passthrough is only available on q35 machines. This does not mean that
> +PCIe capable devices that are passed through as PCI devices will only run at
> +PCI speeds. Passing through devices as PCIe just sets a flag for the guest to
> +tell it that the device is a PCIe device instead of a "really fast legacy PCI
> +device". Some guest applications benefit from this.
> +
> General Requirements
> ~~~~~~~~~~~~~~~~~~~~
>
> -Since passthrough is a feature which also needs hardware support, there are
> -some requirements to check and preparations to be done to make it work.
> -
> +Since passthrough is performed on real hardware, it needs to fulfill some
> +requirements. A brief overview of these requirements is given below, for more
> +information on specific devices, see
> +https://pve.proxmox.com/wiki/PCI_Passthrough[PCI Passthrough Examples].
>
> Hardware
> ^^^^^^^^
> Your hardware needs to support `IOMMU` (*I*/*O* **M**emory **M**anagement
> **U**nit) interrupt remapping, this includes the CPU and the mainboard.
>
> -Generally, Intel systems with VT-d, and AMD systems with AMD-Vi support this.
> +Generally, Intel systems with VT-d and AMD systems with AMD-Vi support this.
> But it is not guaranteed that everything will work out of the box, due
> to bad hardware implementation and missing or low quality drivers.
>
> @@ -35,6 +43,17 @@ hardware, but even then, many modern system can support this.
> Please refer to your hardware vendor to check if they support this feature
> under Linux for your specific setup.
>
> +Determining PCI Card Address
> +^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> +
> +The easiest way is to use the GUI to add a device of type "Host PCI" in the VM's
> +hardware tab. Alternatively, you can use the command line.
> +
> +You can locate your card using
> +
> +----
> + lspci
> +----
>
> Configuration
> ^^^^^^^^^^^^^
> @@ -44,13 +63,12 @@ some configuration to enable PCI(e) passthrough.
>
> .IOMMU
>
> -First, you have to enable IOMMU support in your BIOS/UEFI. Usually the
> -corresponding setting is called `IOMMU` or `VT-d`,but you should find the exact
> +First, you will have to enable IOMMU support in your BIOS/UEFI. Usually the
> +corresponding setting is called `IOMMU` or `VT-d`, but you should find the exact
> option name in the manual of your motherboard.
>
> -For Intel CPUs, you may also need to enable the IOMMU on the
> -xref:sysboot_edit_kernel_cmdline[kernel command line] for older (pre-5.15)
> -kernels by adding:
> +For Intel CPUs, you also need to enable the IOMMU on the
> +xref:sysboot_edit_kernel_cmdline[kernel command line] kernels by adding:
>
> ----
> intel_iommu=on
> @@ -74,14 +92,17 @@ to the xref:sysboot_edit_kernel_cmdline[kernel commandline].
>
> .Kernel Modules
>
> +//TODO: remove `vfio_virqfd` stuff with eol of pve 7
> You have to make sure the following modules are loaded. This can be achieved by
> -adding them to `'/etc/modules''
> +adding them to `'/etc/modules''. In kernels newer than 6.2 ({pve} 8 and onward)
> +the 'vfio_virqfd' module is part of the 'vfio' module, therefore loading
> +'vfio_virqfd' in {pve} 8 and newer is not necessary.
>
> ----
> vfio
> vfio_iommu_type1
> vfio_pci
> - vfio_virqfd
> + vfio_virqfd #not needed if on kernel 6.2 or newer
> ----
>
> [[qm_pci_passthrough_update_initramfs]]
> @@ -92,6 +113,14 @@ After changing anything modules related, you need to refresh your
> # update-initramfs -u -k all
> ----
>
> +To check if the modules are being loaded, the output of
> +
> +----
> +# lsmod | grep vfio
> +----
> +
> +should include the four modules from above.
> +
> .Finish Configuration
>
> Finally reboot to bring the changes into effect and check that it is indeed
> @@ -104,11 +133,16 @@ enabled.
> should display that `IOMMU`, `Directed I/O` or `Interrupt Remapping` is
> enabled, depending on hardware and kernel the exact message can vary.
>
> +For notes on how to troubleshoot or verify if IOMMU is working as intended, please
> +see the https://pve.proxmox.com/wiki/PCI_Passthrough#Verifying_IOMMU_parameters[Verifying IOMMU Parameters]
> +section in our wiki.
> +
> It is also important that the device(s) you want to pass through
> -are in a *separate* `IOMMU` group. This can be checked with:
> +are in a *separate* `IOMMU` group. This can be checked with a call to the {pve}
> +API:
>
> ----
> -# find /sys/kernel/iommu_groups/ -type l
> +# pvesh get /nodes/{nodename}/hardware/pci --pci-class-blacklist ""
> ----
>
> It is okay if the device is in an `IOMMU` group together with its functions
> @@ -159,8 +193,8 @@ PCI(e) card, for example a GPU or a network card.
> Host Configuration
> ^^^^^^^^^^^^^^^^^^
>
> -In this case, the host must not use the card. There are two methods to achieve
> -this:
> +{pve} tries to automatically make the PCI(e) device unavailable for the host.
> +However, if this doesn't work, there are two things that can be done:
>
> * pass the device IDs to the options of the 'vfio-pci' modules by adding
> +
> @@ -175,7 +209,7 @@ the vendor and device IDs obtained by:
> # lspci -nn
> ----
>
> -* blacklist the driver completely on the host, ensuring that it is free to bind
> +* blacklist the driver on the host completely, ensuring that it is free to bind
> for passthrough, with
> +
> ----
> @@ -183,11 +217,49 @@ for passthrough, with
> ----
> +
> in a .conf file in */etc/modprobe.d/*.
> ++
> +To find the drivername, execute
> ++
> +----
> +# lspci -k
> +----
> ++
> +for example:
> ++
> +----
> +# lspci -k | grep -A 3 "VGA"
> +----
> ++
> +will output something similar to
> ++
> +----
> +01:00.0 VGA compatible controller: NVIDIA Corporation GP108 [GeForce GT 1030] (rev a1)
> + Subsystem: Micro-Star International Co., Ltd. [MSI] GP108 [GeForce GT 1030]
> + Kernel driver in use: <some-module>
> + Kernel modules: <some-module>
> +----
> ++
> +Now we can blacklist the drivers by writing them into a .conf file:
> ++
> +----
> +echo "blacklist <some-module>" >> /etc/modprobe.d/blacklist.conf
> +----
>
> For both methods you need to
> xref:qm_pci_passthrough_update_initramfs[update the `initramfs`] again and
> reboot after that.
>
> +Should this not work, you might need to set a soft dependency to load the gpu
> +modules before loading 'vfio-pci'. This can be done with the 'softdep' flag, see
> +also the manpages on 'modprobe.d' for more information.
> +
> +For example, if you are using drivers named <some-module>:
> +
> +----
> +# echo "softdep <some-module> pre: vfio-pci" >> /etc/modprobe.d/<some-module>.conf
> +----
> +
> +
> .Verify Configuration
>
> To check if your changes were successful, you can use
> @@ -208,13 +280,42 @@ passthrough.
> [[qm_pci_passthrough_vm_config]]
> VM Configuration
> ^^^^^^^^^^^^^^^^
> -To pass through the device you need to set the *hostpciX* option in the VM
> +When passing through a GPU, the best compatibility is reached when using
> +'q35' as machine type, 'OVMF' ('UEFI' for VMs) instead of SeaBIOS and PCIe
> +instead of PCI. Note that if you want to use 'OVMF' for GPU passthrough, the
> +GPU needs to have an UEFI capable ROM, otherwise use SeaBIOS instead. To check if
> +the ROM is UEFI capable, see the
> +https://pve.proxmox.com/wiki/PCI_Passthrough#How_to_know_if_a_graphics_card_is_UEFI_.28OVMF.29_compatible[PCI Passthrough Examples]
> +wiki.
> +
> +Furthermore, using OVMF, disabling vga arbitration may be possible, reducing the
> +amount of legacy code needed to be run during boot. To disable vga arbitration:
> +
> +----
> + echo "options vfio-pci ids=<vendor-id>,<device-id> disable_vga=1" > /etc/modprobe.d/vfio.conf
> +----
> +
> +replacing the <vendor-id> and <device-id> with the ones obtained from
> +
> +----
> +# lspci -nn
> +----
> +
> +PCI devices can be added in the web interface in the hardware section of the VM.
> +Alternatively, you can use the command line; set the *hostpciX* option in the VM
> configuration, for example by executing:
>
> ----
> # qm set VMID -hostpci0 00:02.0
> ----
>
> +or by adding a line to the VM configuration file:
> +
> +----
> + hostpci0: 00:02.0
> +----
> +
> +
> If your device has multiple functions (e.g., ``00:02.0`' and ``00:02.1`' ),
> you can pass them through all together with the shortened syntax ``00:02`'.
> This is equivalent with checking the ``All Functions`' checkbox in the
> @@ -262,21 +363,21 @@ For example:
> # qm set VMID -hostpci0 02:00,device-id=0x10f6,sub-vendor-id=0x0000
> ----
>
> -
> -Other considerations
> -^^^^^^^^^^^^^^^^^^^^
> -
> -When passing through a GPU, the best compatibility is reached when using
> -'q35' as machine type, 'OVMF' ('EFI' for VMs) instead of SeaBIOS and PCIe
> -instead of PCI. Note that if you want to use 'OVMF' for GPU passthrough, the
> -GPU needs to have an EFI capable ROM, otherwise use SeaBIOS instead.
> -
> SR-IOV
> ~~~~~~
>
> -Another variant for passing through PCI(e) devices, is to use the hardware
> +Another variant for passing through PCI(e) devices is to use the hardware
> virtualization features of your devices, if available.
>
> +.Enabling SR-IOV
> +[NOTE]
> +====
> +To use SR-IOV, platform support is especially important. It may be necessary
> +to enable this feature in the BIOS/UEFI first, or to use a specific PCI(e) port
> +for it to work. In doubt, consult the manual of the platform or contact its
> +vendor.
> +====
> +
> 'SR-IOV' (**S**ingle-**R**oot **I**nput/**O**utput **V**irtualization) enables
> a single device to provide multiple 'VF' (**V**irtual **F**unctions) to the
> system. Each of those 'VF' can be used in a different VM, with full hardware
> @@ -288,7 +389,6 @@ Currently, the most common use case for this are NICs (**N**etwork
> physical port. This allows using features such as checksum offloading, etc. to
> be used inside a VM, reducing the (host) CPU overhead.
>
> -
> Host Configuration
> ^^^^^^^^^^^^^^^^^^
>
> @@ -326,14 +426,6 @@ After creating VFs, you should see them as separate PCI(e) devices when
> outputting them with `lspci`. Get their ID and pass them through like a
> xref:qm_pci_passthrough_vm_config[normal PCI(e) device].
>
> -Other considerations
> -^^^^^^^^^^^^^^^^^^^^
> -
> -For this feature, platform support is especially important. It may be necessary
> -to enable this feature in the BIOS/EFI first, or to use a specific PCI(e) port
> -for it to work. In doubt, consult the manual of the platform or contact its
> -vendor.
> -
> Mediated Devices (vGPU, GVT-g)
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> @@ -346,7 +438,6 @@ With this, a physical Card is able to create virtual cards, similar to SR-IOV.
> The difference is that mediated devices do not appear as PCI(e) devices in the
> host, and are such only suited for using in virtual machines.
>
> -
> Host Configuration
> ^^^^^^^^^^^^^^^^^^
>
> diff --git a/qm.adoc b/qm.adoc
> index b3c3034..ed804e2 100644
> --- a/qm.adoc
> +++ b/qm.adoc
> @@ -139,7 +139,7 @@ snapshots) more intelligently.
> {pve} allows to boot VMs with different firmware and machine types, namely
> xref:qm_bios_and_uefi[SeaBIOS and OVMF]. In most cases you want to switch from
> the default SeaBIOS to OVMF only if you plan to use
> -xref:qm_pci_passthrough[PCIe pass through]. A VMs 'Machine Type' defines the
> +xref:qm_pci_passthrough[PCIe passthrough]. A VMs 'Machine Type' defines the
> hardware layout of the VM's virtual motherboard. You can choose between the
> 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]
> diff --git a/system-booting.adoc b/system-booting.adoc
> index 0e61e3d..71603b0 100644
> --- a/system-booting.adoc
> +++ b/system-booting.adoc
> @@ -288,6 +288,14 @@ The kernel commandline needs to be placed as one line in `/etc/kernel/cmdline`.
> To apply your changes, run `proxmox-boot-tool refresh`, which sets it as the
> `option` line for all config files in `loader/entries/proxmox-*.conf`.
>
> +A complete list of kernel parameters can be found at
> +'https://www.kernel.org/doc/html/v<YOUR-KERNEL-VERSION>/admin-guide/kernel-parameters.html'.
> +replace <YOUR-KERNEL-VERSION> with the major.minor version (e.g. 5.15). You can
> +find your kernel version by running
> +
> +----
> +# uname -r
> +----
>
> [[sysboot_kernel_pin]]
> Override the Kernel-Version for next Boot
More information about the pve-devel
mailing list