Introduction - Revision history

Arthur at 03:21, 19 January 2023

2023-01-19T03:21:02Z

← Older revision		Revision as of 03:21, 19 January 2023
Line 11:		Line 11:

	=== GPU Drivers ===		=== GPU Drivers ===
	GPUs need software to communicate from a host computer system to the device - this software is called a GPU Driver and it typically interacts with the device using [[wikipedia:Memory-mapped_I/O\|MMIO or Memory Mapped IO]]. MMIO is a set of operations used to allow normal CPU instructions to address registers in RAM in order to communicate with the PCI-e device. In the context of virtualization these registers must be presented into the device in a way that the guest operating system expects. Typically this is accomplished using [https://open-iov.org/index.php/Virtual_I/O_Internals#All_Modes VFIO or Virtual Function IO]. VFIO is used for many methods of virtualizing the device~~. This includes~~ discrete assignment, ~~software virtualization, and firmware virtualization~~.		GPUs need software to communicate from a host computer system to the device - this software is called a GPU Driver and it typically interacts with the device using [[wikipedia:Memory-mapped_I/O\|MMIO or Memory Mapped IO]]. MMIO is a set of operations used to allow normal CPU instructions to address registers in RAM in order to communicate with the PCI-e device. In the context of virtualization these registers must be presented into the device in a way that the guest operating system expects. Typically this is accomplished using [https://open-iov.org/index.php/Virtual_I/O_Internals#All_Modes VFIO or Virtual Function IO]. VFIO is used for many methods of virtualizing the device including discrete assignment of a physical device, or virtual (partitioned) device assignment created with or without hardware assistance.

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware ~~enforcement~~ access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output is known as an IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization (VFIO-Mdev)], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization (SR-IOV)], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforced access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output is known as an IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization (VFIO-Mdev)], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization (SR-IOV)], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* IOMMU */

2022-12-21T23:02:04Z

IOMMU

← Older revision		Revision as of 23:02, 21 December 2022
Line 14:		Line 14:

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output is known as an IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output is known as an IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization (VFIO-Mdev)], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization (SR-IOV)], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* IOMMU */

2022-12-21T23:00:49Z

IOMMU

← Older revision		Revision as of 23:00, 21 December 2022
Line 14:		Line 14:

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output is known as an IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* IOMMU */

2022-12-21T22:59:44Z

IOMMU

← Older revision		Revision as of 22:59, 21 December 2022
Line 14:		Line 14:

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized MMIO (Memory Mapped Input Output) devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* IOMMU */

2022-12-21T22:56:38Z

IOMMU

← Older revision		Revision as of 22:56, 21 December 2022
Line 14:		Line 14:

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (an [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (a [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* IOMMU */

2022-12-21T22:56:23Z

IOMMU

← Older revision		Revision as of 22:56, 21 December 2022
Line 14:		Line 14:

	=== IOMMU ===		=== IOMMU ===
	[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (an [[wikipedia:Input–output_memory_management_unit\|~~Input Output~~ Memory Management Unit or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).		[[wikipedia:Memory_management_unit\|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging\|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (an [[wikipedia:Input–output_memory_management_unit\|Memory Management Unit for Input Output or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===

Arthur: /* GPU Drivers */

2022-12-21T22:54:52Z

GPU Drivers

← Older revision		Revision as of 22:54, 21 December 2022
Line 11:		Line 11:

	=== GPU Drivers ===		=== GPU Drivers ===
	GPUs need software to communicate from a host computer system to the device ~~(of course)~~ - this software is called a GPU Driver and it typically interacts with the device using [[wikipedia:Memory-mapped_I/O\|MMIO or Memory Mapped IO]]. MMIO is a set of operations used to allow normal CPU instructions to address registers in RAM in order to communicate with the PCI-e device. In the context of virtualization these registers must be presented into the device in a way that the guest operating system expects. Typically this is accomplished using [https://open-iov.org/index.php/Virtual_I/O_Internals#All_Modes VFIO or Virtual Function IO]. VFIO is used for many methods of virtualizing the device. This includes discrete assignment, software virtualization, and firmware virtualization.		GPUs need software to communicate from a host computer system to the device - this software is called a GPU Driver and it typically interacts with the device using [[wikipedia:Memory-mapped_I/O\|MMIO or Memory Mapped IO]]. MMIO is a set of operations used to allow normal CPU instructions to address registers in RAM in order to communicate with the PCI-e device. In the context of virtualization these registers must be presented into the device in a way that the guest operating system expects. Typically this is accomplished using [https://open-iov.org/index.php/Virtual_I/O_Internals#All_Modes VFIO or Virtual Function IO]. VFIO is used for many methods of virtualizing the device. This includes discrete assignment, software virtualization, and firmware virtualization.

	=== IOMMU ===		=== IOMMU ===

Arthur at 22:27, 12 December 2022

2022-12-12T22:27:02Z

← Older revision		Revision as of 22:27, 12 December 2022
Line 1:		Line 1:
	This page will provide introductory context on where to begin reading. It will aim to give you a sense of primitives or foundational concepts used for GPU Virtualization.		This page will provide introductory context on where to begin reading. It will aim to give you a sense of primitives or foundational concepts used for GPU Virtualization.

	== ~~Primarives~~ ==		== Terminology ==
			To get started with basic terminology around IO Virtualization read the [[Glossary]] page. This will get you familiar with a few terms and concepts used in IO Virtualization documentation.

			== Primitives ==
	The primitives section will cover a few notes on the basics of virtualizing a GPU or IO device.		The primitives section will cover a few notes on the basics of virtualizing a GPU or IO device.

Arthur: /* Virtualization vs. Paravirtualization */

2022-12-12T17:43:30Z

Virtualization vs. Paravirtualization

← Older revision		Revision as of 17:43, 12 December 2022
Line 14:		Line 14:

	=== Virtualization vs. Paravirtualization ===		=== Virtualization vs. Paravirtualization ===
	Virtualization provides ~~the use of~~ unmodified guest drivers in the context of GPUs.		[[wikipedia:Virtualization\|Virtualization]] provides support for unmodified guest software or drivers in the context of GPUs.

	Paravirtualization provides		[[wikipedia:Paravirtualization\|Paravirtualization]] provides support for "enlightened" guests or modified software which is aware of virtualization.

	=== ~~Software~~ Virtualization vs. ~~Firmware~~ Virtualization ===		=== Unassisted vs. Assisted ===
			Similar to virtualization on CPUs with tools like Xen, IO devices can be virtualized on different kinds of computer hardware. Some devices provide hardware virtualization assistance (such as [https://wiki.xenproject.org/wiki/Xen_Project_Software_Overview#HVM_and_its_variants_.28x86.29 HVM mode in Xen]) and some do not (such as [https://wiki.xenproject.org/wiki/Xen_Project_Software_Overview#PV_.28x86.29 PV mode in Xen]), as a result an [[Virtual I/O Internals\|appropriate mode should be used]] depending on the availability of assistance.

			== Further Reading ==
			This section will cover what you should read to obtain more information about virtualization.

			=== Virtualization Fundamentals ===
			The [[Virtualization Fundamentals]] article provides links to some of the foundational work which underly modern virtualization industry and development today. If you'd like to understand both the theories on which virtualization builds as well as the history of industry you should read several articles here. Notable ''[https://en.m.wikipedia.org/wiki/Popek_and_Goldberg_virtualization_requirements Popek and Goldberg Virtualization Requirements]'' as well as ''[https://www.cl.cam.ac.uk/research/srg/netos/papers/2003-xensosp.pdf Xen and the Art of Virtualization]'' provide a great foundations-first approach to learning about virtualization.

			=== Virtual I/O Internals ===
			The [[Virtual I/O Internals]] article provides a detailed look into the interface layers between IO driver internals, the hypervisor, and the guest OS. This is a great article to read if you'd like to better understand how virtual machines can interact with both discretely assigned devices as well as those which have been partitioned using software or firmware.

			=== GPU Driver Internals ===
			The [[GPU Driver Internals]] article provides information on how GPU drivers are structured internally as well as how they achieve virtualization. This includes high level architecture, device initialization, as well as memory mapping, and scheduling.

Arthur: Created page with "This page will provide introductory context on where to begin reading. It will aim to give you a sense of primitives or foundational concepts used for GPU Virtualization. == Primarives == The primitives section will cover a few notes on the basics of virtualizing a GPU or IO device. === PCI BAR === PCI-e devices such as GPUs are used for input and output to and from a computer system. This input and output occurs using [http://xillybus.com/tutorials/pci-express-tlp-pci..."

2022-12-12T17:04:26Z

Created page with "This page will provide introductory context on where to begin reading. It will aim to give you a sense of primitives or foundational concepts used for GPU Virtualization. == Primarives == The primitives section will cover a few notes on the basics of virtualizing a GPU or IO device. === PCI BAR === PCI-e devices such as GPUs are used for input and output to and from a computer system. This input and output occurs using [http://xillybus.com/tutorials/pci-express-tlp-pci..."

New page

This page will provide introductory context on where to begin reading. It will aim to give you a sense of primitives or foundational concepts used for GPU Virtualization.

== Primarives ==
The primitives section will cover a few notes on the basics of virtualizing a GPU or IO device.

=== PCI BAR ===
PCI-e devices such as GPUs are used for input and output to and from a computer system. This input and output occurs using [http://xillybus.com/tutorials/pci-express-tlp-pcie-primer-tutorial-guide-1 TLPs or Transaction Layer Packets] over the PCI-e packet network. The IO device is presented to the host operating system as [https://wiki.osdev.org/PCI#Base_Address_Registers BARs or Base Address Registers].

=== GPU Drivers ===
GPUs need software to communicate from a host computer system to the device (of course) - this software is called a GPU Driver and it typically interacts with the device using [[wikipedia:Memory-mapped_I/O|MMIO or Memory Mapped IO]]. MMIO is a set of operations used to allow normal CPU instructions to address registers in RAM in order to communicate with the PCI-e device. In the context of virtualization these registers must be presented into the device in a way that the guest operating system expects. Typically this is accomplished using [https://open-iov.org/index.php/Virtual_I/O_Internals#All_Modes VFIO or Virtual Function IO]. VFIO is used for many methods of virtualizing the device. This includes discrete assignment, software virtualization, and firmware virtualization.

=== IOMMU ===
[[wikipedia:Memory_management_unit|MMU or Memory Management Units]] handle hardware enforcement access to [[wikipedia:Memory_paging|memory paging]] in computer systems. In the context of virtualized Memory Mapped Input Output devices an MMU is required to hardware enforce access to virtual devices (an [[wikipedia:Input–output_memory_management_unit|Input Output Memory Management Unit or IOMMU]]). In the case of VFIO, groups of IOMMU devices ([https://vfio.blogspot.com/2014/08/iommu-groups-inside-and-out.html IOMMU Groups]) are used to control access to the device. These IOMMU Groups often represent individual devices or sub-devices (in the context of [https://open-iov.org/index.php/Virtual_I/O_Internals#Mdev_Mode Unassisted Virtualization], [https://open-iov.org/index.php/Virtual_I/O_Internals#SR-IOV_Mode Hardware-Assisted Virtualization], or [https://open-iov.org/index.php/Virtual_I/O_Internals#SIOV_Mode Hardware-Assisted Paravirtualization]) which may be bound to a VFIO bus driver (such as [https://docs.kernel.org/driver-api/vfio.html#vfio-bus-driver-api vfio-pci]) which results in a [https://open-iov.org/index.php/Virtual_I/O_Internals#Binding_VFIO_devices VFIO Group] (unit of ownership).

=== Virtualization vs. Paravirtualization ===
Virtualization provides the use of unmodified guest drivers in the context of GPUs.

Paravirtualization provides

=== Software Virtualization vs. Firmware Virtualization ===