WO2006072101A2 - One step address translation of graphics addresses in virtualization - Google Patents

One step address translation of graphics addresses in virtualization Download PDF

Info

Publication number
WO2006072101A2
WO2006072101A2 PCT/US2005/047683 US2005047683W WO2006072101A2 WO 2006072101 A2 WO2006072101 A2 WO 2006072101A2 US 2005047683 W US2005047683 W US 2005047683W WO 2006072101 A2 WO2006072101 A2 WO 2006072101A2
Authority
WO
WIPO (PCT)
Prior art keywords
graphics
address
memory
guest
translation table
Prior art date
Application number
PCT/US2005/047683
Other languages
French (fr)
Other versions
WO2006072101A3 (en
Inventor
Michael Goldsmith
Kiran Panesar
Original Assignee
Intel Corporation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Intel Corporation filed Critical Intel Corporation
Priority to CN2005800448642A priority Critical patent/CN101088078B/en
Priority to JP2007549696A priority patent/JP5006798B2/en
Priority to EP19198130.7A priority patent/EP3605342A1/en
Priority to EP05856135A priority patent/EP1839158A2/en
Publication of WO2006072101A2 publication Critical patent/WO2006072101A2/en
Publication of WO2006072101A3 publication Critical patent/WO2006072101A3/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • G06F12/1027Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
    • G06F12/1036Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB] for multiple virtual address spaces, e.g. segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • G06F12/1009Address translation using page tables, e.g. page table structures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • G06F12/1081Address translation for peripheral access to main memory, e.g. direct memory access [DMA]
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • G06F12/109Address translation for multiple virtual address spaces, e.g. segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • G06F2009/45579I/O management, e.g. providing access to device drivers or storage
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533Hypervisors; Virtual machine monitors
    • G06F9/45558Hypervisor-specific management and integration aspects
    • G06F2009/45583Memory management, e.g. access or allocation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2212/00Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
    • G06F2212/10Providing a specific technical effect
    • G06F2212/1016Performance improvement
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2212/00Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
    • G06F2212/65Details of virtual memory and virtual address translation
    • G06F2212/657Virtual address space management

Definitions

  • a graphics device may require a contiguous memory range to store graphics data.
  • the graphics data may include, for example, three-dimensional data and instructions used by, for example, a graphics subsystem in a personal computer.
  • the contiguous graphics memory may be correlated to a physical memory using a graphics translation table. However, the physical memory may not be contiguous.
  • Virtualization in some aspects, is a technique to operate one or more guest operating systems (OS) on the same host hardware by presenting multiple copies of a host hardware-software interface to each guest OS.
  • OS guest operating systems
  • a memory address for the guest may be correlated to host physical memory address.
  • the physical host memory may back up the physical guest memory.
  • FIG. 1 is block diagram of a computer system, according to some embodiments hereof;
  • FIG. 2 is an exemplary depiction of a graphics translation table, according to some embodiments herein;
  • FIG. 3 is an exemplary depiction of a virtual machine, according to some embodiments herein;
  • FIG. 4 is an exemplary depiction of physical memory mapping on a virtualized machine, in accordance with some embodiments herein;
  • FIG. 5 is a flow diagram of a memory translation including two translations;
  • FIG. 6 is an exemplary memory translation scheme including one translation operation, according to some embodiments herein;
  • FIG. 7 is an exemplary flow diagram of a process, according to some embodiments herein.
  • FIG. 8 is an exemplary depiction of a logical memory map illustrating a guest graphics memory address to host physical address translation, in accordance with some embodiments herein.
  • FIG. 1 is an exemplary block diagram illustrating physical hardware of a computer system that may be used in some embodiments hereof, generally represented by reference number 100.
  • Computer system 100 includes a processor 105.
  • Processor 105 represents a central processing unit of any type of architecture. Some embodiments hereof may be implemented by computer systems having multiple processors.
  • Computer system 100 includes a chipset 110 that has a number of control circuits and a number of interface circuits to allow processor 105 to access a system memory 115, a system bus 150 (e.g., a Peripheral Component Interconnect (PCI) Express bus), and a graphics device 160.
  • PCI Peripheral Component Interconnect
  • PCI Express devices 155i through 155 n are connected to Express bus 150, as defined by the PCI Special Interest Group (PCI-SIG) in "PCI Express Base Specification, Version 1.0" (July 2002).
  • PCI Express devices 155i through 155 n are input/output (I/O) hardware devices such as, for example, a disk controller/card, a local area network controller/card, etc.
  • chipset 110 may have one or more bus controllers (e.g., PCI Express bus), a graphics controller, a CPU controller, and a memory controller to control data access between various components such as, for example, bus 150, graphics device 160, system memory 115, processor 105, etc.
  • bus controllers e.g., PCI Express bus
  • computer system 100 may include other, additional, or fewer components than those illustrated in FIG. 1 , without departing or altering the scope of the various embodiments herein.
  • System memory 115 represents one or more mechanisms for storing information.
  • system memory 115 may include non-volatile or volatile memories.
  • system memory 115 includes a graphics memory unit 118, a graphics aperture 120 and a main memory 125.
  • Main memory 125 may include an operating system (OS) 130, a memory manager 135, a graphics memory translation table 140, and additional main memory 145 allocated for other information such as, for example, other programs and data.
  • Graphics device 160 may be an add-in device or integrated into computer system 100.
  • graphics device 160 includes a graphics processor 165 and a graphics local memory 170.
  • the memory may be random access memory (RAM) (e.g., extended data out dynamic random access memory (EDO), synchronous graphic random access memory (SGRAM), video random access memory (VRAM)).
  • RAM random access memory
  • EEO extended data out dynamic random access memory
  • SGRAM synchronous graphic random access memory
  • VRAM video random access memory
  • the memory is included, since the video card must be able to remember a complete screen image at any time, and maintain local copies of graphics programs, and graphics objects like triangles, and textures. It is noted that some embodiments herein are also applicable to graphics devices and I/O devices having no local memory.
  • Graphics processor 160 performs graphics functions, such as, 3-D rendering operations, drawings, etc.
  • Graphics processor 165 has access to its own graphics local memory 160.
  • Graphics device 160 may be coupled to chipset 110 via accelerated graphics port (AGP) 175.
  • AGP accelerated graphics port
  • AGP 175 provides a high-speed bus for moving data directly from system memory 115 to graphics device 160.
  • Direct references may be made to system memory 115. Due to the direct references to system memory 115, a contiguous view of system memory 115 may be essential for efficient transfer of information between graphics device 160 and system memory 115.
  • graphics device 160 may be coupled to chipset via PCI express bus 150.
  • a range of system memory 115 is reserved for graphics memory unit 118 including graphics aperture 120.
  • Graphics aperture 120 provides a range of memory addresses used by AGP 175 for graphics information such as, for example, 3-D features and textures.
  • system memory 115 is dynamically allocated for graphics data, it is necessary to provide a graphics mapping table mechanism to map random segments of system memory 115 into a single contiguous, physical space for graphics aperture 120.
  • a graphics memory translation table such as a Graphics Address Re- Mapping Table (GART) or a Graphics Translation Table (GTT) may be used to provide a physically-contiguous view of scattered pages in system memory for direct memory access (DMA) transfers.
  • GART Graphics Address Re- Mapping Table
  • GTT Graphics Translation Table
  • FIG. 2 illustrates an exemplary graphics memory translation table map, generally represented by reference number 200.
  • system memory 115 includes a main memory 125 having an address range from address zero (0) to a top thereof.
  • Graphics aperture 120 may have an address range that begins from the top of main memory 125.
  • Graphics aperture 120 is a virtual memory and maps into the physical address space of main memory 125.
  • Graphics aperture 120 is a portion of system memory 115 that is allocated by operating system 130 for use by graphics device 160. Graphics device 160 is provided access to the reserved graphics aperture 120 to store texture data, front buffer data or other graphics data for faster graphics data processing. Each address Pa in graphics aperture 120 has a corresponding entry mapped into a physical address space Pg of main memory 125.
  • Operating system 130 allocates pages in main memory 125 (of system memory 115) wherever they are found and assigns them for graphics device 160. Graphics device 160 is therefore provided with a continuous block of graphics aperture 120, wherein references pointing from a graphics aperture 120 address Pa to a corresponding address Pg in main memory 125 are stored in graphics memory translation table 140 as a page table entry (PTE).
  • PTE page table entry
  • Virtualization in some aspects, is a technique to operate one or more guest operating systems (OS) on the same native hardware by presenting multiple copies of a host hardware-software interface to each guest OS.
  • the native hardware may be referred to as the host.
  • the multiple guest OSs may even run concurrently.
  • managing memory used by a graphics device or subsystem is further complicated because of a need to translate guest memory addresses to physical host addresses due to the virtualization of the host hardware.
  • FIG. 3 illustrates an exemplary system 300 that supports virtualization.
  • System 300 includes physical host hardware 305, a number of guest virtual machines VMo 31Oo and VMi 310-i, and a virtual machine monitor (VMM) 330.
  • VMM virtual machine monitor
  • Each of VM 0 31O 0 and VMi 31 Oi may be referred to as a guest herein relative to physical host hardware 305.
  • physical host hardware 305 may include a computer system and/or components thereof similar to system 100 illustrated in FIG. 1.
  • Physical host hardware 305 may also be referred to herein as the host.
  • physical host hardware 305 may include a processor, a memory, various I/O devices (e.g., keyboard, monitor, USB controller, network controller etc.), and a graphics device.
  • I/O devices e.g., keyboard, monitor, USB controller, network controller etc.
  • VM 0 31O 0 has a number of addresses 315, a guest OS G1 320, and components 325 associated therewith.
  • VMi 31O 1 has a number of addresses 317, a guest OS G2 322, and virtual components 327. It should be appreciated by those in the art that differences in functionality may exist between the virtual machine(s) supported by system 300.
  • VMM 330 provides a hardware-software interface to each of VMo 31Oo and VMi 310-
  • guest OS G1 320 and guest OS G2 322 may operate concurrently, in part due to each guest OS operating in its own virtual machine.
  • a guest memory address needs to be translated or mapped to a host physical address since the physical guest memory address (Pg) is a virtual memory address that is actually located at a physical host address (Ph). That is, since the guest's resources, including memory, are virtualizations of hardware or the entire hardware environment of the host, a correlation must be established to associate guest memory addresses (Pg) to physical host addresses (Ph).
  • FIG. 4 is an illustration showing guest OS memory mapped to host memory in a system that supports virtualization.
  • guest G1 address space 410 is 128 MB and guest G1 address space 405 is 256 MB.
  • Physical memory address space for guest OS G1 (P g i) and guest OS G2 (P g2 ) are shown mapped to host physical address space (PM) and (P h 2), respectively.
  • guest memory address 415 is mapped to host address 420
  • guest address 425 is mapped to 430.
  • FIG. 4 It should be appreciated that the address locations and sizes shown in FIG. 4 are provided as examples, not necessarily actual memory addresses. Furthermore, the exemplary memory illustrated in FIG. 4 do not represent or imply any limitations to the present disclosure. [0036] It should be appreciated that while various embodiments and aspects of the present disclosure are discussed in the context of a graphics device assigned to a virtual machine and associating a memory address thereof with a physical host address, various aspects and embodiments of the present disclosure may encompass other types of devices. That is, various aspects and embodiments herein may include devices, systems, and subsystems other than a graphics device. For example, an I/O device having a local processor and local memory and forming part of a virtual machine or system that supports virtualization may also benefit from the systems and methods of the present disclosure.
  • FIG. 5 provides an illustrative depiction of a two-step translation process 500 used to map a graphics aperture address (Pa) of a graphics device to a physical guest address (Pg) that is backed up by an actual physical host memory address (Ph) located in main memory of a host system.
  • graphics aperture address (Pa) is translated to a guest physical address (Pg) by chipset 110.
  • Chipset 110 uses a graphics memory translation table mechanism such as a GTT and page table entries (PTEs) to map the graphics aperture address (Pa) to the physical guest address (Pg).
  • PTEs page table entries
  • the guest physical address (Pg) must still be associated with a host physical address (Ph) since the host physical hardware is where the actual physical memory locations exist.
  • a DMA remap mechanism of operation 510 provides the translation from the physical guest address (Pg) to the physical host address (Ph).
  • DMA re-map mechanism 510 may be implemented in hardware or software.
  • FIG. 6 provides an exemplary illustration of a process 605 to map a graphics memory (e.g., aperture) address (Pa) of a guest, virtual machine to a physical host address (Ph) in a single process or operation.
  • a graphics translation table is provided that may use PTEs for mapping a guest graphics memory address (Pa) to physical host address (Ph).
  • the graphics memory translation table may be a GART or a GTT. Furthermore, the graphics memory translation table may be implemented in a chipset, such as, for example, chipset 110 shown in FIG. 1.
  • FIG. 7 is an exemplary flow diagram of a process 700, according to some embodiments herein.
  • FIG. 8 may be referenced in conjunction with FIG. 7 for a better understanding of FIG. 7 and the discussion thereof.
  • FIG. 8 is an exemplary depiction of a logical memory map, generally represented by reference number 800, illustrating a guest graphics memory address (Pa) to host physical address (Ph) translation, facilitated by a PTE to a graphics memory translation table 810.
  • the graphics memory translation table 810 (e.g., a GTT) is used to facilitate a translation or mapping of guest graphics memory address (Pa) 805 to host physical address (Ph) 815.
  • the translation may be facilitated by DMA remapping.
  • a request for memory is made for a graphics device (or other I/O device) assigned to a virtual machine (i.e., guest).
  • a mapping or translation of the guest graphics memory address (Pa) to the host physical address (Ph) is provided.
  • the host physical address Ph is an actual address location of physical memory in system memory of the physical host hardware 410.
  • the guest graphics memory address (Pa) may be mapped to the host physical address (Ph) in GTT 810 using a DMA remapping technique.
  • the graphics device and more generally an I/O device including a processor and local memory that is assigned or supported by virtualization, uses host physical addresses (815) to DMA data from main memory 125 of physical host hardware 330.
  • the process of using a graphics memory translation table 810 to map a guest graphics memory address (Pa) to a host physical address (Ph) may be implemented in software. In some embodiments herein, the process of using graphics memory translation table 810 to map the guest graphics memory address (Pa) to the host physical address (Ph) may be implemented in hardware.
  • the hardware mechanism may use page table translation logic in chipset hardware.
  • a guest OS driver and a VMM cooperate to manage entries to the graphics memory translation table.
  • the guest OS driver provides a physical host address (Ph) directly to the graphics memory translation table.
  • the guest OS driver queries the VMM for a valid physical host address.
  • the VMM provides a valid (e.g., available) host physical address (Ph). That is, the VMM only returns a host physical address that is valid for use by the guest OS.
  • a guest OS driver and a VMM cooperate to manage entries to the graphics memory translation table in which the guest OS driver is aware of the guest-to-host mapping.
  • Graphics memory translation table 810 may be read-only for the guest OS. Accordingly, the guest OS can only read from GTT 810.
  • the guest OS driver may attempt to write a guest graphics memory address to the GTT 810.
  • the VMM may provide a service to validate physical host addresses prior to entering the validated physical host addresses (Ph) into the graphics memory translation table. That is, the VMM validates the physical host addresses and enters the validated physical host addresses into the graphics memory translation table.
  • VMM validation and entry of the physical host addresses into GTT 810 may be provided to offer a level of security to safeguard against a guest OS from accessing a physical host address needed by, for example, another guest OS.
  • writes or installs of physical host memory to GTT 810 may be accomplished as part of a batch process. Accordingly, system resource overhead may be amortized.
  • the VMM sets up the page tables used in the graphics memory translation table, GTT 810.
  • the guest OS is unaware of the guest-to-host mapping functionality or process herein. Any writes intended for GTT 810 by a guest OS driver are captured by the VMM (e.g., 330).
  • the VMM translates a guest graphics memory address (Pa) into a corresponding physical host address (Ph) and installs the corresponding host physical address (Ph) in the graphics memory translation table.
  • Pa guest graphics memory address
  • Ph physical host address
  • the guest OS cannot access or monitor the physical host addresses of other VMs.
  • the functionality of process 700 is implemented in hardware.
  • An address decoder or other hardware devices may be used to detect writes to GTT 810.
  • the hardware may make a query to determine a valid physical host address and install valid physical host address entries to the GTT.
  • the guest OS need not be altered or modified since the functionality of the guest-to-host mapping (i.e., translation) does not depend on the OS driver.
  • a measure of security is provided since the guest OS driver cannot set-up a GTT to map an entry from other VMM's physical memory (i.e., a guest OSS cannot snoop on the activity of another VM's memory).
  • a single translation operation may be provided to translate the guest graphics memory address to host physical address.
  • the number and frequency of translating needed for an I/O device such as, for example, a graphics device, in a system that supports virtualization may be reduced.
  • a graphics memory translation table may be utilized to map a guest graphics aperture address to a host physical address. Accordingly, a need for separate DMA remapping and the associated hardware costs and/or changes to translate a guest address to a host address may be reduced or eliminated. Additionally, hardware and/or software (e.g., a chipset) implementing or including the translation functionality disclosed herein may be, in some embodiments, generalized and used in systems that support virtualization. [0051] In some embodiments herein, the graphics may include at least one chipset address remapping method in addition to the remapping methods disclosed hereinabove. For example, an I/O Memory Management Unit (not shown) may be included.

Abstract

A system and method including, in some embodiments, receiving a request for a graphics memory address for an input/output (I/O) device assigned to a virtual machine in a system that supports virtualization, and installing, in a graphics memory translation table, a physical guest graphics memory address to host physical memory address translation.

Description

ONE STEP ADDRESS TRANSLATION OF GRAPHICS ADDRESSES IN VIRTUALIZATION
BACKGROUND
[0001] A graphics device may require a contiguous memory range to store graphics data. The graphics data may include, for example, three-dimensional data and instructions used by, for example, a graphics subsystem in a personal computer. The contiguous graphics memory may be correlated to a physical memory using a graphics translation table. However, the physical memory may not be contiguous.
[0002] Virtualization, in some aspects, is a technique to operate one or more guest operating systems (OS) on the same host hardware by presenting multiple copies of a host hardware-software interface to each guest OS. In a system that supports virtualization, a memory address for the guest may be correlated to host physical memory address. The physical host memory may back up the physical guest memory.
[0003] Furthermore, in a system having a graphics device and supporting virtualization, it may be necessary to provide two memory address translations. One translation to map a graphics address to a physical guest address and a second translation to translate the physical guest address to a physical host address.
[0004] Thus, there exists a need in the technology to provide a system and method to efficiently translate graphics addresses in a context that supports virtualization.
BRIEF DESCRIPTION OF THE DRAWINGS
[0005] FIG. 1 is block diagram of a computer system, according to some embodiments hereof; [0006] FIG. 2 is an exemplary depiction of a graphics translation table, according to some embodiments herein;
[0007] FIG. 3 is an exemplary depiction of a virtual machine, according to some embodiments herein; [0008] FIG. 4 is an exemplary depiction of physical memory mapping on a virtualized machine, in accordance with some embodiments herein;
[0009] FIG. 5 is a flow diagram of a memory translation including two translations; [0010] FIG. 6 is an exemplary memory translation scheme including one translation operation, according to some embodiments herein;
[0011] FIG. 7 is an exemplary flow diagram of a process, according to some embodiments herein; and
[0012] FIG. 8 is an exemplary depiction of a logical memory map illustrating a guest graphics memory address to host physical address translation, in accordance with some embodiments herein.
DETAILED DESCRIPTION
[0013] FIG. 1 is an exemplary block diagram illustrating physical hardware of a computer system that may be used in some embodiments hereof, generally represented by reference number 100. Computer system 100 includes a processor 105. Processor 105 represents a central processing unit of any type of architecture. Some embodiments hereof may be implemented by computer systems having multiple processors. Computer system 100 includes a chipset 110 that has a number of control circuits and a number of interface circuits to allow processor 105 to access a system memory 115, a system bus 150 (e.g., a Peripheral Component Interconnect (PCI) Express bus), and a graphics device 160. [0014] A number of peripheral component interconnect (PCI Express) devices 155i through 155n are connected to Express bus 150, as defined by the PCI Special Interest Group (PCI-SIG) in "PCI Express Base Specification, Version 1.0" (July 2002). PCI Express devices 155i through 155n are input/output (I/O) hardware devices such as, for example, a disk controller/card, a local area network controller/card, etc. [0015] In some embodiments, chipset 110 may have one or more bus controllers (e.g., PCI Express bus), a graphics controller, a CPU controller, and a memory controller to control data access between various components such as, for example, bus 150, graphics device 160, system memory 115, processor 105, etc.
[0016] It should be appreciated that computer system 100 may include other, additional, or fewer components than those illustrated in FIG. 1 , without departing or altering the scope of the various embodiments herein.
[0017] System memory 115 represents one or more mechanisms for storing information. For example, system memory 115 may include non-volatile or volatile memories. In some embodiments, system memory 115 includes a graphics memory unit 118, a graphics aperture 120 and a main memory 125. Main memory 125 may include an operating system (OS) 130, a memory manager 135, a graphics memory translation table 140, and additional main memory 145 allocated for other information such as, for example, other programs and data. [0018] Graphics device 160 may be an add-in device or integrated into computer system 100. In some embodiments, graphics device 160 includes a graphics processor 165 and a graphics local memory 170. The memory may be random access memory (RAM) (e.g., extended data out dynamic random access memory (EDO), synchronous graphic random access memory (SGRAM), video random access memory (VRAM)). The memory is included, since the video card must be able to remember a complete screen image at any time, and maintain local copies of graphics programs, and graphics objects like triangles, and textures. It is noted that some embodiments herein are also applicable to graphics devices and I/O devices having no local memory. [0019] Graphics processor 160 performs graphics functions, such as, 3-D rendering operations, drawings, etc. Graphics processor 165 has access to its own graphics local memory 160. Graphics device 160 may be coupled to chipset 110 via accelerated graphics port (AGP) 175. AGP 175 provides a high-speed bus for moving data directly from system memory 115 to graphics device 160. Direct references may be made to system memory 115. Due to the direct references to system memory 115, a contiguous view of system memory 115 may be essential for efficient transfer of information between graphics device 160 and system memory 115. [0020] In some embodiments, graphics device 160 may be coupled to chipset via PCI express bus 150.
[0021] In some embodiments, a range of system memory 115 is reserved for graphics memory unit 118 including graphics aperture 120. Graphics aperture 120 provides a range of memory addresses used by AGP 175 for graphics information such as, for example, 3-D features and textures. However, since system memory 115 is dynamically allocated for graphics data, it is necessary to provide a graphics mapping table mechanism to map random segments of system memory 115 into a single contiguous, physical space for graphics aperture 120. [0022] A graphics memory translation table such as a Graphics Address Re- Mapping Table (GART) or a Graphics Translation Table (GTT) may be used to provide a physically-contiguous view of scattered pages in system memory for direct memory access (DMA) transfers. With AGP 175, main memory is specifically used for advanced three-dimensional features, such as textures, alpha buffers, and ZBuffers. As mentioned above, since the AGP generates direct references into system memory, a contiguous view of that space is essential. However, since system memory is dynamically allocated in, for example, random 4K pages, it may be necessary to provide an address mapping mechanism that maps random 4K pages into a single contiguous, physical address space. [0023] FIG. 2 illustrates an exemplary graphics memory translation table map, generally represented by reference number 200. In some embodiments, system memory 115 includes a main memory 125 having an address range from address zero (0) to a top thereof. Graphics aperture 120 may have an address range that begins from the top of main memory 125. Graphics aperture 120 is a virtual memory and maps into the physical address space of main memory 125.
[0024] Graphics aperture 120 is a portion of system memory 115 that is allocated by operating system 130 for use by graphics device 160. Graphics device 160 is provided access to the reserved graphics aperture 120 to store texture data, front buffer data or other graphics data for faster graphics data processing. Each address Pa in graphics aperture 120 has a corresponding entry mapped into a physical address space Pg of main memory 125. [0025] Operating system 130 allocates pages in main memory 125 (of system memory 115) wherever they are found and assigns them for graphics device 160. Graphics device 160 is therefore provided with a continuous block of graphics aperture 120, wherein references pointing from a graphics aperture 120 address Pa to a corresponding address Pg in main memory 125 are stored in graphics memory translation table 140 as a page table entry (PTE). [0026] Virtualization, in some aspects, is a technique to operate one or more guest operating systems (OS) on the same native hardware by presenting multiple copies of a host hardware-software interface to each guest OS. The native hardware may be referred to as the host. The multiple guest OSs may even run concurrently. In a virtualization context, managing memory used by a graphics device or subsystem is further complicated because of a need to translate guest memory addresses to physical host addresses due to the virtualization of the host hardware.
[0027] FIG. 3 illustrates an exemplary system 300 that supports virtualization. System 300 includes physical host hardware 305, a number of guest virtual machines VMo 31Oo and VMi 310-i, and a virtual machine monitor (VMM) 330. Each of VM0 31O0 and VMi 31 Oi may be referred to as a guest herein relative to physical host hardware 305. In some embodiments, physical host hardware 305 may include a computer system and/or components thereof similar to system 100 illustrated in FIG. 1. Physical host hardware 305 may also be referred to herein as the host.
[0028] It should be appreciated that the particular physical host hardware included in host 305 may be varied, much as system 100 may be altered as stated in conjunction with the discussion of FIG. 1 , without departing from and/or altering the scope of the various embodiments herein. [0029] In some embodiments, physical host hardware 305 may include a processor, a memory, various I/O devices (e.g., keyboard, monitor, USB controller, network controller etc.), and a graphics device. [0030] Two virtual machines VM0 31Oo and VM1 31 Oi are shown in system 300. It should be appreciated that more or fewer virtual machines may be included or supported in system 300. VMo 31Oo and VMi 31 Oi are depicted as including similar components. For example, VM0 31O0 has a number of addresses 315, a guest OS G1 320, and components 325 associated therewith. In a similar manner, VMi 31O1 has a number of addresses 317, a guest OS G2 322, and virtual components 327. It should be appreciated by those in the art that differences in functionality may exist between the virtual machine(s) supported by system 300.
[0031] In some embodiments, VMM 330 provides a hardware-software interface to each of VMo 31Oo and VMi 310-|. Each instance of the hardware- software interface may provide an efficient replica of host 305, including the processing, memory, instructions, and other resources thereof (e.g., memory and I/O devices) to the virtual machines connected to or supported by system 300. In some embodiments, guest OS G1 320 and guest OS G2 322 may operate concurrently, in part due to each guest OS operating in its own virtual machine. [0032] In a system that supports virtualization, a guest memory address needs to be translated or mapped to a host physical address since the physical guest memory address (Pg) is a virtual memory address that is actually located at a physical host address (Ph). That is, since the guest's resources, including memory, are virtualizations of hardware or the entire hardware environment of the host, a correlation must be established to associate guest memory addresses (Pg) to physical host addresses (Ph).
[0033] Thus, in a virtualized system or a system that supports virtualization, it may be necessary to translate a graphics address of a virtual, guest machine to an actual physical host address location. In accordance with some embodiments herein, there is provided a method and a system to efficiently translate graphics addresses in a context that supports virtualization.
[0034] In a computing system that supports virtualization, a guest memory address, Pg, may be backed up by (i.e, located at) a physical host address Ph. FIG. 4 is an illustration showing guest OS memory mapped to host memory in a system that supports virtualization. As illustrated, guest G1 address space 410 is 128 MB and guest G1 address space 405 is 256 MB. Physical memory address space for guest OS G1 (Pgi) and guest OS G2 (Pg2) are shown mapped to host physical address space (PM) and (Ph2), respectively. For example, guest memory address 415 is mapped to host address 420, and guest address 425 is mapped to 430. [0035] It should be appreciated that the address locations and sizes shown in FIG. 4 are provided as examples, not necessarily actual memory addresses. Furthermore, the exemplary memory illustrated in FIG. 4 do not represent or imply any limitations to the present disclosure. [0036] It should be appreciated that while various embodiments and aspects of the present disclosure are discussed in the context of a graphics device assigned to a virtual machine and associating a memory address thereof with a physical host address, various aspects and embodiments of the present disclosure may encompass other types of devices. That is, various aspects and embodiments herein may include devices, systems, and subsystems other than a graphics device. For example, an I/O device having a local processor and local memory and forming part of a virtual machine or system that supports virtualization may also benefit from the systems and methods of the present disclosure. As an example, a memory address for a virtual machine I/O device having a processor and a local memory, similar to but not necessarily a graphics device or graphics subsystem, may be used with and/or include aspects of the present disclosure. [0037] FIG. 5 provides an illustrative depiction of a two-step translation process 500 used to map a graphics aperture address (Pa) of a graphics device to a physical guest address (Pg) that is backed up by an actual physical host memory address (Ph) located in main memory of a host system. For example, graphics aperture address (Pa) is translated to a guest physical address (Pg) by chipset 110. Chipset 110 uses a graphics memory translation table mechanism such as a GTT and page table entries (PTEs) to map the graphics aperture address (Pa) to the physical guest address (Pg). However, the guest physical address (Pg) must still be associated with a host physical address (Ph) since the host physical hardware is where the actual physical memory locations exist. A DMA remap mechanism of operation 510 provides the translation from the physical guest address (Pg) to the physical host address (Ph). DMA re-map mechanism 510 may be implemented in hardware or software. [0038] In accordance with some embodiments herein, FIG. 6 provides an exemplary illustration of a process 605 to map a graphics memory (e.g., aperture) address (Pa) of a guest, virtual machine to a physical host address (Ph) in a single process or operation. Per operation 605, a graphics translation table is provided that may use PTEs for mapping a guest graphics memory address (Pa) to physical host address (Ph).
[0039] In some embodiments, the graphics memory translation table may be a GART or a GTT. Furthermore, the graphics memory translation table may be implemented in a chipset, such as, for example, chipset 110 shown in FIG. 1. [0040] FIG. 7 is an exemplary flow diagram of a process 700, according to some embodiments herein. FIG. 8 may be referenced in conjunction with FIG. 7 for a better understanding of FIG. 7 and the discussion thereof. FIG. 8 is an exemplary depiction of a logical memory map, generally represented by reference number 800, illustrating a guest graphics memory address (Pa) to host physical address (Ph) translation, facilitated by a PTE to a graphics memory translation table 810. The graphics memory translation table 810 (e.g., a GTT) is used to facilitate a translation or mapping of guest graphics memory address (Pa) 805 to host physical address (Ph) 815. The translation may be facilitated by DMA remapping.
[0041] At operation 705, in a system that supports virtualization, a request for memory is made for a graphics device (or other I/O device) assigned to a virtual machine (i.e., guest). [0042] At operation 710, a mapping or translation of the guest graphics memory address (Pa) to the host physical address (Ph) is provided. The host physical address Ph is an actual address location of physical memory in system memory of the physical host hardware 410. The guest graphics memory address (Pa) may be mapped to the host physical address (Ph) in GTT 810 using a DMA remapping technique. The graphics device, and more generally an I/O device including a processor and local memory that is assigned or supported by virtualization, uses host physical addresses (815) to DMA data from main memory 125 of physical host hardware 330.
[0043] In some embodiments herein, the process of using a graphics memory translation table 810 to map a guest graphics memory address (Pa) to a host physical address (Ph) may be implemented in software. In some embodiments herein, the process of using graphics memory translation table 810 to map the guest graphics memory address (Pa) to the host physical address (Ph) may be implemented in hardware. The hardware mechanism may use page table translation logic in chipset hardware.
[0044] Regarding some embodiments including a software implementation of the process of using a graphics memory translation table 810 to map a guest graphics memory address (Pa) to a host physical address (Ph), a guest OS driver and a VMM (e.g., 330) cooperate to manage entries to the graphics memory translation table. The guest OS driver provides a physical host address (Ph) directly to the graphics memory translation table. Prior to installing the host physical address (Ph) in the graphics memory translation table, the guest OS driver queries the VMM for a valid physical host address. In response to the query by the guest OS, the VMM provides a valid (e.g., available) host physical address (Ph). That is, the VMM only returns a host physical address that is valid for use by the guest OS. The guest OS driver then installs the valid physical host address (Ph) in the GTT. [0045] In some software implemented embodiments, a guest OS driver and a VMM cooperate to manage entries to the graphics memory translation table in which the guest OS driver is aware of the guest-to-host mapping. Graphics memory translation table 810 may be read-only for the guest OS. Accordingly, the guest OS can only read from GTT 810. In these embodiments, the guest OS driver may attempt to write a guest graphics memory address to the GTT 810. The VMM may provide a service to validate physical host addresses prior to entering the validated physical host addresses (Ph) into the graphics memory translation table. That is, the VMM validates the physical host addresses and enters the validated physical host addresses into the graphics memory translation table. The VMM validation and entry of the physical host addresses into GTT 810 may be provided to offer a level of security to safeguard against a guest OS from accessing a physical host address needed by, for example, another guest OS. [0046] In some software implemented embodiments herein, writes or installs of physical host memory to GTT 810 may be accomplished as part of a batch process. Accordingly, system resource overhead may be amortized.
[0047] In some software implemented embodiments herein, the VMM sets up the page tables used in the graphics memory translation table, GTT 810. In these embodiments, the guest OS is unaware of the guest-to-host mapping functionality or process herein. Any writes intended for GTT 810 by a guest OS driver are captured by the VMM (e.g., 330). The VMM translates a guest graphics memory address (Pa) into a corresponding physical host address (Ph) and installs the corresponding host physical address (Ph) in the graphics memory translation table. In these embodiments, it is not necessary to alter, adapt, or modify the guest OS driver to accommodate guest-to-host translations since the OS driver does not translate or make entries to the GTT. Additionally, the guest OS cannot access or monitor the physical host addresses of other VMs. [0048] In some embodiments herein, the functionality of process 700 is implemented in hardware. An address decoder or other hardware devices may be used to detect writes to GTT 810. Upon detection of an attempted write to the GTT by a guest OS driver, the hardware may make a query to determine a valid physical host address and install valid physical host address entries to the GTT. Hereto, as in some of the software implementation embodiments discussed above, the guest OS need not be altered or modified since the functionality of the guest-to-host mapping (i.e., translation) does not depend on the OS driver. As in some software embodiments discussed above, a measure of security is provided since the guest OS driver cannot set-up a GTT to map an entry from other VMM's physical memory (i.e., a guest OSS cannot snoop on the activity of another VM's memory).
[0049] Accordingly, a single translation operation may be provided to translate the guest graphics memory address to host physical address. In this manner, the number and frequency of translating needed for an I/O device such as, for example, a graphics device, in a system that supports virtualization may be reduced.
[0050] In accordance with the hereinabove disclosure, a graphics memory translation table may be utilized to map a guest graphics aperture address to a host physical address. Accordingly, a need for separate DMA remapping and the associated hardware costs and/or changes to translate a guest address to a host address may be reduced or eliminated. Additionally, hardware and/or software (e.g., a chipset) implementing or including the translation functionality disclosed herein may be, in some embodiments, generalized and used in systems that support virtualization. [0051] In some embodiments herein, the graphics may include at least one chipset address remapping method in addition to the remapping methods disclosed hereinabove. For example, an I/O Memory Management Unit (not shown) may be included.
[0052] The several embodiments described herein are solely for the purpose of illustration. The various features described herein need not all be used together, and any one or more of those features may be incorporated in a single embodiment. Therefore, persons in the art will recognize from this description that other embodiments may be practiced with various modifications and alterations.

Claims

WHAT IS CLAIMED IS:
1. A method comprising: receiving a request for a guest graphics memory address for an input/output (I/O) device assigned to a virtual machine in a system that supports virtualization; and installing, in a graphics memory translation table, a physical guest graphics memory address to host physical memory address translation.
2. The method of claim 1 , wherein the graphics memory translation table comprises at least one of: a graphics translation table and a graphics address remapping table.
3. The method of claim 1 , wherein the I/O device comprises a graphics device.
4. The method of claim 1 , wherein the I/O device comprises a processor or a controller.
5. The method of claim 4, wherein the I/O device comprises a local memory associated with the processor or the controller thereof.
6. The method of claim 1 , wherein the translation table includes at least one additional chipset address remapping method.
7. The method of claim 1 , further comprising performing the installing in a batch operation.
8. The method of claim 1 , wherein a guest operating system (OS) driver facilitates entry of a host address into the graphics memory translation table.
9. The method of claim 1 , wherein a virtual machine monitor (VMM) facilitates entry of a host address into the graphics memory translation table.
10. The method of claim 9, wherein a guest operating system (OS) driver invokes the VMM for installing the translation in the translation table.
11. The method of claim 9, wherein the VMM traps an update to the graphics translation table.
12. The method of claim 9, further comprising validating, at least, an availability and an assignment of the host physical address by the VMM.
13. The method of claim 9, wherein the graphics memory translation table is read-only for a guest OS driver.
14. The method of claim 1 , wherein the installing is facilitated by hardware mechanism in a chipset.
15. The method of claim 14, facilitated by an address decoder in the chipset.
16. A system comprising: an input/output (I/O) device assigned to a virtual machine in a system that supports virtualization; a graphics memory translation table to associate a physical guest graphics memory address with a host physical address; a graphics memory unit; and a memory associated with the graphics memory unit and storing instructions that when executed by a machine result in the following: installing, in the graphics memory translation table, a physical guest graphics memory address to host physical memory address translation.
17. The system of claim 16, wherein the graphics memory translation table comprises at least one of: a graphics translation table and a graphics address remapping table.
18. The system of claim 16, wherein the I/O device comprises a graphics device.
19. The system of claim 16, wherein I/O device comprises a processor or a controller.
20. The system of claim 19, wherein the I/O device comprises a local memory associated with the processor or the controller thereof.
21. The system of claim 16, further comprising a virtual machine monitor
(VMM) to facilitate entry of a host address into the graphics memory translation table.
22. The system of claim 21 , wherein the VMM validates, at least, an availability and an assignment of the host physical address.
23. The system of claim 21 , wherein the graphics memory translation table is read-only for a guest OS driver.
24. The system of claim 16, further comprising a guest operating system
(OS) driver to facilitate entry of a host address into the graphics memory translation table.
25. The system of claim 16, wherein the installing is performed in a batch operation.
26. A system comprising: an input/output (I/O) device assigned to a virtual machine in a system that supports virtualization; memory to store a graphics memory translation table associating a physical guest graphics memory address with a host physical address; and a graphics memory unit storing instructions that when executed by a machine result in the following: installing, in the graphics memory translation table, a physical guest graphics memory address to host physical memory address translation.
27. The system of claim 26, wherein the graphics memory translation table comprises at least one of: a graphics translation table and a graphics address remapping table. 8. The system of claim 26, wherein the I/O device comprises a graphics device.
PCT/US2005/047683 2004-12-29 2005-12-29 One step address translation of graphics addresses in virtualization WO2006072101A2 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN2005800448642A CN101088078B (en) 2004-12-29 2005-12-29 One step address translation method and system for graphics addresses in virtualization
JP2007549696A JP5006798B2 (en) 2004-12-29 2005-12-29 One-step address conversion of virtualized graphics address
EP19198130.7A EP3605342A1 (en) 2004-12-29 2005-12-29 One step address translation of graphics addresses in virtualization
EP05856135A EP1839158A2 (en) 2004-12-29 2005-12-29 One step address translation of graphics addresses in virtualization

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US11/025,126 US9058292B2 (en) 2004-12-29 2004-12-29 System and method for one step address translation of graphics addresses in virtualization
US11/025,126 2004-12-29

Publications (2)

Publication Number Publication Date
WO2006072101A2 true WO2006072101A2 (en) 2006-07-06
WO2006072101A3 WO2006072101A3 (en) 2006-11-23

Family

ID=36481388

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2005/047683 WO2006072101A2 (en) 2004-12-29 2005-12-29 One step address translation of graphics addresses in virtualization

Country Status (7)

Country Link
US (4) US9058292B2 (en)
EP (3) EP3073385A1 (en)
JP (1) JP5006798B2 (en)
KR (2) KR20070086791A (en)
CN (2) CN101923520B (en)
TW (1) TWI300541B (en)
WO (1) WO2006072101A2 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11243797B2 (en) 2017-01-31 2022-02-08 Opensynergy Gmbh Method for operating a control device, control device and computer program product

Families Citing this family (40)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7886126B2 (en) 2005-01-14 2011-02-08 Intel Corporation Extended paging tables to map guest physical memory addresses from virtual memory page tables to host physical memory addresses in a virtual machine system
US8200796B1 (en) * 2005-05-05 2012-06-12 Digital Display Innovations, Llc Graphics display system for multiple remote terminals
US8019883B1 (en) 2005-05-05 2011-09-13 Digital Display Innovations, Llc WiFi peripheral mode display system
US11733958B2 (en) 2005-05-05 2023-08-22 Iii Holdings 1, Llc Wireless mesh-enabled system, host device, and method for use therewith
US7499057B2 (en) * 2005-09-09 2009-03-03 Microsoft Corporation Address translation in an integrated graphics environment
US7555628B2 (en) 2006-08-15 2009-06-30 Intel Corporation Synchronizing a translation lookaside buffer to an extended paging table
KR100889602B1 (en) * 2006-12-05 2009-03-20 한국전자통신연구원 Apparatus and method of ray-triangle collision detection for ray-tracing
US7685401B2 (en) * 2006-12-27 2010-03-23 Intel Corporation Guest to host address translation for devices to access memory in a partitioned system
US20080229053A1 (en) * 2007-03-13 2008-09-18 Edoardo Campini Expanding memory support for a processor using virtualization
US8893126B2 (en) * 2008-02-01 2014-11-18 International Business Machines Corporation Binding a process to a special purpose processing element having characteristics of a processor
US8087269B2 (en) 2008-02-07 2012-01-03 Checkpoint Systems, Inc. Cable wrap security device
US20110102443A1 (en) * 2009-11-04 2011-05-05 Microsoft Corporation Virtualized GPU in a Virtual Machine Environment
KR101620058B1 (en) 2009-11-23 2016-05-24 삼성전자주식회사 Apparatus for switching screen between virtual machines and method thereof
US9672583B2 (en) * 2011-12-21 2017-06-06 Intel Corporation GPU accelerated address translation for graphics virtualization
US9317441B2 (en) 2011-12-22 2016-04-19 Intel Cororation Indexed page address translation to reduce memory footprint in virtualized environments
TWI510056B (en) * 2011-12-27 2015-11-21 Hon Hai Prec Ind Co Ltd 3d imaging module and 3d imaging method
CN102567217B (en) * 2012-01-04 2014-12-24 北京航空航天大学 MIPS platform-oriented memory virtualization method
US9727345B2 (en) 2013-03-15 2017-08-08 Intel Corporation Method for booting a heterogeneous system and presenting a symmetric core view
WO2015000106A1 (en) 2013-07-01 2015-01-08 Intel Corporation Efficient graphics virtualization with address ballooning
US8719374B1 (en) 2013-09-19 2014-05-06 Farelogix, Inc. Accessing large data stores over a communications network
US9792222B2 (en) 2014-06-27 2017-10-17 Intel Corporation Validating virtual address translation by virtual machine monitor utilizing address validation structure to validate tentative guest physical address and aborting based on flag in extended page table requiring an expected guest physical address in the address validation structure
US10353680B2 (en) * 2014-07-25 2019-07-16 Intel Corporation System converter that implements a run ahead run time guest instruction conversion/decoding process and a prefetching process where guest code is pre-fetched from the target of guest branches in an instruction sequence
US11281481B2 (en) 2014-07-25 2022-03-22 Intel Corporation Using a plurality of conversion tables to implement an instruction set agnostic runtime architecture
CN107111498B (en) * 2014-11-12 2020-12-29 英特尔公司 Live migration of virtual machines from/to a host computer using graphical virtualization
KR101751629B1 (en) 2014-12-24 2017-06-27 인텔 코포레이션 Hybrid on-demand graphics translation table shadowing
KR101672119B1 (en) * 2015-02-16 2016-11-04 (주) 퓨전데이타 Server Device and Method for outputting High Definition Video in a Virtual Environment
WO2016149935A1 (en) * 2015-03-26 2016-09-29 Intel Corporation Computing methods and apparatuses with graphics and system memory conflict check
WO2016205975A1 (en) * 2015-06-26 2016-12-29 Intel Corporation Apparatus and method to improve scalability of graphics processor unit (gpu) virtualization
US10055242B2 (en) 2015-10-16 2018-08-21 Microsoft Technology Licensing, Llc Virtualizing audio decoding hardware
US10853118B2 (en) * 2015-12-21 2020-12-01 Intel Corporation Apparatus and method for pattern-driven page table shadowing for graphics virtualization
EP3188014B1 (en) * 2015-12-29 2022-07-13 Dassault Systèmes Management of a plurality of graphic cards
WO2018053829A1 (en) * 2016-09-26 2018-03-29 Intel Corporation Apparatus and method for hybrid layer of address mapping for virtualized input/output implementation
WO2018119712A1 (en) * 2016-12-27 2018-07-05 深圳前海达闼云端智能科技有限公司 Video display method, device, electronic apparatus, and computer program product
JP2020503609A (en) * 2016-12-27 2020-01-30 深▲せん▼前海達闥雲端智能科技有限公司Cloudminds (Shenzhen) Robotics Systems Co., Ltd. Memory access method, apparatus and electronic equipment for multi-operating system
US10733112B2 (en) 2017-06-16 2020-08-04 Alibaba Group Holding Limited Method and apparatus for hardware virtualization
US11469953B2 (en) 2017-09-27 2022-10-11 Intel Corporation Interworking of legacy appliances in virtualized networks
EP3704602B1 (en) * 2017-10-24 2023-09-06 INTEL Corporation Hardware assisted virtual switch
CN110874258B (en) * 2018-08-31 2023-05-26 阿里巴巴集团控股有限公司 Physical machine changing method and device
US11228561B2 (en) * 2019-04-23 2022-01-18 Red Hat, Inc. Safe MAC programming support for vNIC
US10872458B1 (en) * 2019-09-06 2020-12-22 Apple Inc. Graphics surface addressing

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4456954A (en) * 1981-06-15 1984-06-26 International Business Machines Corporation Virtual machine system with guest architecture emulation using hardware TLB's for plural level address translations
US20010028355A1 (en) * 1997-06-25 2001-10-11 Porterfield A. Kent Apparatus for graphic address remapping
US20020082824A1 (en) * 2000-12-27 2002-06-27 Gilbert Neiger Virtual translation lookaside buffer

Family Cites Families (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5454536A (en) * 1977-10-08 1979-04-28 Fujitsu Ltd Data processor
JPS608971A (en) * 1983-06-29 1985-01-17 Toshiba Corp Central processing unit
JPH0658650B2 (en) * 1986-03-14 1994-08-03 株式会社日立製作所 Virtual computer system
US5577231A (en) * 1994-12-06 1996-11-19 International Business Machines Corporation Storage access authorization controls in a computer system using dynamic translation of large addresses
US6073226A (en) * 1997-03-19 2000-06-06 Microsoft Corporation System and method for minimizing page tables in virtual memory systems
US5933158A (en) * 1997-09-09 1999-08-03 Compaq Computer Corporation Use of a link bit to fetch entries of a graphic address remapping table
US6112285A (en) * 1997-09-23 2000-08-29 Silicon Graphics, Inc. Method, system and computer program product for virtual memory support for managing translation look aside buffers with multiple page size support
US6192455B1 (en) * 1998-03-30 2001-02-20 Intel Corporation Apparatus and method for preventing access to SMRAM space through AGP addressing
US6362826B1 (en) * 1999-01-15 2002-03-26 Intel Corporation Method and apparatus for implementing dynamic display memory
US6370633B2 (en) * 1999-02-09 2002-04-09 Intel Corporation Converting non-contiguous memory into contiguous memory for a graphics processor
JP2001051900A (en) * 1999-08-17 2001-02-23 Hitachi Ltd Information processing unit and processor for virtual computer system
US6970992B2 (en) * 1999-10-04 2005-11-29 Intel Corporation Apparatus to map virtual pages to disparate-sized, non-contiguous real pages and methods relating thereto
US6628294B1 (en) * 1999-12-31 2003-09-30 Intel Corporation Prefetching of virtual-to-physical address translation for display data
ES2312483T3 (en) 2000-07-14 2009-03-01 Irdeto Access B.V. ARCHITECTURE OF SECURE DATA DISSEMINATION BY PACKAGES.
US6725289B1 (en) * 2002-04-17 2004-04-20 Vmware, Inc. Transparent address remapping for high-speed I/O
US7953588B2 (en) * 2002-09-17 2011-05-31 International Business Machines Corporation Method and system for efficient emulation of multiprocessor address translation on a multiprocessor host
US6895491B2 (en) * 2002-09-26 2005-05-17 Hewlett-Packard Development Company, L.P. Memory addressing for a virtual machine implementation on a computer processor supporting virtual hash-page-table searching
US7069413B1 (en) * 2003-01-29 2006-06-27 Vmware, Inc. Method and system for performing virtual to physical address translations in a virtual machine monitor
US7243208B2 (en) * 2003-08-13 2007-07-10 Renesas Technology Corp. Data processor and IP module for data processor
US7334108B1 (en) * 2004-01-30 2008-02-19 Nvidia Corporation Multi-client virtual address translation system with translation units of variable-range size
US7213125B2 (en) * 2004-07-31 2007-05-01 Hewlett-Packard Development Company, L.P. Method for patching virtually aliased pages by a virtual-machine monitor
US20060112212A1 (en) * 2004-11-23 2006-05-25 Hob Gmbh & Co. Kg Virtual machine computer system for running guest operating system on a central processing means virtualized by a host system using region ID virtual memory option
US7243209B2 (en) 2005-01-27 2007-07-10 International Business Machines Corporation Apparatus and method for speeding up access time of a large register file with wrap capability

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4456954A (en) * 1981-06-15 1984-06-26 International Business Machines Corporation Virtual machine system with guest architecture emulation using hardware TLB's for plural level address translations
US20010028355A1 (en) * 1997-06-25 2001-10-11 Porterfield A. Kent Apparatus for graphic address remapping
US20020082824A1 (en) * 2000-12-27 2002-06-27 Gilbert Neiger Virtual translation lookaside buffer

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11243797B2 (en) 2017-01-31 2022-02-08 Opensynergy Gmbh Method for operating a control device, control device and computer program product

Also Published As

Publication number Publication date
JP5006798B2 (en) 2012-08-22
CN101923520B (en) 2012-11-28
US20200356490A1 (en) 2020-11-12
WO2006072101A3 (en) 2006-11-23
US20060139360A1 (en) 2006-06-29
US9058292B2 (en) 2015-06-16
TWI300541B (en) 2008-09-01
US10133674B2 (en) 2018-11-20
KR20090007494A (en) 2009-01-16
JP2008527508A (en) 2008-07-24
US20150301945A1 (en) 2015-10-22
US20190213138A1 (en) 2019-07-11
TW200634662A (en) 2006-10-01
EP3073385A1 (en) 2016-09-28
US10671541B2 (en) 2020-06-02
EP1839158A2 (en) 2007-10-03
KR20070086791A (en) 2007-08-27
CN101088078B (en) 2010-06-09
US11023385B2 (en) 2021-06-01
KR100955111B1 (en) 2010-04-28
CN101923520A (en) 2010-12-22
EP3605342A1 (en) 2020-02-05
CN101088078A (en) 2007-12-12

Similar Documents

Publication Publication Date Title
US11023385B2 (en) System and method for one step address translation of graphics addresses in virtualization
US9817770B2 (en) Memory address re-mapping of graphics data
CN111190752B (en) Method and device for sharing kernel memory of virtual machine
US8392628B2 (en) Sharing memory spaces for access by hardware and software in a virtual machine environment
US6880022B1 (en) Transparent memory address remapping
US6049854A (en) System and method for sharing physical memory among distinct computer environments
JPH0816479A (en) Memory-address space management
US11194735B2 (en) Technologies for flexible virtual function queue assignment
US7426625B2 (en) Data processing system and computer program product for support of system memory addresses with holes
US11494092B2 (en) Address space access control
US20220179677A1 (en) Memory protection for virtual machines

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 200580044864.2

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 2007549696

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 1020077014875

Country of ref document: KR

WWE Wipo information: entry into national phase

Ref document number: 2005856135

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 1020087030928

Country of ref document: KR