US20140164716A1 - Override system and method for memory access management - Google Patents
Override system and method for memory access management Download PDFInfo
- Publication number
- US20140164716A1 US20140164716A1 US13/960,739 US201313960739A US2014164716A1 US 20140164716 A1 US20140164716 A1 US 20140164716A1 US 201313960739 A US201313960739 A US 201313960739A US 2014164716 A1 US2014164716 A1 US 2014164716A1
- Authority
- US
- United States
- Prior art keywords
- memory
- memory management
- context
- fault
- access
- Prior art date
- Legal status (The legal status 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 status listed.)
- Abandoned
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/10—Address translation
- G06F12/1027—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
- G06F11/073—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in a memory management context, e.g. virtual memory or cache management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0793—Remedial or corrective actions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0614—Improving the reliability of storage systems
- G06F3/0619—Improving the reliability of storage systems in relation to data integrity, e.g. data losses, bit errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0673—Single storage device
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/68—Details of translation look-aside buffer [TLB]
- G06F2212/681—Multi-level TLB, e.g. microTLB and main TLB
Abstract
A memory management system and method are described. In one embodiment, a memory management system includes a memory management unit for virtualizing context memory storage and independently controlling access to the context memory without interference from other engine activities. The shared resource management unit overrides a stream of access denials (e.g., NACKs) associated with an access problem. The memory management system and method facilitate efficient and flexible access to memory while controlling translation between virtual and physical memory “spaces”. In one embodiment the memory management system includes a translation lookaside buffer and a fill component. The translation lookaside buffer tracks information associating a virtual memory space with a physical memory space. The fill component tracks the status of an access request progress from a plurality of engines independently and faults that occur in attempting to access a memory space
Description
- This is a continuation and claims the benefit of and priority to U.S. application Ser. No. 11/591,857 entitled “Override System and Method for Memory Access Management” filed on Nov. 1, 2006, which is incorporated herein by reference.
- The present invention relates to the field of information processing. In particular, the present invention relates to a system and method for shared resource access management.
- Electronic systems and circuits have made a significant contribution towards the advancement of modern society and are utilized in a number of applications to achieve advantageous results. Numerous electronic technologies such as digital computers, calculators, audio devices, video equipment, and telephone systems have facilitated increased productivity and reduced costs in analyzing and communicating data in most areas of business, science, education and entertainment. Electronic systems providing these advantageous results often include shared resources. However, designing and building systems that efficiently access and utilize a shared resource is often challenging. Coordinating the access to shared resources can be highly complex and resource intensive.
- Memory or storage is usually a very important component of a number of systems. Memories typically store information utilized by a system in performance of a number of different tasks. Other components of a system typically request access to a memory in order to retrieve (e.g., “read”) information from or forward (e.g., “write’) information to the memory. Different types of memories (e.g., bulk storage, main memory, removable memory etc.) and or memory “spaces” (e.g., virtual, physical, etc.) can be utilized to support information storage.
- Different types of memory can potentially offer different features. For example, different types of memories typically have different relative storage capacities and relatively different access speeds. Traditionally, systems that have relatively large storage capacity have relatively slow access speeds and systems that have relatively fast access speeds have relatively small storage capacities. For example, main system memories are relatively fast compared to bulk storage memories but typically store less information. A number of systems transfer chunks of information between relatively fast small memories and relatively slow bulk memories in an attempt to optimize speed and capacity.
- Traditional computer systems also often run programs that utilize virtual memory space addressing. Typically the computer system performs a translation between the virtual memory space addressing and physical memory space addressing. However, errors or faults can occur during the translation between virtual memory space addressing and physical memory space addressing. For example, when performing virtual to physical memory space translation a fault can occur when a translation map or a physical page is not resident. Faults or errors in storing information can cause problems for continued proper operation and/or accurate results.
- A memory management system and method are described. In one embodiment, a memory management system includes a memory management unit for virtualizing context memory storage and independently controlling access to the context memory without interference from other engine activities. The shared resource management unit overrides a stream of access denials (e.g., NACKs) associated with an access problem. The memory management system and method facilitate efficient and flexible access to memory while controlling translation between virtual and physical memory “spaces”. In one embodiment the memory management system includes a translationlookaside buffer and a fill component. The translation lookaside buffer tracks information associating a virtual memory space with a physical memory space. The fill component tracks the status of an access request progress from a plurality of engines independently and faults that occur in attempting to access a memory space.
- In one embodiment accesses to store the context information are immune to post-fault expediting responses. For example, a memory management unit does not issue a pre-emptive negative acknowledgement (NACK) for memory accesses associated with context information. In one exemplary implementation, a memory management unit handles a plurality of contexts.
- The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments of the invention by way of example and not by way of limitation. The drawings referred to in this specification should be understood as not being drawn to scale except if specifically noted.
-
FIG. 1 is a block diagram of an exemplary shared resource management system in accordance with one embodiment of the present invention. -
FIG. 2 is a block diagram of an exemplary memory management system in accordance with one embodiment of the present invention -
FIG. 3 is a block diagram of a computer system, one embodiment of a computer system upon which embodiments of the present invention can be implemented. -
FIG. 4 is a flow chart of an exemplary memory management method in accordance with one embodiment of the present invention. -
FIG. 5 is a flow chart of an exemplary memory management method in accordance with one embodiment of the present invention. -
FIG. 6 is a flow chart of an exemplary management method in accordance with one embodiment of the present invention. -
FIG. 7A is a flow chart of an exemplary memory management method in accordance with one embodiment of the present invention. -
FIG. 7B is a flow chart of an exemplary engine initiated block bind method in accordance with one embodiment of the present invention. -
FIG. 8 is a block diagram illustration of an exemplary context specifier in accordance with one embodiment of the present invention. -
FIG. 9 is a block diagram of an exemplary page table data structure in accordance with one embodiment of the present invention. -
FIG. 10 is a block diagram of an exemplary system for translating virtual addresses to physical addresses in accordance with one embodiment of the present invention. - Reference will now be made in detail to the preferred embodiments of the invention, examples of which are illustrated in the accompanying drawings. While the invention will be described in conjunction with the preferred embodiments, it will be understood that they are not intended to limit the invention to these embodiments. On the contrary, the invention is intended to cover alternatives, modifications and equivalents, which may be included within the spirit and scope of the invention as defined by the appended claims. Furthermore, in the following detailed description of the present invention, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be obvious to one of ordinary skill in the art that the present invention may be practiced without these specific details. In other instances, well known methods, procedures, components, and circuits have not been described in detail as not to unnecessarily obscure aspects of the present invention.
- Some portions of the detailed descriptions which follow are presented in terms of procedures, logic blocks, processing, and other symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the means generally used by those skilled in data processing arts to effectively convey the substance of their work to others skilled in the art. A procedure, logic block, process, etc., is here, and generally, conceived to be a self-consistent sequence of steps or instructions leading to a desired result. The steps include physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical, magnetic, optical, or quantum signals capable of being stored, transferred, combined, compared, and otherwise manipulated in a computer system. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
- It should be borne in mind, however, that all of these and similar terms are associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise as apparent from the following discussions, it is appreciated that throughout the present application, discussions utilizing terms such as “processing”, “computing”, “calculating”, “determining”, “displaying” or the like, refer to the action and processes of a computer system, or similar processing device (e.g., an electrical, optical, or quantum, computing device), that manipulates and transforms data represented as physical (e.g., electronic) quantities. The terms refer to actions and processes of the processing devices that manipulate or transform physical quantities within a computer system's component (e.g., registers, memories, other such information storage, transmission or display devices, etc.) into other data similarly represented as physical quantities within other components.
-
FIG. 1 is a block diagram of an exemplary sharedresource management system 100 in accordance with one embodiment of the present invention. Exemplary sharedresource management system 100 includesengine 111,engine 112,engine 113, sharedresource management unit 120 and sharedresource 130.Engine 111,engine 112, andengine 113 are communicatively coupled to sharedresource management unit 120 which in turn is communicatively coupled to sharedresource 130. - The components of exemplary shared
resource management system 100 cooperatively operate to facilitate efficient management of access to a shared resource by a plurality of engines.Engine 111,engine 112 andengine 113 perform various processing activities. In one embodiment, each engine can perform processing activities for a variety of clients.Shared resource 130 supports the processing byengines resource 130 is a memory for storing information utilized byengines resource management unit 120 independently controls access to the sharedresource 130. - By independently controlling access to the shared
resource 130, sharedresource management system 100 facilitates efficient resolution of access problems. For example, the sharedresource management unit 120 allows a second one of the plurality of engines to access the shared resource even if a fault occurs while a first one of the plurality of engines is attempting to access the shared resource. In one exemplary implementation, sharedresource management unit 120 permits a first one of the plurality of engines to access sharedresource 130 until a problem occurs then prevents further access by the first one of the plurality of engines until the problem is resolved. For example, sharedresource management unit 120 issues NACKs in response to access requests from the first one of the plurality of engines until the problem is resolved. Sharedresource management unit 120 can permit other engines to access the sharedresource 130 even while issuing NACKs to the first one of the plurality of engines. In one embodiment, sharedresource management system 100 manages accesses to a shared memory and the sharedresource management unit 120 handles a plurality of contexts. -
FIG. 2 is a block diagram of exemplarymemory management system 200, in accordance with one embodiment of the present invention. In one exemplary implementation,memory management system 200 is referred to as a memory management unit (MMU).Memory management system 200 facilitates efficient memory access.Memory management system 200 includes shared translation buffers (e.g., hub translation buffer TLB 225 and TLB 239), dedicated micro translation buffers (e.g.,TLBs 231, 232 and 233) and fillcomponent 240. Shared translation buffers and dedicated micro translation buffers translate virtual addresses to physical addresses.Fill component 240 coordinates memory translation lookaside buffer invalidation and fetches page table entry information for the translation look aside buffers (e.g., on TLB misses). If there is a miss in the translation (e.g., information is not stored in a current cache location) a notification is sent to thefill component 240. Thefill component 240 puts together a memory request packet and forwards it to a frame buffer to retrieve translation data. In one embodiment, thefill component 240 can put together memory request packets associated with page table entries for translating between virtual and physical memory spaces. - Notifications of faults are forwarded to the
fill component 240.Fill component 240 forwards a post-fault expediting response. In one exemplary implementation, when an access request from an engine (e.g., 211) encounters a fault, fillcomponent 240 forwards Negative Acknowledgements (NACKs) in response to that request and additional memory access requests from the engine (e.g., 211) until the fault is resolved.Fill component 240 can permit other engines (e.g., 212, 213) to continue memory accesses normally unless a respective independent request from the other engines encounters a fault. -
Fill component 240 can implement the fault occurrence notification through the hierarchy in a reactive manner or a “lazy” notification manner. For example, in a reactive notification scheme a first engine (e.g., 211) memory access request fault notification discovered in Hub TLB 225 is forwarded both up the hierarchy (e.g. to fill component 240) and “directly” to other TLB's in other branches of the hierarchy (e.g.,MTLB 231, 232, 233) and the other TLB's also issue NACKS in response to access requests associated with the first engine's (e.g., 211) activities. By issuing the notification reactively the other TLB's do not have to wait for notification to “traverse” up the hierarchy (e.g. up to fill unit 240) and back down. In a “lazy” notification scheme for the memory access request fault notification discovered in Hub TLB 225, while notification is forwarded up the hierarchy, notification is not forwarded “directly” to other TLB's in other branches of the hierarchy (e.g.,MTLB 231, 232, 233) and the other TLB's continue to issue ACKS until encountering a fault themselves or receiving notification from higher up in the hierarchy. - In one embodiment,
memory management unit 200 facilitates virtualization of context information storage.Memory management unit 200 can permit context information to be stored and retrieved from bulk storage on the fly.Memory management unit 200 allows context information to be stored in varying physical memory locations. In one embodiment accesses to store the context information are immune to post-fault response. For example,memory management unit 200 does not issue a pre-emptive NACK for memory accesses associated with context information. In one exemplary implementation,memory management unit 200 handles a plurality of contexts. - In one embodiment,
memory management unit 200 coordinates binding of context information associated with memory access requests from the plurality of engines. In one exemplary implementation the context information is included in a block and the memory management unit binds the block to a particular engine and stores the information in an instance memory. The instance memory can be protected memory. The instance memory can also support multiple channels associated with the plurality of engines. In one exemplary implementation, the instance memory includes a pointer to a page table. The instance memory can also include context save and restore data and each one of the plurality of engines initiates a unique block bind by indicating an association between their engine ID and a given block of instance memory. - In one embodiment,
memory management unit 200 stores the address of a block of translation data associated with each engine. In one exemplary implementation, each engine participates in a binding mechanism in which each engine specifies tomemory management unit 200 the base address of a memory region in which the virtual address to physical address translation data is stored. An engine which uses a single virtual memory space and hence a single set of translation data, can prior to issuing any translation requests send a block bind operation tomemory management unit 200. In one embodiment, a segmentation data binding mechanism provides an address that is an offset to the base address. One exemplary segmentation caching mechanism is described in concurrent patent application entitled “Direct Memory Access Techniques That Include Caching Segmentation Data”, patent application Ser. No. 11/523,830, filed Sep. 19, 2006, which is incorporated herein by this reference. In one exemplary implementation, a first level of the page tables is also stored at a fixed offset from the base address. - In one embodiment, an engine sends memory management unit 200 a new block bind when the engine switches context. Page tables and segmentation data need not be swapped in and out to switch contexts. In addition, a block of data specifying the translation data can be moved or swapped out of the memory space. In one exemplary implementation, if the translation data is resident in memory, an engine can be given a base address in a block bind operation to restart a context.
- In one embodiment,
memory management unit 200 concurrently manages memory accesses for the plurality of engines.Memory management unit 200 facilitates concurrent memory access translations for the plurality of engines. In one exemplary implementation, the memory management unit tracks an identifier for each of the plurality of engines making a memory access request. The memory management unit associates each of the plurality of engines with particular translation information respectively. In one embodiment, the translation information is stored in a portion of instance memory.Memory management unit 200 can be non-blocking.Memory management unit 200 can also permit a hit under miss. - In one embodiment,
memory management system 200 is included in a graphics system. The graphics system includesengine 211, engine 212, andengine 213,buffer interface block 220,acceleration component 230, fillcomponent 240 andhost 250. In one embodiment, theacceleration component 230 is dedicated to serving the graphics engine.Engine 211, engine 212, andengine 213, are communicatively coupled to framebuffer interface block 220, which is communicatively coupled to fillcomponent 240.Fill component 240 in turn is communicatively coupled toacceleration component 230 andhost 250. - The components of the graphics system cooperatively operate to perform graphics processing memory accesses.
Engines buffer interface block 220 enables efficient and independent access to a frame buffer. Arbiter 221 selectively arbitrates access requests fromengines Acceleration component 230 enables accelerated processing of graphics information. In one embodiment,MTLBs level 1 cache andTLB 239 handles memory translations between virtual and physical memory spaces in alevel 2 cache.Fill component 240 coordinates memory accesses independently as set forth above. Host 250 issues interrupts including interrupts for handling memory access faults. -
FIG. 3 is a block diagram of acomputer system 300, one embodiment of a computer system upon which embodiments of the present invention can be implemented.Computer system 300 includescentral processor unit 301, main memory 302 (e.g., random access memory), chip set 320 withnorth bridge 321 andsouth bridge 325, removabledata storage device 304,input device 307,signal communications port 308, and graphics subsystem 350 which is coupled todisplay 370.Computer system 300 includes several busses for communicatively coupling the components ofcomputer system 300. Communication bus 391 (e.g., a front side bus) couplesnorth bridge 321 ofchipset 320 tocentral processor unit 301. Communication bus 392 (e.g., a main memory bus) couplesnorth bridge 321 ofchipset 320 tomain memory 302. Communication bus 393 (e.g., the Advanced Graphics Port interface) couples north bridge ofchipset 320 tographic subsystem 350. Communication buses 394-397 (e.g., a PCI bus) couple south bridge 325 of chip set 320 to removabledata storage device 304,input device 307,signal communications port 308 respectively. Graphics subsystem 350 includesgraphics processor 351,memory management unit 355 andgraphics buffer 359. - The components of
computer system 300 cooperatively operate to perform a variety of processing tasks and facilitate efficient memory accesses.Communications bus Central processor 301 processes information.Main memory 302 stores information and instructions for thecentral processor 301. Removabledata storage device 304 also stores information and instructions (e.g., functioning as a large information reservoir).Input device 307 provides a mechanism for inputting information and/or for pointing to or highlighting information ondisplay 370.Signal communication port 308 provides a communication interface to exterior devices (e.g., an interface with a network).Display device 370 displays information in accordance with data stored inframe buffer 359.Graphics processor 351 processes graphics commands fromcentral processor 301 and provides the resulting data tographics buffers 359 for storage and retrieval bydisplay monitor 370.Memory management unit 355 handles the memory access requests betweengraphics processor 351 and graphics buffers 359. In one embodiment,memory management unit 355 is similar tomemory management unit 200. It is appreciated that similar memory management units can be implemented to facilitate efficient and independent access requests to other memory components ofcomputer system 300, includingmain memory 302 andbulk data storage 304. - It is appreciated that the present invention can be implemented in a variety of embodiments. In one exemplary implementation the present invention can be utilized in processing systems utilized to provide a variety of graphics applications including video games. For example, the present invention can be utilized to facilitate virtual address to physical address translations in a game console, personal computer, personal digital assistant, cell phone or any number of platforms for implementing a video game. It is also appreciated that references to video game application implementations are exemplary and the present invention is not limited to these implementations.
-
FIG. 4 is a flow chart of an exemplarymemory management method 400 in accordance with one embodiment of the present invention.Memory management method 400 facilitates efficient and independent memory access fault recovery. - At
block 410, a plurality of access requests are received. In one embodiment, the plurality of access requests are from a plurality of clients. It is appreciated the access requests can be directed to various MMU units. In one embodiment,level 1 cache memory units (e.g., micro TLB, Hub TLB, etc.) receive access requests and after MMU translation the request has an aperture field which notifies a frame buffer interface which destination the request should be forwarded to (e.g., frame buffer memory, dedicated memory, main system memory, etc.). One of the pluralities of access requests can be from a first engine and another one of the plurality of access requests can be from a second engine - A fault associated with one of the plurality of access requests from a first engine is encountered at
block 420. It is appreciated that the fault can include a variety of memory fault types. For example, a page directory entry or a page table entry can be invalid, a page size fault, out-of-bound pointer instance fault, invalid instance pointer fault, an instance bind fault, etc. - In
block 430, the fault is handled independently. In one embodiment, handling the fault includes tracking faults from multiple engines independently (e.g. on a per engine basis). Handling the fault can include dropping subsequent access requests and preparing for a context swap. In one exemplary implementation, a stream of NACKs are forwarded in response to additional access requests from the one of the plurality of access requests from the first engine. - In one embodiment, the handling of the fault can be performed reactively or lazily. In both reactive and lazy approaches components further up a memory management hierarchy are notified of the fault. The components further up the memory management hierarchy also forward a stream of NACKs to other branches of the hierarchy in response to additional access requests from the first engine. In a lazy approach components in other branches of the memory management hierarchy are given a greater opportunity to continue to processing additional access requests before being notified of a fault and/or the context swap occurs. In a reactive approach components in other branches of the hierarchy are notified faster of a fault or context swap. For example, in a reactive approach the fault notification is provided more directly to components in other branches without having to wait for the fault to propagate up to the top of the hierarchy (e.g., a fill component) and then back down. In one embodiment, whether the fault handling is performed reactively or lazily is predetermined. In another embodiment, whether the fault handling is performed reactively or lazily is programmable.
- In one embodiment, handling the fault also includes recovering from a fault indication. For example, an engine client stops operations, is swapped out, the fault is fixed, and the client and/or engine restarts. In one exemplary implementation, interrupts are independently issued and an interrupt service routine is performed. The first engine independently restarts after the fault.
- At
block 440, one of the plurality of access requests from a second engine is proceeded with independently. In one embodiment, processing of a second one of the plurality of access requests from a second engine continues while handling the fault associated with the first one of the plurality of access requests. - In one embodiment, a shared resource management method selectively issues the stream of NACKs.
FIG. 5 is a flow chart of exemplarymemory management method 500 in accordance with one embodiment of the present invention. In one embodiment, an attribute is added to segmentation data or context specifier data which specifies the semantics of a management memory unit if a previous fault has been encountered. The bit can specify if a translation request should be faulted if a previous request has faulted or if the translation request should be translated. - At
block 510 an indication of a context session initiation is received. In one embodiment, the indication of a context session initiation is associated with a context switch after a fault. - A physical memory space is assigned to the context session at
block 520. In one embodiment context save is directed tolevel 1 cache memory (e.g., micro TLB, hub TLB, etc.). - In
block 530 virtual memory address to the assigned physical memory address translation is performed for context information associated with the context session. - In one embodiment, post-fault expediting responses are prevented from impacting accesses associated with the context information. In one embodiment, preventing a negative-acknowledgement (NACK) corresponding to a post-fault expediting response from being sent in response to an identified context information access request. In one embodiment, the MMU ignores a pending NACK if the attribute indicates the request is associated with a context specifier access.
- In one embodiment context information access requests are identified. In one exemplary implementation an attribute is added to the segmentation data which specifies the semantics of the MMU if a previous fault is encountered. For example, an attribute is added to a context specifier structure which indicates a NACK should not be issued if a previous fault is encountered.
- In one embodiment the restart includes a block bind.
FIG. 6 is a flow chart ofexemplary management method 600 in accordance with one embodiment of the present invention. In one embodimentmemory management method 600 facilitates service of multiple individual engines with multiple contexts. - At
block 610 an indication of a context session initiation is received. In one embodiment, wherein the context is stored in instance memory. In one exemplary implementation a notification of an engine changing channels is received. The indication of the context session initiation can be associated with a context time slice switch. - A block bind is performed at
block 620. In one embodiment, the block bind includes associating an engine and a channel with a context. In one embodiment the channel includes a state vector that defines the state of the channel. For example, each channel is associated with a particular channel state and address translation. - In one embodiment, TLB entries corresponding to an engine initiating a block bind are invalidated and replaced with information utilized to direct virtual space to physical space translations associated with a new context. In one exemplary implementation the information is demand loaded (e.g., when a TLB needs the translation information it is fetched). In one exemplary implementation, a context specifier includes a block pointer or indication to an instance memory location storing information associated with the new context.
- At block 630 a virtual memory address to physical memory address translation is performed for context information associated with the context session.
- In one embodiment,
management method 600 also includes coordinating multiple engines running on the same channel. In one exemplary implementation the coordinating includes avoiding save collisions. In one exemplary implementation, the coordinating includes initial memory allocation before actual execution. The initial allocation utilizes an “unvalidated” random pointer and a “real” value is put in when execution actually takes place. - In one embodiment, translations are performed concurrently.
FIG. 7A is a flow chart ofmemory management method 700 in accordance with one embodiment of the present invention. - At
block 710 an indication of a context session initiation is received. In one embodiment, an indication of a context session initiation is received from an engine. In one embodiment, the a host decides when to end a current channel's time slice and an associated engine stores current context state and proceed notifies a memory management unit of the context switch. An indication of an invalidation can also come from software. In one exemplary implementation, software has changed some instance memory information and forces an invalidation of the information in a TLB without forcing the engine to change to a different context. - Address translation mapping information is ascertained on a per engine basis in
block 720. For example, the translation information is retrieved from memory. The mapping can be maintained for each engine. In one embodiment, the mapping includes an engine identification and associated translation information locator address. In one exemplary implementation, the associated translation information locator address includes a base address of translation information. - In one embodiment address translation mapping information is cached with an engine identifier included in a tag of each entry. The tag is used to look up the base address of the translation information in memory.
- In
block 730 virtual memory address to physical memory address translation is performed in accordance with the translation information. The translation occurs across engines without interference. In one exemplary implementation, a request for translation includes a virtual address and an engine identifier. The pair is used to match against entries in a TLB. If no entry is found the engine identifier is utilized to look up the base address of the translation data in memory. - In one embodiment, MMU allocation bandwidth is dynamically adjusted between a plurality of engines.
-
FIG. 7B is a flow chart of exemplary engine initiatedblock bind method 790 in accordance with one embodiment of the present invention. - In
block 791 an indication to change to a second channel is received. In one embodiment, an indication to change to a second channel is received from an engine. The indication to change to a second channel can also come from a host. - Work associated with a first channel is finished and flushed in
block 792. In one embodiment the finishing and flushing work includes saving current context for information associated with the first channel. In one exemplary implementation the save is to a reserved area of instance memory. - In block 793 a memory management unit is notified of block bind information associated with a change to the second channel. In one embodiment, notifying includes identifying an instance memory block associated with the second channel. In one exemplary implementation the block bind information includes a base address of instance memory storing new translation information for a new context.
- In
block 794, new virtual space to physical space mapping information is retrieved from instance memory in accordance with block binding by a memory management unit. In one embodiment the block bind binding the base address of instance memory storing new translation information to a particular engine associated with the channel change. - New commands associated with the second channel are processed at
block 795. - Referring now to
FIG. 8 , a block diagram illustration of acontext specifier 800, in accordance with one embodiment of the present invention, is shown. Thecontext specifier 800 is a data structure that stores abase address 810, alimit 820, and one ormore attributes -
FIG. 9 is a block diagram of an exemplary pagetable data structure 900 in accordance with one embodiment of the present invention. The illustrated two-level pagetable data structure 900 may include a page directory 910 and one or more page tables 920-950. The page directory 910 includes a plurality of page directory entries (PDE). Each PDE includes the address of a corresponding page table 920-950. Each PDE may also include a plurality of attributes. Each page table 920-950 includes one or more page table entries (PTE). Each PTE includes a corresponding frame number (e.g., physical page number). Each PTE may also include a plurality of attributes such as a dirty bit, an accessed bit, a page check disable bit, a page write transparent bit, a user accessible bit, a writeable bit, a present bit, a referenced bit, a hash function identification bit, a valid bit, an address compare bit, a changed bit, storage control bits, a no execute bit, page protection bits and/or the like. One or more of the attributes in the PDE and/or PTE may be the same attributes contained in thecontext specifier 800. For example, in one implementation, the attributes in the context specifier includes all of the attributes in the PTEs. - With reference now to
FIG. 10 , a block diagram of an exemplary system for translating virtual addresses to physical addresses in accordance with one embodiment of the present invention is shown. In one exemplary implementation, avirtual address 1005 and acontext specifier 1010 are received. Thecontext specifier 1010 includes one or more attributes 1015. Thevirtual address 1005 includes a virtual page number. - In one exemplary implementation, the virtual address 1050 is translated using a page
table data structure 1045 which may be cached inTLB 1040. In particular, if aTLB 1040 includes a valid virtual-to-physical page mapping for the virtual page number 1025 (e.g., TLB hit) the appropriatephysical page number 1035 is output from theTLB 1040. When thevirtual page number 1025 is mapped to a givenphysical page number 1035 using theTLB 1040, the attributes used in the address translation process and/or output for use in accessing the memory may useattributes 1015 contained in theTLB 1040 or attributes of thecontext specifier 1010. - However, if the
TLB 1040 does not include a valid mapping (e.g., TLB miss), thevirtual address 1005 is translated using a pagetable data structure 1045. More specifically, thevirtual page number 1025 includes a page directory index 1050 and a page table index 1055. The page directory index 1050 in thevirtual address 1005 is used to index a given page directory 1060. The page directory entry (PDE) indexed by the page directory index 1050 includes an address of a given page table 1065. The page table index 1055 in thevirtual address 1005 is used to index the given page table 1065 located at the address specified in the PDE. The page table entry (PTE) indexed by the page table index 1055 includes thephysical page number 1035. When the virtual page number is translated to a givenphysical page number 1035 using the pagetable data structure 1045, theattributes 1015 used in the address translation process and/or output for use in accessing the memory may beattributes 1015 specified in the given PDE and/or PTE, or attributes 1015 of thecontext specifier 1010. - In one embodiment, the present invention includes a page table entry override capability. If the page table entry override capability is engaged an adjustment in the attributes of a context specifier is utilized in determining the physical page number.
- It is appreciated, embodiments of the present invention can facilitate virtual address translation to physical addresses of information located in a variety of memory types including,
level 1 cache memory,level 2 cache memory, frame buffer memory, dedicated memory, main system memory, and etcetera. - The foregoing descriptions of specific embodiments of the present invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise forms disclosed, and obviously many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the Claims appended hereto and their equivalents.
Claims (20)
1. A shared resource management system comprising:
a plurality of engines for performing processing;
a memory for supporting said processing; and
a memory management unit for virtualizing context information memory storage.
2. A shared resource management system of claim 1 wherein said memory management unit permits context information to be stored and retrieved from bulk storage on the fly.
3. A shared resource management system of claim 1 wherein said memory management unit allows context information to be stored in varying physical memory locations.
4. A shared resource management system of claim 1 wherein accesses to store said context information are immune to fault expediting access prevention indications.
5. A shared resource management system of claim 1 wherein said shared resource is a translation buffer resource.
6. A shared resource management system of claim 1 wherein said engines have a fault NACK restart capability.
7. A shared resource management system of claim 1 wherein said memory management unit handles a plurality of contexts.
8. A memory management method comprising:
receiving an indication of a context session initiation;
assigning a physical memory space to said context session;
performing virtual memory address to said assigned physical memory address translation for context information associated with said context session.
9. A memory management method of claim 8 further comprising preventing fault expediting access prevention indications from impacting accesses associated with said context information.
10. A memory management method of claim 8 further comprising identifying context information access requests
11. A memory management method of claim 8 further comprising preventing a negative-acknowledgement (NACK) corresponding to a fault expediting access prevention indication from being sent in response to an identified context information access request.
12. A memory management method of claim 8 wherein an attribute is added to the segmentation data or command which specifies the semantics of a MMU if a previous fault is encountered.
13. A memory management method of claim 8 wherein the MMU ignores a pending NACK if the attribute indicates the request is associated with a context specifier access.
14. A memory management method of claim 8 wherein an attribute is added to a context specifier structure which indicates a NACK should not be issued if a previous fault is encountered.
15. A memory management method of claim 8 wherein said indication of a context session initiation is associated with a context switch after a fault.
16. A memory management method of claim 8 wherein a context save is directed to a level 1 MMU unit.
17. A memory management component comprising:
a shared translation buffer for translating virtual addresses to physical addresses including translation of address associated with context information;
a fill component for coordinating page table data fetching; and
a host for sending interrupts.
18. A memory management component of claim 17 further comprising permitting memory access requests from a first requester to proceed while handling a fault from a second requester.
19. A memory management component of claim 17 wherein said shared translation buffer is a hub translation buffer.
20. A memory management component of claim 17 wherein said shared translation buffer is a level 2 translation buffer
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/960,739 US20140164716A1 (en) | 2006-11-01 | 2013-08-06 | Override system and method for memory access management |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/591,857 US8504794B1 (en) | 2006-11-01 | 2006-11-01 | Override system and method for memory access management |
US13/960,739 US20140164716A1 (en) | 2006-11-01 | 2013-08-06 | Override system and method for memory access management |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/591,857 Continuation US8504794B1 (en) | 2006-11-01 | 2006-11-01 | Override system and method for memory access management |
Publications (1)
Publication Number | Publication Date |
---|---|
US20140164716A1 true US20140164716A1 (en) | 2014-06-12 |
Family
ID=48876512
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/591,857 Active 2028-06-18 US8504794B1 (en) | 2006-11-01 | 2006-11-01 | Override system and method for memory access management |
US13/960,739 Abandoned US20140164716A1 (en) | 2006-11-01 | 2013-08-06 | Override system and method for memory access management |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/591,857 Active 2028-06-18 US8504794B1 (en) | 2006-11-01 | 2006-11-01 | Override system and method for memory access management |
Country Status (1)
Country | Link |
---|---|
US (2) | US8504794B1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140089450A1 (en) * | 2012-09-27 | 2014-03-27 | Mellanox Technologies Ltd. | Look-Ahead Handling of Page Faults in I/O Operations |
US9256545B2 (en) | 2012-05-15 | 2016-02-09 | Mellanox Technologies Ltd. | Shared memory access using independent memory maps |
US9639464B2 (en) | 2012-09-27 | 2017-05-02 | Mellanox Technologies, Ltd. | Application-assisted handling of page faults in I/O operations |
US10031857B2 (en) | 2014-05-27 | 2018-07-24 | Mellanox Technologies, Ltd. | Address translation services for direct accessing of local memory over a network fabric |
US10120832B2 (en) | 2014-05-27 | 2018-11-06 | Mellanox Technologies, Ltd. | Direct access to local memory in a PCI-E device |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8898430B2 (en) * | 2012-12-05 | 2014-11-25 | Arm Limited | Fault handling in address translation transactions |
US10404520B2 (en) * | 2013-05-29 | 2019-09-03 | Microsoft Technology Licensing, Llc | Efficient programmatic memory access over network file access protocols |
US9641614B2 (en) | 2013-05-29 | 2017-05-02 | Microsoft Technology Licensing, Llc | Distributed storage defense in a cluster |
US10606651B2 (en) | 2015-04-17 | 2020-03-31 | Microsoft Technology Licensing, Llc | Free form expression accelerator with thread length-based thread assignment to clustered soft processor cores that share a functional circuit |
US10452995B2 (en) | 2015-06-29 | 2019-10-22 | Microsoft Technology Licensing, Llc | Machine learning classification on hardware accelerators with stacked memory |
US10540588B2 (en) | 2015-06-29 | 2020-01-21 | Microsoft Technology Licensing, Llc | Deep neural network processing on hardware accelerators with stacked memory |
US10203893B2 (en) | 2015-10-22 | 2019-02-12 | American Megatrends, Inc. | Memory channel storage device detection |
US10558468B2 (en) * | 2015-10-22 | 2020-02-11 | American Megatrends International, Llc | Memory channel storage device initialization |
US10249017B2 (en) | 2016-08-11 | 2019-04-02 | Intel Corporation | Apparatus and method for shared resource partitioning through credit management |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030070058A1 (en) * | 2001-10-09 | 2003-04-10 | Boris Ostrovsky | Method and device for a context-based memory management system |
US20060195683A1 (en) * | 2003-08-28 | 2006-08-31 | Mips Technologies, Inc. | Symmetric multiprocessor operating system for execution on non-independent lightweight thread contexts |
US20070050580A1 (en) * | 2005-08-31 | 2007-03-01 | Lewis Russell L | Apparatus, system, and method for implementing protected virtual memory subcontexts |
US20070061547A1 (en) * | 2005-09-09 | 2007-03-15 | Sun Microsystems, Inc. | Hardware demapping of TLBs shared by multiple threads |
Family Cites Families (82)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3815101A (en) | 1972-11-08 | 1974-06-04 | Sperry Rand Corp | Processor state and storage limits register auto-switch |
US3950729A (en) | 1973-08-31 | 1976-04-13 | Nasa | Shared memory for a fault-tolerant computer |
US4654790A (en) | 1983-11-28 | 1987-03-31 | Amdahl Corporation | Translation of virtual and real addresses to system addresses |
US4812981A (en) | 1985-10-24 | 1989-03-14 | Prime Computer, Inc. | Memory management system improving the efficiency of fork operations |
US4797814A (en) | 1986-05-01 | 1989-01-10 | International Business Machines Corporation | Variable address mode cache |
JP2589713B2 (en) | 1987-11-20 | 1997-03-12 | 株式会社日立製作所 | Data processor and data processing system |
US5179669A (en) | 1988-08-22 | 1993-01-12 | At&T Bell Laboratories | Multiprocessor interconnection and access arbitration arrangement |
JPH02288927A (en) | 1989-01-18 | 1990-11-28 | Nec Corp | Shared memory management system |
CA2011807C (en) | 1989-03-20 | 1999-02-23 | Katsumi Hayashi | Data base processing system using multiprocessor system |
JPH0354660A (en) | 1989-07-21 | 1991-03-08 | Nec Corp | Shared memory management system for multiprocessor system |
US5123094A (en) | 1990-01-26 | 1992-06-16 | Apple Computer, Inc. | Interprocessor communications includes second CPU designating memory locations assigned to first CPU and writing their addresses into registers |
JPH04182858A (en) | 1990-11-19 | 1992-06-30 | Mitsubishi Electric Corp | Shared memory management system |
US5245702A (en) | 1991-07-05 | 1993-09-14 | Sun Microsystems, Inc. | Method and apparatus for providing shared off-screen memory |
US5696925A (en) | 1992-02-25 | 1997-12-09 | Hyundai Electronics Industries, Co., Ltd. | Memory management unit with address translation function |
US5414824A (en) | 1993-06-30 | 1995-05-09 | Intel Corporation | Apparatus and method for accessing a split line in a high speed cache |
US5446854A (en) | 1993-10-20 | 1995-08-29 | Sun Microsystems, Inc. | Virtual memory computer apparatus and address translation mechanism employing hashing scheme and page frame descriptor that support multiple page sizes |
US5649102A (en) | 1993-11-26 | 1997-07-15 | Hitachi, Ltd. | Distributed shared data management system for controlling structured shared data and for serializing access to shared data |
US5526504A (en) | 1993-12-15 | 1996-06-11 | Silicon Graphics, Inc. | Variable page size translation lookaside buffer |
US5963984A (en) | 1994-11-08 | 1999-10-05 | National Semiconductor Corporation | Address translation unit employing programmable page size |
US6813699B1 (en) | 1995-06-02 | 2004-11-02 | Transmeta Corporation | Speculative address translation for processor using segmentation and optional paging |
US5999189A (en) | 1995-08-04 | 1999-12-07 | Microsoft Corporation | Image compression to reduce pixel and texture memory requirements in a real-time image generator |
US5949785A (en) | 1995-11-01 | 1999-09-07 | Whittaker Corporation | Network access communications system and methodology |
US6298390B1 (en) | 1995-11-22 | 2001-10-02 | Sun Microsystems, Inc. | Method and apparatus for extending traditional operating systems file systems |
US6012132A (en) | 1997-03-31 | 2000-01-04 | Intel Corporation | Method and apparatus for implementing a page table walker that uses a sliding field in the virtual addresses to identify entries in a page table |
GB9724031D0 (en) | 1997-11-13 | 1998-01-14 | Advanced Telecommunications Mo | Cache memory operation |
US6591355B2 (en) | 1998-09-28 | 2003-07-08 | Technion Research And Development Foundation Ltd. | Distributed shared memory system with variable granularity |
US7007075B1 (en) | 1998-12-09 | 2006-02-28 | E-Lysium Transaction Systems Inc. | Flexible computer resource manager |
US6362826B1 (en) | 1999-01-15 | 2002-03-26 | Intel Corporation | Method and apparatus for implementing dynamic display memory |
US7275246B1 (en) | 1999-01-28 | 2007-09-25 | Ati International Srl | Executing programs for a first computer architecture on a computer of a second architecture |
US6574749B1 (en) | 1999-10-29 | 2003-06-03 | Nortel Networks Limited | Reliable distributed shared memory |
US6499090B1 (en) | 1999-12-28 | 2002-12-24 | Intel Corporation | Prioritized bus request scheduling mechanism for processing devices |
US6625715B1 (en) | 1999-12-30 | 2003-09-23 | Intel Corporation | System and method for translation buffer accommodating multiple page sizes |
US6457115B1 (en) | 2000-06-15 | 2002-09-24 | Advanced Micro Devices, Inc. | Apparatus and method for generating 64 bit addresses using a 32 bit adder |
ATE259081T1 (en) | 2000-07-06 | 2004-02-15 | Texas Instruments Inc | MULTI-PROCESSOR SYSTEM TEST CIRCUIT |
US6636223B1 (en) | 2000-08-02 | 2003-10-21 | Ati International. Srl | Graphics processing system with logic enhanced memory and method therefore |
EP1182571B1 (en) | 2000-08-21 | 2011-01-26 | Texas Instruments Incorporated | TLB operations based on shared bit |
US6742104B2 (en) | 2000-08-21 | 2004-05-25 | Texas Instruments Incorporated | Master/slave processing system with shared translation lookaside buffer |
EP1213650A3 (en) | 2000-08-21 | 2006-08-30 | Texas Instruments France | Priority arbitration based on current task and MMU |
US6883079B1 (en) | 2000-09-01 | 2005-04-19 | Maxtor Corporation | Method and apparatus for using data compression as a means of increasing buffer bandwidth |
US6859208B1 (en) | 2000-09-29 | 2005-02-22 | Intel Corporation | Shared translation address caching |
US6925547B2 (en) | 2000-12-14 | 2005-08-02 | Silicon Graphics, Inc. | Remote address translation in a multiprocessor system |
US7356026B2 (en) | 2000-12-14 | 2008-04-08 | Silicon Graphics, Inc. | Node translation and protection in a clustered multiprocessor system |
US6549997B2 (en) | 2001-03-16 | 2003-04-15 | Fujitsu Limited | Dynamic variable page size translation of addresses |
US7073044B2 (en) | 2001-03-30 | 2006-07-04 | Intel Corporation | Method and apparatus for sharing TLB entries |
US6658538B2 (en) | 2001-06-21 | 2003-12-02 | International Business Machines Corporation | Non-uniform memory access (NUMA) data processing system having a page table including node-specific data storage and coherency control |
US6523104B2 (en) | 2001-07-13 | 2003-02-18 | Mips Technologies, Inc. | Mechanism for programmable modification of memory mapping granularity |
US6823433B1 (en) | 2001-11-13 | 2004-11-23 | Advanced Micro Devices, Inc. | Memory management system and method for providing physical address based memory access security |
US6877077B2 (en) | 2001-12-07 | 2005-04-05 | Sun Microsystems, Inc. | Memory controller and method using read and write queues and an ordering queue for dispatching read and write memory requests out of order to reduce memory latency |
AU2003231070A1 (en) | 2002-04-18 | 2003-11-03 | Advanced Micro Devices Inc. | A computer system including a secure execution mode - capable cpu and a security services processor connected via a secure communication path |
US8285743B2 (en) | 2002-06-24 | 2012-10-09 | International Business Machines Corporation | Scheduling viewing of web pages in a data processing system |
EP1391820A3 (en) | 2002-07-31 | 2007-12-19 | Texas Instruments Incorporated | Concurrent task execution in a multi-processor, single operating system environment |
GB2392998B (en) | 2002-09-16 | 2005-07-27 | Advanced Risc Mach Ltd | Handling interrupts during multiple access program instructions |
US7398525B2 (en) | 2002-10-21 | 2008-07-08 | International Business Machines Corporation | Resource scheduling in workflow management systems |
US7168077B2 (en) | 2003-01-31 | 2007-01-23 | Handysoft Corporation | System and method of executing and controlling workflow processes |
US6963963B2 (en) | 2003-03-25 | 2005-11-08 | Freescale Semiconductor, Inc. | Multiprocessor system having a shared main memory accessible by all processor units |
US7003647B2 (en) | 2003-04-24 | 2006-02-21 | International Business Machines Corporation | Method, apparatus and computer program product for dynamically minimizing translation lookaside buffer entries across contiguous memory |
US7107441B2 (en) | 2003-05-21 | 2006-09-12 | Intel Corporation | Pre-boot interpreted namespace parsing for flexible heterogeneous configuration and code consolidation |
US7082508B2 (en) | 2003-06-24 | 2006-07-25 | Intel Corporation | Dynamic TLB locking based on page usage metric |
US7124255B2 (en) | 2003-06-30 | 2006-10-17 | Microsoft Corporation | Message based inter-process for high volume data |
US20050050013A1 (en) | 2003-08-28 | 2005-03-03 | Sharp Laboratories Of America, Inc. | System and method for policy-driven device queries |
US7321958B2 (en) | 2003-10-30 | 2008-01-22 | International Business Machines Corporation | System and method for sharing memory by heterogeneous processors |
US7159095B2 (en) | 2003-12-09 | 2007-01-02 | International Business Machines Corporation | Method of efficiently handling multiple page sizes in an effective to real address translation (ERAT) table |
US7107411B2 (en) | 2003-12-16 | 2006-09-12 | International Business Machines Corporation | Apparatus method and system for fault tolerant virtual memory management |
US7383414B2 (en) | 2004-05-28 | 2008-06-03 | Oracle International Corporation | Method and apparatus for memory-mapped input/output |
US7234038B1 (en) | 2004-05-28 | 2007-06-19 | Sun Microsystems, Inc. | Page mapping cookies |
US20060004984A1 (en) | 2004-06-30 | 2006-01-05 | Morris Tonia G | Virtual memory management system |
US8001294B2 (en) | 2004-09-28 | 2011-08-16 | Sony Computer Entertainment Inc. | Methods and apparatus for providing a compressed network in a multi-processing system |
US8843727B2 (en) | 2004-09-30 | 2014-09-23 | Intel Corporation | Performance enhancement of address translation using translation tables covering large address spaces |
US7437517B2 (en) | 2005-01-11 | 2008-10-14 | International Business Machines Corporation | Methods and arrangements to manage on-chip memory to reduce memory latency |
US7948896B2 (en) | 2005-02-18 | 2011-05-24 | Broadcom Corporation | Weighted-fair-queuing relative bandwidth sharing |
US7447869B2 (en) | 2005-04-07 | 2008-11-04 | Ati Technologies, Inc. | Method and apparatus for fragment processing in a virtual memory system |
US7299337B2 (en) | 2005-05-12 | 2007-11-20 | Traut Eric P | Enhanced shadow page table algorithms |
US7739668B2 (en) | 2005-05-16 | 2010-06-15 | Texas Instruments Incorporated | Method and system of profiling applications that use virtual memory |
US7814292B2 (en) | 2005-06-14 | 2010-10-12 | Intel Corporation | Memory attribute speculation |
US20070067505A1 (en) | 2005-09-22 | 2007-03-22 | Kaniyur Narayanan G | Method and an apparatus to prevent over subscription and thrashing of translation lookaside buffer (TLB) entries in I/O virtualization hardware |
US7739476B2 (en) | 2005-11-04 | 2010-06-15 | Apple Inc. | R and C bit update handling |
US7616218B1 (en) | 2005-12-05 | 2009-11-10 | Nvidia Corporation | Apparatus, system, and method for clipping graphics primitives |
US7519781B1 (en) | 2005-12-19 | 2009-04-14 | Nvidia Corporation | Physically-based page characterization data |
US7512767B2 (en) | 2006-01-04 | 2009-03-31 | Sony Ericsson Mobile Communications Ab | Data compression method for supporting virtual memory management in a demand paging system |
US7653803B2 (en) | 2006-01-17 | 2010-01-26 | Globalfoundries Inc. | Address translation for input/output (I/O) devices and interrupt remapping for I/O devices in an I/O memory management unit (IOMMU) |
JP4890033B2 (en) | 2006-01-19 | 2012-03-07 | 株式会社日立製作所 | Storage device system and storage control method |
US7545382B1 (en) | 2006-03-29 | 2009-06-09 | Nvidia Corporation | Apparatus, system, and method for using page table entries in a graphics system to provide storage format information for address translation |
-
2006
- 2006-11-01 US US11/591,857 patent/US8504794B1/en active Active
-
2013
- 2013-08-06 US US13/960,739 patent/US20140164716A1/en not_active Abandoned
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030070058A1 (en) * | 2001-10-09 | 2003-04-10 | Boris Ostrovsky | Method and device for a context-based memory management system |
US20060195683A1 (en) * | 2003-08-28 | 2006-08-31 | Mips Technologies, Inc. | Symmetric multiprocessor operating system for execution on non-independent lightweight thread contexts |
US20070050580A1 (en) * | 2005-08-31 | 2007-03-01 | Lewis Russell L | Apparatus, system, and method for implementing protected virtual memory subcontexts |
US20070061547A1 (en) * | 2005-09-09 | 2007-03-15 | Sun Microsystems, Inc. | Hardware demapping of TLBs shared by multiple threads |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9256545B2 (en) | 2012-05-15 | 2016-02-09 | Mellanox Technologies Ltd. | Shared memory access using independent memory maps |
US20140089450A1 (en) * | 2012-09-27 | 2014-03-27 | Mellanox Technologies Ltd. | Look-Ahead Handling of Page Faults in I/O Operations |
US8914458B2 (en) * | 2012-09-27 | 2014-12-16 | Mellanox Technologies Ltd. | Look-ahead handling of page faults in I/O operations |
US9639464B2 (en) | 2012-09-27 | 2017-05-02 | Mellanox Technologies, Ltd. | Application-assisted handling of page faults in I/O operations |
US10031857B2 (en) | 2014-05-27 | 2018-07-24 | Mellanox Technologies, Ltd. | Address translation services for direct accessing of local memory over a network fabric |
US10120832B2 (en) | 2014-05-27 | 2018-11-06 | Mellanox Technologies, Ltd. | Direct access to local memory in a PCI-E device |
Also Published As
Publication number | Publication date |
---|---|
US8504794B1 (en) | 2013-08-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8504794B1 (en) | Override system and method for memory access management | |
US8601235B2 (en) | System and method for concurrently managing memory access requests | |
US8706975B1 (en) | Memory access management block bind system and method | |
US7904692B2 (en) | Iommu with translation request management and methods for managing translation requests | |
EP2548124B1 (en) | Address mapping in virtualized processing system | |
US9298642B2 (en) | Sharing address translation between CPU and peripheral devices | |
US6804741B2 (en) | Coherent memory mapping tables for host I/O bridge | |
US7415575B1 (en) | Shared cache with client-specific replacement policy | |
US7623134B1 (en) | System and method for hardware-based GPU paging to system memory | |
US9569349B2 (en) | Method and apparatus for reallocating memory content | |
EP2936322B1 (en) | Processing device with address translation probing and methods | |
US8868883B1 (en) | Virtual memory management for real-time embedded devices | |
US10241925B2 (en) | Selecting a default page size in a variable page size TLB | |
US20070130372A1 (en) | I/O address translation apparatus and method for specifying a relaxed ordering for I/O accesses | |
KR101893966B1 (en) | Memory management method and device, and memory controller | |
CN110196757A (en) | TLB filling method, device and the storage medium of virtual machine | |
US8607008B1 (en) | System and method for independent invalidation on a per engine basis | |
US20220012183A1 (en) | Methods and systems for translating virtual addresses in a virtual memory based system | |
US8700865B1 (en) | Compressed data access system and method | |
US10754791B2 (en) | Software translation prefetch instructions | |
CA2816443A1 (en) | Secure partitioning with shared input/output | |
US8533425B1 (en) | Age based miss replay system and method | |
US20120110297A1 (en) | Secure partitioning with shared input/output | |
US11636045B2 (en) | Translating virtual addresses in a virtual memory based system | |
WO2023064609A1 (en) | Translation tagging for address translation caching |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |