System and method for invalidating and updating individual GART table entries for accelerated graphics port transaction requests

• US 5,914,730 A
• Filed: 09/09/1997
• Issued: 06/22/1999
• Est. Priority Date: 09/09/1997
• Status: Expired due to Term

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Compaq Computer Corporation (HP Inc.)
• Inventors: Elliott, Robert C., Santos, Gregory N.
• Primary Examiner(s): Tung, Kee M.
• Assistant Examiner(s): Luu, Sy D.
• Application Number:
  US08/926,421
• Time in Patent Office: 651 Days
• Field of Search: 345/503, 345/520-521, 345/514, 345/522, 345/526, 345/515, 345/516, 345/507, 345/509, 711/206-207, 711/118, 395/306-309, 395/310
• US Class Current: 345/531
• CPC Class Codes: G06F 12/1027 using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
• CPC Class Codes: G06F 12/1081 for peripheral access to main memory, e.g. direct memory access [DMA]
• CPC Class Codes: G06F 3/14 Digital output to display device ; Cooperation and interconnection of the display device with other functional units
• CPC Class Codes: G09G 2360/121 using a cache memory
• CPC Class Codes: G09G 2360/125 using unified memory architecture [UMA]
• CPC Class Codes: G09G 5/363 Graphics controllers
• CPC Class Codes: G09G 5/39 Control of the bit-mapped memory

Abstract

A computer system having a core logic chipset that functions as a bridge between an Accelerated Graphics Port ("AGP") bus device such as a graphics controller, and a host processor and computer system memory wherein a Graphics Address Remapping Table ("GART table") is used by the core logic chipset to remap virtual memory addresses used by the AGP graphics controller into physical memory addresses that reside in the computer system memory. The GART table enables the AGP graphics controller to work in contiguous virtual memory address space, but actually use non-contiguous blocks or pages of physical system memory to store textures, command lists and the like. The GART table is made up of a plurality of entries, each entry comprising an address pointer to a base address of a page of graphics data in memory, and feature flags that may be used to customize the associated page. The core logic chipset may cache a subset of the most recently used GART table entries to increase AGP performance when performing the address translation. A GART cache entry control register is used by an application programming interface, such as a GART miniport driver, to indicate to the core logic chipset that an individual GART table entry in the chipset cache should be invalidated and/or updated. The core logic chipset may then perform the required invalidate and/or update operation on the individual GART table entry without having to flush or otherwise disturb the other still relevant GART table entries stored in the cache.

Claims

1. A computer system, comprising:

a system processor executing software instructions and generating graphics data;

a system memory having an addressable memory space comprising a plurality of bytes of storage, wherein each of the plurality of bytes of storage has a unique address;

the software instructions and the graphics data being stored in some of the plurality of bytes of storage of said system memory, wherein the graphics data is stored in a plurality of pages of graphics data, each of the plurality of pages of graphics data comprising a number of the plurality of bytes of storage;

a graphics processor generating video display data from the graphics data and adapted for connection to a video display to display the video display data;

a first interface logic for connecting said system processor to said system memory;

a second interface logic for connecting said system processor and said system memory to said graphics processor;

said second interface logic having a cache memory and a cache entry control register;

said cache memory having a plurality of storage locations, each of the plurality of storage locations comprising an address portion, an entry portion, an entry update portion and an entry invalidate portion;

a graphics address remapping table (GART table) having a plurality of entries, each of the plurality of GART table entries comprising an address pointer to a corresponding one of the plurality of pages of graphics data; and

said second interface logic reading selected ones of the plurality of GART table entries and storing the selected ones in the entry portions of the plurality of storage locations of said cache memory, the storage locations being associated with graphics device addresses asserted by said graphics processor; and

said cache entry control register adapted to receive information for a graphics device address, an entry update and an entry invalidate from an applications programming interface (API) of the software instructions;

wherein,

if the received information through said cache entry control register causes the entry update portion to be set to a first logic level, said second interface logic will read the plurality of GART entries and update a one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

if the received information through said cache entry control register causes the entry invalidate portion to be set to the first logic level, said second interface logic will invalidate the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

2. The computer system of claim 1, further comprising:

a cache entry update bit in said cache entry control register that can be set to the first logic level by the API writing to said cache entry control register and can be read by the API to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry update bit to the first logic level by the API causes said second interface logic to update from the plurality of GART table entries stored in said system memory the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register from the API; and

said second interface logic clearing the cache entry update bit to the second logic level after updating the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

3. The computer system of claim 1, further comprising:

a cache entry invalidate bit in said cache entry control register that can be set to the first logic level by the API writing to said cache entry control register and can be read by the API to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry invalidate bit to the first logic level by the API causes said second interface logic to invalidate the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register from the API; and

said second interface logic clearing the cache entry invalidate bit to the second logic level after invalidating the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

4. The computer system of claim 1, wherein the applications programming interface (API) is a GART miniport driver.

5. The computer system of claim 1, wherein said second interface logic uses the selected ones of the plurality of GART table entries stored in said cache memory to point to addresses of associated pages of a first portion of the graphics data stored in said system memory, the associated pages of the first portion of the graphics data being read by said graphics processor to generate the video display data.

6. The computer system of claim 5, further comprising a local frame buffer memory connected to said graphics processor, said local frame buffer storing a second portion of the graphics data from said system memory.

7. The computer system of claim 6, wherein said local frame buffer memory stores the second portion of the graphics data in contiguous virtual address space and said graphics processor accesses the first portion of the graphics data in contiguous virtual address space by using the selected ones of the plurality of GART table entries stored in said cache memory and accesses the second portion of the graphics data from said local frame buffer memory.

8. The computer system of claim 7, wherein said graphics processor reads the first and second portions of the graphics data in contiguous virtual address space.

9. The computer system of claim 5, wherein the associated pages of the first portion of the graphics data are stored in random non-contiguous pages of the plurality of pages of graphics data.

10. The computer system of claim 1, wherein the first logic level is a logic 1 and the second logic level is a logic 0.

11. The computer system of claim 1, wherein the first logic level is a logic 0 and the second logic level is a logic 1.

12. The computer system of claim 1, further comprising a third interface logic for connecting said system processor and said system memory to input-output devices.

13. The computer system of claim 1, further comprising a fourth interface logic for connecting said system processor and said system memory to storage devices.

14. The computer system of claim 1, wherein the plurality of GART table entries are stored in said system memory.

15. The computer system of claim 1, wherein the plurality of GART table entries are stored in a plurality of pages of GART table entries in said system memory.

16. The computer system of claim 15, wherein the plurality of pages of GART table entries are stored in said system memory in a non-contiguous and random order.

17. The computer system of claim 16, further comprising a GART directory having a plurality of entries, each of the plurality of GART directory entries comprising an address pointer to a corresponding one of the plurality of pages of GART table entries, wherein said second interface logic uses the plurality of GART directory entries for locating the plurality of pages of GART table entries in said system memory.

18. The computer system of claim 1, wherein the number of the plurality of bytes of storage in each of the plurality of pages of graphics data is 4096 bytes.

19. The computer system of claim 1, further comprising a video display.

20. A computer system having a core logic chipset which connects a central processing unit and random access memory to an accelerated graphics port (AGP) bus, said system comprising:

a central processing unit connected to a host bus;

a random access memory connected to a random access memory bus;

a core logic chipset connected to the host bus and the random access memory bus;

said core logic chipset having a first interface bridge for connecting the host bus to the random access memory bus;

said core logic chipset having a second interface bridge for connecting the host bus to an accelerated graphics port (AGP) bus;

said core logic chipset having a third interface bridge for connecting the random access memory bus to the AGP bus;

said core logic chipset having a cache memory and a cache entry control register;

said cache memory having a plurality of storage locations, each of the plurality of storage locations comprising an address portion, an entry portion, an entry update portion and an entry invalidate portion;

said core logic chipset using a graphics address remapping table (GART table) having a plurality of entries, each of the plurality of GART table entries comprising an address pointer to a corresponding one of a plurality of pages of graphics data stored in said random access memory;

said core logic chipset reading selected ones of the plurality of GART table entries stored in said random access memory and storing the selected ones of the plurality of GART table entries in the entry portions of the plurality of storage locations of said cache memory, each of the entry portions associated with a one of the address portions; and

said cache entry control register adapted to receive information for a graphics device address, an entry update and an entry invalidate;

wherein,

if the received information through said cache entry control register causes the entry update portion to be set to a first logic level, said core logic chipset will read the plurality of GART entries and update a one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

if the received information through said cache entry control register causes the entry invalidate portion to be set to the first logic level, said core logic chipset will invalidate the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

21. The computer system of claim 20, wherein the central processing unit is a plurality of central processing units.

22. The computer system of claim 20, wherein the plurality of pages of graphics data are stored in said random access memory in a non-contiguous and random order.

23. The computer system of claim 22, wherein each one of the plurality of GART table entries comprises a plurality of binary bits and each one of the plurality of pages of graphics data is associated with the each one of the plurality of GART table entries such that a first number of most significant bits of the plurality of binary bits comprise a base address of the associated each one of the plurality of pages of graphics data.

24. The computer system of claim 20, further comprising:

a cache entry update bit in said cache entry control register that can be set to the first logic level and can be read to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry update bit to the first logic level causes said core logic chipset to update from the plurality of GART table entries stored in said system memory the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

said core logic chipset clearing the cache entry update bit to the second logic level after updating the one of the plurality of storage locations associated with the device address received by said cache entry control register.

25. The computer system of claim 20, further comprising:

a cache entry invalidate bit in said cache entry control register that can be set to the first logic level and can be read to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry invalidate bit to the first logic level causes said core logic chipset to invalidate the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

said core logic chipset clearing the cache entry invalidate bit to the second logic level after invalidating the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

26. The computer system of claim 20, wherein the plurality of GART table entries are stored in at least one page of said random access memory.

27. The computer system of claim 20, wherein each of the plurality of pages of graphics data is 4096 bytes.

28. The computer system of claim 20, wherein said core logic chipset is at least one integrated circuit.

29. The computer system of claim 28, wherein said at least one integrated circuit core logic chipset is at least one application specific integrated circuit.

30. The computer system of claim 28, wherein said at least one integrated circuit core logic chipset is at least one programmable logic array integrated circuit.

31. The computer system of claim 20, wherein said central processing unit executes software instructions and generates the graphics data.

32. The computer system of claim 20, further comprising a graphics processor for generating video display data based upon the graphics data.

33. The computer system of claim 32, further comprising a local frame buffer memory coupled to said graphics processor, wherein said graphics processor combines video data stored in said local frame buffer memory with the associated ones of the plurality of pages of graphics data read from said random access memory based upon the selected ones of the plurality of GART table entries stored in said cache memory to generate video display data.

34. The computer system of claim 20, further comprising said core logic chipset having a fourth interface bridge for connecting the host bus to a peripheral component interconnect (PCI) bus adapted for connection to input-output devices.

35. The computer system of claim 34, further comprising said core logic chipset having a fifth interface bridge for connecting the random access memory bus to the PCI bus.

36. The computer system of claim 34, further comprising said core logic chipset having a sixth interface bridge for connecting the AGP bus to the PCI bus.

37. The computer system of claim 20, further comprising a network interface card, a hard disk, a floppy disk drive, a modem, a keyboard and a mouse.

38. The computer system of claim 20, further comprising a serial port, a parallel port and a real time clock.

39. The computer system of claim 20, further comprising a read only memory basic input-output system (ROM BIOS), a non-volatile random access memory (NVRAM), a tape drive and a CD ROM drive.

40. A method, in a computer system, of updating and invalidating individual selected ones of a plurality of graphics address remapping table (GART table) entries stored in a cache memory, said method comprising the steps of:

storing a plurality of pages of graphics data in any order in a computer system memory;

storing a plurality of entries of a graphics address remapping table (GART table) in the computer system memory, wherein each one of the plurality of GART table entries corresponds to a one of the plurality of pages of graphics data stored in the computer system memory;

reading selected ones of the plurality of GART table entries stored in the computer system memory;

storing the selected ones read from the computer system memory into a cache memory, wherein the cache memory has a plurality of storage locations, each of the plurality of storage locations comprising a graphics device address portion, an entry portion, an entry update portion and an entry invalidate portion, wherein the selected ones are stored in the entry portions;

writing a first logic level to the entry update portion of a one of the plurality of storage locations when an associated one of the selected ones requires updating from the computer system memory; and

writing the first logic level to the entry invalidate portion of a one of the plurality of storage locations when an associated one of the selected ones is invalid.

41. The method of claim 40, further comprising the steps of:

reading the entry update portions of the plurality of storage locations of the cache memory;

reading a new selected one of the plurality of GART table entries in the computer system memory for each of the entry update portions containing the first logic level;

storing the new selected one in the entry portion associated with each of the entry update portions containing the first logic level; and

resetting each of the entry update portions to a second logic level after storing the new selected one in the entry portion.

42. The method of claim 41, further comprising the step of writing to a cache entry control register a graphics device address of a selected one of the plurality of GART table entries stored in the cache memory and the first logic level to a cache entry update bit of the cache entry control register, wherein the first logic level is written to the entry update portion of the one of the plurality of storage locations associated with the graphics device address.

43. The method of claim 42, further comprising the step of clearing the cache entry update bit from the first logic level to a second logic level after storing the associated new selected one.

44. The method of claim 40, further comprising the steps of:

reading the entry invalidate portions of the plurality of storage locations of the cache memory;

invalidating the storage location associated with each of the entry invalidate portions containing the first logic level; and

resetting each of the entry invalidate portions to a second logic level after invalidating the storage location associated therewith.

45. The method of claim 44, further comprising the step of writing to a cache entry control register a graphics device address of a selected one of the plurality of GART table entries stored in the cache memory and the first logic level to a cache entry invalidate bit of the cache entry control register, wherein the first logic level is written to the entry invalidate portion of the one of the plurality of storage locations associated with the graphics device address.

46. The method of claim 45, further comprising the step of clearing the cache entry invalidate bit from the first logic level to a second logic level after invalidating the storage location associated therewith.

47. The method of claim 40, further comprising the step of reading associated ones of the plurality of pages of graphics data in an order determined by the selected ones of the plurality of GART table entries stored in the cache memory.

48. The method of claim 40, wherein a system memory address is determined for each byte of graphics data stored in the plurality of pages of graphics data by a base address stored in the associated one of the plurality of GART table entries and an offset address added to the base address.

49. The method of claim 40, further comprising the step of allocating memory locations in the computer system for storing the plurality of GART table entries during initialization of the computer system.

50. A core logic chipset adapted for connection to a computer central processing unit and random access memory, an accelerated graphics port (AGP) bus and a peripheral component interconnect (PCI) bus, comprising:

an accelerated graphics port (AGP) request queue;

an AGP reply queue;

an AGP data and control logic;

said AGP data and control logic having an AGP cache entry control register;

an AGP cache memory;

said AGP cache memory having a plurality of storage locations, each of the plurality of storage locations comprising a graphics device address portion, an entry portion, an entry update portion and an entry invalidate portion;

an AGP arbiter;

a host to peripheral component interconnect (PCI) bridge;

a PCI to PCI bridge;

a memory interface and control logic adapted for connecting to a computer system random access memory; and

a host bus interface adapted for connecting to a computer system host bus having at least one central processing unit connected thereto;

wherein,

said AGP request and reply queues are connected to said memory interface and control logic;

said AGP data and control logic is connected to said memory and interface control logic;

said AGP data and control logic is connected to the host bus interface;

said host to PCI bus bridge is connected to the host bus interface and is adapted for connection to a PCI bus;

said PCI to PCI bridge is connected to said AGP data and control logic, wherein said PCI to PCI bridge transfers PCI information transactions between said Host to PCI bus bridge and said AGP data and control logic;

said AGP data and control logic and said AGP arbiter adapted for connection to an AGP bus having an AGP device;

wherein

said AGP data and control logic is adapted to use a graphics address remapping table (GART table) having a plurality of entries, each of the plurality of GART table entries comprising an address pointer to a one of a plurality of pages of graphics data stored in the computer system random access memory;

said AGP data and control logic is adapted to read selected ones of the plurality of GART table entries stored in said random access memory and is adapted to store the selected ones of the plurality of GART table entries in the entry portions of the plurality of storage locations of said cache memory, each of the entry portions associated with a one of the graphics device address portions; and

said cache entry control register adapted to receive information for a graphics device address, an entry update and an entry invalidate;

wherein,

if the entry update of the received information causes the entry update portion to be set to a first logic level, said AGP data and control logic is adapted to update from the GART table entries stored in said random access memory the one of the plurality of storage locations associated with the graphics device address; and

if the entry invalidate of the received information causes the entry invalidate portion to be set to the first logic level, said AGP data and control logic is adapted to invalidate the one of the plurality of storage locations associated with the graphics device address.

51. The core logic chipset of claim 50, further comprising:

a cache entry update bit in said cache entry control register that can be set to the first logic level and can be read to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry update bit to the first logic level causes said AGP data and control logic to update from the GART table entries stored in said system memory the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

said AGP data and control logic clearing the cache entry update bit to the second logic level after updating the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.

52. The core logic chipset of claim 50, further comprising:

a cache entry invalidate bit in said cache entry control register that can be set to the first logic level and can be read to determine if set to the first logic level or cleared to a second logic level, wherein setting the cache entry invalidate bit to the first logic level causes said AGP data and control logic to invalidate the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register; and

said AGP data and control logic clearing the cache entry invalidate bit to the second logic level after invalidating the one of the plurality of storage locations associated with the graphics device address received by said cache entry control register.