Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
MULTIPLE BLOCK SEQUENTIAL MEMORY MANAGEMENT
Document Type and Number:
WIPO Patent Application WO/2002/008907
Kind Code:
A2
Abstract:
Systems and methods are described for multiple block sequential memory management. A method includes: partitioning a block of memory into a plurality of shared memory segments; and providing a processor with accessibility to each of the plurality of shared memory segments. An apparatus includes: a computer system; a block of memory including a pluriel of shared memory segments; and a processor.

Inventors:
WEST KARLON (US)
Application Number:
PCT/US2001/023863
Publication Date:
January 31, 2002
Filing Date:
July 26, 2001
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
TIMES N SYSTEMS INC (US)
WEST KARLON (US)
International Classes:
G05D11/00; G06F9/46; G06F12/00; G06F13/16; (IPC1-7): G06F12/00
Domestic Patent References:
WO2000036509A22000-06-22
WO1999017196A11999-04-08
Foreign References:
EP0917056A21999-05-19
Other References:
LIVIU IFTODE, JASWINDER PAL SINGH, KAI LI: "SCOPE CONSISTENCY: A BRIDGE BETWEEN RELEASE CONSISTENCY AND ENTRY CONSISTENCY" 8TH. ANNUAL ACM SYMPOSIUM ON PARALLEL ALGORITHMS AND ARCHITECTURES. SPAA '96. PADUA, JUNE 24 - 26, 1996, ANNUAL ACM SYMPOSIUM ON PARALLEL ALGORITHMS AND ARCHITECTURES (SPAA), NEW YORK, IEEE, US, vol. SYMP. 8, 24 June 1996 (1996-06-24), pages 277-287, XP000697185 ISBN: 0-89791-809-6
G. CATTANEO, G. DI GIORE, M. RUOTOLO: "ANOTHER C THREADS LIBRARY" ACM SIGPLAN NOTICES, ASSOCIATION FOR COMPUTING MACHINERY, NEW YORK, US, vol. 27, no. 12, 1 December 1992 (1992-12-01), pages 81-90, XP000330885 ISSN: 0362-1340
WILLIAM J. BOLOSKY, ROBERT P. FITZGERALD, MICHAEL L. SCOTT: "SIMPLE BUT EFFECTIVE TECHNIQUES FOR NUMA MEMORY MANAGEMENT" OPERATING SYSTEMS REVIEW (SIGOPS), ACM HEADQUARTER. NEW YORK, US, vol. 23, no. 5, 1989, pages 19-31, XP000115485
Attorney, Agent or Firm:
Bruckner, John J. (Llp 600 Congress Avenue Suite 2400 Austin, TX, US)
Download PDF:
Claims:
CLAIMS What is claimed is:
1. A method, comprising: partitioning a block of memory into a plurality of shared memory segments; and providing a processor with accessibility to each of the plurality of shared memory segments.
2. The method of claim 1, wherein partitioning includes logically partitioning.
3. The method of claim 1, further comprising partitioning another block of memory.
4. The method of claim 1, further comprising providing another processor with accessibility to each of the plurality of memory segments.
5. The method of claim 1, wherein accessibility to each of the plurality of shared memory segments includes accessibility to a plurality of shared memory management data structures.
6. The method of claim 5, further comprising subpartitioning the plurality of shared memory management data structures.
7. The method of claim 1, wherein accessibility to the plurality of shared memory segments includes accessibility to a plurality of shared memory lock structures.
8. The method of claim 7, further comprising subpartitioning the plurality of shared memory lock structures.
9. The method of claim 1, wherein accessibility includes mutually exclusive accessibility during a period of time.
10. The method of claim 9, wherein mutually exclusive accessibility is achieved by utilization of a synchronization technique.
11. The method of claim 10, wherein the synchronization technique includes at least one synchronization technique selected from the group consisting of : spin locking, bus locking and token passing.
12. The method of claim 1, wherein accessibility includes accessibility utilizing a shared memory segment selection technique.
13. The method of claim 12, wherein the shared memory segment selection technique includes at least one shared memory segment selection technique selected from the group consisting of : distributed RoundRobin algorithms, shared RoundRobin algorithms, processor identification algorithms, allocation length algorithms; and hashing algorithms.
14. The method of claim 1, wherein the plurality of shared memory segments include a plurality of shared memory segments of different sizes.
15. The method of claim 1, further comprising dynamically increasing and decreasing the number of shared memory segments.
16. The method of claim 1, further comprising dynamically merging and splitting a plurality of shared memory segments.
17. A computer program, comprising computer or machine readable program elements translatable for implementing the method of claim 1.
18. An apparatus for performing the method of claim 1.
19. An apparatus, comprising: a computer system; a block of memory including a plurality of shared memory segments; and a processor.
20. The apparatus of claim 19, further comprising another block of memory.
21. The apparatus of claim 19, further comprising another processor.
22. The apparatus of claim 19, wherein the processor is coupled to each of the plurality of shared memory segments.
23. The apparatus of claim 19, further comprising a plurality of shared memory management data structures.
24. The apparatus of claim 23, wherein the processor is coupled to each of the plurality of shared memory management data structures.
25. The apparatus of claim 23, wherein the plurality of shared memory segments is coupled to the plurality of shared memory data structures.
26. The apparatus of claim 19, further comprising a plurality of shared memory lock structures.
27. The apparatus of claim 26, wherein the processor is coupled to each of the plurality of shared memory lock structures.
28. The apparatus of claim 26, wherein the plurality of shared memory segments is coupled to the plurality of shared memory lock structures.
29. The apparatus of claim 26, wherein the plurality of shared memory data structures is coupled to the plurality of shared memory lock structures.
30. A network, comprising the apparatus of claim 19.
31. An electronic media, comprising the computer program of claim 17.
32. An apparatus, comprising the electronic media of claim 31. 33.
33. A method, comprising deploying the electronic media of claim 31.
34. A kit, comprising the apparatus of claim 19 and the apparatus of claim 31.
35. The kit of claim 34, further comprising instructions.
Description:
DESCRIPTION MULTIPLE BLOCK SEQUENTIAL MEMORY MANAGEMENT CROSS-REFERENCES TO RELATED APPLICATIONS This application is a continuation-in-part of, and claims a benefit of priority under 35 U. S. C. 119 (e) and/or 35 U. S. C. 120 from, copending U. S. Ser. No. 60/220,974, filed July 26,2000; 60/220,748, filed July 26,2000; and filed July 25,2001, the entire contents of both of which are hereby expressly incorporated by reference for all purposes.

BACKGROUND OF THE INVENTION 1. Field of the Invention The invention relates generally to the field of computer systems. More particularly, the invention relates to computer systems where one or more central processing units (CPUs) are connected to one or more memory (RAM) subsystems, or portions thereof.

2. Discussion of the Related Art In a typical computing system, every CPU can access all of the RAM, either directly with Load and Store instructions, or indirectly, such as with a message passing scheme.

When more than one CPU can access or manage the RAM subsystem or a portion thereof, certain accesses to that RAM, specifically allocation and deallocation of RAM for use by the Operating System or some application, must be synchronized to ensure mutually exclusive access to those data structures tracking memory allocation and deallocation by no more that one CPU at a time.

This technology, in turn, generates contention for those data structures by multiple CPUs and thereby reduces overall system performance. What is required is a solution that increases system performance by reducing contention for those data structures by multiple CPUs.

Heretofore, the requirements of reducing contention for those data structures by multiple CPUs and increasing system performance referred to above have not been fully met. What is needed is a solution that addresses these requirements.

SUMMARY OF THE INVENTION There is a need for the following embodiments. Of course, the invention is not limited to these embodiments.

According to a first aspect of the invention, a method comprises: partitioning a block of memory into a plurality of shared memory segments; and providing a processor with accessibility to each of the plurality of shared memory segments. According to a second aspect of the invention, an apparatus comprises: a computer system; a block of memory including a plurality of shared memory segments; and a processor.

These, and other, embodiments of the invention will be better appreciated and understood when considered in conjunction with the following description and the accompanying drawings. It should be understood, however, that the following description, while indicating various embodiments of the invention and numerous specific details thereof, is given by way of illustration and not of limitation. Many substitutions, modifications, additions and/or rearrangements may be made within the scope of the invention without departing from the spirit thereof, and the invention includes all such substitutions, modifications, additions and/or rearrangements.

BRIEF DESCRIPTION OF THE DRAWINGS The drawings accompanying and forming part of this specification are included to depict certain aspects of the invention. A clearer conception of the invention, and of the components and operation of systems provided with the invention, will become more readily apparent by referring to the exemplary, and therefore nonlimiting, embodiments illustrated in the drawings, wherein like reference numerals (if they occur in more than one view) designate the same elements. The invention may be better understood by reference to one or more of these drawings in combination with the description presented herein. It should be noted that the features illustrated in the drawings are not necessarily drawn to scale.

FIG. 1 illustrates a block diagram of a computer system representing an embodiment of the invention.

FIG. 2 illustrates a flowchart diagram of a process that can be implemented by a computer program, representing an embodiment of the invention.

FIG. 3 illustrates a flowchart diagram of a shared Round-Robin memory segment selection scheme, representing an embodiment of the invention.

DESCRIPTION OF PREFERRED EMBODIMENTS The invention and the various features and advantageous details thereof are explained more fully with reference to the nonlimiting embodiments that are illustrated in the accompanying drawings and detailed in the following description. Descriptions of well known components and processing techniques are omitted so as not to unnecessarily obscure the invention in detail. It should be understood, however, that the detailed description and the specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only and not by way of limitation. Various substitutions, modifications, additions and/or rearrangements within the spirit and/or scope of the underlying inventive concept will become apparent to those skilled in the art from this detailed description.

The below-referenced U. S. Patent Applications disclose embodiments that were satisfactory for the purposes for which they are intended. The entire contents of U. S. Serial Numbers 09/273, 430, filed March 19,1999; 09/859,193, filed May 15,2001; 09/854,351, filed May 10,2001; 09/672,909, filed September 28, 2000; 09/653,189, filed August 31, 2000; 09/652,815, filed August 31,2000 ; 09/653,183, filed August 31,2000; 09/653,425, filed August 31,2000; 09/653,421, filed August 31,2000; 09/653,557, filed August 31, 2000; 09/653,475, filed August 31,2000; 09/653,429, filed August 31,2000; 09/653,502, filed August 31,2000; (Attorney Docket No. TNSY: 017US), filed July 25,2001; (Attorney Docket No. TNSY: 018US), filed July 25,2001; (Attorney Docket No. TNSY: 019US), filed July 25,2001; (Attorney Docket No.

TNSY: 020US), filed July 25,2001; (Attorney Docket No. TNSY: 021US), filed July 25,2001; (Attorney Docket No. TNSY: 022US), filed July 25,2001 (Attorney Docket No. TNSY : 023US), filed July 25,2001; (Attorney Docket No. TNSY : 024US), filed July 25,2001; and (Attorney Docket No.

TNSY: 026US), filed July 25,2001 are hereby expressly incorporated by reference herein for all purposes.

In a computing system for which the memory (RAM) subsystem or a portion of the RAM subsystem is connected to one or more central processing units (CPU) a technique is taught for reducing RAM subsystem contention and methods to efficiently and correctly process memory allocation and deallocation from the RAM subsystem.

In a compute system where more than one CPU has access to the RAM subsystem, or portions thereof, some methods of providing mutually exclusive access to the data structures used to track memory allocation and deallocation among the multiple CPUs must be provided. Traditionally, this is done with spinlocks, Test-And-Set registers, or bus locking mechanisms. In any of these scenarios, while a CPU is manipulating these specific data structures and another CPU needs to manipulate these data structures, the other CPU (s) must wait until the first CPU is finished, thus keeping the other CPUs from performing other work and thereby reducing the performance of the overall system.

In a computing system where each CPU has shared access to the RAM subsystem or a portion thereof, a methodology can be designed where more than one CPU can access the memory management data structures simultaneously while reducing contention for those data structures and, thus, increasing overall system performance.

Scardamalia et al [U. S. Serial No. 09/273, 430, filed March 19,1999] have described a system in which each compute node has its own private memory, but in which there is also provided a shared global memory accessible by all compute nodes. In this case, contention for shared memory data structures only occurs when more than one node is attempting to allocate or deallocate some shared memory at the same time. It is possible in a traditional symmetric multiprocessor (SMP) where all memory is shared among all CPU that the techniques described by this invention will also apply. It is also obvious to one skilled in the art that other distributed, share everything compute systems, including but not limited to cc-NUMA, benefit from the techniques taught herein.

A compute system of the type described in U. S. Serial No. 09/273,430, filed March 19, 1999 can be designed with each CPU able to allocate or reserve, and deallocate or release global shared memory for its use. The data structures describing the usable shared memory may reside in shared memory, though that is not necessary. The key concept of this invention is that the sum total of shared memory can be partitioned into two or more non-overlapping memory segments, each segment utilizing its own data structures used for managing shared memory allocation and deallocation, such that more than on CPU can

allocate or deallocate shared memory simultaneously without introducing mutual exclusion, thereby reducing RAM contention and increasing overall system performance.

When a CPU allocates or deallocates shared memory, some form of inter-CPU synchronization for purposes of mutual exclusion must be used to maintain the integrity of the data structures involved. However, if there are two or more sets of data structures managing different segments of memory, then access to all sets of memory segment data structures can be performed in parallel. FIG. 1 shows an example of such a compute system, with multiple CPUs, where global shared memory has been partitioned into two memory segments, and where the data structures for managing each shared memory segment as well as the synchronization primitives required for the management may be located in such system. However, the means described by this disclosure equally apply to compute systems where the data structures used to manage multiple shared memory segments and/or the synchronization primitives are not located in global shared memory.

Referring to FIG. 1, a compute system is depicted. Element 101 is a standard CPU.

This figure represents a two CPU compute system, namely elements 101 and 108, but it is obvious to one skilled in the art that this figure can contain many more that just two CPUs.

Element 108 is also a standard CPU.

Element 103 is the global shared memory that is accessible, and accessed, by a plurality of CPUs. Even though this invention applies to single CPU compute systems, the benefits of this invention are not realized in such a configuration since the contention for memory by more that one CPU never occurs.

It is possible to extend the techniques taught by this invention down to the process level, or even thread level, where a given process or thread may deterministically allocate or release shared memory from multiple memory segments in such a way as to reduce inter-process or inter-thread contention of the memory management data structures where the processes or threads are running on a single CPU system or on a multiple CPU system.

Element 102 is one of the shared memory segments, and element 106 is the other memory segment as described in this example. It is obvious to one skilled in the art that this figure may contain many more that two memory segments, each with its own data structures and synchronization primitives. The memory segments may be identical in size, but that is not required. An alternative enhancement to this technique is to have differently sized memory segments, such that memory allocations of a given size determine which

memory segment to utilize, reducing not only memory contention, but also reducing the occurrence of memory fragmentation.

Element 104 shows that the data structures for managing the allocation and deallocation in this compute system for memory segment 1 are actually located in the memory segment 1. However, it is obvious to one skilled in the art that the data structures used to manage allocation and deallocation from the different global shared memory segments may be located in any memory area of a single CPU or even distributed and synchronized across a plurality of CPUs. Element 109 is the data structure for memory segment 2.

Element 105 shows the synchronization mechanism used in this compute system for enforcing mutually exclusive access to the data structures used to manage shared memory segment 1 allocation and deallocations is a set of one or more locks, located in the global shared segment 1, accessible to all CPUs. It is obvious to one skilled in the art that the synchronization mechanism may be implemented using a bus locking mechanism on element 107, a token passing scheme used to coordinate access to the shared data structures among the different CPUs, or any of a number of different synchronization techniques. This invention does not depend on the synchronization technique used, but it is more easily described while referencing a give technique. Element 110 shows the location of the locks used to synchronize allocation and deallocation from memory segment 2.

Referring to FIG. 2, a decision flow of an application attempting to allocate global shared memory is depicted. Element 201 is the actual function call the application makes.

There are various and sundry parameters associated with the call, but, for the purposes of this invention, the parameters values themselves are not relevant. However, it is obvious to one skilled in the art that numerous methods for determining which segment to attempt an allocation from may be based on the characteristics of one or more of the parameter values. These characteristics include, but are not limited to, exclusive versus shared use, cached versus non-cached shared memory, memory ownership flags, length, etc.

Element 202 implements the selection of available memory segments. A simplistic solution uses a distributed round-robin scheme, such that any memory allocation from a given CPU selects the next memory segment in a circular fashion from the memory segment last used by that CPU. While this scheme reduces contention for the shared memory data structures, it is likely that multiple CPUs will still try to access the same

memory segment at any given time, limiting the effectiveness of this method. A better solution implements a shared round-robin scheme, such that every CPU accesses the same circular index used to determine which memory segment to allocate from the same memory segment at the same time.

Referring to FIG. 3, a Shared Round-Robin memory segment selection scheme is depicted. In such scheme, each processor can keep an ordered array of the available shared memory segments, dynamically updated whenever any processor performs memory allocation. Element 300 reads a well-known location stored in shared memory that contains the current shared memory segment array index. Element 301 increments the current index value to a new index value for the next future shared memory allocation.

Element 302 returns the segment at the index that was just modified. Finally, the allocation is from the index that was selected. In this manner, each processor that is performing shared memory allocations will use a different shared memory segment from the last shared memory allocation.

It is obvious to one skilled in the art that there are many more memory selection schemes than the ones mentioned here, including but not limited to: (i) selecting memory segments based on a unique CPU identification, providing a set associative scheme for allocating memory; (ii) selecting memory segments based on allocation length, also giving a set associated scheme with reduction in memory fragmentation; and (iii) selecting memory segments based on a hashing algorithm, randomly choosing the next memory segment, reducing system overhead in the selection process.

Element 203 is just the normal shared memory allocation procedure, which may include all of the synchronization and potential contention described above, but based on a given memory segment. The intent of this invention is to satisfy as many shared memory allocation requests by different memory segments simultaneously thereby reducing contention as much as possible. If in fact no simultaneous shared memory allocation request is ever satisfied by element 203, then a negligible amount of system overhead, and no additional contention is introduced by this invention. Therefore, in a worst case scenario, overall system performance is basically unaffected, but with a best case possibility of reducing shared memory data structure contention to almost zero.

Element 204 is a decision of whether the shared memory allocation from the selected segment succeeded or not. If the allocation succeeded, control is transferred to

element 206, and the shared memory address is returned to the caller. If the allocation failed, control is given to the decision in element 205.

Element 205 is a decision of whether all memory segments have been tried yet, or if there are still some available memory segments left. If there are one or more memory segments left, then control is returned back to element 302, and another attempt to allocate shared memory is made. If all memory segments have been tried with no success, then control is transferred to element 206, and the calling function is notified of status failure.

It is obvious to one skilled in the art that certain enhancements of the data flow described in FIG. 2 are possible, including: (i) dynamically adjusting the memory segment boundaries; (ii) dynamically increasing or decreasing the number of available memory segments; (iii) merging memory segments when multiple segments were tried in a an attempt to allocate shared memory with not success, and etc.

The context of the invention can include computer systems. The context of the invention can also include computer systems for which the RAM subsystem or a portion thereof is connected to one or more CPUs.

The invention can also be included in a kit. The kit can include some, or all, of the components that compose the invention. The kit can be an in-the-field retrofit kit to improve existing systems that are capable of incorporating the invention. The kit can include software, firmware and/or hardware for carrying out the invention. The kit can also contain instructions for practicing the invention. Unless otherwise specified, the components, software, firmware, hardware and/or instructions of the kit can be the same as those used in the invention.

The term approximately, as used herein, is defined as at least close to a given value (e. g., preferably within 10% of, more preferably within 1% of, and most preferably within 0.1% of). The term substantially, as used herein, is defined as at least approaching a given state (e. g., preferably within 10% of, more preferably within 1% of, and most preferably within 0.1% of). The term coupled, as used herein, is defined as connected, although not necessarily directly, and not necessarily mechanically. The term deploying, as used herein, is defined as designing, building, shipping, installing and/or operating. The term means, as used herein, is defined as hardware, firmware and/or software for achieving a result. The term program or phrase computer program, as used herein, is defined as a sequence of instructions designed for execution on a computer system. A program, or computer

program, may include a subroutine, a function, a procedure, an object method, an object implementation, an executable application, an applet, a servlet, a source code, an object code, a shared library/dynamic load library and/or other sequence of instructions designed for execution on a computer system. The terms including and/or having, as used herein, are defined as comprising (i. e., open language). The terms a or an, as used herein, are defined as one or more than one. The term another, as used herein, is defined as at least a second or more.

Practical Applications of the Invention A practical application of the invention that has value within the technological arts is improving system performance in an environment where there are multiple compute nodes, each with one or more CPUs, where there are one or more shared RAM units which are accessible by some or all of the compute nodes. There are virtually innumerable uses for the invention, all of which need not be detailed here.

Advantages of the Invention A multiple block sequential memory management technique, representing an embodiment of the invention, can be cost effective and advantageous for at least the following reasons. The invention improves quality and/or reduces costs compared to previous approaches.

All the disclosed embodiments of the invention disclosed herein can be made and used without undue experimentation in light of the disclosure. Although the best mode of carrying out the invention contemplated by the inventor (s) is disclosed, practice of the invention is not limited thereto. Accordingly, it will be appreciated by those skilled in the art that the invention may be practiced otherwise than as specifically described herein.

Further, the individual components need not be formed in the disclosed shapes, or combined in the disclosed configurations, but could be provided in virtually any shapes, and/or combined in virtually any configuration. Further, the individual components need not be fabricated from the disclosed materials, but could be fabricated from virtually any suitable materials.

Further, variation may be made in the steps or in the sequence of steps composing methods described herein.

Further, although the multiple block sequential memory management technique described herein can be a separate module, it will be manifest that the multiple block sequential memory management technique may be integrated into the system with which it is associated. Furthermore, all the disclosed elements and features of each disclosed embodiment can be combined with, or substituted for, the disclosed elements and features of every other disclosed embodiment except where such elements or features are mutually exclusive.

It will be manifest that various substitutions, modifications, additions and/or rearrangements of the features of the invention may be made without deviating from the spirit and/or scope of the underlying inventive concept. It is deemed that the spirit and/or scope of the underlying inventive concept as defined by the appended claims and their equivalents cover all such substitutions, modifications, additions and/or rearrangements.

The appended claims are not to be interpreted as including means-plus-function limitations, unless such a limitation is explicitly recited in a given claim using the phrase (s)"means for"and/or"step for."Subgeneric embodiments of the invention are delineated by the appended independent claims and their equivalents. Specific embodiments of the invention are differentiated by the appended dependent claims and their equivalents.