CN101419561A - Resource management method and system in isomerization multicore system - Google Patents
Resource management method and system in isomerization multicore system Download PDFInfo
- Publication number
- CN101419561A CN101419561A CNA2007101764277A CN200710176427A CN101419561A CN 101419561 A CN101419561 A CN 101419561A CN A2007101764277 A CNA2007101764277 A CN A2007101764277A CN 200710176427 A CN200710176427 A CN 200710176427A CN 101419561 A CN101419561 A CN 101419561A
- Authority
- CN
- China
- Prior art keywords
- nuclear
- operating system
- status information
- resource status
- resource
- 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.)
- Pending
Links
Images
Landscapes
- Debugging And Monitoring (AREA)
Abstract
The invention relates to resource management technologies of computer operating systems, in particular to a management method of CPU resources in a heterogeneous multi-core system and a system thereof. The method and the system help solve the technical problem that management and allocation of the resources by hardware can not adapt to state requirement changes during system operation in a real-time manner in the heterogeneous multi-core system. The real-time management of a master core on the resource state of slave cores is realized by the following steps: a slave operating system reports the system resource state information thereof to a master core operating system; the master core operating system analyzes service states of various resources of the slave core system according to the reported system resource state information; and the master core operating system adjusts a resource allocation strategy according to the analysis results. The method and the system avoid a design limitation of a one-off resource allocation strategy of the hardware, increase system management means, and enhances the adaptability of the systems to real-time situations and the capacity of the systems to handle emergencies.
Description
Technical field
The present invention relates to the resource management techniques in the computer operating system, relate in particular to the management method and the system of cpu resource in a kind of heterogeneous multi-core system.
Background technology
Along with the raising of chip design level and the lifting of application demand, multi-core CPU more and more is applied in the product.Adopt at present the isomery system of multi-core CPU to select for use a nuclear among the CPU as main nuclear mostly, bear main resource management and complex logic computing, remaining nuclear is as bear the simple but big work of data volume of logic from nuclear.
In the isomery system of multi-core CPU, each nuclear operation operating system separately.Modern multiple task operating system provides means such as fairly perfect management of process, process scheduling, has realized in an operating system reasonably distributing system resource.Operating system can adopt timeslice, process level ladder of management such as priority manage scheduling at the process that belongs in its range of management, take processor and make the process that must use peripherals have no chance to start peripherals so that the process that prevents from seldom to use peripherals is long.
But in heterogeneous multi-core system, each nuclear among the CPU has adopted different operating system, for lacking effective ladder of management between the different operating system.Management between the operating system has exceeded the range of management of operating system itself.Operating system on each CPU of present heterogeneous system is to finish by the disposable configuration to hardware for the strategy that takies and divide of resource substantially, as the message that enters from the external world being distributed to buffer size that each nuclear the time distributed etc..To after the CPU configured in one piece, CPU all adopts same strategy to examine Resources allocation for its each in this powers on the cycle when system initialization.Also we can say, just operating system is being managed at present by hardware.And the ladder of management that general CPU hardware provides itself is limited, can't use complicated application demand flexibly.And hardware initialization determined that with regard to unique each checks the allocation strategy of resource after finishing, in real time the state variation of adaptive system run duration.
Summary of the invention
Have mirror therewith, one of purpose of the present invention provides the method for managing resource in a kind of heterogeneous multi-core system.Can solve the technical matters that by hardware means resource is managed distribution adaptive system run duration status requirements variation in real time in the heterogeneous multi-core system.For realizing above technical purpose, the present invention adopts following technical scheme:
Method for managing resource in a kind of heterogeneous multi-core system may further comprise the steps at least:
A, report its system resource status information to main nuclear operating system from nuclear operating system;
The described system resource status information that B, main nuclear operating system basis report is to respectively analyzing from nuclear operating-system resources user mode;
C, main nuclear operating system are adjusted resource allocation policy according to the result of described analysis.
Further, describedly regularly and/or according to alarming threshold report its system resource status information to described main nuclear operating system from nuclear operating system.
Further, the described internuclear means of communication that provides from employing hardware between nuclear operating system and the described main nuclear operating system communicates as carrier.
Further, in the step 1, also comprise the step of described system resource status information being carried out buffer memory.
Based on such scheme, further, for described buffer memory is provided with a buffer threshold, when the described system resource status information that reports reached described buffer threshold, described main nuclear operating system was carried out analyzing and processing in batches.
Based on such scheme, further, when the described system resource status information that reports from nuclear operating system belongs to emergence message, described main nuclear operating system will be analyzed this system resource status information that belongs to emergence message at once.
Based on such scheme, further, described main nuclear operating system also comprises the step of the system resource status information of buffer memory being carried out analyzing and processing when the described system resource status information that belongs to emergence message of analyzing and processing.
Another object of the present invention is to provide the resource management system in a kind of heterogeneous multi-core system, scheme is:
Resource management system in a kind of heterogeneous multi-core system comprises:
From the nuclear information reporting module, be used to collect from examining current system resource status information, and report main nuclear operating system;
From nuclear resource state adjusting module, be used for resource status being adjusted according to the resource adjustment instruction that main nuclear operating system sends;
From the nuclear information cache module, be used for the described system resource status information that reports from nuclear is carried out buffer memory;
From the nuclear information analysis module, be used for the described system resource status information that reports from nuclear is analyzed;
The resource allocation policy adjusting module is used for adjusting resource allocation policy according to described analysis result from the nuclear information analysis module, to sending resource adjustment instruction from nuclear operating system.
Further, describedly comprise from the nuclear information cache module:
Urgent judge module is used to judge whether the described system resource status information that reports from nuclear of reception belongs to emergence message, if then described emergence message is sent to described from the nuclear information analysis module; If not, then the described system resource status information that receives is cached in the cache module;
Cache module is used for the described system resource status information that buffer memory receives.
Further, describedly also comprise a thresholding judge module from the nuclear information cache module, be used for the system resource amount of state information that reports from nuclear of described buffer memory is monitored, the system resource status information with described buffer memory when reaching thresholding sends to described from the nuclear information analysis module.
The present invention has introduced than the more higher leveled operating system management mechanism of management of process in the operating system, main nuclear operating system in the isomery system can be adjusted resource allocation policy at current operation conditions, also play simultaneously management role from nuclear operating system.The present invention has broken away from the limitation of design of the disposable resource allocation policy of hardware, increased the ladder of management of system, has improved system to the adaptability of real-time condition with to the processing power of emergency situations.
Description of drawings
Fig. 1 receives and handles from the process flow diagram of nuclear operating system reporting information for the main nuclear operating system of the present invention;
Fig. 2 is the module composition diagram of system resource management system in the heterogeneous multi-core system of the present invention.
Embodiment
Below in conjunction with specific embodiment and accompanying drawing method and system of the present invention is elaborated.
Core concept of the present invention is, utilize internuclear means of communication, adopt the self-defined message form, by respectively from examining dynamically, initiatively report in real time respectively system resource status information from nuclear, main nuclear operating system receives from the reporting information of nuclear, and analyze, according to analysis result resource allocation policy is adjusted dynamically.
Fig. 1 receives and the process flow diagram of handling the system resource status information that reports from nuclear operating system for the main nuclear operating system of the present invention, and is mainly as follows:
Step 1: the system resource status information that the internuclear means of communication that main nuclear operating system provides by CPU hardware reports from nuclear from the communications carrier acquisition.
The internuclear means of communication that adopts hardware to provide between the operating system on the different IPs is provided to be communicated as carrier.The common protocol format that consults that adopts of operating system on principal and subordinate's nuclear makes an explanation to Content of Communication.
Can adopt following dual mode individually or simultaneously during from nuclear reporting system resource state information:
Timing cycle reports mode, from the nuclear end is set one timing cycle, periodically reports this system resource status information from nuclear.
Based on the alarming threshold mode, from the nuclear end urgency level of system resource status information is carried out classification, set one and distinguish the emergence message thresholding, the urgency level grade belongs to emergence message greater than the system resource status information of this thresholding, belongs to general information less than the information of this thresholding.For emergence message, require main nuclear end to respond immediately, can further satisfy specific demand like this.
Step 2: the communication protocol that main nuclear operating system adopts principal and subordinate's nuclear to decide through consultation judges whether the system resource status information that reports from nuclear belongs to emergence message, if execution in step 6 then; Execution in step 3 then if not.
Step 3: will be cached to buffer area from the system resource status information that nuclear reports.
Step 4: whether the quantity of judging in the buffer area system resource status information that reports from nuclear has reached buffer threshold, if reach, changes step 5; If do not reach, change step 1.
Step 5: main nuclear operating system is carried out the batch quantity analysis processing to the system resource status information that respectively reports from nuclear in the buffer area, and execution in step 7 then.
Step 6: main nuclear operating system is carried out analyzing and processing to the system resource status information that belongs to emergence message, and execution in step 7 then.
Step 7: main nuclear operating system judges whether needs adjustment resource allocation policy according to analysis result, if need then execution in step 8, if do not need then to change step 1.
Step 8: main nuclear operating system is according to respectively from the resource operating position of nuclear resource allocation policy being adjusted.
Main nuclear operating system is carried out sending from nuclear operating system of resource status adjustment to needs and is adjusted instruction, current resource user mode is adjusted according to adjusting instruction from examining operating system.
Fig. 2 is the module composition diagram of system resource management system in the heterogeneous multi-core system of the present invention, comprise from nuclear resource administration module and main nuclear resource administration module, operate on nuclear operating system from the nuclear resource administration module, main nuclear resource administration module operates on the main nuclear operating system.
Comprise from the nuclear information reporting module, from nuclear resource state adjusting module from the nuclear resource administration module.
Main nuclear resource administration module comprises from the nuclear information cache module, from nuclear information analysis module, resource allocation policy adjusting module.
Be used for the system resource status information that buffer memory reports from nuclear from the nuclear information cache module, from the nuclear information cache module system resource status information that receives is stored in the special memory headroom, after arriving certain thresholding, gives from the nuclear information analysis module in batches and handle.Further comprise urgent judge module, cache module, thresholding judge module from the nuclear information cache module.Urgent judge module is used to judge whether the system resource status information that reports from nuclear belongs to emergence message, and if belonged to directly notice carry out analyzing and processing from the nuclear information analysis module, store if do not belong to then send it to cache module.Whether the quantity of the system resource status information that the thresholding judge module is used for judging that cache module is stored has surpassed buffer threshold, and if surpassed directly notice carry out analyzing and processing from the nuclear information analysis module, if with ing above do not process.
Be used for the system resource status information that reports from nuclear single or in batches is analyzed from the nuclear information analysis module, and produce analysis result funding source allocation strategy adjusting module and use.
The resource allocation policy adjusting module is used for the system resource status information that reports from nuclear operating system is analyzed, and according to analysis result the allocation strategy of resource is dynamically adjusted.
The present invention also can have other various embodiments; under the situation that does not deviate from spirit of the present invention and essence thereof; those skilled in the art work as can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection domain of the appended claim of the present invention.
Claims (10)
1, the method for managing resource in a kind of heterogeneous multi-core system is characterized in that, may further comprise the steps at least:
A, report its system resource status information to main nuclear operating system from nuclear operating system;
The described system resource status information that B, main nuclear operating system basis report is to respectively analyzing from nuclear operating-system resources user mode;
C, main nuclear operating system are adjusted resource allocation policy according to the result of described analysis.
2, method according to claim 1 is characterized in that, describedly regularly and/or according to alarming threshold reports its system resource status information to described main nuclear operating system from nuclear operating system.
3, method according to claim 1 is characterized in that, the described internuclear means of communication that provides from employing hardware between nuclear operating system and the described main nuclear operating system communicates as carrier.
4, method according to claim 1 is characterized in that, in the step 1, also comprises the step of described system resource status information being carried out buffer memory.
5, method according to claim 4 is characterized in that, for described buffer memory is provided with a buffer threshold, when the described system resource status information that reports reached described buffer threshold, described main nuclear operating system was carried out analyzing and processing in batches.
6, method according to claim 4 is characterized in that, when the described system resource status information that reports from nuclear operating system belongs to emergence message, described main nuclear operating system will be analyzed this system resource status information that belongs to emergence message at once.
7, method according to claim 6 is characterized in that, described main nuclear operating system also comprises the step of the system resource status information of buffer memory being carried out analyzing and processing when the described system resource status information that belongs to emergence message of analyzing and processing.
8, the resource management system in a kind of heterogeneous multi-core system is characterized in that, comprising:
From the nuclear information reporting module, be used to collect from examining current system resource status information, and report main nuclear operating system;
From nuclear resource state adjusting module, be used for resource status being adjusted according to the resource adjustment instruction that main nuclear operating system sends;
From the nuclear information cache module, be used for the described system resource status information that reports from nuclear is carried out buffer memory;
From the nuclear information analysis module, be used for the described system resource status information that reports from nuclear is analyzed;
The resource allocation policy adjusting module is used for adjusting resource allocation policy according to described analysis result from the nuclear information analysis module, to sending resource adjustment instruction from nuclear operating system.
9, system according to claim 1 is characterized in that, describedly comprises from the nuclear information cache module:
Urgent judge module is used to judge whether the described system resource status information that reports from nuclear of reception belongs to emergence message, if then described emergence message is sent to described from the nuclear information analysis module; If not, then the described system resource status information that receives is cached in the cache module;
Cache module is used for the described system resource status information that buffer memory receives.
10, system according to claim 1, it is characterized in that, describedly also comprise a thresholding judge module from the nuclear information cache module, be used for the system resource amount of state information that reports from nuclear of described buffer memory is monitored, the system resource status information with described buffer memory when reaching thresholding sends to described from the nuclear information analysis module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007101764277A CN101419561A (en) | 2007-10-26 | 2007-10-26 | Resource management method and system in isomerization multicore system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007101764277A CN101419561A (en) | 2007-10-26 | 2007-10-26 | Resource management method and system in isomerization multicore system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101419561A true CN101419561A (en) | 2009-04-29 |
Family
ID=40630357
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007101764277A Pending CN101419561A (en) | 2007-10-26 | 2007-10-26 | Resource management method and system in isomerization multicore system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101419561A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102096605A (en) * | 2011-02-17 | 2011-06-15 | 成电汽车电子产业园(昆山)有限公司 | Multi-level resource management implementation method in embedded real-time operation system |
| WO2011079424A1 (en) * | 2009-12-31 | 2011-07-07 | 西门子公司 | Method and compiler for parallelizing automatic control program |
| CN102231684A (en) * | 2011-06-16 | 2011-11-02 | 华为技术有限公司 | Interface board state detection method, multi-core central processing unit, interface board and router |
| CN102238549A (en) * | 2010-04-30 | 2011-11-09 | 索尼公司 | System and method for managing resources in heterogeneous network |
| CN102324967A (en) * | 2011-05-24 | 2012-01-18 | 中兴通讯股份有限公司 | Method and system for mutual scheduling of equipment resources |
| CN102573072A (en) * | 2010-12-29 | 2012-07-11 | 中国移动通信集团湖南有限公司 | Self-adaptive regulating method and device of network resource allocation as well as communication system |
| CN102722414A (en) * | 2012-05-22 | 2012-10-10 | 中国科学院计算技术研究所 | Input/output (I/O) resource management method for multi-root I/O virtualization sharing system |
| CN102736953A (en) * | 2011-04-08 | 2012-10-17 | 腾讯科技（深圳）有限公司 | System resource allocation method and equipment |
| CN103279389A (en) * | 2013-04-28 | 2013-09-04 | 中国工商银行股份有限公司 | Scheduling method and scheduling system of host computer batch loads |
| CN103729248A (en) * | 2012-10-16 | 2014-04-16 | 华为技术有限公司 | Method and device for determining tasks to be migrated based on cache perception |
| CN104714924A (en) * | 2013-12-12 | 2015-06-17 | 华为技术有限公司 | Resource control method and device |
| CN104714821A (en) * | 2013-12-17 | 2015-06-17 | 华为技术有限公司 | Operating system instance establishing method and device |
| CN104932945A (en) * | 2015-06-18 | 2015-09-23 | 合肥工业大学 | Task-level out-of-order multi-issue scheduler and scheduling method thereof |
| WO2017114061A1 (en) * | 2015-12-28 | 2017-07-06 | 中兴通讯股份有限公司 | Asymmetric multiprocessing system and method for managing hardware resource thereof |
| CN104838360B (en) * | 2012-09-04 | 2018-09-07 | 微软技术许可有限责任公司 | Resource management based on quota |
| CN109189580A (en) * | 2018-09-17 | 2019-01-11 | 武汉虹旭信息技术有限责任公司 | A kind of multitask development model and its method based on multi-core platform |
| WO2022204873A1 (en) * | 2021-03-29 | 2022-10-06 | 华为技术有限公司 | Electronic apparatus, system on chip, and physical core allocation method |
-
2007
- 2007-10-26 CN CNA2007101764277A patent/CN101419561A/en active Pending
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011079424A1 (en) * | 2009-12-31 | 2011-07-07 | 西门子公司 | Method and compiler for parallelizing automatic control program |
| CN106131858B (en) * | 2010-04-30 | 2019-12-13 | 索尼公司 | system and method for managing resources in heterogeneous network |
| CN102238549A (en) * | 2010-04-30 | 2011-11-09 | 索尼公司 | System and method for managing resources in heterogeneous network |
| CN106131858A (en) * | 2010-04-30 | 2016-11-16 | 索尼公司 | The system and method for resource is managed in heterogeneous network |
| CN102238549B (en) * | 2010-04-30 | 2016-09-14 | 索尼公司 | The system and method for resource is managed in heterogeneous network |
| CN102573072A (en) * | 2010-12-29 | 2012-07-11 | 中国移动通信集团湖南有限公司 | Self-adaptive regulating method and device of network resource allocation as well as communication system |
| CN102096605B (en) * | 2011-02-17 | 2012-10-03 | 成电汽车电子产业园(昆山)有限公司 | Multi-level resource management implementation method in embedded real-time operation system |
| CN102096605A (en) * | 2011-02-17 | 2011-06-15 | 成电汽车电子产业园(昆山)有限公司 | Multi-level resource management implementation method in embedded real-time operation system |
| CN102736953A (en) * | 2011-04-08 | 2012-10-17 | 腾讯科技（深圳）有限公司 | System resource allocation method and equipment |
| CN102736953B (en) * | 2011-04-08 | 2015-06-17 | 腾讯科技（深圳）有限公司 | System resource allocation method and equipment |
| CN102324967A (en) * | 2011-05-24 | 2012-01-18 | 中兴通讯股份有限公司 | Method and system for mutual scheduling of equipment resources |
| CN102231684A (en) * | 2011-06-16 | 2011-11-02 | 华为技术有限公司 | Interface board state detection method, multi-core central processing unit, interface board and router |
| CN102231684B (en) * | 2011-06-16 | 2013-11-06 | 华为技术有限公司 | Interface board state detection method, multi-core central processing unit, interface board and router |
| CN102722414A (en) * | 2012-05-22 | 2012-10-10 | 中国科学院计算技术研究所 | Input/output (I/O) resource management method for multi-root I/O virtualization sharing system |
| CN104838360B (en) * | 2012-09-04 | 2018-09-07 | 微软技术许可有限责任公司 | Resource management based on quota |
| CN103729248B (en) * | 2012-10-16 | 2017-12-15 | 华为技术有限公司 | A kind of method and apparatus of determination based on cache perception task to be migrated |
| US9483321B2 (en) | 2012-10-16 | 2016-11-01 | Huawei Technologies Co., Ltd. | Method and apparatus for determining to-be-migrated task based on cache awareness |
| CN103729248A (en) * | 2012-10-16 | 2014-04-16 | 华为技术有限公司 | Method and device for determining tasks to be migrated based on cache perception |
| CN103279389A (en) * | 2013-04-28 | 2013-09-04 | 中国工商银行股份有限公司 | Scheduling method and scheduling system of host computer batch loads |
| CN103279389B (en) * | 2013-04-28 | 2016-06-22 | 中国工商银行股份有限公司 | The dispatching method of a kind of host batch load and dispatching patcher |
| CN104714924B (en) * | 2013-12-12 | 2018-07-13 | 华为技术有限公司 | A kind of resource control method and device |
| CN104714924A (en) * | 2013-12-12 | 2015-06-17 | 华为技术有限公司 | Resource control method and device |
| CN104714821A (en) * | 2013-12-17 | 2015-06-17 | 华为技术有限公司 | Operating system instance establishing method and device |
| CN104714821B (en) * | 2013-12-17 | 2018-07-31 | 华为技术有限公司 | Operation system example creation method and device |
| CN104932945B (en) * | 2015-06-18 | 2018-05-18 | 合肥工业大学 | A kind of out of order multi-emitting scheduler of task level and its dispatching method |
| CN104932945A (en) * | 2015-06-18 | 2015-09-23 | 合肥工业大学 | Task-level out-of-order multi-issue scheduler and scheduling method thereof |
| WO2017114061A1 (en) * | 2015-12-28 | 2017-07-06 | 中兴通讯股份有限公司 | Asymmetric multiprocessing system and method for managing hardware resource thereof |
| CN109189580A (en) * | 2018-09-17 | 2019-01-11 | 武汉虹旭信息技术有限责任公司 | A kind of multitask development model and its method based on multi-core platform |
| WO2022204873A1 (en) * | 2021-03-29 | 2022-10-06 | 华为技术有限公司 | Electronic apparatus, system on chip, and physical core allocation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101419561A (en) | Resource management method and system in isomerization multicore system | |
| US10445850B2 (en) | Technologies for offloading network packet processing to a GPU | |
| CN103841052B (en) | A kind of bandwidth resource allocation System and method for | |
| CN111198759B (en) | Memory optimization method, system, terminal equipment and readable storage medium | |
| KR101565865B1 (en) | Method, apparatus, and system for scheduling processor core in multiprocessor core system | |
| CN102999377B (en) | Service concurrent access control method and device | |
| US9652150B2 (en) | Global memory sharing method and apparatus, and communications system | |
| CN101547150B (en) | method and device for scheduling data communication input port | |
| CN103684916A (en) | Method and system for intelligent monitoring and analyzing under cloud computing | |
| CN107766145B (en) | Memory management method and device under dual systems | |
| CN109067597A (en) | A kind of distributed system dynamic and intelligent service administering method | |
| CN103812750B (en) | System and method for protecting data communication equipment CPU receiving and transmitting message | |
| CN102567072A (en) | Resource allocation method, resource allocation device and resource allocation system | |
| CN103365726A (en) | Resource management method and system facing GPU (Graphic Processing Unit) cluster | |
| CN102081554A (en) | Cloud computing operating system as well as kernel control system and method thereof | |
| CN108028850B (en) | Private cloud management platform | |
| CN111200541A (en) | Network data processing method and device | |
| CN111586140A (en) | Data interaction method and server | |
| CN113867941A (en) | Task management system applied to server cluster | |
| CN101442439B (en) | Method for reporting interruption and PCI bus system | |
| CN110838987B (en) | Queue current limiting method and storage medium | |
| CN1988684B (en) | Network process managing method and device | |
| CN102143053B (en) | Method, device and system for transmitting data | |
| CN113905091B (en) | Method and device for processing access request | |
| CN106570011A (en) | Distributed crawler URL seed distribution method, dispatching node, and grabbing node |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20090429 |