CN108228333A - A kind of internuclear resource isolation method of multiple nucleus system - Google Patents
A kind of internuclear resource isolation method of multiple nucleus system Download PDFInfo
- Publication number
- CN108228333A CN108228333A CN201611154106.2A CN201611154106A CN108228333A CN 108228333 A CN108228333 A CN 108228333A CN 201611154106 A CN201611154106 A CN 201611154106A CN 108228333 A CN108228333 A CN 108228333A
- Authority
- CN
- China
- Prior art keywords
- processor core
- resource
- private data
- multiple nucleus
- core
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5066—Algorithms for mapping a plurality of inter-dependent sub-tasks onto a plurality of physical CPUs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/16—Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Storage Device Security (AREA)
Abstract
The invention belongs to computer system software professional skill fields, are related to the internuclear protection of resources method of multi-core operation system.Internuclear resource in the present invention refers to the private privileges for belonging to each processor core in multiple nucleus system, can only be accessed by present processor core.The present invention defines a private data section, in systems by the way that data attribute mode is set to belong to the resource of processor core in private data section.In operating system system initialization, private data section is mapped to different physical address spaces by each processor core by memory management unit function.By this mapping mode, although the access of each processor core is same logical address, corresponding different physical address, so actually accessing different data resources.The present invention realizes the security isolation protection of resource between multiple nucleus system center, improves safety and the access performance of internuclear resource.
Description
Technical field
The invention belongs to computer system software professional skill field, it is related to privately owned between the processor core of multi-core operation system
Resource isolation guard method.
Background technology
Internuclear resource in multiple nucleus system refers to the private privileges for belonging to each processor core, can only be visited by present processor core
It asks.The design of multi-core operation system is complex at present, about the separation and protection of internuclear resource particularly in system.Multinuclear system
The internuclear resource resource security of system influences safety, stability and the performance of whole system.Some existing multi-core operation systems
Typically by resource privatization, with the formal definition of every processor core variable, processor core when accessing the private privileges of oneself,
Every processor core variable is indexed by processor core number.This mode designs simply, and autgmentability is also fine, but safety is not
Height, performance are also bad.Although data definition is indexed access, reality into every processor core variable by processor core number
Upper or global resource, a processor core can be indexed using arbitrary core number accesses random processor nuclear resource, including
Present processor core, the also resource including other processor cores.Therefore this design method could not play isolation processing device core money
The effect in source.On the other hand it is exactly resource access performance, each processor core goes index resource again by obtaining core number, added
It obtains core number operation and affects system performance;Secondly it many data resources and non-critical is defined, therefore deposit by caching block size
When a core accesses the private privileges of oneself, additional the private privileges of other processor cores can be loaded into caching, when
During other processor cores update fornication resource, and buffer consistency affairs can be triggered, therefore reduce processor core private privileges
Access performance.
Invention content
The purpose of the present invention:
The present invention is in order to solve the safety issue of processor core resource in multi-core operation system, it is proposed that a kind of processor
The insulation blocking method of nuclear resource.
Technical scheme of the present invention defines:
Technical scheme of the present invention includes the design of multi-core operation internal system and the storage management list of hardware processor core
First mapping relations definition.Resource refers to the data for belonging to some processor core between processor core, it can only be visited by present processor core
It asks, scheduling queue, scheduling identification including processor each in system, executable task dispatching.Whether one data is processor core
Between resource, by operating system design, personnel determine.
Present invention private data section defined in multi-core operation system, the section are responsible for storing processor nuclear resource.System
If the data in system belong to processor core resource, by the data definition in private data section.When system starts, processor core
The MMU memory management unit of oneself is initialized, private data section is mapped to one section of physical space, the logic of each processor nuclear mapping
Address is the same, is all privately owned section of address, but physical address is different, is all one section of independent space.In system operation,
Each processor core accesses private data section by logical address, accesses that section of physical space of oneself mapping respectively, therefore practical
On be to access different address, so as to access respective private data.
Technical solution:
A kind of internuclear resource isolation method of multiple nucleus system, including following steps;(1) compiling of the multiple nucleus system
A private data section defined in script is linked, the private data segment base and size are with 4K byte-aligneds;(2) in institute
It states in multiple nucleus system program, the private privileges of processor core is placed them in by attribute definition mode in data segment;
According to the number of processor cores N of system in the compiling link script of the multiple nucleus system, it is big to reserve N number of private data section for system
Small space;(3) executable file is generated after the completion of multiple nucleus system compiling link, determines private data section size and reserved sky
Between address range;(4) when multiple nucleus system starts, each processor core maps private data section when initializing memory management unit
To reserved address space, private data section is mapped on i-th piece of reserved space by processor core i;Each processor core is completed
After the mapping of private data section, private data resource may have access to.
Necessary step behavioral illustrations:
The key step of the present invention is to define private data section in systems first, secondly utilizing hardware processor core
Private data section is mapped in respective physical space by MMU memory management unit function.
The present invention has the advantage that effect:
The present invention can realize the security isolation of resource between processor core first, and each processor core can only access oneself mapping
That section of physical space in data, resource is invisible between the processor core of other cores;Secondly access performance is improved, it is each to handle
Device core is all by the unified internuclear resource of logical address access process device, does not need to be indexed by core number;It improves again
Set expandability, the processor check figure that system is supported have no effect on the definition of resource between processor core, only influence needed for
Private data section is mapped to the privately owned physical space of oneself by the privately owned physical space quantity wanted, each processor core.
Description of the drawings
Fig. 1 is the internuclear resource space distribution of each processor core with accessing schematic diagram;
Fig. 2 multi-core processor cores private privileges isolation figure;
Specific embodiment
The present invention is as follows in the embodiment of some real-time multi-core operation system:
A) private data section, entitled pdatasec are defined in systems;
B) the data data_A and data_B in system belong to resource between processor core;
C) when defining data data_A and data_B, the two data are placed in pdatasec sections;
D) the pdatasec segment bases for assuming System build generation are 0XA00000, size 4KB, data data_A
Address for 0XA00000, the address of data data_B is 0XA000600;
E) when system starts, pdatasec is mapped to respectively privately owned by the storage management Elementary Function of each processor core oneself
Physical space, processor core 0 are 0x1000000, and core 1 is 0x1001000, and core 2 is 0x1002000, and core 3 is 0x1003000;
F) all using logical address 0XA00000 when all processor cores access data data_A during system operation, number is accessed
According to during data_B use logical address 0XA000600.But the physical address of actual access is different, when accessing data data_A,
Processor core 0 accesses physical address 0x1000000, and core 1 accesses physical address 0x1001000, and core 1 accesses physical address
0x1002000, core 1 access physical address 0x1003000;When accessing data data_A, processor core 0 accesses physical address
0x1000600, core 1 access physical address 0x1001600, and core 1 accesses physical address 0x1002600, and core 1 accesses physical address
0x1003600。
Each processor core accesses different physical address using same logical address, due to the storage management of processor core
Elementary Function so that core 0 can only access [0x1000000,0x1001000), other processors invisible to other address spaces
Core is also the same, therefore has achieved the purpose that the real internuclear resource of isolation processing device, ensures to improve access property while safety
Energy and autgmentability.
Claims (1)
1. a kind of internuclear resource isolation method of multiple nucleus system, it is characterised in that:Including following steps;
(1) a private data section defined in the compiling link script of the multiple nucleus system, the private data segment base
And size is with 4K byte-aligneds;
(2) in the multiple nucleus system program, the private privileges of processor core are placed them in by attribute definition mode
In data segment;According to the number of processor cores N of system in the compiling link script of the multiple nucleus system, reserved for system N number of
The space of private data section size;
(3) generate executable file after the completion of multiple nucleus system compiling link, determine private data section size and it is reserved spatially
Location range;
(4) multiple nucleus system start when, each processor core mapped when initialize memory management unit private data section to reserve
Address space, private data section is mapped on i-th piece of reserved space by processor core i;Each processor core completes privately owned number
After being mapped according to section, private data resource may have access to.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611154106.2A CN108228333A (en) | 2016-12-14 | 2016-12-14 | A kind of internuclear resource isolation method of multiple nucleus system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611154106.2A CN108228333A (en) | 2016-12-14 | 2016-12-14 | A kind of internuclear resource isolation method of multiple nucleus system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108228333A true CN108228333A (en) | 2018-06-29 |
Family
ID=62638459
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611154106.2A Pending CN108228333A (en) | 2016-12-14 | 2016-12-14 | A kind of internuclear resource isolation method of multiple nucleus system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108228333A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109032938A (en) * | 2018-07-17 | 2018-12-18 | 中国航空无线电电子研究所 | Multi-core DSP program development adjustment method, documentation of program and loading method |
CN114327777A (en) * | 2021-12-30 | 2022-04-12 | 元心信息科技集团有限公司 | Method and device for determining global page directory, electronic equipment and storage medium |
US11755785B2 (en) | 2020-08-03 | 2023-09-12 | Nxp Usa, Inc. | System and method of limiting access of processors to hardware resources |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101477477A (en) * | 2009-01-12 | 2009-07-08 | 华为技术有限公司 | Kernel spacing isolation method, spacing management entity and system |
CN102713847A (en) * | 2009-12-29 | 2012-10-03 | 超威半导体公司 | Hypervisor isolation of processor cores |
CN102938035A (en) * | 2012-11-08 | 2013-02-20 | 西安交通大学 | Driving separation system inside virtual machine and method |
CN103608792A (en) * | 2013-05-28 | 2014-02-26 | 华为技术有限公司 | Method and system for supporting resource isolation under multi-core architecture |
CN104657193A (en) * | 2013-11-21 | 2015-05-27 | 华为技术有限公司 | Method and device for accessing to physical resources |
-
2016
- 2016-12-14 CN CN201611154106.2A patent/CN108228333A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101477477A (en) * | 2009-01-12 | 2009-07-08 | 华为技术有限公司 | Kernel spacing isolation method, spacing management entity and system |
CN102713847A (en) * | 2009-12-29 | 2012-10-03 | 超威半导体公司 | Hypervisor isolation of processor cores |
CN102938035A (en) * | 2012-11-08 | 2013-02-20 | 西安交通大学 | Driving separation system inside virtual machine and method |
CN103608792A (en) * | 2013-05-28 | 2014-02-26 | 华为技术有限公司 | Method and system for supporting resource isolation under multi-core architecture |
CN104657193A (en) * | 2013-11-21 | 2015-05-27 | 华为技术有限公司 | Method and device for accessing to physical resources |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109032938A (en) * | 2018-07-17 | 2018-12-18 | 中国航空无线电电子研究所 | Multi-core DSP program development adjustment method, documentation of program and loading method |
US11755785B2 (en) | 2020-08-03 | 2023-09-12 | Nxp Usa, Inc. | System and method of limiting access of processors to hardware resources |
CN114327777A (en) * | 2021-12-30 | 2022-04-12 | 元心信息科技集团有限公司 | Method and device for determining global page directory, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108369507A (en) | For using the method and apparatus for handling process instruction in memory | |
US9208088B2 (en) | Shared virtual memory management apparatus for providing cache-coherence | |
CN103729238B (en) | Method and system for managing a lock to a resource shared among a plurality of processors | |
US10061701B2 (en) | Sharing of class data among virtual machine applications running on guests in virtualized environment using memory management facility | |
DE112020004181T5 (en) | PROVIDE DIRECT DATA ACCESS BETWEEN ACCELERATORS AND STORAGE IN A COMPUTING ENVIRONMENT | |
CN110663025A (en) | Core mapping | |
CN113094700A (en) | System for executing safety operation and method for executing safety operation by system | |
US10331591B2 (en) | Logical-to-physical block mapping inside the disk controller: accessing data objects without operating system intervention | |
CN103065084A (en) | Windows hidden process detection method performed at external machine of virtual machine | |
US11500828B1 (en) | Method and device for constructing database model with ID-based data indexing-enabled data accessing | |
Gaud et al. | Efficient workstealing for multicore event-driven systems | |
CN108228333A (en) | A kind of internuclear resource isolation method of multiple nucleus system | |
Skarlatos et al. | Babelfish: Fusing address translations for containers | |
EP4209914A1 (en) | Reconfigurable cache architecture and methods for cache coherency | |
Hamidouche et al. | Gpu initiated openshmem: correct and efficient intra-kernel networking for dgpus | |
Chen et al. | Concurrent hash tables on multicore machines: Comparison, evaluation and implications | |
CN105278880A (en) | Cloud computing virtualization-based memory optimization device and method | |
Fahmy et al. | CoS-HDFS: Co-locating geo-distributed spatial data in hadoop distributed file system | |
Kommareddy et al. | Exploring allocation policies in disaggregated non-volatile memories | |
CN108885569A (en) | Operation when identifying enhanced simultaneously operating result to improve operation | |
WO2023029756A1 (en) | Sleep / wake-up performance enhancing for simultaneous address translation table walks | |
CN110336894A (en) | A kind of virtual machine user space address is to host kernel spacing address conversion method | |
Addisie et al. | Collaborative accelerators for streamlining MapReduce on scale-up machines with incremental data aggregation | |
CN109582603A (en) | For providing the technology of access protection to shared virtual memory | |
CN113986775B (en) | Page table item generation method, system and device in RISC-V CPU verification |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180629 |
|
RJ01 | Rejection of invention patent application after publication |