CN103269317A - Lock-free communication method and system based on symmetric multiprocessing SMP system - Google Patents
Lock-free communication method and system based on symmetric multiprocessing SMP system Download PDFInfo
- Publication number
- CN103269317A CN103269317A CN2013101417629A CN201310141762A CN103269317A CN 103269317 A CN103269317 A CN 103269317A CN 2013101417629 A CN2013101417629 A CN 2013101417629A CN 201310141762 A CN201310141762 A CN 201310141762A CN 103269317 A CN103269317 A CN 103269317A
- Authority
- CN
- China
- Prior art keywords
- data flow
- snat
- port
- socket
- lock
- 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.)
- Granted
Links
Images
Landscapes
- Computer And Data Communications (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a lock-free communication method based on an SMP system. The SMP system comprises a plurality of sockets, wherein each socket is of an SMP structure, and ports of the sockets are independent. Lock-free communication carried out by each socket comprises the following steps that ips in ip sections are allocated to various data flow processing cores through the Hash algorithm, a plurality of ip sections of snat are allocated on each port of the sockets, and the ips of the snat are inserted into a maintenance table of a receiver; when the receiver detects that a reverse data flow returns back, the receiver detects whether the destination address of the reverse data flow is the ip section of the snat at the port where the reverse data flow exists or not; if the destination address is the ip section of the snat at the port where the reverse data flow exists, the reverse data flow is allocated to the reverse data flow core comprising a private address by adoption of the ips through the Hash algorithm. The lock-free communication method achieves bidirectional flow affinity, is programmable and strong in controllability, a lock-free queue avoids competition expenses, and system performance is optimized. The invention further discloses a lock-free communication system based on the SMP system.
Description
Technical field
The present invention relates to field of computer technology, particularly a kind of nothing lockization communication means and system based on symmetrical multiprocessing smp system.
Background technology
SMP(Symmetrical Multi-Processing, symmetric multiprocessor system) in the framework, can there be the situation of shared resource in nucleus module.The process executed in parallel of moving on a plurality of processors may have several processes and visit certain shared resource simultaneously, need consider the lock expense that shared drive brings when programming.
Among the message reception technique RSS (Receive-Side Scaling) of main flow network interface card, hash algorithm is to programme in network interface card hardware at present, therefore can't externally carry out custom programming, can not adjust hash algorithm as required.The RSS technology can guarantee the unidirectional compatibility of data flow, therefore, for the network equipment of only handling unidirectional traffic, can accomplish to avoid shared drive, but RSS can not guarantee that bidirectional traffic is affine.
For general device for processing streams, especially has NAT (Network Address Translation, network address translation) equipment of function, if can accomplish that bidirectional flow is affine, guarantee that namely same flows to out all by same cpu processing, just can avoid examining with internuclear the sharing of internal memory, thereby eliminate the lock expense that shared drive brings.But also there is not good solution at present.
Summary of the invention
The present invention is intended to solve at least one of technical problem that exists in the prior art.For this reason, one object of the present invention is to propose a kind of nothing lockization communication means based on symmetrical multiprocessing smp system.This method has realized that bidirectional flow is affine, and able to programme, controllability is strong, and lock-free queue avoids competing expense, has optimized systematic function.
Second purpose of the present invention is to propose a kind of nothing lockization communication system based on symmetrical multiprocessing smp system.
For achieving the above object, the embodiment of first aspect present invention has proposed a kind of nothing lockization communication means based on symmetrical multiprocessing smp system, described smp system comprises a plurality of socket, wherein each described socket is the SMP structure, port between described a plurality of socket is independent, wherein, each described socket does not have lockization communication and comprises the steps:
Ip in the described ip section is dispensed on each Data Stream Processing core by hash algorithm, and to the ip section of a plurality of source addresses of each port arrangement of described socket conversion snat, the ip of snat is inserted in the Maintenance Table of receiver, wherein, the source address of forward data stream is set to the privately owned address of current place Data Stream Processing core; When described receiver had detected reverse data flow and returns, whether the destination address that described receiver detects described reverse data flow was the ip section of the snat of self place port; If then adopt ip distributing to of described reverse data flow to be comprised the reverse data flow core of described privately owned address by hash algorithm.
The nothing lockization communication means based on symmetrical multiprocessing smp system according to the embodiment of the invention, by hash algorithm and lock-free queue technology, guarantee the two-way compatibility of data flow, being same flows to out by same cpu processing, the lock competition that internal memory between avoiding examining and examining is shared with and brought, eliminated the lock expense that shared drive brings, to user transparent, data flow has been had highly controlled.Able to programme, flexibility and controllability are strong, and the effect of performance optimization simultaneously is obvious.
In one embodiment of the invention, described hash algorithm is to get surplus hash algorithm.
In one embodiment of the invention, in second address of the ip section of the snat of the described port correspondence of each port arrangement of described socket.
In one embodiment of the invention, be provided with among each described socket and have receiving data stream and handle core, described receiving data stream is handled core and by lock-free queue the data flow that receives is assigned on each Data Stream Processing core according to the five-tuple Hash.
The embodiment of second aspect present invention has proposed a kind of nothing lockization communication system based on symmetrical multiprocessing smp system, described smp system comprises a plurality of socket, wherein each described socket is the SMP structure, port between described a plurality of socket is independent, described nothing lockization communication system based on symmetrical multiprocessing smp system does not have the lockization communication process to each described socket, and wherein said nothing lockization communication system based on symmetrical multiprocessing smp system comprises: distributor, source address conversion snat device and receiver.
Distributor is used for the ip of described ip section is dispensed to each Data Stream Processing core by hash algorithm; Source address conversion snat device is used for the ip section to a plurality of source address conversion of each port arrangement of described socket snat, the ip of snat is inserted in the Maintenance Table of receiver, wherein, the source address of forward data stream is set to the privately owned address of current place Data Stream Processing core; Described receiver is used for the ip of storage snat, and when having detected reverse data flow and return, and whether the destination address that described receiver detects described reverse data flow is the ip section of the snat of self place port; When described distributor also is used for destination address in described reverse data flow and is the ip section of snat of self place port, adopt ip distributing to of described reverse data flow to be comprised the reverse data flow core of described privately owned address by hash algorithm.
The nothing lockization communication system based on symmetrical multiprocessing smp system according to the embodiment of the invention, by hash algorithm and lock-free queue technology, guarantee the two-way compatibility of data flow, being same flows to out by same cpu processing, the lock competition that internal memory between avoiding examining and examining is shared with and brought, eliminated the lock expense that shared drive brings, to user transparent, data flow has been had highly controlled.Able to programme, flexibility and controllability are strong, and the effect of performance optimization simultaneously is obvious.
In one embodiment of the invention, described hash algorithm is to get surplus hash algorithm.
In one embodiment of the invention, also comprise configuration module, be used for second address in the ip section of the snat of the described port correspondence of each port arrangement of described socket.
Additional aspect of the present invention and advantage part in the following description provide, and part will become obviously from the following description, or recognize by practice of the present invention.
Description of drawings
Above-mentioned and/or additional aspect of the present invention and advantage are from obviously and easily understanding becoming the description of embodiment in conjunction with following accompanying drawing, wherein:
Fig. 1 is the flow chart based on the nothing lockization communication means of symmetrical multiprocessing smp system according to the embodiment of the invention;
Fig. 2 is the Organization Chart according to the symmetrical multiprocessing smp system of the embodiment of the invention;
Fig. 3 is the distribution schematic diagram according to the snat ip of the embodiment of the invention; With
Fig. 4 is the structural representation based on the nothing lockization communication system of symmetrical multiprocessing smp system according to the embodiment of the invention.
Embodiment
Describe embodiments of the invention below in detail, the example of described embodiment is shown in the drawings, and wherein identical or similar label is represented identical or similar elements or the element with identical or similar functions from start to finish.Be exemplary below by the embodiment that is described with reference to the drawings, only be used for explaining the present invention, and can not be interpreted as limitation of the present invention.
Below with reference to the nothing lockization communication means based on symmetrical multiprocessing smp system of Fig. 1 description according to the embodiment of the invention, smp system comprises a plurality of socket, wherein each socket is the SMP structure, port between a plurality of socket is independent, wherein, each socket does not have lockization communication and comprises the steps:
Step S110: the ip in the ip section is dispensed on each Data Stream Processing core by hash algorithm, and to the ip section of a plurality of source addresses of each port arrangement of socket conversion snat, the ip of snat is inserted in the Maintenance Table of receiver, wherein, the source address of forward data stream is set to the privately owned address of current place Data Stream Processing core.
Wherein, be provided with among each socket and have receiving data stream and handle core, receiving data stream is handled core and by lock-free queue the data flow that receives is assigned on each Data Stream Processing core according to the five-tuple Hash.
Step S120: when receiver had detected reverse data flow and returns, whether the destination address that receiver detects reverse data flow was the ip section of the snat of self place port.
Step S130: if then adopt ip distributing to of reverse data flow to be comprised the reverse data flow core of privately owned address by hash algorithm.
Wherein, hash algorithm is to get surplus hash algorithm.
Wherein, in second address of the ip section of the snat of each port arrangement port correspondence of socket.
Specific embodiment below with reference to Fig. 2 describes this method.Fig. 2 is the Organization Chart of symmetrical multiprocessing smp system, as shown in the figure, server in the present embodiment adopts NUMA(Non Uniform MemoryAccess Achitecture) framework, the server in the smp system has two socket, is respectively socketS1 and socket S2.Each socket is the SMP framework, and the port between a plurality of socket is independent.Concrete server configuration parameter is as follows:
Cpu：westmere?E56452.4GHZ
Socket?number:2
Cpu?number:2
Cpu?cores：6
L1cachesize:32k
L2cachesize:256k
L3cachesize:12288k
Memory:96GB
Smp system adopts the message tupe of pipe-line.Because the NUMA structure has symmetry, so only analyzes socket S1 in the following Example, another socket S2 by that analogy.Core0 is that the receiving data stream of socket S1 is handled core, be used for packet receiving, by lock-free queue Q1 the packet of receiving is assigned on each Data Stream Processing core according to the five-tuple Hash, in Fig. 2, be assigned to each Data Stream Processing core among the socket S1 and be respectively core1, core2, core3, core4, core5.Thereby bonding to stream affine.
The Data Stream Processing core needs to be source address conversion snat after handle packet, the source address of forward data stream is set to the privately owned address of current place Data Stream Processing core.Receiving data stream is handled core and is being received with this after privately owned address is the packet of destination address, can use the hash algorithm identical with distributing privately owned address, give the Data Stream Processing core that has this privately owned address with allocation of packets, thereby realize that bidirectional flow is affine.
Particularly, as shown in Figure 3, the realization that bidirectional flow is affine further comprises:
Dispose the ip section of a plurality of source addresses conversion snat at each port (port) at socket by configuration file.As shown in Figure 3, the ip section that comprises respectively of port0 and port1 is snat ip list1 and snat ip list2.During initialization, the ip in the ip section by getting surplus hash algorithm, is dispensed on each Data Stream Processing core core a as shown in Figure 3, core b, core c.The ip section of snat is inserted in the Maintenance Table of receiver Receiver/Dispatcher, as T1, T2, T3.When reverse data flow was returned, receiver was identified destination address, judged whether to belong to the snat ip section of self place port.If, adopt ip by getting surplus hash algorithm, with this allocation of packets to the reverse data flow core that comprises privately owned address, thereby guarantee that same data flow turnover direction is on same core.
For the reverse flow packet is received from identical port with the positive flow packet, dispose second address of snat ip section at each port of socket.For example the snat ip section on the port port0 is 1.1.1.0/24, and then on port0 after the configuration address 1.1.1.2/24, destination address belongs to the packet of 1.1.1.0/24 can be by direct-connected routing forwarding to port0.
In order to reduce the loss that competition brings to greatest extent, packet receives the lock-free queue that adopts single producer, single consumer between core and Data Stream Processing core and realizes, and has adopted batch processing and means such as look ahead are optimized.Be understandable that above-mentioned explanation is order ground presented for purpose of illustration only, embodiments of the invention are not limited thereto.
The nothing lockization communication means based on symmetrical multiprocessing smp system of the embodiment of the invention, a kind of hash method of handling based on the nothing lockization stream of SMP structure has been proposed, able to programme, guaranteed the two-way compatibility of data flow, having avoided internal memory to share the lock that brings competes, to user transparent, data flow is had highly controlled, simultaneously the performance optimization effect is obvious.
Below with reference to the nothing lockization communication system 100 based on symmetrical multiprocessing smp system of Fig. 4 description according to the embodiment of the invention, smp system comprises a plurality of socket, wherein each socket is the SMP structure, port between a plurality of socket is independent, nothing lockization communication system based on symmetrical multiprocessing smp system does not have the lockization communication process to each socket, and wherein the nothing lockization communication system based on symmetrical multiprocessing smp system comprises:
Wherein, the nothing lockization communication system 100 based on symmetrical multiprocessing smp system comprises: distributor 110, source address conversion snat device 120, receiver distributor 130 and configuration module 140.
Configuration module is used for second address in the ip section of the snat of each port arrangement port correspondence of socket.
Wherein, hash algorithm is to get surplus hash algorithm.
Specific embodiment below with reference to Fig. 2 describes native system.As shown in Figure 2, the server in the present embodiment adopts NUMA(Non Uniform Memory Access Achitecture) framework, the server in the smp system has two socket, is respectively socket S1 and socket S2.Each socket is the SMP framework, and the port between a plurality of socket is independent.Concrete server configuration parameter is as follows:
Cpu：westmere?E56452.4GHZ
Socket?number:2
Cpu?number:2
Cpu?cores：6
L1cachesize:32k
L2cachesize:256k
L3cachesize:12288k
Memory:96GB
Smp system adopts the message tupe of pipe-line.Because the NUMA structure has symmetry, so only analyzes socket S1 in the following Example, another socket S2 by that analogy.Core0 is that the receiving data stream of socket S1 is handled core, be used for packet receiving, distributor 110 is assigned to the packet of receiving on each Data Stream Processing core according to the five-tuple Hash by lock-free queue Q1, in Fig. 2, distributor 110 is assigned to each Data Stream Processing core and is respectively core1, core2, core3, core4, core5, thus bonding to stream affine.
Source address conversion snat device 120 is source address conversion snat behind Data Stream Processing core processing packet, the source address of forward data stream is set to the privately owned address of current place Data Stream Processing core.Receiving data stream is handled core and is being received with this after privately owned address is the packet of destination address, can use the hash algorithm identical with distributing privately owned address, give the Data Stream Processing core that has this privately owned address with allocation of packets, thereby realize that bidirectional flow is affine.
Particularly, as shown in Figure 3, the realization that bidirectional flow is affine further comprises:
Source address conversion snat device 120 disposes the ip section of a plurality of source addresses conversion snat at each port (port) at socket by configuration file.As shown in Figure 3, the ip section that comprises respectively of port0 and port1 is snat ip list1 and snat ip list2.During initialization, distributor 110 by getting surplus hash algorithm, is dispensed on each Data Stream Processing core core a as shown in Figure 3, core b, core c with the ip in the ip section.Source address conversion snat device 120 inserts the ip section of snat in the Maintenance Table of receiver 130Receiver/Dispatcher, as T1, T2, T3.When reverse data flow was returned, 130 pairs of destination addresses of receiver were identified, and judged whether to belong to the snatip section of self place port.If distributor 110 adopts ip by getting surplus hash algorithm, with this allocation of packets to the reverse data flow core that comprises privately owned address, thereby guarantee that same data flow turnover direction is on same core.
For the reverse flow packet is received from identical port with the positive flow packet, configuration module is in second address of each port configuration snat ip section of socket.For example the snatip section on the port port0 is 1.1.1.0/24, and then on port0 after the configuration address 1.1.1.2/24, destination address belongs to the packet of 1.1.1.0/24 can be by direct-connected routing forwarding to port0.
In order to reduce the loss that competition brings to greatest extent, packet receives the lock-free queue that adopts single producer, single consumer between core and Data Stream Processing core and realizes, and has adopted batch processing and means such as look ahead are optimized.Be understandable that above-mentioned explanation is order ground presented for purpose of illustration only, embodiments of the invention are not limited thereto.
The nothing lockization communication system based on symmetrical multiprocessing smp system of the embodiment of the invention, a kind of hash method of handling based on the nothing lockization stream of SMP structure has been proposed, able to programme, guaranteed the two-way compatibility of data flow, having avoided internal memory to share the lock that brings competes, to user transparent, data flow is had highly controlled, simultaneously the performance optimization effect is obvious.
In the description of this specification, concrete feature, structure, material or characteristics that the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means in conjunction with this embodiment or example description are contained at least one embodiment of the present invention or the example.In this manual, the schematic statement to above-mentioned term not necessarily refers to identical embodiment or example.And concrete feature, structure, material or the characteristics of description can be with the suitable manner combination in any one or more embodiment or example.
Although illustrated and described embodiments of the invention, for the ordinary skill in the art, be appreciated that without departing from the principles and spirit of the present invention and can carry out multiple variation, modification, replacement and modification to these embodiment that scope of the present invention is by claims and be equal to and limit.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310141762.9A CN103269317B (en) | 2013-04-22 | 2013-04-22 | Nothing lockization communication means based on symmetric multi-processors smp system and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310141762.9A CN103269317B (en) | 2013-04-22 | 2013-04-22 | Nothing lockization communication means based on symmetric multi-processors smp system and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103269317A true CN103269317A (en) | 2013-08-28 |
| CN103269317B CN103269317B (en) | 2016-12-28 |
Family
ID=49012926
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310141762.9A Active CN103269317B (en) | 2013-04-22 | 2013-04-22 | Nothing lockization communication means based on symmetric multi-processors smp system and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103269317B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105634958A (en) * | 2015-12-24 | 2016-06-01 | 东软集团股份有限公司 | Packet forwarding method and device based on multi-core system |
| CN105871741A (en) * | 2015-01-23 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Packet dispersion method and device |
| CN105871725A (en) * | 2015-01-23 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Message shunt method and device |
| CN108011991A (en) * | 2017-11-30 | 2018-05-08 | 新华三技术有限公司 | Stream compression forwarding method, master control borad, interface board, engine plate and distributed fire wall |
| CN116248590A (en) * | 2022-12-16 | 2023-06-09 | 中国联合网络通信集团有限公司 | Data forwarding method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060221832A1 (en) * | 2005-04-04 | 2006-10-05 | Sun Microsystems, Inc. | Virtualized partitionable shared network interface |
| CN101964749A (en) * | 2010-09-21 | 2011-02-02 | 北京网康科技有限公司 | Message retransmission method and system based on multi-core architecture |
| CN102549984A (en) * | 2009-05-05 | 2012-07-04 | 思杰系统有限公司 | Systems and methods for packet steering in a multi-core architecture |
| US8244864B1 (en) * | 2001-03-20 | 2012-08-14 | Microsoft Corporation | Transparent migration of TCP based connections within a network load balancing system |
-
2013
- 2013-04-22 CN CN201310141762.9A patent/CN103269317B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8244864B1 (en) * | 2001-03-20 | 2012-08-14 | Microsoft Corporation | Transparent migration of TCP based connections within a network load balancing system |
| US20060221832A1 (en) * | 2005-04-04 | 2006-10-05 | Sun Microsystems, Inc. | Virtualized partitionable shared network interface |
| CN102549984A (en) * | 2009-05-05 | 2012-07-04 | 思杰系统有限公司 | Systems and methods for packet steering in a multi-core architecture |
| CN101964749A (en) * | 2010-09-21 | 2011-02-02 | 北京网康科技有限公司 | Message retransmission method and system based on multi-core architecture |
Non-Patent Citations (1)
| Title |
|---|
| 姚崎: "基于队列的处理器亲和机制及在网络转发中的应用", 《北京交通大学学报》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105871741A (en) * | 2015-01-23 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Packet dispersion method and device |
| CN105871725A (en) * | 2015-01-23 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Message shunt method and device |
| CN105871741B (en) * | 2015-01-23 | 2018-12-25 | 阿里巴巴集团控股有限公司 | A kind of message diversion method and device |
| CN105871725B (en) * | 2015-01-23 | 2019-02-05 | 阿里巴巴集团控股有限公司 | A kind of message diversion method and device |
| CN105634958A (en) * | 2015-12-24 | 2016-06-01 | 东软集团股份有限公司 | Packet forwarding method and device based on multi-core system |
| CN105634958B (en) * | 2015-12-24 | 2019-05-31 | 东软集团股份有限公司 | Message forwarding method and device based on multiple nucleus system |
| CN108011991A (en) * | 2017-11-30 | 2018-05-08 | 新华三技术有限公司 | Stream compression forwarding method, master control borad, interface board, engine plate and distributed fire wall |
| CN116248590A (en) * | 2022-12-16 | 2023-06-09 | 中国联合网络通信集团有限公司 | Data forwarding method, device, equipment and storage medium |
| CN116248590B (en) * | 2022-12-16 | 2024-05-10 | 中国联合网络通信集团有限公司 | Data forwarding method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103269317B (en) | 2016-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113553185B (en) | Data processing system with hardware acceleration plane and software plane | |
| CN107066319B (en) | Multi-dimensional scheduling system for heterogeneous resources | |
| CN104166597B (en) | A kind of method and device for distributing long-distance inner | |
| US8379642B2 (en) | Multicasting using a multitiered distributed virtual bridge hierarchy | |
| EP2880828B1 (en) | System and method for virtual ethernet interface binding | |
| CN113259503B (en) | Method and system for realizing cross-network communication among different containers based on Infiniband | |
| CN105791448B (en) | An address allocation method, CGN equipment and CGN dual-master system | |
| US8594100B2 (en) | Data frame forwarding using a distributed virtual bridge | |
| US20140068132A1 (en) | Automatic construction of deadlock free interconnects | |
| KR20200135780A (en) | Mediating parts of a transaction through a virtual channel associated with the interconnect | |
| CN103117947B (en) | A kind of load sharing method and device | |
| CN102594660B (en) | A kind of virtual interface exchange method, Apparatus and system | |
| CN103269317A (en) | Lock-free communication method and system based on symmetric multiprocessing SMP system | |
| US20110261687A1 (en) | Priority Based Flow Control Within a Virtual Distributed Bridge Environment | |
| US11233752B2 (en) | Packet forwarding | |
| US10915370B2 (en) | Inter-host communication without data copy in disaggregated systems | |
| JP5834178B2 (en) | Semiconductor circuit bus system | |
| WO2025082101A1 (en) | Distributed computing method, apparatus, device and system, and readable storage medium | |
| US11875839B2 (en) | Flow based rate limit | |
| WO2017000823A2 (en) | Openflow protocol resources control method, device and system | |
| CN107547247B (en) | IP address allocation method and device for three-layer management network in intelligent elastic architecture | |
| CN103457862A (en) | Three-layer multicast achieving method in software longitudinal stacking system and equipment | |
| CN104360980A (en) | Multi-plane separating type multi-segment computer architecture | |
| TWI629887B (en) | A reconfigurable interconnect element with local lookup tables shared by multiple packet processing engines | |
| CN104092558A (en) | Multi-processor equipment IP configuration method and multi-processor equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |