CN104503928A - Random memory circuit based on queue management - Google Patents

Random memory circuit based on queue management Download PDF

Info

Publication number
CN104503928A
CN104503928A CN201410749390.2A CN201410749390A CN104503928A CN 104503928 A CN104503928 A CN 104503928A CN 201410749390 A CN201410749390 A CN 201410749390A CN 104503928 A CN104503928 A CN 104503928A
Authority
CN
China
Prior art keywords
buffering
buffer cell
scheduler
fifo
buffer
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
Application number
CN201410749390.2A
Other languages
Chinese (zh)
Inventor
张利洲
蒲恺
冯晓东
刘陈
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
AVIC No 631 Research Institute
Original Assignee
AVIC No 631 Research Institute
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by AVIC No 631 Research Institute filed Critical AVIC No 631 Research Institute
Priority to CN201410749390.2A priority Critical patent/CN104503928A/en
Publication of CN104503928A publication Critical patent/CN104503928A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/14Handling requests for interconnection or transfer
    • G06F13/16Handling requests for interconnection or transfer for access to memory bus
    • G06F13/1605Handling requests for interconnection or transfer for access to memory bus based on arbitration
    • G06F13/1642Handling requests for interconnection or transfer for access to memory bus based on arbitration with request queuing

Abstract

The invention provides a random memory circuit suitable for an FPGA logic circuit. A queue marker register is additionally arranged based on the conventional FIFO buffering interaction mode, and is used for realizing state description of a buffering unit; assurance of data submitting order is realized through the FIFO circuit; software can use buffering dynamically according to actual requirements; sending order management mechanism can be realized by operating the FIFO. Under conventional random access and queue access information interaction modes, the random memory circuit effectively combines the two modes, meets the management mode that random access is adopted at one end while queue dispatching is adopted at the other end, and effectively satisfies system application.

Description

A kind of based on queue management can memory circuit at random
Technical field:
The invention belongs to design of computer hardware field, relate to a kind of be applicable to fpga logic circuit can memory circuit at random.
Background technology:
When design height communication network board, need data-interface at a high speed to realize data interaction access between main frame and board, and interaction mechanism is simple and reliable, to meet the requirement of system for communication efficiency and application model.
Traditional interactive mode generally comprises the following two kinds from using forestland:
Whether one is random access mode, and in this mode, host side can random access buffer cell, by management buffer cell flag notification logical circuit, have data to need mutual by the state recognition of logical circuit judgement symbol.Shortcoming is complex management, None-identified data submission order, and for the complex management indicated, advantage is that main frame uses dirigibility high.
Another is the pattern based on queue, in this mode, host side and logical circuit are by safeguarding that queue pointer's mode realizes queue management, it adopts mechanism for FIFO (first in first out), host side and logical circuit must be merely able to sequential storage data, advantage is that structure is simple, and be easy to realize, shortcoming is that supported application model is single.
Above two kinds of modes, no matter which kind of, the equal Shortcomings when meeting following application model:
Host side Random Access Data, but require that hardware processes according to the order submitted to, and after processing is completed, corresponding buffer zone can be discharged according to virtual condition data.
Summary of the invention:
The object of this invention is to provide one can memory circuit at random, to meet above-mentioned application model, can either ensure that hardware circuit is ensureing the situation dispatching data of sequencing according to queue mode, also can meet the requirement that data random access and buffer zone discharge at random simultaneously.
Technical scheme of the present invention is as follows:
Based on queue management can a memory circuit at random, its special character is: comprise
N buffer cell, and set up data interaction between main frame;
The degree of depth is the FIFO memory of n, and the buffering corresponding to each buffer cell that host side sends is numbered by this FIFO memory sequence delivery to scheduler;
Width is the buffer cell Status Flag register of n bits, and each bit wherein, for describing the state of corresponding buffer cell, is responsible for set and clearing respectively by described host side and scheduler; And
Described scheduler, in order to read the content of described FIFO memory and buffer cell and to judge bit clearing specific in buffer cell Status Flag register.
Above-mentioned can the access implementation method of memory circuit at random, comprise the following steps:
1) host side distributing buffer unit, and in buffer cell Status Flag register, fill in buffering service marking according to arranging to the use of buffer cell, by FIFO memory transmission buffering numbering;
2) dispatch scheduler inquiry FIFO memory state, read buffering numbering, carry out digital independent and process;
3), after scheduler completes process, reset buffering service marking, buffer release unit, completes a data interaction.
The present invention by increase flag register, as the mark of buffering using state, and in conjunction with the use of FIFO, both can adapt to the requirement of traditional page management approach, also can meet the requirement of system random access concurrent processing.Specifically there is following advantage:
Be simple and easy to use: effectively can simplify amortization management method when fpga logic designs, save logical resource, host software can realize random access capabilities in a straightforward manner simultaneously;
Versatility is good: can meet queue application mode, also can meet random administration mode, and only need revise software can adapt to, and the versatility of hardware improves;
High efficient and flexible: the mutual dispatch deal of high speed that can realize data with less software and hardware expense, can meet the data coupling requirement of high-speed data communication.
Accompanying drawing explanation
Fig. 1 is circuit diagram of the present invention.
Embodiment
Embodiments of the invention as shown in Figure 1, under the buffering interactive mode of tradition based on FIFO, increase a queue flag register, for realizing the state description of buffer cell, the guarantee of data submission order is realized by fifo circuit, software can use buffering according to actual needs dynamically, and realizes the sequence management mechanism of transmission by operation FIFO simultaneously.Both, under the information interaction pattern of traditional random access and queue access, effectively combine, meet one end random access by the present invention, and the way to manage of one end employings degree row scheduling, the system that effectively meets is applied.
1. logic circuit unit:
1) degree of depth be the fifo circuit of n 1., mutual for the queuing realizing available buffer;
2) width be the buffer status register of n bits 2., each bit is for describing the state of corresponding buffer cell;
3) pool of buffer unit 3., user interactive data, can be the descriptor of buffer cell, also can be physically actual storage block; Following treatment scheme is described for descriptor.
4) scheduler 4., realizes FIFO content and read scheduling and buffering Single Component Management.
2. treatment scheme:
A. after powering on or resetting, be 1. dummy status, 2., the value of all positions is " 0 ", and represent that corresponding buffering is idle, descriptor is 3. invalid, represents that buffering does not configure, 4. detects empty full scale will 1..
B. host software performs buffering batch operation, upgrades descriptor 3., for identifying each address cushioned;
C. host software selects a buffering preparing to use, and writes data, after write, the respective flag of buffering in is 2. set to " 1 ", and the sequence number of buffering is write 1.;
D. 4. detect 1. for after " non-NULL " state, from FIFO, read queue sequence number, and according to the serial number read, from correspondence buffering, extract data process;
E. 4. after data processing completes, according to buffer cell serial number, the value of corresponding bit in is 2. set to " 0 " by scheduler again, discharges this buffering, completes an interactive operation of this element data.
The present invention can realize the FIFO management of data interaction, and supports phase of depositing the at random processing capacity of main frame and logic, effectively can simplify Logic Circuit Design simultaneously.

Claims (2)

1. based on queue management can a memory circuit at random, it is characterized in that: comprise
N buffer cell, and set up data interaction between main frame;
The degree of depth is the FIFO memory of n, and the buffering corresponding to each buffer cell that host side sends is numbered by this FIFO memory sequence delivery to scheduler;
Width is the buffer cell Status Flag register of n bits, and each bit wherein, for describing the state of corresponding buffer cell, is responsible for set and clearing respectively by described host side and scheduler; And
Described scheduler, in order to read the content of described FIFO memory and buffer cell and to judge bit clearing specific in buffer cell Status Flag register.
2. can the access implementation method of memory circuit at random described in claim 1, comprise the following steps:
1) host side distributing buffer unit, and in buffer cell Status Flag register, fill in buffering service marking according to arranging to the use of buffer cell, by FIFO memory transmission buffering numbering;
2) dispatch scheduler inquiry FIFO memory state, read buffering numbering, carry out digital independent and process;
3), after scheduler completes process, reset buffering service marking, buffer release unit, completes a data interaction.
CN201410749390.2A 2014-12-05 2014-12-05 Random memory circuit based on queue management Pending CN104503928A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410749390.2A CN104503928A (en) 2014-12-05 2014-12-05 Random memory circuit based on queue management

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410749390.2A CN104503928A (en) 2014-12-05 2014-12-05 Random memory circuit based on queue management

Publications (1)

Publication Number Publication Date
CN104503928A true CN104503928A (en) 2015-04-08

Family

ID=52945327

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410749390.2A Pending CN104503928A (en) 2014-12-05 2014-12-05 Random memory circuit based on queue management

Country Status (1)

Country Link
CN (1) CN104503928A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113961148A (en) * 2021-10-28 2022-01-21 苏州浪潮智能科技有限公司 Data interaction method, system and device for host and storage system
CN114936171A (en) * 2022-06-14 2022-08-23 深存科技(无锡)有限公司 Memory access controller architecture

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1904854A (en) * 2005-07-28 2007-01-31 光宝科技股份有限公司 Stream data buffer unit and access method thereof
US20070255874A1 (en) * 2006-04-28 2007-11-01 Jennings Kevin F System and method for target device access arbitration using queuing devices
CN101071404A (en) * 2006-05-09 2007-11-14 大唐移动通信设备有限公司 Small-capacity FIFO storage device data-moving trigger and method
CN101567849A (en) * 2009-04-30 2009-10-28 炬才微电子(深圳)有限公司 Data buffer caching method and device

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1904854A (en) * 2005-07-28 2007-01-31 光宝科技股份有限公司 Stream data buffer unit and access method thereof
US20070255874A1 (en) * 2006-04-28 2007-11-01 Jennings Kevin F System and method for target device access arbitration using queuing devices
CN101071404A (en) * 2006-05-09 2007-11-14 大唐移动通信设备有限公司 Small-capacity FIFO storage device data-moving trigger and method
CN101567849A (en) * 2009-04-30 2009-10-28 炬才微电子(深圳)有限公司 Data buffer caching method and device

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113961148A (en) * 2021-10-28 2022-01-21 苏州浪潮智能科技有限公司 Data interaction method, system and device for host and storage system
CN113961148B (en) * 2021-10-28 2023-08-08 苏州浪潮智能科技有限公司 Data interaction method, system and device for host and storage system
CN114936171A (en) * 2022-06-14 2022-08-23 深存科技(无锡)有限公司 Memory access controller architecture
CN114936171B (en) * 2022-06-14 2023-11-14 深存科技(无锡)有限公司 Storage access controller architecture

Similar Documents

Publication Publication Date Title
US9792235B2 (en) Optimized credit return mechanism for packet sends
CN106415515B (en) Grouping is sent using the PIO of the optimization without SFENCE write-in sequence
US7496699B2 (en) DMA descriptor queue read and cache write pointer arrangement
CN103052949B (en) The mechanism of process peripherals page fault
CN100345130C (en) Multi-channel interface for communications between devices
CN100527112C (en) Controller apparatus and method for improved data transfer
KR102427550B1 (en) QoS-AWARE IO MANAGEMENT FOR PCIe STORAGE SYSTEM WITH RECONFIGURABLE MULTI-PORTS
CN110083461B (en) Multitasking system and method based on FPGA
CN103647807A (en) Information caching method, device and communication apparatus
CN102724035B (en) Encryption and decryption method for encrypt card
CN103763173A (en) Data transmission method and computing node
CN103562895B (en) It is easy to route by selective polymerisation adjacent data cell
CN102541779A (en) System and method for improving direct memory access (DMA) efficiency of multi-data buffer
CN105740178B (en) Chip network system with and forming method thereof
CN105260332A (en) Method and system for orderly storing CPLD data packets
CN100476775C (en) Host computer controller used for bus communication equipment and bus communication device
CN101877666A (en) Method and device for receiving multi-application program message based on zero copy mode
CN103577469B (en) Database connection multiplexing method and apparatus
US8032675B2 (en) Dynamic memory buffer allocation method and system
CN104503928A (en) Random memory circuit based on queue management
US9288163B2 (en) Low-latency packet receive method for networking devices
CN112181887A (en) Data transmission method and device
CN107066413A (en) A kind of method and its bus system for being used to handle multiple bus apparatus data
US20070028015A1 (en) System and method for processing data streams
CN101577712A (en) Service front-end processor supporting remote access of multiple terminals and network interface card thereof

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20150408

RJ01 Rejection of invention patent application after publication