JPH09259030A - Multi-cpu device - Google Patents

Multi-cpu device

Info

Publication number
JPH09259030A
JPH09259030A JP7230796A JP7230796A JPH09259030A JP H09259030 A JPH09259030 A JP H09259030A JP 7230796 A JP7230796 A JP 7230796A JP 7230796 A JP7230796 A JP 7230796A JP H09259030 A JPH09259030 A JP H09259030A
Authority
JP
Japan
Prior art keywords
signal
cpu
state
control
circuit
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
JP7230796A
Other languages
Japanese (ja)
Inventor
Yasunari Kanzawa
泰成 神澤
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.)
Oki Electric Industry Co Ltd
Original Assignee
Oki Electric Industry Co Ltd
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 Oki Electric Industry Co Ltd filed Critical Oki Electric Industry Co Ltd
Priority to JP7230796A priority Critical patent/JPH09259030A/en
Publication of JPH09259030A publication Critical patent/JPH09259030A/en
Pending legal-status Critical Current

Links

Landscapes

  • Multi Processors (AREA)

Abstract

PROBLEM TO BE SOLVED: To provide a multi-CPU device where plural CPU can easily access the storage circuit (S-RAM circuit and D-RAM circuit, for example) of a single port through a bus line of one system with a simple constitution. SOLUTION: A control preferential selection circuit A monitors a chip selection signal outputted from CPU 1 and CPU 2, decides control priority right, outputs a master 1-n signal or a master 2-n signal, gives the master 1-n signal to a generation circuit B, a generation circuit C and a generation circuit D and gives the master 2-n signal to a generation circuit E and a generation circuit G. A CPU 1 READY response timing generation part B decides timing for terminating the access cycle of CPU 1, generates a READY signal (b) and gives it to CPU 1.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【発明の属する技術分野】本発明は、マルチCPU装置
に関し、例えば、複数のCPUが一つのS(Stati
c)−RAM回路にアクセスする装置に適用し得るもの
である。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-CPU device, for example, a plurality of CPUs having one S (Static).
c) -Applicable to devices that access RAM circuits.

【0002】[0002]

【従来の技術】近年、メモリ素子の開発が盛んに行わ
れ、S−RAM、D(Dynamic)−RAMや、D
PM(Dual Port MEMORY)などが普及
している。
2. Description of the Related Art In recent years, memory devices have been actively developed, and S-RAM, D (Dynamic) -RAM and D-RAM have been developed.
PM (Dual Port MEMORY) and the like are widely used.

【0003】例えば、S−RAMは、書き込み、読み出
しが1ポートで、大容量化することが比較的に容易で、
安価である。一方、DPMは書き込み、読み出しが2ポ
ートであり、2種類のマイクロプロセッサ(MPU)マ
スタが同時にアクセス可能であり、小容量で高価であ
る。
For example, the S-RAM has one port for writing and reading, and it is relatively easy to increase the capacity.
It is cheap. On the other hand, the DPM has two ports for writing and reading, can be simultaneously accessed by two types of microprocessor (MPU) masters, and has a small capacity and is expensive.

【0004】[0004]

【発明が解決しようとする課題】上述のようにデュアル
ポートメモリは、2種類のマイクロプロセッサのマスタ
によるアクセスが可能であるが、小容量であり、高価で
ある。従って、デュアルポートメモリで大きな情報を扱
う場合は、デュアルポートメモリを複数個用いることで
実現できるが、かなり高価になってしまうという問題が
ある。
As described above, the dual port memory can be accessed by the masters of two types of microprocessors, but it has a small capacity and is expensive. Therefore, when a large amount of information is handled by the dual port memory, it can be realized by using a plurality of dual port memories, but there is a problem that it becomes considerably expensive.

【0005】このため、大容量で安価なS−RAMを用
いて、複数のCPU(又はマイクロプロセッサ)からす
ると、あたかもデュアルポートメモリにアクセスしてい
るかのようにアクセスすることができるマルチCPU装
置が望まれていた。
Therefore, a large-capacity and inexpensive S-RAM is used to provide a multi-CPU device which can be accessed from a plurality of CPUs (or microprocessors) as if they were accessing a dual-port memory. Was wanted.

【0006】このようなことから、シングルポートの記
憶回路(例えば、S−RAM回路や、D−RAM回路な
ど)に対して一系統のバスラインを通じて複数のCPU
からアクセスすることを簡単な構成で容易に行い得るマ
ルチCPU装置の実現が要請されている。
For this reason, a plurality of CPUs are connected to a single-port memory circuit (for example, an S-RAM circuit or a D-RAM circuit) through a bus line of one system.
There is a demand for realization of a multi-CPU device that can be easily accessed from a simple configuration.

【0007】[0007]

【課題を解決するための手段】そこで、本発明は、複数
のCPUが一系統のバスラインを通じてシングルポート
の記憶回路にアクセスするマルチCPU装置において、
以下の特徴的な構成で上述の課題を解決する。
Therefore, the present invention provides a multi-CPU device in which a plurality of CPUs access a single-port memory circuit through a bus line of one system.
The above-mentioned problems are solved by the following characteristic configurations.

【0008】即ち、本発明は、(1)上記各CPUから
出力されるチップセレクト信号のアクティブ/非アクテ
ィブの状態を監視し、同じ時期に2以上のCPUからの
チップセレクト信号が競合した場合に、これらの複数の
チップセレクト信号に対する制御要請に対して代わるが
わる制御権を与えて制御権信号を出力し、この制御権を
与えられない期間は制御要請に対して制御待ちさせる
『調停手段』と、(2)制御権信号に対応するCPUへ
の制御許可を与えるための制御許可応答信号を生成して
CPUへ与え、制御待ちさせているチップセレクト信号
を出力しているCPUに対して制御待ち応答信号を与え
る『制御許可/許可待ち応答信号生成手段』と、(3)
制御許可を与えられたCPUからのアドレスをバスライ
ンを通じて記憶回路に与えると共に、制御権信号に基づ
き記憶回路のアクセスタイムに対応するように記憶回路
用のチップセレクト信号を生成して記憶回路に与える
『記憶回路用チップセレクト信号生成手段』とを備え
る。
That is, the present invention (1) monitors the active / inactive state of the chip select signals output from the CPUs, and when the chip select signals from two or more CPUs compete at the same time, , "Arbitration means" which gives a control right instead of the control request for these plurality of chip select signals and outputs a control right signal, and waits for the control request during the period when the control right is not given. And (2) generate a control permission response signal for giving control permission to the CPU corresponding to the control right signal, give it to the CPU, and control the CPU outputting the chip select signal waiting for control. “Control permission / permission waiting response signal generating means” for giving a waiting response signal, and (3)
An address from the CPU to which control permission is given is given to the memory circuit through the bus line, and a chip select signal for the memory circuit is generated based on the control right signal so as to correspond to the access time of the memory circuit and given to the memory circuit. "Memory circuit chip select signal generating means".

【0009】このような構成で(1)の『調停手段』
が、複数のCPUからの同じ時期に発生したチップセレ
クト信号によって制御要請がされた場合に、制御要請し
ているCPUに対して、代わるがわる制御権を与えるこ
とで、複数のCPUに対する制御権を平等に与えて、複
数の処理を同時に行うことが容易になる。
With such a structure, the "arbitration means" of (1)
However, when a control request is issued by a chip select signal generated from a plurality of CPUs at the same time, a control right for the plurality of CPUs is given instead by giving an alternative control right to the CPU requesting the control. Can be given evenly and a plurality of processes can be easily performed at the same time.

【0010】そして、(2)の『制御許可/許可待ち応
答信号生成手段』が、各CPUに対する代わるがわるの
制御権信号によって、制御許可応答信号を生成して各C
PUに与える。この制御許可応答信号は、例えば、レデ
ィ(Ready)信号又はイネーブル(Enable)
信号などが好ましい。一方、制御要請している各CPU
に代わるがわる制御許可応答信号が与えられているとき
に、制御許可が与えられていない期間には、制御許可待
ち応答信号を与えることでCPUは制御待ちを行う。
Then, the "control permission / permission waiting response signal generating means" of (2) generates a control permission response signal in response to the control authority signal for each CPU instead of each CPU,
Give to PU. The control permission response signal is, for example, a ready signal or an enable signal.
Signals etc. are preferred. On the other hand, each CPU requesting control
In place of the control permission response signal, the CPU waits for control by giving the control permission wait response signal during the period when the control permission is not given.

【0011】更に、(3)の『記憶回路用チップセレク
ト信号生成手段』によって、制御許可応答信号を受けた
CPUがアドレスとチップセレクト信号とを生成してバ
スラインを通じて記憶回路に与えることで、複数のCP
Uからのアドレスやチップセレクト信号が衝突すること
なく記憶回路に対して整然とアクセスすることができ
る。
Furthermore, by (3) "memory circuit chip select signal generating means", the CPU which receives the control permission response signal generates an address and a chip select signal and gives them to the memory circuit through the bus line. Multiple CPs
The memory circuit can be accessed in an orderly manner without collision of the address and the chip select signal from U.

【0012】[0012]

【発明の実施の形態】次に本発明の好適な実施の形態を
図面を用いて説明する。そこで、本実施の形態では、複
数のマイクロプロセッサであるマスタからのチップセレ
クト信号を監視して、マスタに対して制御権付与を行う
制御優先選択部Aと、READY応答タイミング生成部
B、Eと、制御権が与えられたマスタのアドレスをS−
RAMに対して有意とするアドレスバッファ制御タイミ
ング生成部C、Fと、S−RAMのアクセスタイムを考
慮したチップセレクト信号を生成するチップセレクト入
力タイミング生成部D、Gを備えるバス調停装置を構成
する。
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, a preferred embodiment of the present invention will be described with reference to the drawings. Therefore, in this embodiment, a control priority selection unit A that monitors chip select signals from a master, which is a plurality of microprocessors, and gives control right to the master, and READY response timing generation units B and E. , S- is the address of the master to which control is given.
A bus arbitration device including address buffer control timing generation units C and F that are significant for RAM and chip select input timing generation units D and G that generate a chip select signal considering the access time of S-RAM is configured. .

【0013】図1は、マルチCPU装置の機能構成図で
ある。この図1において、マルチCPU装置は、主にC
PU1と、CPU2と、バス調停回路10と、CPU1
アドレスバッファ回路Hと、AND回路9と、CPU2
アドレスバッファ回路Iと、S−RAM回路8と、CP
U1データバッファ回路Jと、CPU2データバッファ
回路Kとから構成されている。
FIG. 1 is a functional block diagram of a multi-CPU device. In FIG. 1, the multi-CPU device is mainly C
PU1, CPU2, bus arbitration circuit 10, CPU1
Address buffer circuit H, AND circuit 9, CPU 2
Address buffer circuit I, S-RAM circuit 8, CP
It is composed of a U1 data buffer circuit J and a CPU2 data buffer circuit K.

【0014】バス調停回路10は、制御優先選択回路A
と、CPU1READY応答タイミング生成部Bと、C
PU1アドレスバッファ制御タイミング生成回路Cと、
CPU1チップセレクト入力タイミング生成回路Dと、
CPU2READY応答タイミング生成回路Eと、CP
U2アドレスバッファ制御タイミング生成回路Fと、C
PU2チップセレクト入力タイミング生成回路Gとから
構成されている。
The bus arbitration circuit 10 includes a control priority selection circuit A.
And CPU1READY response timing generator B and C
PU1 address buffer control timing generation circuit C,
A CPU1 chip select input timing generation circuit D,
CPU2READY response timing generation circuit E and CP
U2 address buffer control timing generation circuit F and C
It comprises a PU2 chip select input timing generation circuit G.

【0015】この図1の構成においては、主に、2種類
のバスマスタである、CPU1、CPU2が備えられ、
それぞれが共通メモリであるS−RAM回路8に対して
アクセスするとき、それぞれのCPUの優先権を選択す
る手段と、優先権を獲得したときのアドレスバスドライ
バ制御タイミング手段と、READY応答のタイミング
生成手段とを備えて、バス調停を行うように構成してい
る。
The configuration of FIG. 1 is mainly provided with two types of bus masters, CPU1 and CPU2,
When accessing the S-RAM circuit 8, which is a common memory, means for selecting the priority of each CPU, address bus driver control timing means when the priority is acquired, and timing generation of the READY response And means for performing bus arbitration.

【0016】CPU1アドレスバッファ回路Hと、CP
U2アドレスバッファ回路Iと、S−RAM回路8との
間には、アドレスバスが形成されている。また、CPU
1データバッファ回路Jと、CPU2データバッファ回
路Kと、S−RAM回路8との間にも、データバスが形
成されている。これらのバスに対するCPU1、CPU
2からのアクセスの調停をバス調停回路10によって調
停を行っている。
CPU1 address buffer circuit H and CP
An address bus is formed between the U2 address buffer circuit I and the S-RAM circuit 8. Also, CPU
A data bus is also formed between the 1-data buffer circuit J, the CPU 2 data buffer circuit K, and the S-RAM circuit 8. CPU1 and CPU for these buses
The bus arbitration circuit 10 arbitrates the access from the access point 2.

【0017】CPU1は、内部にマイクロプロセッサと
して、例えば、NEC製MPUでV.53を使用し、こ
のMPUはシステムクロックを10MHzとし、WAI
Tなし1サイクルを2システムクロックとし、アドレス
をCPU1アドレスバッファ回路Hに与え、データをC
PU1データバッファ回路Jに与え、チップセレクト信
号master1cs−nを制御優先選択回路Aに与
え、CPU1READY応答タイミング生成回路Bから
READY信号を受けるものである。
The CPU 1 is an internal microprocessor, for example, an NEC MPU V.I. 53, this MPU sets the system clock to 10MHz, WAI
One cycle without T is set as two system clocks, an address is given to the CPU1 address buffer circuit H, and data is C
It is supplied to the PU1 data buffer circuit J, the chip select signal master1cs-n is supplied to the control priority selection circuit A, and the READY signal is received from the CPU1READY response timing generation circuit B.

【0018】CPU2は、内部にマイクロプロセッサと
して、例えば、Intel社製MPUで80186XL
を使用し、このMPUはシステムクロックとして16M
Hzを使用し、WAITなし1サイクルを4システムク
ロックとし、アドレスをCPU2アドレスバッファ回路
Iに与え、データをCPU2データバッファ回路Kに与
え、チップセレクト信号master2cs−nを制御
優先選択回路Aに与え、CPU2READY応答タイミ
ング生成回路EからREADY信号を受ける。
The CPU 2 is an internal microprocessor, for example, an Intel MPU 80186XL.
This MPU uses 16M as the system clock.
1 Hz is used as 4 system clocks, an address is supplied to the CPU2 address buffer circuit I, data is supplied to the CPU2 data buffer circuit K, and a chip select signal master2cs-n is supplied to the control priority selection circuit A. The READY signal is received from the CPU2READY response timing generation circuit E.

【0019】制御優先選択回路Aは、CPU1、CPU
2から出力されるチップセレクト信号master1c
s−n、master2cs−nを監視し、制御優先権
を決定し、master1−n信号又はmaster2
−n信号を出力し、master1−n信号をCPU1
READY応答タイミング生成部Bと、CPU1アドレ
スバッファ制御タイミング生成回路Cと、CPU1チッ
プセレクト入力タイミング生成回路Dとに与え、mas
ter2−n信号をCPU2READY応答タイミング
生成回路Eと、CPU2アドレスバッファ制御タイミン
グ生成回路Fと、CPU2チップセレクト入力タイミン
グ生成回路Gとに与える。
The control priority selection circuit A includes a CPU 1, a CPU
2 outputs the chip select signal master1c
s-n, master2cs-n are monitored, control priority is determined, master1-n signal or master2
-N signal is output, and master1-n signal is output to CPU1.
It is given to the READY response timing generation unit B, the CPU1 address buffer control timing generation circuit C, and the CPU1 chip select input timing generation circuit D.
The ter2-n signal is given to the CPU2READY response timing generation circuit E, the CPU2 address buffer control timing generation circuit F, and the CPU2 chip select input timing generation circuit G.

【0020】CPU1READY応答タイミング生成回
路Bは、CPU1のアクセスタイムを終了させるための
タイミングを決定して、READY信号bを生成してC
PU1へ与える。
The CPU1 READY response timing generation circuit B determines the timing for ending the access time of the CPU 1, generates the READY signal b, and outputs the C signal.
Give to PU1.

【0021】CPU1アドレスバッファ制御タイミング
生成回路Cは、S−RAM回路8に入力するアドレス、
その他の必要な信号をイネーブルするタイミングを決定
し、タイミング信号cをCPU1アドレスバッファ回路
HとCPU1データバッファ回路Jとに与える。
The CPU 1 address buffer control timing generation circuit C has an address input to the S-RAM circuit 8,
The timing for enabling other necessary signals is determined, and the timing signal c is given to the CPU1 address buffer circuit H and the CPU1 data buffer circuit J.

【0022】CPU1チップセレクト入力タイミング生
成回路Dは、S−RAM回路8に対するアクセスタイミ
ングを考慮してチップセレクト入力のタイミングを決定
してタイミング信号dをAND回路9に与える。
The CPU 1 chip select input timing generation circuit D determines the chip select input timing in consideration of the access timing to the S-RAM circuit 8 and supplies the timing signal d to the AND circuit 9.

【0023】CPU2READY応答タイミング生成回
路Eは、CPU2のアクセスサイクロを終了させるため
のタイミングを決定してREADY信号eをCPU2へ
与える。
The CPU2 READY response timing generation circuit E determines a timing for ending the access cyclo of the CPU 2 and gives a READY signal e to the CPU 2.

【0024】CPU2アドレスバッファ制御タイミング
生成回路Fは、S−RAM回路8に入力するアドレス、
その他の必要な信号をイネーブルするためのタイミング
を決定してタインミング信号fをCPU2アドレスバッ
ファ回路Iと、CPU2データバッファ回路Kとに与え
る。
The CPU 2 address buffer control timing generation circuit F has an address input to the S-RAM circuit 8,
The timing for enabling other necessary signals is determined and the timing signal f is given to the CPU2 address buffer circuit I and the CPU2 data buffer circuit K.

【0025】CPU2チップセレクト入力タイミング生
成回路Gは、S−RAM8回路のアクセスタイムを考慮
して、チップセレクト入力のタイミングを決定してタイ
ミング信号gをAND回路9に与える。
The CPU2 chip select input timing generation circuit G determines the chip select input timing in consideration of the access time of the S-RAM8 circuit and supplies the timing signal g to the AND circuit 9.

【0026】CPU1アドレスバッファ回路Hは、CP
U1アドレスバッファ制御タイミング生成回路Cによっ
て生成されたタイミング信号cによって制御され、共通
的なS−RAM回路8に対してアドレスバスを通じてア
ドレスを与える。
The CPU1 address buffer circuit H has a CP
It is controlled by the timing signal c generated by the U1 address buffer control timing generation circuit C and gives an address to the common S-RAM circuit 8 through the address bus.

【0027】CPU2アドレスバッファ回路Iは、CP
U2アドレスバッファ制御タイミング生成回路Fによっ
て生成されたタイミング信号fによって制御され、共通
的なS−RAM回路8に対してアドレスバスを通じてア
ドレスを与える。
The CPU2 address buffer circuit I has a CP
It is controlled by the timing signal f generated by the U2 address buffer control timing generation circuit F and gives an address to the common S-RAM circuit 8 through the address bus.

【0028】AND回路9は、CPU1チップセレクト
入力タイミング生成回路Dからのタイミング信号dと、
CPU2チップセレクト入力タイミング生成回路Gから
のタイミング信号gとの論理積を行って、この論理積信
号ramcs−nをS−RAM回路8に対するチップセ
レクト信号として与える。
The AND circuit 9 receives the timing signal d from the CPU1 chip select input timing generation circuit D and
A logical product is performed with the timing signal g from the CPU2 chip select input timing generation circuit G, and this logical product signal ramcs-n is given as a chip select signal to the S-RAM circuit 8.

【0029】S−RAM回路8は、1又は2以上のS−
RAMICによって構成され、アドレスバスによってC
PU1アドレスバッファ回路H又はCPU2アドレスバ
ッファ回路Iからアドレスを受けると共に、チップセレ
クト信号をAND回路9から論理積信号ramcs−n
として受け、データバスを通じてCPU1データバッフ
ァ回路J又はCPU2データバッファ回路Kからのデー
タを受けて記憶したり、データを読み出したりする。
The S-RAM circuit 8 includes one or more S-RAM circuits.
It is composed of RAMIC and C by address bus
The address is received from the PU1 address buffer circuit H or the CPU2 address buffer circuit I, and the chip select signal is sent from the AND circuit 9 as the logical product signal ramcs-n.
And receives and stores the data from the CPU1 data buffer circuit J or the CPU2 data buffer circuit K through the data bus and reads the data.

【0030】CPU1データバッファ回路Jと、CPU
1アドレスバッファ制御タイミング生成回路Cからのタ
イミング信号cによって制御され、データバスの制御を
行うものであって、CPU1からのデータをS−RAM
回路8に与え、また、S−RAM回路8からのデータを
CPU1に与える。
CPU1 data buffer circuit J and CPU
The data bus is controlled by the timing signal c from the 1 address buffer control timing generation circuit C, and the data from the CPU 1 is stored in the S-RAM.
The data is supplied to the circuit 8 and the data from the S-RAM circuit 8 is supplied to the CPU 1.

【0031】CPU2データバッファ回路Kは、CPU
2アドレスバッファ制御タイミング生成回路Fからのタ
イミング信号fによって制御され、データバスの制御を
行うものであって、CPU2からのデータをS−RAM
回路8に与え、また、S−RAM回路8からのデータを
CPU2に与える。
The CPU2 data buffer circuit K is a CPU
Controlled by the timing signal f from the 2-address buffer control timing generation circuit F, the data bus is controlled, and the data from the CPU 2 is stored in the S-RAM.
The data is supplied to the circuit 8 and the data from the S-RAM circuit 8 is supplied to the CPU 2.

【0032】図3は、制御優先選択回路Aの一例の詳細
な回路構成である。この図3において、制御優先選択回
路Aは、論理回路R1と、レジスタ回路REG1〜RE
G4と、デコーダ回路A1とから構成されている。
FIG. 3 is a detailed circuit configuration of an example of the control priority selection circuit A. In FIG. 3, the control priority selection circuit A includes a logic circuit R1 and register circuits REG1 to RE.
It is composed of G4 and a decoder circuit A1.

【0033】論理回路R1には、CPU1からチップセ
レクト信号master1cs−nを受け、更に、CP
U2からチップセレクト信号master2cs−nを
受け、制御優先権の調停を行って、レジスタREG1〜
REG4に対して、10MHzのクロックを用いて制御
データを設定する。デコーダ回路A1は、レジスタRE
G1〜REG4の制御データをデコードしてデコード信
号master1−n又はデコード信号master2
−nを生成して出力する。
The logic circuit R1 receives the chip select signal master1cs-n from the CPU1 and further receives CP.
It receives the chip select signal master2cs-n from U2, arbitrates the control priority, and registers REG1 to REG1.
Control data is set for REG4 using a clock of 10 MHz. The decoder circuit A1 includes a register RE
The control data of G1 to REG4 is decoded to decode the decode signal master1-n or the decode signal master2.
-N is generated and output.

【0034】上述のバス調停回路10の各構成回路は、
与えられる任意のクロックによって、制御権信号である
デコード信号master1−n及びデコード信号ma
ster2−nに対応して同期制御される。即ち、CP
U1、2からのチップセレクト信号master1cs
−n、master2cs−nが競合した場合に、これ
らのチップセレクト信号に対する制御要請に対して代わ
るがわる制御権を与えて制御権信号master1−
n、master2−nを出力し、制御権を与えられな
い期間は制御要請に対して制御待ちさせる。
Each constituent circuit of the bus arbitration circuit 10 described above is
The decode signals master1-n and the decode signal ma, which are control signals, are controlled by an arbitrary clock applied.
Synchronous control is performed corresponding to star2-n. That is, CP
Chip select signal master1cs from U1 and U2
-N and master2cs-n conflict with each other, a control right is given instead of a control request for these chip select signals to give a control right signal master1-.
n and master2-n are output, and the control request is waited for during the period when the control right is not given.

【0035】(動作): 次に、図1のマルチCPU
装置の動作を説明する。先ず、2種類のバスマスタであ
るCPU1、2から出力される、共通のS−RAM回路
8に対するチップセレクト信号master1cs−
n、master2cs−nを制御優先選択回路Aが監
視する。2種類のバスマスタであCPU1、CPU2が
同時にチップセレクト信号master1cs−n、m
aster2cs−nを出力した場合は、CPU1に制
御権を与え、同時でない場合には、出力してきたCPU
に制御権を与え、両方のチップセレクト信号が、有意状
態に固定されるときには、2つのCPU1、2に対して
交互に制御権を与える。
(Operation): Next, the multi CPU of FIG.
The operation of the device will be described. First, the chip select signal master1cs- for the common S-RAM circuit 8 output from the CPUs 1 and 2 which are two types of bus masters.
The control priority selection circuit A monitors n and master2cs-n. The two types of bus masters, CPU1 and CPU2, simultaneously perform chip select signals master1cs-n, m.
When the aster2cs-n is output, the control right is given to the CPU1, and when it is not simultaneous, the CPU that has output
To the two CPUs 1 and 2 when both chip select signals are fixed to the significant state.

【0036】この制御権付与の選択を制御優先選択回路
Aが行う。この制御優先選択回路Aの付与によってマス
タのCPUが決定されると、共通のS−RAM回路8の
アクセスタイム、CPUのアクセスタイム中のデータサ
ンプリング位置を考慮した時間分の、CPU1アドレス
バッファ制御タイミング生成回路Cのタイミング信号c
及びCPU1チップセレクト入力タイミング生成回路D
のタイミング信号dが生成され、又はCPU2アドレス
バッファ制御タイミング生成回路Fのタイミング信号及
びCPU2チップセレクト入力タイミング生成回路Gの
タイミング信号が生成され、マスタのCPUのアクセス
タイム終了時間にCPU1READY応答タイミング生
成回路BのREADY信号b又はCPU2READY応
答タイミング生成回路EのREADY信号eが生成され
る。
The control priority selection circuit A selects the control right grant. When the master CPU is determined by the provision of the control priority selection circuit A, the CPU 1 address buffer control timing corresponding to the access time of the common S-RAM circuit 8 and the data sampling position in the access time of the CPU is considered. Timing signal c of generation circuit C
And CPU1 chip select input timing generation circuit D
Or the timing signal of the CPU2 address buffer control timing generation circuit F and the timing signal of the CPU2 chip select input timing generation circuit G are generated, and the CPU1READY response timing generation circuit is generated at the access time end time of the master CPU. The READY signal b of B or the READY signal e of the CPU2READY response timing generation circuit E is generated.

【0037】このように、本実施の形態のマルチCPU
装置の構成においては、『制御待ちにあるCPU1又は
2に対しては、CPU1READY応答タイミング生成
回路BのREADY信号b又はCPU2READY応答
タイミング生成回路EのREADY信号eの応答を遅ら
せることで、データバス及びアドレスバスの競合を調停
するように構成している』ことが特徴である。
As described above, the multi-CPU of this embodiment
In the configuration of the device, “for the CPU 1 or 2 waiting for control, by delaying the response of the READY signal b of the CPU1READY response timing generation circuit B or the READY signal e of the CPU2READY response timing generation circuit E, the data bus and It is configured to arbitrate address bus contention. ”

【0038】図2は、制御優先選択回路Aの一例の動作
状態の詳細な遷移図である。この図2において、制御優
先選択回路Aの動作状態として、動作状態ST1〜ST
8が存在する。動作状態ST1は、アクセスなし状態で
あり、ここで、(0.0.0.0)はレジスタREG1
〜REG4の制御データの状態を表す。つまり、CPU
1又は2によって、リセットされると、このアクセスな
し状態(ST1)になる。この図2において、アクセス
なし状態(ST1)で、(1.1)と記述しているもの
は、(CPU2のチップセレクトの状態、CPU1のチ
ップセレクトの状態)を表すもので、(1.1)はCP
U1、2共にチップセレクトが出されていない状態を表
すものである。尚、(1.1)は負論理で表している。
FIG. 2 is a detailed transition diagram of an operation state of an example of the control priority selection circuit A. In FIG. 2, as the operating state of the control priority selection circuit A, operating states ST1 to ST
There are eight. The operating state ST1 is a state without access, where (0.0.0.0) is the register REG1.
~ Indicates the state of REG4 control data. That is, CPU
When reset by 1 or 2, this no access state (ST1) is set. In FIG. 2, in the no access state (ST1), what is described as (1.1) indicates (the state of the chip select of the CPU2 and the state of the chip select of the CPU1). ) Is CP
Both U1 and U2 represent a state in which the chip select is not issued. Incidentally, (1.1) is represented by negative logic.

【0039】動作状態ST2は、CPU1アクセス状態
を表すもので、レジスタREG1〜REG4の状態は
(1.0.0.0)であり、アクセスなし状態(ST
1)からチップセレクト信号が(1.0)又は(0.
0)になると、このCPU1アクセス状態(ST2)に
なる。また、このCPU1アクセス状態(ST2)から
チップセレクト信号が(1.1)となると、再びアクセ
スなし状態(ST1)に状態遷移する。また、このCP
U1アクセス状態(ST2)からチップセレクト信号が
(0.0)又は(0.1)になると、CPU1アクセス
・CPU2アクセス待ち状態(ST4)に状態遷移する
のである。このCPU1アクセス状態(ST2)におい
て、チップセレクト信号が(1.0)の場合は、この状
態を維持する。
The operating state ST2 represents the CPU1 access state, the states of the registers REG1 to REG4 are (1.0.0.0), and the no access state (ST
The chip select signal from (1) to (1.0) or (0.
0), the CPU 1 access state (ST2) is set. Further, when the chip select signal becomes (1.1) from the CPU1 access state (ST2), the state transitions again to the no access state (ST1). Also, this CP
When the chip select signal becomes (0.0) or (0.1) from the U1 access state (ST2), the state transitions to the CPU1 access / CPU2 access wait state (ST4). In this CPU1 access state (ST2), if the chip select signal is (1.0), this state is maintained.

【0040】動作状態ST3は、CPU2アクセス状態
を表すもので、レジスタREG1〜REG4の状態は、
(0.0.0.1)であり、アクセスなし状態(ST
1)からチップセレクト信号が(0.1)となると、こ
のCPU2アクセス状態(ST3)になる。また、この
CPU2アクセス状態(ST3)からチップセレクト信
号が(1.1)となると、再びアクセスなし状態(ST
1)に状態遷移する。また、このCPU2アクセス状態
(ST3)からチップセレクト信号が(1.0)又は
(0.0)となると、CPU2空アクセス状態(ST
8)へ状態遷移する。
The operating state ST3 represents the CPU2 access state, and the states of the registers REG1 to REG4 are:
(0.0.0.1), no access state (ST
When the chip select signal becomes (0.1) from 1), the CPU2 access state (ST3) is set. When the chip select signal becomes (1.1) from the CPU2 access state (ST3), the no access state (ST) is restored again.
State transition to 1). When the chip select signal becomes (1.0) or (0.0) from the CPU2 access state (ST3), the CPU2 empty access state (ST
State transition to 8).

【0041】動作状態ST4は、CPU1アクセス・C
PU2アクセス待ち状態を表すもので、レジスタREG
1〜REG4の状態は、(1.0.1.0)であり、C
PU1アクセス状態(ST2)からチップセレクト信号
が(0.1)になると、このCPU1アクセス・CPU
2アクセス待ち状態(ST4)になる。また、このCP
U1アクセス・CPU2アクセス待ち状態(ST4)か
らチップセレクト信号が(1.0)になると、再び、C
PU1アクセス状態(ST2)に状態遷移する。また、
このCPU1アクセス・CPU2アクセス待ち状態(S
T4)からチップセレクト信号が(1.1)になると、
再び、アクセスなし状態(ST1)に状態遷移する。ま
た、このCPU1アクセス・CPU2アクセス待ち状態
(ST4)からチップセレクト信号が(0.1)又は
(0.0)になると、CPU2アクセス待ち状態(ST
7)に状態遷移する。
The operating state ST4 is the CPU1 access C
It shows the state of waiting for PU2 access,
The states of 1 to REG4 are (1.0.1.0), and C
When the chip select signal becomes (0.1) from the PU1 access state (ST2), the CPU1 access / CPU
2 Access wait state (ST4) is entered. Also, this CP
When the chip select signal becomes (1.0) from the U1 access / CPU2 access wait state (ST4), C
The state transitions to the PU1 access state (ST2). Also,
This CPU1 access / CPU2 access waiting state (S
When the chip select signal becomes (1.1) from T4,
The state transits to the no access state (ST1) again. When the chip select signal becomes (0.1) or (0.0) from the CPU1 access / CPU2 access wait state (ST4), the CPU2 access wait state (ST
The state transitions to 7).

【0042】動作状態ST5は、CPU2アクセス・C
PU1アクセス待ち状態を表すもので、レジスタREG
1〜REG4の状態は、(0.1.0.1)であり、C
PU2空アクセスの状態(ST8)からチップセレクト
信号が(0.0)又は(1.0)となると、このCPU
2アクセス・CPU1アクセス待ち状態(ST5)にな
る。また、このCPU2アクセス・CPU1アクセス待
ち状態(ST5)からチップセレクト信号が(1.1)
になると、再びアクセスなし状態(ST1)に状態遷移
する。また、このCPU2アクセス・CPU1アクセス
待ち状態(ST5)からチップセレクト信号が(0.
1)になると、再びCPU2アクセス状態(ST3)に
状態遷移する。また、このCPU2アクセス・CPU1
アクセス待ち状態(ST5)からチップセレクト信号が
(1.0)又は(0.0)になると、CPU1アクセス
待ち状態(ST6)に状態遷移する。
The operating state ST5 is CPU2 access C
It indicates the PU1 access wait state and
The states of 1 to REG4 are (0.1.0.1), and C
When the chip select signal becomes (0.0) or (1.0) from the PU2 empty access state (ST8), this CPU
2 access / CPU 1 access wait state (ST5). In addition, the chip select signal is (1.1) from the CPU2 access / CPU1 access wait state (ST5).
Then, the state transits to the no access state (ST1) again. Further, the chip select signal (0.
When it becomes 1), the state transits to the CPU2 access state (ST3) again. Also, this CPU2 access / CPU1
When the chip select signal becomes (1.0) or (0.0) from the access wait state (ST5), the state transitions to the CPU1 access wait state (ST6).

【0043】動作状態ST6は、CPU1アクセス待ち
状態を表すものであり、レジスタREG1〜REG4の
状態は、(0.1.0.0)であり、CPU2アクセス
・CPU1アクセス待ち状態(ST5)からチップセレ
クト信号が(1.0)又は(0.0)となると、このC
PU1アクセス待ち状態(ST6)になる。このCPU
1アクセス待ち状態(ST6)からチップセレクト信号
が(1.0)又は(0.0)となると、CPU1アクセ
ス状態(ST2)に状態遷移する。また、このCPU1
アクセス待ち状態(ST6)からチップセレクト信号が
(1.1)に変化すると、アクセスなし状態(ST1)
に状態遷移する。また、このCPU1アクセス待ち状態
(ST6)からチップセレクト信号が(0.1)に変化
すると、CPU2アクセス状態(ST3)に状態遷移す
る。
The operating state ST6 represents the CPU1 access wait state, the states of the registers REG1 to REG4 are (0.1.0.0), and the CPU2 access / CPU1 access wait state (ST5) changes from the chip to the chip. When the select signal becomes (1.0) or (0.0), this C
The PU1 access waiting state (ST6) is entered. This CPU
When the chip select signal becomes (1.0) or (0.0) from the one access waiting state (ST6), the state transitions to the CPU1 access state (ST2). Also, this CPU 1
When the chip select signal changes from the access wait state (ST6) to (1.1), no access state (ST1)
The state transitions to. Further, when the chip select signal changes from this CPU1 access waiting state (ST6) to (0.1), the state transitions to the CPU2 access state (ST3).

【0044】動作状態ST7は、CPU2アクセス待ち
状態を表すものであり、レジスタREG1〜REG4の
状態は、(0.0.1.0)であり、CPU1アクセス
・CPU2アクセス待ち状態(ST4)からチップセレ
クト信号が(0.1)又は(0.0)に変化すると、こ
のCPU2アクセス待ち状態(ST7)に状態遷移す
る。また、このCPU2アクセス待ち状態(ST7)か
らチップセレクト信号が(1.0)に変化すると、CP
U1アクセス状態(ST2)に状態遷移する。また、こ
のCPU2アクセス待ち状態(ST7)からチップセレ
クト信号が(1.1)に変化すると、再びアクセスなし
状態(ST1)に状態遷移する。また、このCPU2ア
クセス待ち状態(ST7)からチップセレクト信号が
(0.1)又は(0.0)に変化すると、CPU2アク
セス状態(ST3)に状態遷移する。
The operating state ST7 represents the CPU2 access wait state, the states of the registers REG1 to REG4 are (0.0.1.0), and the CPU1 access / CPU2 access wait state (ST4) changes from the chip to the chip. When the select signal changes to (0.1) or (0.0), the state transitions to the CPU2 access waiting state (ST7). When the chip select signal changes from this CPU2 access wait state (ST7) to (1.0), CP
The state transitions to the U1 access state (ST2). When the chip select signal changes from this CPU2 access wait state (ST7) to (1.1), the state transitions again to the no access state (ST1). Further, when the chip select signal changes from (0.1) or (0.0) from the CPU2 access wait state (ST7), the state transitions to the CPU2 access state (ST3).

【0045】動作状態ST8は、CPU2空アクセス状
態を表すものであり、レジスタREG1〜REG4の状
態は、(0.1.1.1)であり、CPU2アクセス状
態(ST3)からチップセレクト信号が(0.0)又は
(1.0)に変化すると、このCPU2空アクセス状態
(ST8)に状態遷移する。また、このCPU2空アク
セス状態(ST8)からチップセレクト信号が(0.
0)又は(1.0)に変化すると、CPU2アクセス・
CPU1アクセス待ち状態(ST5)に状態遷移する。
また、このCPU2空アクセス状態(ST8)からチッ
プセレクト信号が(1.1)に変化すると、再び、アク
セスなし状態(ST1)に状態遷移する。
The operating state ST8 represents the CPU2 idle access state, the states of the registers REG1 to REG4 are (0.1.1.1), and the chip select signal (from the CPU2 access state (ST3) is ( When it is changed to 0.0) or (1.0), the state transitions to this CPU2 idle access state (ST8). Further, the chip select signal (0.
When it changes to 0) or (1.0), CPU2 access
The state transits to the CPU1 access waiting state (ST5).
When the chip select signal changes from this CPU2 idle access state (ST8) to (1.1), the state transitions again to the no access state (ST1).

【0046】レジスタREG1〜REG4の信号状態と
して、上述の図2の動作状態ST1〜ST8以外の状態
として、規定外状態として、(0.0.1.1)、
(0.1.1.0)、(1.0.0.1)、(1.0.
1.1)、(1.1.0.0)、(1.1.0.1)、
(1.1.1.0)、(1.1.1.1)が存在する。
As the signal states of the registers REG1 to REG4, the states other than the operation states ST1 to ST8 of FIG.
(0.1.1.0), (1.0.0.1), (1.0.
1.1), (1.1.0.0), (1.1.0.1),
There are (1.1.1.0) and (1.1.1.1).

【0047】このような構成によって、CPU1、CP
U2が制御を行う上で、上述の動作状態ST1〜ST8
の状態が存在し、CPU1だけによる制御動作の場合は
状態ST1と状態ST2との間で動作を行い、CPU2
だけによる制御動作の場合は状態ST1と状態ST3と
の間で動作を行う。
With such a configuration, the CPU 1 and CP
In controlling the U2, the above-mentioned operating states ST1 to ST8
If the control operation is performed only by the CPU1, the operation is performed between the state ST1 and the state ST2.
In the case of the control operation only by the operation, the operation is performed between the states ST1 and ST3.

【0048】また、図2の例において、アクセスタイム
の長いS−RAM回路を使用する場合は、アクセス待ち
状態を増やすか、状態遷移の際の同期クロックを低速に
すればよく、また、アクセスタイムの短いS−RAM回
路を使用する場合は、アクセス待ち状態を少なくする
か、状態遷移の際の同期クロックを高速にすればよい。
Further, in the example of FIG. 2, when the S-RAM circuit having a long access time is used, the number of access wait states may be increased or the synchronous clock at the time of state transition may be slowed down. When using a short S-RAM circuit, the access wait state may be reduced, or the synchronous clock at the time of state transition may be increased in speed.

【0049】図4は、CPU1、CPU2との両方のチ
ップセレクト信号が有意状態に固定され続けた場合の動
作タイミングチャートである。図5は、CPU1だけの
チップセレクトが有意状態に固定された場合の動作タイ
ミングチャートである。図6は、CPU2だけのチップ
セレクトが有意状態に固定された場合の動作タイミング
チャートである。
FIG. 4 is an operation timing chart when the chip select signals of both the CPU1 and the CPU2 continue to be fixed in the significant state. FIG. 5 is an operation timing chart when the chip select of only the CPU 1 is fixed to the significant state. FIG. 6 is an operation timing chart when the chip select of only the CPU 2 is fixed to the significant state.

【0050】これらの図4〜図6において、信号(a)
〜信号(zd)は共通的な信号である。(a)は、SY
NC−CLK2Nであり、CPU2側に対応したバス調
停回路10の10MHzの同期クロックの波形を表して
いる。(b)は、CSCRAM−N(ロウレベルのとき
にアクティブ)であり、CPU1が出力するチップセレ
クト信号master1cs−nの信号を表している。
(c)は、ALMCS−N(ロウレベルのときにアクテ
ィブ)であり、CPU2が出力するチップセレクト信号
master2cs−nの信号を表している。
In FIGS. 4 to 6, the signal (a)
~ Signal (zd) is a common signal. (A) is SY
NC-CLK2N, which represents the waveform of the 10 MHz synchronous clock of the bus arbitration circuit 10 corresponding to the CPU 2 side. (B) is CSCRAM-N (active when low level), and represents the signal of the chip select signal master1cs-n output by the CPU 1.
(C) is ALMCS-N (active when low level), and represents the signal of the chip select signal master2cs-n output by the CPU 2.

【0051】(d)は、REG1OUTであり、レジス
タREG1の出力信号の波形を表す。(e)は、REG
2OUTであり、レジスタREG2の出力信号の波形を
表す。(f)は、REG3OUTであり、レジスタRE
G3の出力信号の波形を表す。(g)は、REG4OU
Tであり、レジスタREG4の出力信号の波形を表す。
(h)は、D0000であり、状態(0000)、アク
セスなし状態(図2のST1)のときの波形を表す。
(i)は、ALM0001−N(ロウレベルのときにア
クティブ)であり、CPU2に制御権付与をしていると
きの信号波形を表す。
(D) is REG1OUT, which represents the waveform of the output signal of the register REG1. (E) is REG
2OUT, which represents the waveform of the output signal of the register REG2. (F) is REG3OUT, and register RE
The waveform of the output signal of G3 is shown. (G) is REG4OU
T, which represents the waveform of the output signal of the register REG4.
(H) is D0000, and represents a waveform in the state (0000) and the no access state (ST1 in FIG. 2).
(I) is a signal waveform when ALM0001-N (active at a low level) is given and the control right is given to the CPU 2.

【0052】(j)は、D0010であり、CPU2が
アクセス待ち状態(図2のST7)のときの波形であ
る。(k)は、D0011であり、図2の規定外状態の
ときの波形である。(l)は、D0100であり、CP
U1アクセス待ち状態(図2のST6)のときの波形で
ある。(m)は、D0101であり、CPU2アクセス
・CPU1アクセス待ち状態(図2のST5)のときの
波形である。(n)は、D0110であり、図2の規定
外状態のときの波形である。(o)は、D0111であ
り、CPU2空アクセス状態(図2のST8)のときの
波形である。(p)は、CAP1000−N(ロウレベ
ルのときにアクティブ)であり、CPU1に制御権付与
したときの信号波形を表す。
(J) is a waveform when D0010 is the CPU 2 in the access waiting state (ST7 in FIG. 2). (K) is D0011, which is a waveform in the non-specified state of FIG. (L) is D0100 and CP
This is a waveform in the U1 access wait state (ST6 in FIG. 2). (M) is a waveform of D0101 in the CPU2 access / CPU1 access waiting state (ST5 in FIG. 2). (N) is D0110, which is the waveform in the non-specified state of FIG. (O) is a waveform when D0111 is in the CPU2 idle access state (ST8 in FIG. 2). (P) is CAP1000-N (active at low level), and represents a signal waveform when the control right is given to the CPU 1.

【0053】(q)は、D1001であり、図2の規定
外状態のときの波形を表す。(r)は、D1010であ
り、CPU1アクセス・CPU2アクセス待ち状態(図
2のST4)のときの波形を表す。(s)は、D101
1であり、図2の規定外状態のときの波形を表す。
(t)は、D1100であり、図2の規定外状態のとき
の波形を表す。(u)は、D1101であり、図2の規
定外状態のときの波形を表す。
(Q) is D1001, and represents the waveform in the non-specified state of FIG. (R) is D1010 and represents a waveform in the CPU1 access / CPU2 access waiting state (ST4 in FIG. 2). (S) is D101
1 and represents the waveform in the non-specified state of FIG.
(T) is D1100 and represents the waveform in the non-specified state of FIG. (U) is D1101, and represents the waveform in the non-specified state of FIG.

【0054】(v)は、D1110であり、図2の規定
外状態のときの波形を表す。(w)は、D1111であ
り、図2の規定外状態のときの波形を表す。(x)は、
16Mであり、CPU2の16MHzクロックの信号波
形を表している。(y)は、SYNC−CLK1であ
り、CPU1側に対応したバス調停回路10の10MH
zの同期クロックの波形である。(z)は、CAPAC
CCS−N(ロウレベルのときにアクティブ)であり、
CPU1アドレスバッファイネーブル信号cの波形を表
す。(za)は、CAPCRAMRDY−N(ロウレベ
ルのときにアクティブ)であり、CPU1応答READ
Y信号bの波形を表す。
(V) is D1110, which represents the waveform in the non-specified state of FIG. (W) is D1111 and represents the waveform in the non-specified state of FIG. (X) is
16M, which represents the signal waveform of the 16 MHz clock of the CPU 2. (Y) is SYNC-CLK1, which is 10 MH of the bus arbitration circuit 10 corresponding to the CPU 1 side.
It is a waveform of a synchronous clock of z. (Z) is CAPAC
CCS-N (active at low level),
The waveform of the CPU1 address buffer enable signal c is shown. (Za) is CAPCRAMRDY-N (active at low level), and CPU1 response READ
The waveform of the Y signal b is shown.

【0055】(zb)は、ALMACCCS−N(ロウ
レベルのときにアクティブ)であり、CPU2のアドレ
スバッファイネーブル信号fの波形を表す。(zc)
は、ALMRDYであり、ハイレベルのときにアクティ
ブであり、CPU2に対する応答READY信号eの波
形を表す。(zd)は、CRAMCS−N(ロウレベル
のときにアクティブ)であり、S−RAM回路8への入
力チップセレクト信号ramcs−nの波形を表す。
(Zb) is ALMACCCS-N (active at low level) and represents the waveform of the address buffer enable signal f of the CPU 2. (Zc)
Represents the waveform of the response READY signal e to the CPU 2, which is ALMRDY and is active when it is at a high level. (Zd) is CRAMCS-N (active when low level) and represents the waveform of the input chip select signal ramcs-n to the S-RAM circuit 8.

【0056】(CPU1だけのチップセレクトが有意状
態に固定された場合の動作):図5において、(b)の
CPU1が出力するチップセレクト信号master1
cs−nである、CSCRAM−N信号は、CPU1が
制御権を獲得するためにロウレベル(アクティブ)を出
力している。一方、(c)のCPU2が出力するチップ
セレクト信号master2cs−nである、ALMC
S−N信号は、制御権を獲得しようとしていないのでハ
イレベル(非アクティブ)である。(d)〜(g)のレ
ジスタREG1〜REG4の出力信号である、REG1
OUT〜REG4OUTは、REG1OUTだけがハイ
レベル(論理1)で、他は全てロウレベル(論理0)を
維持する。
(Operation when the chip select of only the CPU1 is fixed to the significant state): In FIG. 5, the chip select signal master1 output from the CPU1 of FIG.
The CSCRAM-N signal, which is cs-n, outputs a low level (active) for the CPU 1 to acquire the control right. On the other hand, the chip select signal master2cs-n output from the CPU 2 in FIG.
The SN signal is at a high level (inactive) because it is not trying to acquire the control right. REG1 which is an output signal of the registers REG1 to REG4 in (d) to (g)
Regarding OUT to REG4OUT, only REG1OUT is at a high level (logic 1), and the rest of them maintain a low level (logic 0).

【0057】(h)のアクセスなし状態(0000)を
表す、D0000信号は、ハイレベルを維持している。
(i)のCPU2に制御権付与をしていることを表す、
ALM0001−N信号は、ハイレベル(非アクティ
ブ)となっているので制御権を付与していない。(j)
のCPU2がアクセス待ち状態を表す、D0010信号
はハイレベル(非アクティブ)となっている。(k)の
規定外状態である、D0011信号であり、ハイレベル
を維持している。(l)のCPU1アクセス待ち状態を
表す、D0100信号はハイレベル(非アクティブ)と
なっている。
The D0000 signal, which represents the no access state (0000) in (h), maintains a high level.
It represents that the control right is given to the CPU 2 of (i),
Since the ALM0001-N signal is at high level (inactive), the control right is not given. (J)
The D0010 signal indicating that the CPU 2 is waiting for access is at a high level (inactive). It is the D0011 signal which is in the out-of-specification state of (k) and maintains a high level. The D0100 signal indicating the CPU1 access waiting state of (l) is at a high level (inactive).

【0058】(m)のCPU2アクセス・CPU1アク
セス待ち状態を表す、D0101信号はハイレベル(非
アクティブ)となっている。(n)の規定外状態を表す
D0110信号はハイレベルを維持している。(o)の
CPU2空アクセス状態を表す、D0111信号はハイ
レベル(非アクティブ)となっている。(p)のCPU
1に制御権付与したときを表す、CAP1000−N信
号は、ロウレベル(アクティブ)となっている。(q)
の規定外状態を表すD1001信号は、ハイレベルを維
持している。(r)のCPU1アクセス・CPU2アク
セス待ち状態を表す、D1010信号は、ハイレベル
(非アクティブ)となっている。
The D0101 signal indicating the CPU2 access / CPU1 access waiting state of (m) is at a high level (inactive). The D0110 signal indicating the non-regulated state of (n) maintains the high level. The D0111 signal representing the CPU2 empty access state of (o) is at a high level (inactive). (P) CPU
The CAP1000-N signal, which indicates when the control right is given to 1, is at a low level (active). (Q)
The D1001 signal representing the non-regulated state of is maintained at the high level. The D1010 signal indicating the CPU1 access / CPU2 access waiting state of (r) is at a high level (inactive).

【0059】(s)の規定外状態を表すD1011信号
は、ハイレベル(非アクティブ)を維持している。
(t)の規定外状態を表すD1100信号もハイレベル
(非アクティブ)を維持している。(u)の規定外状態
を表すD1101もハイレベル(非アクティブ)を維持
している。(v)の規定外状態を表すD1110もハイ
レベル(非アクティブ)を維持している。(w)の規定
外状態を表すD1111もハイレベル(非アクティブ)
を維持している。(z)のCPU1アドレスバッファイ
ネーブル信号cを表す、CAPACCCS−N信号は、
ハイレベル(非アクティブ)とロウレベル(アクティ
ブ)とを繰り返している。ロウレベルの期間が長く、ハ
イレベルになるのは、3〜4クロック周期の期間であ
る。
The D1011 signal indicating the out-of-specification state of (s) is maintained at a high level (inactive).
The D1100 signal indicating the out-of-specification state of (t) also maintains the high level (inactive). D1101 representing the non-regulated state of (u) also maintains a high level (inactive). D1110, which represents the non-regulated state of (v), also maintains a high level (inactive). D1111, which represents the non-regulated state in (w), is also high level (inactive)
Has been maintained. The CAPACCCS-N signal representing the CPU1 address buffer enable signal c of (z) is
High level (inactive) and low level (active) are repeated. The low-level period is long and the high-level period is 3 to 4 clock cycles.

【0060】(za)のCPU1応答READY信号b
を表すCAPCRAMRDY−N信号は、ハイレベル
(非アクティブ)とロウレベル(アクティブ)とを繰り
返している。ハイレベルの期間が長く、ロウレベルにな
るのは、2クロック周期程度の幅である。(zb)のC
PU2のアドレスバッファイネーブル信号fである、A
LMACCCS−N信号は、ハイレベル(非アクティ
ブ)となっている。(zc)のCPU2に対する応答R
EADY信号eを表す、ALMRDY信号は、ロウレベ
ル(非アクティブ)となっている。(zd)のS−RA
M回路8への入力チップセレクト信号ramcs−nで
ある、CRAMCS−N信号は、ハイレベル(非アクテ
ィブ)とロウレベル(アクティブ)とを繰り返してい
る。ハイレベルとロウレベルの期間は、ほぼ同じであり
2クロック周期程度の幅である。
(Za) CPU1 response READY signal b
The CAPCRAMRDY-N signal that represents is repeatedly high level (inactive) and low level (active). The high-level period is long and the low-level period has a width of about two clock cycles. C of (zb)
A, which is the address buffer enable signal f of PU2,
The LMACCCS-N signal is at high level (inactive). Response R of (zc) to CPU2
The ALMRDY signal, which represents the EADY signal e, is at a low level (inactive). (Zd) S-RA
The CRAMCS-N signal, which is the input chip select signal ramcs-n to the M circuit 8, repeats high level (inactive) and low level (active). The high level period and the low level period are almost the same and have a width of about two clock cycles.

【0061】以上のようにして、CPU2は待機し、C
PU1だけのチップセレクトが有意状態に固定された場
合は、CPU1だけに対する制限権の付与動作となり、
単純な動作となる。
As described above, the CPU 2 waits and the C
When the chip select of only PU1 is fixed to the significant state, the restricting right is granted to only CPU1.
It is a simple operation.

【0062】(CPU2だけのチップセレクトが有意状
態に固定された場合の動作):図6において、(b)の
CPU1が出力するチップセレクト信号master1
cs−nである、CSCRAM−N信号は、CPU1が
制御権を獲得しないためにハイレベル(非アクティブ)
を出力している。(c)のCPU2が出力するチップセ
レクト信号master2cs−nである、ALMCS
−N信号は、制御権を獲得するためにロウレベル(アク
ティブ)である。(d)〜(g)のレジスタREG1〜
REG4の出力信号であるREG1OUT〜REG4O
UTは、REG4OUTだけがハイレベル(論理1)
で、他は全てロウレベル(論理0)を維持する。
(Operation when the chip select of only the CPU2 is fixed to the significant state): In FIG. 6, the chip select signal master1 output from the CPU1 of (b).
The CSCRAM-N signal, which is cs-n, is high level (inactive) because the CPU 1 does not acquire the control right.
Is output. ALMCS, which is the chip select signal master2cs-n output by the CPU 2 in (c).
The -N signal is at a low level (active) to acquire the control right. Registers REG1 to (d) to (g)
REG1OUT to REG4O which are output signals of REG4
In UT, only REG4OUT is at high level (logic 1)
Then, all others are kept at the low level (logic 0).

【0063】(h)のアクセスなし状態(0000)を
表す、D0000信号は、ハイレベルを維持している。
(i)のCPU2に制御権付与をしていることを表す、
ALM0001−N信号は、ロウレベル(アクティブ)
となっている。(j)のCPU2がアクセス待ち状態を
表す、D0010信号はハイレベル(非アクティブ)と
なっている。(k)の規定外状態である、D0011信
号であり、ハイレベルを維持している。(l)のCPU
1アクセス待ち状態を表す、D0100信号はハイレベ
ル(非アクティブ)となっている。(m)のCPU2ア
クセス・CPU1アクセス待ち状態を表す、D0101
信号はハイレベル(非アクティブ)となっている。
(n)の規定外状態を表すD0110信号はハイレベル
を維持している。(o)のCPU2空アクセス状態を表
す、D0111信号はハイレベル(非アクティブ)とな
っている。
The D0000 signal, which represents the no access state (0000) in (h), maintains a high level.
It represents that the control right is given to the CPU 2 of (i),
ALM0001-N signal is low level (active)
It has become. The D0010 signal indicating that the CPU 2 in (j) is in an access waiting state is at a high level (inactive). It is the D0011 signal which is in the out-of-specification state of (k) and maintains a high level. (L) CPU
The D0100 signal, which indicates a one-access waiting state, is at a high level (inactive). D0101 showing the CPU2 access / CPU1 access waiting state of (m)
The signal is high level (inactive).
The D0110 signal indicating the non-regulated state of (n) maintains the high level. The D0111 signal representing the CPU2 empty access state of (o) is at a high level (inactive).

【0064】(p)のCPU1に制御権付与したときを
表す、CAP1000−N信号は、ハイレベル(非アク
ティブ)となっている。(q)の規定外状態を表すD1
001信号は、ハイレベルを維持している。(r)のC
PU1アクセス・CPU2アクセス待ち状態を表す、D
1010信号は、ハイレベル(非アクティブ)となって
いる。(s)の規定外状態を表すD1011信号は、ハ
イレベル(非アクティブ)を維持している。(t)の規
定外状態を表すD1100信号もハイレベル(非アクテ
ィブ)を維持している。(u)の規定外状態を表すD1
101もハイレベル(非アクティブ)を維持している。
(v)の規定外状態を表すD1110もハイレベル(非
アクティブ)を維持している。(w)の規定外状態を表
すD1111もハイレベル(非アクティブ)を維持して
いる。
The CAP1000-N signal, which indicates when the control right is given to the CPU 1 in (p), is at a high level (inactive). D1 representing the non-regulated state of (q)
The 001 signal maintains the high level. (R) C
D, which represents the state of waiting for PU1 access / CPU2 access
The 1010 signal is at high level (inactive). The D1011 signal indicating the non-regulated state of (s) is maintained at the high level (inactive). The D1100 signal indicating the out-of-specification state of (t) also maintains the high level (inactive). D1 representing the non-regulated state of (u)
101 also maintains a high level (inactive).
D1110, which represents the non-regulated state of (v), also maintains a high level (inactive). D1111, which represents the non-specified state in (w), also maintains a high level (inactive).

【0065】(z)のCPU1アドレスバッファイネー
ブル信号cを表す、CAPACCCS−N信号は、ハイ
レベル(非アクティブ)となる。(za)のCPU1応
答READY信号bを表す、CAPCRAMRDY−N
信号は、ハイレベル(非アクティブ)となる。(zb)
のCPU2のアドレスバッファイネーブル信号fであ
る、ALMACCCS−N信号は、ハイレベル(非アク
ティブ)とロウレベル(アクティブ)とを繰り返す。ほ
とんどの期間はロウレベル(アクティブ)となる。(z
c)のCPU2に対する応答READY信号eを表す、
ALMRDY信号は、ロウレベル(非アクティブ)とハ
イレベル(アクティブ)とを繰り返す。ほとんど期間は
ロウレベル(アクティブ)となる。
The CAPACCCS-N signal (z), which represents the CPU1 address buffer enable signal c, becomes high level (inactive). CAPCRAMRDY-N representing the CPU1 response READY signal b of (za)
The signal becomes high level (inactive). (Zb)
The ALMACCCS-N signal, which is the address buffer enable signal f of the CPU 2 of, repeats high level (inactive) and low level (active). It is low level (active) for most of the period. (Z
c) represents a response READY signal e to the CPU 2,
The ALMRDY signal repeats low level (inactive) and high level (active). It is at low level (active) for most of the period.

【0066】(zd)のS−RAM回路8への入力チッ
プセレクト信号ramcs−nである、CRAMCS−
N信号は、ハイレベル(非アクティブ)とロウレベル
(アクティブ)とを繰り返している。ハイレベルとロウ
レベルの期間は、ほぼ同じであり2クロック周期程度の
幅である。
(Zd) CRAMCS-, which is the input chip select signal ramcs-n to the S-RAM circuit 8.
The N signal repeats high level (inactive) and low level (active). The high level period and the low level period are almost the same and have a width of about two clock cycles.

【0067】以上のようにして、CPU1を待機させ、
CPU2だけのチップセレクトが有意状態に固定された
場合は、CPU2だけに対する制御権の付与動作とな
り、単純な動作となる。
As described above, the CPU 1 is made to stand by,
When the chip select of only the CPU 2 is fixed to the significant state, the operation of giving the control right to only the CPU 2 is performed, which is a simple operation.

【0068】(CPU1、CPU2との両方のチップセ
レクト信号が有意状態に固定され続けた場合の動作):
図4において、(b)のCPU1が出力するチップ
セレクト信号master1cs−nである、CSCR
AM−N信号は、ロウレベル(アクティブ)である。
(c)のCPU2が出力するチップセレクト信号mas
ter2cs−nである、ALMCS−N信号もロウレ
ベル(アクティブ)である。ここで、CPU1、2のチ
ップセレクト信号が同じ時期に発生している。(d)〜
(g)の、レジスタREG1〜REG4の出力信号であ
る、REG1OUT〜REG4OUTは、ハイレベルと
ロウレベルとを繰り返している。(h)のアクセスなし
状態(0000)を表す、D0000信号は、ハイレベ
ルを維持している。
(Operation when both chip select signals of CPU1 and CPU2 are continuously fixed to the significant state):
In FIG. 4, the CSCR, which is the chip select signal master1cs-n output from the CPU 1 in (b),
The AM-N signal is low level (active).
(C) Chip select signal mas output by CPU2
The ALMCS-N signal, which is ter2cs-n, is also low level (active). Here, the chip select signals of the CPUs 1 and 2 are generated at the same time. (D) ~
The output signals REG1OUT to REG4OUT of the registers REG1 to REG4 in (g) repeat high level and low level. The D0000 signal indicating the no access state (0000) in (h) maintains the high level.

【0069】(i)のCPU2に制御権付与を行ってい
ることを表す、ALM0001−N信号は、ハイレベル
(非アクティブ:非制御権付与)とロウレベル(アクテ
ィブ:制御権付与)とを代わるがわる繰り返している。
(j)のCPU2がアクセス待ち状態を表す、D001
0信号はハイレベル(非アクティブ:非制御権付与)と
ロウレベル(アクティブ:制御権付与)とを繰り返して
いる。ハイレベルの期間が長く、ロウレベル(制御権付
与)になるのは、1クロック周期の幅で2回である。
(k)は規定外状態である、D0011信号であり、ハ
イレベルを維持している。
The ALM0001-N signal (i), which indicates that the CPU 2 is given the control right, switches between a high level (inactive: non-control right granted) and a low level (active: control right granted). It's repeating badly.
The CPU 2 of (j) represents the access waiting state, D001
The 0 signal repeats high level (inactive: non-control right grant) and low level (active: control right grant). The high level period is long and the low level (granting control right) occurs twice in the width of one clock cycle.
(K) is a D0011 signal that is in an out-of-specification state and maintains a high level.

【0070】(l)のCPU1アクセス待ち状態を表
す、D0100信号はハイレベル(非アクティブ)とロ
ウレベル(アクティブ)とを繰り返している。ハイレベ
ルの期間が長く、ロウレベルになるのは、1クロック周
期の幅で2回である。(m)のCPU2アクセス・CP
U1アクセス待ち状態を表す、D0101信号はハイレ
ベル(非アクティブ)とロウレベル(アクティブ)とを
繰り返している。ハイレベルの期間が長く、ロウレベル
になるのは、1クロック周期の幅で2回である。(n)
の規定外状態を表すD0110信号はハイレベルを維持
している。
The signal D0100, which represents the CPU 1 access waiting state (1), repeatedly repeats high level (inactive) and low level (active). The high-level period is long and the low-level period is twice in the width of one clock cycle. (M) CPU2 access / CP
The D0101 signal, which represents the U1 access wait state, repeats high level (inactive) and low level (active). The high-level period is long and the low-level period is twice in the width of one clock cycle. (N)
The D0110 signal representing the non-regulated state of is maintained at the high level.

【0071】(o)のCPU2空アクセス状態を表す、
D0111信号はハイレベル(非アクティブ)とロウレ
ベル(アクティブ)とを繰り返している。ハイレベルの
期間が長く、ロウレベルになるのは、1クロック周期の
幅で2回である。(p)のCPU1に制御権付与したと
きを表す、CAP1000−N信号は、ハイレベル(非
アクティブ)とロウレベル(アクティブ:制御権付与)
とを繰り返している。ハイレベルの期間が長く、ロウレ
ベルになるのは、1クロック周期の幅で2回である。
(q)の規定外状態を表すD1001信号は、ハイレベ
ルを維持している。(r)のCPU1アクセス・CPU
2アクセス待ち状態を表す、D1010信号は、ハイレ
ベル(非アクティブ)とロウレベル(アクティブ)とを
繰り返している。ハイレベルの期間が長く、ロウレベル
になるのは、1クロック周期の幅で2回である。
(O) represents the CPU2 idle access state,
The D0111 signal repeats high level (inactive) and low level (active). The high-level period is long and the low-level period is twice in the width of one clock cycle. The CAP1000-N signal, which indicates when the control right is given to the CPU 1 in (p), is at a high level (inactive) and a low level (active: control right is given).
Is repeated. The high-level period is long and the low-level period is twice in the width of one clock cycle.
The D1001 signal representing the out-of-specification state of (q) maintains the high level. (R) CPU1 access CPU
The D1010 signal, which represents the two-access waiting state, repeats high level (inactive) and low level (active). The high-level period is long and the low-level period is twice in the width of one clock cycle.

【0072】(s)の規定外状態を表すD1011信号
は、ハイレベル(非アクティブ)を維持している。
(t)の規定外状態を表すD1100信号もハイレベル
(非アクティブ)を維持している。(u)の規定外状態
を表すD1101もハイレベル(非アクティブ)を維持
している。(v)の規定外状態を表すD1110もハイ
レベル(非アクティブ)を維持している。(w)の規定
外状態を表すD1111もハイレベル(非アクティブ)
を維持している。
The D1011 signal indicating the non-specified state of (s) is maintained at the high level (inactive).
The D1100 signal indicating the out-of-specification state of (t) also maintains the high level (inactive). D1101 representing the non-regulated state of (u) also maintains a high level (inactive). D1110, which represents the non-regulated state of (v), also maintains a high level (inactive). D1111, which represents the non-regulated state in (w), is also high level (inactive)
Has been maintained.

【0073】(z)のCPU1アドレスバッファイネー
ブル信号cを表す、CAPACCCS−N信号は、ハイ
レベル(非アクティブ)とロウレベル(アクティブ:イ
ネーブブル)とを繰り返している。ハイレベルの期間が
長く、ロウレベルになるのは、3〜4クロック周期の幅
で2回程度である。(za)のCPU1応答READY
信号bを表す、CAPCRAMRDY−N信号は、ハイ
レベル(非アクティブ)とロウレベル(アクティブ:イ
ネーブル)とを繰り返している。ハイレベルの期間が長
く、ロウレベルになるのは、1クロック周期の幅で2回
程度である。
The CAPACCCS-N signal (z) representing the CPU1 address buffer enable signal c repeats high level (inactive) and low level (active: enable). The high-level period is long, and the high-level period is low level about twice in the width of 3 to 4 clock cycles. (Za) CPU1 response READY
The CAPCRAMRDY-N signal representing the signal b repeats high level (inactive) and low level (active: enable). The high-level period is long and the low-level period is about twice in one clock cycle.

【0074】(zb)のCPU2のアドレスバッファイ
ネーブル信号fである、ALMACCCS−N信号は、
ハイレベル(非アクティブ)とロウレベル(アクティ
ブ:イネーブル)とを繰り返している。ハイレベルとロ
ウレベルの期間はほぼ同じであり、4クロック周期程度
の幅である。(zc)のCPU2に対する応答READ
Y信号eを表す、ALMRDY信号は、ハイレベル(ア
クティブ)とロウレベル(非アクティブ)とを繰り返し
ている。ハイレベルの期間が1クロック幅程度である。
(zd)のS−RAM回路8への入力チップセレクト信
号ramcs−nである、CRAMCS−N信号は、ハ
イレベル(非アクティブ)とロウレベル(アクティブ)
とを繰り返している。ハイレベルとロウレベルの期間
は、ほぼ同じであり2クロック周期程度の幅である。
The ALMACCCS-N signal, which is the address buffer enable signal f of the CPU 2 in (zb), is
High level (inactive) and low level (active: enable) are repeated. The high level period and the low level period are almost the same and have a width of about 4 clock cycles. (Zc) response to CPU2 READ
The ALMRDY signal, which represents the Y signal e, repeats high level (active) and low level (inactive). The high level period is about one clock width.
The CRAMCS-N signal, which is the input chip select signal ramcs-n to the (zd) S-RAM circuit 8, is at a high level (inactive) and a low level (active).
Is repeated. The high level period and the low level period are almost the same and have a width of about two clock cycles.

【0075】図4において、図4の(b)のCSCRA
M−Nと、図4の(c)のALMCS−Nとの両方の信
号が有意状態に(アクティブ:ロウレベルに)固定さ
れ、つまり、2種類のCPU1、2がS−RAM回路8
に対してアクセスしようとしているときに、図4の
(z)のCPU1アドレスバッファイネーブル信号cで
あるCAPACCCS−N信号と、図4の(zb)のC
PU2のアドレスバッファイネーブル信号fであるAL
MACCCS−N信号に注目すると、交互に有意状態
(アクティブ:ロウレベル)になっている。このことは
S−RAM回路8に対するアドレスバス、データバス、
その他に対して、2種類のCPU1、2が競合していな
いことと、両方のCPU1、2に対して、図(i)、
(p)に示すように平等に制御権が与えられていること
を示している。少しロウレベル又はハイレベルの信号の
長さに違いがあるのは、今回の場合は、CPU1、2の
システムクロックサイクルの長さが異なるためである。
即ち、CPU1は、1サイクルを2システムクロックと
し、CPU2は、1サイクルを4システムクロックサイ
クルとしている。
In FIG. 4, the CSCR A of FIG.
The signals of both MN and ALMCS-N of FIG. 4C are fixed to a significant state (active: low level), that is, the two types of CPUs 1 and 2 are S-RAM circuits 8
4C, the CAPACCCS-N signal which is the CPU1 address buffer enable signal c of FIG. 4Z and the C of FIG.
AL which is the address buffer enable signal f of PU2
Focusing on the MACCCS-N signal, they are alternately in a significant state (active: low level). This means that the address bus, data bus, and
On the other hand, the two types of CPUs 1 and 2 do not compete with each other, and FIG.
As shown in (p), it is shown that the control right is given equally. There is a slight difference in the lengths of the low-level and high-level signals because the lengths of the system clock cycles of the CPUs 1 and 2 are different in this case.
That is, the CPU 1 has one cycle as two system clocks, and the CPU 2 has one cycle as four system clock cycles.

【0076】(本発明の実施の形態の効果): 以上
の本発明の実施の形態によれば、S−RAM回路を用い
てマルチポートメモリ回路と同じ機能を果たす回路を構
成することができ、CPU又はマイクロプロセッサであ
るマスタは、共通のS−RAM回路8のバスの解放状態
を確認することなく、任意の時間にアクセスすることが
でき、特別な動作、例えば、バス調停回路10等のファ
ームウエア的な確認動作が不必要となる。
(Effects of the Embodiment of the Present Invention) According to the embodiments of the present invention described above, it is possible to configure a circuit having the same function as that of the multiport memory circuit by using the S-RAM circuit. A master, which is a CPU or a microprocessor, can access at any time without checking the release state of the bus of the common S-RAM circuit 8, and can perform a special operation, for example, a firmware of the bus arbitration circuit 10 or the like. The checking operation in terms of wear is unnecessary.

【0077】また、制御待ち時間は、最大でも制御優先
権選択部の動作時間と相手側のマスタのCPUの1クロ
ックサイクル分の時間である。更に、CPU1、2の処
理サイクル周期が異なっても、制御要請に応じて代わる
がわる制御権を付与して処理することができるので、効
率的に処理することができる。
The control waiting time is at most the operation time of the control priority selection unit and one clock cycle of the CPU of the master on the other side. Further, even if the processing cycle periods of the CPUs 1 and 2 are different, the control right can be given and processed in response to the control request, so that the processing can be performed efficiently.

【0078】従って、シングルポートのS−RAM回路
8に対して一系統のバスラインを通じて複数のCPUか
らアクセスすることが簡単な構成で容易に行うマルチC
PU装置を実現することができる。
Therefore, the multi-C can easily access the single-port S-RAM circuit 8 from a plurality of CPUs through one bus line.
A PU device can be realized.

【0079】(他の実施の形態): (1)尚、以上
の実施の形態においては、マイクロプロセッサのマスタ
には、80186とを使用したものを示したが、これ以
外の規格、仕様のマイクロプロセッサを使用した場合も
タイミング生成部の生成時間を変更することで適用可能
である。
Other Embodiments: (1) In the above embodiments, the microprocessor of 80186 is used as the master of the microprocessor. However, the microprocessor of other standards or specifications is used. Even when the processor is used, it can be applied by changing the generation time of the timing generation unit.

【0080】(2)また、制御優先選択部を変更するこ
とによって更に多くのマイクロプロセッサが一つのS−
RAM回路に対してアクセスする構成も可能である。
(2) Further, by changing the control priority selection section, more microprocessors can use one S-.
A configuration for accessing the RAM circuit is also possible.

【0081】(3)更に、S−RAM回路の代わりにD
(ダイナミック)−RAM回路でも良いし、S−RAM
回路とD−RAM回路との混合構成のメモリ回路でもよ
いし、レジスタで構成されるレジスタ回路でもよい。D
−RAMは記憶保持動作が必要なメモリだが、S−RA
Mはこの動作が不必要であるので、構成が簡単になる。
D−RAMにおいては、シンクロナスD−RAMを使用
することで高速動作を実現することもできる。
(3) Further, instead of the S-RAM circuit, D
(Dynamic) -RAM circuit, S-RAM
A memory circuit having a mixed configuration of a circuit and a D-RAM circuit may be used, or a register circuit including a register may be used. D
-RAM is a memory that requires a memory holding operation, but S-RA
Since M does not need this operation, the configuration is simple.
In the D-RAM, high speed operation can be realized by using the synchronous D-RAM.

【0082】(4)更にまた、上述のマルチCPUシス
テムは、例えば、次のようなシステムに適用すること
で、小型のシステムを能率的、機能的にする上で好適で
ある。即ち、伝送システムの故障異常を監視し、この監
視状態を通報する監視装置においては、各部の回路の状
態を監視して、異常箇所を検出するための第1のCPU
と、異常箇所を検出すると、この異常箇所を情報として
蓄積するためのS−RAM回路と、このS−RAM回路
に蓄積されている情報を時事刻々ホスト側に伝送するた
めの第2のCPUとから構成することも好ましい。
(4) Furthermore, the multi-CPU system described above is suitable for making a small system efficient and functional by applying it to the following system, for example. That is, in the monitoring device for monitoring the failure abnormality of the transmission system and notifying the monitoring state, the first CPU for monitoring the state of the circuit of each part and detecting the abnormal portion
When an abnormal point is detected, an S-RAM circuit for storing the abnormal point as information, and a second CPU for transmitting the information stored in the S-RAM circuit to the host side every moment. It is also preferable to be composed of

【0083】[0083]

【発明の効果】以上で説明したように本発明は、複数の
CPUからのチップセレクト信号が競合した場合に、こ
れらの複数のチップセレクト信号に対する制御要請に対
して代わるがわる制御権を与えて制御権信号を出力し、
この制御権を与えられない期間は制御要請に対して制御
待ちさせ、制御権信号に対応するCPUへの制御許可を
与えるための制御許可応答信号を生成してCPUへ与
え、制御待ちさせているチップセレクト信号を出力して
いるCPUに対して制御待ち応答信号を与え、制御権信
号に基づき記憶回路のアクセスタイムに対応するように
記憶回路用のチップセレクト信号を生成して与えること
で、シングルポートの記憶回路(例えば、S−RAM回
路や、D−RAM回路など)に対して一系統のバスライ
ンを通じて複数のCPUからアクセスすることを簡単な
構成で容易に行うマルチCPU装置を実現することがで
きる。
As described above, according to the present invention, when chip select signals from a plurality of CPUs compete with each other, the control right is given instead of the control request for the plurality of chip select signals. Output control right signal,
During the period in which the control right is not given, the control is made to wait for the control request, and the control permission response signal for giving the control permission to the CPU corresponding to the control right signal is generated and given to the CPU to wait for the control. By giving a control wait response signal to the CPU that is outputting the chip select signal, and generating and giving a chip select signal for the memory circuit so as to correspond to the access time of the memory circuit based on the control right signal, a single signal is obtained. To realize a multi-CPU device that easily allows a plurality of CPUs to access a port storage circuit (for example, an S-RAM circuit, a D-RAM circuit, etc.) through a single bus line. You can

【図面の簡単な説明】[Brief description of drawings]

【図1】本発明の実施の形態のマルチCPU装置の機能
構成図である。
FIG. 1 is a functional configuration diagram of a multi-CPU device according to an embodiment of the present invention.

【図2】図1の実施の形態のマルチCPU装置の制御優
先選択回路Aの一例の動作状態の詳細な遷移図である。
FIG. 2 is a detailed transition diagram of an operation state of an example of the control priority selection circuit A of the multi CPU device according to the embodiment of FIG.

【図3】図1の制御優先選択回路Aの一例の詳細な回路
構成である。
3 is a detailed circuit configuration of an example of a control priority selection circuit A of FIG.

【図4】実施の形態のCPU1、CPU2との両方のチ
ップセレクト信号が有意状態に固定され続けた場合の動
作タイミングチャートである。
FIG. 4 is an operation timing chart when the chip select signals of both CPU1 and CPU2 of the embodiment continue to be fixed to a significant state.

【図5】実施の形態のCPU1だけのチップセレクトが
有意状態に固定された場合の動作タイミングチャートで
ある。
FIG. 5 is an operation timing chart when the chip select of only the CPU 1 of the embodiment is fixed to a significant state.

【図6】実施の形態のCPU2だけのチップセレクトが
有意状態に固定された場合の動作タイミングチャートで
ある。
FIG. 6 is an operation timing chart when the chip select of only the CPU 2 of the embodiment is fixed to a significant state.

【符号の説明】[Explanation of symbols]

1、2…CPU、8…S−RAM回路、9…AND回
路、10…バス調停回路、A…制御優先選択回路、B…
CPU1READY応答タイミング生成回路、C…CP
U1アドレスバッファ制御タイミング生成回路、D…C
PU1チップセレクト入力タイミング生成回路、E…C
PU2READY応答タイミング生成回路、F…CPU
2アドレスバッファ制御タイミング生成回路、G…CP
U2チップセレクト入力タイミング生成回路、H…CP
U1アドレスバッファ回路、I…CPU2アドレスバッ
ファ回路、J…CPU1データバッファ回路、K…CP
U2データバッファ回路。
1, 2 ... CPU, 8 ... S-RAM circuit, 9 ... AND circuit, 10 ... Bus arbitration circuit, A ... Control priority selection circuit, B ...
CPU1READY response timing generation circuit, C ... CP
U1 address buffer control timing generation circuit, D ... C
PU1 chip select input timing generation circuit, E ... C
PU2READY response timing generation circuit, F ... CPU
2 address buffer control timing generation circuit, G ... CP
U2 chip select input timing generation circuit, H ... CP
U1 address buffer circuit, I ... CPU2 address buffer circuit, J ... CPU1 data buffer circuit, K ... CP
U2 data buffer circuit.

Claims (3)

【特許請求の範囲】[Claims] 【請求項1】 複数のCPUが一系統のバスラインを通
じてシングルポートの記憶回路にアクセスするマルチC
PU装置において、 上記各CPUから出力されるチップセレクト信号のアク
ティブ/非アクティブの状態を監視し、同じ時期に2以
上のCPUからのチップセレクト信号が競合した場合
に、これらの複数のチップセレクト信号に対する制御要
請に対して代わるがわる制御権を与えて制御権信号を出
力し、この制御権を与えられない期間は制御要請に対し
て制御待ちさせる調停手段と、 上記制御権信号に対応するCPUへの制御許可を与える
ための制御許可応答信号を生成して上記CPUへ与え、
上記制御待ちさせているチップセレクト信号を出力して
いるCPUに対して制御待ち応答信号を与える制御許可
/許可待ち応答信号生成手段と、 上記制御許可を与えられた上記CPUからのアドレスを
上記バスラインを通じて上記記憶回路に与えると共に、
上記制御権信号に基づき上記記憶回路のアクセスタイム
に対応するように上記記憶回路用のチップセレクト信号
を生成して上記記憶回路に与える記憶回路用チップセレ
クト信号生成手段とを備えたことを特徴とするマルチC
PU装置。
1. A multi-C in which a plurality of CPUs access a single-port memory circuit through a single bus line.
In the PU device, the active / inactive state of the chip select signals output from each of the CPUs is monitored, and if chip select signals from two or more CPUs compete at the same time, these multiple chip select signals are detected. Arbitration means for giving a control right instead of the control request to the control request signal, outputting a control right signal, and waiting for the control request during the period when the control right is not given, and a CPU corresponding to the control right signal. Generate a control permission response signal for giving control permission to
A control permission / permission waiting response signal generating means for giving a control waiting response signal to the CPU outputting the chip select signal for making the control wait, and an address from the CPU given the control permission for the bus. While giving to the above memory circuit through the line,
A memory circuit chip select signal generating means for generating a memory circuit chip select signal corresponding to the access time of the memory circuit based on the control right signal and giving the memory circuit chip select signal to the memory circuit. Multi C
PU device.
【請求項2】 上記制御許可を与えられた上記CPUか
らのアドレスをアドレスバスライン部に出力すると共
に、上記CPUと上記記憶回路との間のデータバスライ
ン部を通じてデータの授受を行うことを特徴とする請求
項1記載のマルチCPU装置。
2. The address from the CPU to which the control permission is given is output to an address bus line unit, and data is transmitted and received through the data bus line unit between the CPU and the memory circuit. The multi-CPU device according to claim 1.
【請求項3】 上記各CPUの処理サイクル周期がそれ
ぞれ異なることを特徴とする請求項1又は2記載のマル
チCPU装置。
3. The multi-CPU device according to claim 1, wherein the processing cycle period of each CPU is different.
JP7230796A 1996-03-27 1996-03-27 Multi-cpu device Pending JPH09259030A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP7230796A JPH09259030A (en) 1996-03-27 1996-03-27 Multi-cpu device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP7230796A JPH09259030A (en) 1996-03-27 1996-03-27 Multi-cpu device

Publications (1)

Publication Number Publication Date
JPH09259030A true JPH09259030A (en) 1997-10-03

Family

ID=13485491

Family Applications (1)

Application Number Title Priority Date Filing Date
JP7230796A Pending JPH09259030A (en) 1996-03-27 1996-03-27 Multi-cpu device

Country Status (1)

Country Link
JP (1) JPH09259030A (en)

Similar Documents

Publication Publication Date Title
JP4024875B2 (en) Method and apparatus for arbitrating access to shared memory for network ports operating at different data rates
US5710891A (en) Pipelined distributed bus arbitration system
US20010042178A1 (en) Data path architecture and arbitration scheme for providing access to a shared system resource
EP1564646A2 (en) Configurable embedded processor
JPH1021184A (en) Single-chip microcomputer with built-in dma
KR100708096B1 (en) Bus system and execution scheduling method for access commands thereof
JPH06231074A (en) Multiple access system for system bus
US5307466A (en) Distributed programmable priority arbitration
US7062588B2 (en) Data processing device accessing a memory in response to a request made by an external bus master
KR100487218B1 (en) Apparatus and method for interfacing an on-chip bus
JP4151362B2 (en) Bus arbitration method, data transfer device, and bus arbitration method
JPH09259030A (en) Multi-cpu device
JP2005503612A (en) Method of interfacing between external memory and processor supporting burst mode
JP3240863B2 (en) Arbitration circuit
JP2007052685A (en) Microcontroller
JPH0343804A (en) Sequence controller
JP2004213666A (en) Dma module and its operating method
US6499087B1 (en) Synchronous memory sharing based on cycle stealing
JPH064401A (en) Memory access circuit
JPH1139266A (en) Multiprocessor device
JPS6341973A (en) Multi-processor system
JPH08129523A (en) Computer system
JPH0333966A (en) Memory control circuit
JPH0764849A (en) Shared memory controller for processor
JPH03201054A (en) Method and device for control of common bus, master device, and computer system