JPS5952493A - Information processor - Google Patents

Information processor

Info

Publication number
JPS5952493A
JPS5952493A JP57160974A JP16097482A JPS5952493A JP S5952493 A JPS5952493 A JP S5952493A JP 57160974 A JP57160974 A JP 57160974A JP 16097482 A JP16097482 A JP 16097482A JP S5952493 A JPS5952493 A JP S5952493A
Authority
JP
Japan
Prior art keywords
common memory
address
clock signal
memory
access
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
JP57160974A
Other languages
Japanese (ja)
Inventor
Tetsuya Ikeda
哲也 池田
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.)
Hitachi Ltd
Original Assignee
Hitachi 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 Hitachi Ltd filed Critical Hitachi Ltd
Priority to JP57160974A priority Critical patent/JPS5952493A/en
Publication of JPS5952493A publication Critical patent/JPS5952493A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/21Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
    • G11C11/34Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)

Abstract

PURPOSE:To prevent the operating efficiency of a processor from being decreased, by applying a clock signal with changed phase at each processor and switching a system bus of each processor with the clock signal so as to access respectively a common memory. CONSTITUTION:An MPU clock signal generating circuit 3 applies the clock signal directly to an MPU2 and to an MPU1 via a phase inverting circuit 4. Thus, since an address switching circuit 5 and a data switching circuit 9 switch address buses 11, 21 and data buses 12, 22 in the MPU1 and MPU2 and apply the signal to the common memory 8, the MPU1 can access the common memory 8 at a half period of the clock signal and the MPU2 can access the memory at the remaining half period. Thus, when one MPU accesses the common memory 8, the other MPU is at an idle time and both the MPUs can access the common memory 8 independently.

Description

【発明の詳細な説明】 本発明tま、複数のマイクロプロセッサを有するマルチ
プロセッサ方式の情報処Jjl装置に係り、%に共通メ
モリとして使用ず石ダイナミックメモリをリフレッシユ
するのに好適力情報処j1)!装置に関する。
DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a multiprocessor type information processing device having a plurality of microprocessors, and is suitable for refreshing dynamic memory without using it as a common memory. ! Regarding equipment.

一般に複数のマイクロプロセッサを有するマルチプロセ
ッサシステムでれ、プロセラ・ν゛間のデータの授受に
は各々のプロセラ・ν°からアクセス可能外共通メモリ
がm−られる。この:IIE通メモリ社各々のプロ・ヒ
ップから同時にアクセスナ・ることはできないため、プ
ロセッサに俊度度を伺けて、優先度の高いプロセッサが
共通メモリをアクセスする揚台は、それ以外のプロセッ
サは停止させている。しかしながらこの方式は、プロセ
ッサ間のデータ転送が頻繁になってJ(通。
Generally, in a multiprocessor system having a plurality of microprocessors, a common memory m- which is not accessible from each processor ν° is used for exchanging data between the processors ν°. This: Since it is not possible to access the common memory from each of the IIE Tsumemory companies at the same time, it is possible to increase the agility of the processor and allow the high-priority processor to access the common memory from other processors. Processor is stopped. However, this method requires frequent data transfer between processors.

メモリのアクセス回数が増えると、プロセッツの稼動効
率が低下す石という欠点がある。
The disadvantage is that as the number of memory accesses increases, the operating efficiency of Prosets decreases.

さらに共通メモリを大容預、化するためにコスト的に有
利なダイナミックメモリを採用する場合にはメモリのア
ドレスを一定周期ごとに読み出すりフレッシー動作を必
要とする。この場合共通メモリはそれぞれのプn士ツリ
゛からのアクセスの他に、メモリをリフレッシj、する
ためのりフレッシコアドレスの供給を必要とし、この場
合リフレッシ、wb作はメモリの内容保存のためにはプ
ロセッサからのアクセス要求よυ優先すべき処理である
ため、プロセッサの稼動効率はさらに低下する。
Furthermore, if a cost-effective dynamic memory is used to increase the storage capacity of a common memory, it is necessary to read the address of the memory at regular intervals or perform a fresh operation. In this case, the common memory requires the supply of a fresh address in order to refresh the memory in addition to access from each network memory, and in this case, the refresh and wb operations are performed in order to save the contents of the memory. Since υ is a process that should be prioritized over access requests from the processor, the operating efficiency of the processor further decreases.

本発明の目的は、上記した従来の欠点をなくし、共通メ
モリにダイナミックメモリを用いたマルチプロセッサシ
ステムにおいて、プロセッサの1つが、共通メモリをア
クセスしても他のプロセッサを停止させることなく、寸
だ、共通メモリのりフレッシ轟動作のために、プロセッ
サの稼動効率を低下させる仁とのない情報処理装置を提
供することKある。
SUMMARY OF THE INVENTION An object of the present invention is to eliminate the above-mentioned conventional drawbacks, and in a multiprocessor system using dynamic memory as a common memory, even if one of the processors accesses the common memory, the other processors do not stop. Another object of the present invention is to provide an information processing device that does not reduce the operating efficiency of a processor due to the common memory's fresh operation.

、上記目的を達成するため本発明でtよ、すべてのプn
セッリ゛のクロック同期をとるために、同一のクロック
信号発生手段からのりnツク(17号をアクセス“す゛
ごとに位相を変えて供給12、共通メモリに供給される
各々のプj+七ッνのシスデノ、バスをクロック信号で
切JfJ8える?−とで、どのプロセッサも他のプロセ
ッサに4引)を力える仁となく共通メモリをアクー童−
スできるようにし、さらにプロセッサのどれもが共通メ
モリをアクセスしていな諭期間を検出して、その期間に
共洒メモリにリフレッシュアドレスを供給することで、
プロセッサの稼動効率を低下さぜないように17だもの
である。
In order to achieve the above object, the present invention provides all
In order to synchronize the clocks of the memory cells, the same clock signal generating means supplies a clock signal (17) with a different phase for each access, and a signal of Sysdeno, disconnecting the bus with a clock signal JfJ8 ?
Furthermore, by detecting a period when none of the processors is accessing the common memory and supplying a refresh address to the common memory during that period,
17 so as not to reduce the operating efficiency of the processor.

以下本発明を第1図および第2図を用いて詳n11に説
明する。第1図は本発明による情報処理装置の一実施例
を示すプnツク図である。第1図においてlおよび2は
マイクロプロセッサ(MJと略す)、3はMPUクロッ
ク信号発生回路、4は位相反転回路、5およびフはアド
レス切換回路、6)より7レツシご、アドレス発生回路
、8Ll、共通メモリ、9はデータ切換回路、〕、00
1M1UlおJ、び2かり1′:通メモリ8を選択しで
いるかどうかを検出する共通メモリ選択検出回路である
The present invention will be explained in detail below using FIGS. 1 and 2. FIG. 1 is a block diagram showing an embodiment of an information processing apparatus according to the present invention. In Fig. 1, 1 and 2 are microprocessors (abbreviated as MJ), 3 is an MPU clock signal generation circuit, 4 is a phase inversion circuit, 5 and , common memory, 9 is a data switching circuit, ], 00
1M1UlOJ, 2Kari1': This is a common memory selection detection circuit that detects whether or not the common memory 8 has been selected.

′!F、た11および12t」、AfpUl(7)7ド
レスハスオヨヒデータバス、2]、および22 i: 
M11U2のアドレスバスおよびデータバスである。
′! F, 11 and 12t'', AfpUl (7) 7 Dresshas Oyohi Data Bus, 2], and 22 i:
These are the address bus and data bus of M11U2.

第1図に示すマルチ・ンロセッザシステムの情報処理装
置の動作1d:以下の如くである。第3図においてAf
PUクロック(N号発生回路3は、MPU1およびMI
Ylaに供給するり11ツ・り信号を発生l2、MpH
2には直接に、MPUIIICItよ位相反転回路4を
介してそれぞれ供給する。したがってMI’l11のア
ドレスバス]1およびデータバス12に現われるアドレ
ス信号とデータ、MIUf;Aのアドレスバス21オよ
び22に現われるアドレス46月とデータは互いにクロ
ック信号の半周期だけずれてバス上に出□ 力すること
になる。アドレス切換回路5およびデータ切換回路9は
りrツク信号によってAll”Ulのアドレスバス11
およびデータバス12とMr′cI2、の7ド1/スバ
ス21およびデータバス22とを切換えて共通メモリ8
に供給するので、クロック信号の半分の周期ではMI’
l13がye bの半分の周期でit A(rYJBが
共通メモリ8をアクセスすることができる。このため、
片方のAIr′Uが共通メモリ8をアクセスする期間に
tユもう一方のAirUは休止期間となるので、MI’
Uを停止させる必要はたくなり、双方の1M)は互いに
独立して自由に共通メモリ8をアクセスすることができ
る。
The operation 1d of the information processing device of the multi-processor system shown in FIG. 1 is as follows. In Figure 3, Af
PU clock (N number generation circuit 3 is connected to MPU1 and MI
Generates a signal to be supplied to Yla, l2, MpH
2 is directly supplied to MPUIIICIt and via a phase inversion circuit 4, respectively. Therefore, the address signals and data appearing on the address bus 11 and data bus 12 of MIUf; address and data appearing on the address buses 21 and 22 of MIUf are shifted from each other by half a period of the clock signal on the bus. Output □ Output. The address switching circuit 5 and the data switching circuit 9 are connected to the address bus 11 of All"U1 by the r-check signal.
The common memory 8
Therefore, in half the period of the clock signal, MI'
It A (rYJB can access the common memory 8 in half the period of ye b. Therefore,
During the period when one AirU accesses the common memory 8, the other AirU is in a dormant period, so MI'
There is no longer a need to stop U, and both 1M) can freely access the common memory 8 independently of each other.

次に共通メモリ8のリフ1/ツシj71IIJ作r(つ
いて説明する。アト1/ス切換回路)はアドレス切換回
路5によって切換えられた、Afl’UliたはAir
U2からのアドレスξ、リンレッジ具アトtノス発生回
路6からのリフレッシュアドレスとを、共通メモリ選哲
検出回路10の検出出力に応じで、切換える。検出回路
10はアト1/スバス、’L、1.21 Jlのアドレ
ス信号が共通メモリ8に割病てたアドレスを指定するも
のであるか百かを判別する。
Next, the riff 1/switching circuit of the common memory 8 (this will be explained) is switched by the address switching circuit 5, Afl'Uli or Air.
The address ξ from U2 and the refresh address from the linkage device attnos generation circuit 6 are switched according to the detection output of the common memory selection detection circuit 10. The detection circuit 10 determines whether the address signals of AT1/subas, 'L, and 1.21 Jl designate an address that has been stored in the common memory 8 or not.

このためM2図のタイミング図忙示したようにA、fl
’U1.およびAfl’U2が共通メモリ8をアクセス
・ノる場合(第2図■、0に示したアドレスの場合)に
は共通メモリ8にはMr’U1または2のアドレスが供
給されて共通メモリ8を読み宵きすることができ、−そ
れ以外のMl’r11とMI’U2の両方ともが共通メ
モリ8をアクセスしない場合にはリフレ=ッシ^アドレ
ス発生回路6からのりフレツシコ。
Therefore, as shown in the timing diagram of M2 diagram, A, fl
'U1. And when Afl'U2 accesses the common memory 8 (in the case of the address shown in Figure 2, ■, 0), the address of Mr'U1 or 2 is supplied to the common memory 8, and the common memory 8 is - If neither Ml'r11 nor MI'U2 accesses the common memory 8, a refresh command is sent from the address generation circuit 6.

アドレスが共通メモリ8に供給されて、共通メモリをリ
フレッシュできる。リフレッシ5.アドレス発生回路6
において発生されるリフレッシュアト1/スはクロック
信号の周期よシ長い期間(例えば4倍〜16倍)、同じ
アドレスを出力するようにすれば、All’U1とMP
t12が続けて共通メモリをアクセスする場合でも、必
ずリフレッシエアドレスが共通メモリ8に化5給できる
ので、リフレッシツ、ミスは生じることはない。(通常
プロセッサが共通メモリを読み書きする命令が必要とす
る全マシンーリ・イクル数のうち、1マシン・す・イク
ルの期間だ&Jが共通メモリをアクセスするために使用
される。) 以上説明したように本実施例によれば2つのMI’Uが
互いにクロック信号の″1′周期ごと共通メモリをアク
セスできる構成となっているため、仙のAlr’lJを
制限することなく、@111.171.て1(通メモリ
をアクセスでき、共通メモリのリフレッシュも2つのA
、fl’Uのあき時間を利用12で有効に行なっている
だめ、それぞれのAfplJの稼QH効率を低下させる
ことνよない。
The address is provided to the common memory 8 so that the common memory can be refreshed. Refresh 5. Address generation circuit 6
If the refresh at1/s generated in 1/2 is made to output the same address for a period longer than the period of the clock signal (for example, 4 times to 16 times), All'U1 and MP
Even if t12 continues to access the common memory, the refresher address can always be transferred to the common memory 8, so refreshes and misses will not occur. (Usually, out of the total number of machine cycles that a processor requires for instructions to read and write common memory, 1 machine cycle is used to access common memory.) As explained above, According to this embodiment, since the two MI'Us are configured to be able to access the common memory every "1" period of the clock signal, @111.171. 1 (The common memory can be accessed, and the common memory can also be refreshed using two A
, fl'U's free time is effectively utilized 12, so that the operating QH efficiency of each AfplJ is not reduced.

外お本実施例ではA4PUを2つ用いた場合であるが、
All’lJの数置4が増えても、伺ら問題v、1、な
い。
In this example, two A4PUs are used.
Even if the number 4 in All'lJ increases, there is no problem v, 1.

以−)二述べたように本発明によるマルチプロセッサシ
ステム・の情報処理装置て!あJlば、名々のプロセッ
サをクロック信号の6ン相差ごとに、動作させることが
できるのでノ1ミ通メモリを各々のプロセッサが独立し
て自由にアクセスでき、またブロセッ°すによるアクセ
スのあき時間を利用して共通メモリのリフレッシュがで
き、プロセッサの稼動効率を低下させることがなくなる
As mentioned above, there is an information processing device for a multiprocessor system according to the present invention! In other words, since various processors can be operated according to six phase differences in clock signals, each processor can freely access the memory independently, and the access gap caused by processors can be reduced. The common memory can be refreshed using time, and the operating efficiency of the processor will not be reduced.

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

第1図は本発明の一実施例を・示すマルチプロセッサシ
ステムの情報処理装置のブロック図、第2図t」、プロ
ヒツサによる共通メモリのアクセ、−(JIJJ 間お
よび共通メモリのりフレフシ−0期間を示すタイミング
図である。 ]、2・・・AfrlJ 3・・・り1ツク信号発生回路 4・・・位相反転回路 へり・・・アドレス切換回路 6・・・リフレッシ−、アドレス切換回路8・・・共通
メモリ 9・・・データ切換回路 10・・・共通メモリ選択検出回路 代理人弁理士 薄 1)利、+九 第1図
FIG. 1 is a block diagram of an information processing device of a multiprocessor system showing an embodiment of the present invention. ], 2... AfrlJ 3... 1-turn signal generation circuit 4... Phase inversion circuit edge... Address switching circuit 6... Refresh, address switching circuit 8...・Common memory 9...Data switching circuit 10...Common memory selection detection circuit Attorney Susuki 1) Interest, +9 Figure 1

Claims (1)

【特許請求の範囲】[Claims] 複数の演算処理装置と該演算処理装置の各々によって共
通忙使用され石共通メモリを該共通メモリをリフレッシ
五するためのす7レツシエアドレスを発生するす7レツ
シβ、アドレス発生部とを有する情報処理装置において
、上記複数の演算処理装置のすべてを同期して動作させ
るためのりnツク信号を発生ずるクロック信号発生手段
と、該クロック信号の位相を変えてそれぞれの演舞処理
装*に供給する位相シフト手段と、演算処理装置のそれ
ぞれのシステムパスラフ1ツク信号の位相によって切換
えて共通メモリに供給するパス切換手段と、共通□メモ
リが演算処理装置によって、アクセスされているかどう
かを検出する共通メモリ選択検出手段と、共通メモリの
アクセス時は共通メモリを選択している演算処理装置か
らのアドレスを共通メモリに供給し、共通メモリの非ア
クセス時は上記リフレッシコ、アドレス発生部からのり
7レツシj、アドレスを共通メモリに供給するようにア
ドレス信号を切換え石アドレス切換手段とを設けること
を特徴とす石情報処理装置。
Information comprising a plurality of arithmetic processing units and an address generation unit that generates a address for refreshing a common memory commonly used by each of the arithmetic processing units. In the processing device, a clock signal generating means for generating a clock signal for operating all of the plurality of arithmetic processing devices in synchronization, and a phase changer for changing the phase of the clock signal and supplying it to each performance processing device*. a shift means, a path switching means that switches according to the phase of each system path rough signal of the arithmetic processing unit and supplies it to the common memory, and a common memory that detects whether the common memory is being accessed by the arithmetic processing unit. a selection detection means; when the common memory is accessed, the address from the arithmetic processing unit selecting the common memory is supplied to the common memory; when the common memory is not accessed, the above refresh code is sent from the address generator; 1. A stone information processing device comprising: a stone address switching means for switching an address signal so as to supply an address to a common memory.
JP57160974A 1982-09-17 1982-09-17 Information processor Pending JPS5952493A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP57160974A JPS5952493A (en) 1982-09-17 1982-09-17 Information processor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP57160974A JPS5952493A (en) 1982-09-17 1982-09-17 Information processor

Publications (1)

Publication Number Publication Date
JPS5952493A true JPS5952493A (en) 1984-03-27

Family

ID=15726171

Family Applications (1)

Application Number Title Priority Date Filing Date
JP57160974A Pending JPS5952493A (en) 1982-09-17 1982-09-17 Information processor

Country Status (1)

Country Link
JP (1) JPS5952493A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6262496A (en) * 1985-09-13 1987-03-19 Hitachi Ltd Processor selecting device

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5394745A (en) * 1977-01-31 1978-08-19 Copal Co Ltd Method of processing data

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5394745A (en) * 1977-01-31 1978-08-19 Copal Co Ltd Method of processing data

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6262496A (en) * 1985-09-13 1987-03-19 Hitachi Ltd Processor selecting device

Similar Documents

Publication Publication Date Title
KR940007690A (en) Processor system with synchronous dynamic memory
JPH03137757A (en) Priority control system
JPH0479026B2 (en)
JPS5952493A (en) Information processor
JPS61114362A (en) Access control system for share memory
JPH0628850A (en) Control circuit for dynamic ram
US7707450B1 (en) Time shared memory access
JP2617132B2 (en) Direct memory access method
JP2882831B2 (en) Bus interleave control method
JPH02129750A (en) Storage device
JPS58159292A (en) Memory refreshing method
JPH04372030A (en) Memory access system for processor
JPH06103026A (en) Memory system
JPS6243737A (en) Interruption control system
JPS6143741B2 (en)
JPS60151894A (en) Refresh circuit of dynamic ram
JPS59121455A (en) Prefixing system
JPS63201810A (en) Time system for information processing system
JPH04362594A (en) Memory backup controller
JPH0660645A (en) Power saving storage device
JPH0594363A (en) Storage device
JPH11176155A (en) Dram-refreshing circuit
JP2004102808A (en) Memory controller
JPS60124139A (en) Communication system of buffer memory
JPH03164962A (en) Memory control system