JPS5819785A - Memory access controlling system - Google Patents

Memory access controlling system

Info

Publication number
JPS5819785A
JPS5819785A JP56119700A JP11970081A JPS5819785A JP S5819785 A JPS5819785 A JP S5819785A JP 56119700 A JP56119700 A JP 56119700A JP 11970081 A JP11970081 A JP 11970081A JP S5819785 A JPS5819785 A JP S5819785A
Authority
JP
Japan
Prior art keywords
address
data
signal
memory
cache memory
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
JP56119700A
Other languages
Japanese (ja)
Other versions
JPH0250498B2 (en
Inventor
Masaaki Kobayashi
正明 小林
Shigeru Hashimoto
繁 橋本
Takumi Kishino
琢己 岸野
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.)
Fujitsu Ltd
Original Assignee
Fujitsu 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 Fujitsu Ltd filed Critical Fujitsu Ltd
Priority to JP56119700A priority Critical patent/JPS5819785A/en
Publication of JPS5819785A publication Critical patent/JPS5819785A/en
Publication of JPH0250498B2 publication Critical patent/JPH0250498B2/ja
Granted legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/0802Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
    • G06F12/0862Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches with prefetch

Abstract

PURPOSE:To reduce the access time from a processor to cache memory and to increase the processing efficiency for a memory access controlling system of the processor, etc., by fetching in advance the address counter of the cache memory. CONSTITUTION:A main storage device 18 and a cache memory 3 are connected to a processor 6. An adder memory 2 stores the addition given to an address counter 1 and delivers an added signal G. The memory 3 comprises an address part (a) and a data part (d) and, to the beginning, and accessed by a lower level address a2 of the counter 1. The read-out address data a1 is compared 5 with an upper level address a1, and a hit signal H1 obtained upon attaining a comparison agreement reads out the data d1 to transfer it to the processor 6. A discriminating part 17 compares the signal H1 with the signal G, and an output signal B is fed to a comparing part 4. The addresses a1 and a2 are compared with an address data A coming next which and the data d1 of the memory 3 is read out by the signal H2 upon attaining comparison agreement to be fed to the processor 6.

Description

【発明の詳細な説明】 本発明はプロセサへ下におけるメモリへのアクセスを制
御するメモリアクセス制御方式に門する。
DETAILED DESCRIPTION OF THE INVENTION The present invention is directed to a memory access control scheme for controlling access to underlying memory to a processor.

キャシュメモリが設けられた処理装置においては、主記
憶の一部の情報は前記キャシュメモリに蓄えられている
。このキャシュメモリには高速の集積回路(IC)が用
いられるので、処理速度が著しく向上する。一般にキャ
シュメモリのアクセス時間は、主記憶のアクセス時間よ
92倍以上速い。このため主記憶の1サイクルタイム内
に、キャシュメモリへは少くも2回アクセスが可能であ
る。従ってキャシュメモリ制御部は、処理装置から第1
のアドレスデータA、を受けて制御を行ったのち、直ち
にキャシュメそりのアドレスカウンタをJtj ltl
、させ、その値をA1+1としておく。次に処理装置か
ら第2のアドレスデータA!を受けた際、この人、と前
記アドレスカウンタの値(A1+1)との比較1判別を
行い、もし人=(A1+1)であれば、キャシュメモリ
から直ちにデータの読出しが可能となる。すなわちメモ
リアクセスの先行詞’l1il+を行うことができる。
In a processing device equipped with a cache memory, part of the information in the main memory is stored in the cache memory. Since this cache memory uses a high-speed integrated circuit (IC), processing speed is significantly improved. Generally, the access time of cache memory is 92 times faster than the access time of main memory. Therefore, the cache memory can be accessed at least twice within one cycle time of the main memory. Therefore, the cache memory control unit receives the first
After receiving and controlling the address data A, the address counter of the cash register is immediately set to
, and set its value to A1+1. Next, the processing device sends second address data A! When received, this person is compared with the value (A1+1) of the address counter to make a determination, and if person=(A1+1), data can be immediately read from the cache memory. In other words, the antecedent 'l1il+ of memory access can be performed.

本発明は上記の点に着目したものであり、処理装置の処
理効率を向上するメモリアクセスflilJ O1l方
式の提供を目的とする。
The present invention has focused on the above points, and aims to provide a memory access flilJO11 method that improves the processing efficiency of a processing device.

本発明は、主記憶装置と、キャシュメモリと、前記キャ
シュメモリのアドレスを定めるアドレスカウンタと、処
理装置とで構成され、第1の時点に設定された前記アド
レスカウンタの第1アドレスと前記キャシュメモリから
読出された第2のアドレスとを比較する第1の比較手段
と、前記第1のアドレスと第2の時点に前記処理装置か
ら発せられる第3のアドレスとを比較する第2の比較手
段と、前記アドレスカウンタをプラス1だけ歩進せしめ
る手段と、該歩進法みの信号と前記第1の比較手段の出
力信号とを判別する手段とを備え、前記第1のアドレス
と第2のアドレスが一致したとき且つ前記第1のアドレ
スと前記第3のアドレスとが一致したとき該第1のアド
レスで定まるキャシュメモリ内のデータを前記処理装置
へ転送せしめることを特徴とするメモリ制御方式である
The present invention comprises a main storage device, a cache memory, an address counter that determines an address of the cache memory, and a processing device, and the first address of the address counter set at a first time and the cache memory a first comparing means for comparing the first address with a second address read from the processing device; and a second comparing means for comparing the first address and a third address issued from the processing device at a second time point. , comprising means for incrementing the address counter by +1, and means for discriminating between the incrementing signal and the output signal of the first comparing means, the first address and the second address and when the first address and the third address match, the data in the cache memory determined by the first address is transferred to the processing device. .

以下、本発明を図面によって説明する。第1図は本発明
の一実施例を説明するブロック図、第20は本発明の一
実施例をπI(1明するフローチャートであり、1はア
ドレスカウンタ、2は加算メモリ、3はキャシュメモリ
、4,5は比較部、6は処理装置、7,13はOR回路
、8,9.12はM山口路、10はセットパルス発生部
、11はアドレス発生部、14は読取指令発生部、15
はレジスタ部、16は処J・U部、17は判別部、18
は上記t(検装置j:i 、 A I−jアドレスデー
タ、Bは出力信号、Cは読取指令、E、Fは不一致信号
、Gは加算済み信号、II、  、H,はヒツト信号、
Sはセット信号、aはアドレス部、alは上位アドレス
、afはアドレスデータ、a2’!:l:下位アドレス
、dはデータ部である。第1図における加nメモリ2は
、アドレスカウンタが加算(+1)されたことを記憶す
るメモリであり、加算済み信号Gは、その加算済み状態
を示す信号である。壕だキャシュメモリ3はアドレス部
aとデータ部dとで構成され、アドレス部aには、アド
レスデータAの上位バイトが格 3− 悄されている。第1図において、最初C第1の時点)に
、アドレスカウンタ内の下位アドレスa2によシキャシ
瓢メモリ3にアクセスし、アドレス部Rからアドレスデ
ータa/、を読出し、このal、と前記アドレスカウン
タの上位アドレスa、との比較を、比較部5で行う。上
述のようにアドレスa1はアドレスデータAの上位バイ
トであるので、alとal、  の両者が一致した場合
には比較部5からヒツト信号H1を発する。このヒツト
信号H,は、キャシュメモリ3のアドレスa2のデータ
部dより、データd1を読出し、これを処理装置6へ転
送する。データd1はAND回路9.OR回路13を経
てレジスタ部15にセットされる。次に判別部17はヒ
ツト信号H,と、加算済み信号Gとを判別する。すなわ
ち比較部5からヒツト信号H8が出力され、しかもアド
レスカウンタ1が加算(+1)済みであれば、判別部1
7から出力信号Bを出力する。この出力信号Bによシ、
比較部4において、アドレスカウンタ1のアドレス(a
X  及4− トさ7するアドレスデータAとの比較を行う。この両者
が一致したとき、ヒツト信号■I2が発せられ、とのた
めキャシュメモリ内のデータd1が取出されて、り1理
装置6へ転送される。データd、はAND回路8、OR
回路13を経て、レジスタ部15にセットされる。換言
す牡ば、処理装置6からのアドレスデータ四をアドレス
カウンタ1にセットして、キャシュメモリ3からデータ
を読出したのち、該アドレスカウンタ1を歩進(プラス
1)しておく。次に処理装置6から発せられるアドレス
データ(At)とアドレスデータ(At)とが一致した
のでちるから、直ちにキャシュメモリ3からデータの読
出しを行うものである。
Hereinafter, the present invention will be explained with reference to the drawings. FIG. 1 is a block diagram illustrating an embodiment of the present invention, and 20th is a flowchart illustrating an embodiment of the present invention, in which 1 is an address counter, 2 is an addition memory, 3 is a cache memory, 4 and 5 are comparison sections, 6 is a processing device, 7 and 13 are OR circuits, 8, 9.12 are M Yamaguchi paths, 10 is a set pulse generation section, 11 is an address generation section, 14 is a read command generation section, 15
is a register section, 16 is a processing J/U section, 17 is a discrimination section, 18
is the above t (inspection device j: i, A I-j address data, B is the output signal, C is the reading command, E, F is the mismatch signal, G is the added signal, II, , H, is the hit signal,
S is a set signal, a is an address part, al is an upper address, af is address data, a2'! :l: Lower address, d is the data part. The addition memory 2 in FIG. 1 is a memory that stores the fact that the address counter has been added (+1), and the addition completed signal G is a signal indicating the addition completed state. The deep cache memory 3 is composed of an address section a and a data section d, and the upper byte of address data A is stored in the address section a. In FIG. 1, at the first time C1), the memory 3 is accessed using the lower address a2 in the address counter, address data a/, is read from the address section R, and this al and the address counter The comparator 5 compares the address with the upper address a. As mentioned above, since the address a1 is the upper byte of the address data A, when both al and al match, the comparing section 5 generates the hit signal H1. This hit signal H, reads data d1 from the data section d of address a2 of the cache memory 3, and transfers it to the processing device 6. The data d1 is the AND circuit 9. It is set in the register section 15 via the OR circuit 13. Next, the determining section 17 determines between the hit signal H and the added signal G. That is, if the hit signal H8 is output from the comparator 5 and the address counter 1 has already been added (+1), the discriminator 1
Output signal B is output from 7. According to this output signal B,
In the comparator 4, the address (a
A comparison is made with the address data A to be compared. When the two match, a hit signal I2 is generated, and the data d1 in the cache memory is taken out and transferred to the controller 6. Data d is AND circuit 8, OR
It is set in the register section 15 via the circuit 13. In other words, address data 4 from the processing device 6 is set in the address counter 1, the data is read from the cache memory 3, and then the address counter 1 is incremented (plus 1). Next, since the address data (At) issued from the processing device 6 and the address data (At) match, data is immediately read from the cache memory 3.

比較部5において、不一致信号Eが発せられた場合には
、この不一致信号EはOR回路7を経てセットパルス発
生部をトリガーする。発生されたセットパルスSはアド
レスカウンタ1に達し、処理装置6からのアドレスデー
タAをセットする。
When the comparison section 5 generates a mismatch signal E, the mismatch signal E passes through the OR circuit 7 and triggers the set pulse generation section. The generated set pulse S reaches the address counter 1 and sets the address data A from the processing device 6.

一方、不一致信号は処理装置6へ送られ、読取指ら牡、
主記憶装置18のデータd2はAND回路12、OR,
回路13を経て、レジスタ部15にセットされる。
On the other hand, the mismatch signal is sent to the processing device 6,
The data d2 in the main memory device 18 is input to the AND circuit 12, OR,
It is set in the register section 15 via the circuit 13.

比較部4において、不一致信号Fが発せられた場合、こ
の信号はOR回路7を経てセットパルス発生部をトリガ
ーするので、発生されたセットパルスSはアドレスカウ
ンタ1(で達し、アドレスデータAをセットする。一方
、不一ν(信号Fは比較部5に達して上位アドレスa1
とアドレスデータa1との比較を行う。この比較動作は
既述と同じである。第2図は以上に説明した動作順序を
示すフローチャートである。
When a mismatch signal F is generated in the comparator 4, this signal passes through the OR circuit 7 and triggers the set pulse generator, so the generated set pulse S reaches the address counter 1 and sets the address data A. On the other hand, the difference ν (signal F reaches the comparator 5 and the upper address a1
and the address data a1 are compared. This comparison operation is the same as described above. FIG. 2 is a flowchart showing the sequence of operations described above.

以上のように本発明は、キャシュメモリの高速性を生か
し、キャシュメモリのアドレスカウンタを予め歩進させ
ておくことにより、処理装置からのキャシュメモリへの
アクセス時間を短縮せしめたものであシ、処理効率を著
しく向上しうる利点を有する。
As described above, the present invention takes advantage of the high speed of cache memory and increments the address counter of the cache memory in advance to shorten the time required for accessing the cache memory from the processing device. It has the advantage of significantly improving processing efficiency.

【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説明するブロック図、ff
+’++ 2図は不発明の一笑施例を説明するフローチ
ャートであシ、図中に用いた符号は次の通である0 1ζ−1ニアドレスカウンタ 2は力117:メモリ、
3はキャシュメモリ、4,5は比較部、6は処理装仕孕
、7.13はOR回路、8,9.12はAND回路、1
0はセットパルス発生部、11はアドレス発生部、14
は読取指令発生部、15はレジスタ部、16は処理部、
17は判別部、18は主記憶装着、Aはアドレスデータ
、Bは出力信号、ci、[I’W取指全指令、Fは不一
致信号、Gは力峙γ済み信号、Hl。 ■■2はヒツト信号、Sはセット信号、aけアドレス部
、aIは上位アドレス、a′、はアドレスデータ、a、
は下位アドレス、dはデータ部を示す。
FIG. 1 is a block diagram explaining one embodiment of the present invention, ff
+'++ Figure 2 is a flowchart explaining an embodiment of the uninvented joke, and the symbols used in the figure are as follows: 0 1ζ-1 Near address counter 2: Force 117: Memory;
3 is a cache memory, 4 and 5 are comparison sections, 6 is a processing device, 7.13 is an OR circuit, 8, 9.12 is an AND circuit, 1
0 is a set pulse generator, 11 is an address generator, 14
15 is a register unit, 16 is a processing unit,
17 is a determination unit, 18 is a main memory installed, A is address data, B is an output signal, ci, [I'W take all commands, F is a mismatch signal, G is a force γ completed signal, Hl. ■■2 is a hit signal, S is a set signal, a digit address part, aI is an upper address, a' is address data, a,
indicates a lower address, and d indicates a data portion.

Claims (1)

【特許請求の範囲】[Claims] 主記憶装置と、キャシュメモリと、前記キャシュメモリ
のアドレスを定めるアドレスカウンタと、処理装置とで
構成され、第1の時点に設定された前記アドレスカウン
タの第1アドレスと前記キャシュメモリから読出された
第2のアドレスとを比較する第1の比較手段と、前記第
1のアドレスと第2の時点に前記処理装置から発せられ
る第3のアドレスとを比較する第2の比較手段と、前記
アドレスカウンタをプラス1だけ歩進せしめる手段と、
該歩進済みの信号と前記第1の比較手段の出力信号とを
判別する手段とを備え、前記第1のアドレスと第2のア
ドレスが一致したとき又は前記第1のアドレスと前記第
3のアドレスとが一致したとき該第1のアドレスで定ま
るキヤメモリ内のデータを前記処理装置へ転送せしめる
ことを特徴とするメモリ制御方式。
The device includes a main storage device, a cache memory, an address counter that determines the address of the cache memory, and a processing device, and the first address of the address counter set at a first time point and the first address read from the cache memory are a second comparing means for comparing the first address with a third address issued from the processing device at a second time; A means to advance by +1,
means for discriminating between the incremented signal and the output signal of the first comparing means, and when the first address and the second address match, or when the first address and the third A memory control method characterized in that when the first address matches the first address, data in the cache memory determined by the first address is transferred to the processing device.
JP56119700A 1981-07-30 1981-07-30 Memory access controlling system Granted JPS5819785A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP56119700A JPS5819785A (en) 1981-07-30 1981-07-30 Memory access controlling system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP56119700A JPS5819785A (en) 1981-07-30 1981-07-30 Memory access controlling system

Publications (2)

Publication Number Publication Date
JPS5819785A true JPS5819785A (en) 1983-02-04
JPH0250498B2 JPH0250498B2 (en) 1990-11-02

Family

ID=14767905

Family Applications (1)

Application Number Title Priority Date Filing Date
JP56119700A Granted JPS5819785A (en) 1981-07-30 1981-07-30 Memory access controlling system

Country Status (1)

Country Link
JP (1) JPS5819785A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS61258063A (en) * 1985-05-07 1986-11-15 日本染色機械株式会社 Apparatus for treating long fiber product

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5099046A (en) * 1973-12-28 1975-08-06
JPS5265628A (en) * 1975-11-28 1977-05-31 Hitachi Ltd Information processing device
JPS53134335A (en) * 1977-04-28 1978-11-22 Fujitsu Ltd Memory control system
JPS5680871A (en) * 1979-12-06 1981-07-02 Fujitsu Ltd Buffer memory control system

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5099046A (en) * 1973-12-28 1975-08-06
JPS5265628A (en) * 1975-11-28 1977-05-31 Hitachi Ltd Information processing device
JPS53134335A (en) * 1977-04-28 1978-11-22 Fujitsu Ltd Memory control system
JPS5680871A (en) * 1979-12-06 1981-07-02 Fujitsu Ltd Buffer memory control system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS61258063A (en) * 1985-05-07 1986-11-15 日本染色機械株式会社 Apparatus for treating long fiber product

Also Published As

Publication number Publication date
JPH0250498B2 (en) 1990-11-02

Similar Documents

Publication Publication Date Title
EP0432575A2 (en) Data processor having wait state control unit
US4757439A (en) Memory bus architecture
JPS5819785A (en) Memory access controlling system
US5388229A (en) Data transmission system which prefetches data to be transferred
KR920010977B1 (en) Memory bus architecture
JPH0778106A (en) Data processing system
JP2581484B2 (en) Data processing system
JP2003051190A (en) Semiconductor memory device, device and method for setting burst mode to the memory device
JP2805786B2 (en) Information processing device
JPS58220567A (en) Picture data compressing system
JP2906440B2 (en) Memory access method for multiplex transmission equipment
JPH01193961A (en) Address converting device
KR980011450A (en) High-speed memory device
JPH0721083A (en) Memory reader
JP2684818B2 (en) Distance value calculation / extraction circuit in character recognition device
JPS59140685A (en) Data processor having hierarchical cache memory
JPS58168133A (en) Converting system of channel dynamic address
JP2000059283A (en) Frame storage device and method
JPH04361343A (en) Parity check circuit
JPH05298179A (en) Memory control system
JPH04239942A (en) Address conversion buffer device
JPH11184751A (en) Memory control circuit and memory device
JPH03237537A (en) Memory read system
JPH02250135A (en) Control method for cache memory system and timing control circuit for its method
JPH0352066A (en) Block data nullification processing system