JPH0432415B2 - - Google Patents
Info
- Publication number
- JPH0432415B2 JPH0432415B2 JP56214134A JP21413481A JPH0432415B2 JP H0432415 B2 JPH0432415 B2 JP H0432415B2 JP 56214134 A JP56214134 A JP 56214134A JP 21413481 A JP21413481 A JP 21413481A JP H0432415 B2 JPH0432415 B2 JP H0432415B2
- Authority
- JP
- Japan
- Prior art keywords
- address
- address translation
- buffer
- instruction
- logical
- 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.)
- Expired - Lifetime
Links
- 238000006243 chemical reaction Methods 0.000 claims description 32
- 230000007246 mechanism Effects 0.000 claims description 21
- 230000010365 information processing Effects 0.000 claims description 11
- 230000004913 activation Effects 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000000034 method Methods 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
Landscapes
- Advance Control (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Description
【発明の詳細な説明】
本発明は、論理アドレス方式における論理アド
レスと実アドレスとの対を複数個有する変換バツ
フアをもつ情報処理装置に関する。DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an information processing apparatus having a conversion buffer having a plurality of pairs of logical addresses and real addresses in a logical addressing system.
高度の先行制御を行ない、キヤツシユメモリを
備え論理アドレス方式の論理アドレスと実アドレ
スの対を変換バツフアを持つ情報処理装置におい
ては、従来より命令語の先取り動作と命令の実行
動作を並列的に行ない命令の実行時間に命令語の
メモリ系からの読出しに費した時間がなるべく影
響を与えないような制御方式がとりいれられてい
る。 Information processing equipment that performs advanced advance control, is equipped with a cache memory, and has a buffer for converting pairs of logical addresses and real addresses in the logical addressing system, has conventionally performed the prefetching of instruction words and the execution of instructions in parallel. A control method is adopted in which the time spent reading the instruction word from the memory system has as little influence as possible on the execution time of the instruction.
すなわち、先行する分岐命令(以下先行命令と
称す。)の実行結果が判る前に分岐先予測命令
(以下後続命令と称す。)の命令語の読出し要求を
メモリ系に行なう場合が多い。 That is, in many cases, a request to read the instruction word of a branch destination prediction instruction (hereinafter referred to as a succeeding instruction) is made to the memory system before the execution result of a preceding branch instruction (hereinafter referred to as a preceding instruction) is known.
そのため、先行命令の実行結果の如何によつて
は、先行命令以降の後続命令の処理を無効とし、
先行命令に連続する命令から実行を再開するよう
なことが度々生じる。このような場合、後続命令
の読出し要求による論理アドレスに対する実アド
レスが変換バツフアに存在しないこともあり、こ
のときアドレス変換機構は実行とは無関係なアド
レス変換を行なうことを強いられ、先行制御部お
よび実行制御部より見たメモリ系を閉じてしま
い、後続命令の命令語の論理アドレスに対応する
実アドレスを作成する無駄な処理が実行時間を遅
らす要因となつている。 Therefore, depending on the execution result of the preceding instruction, the processing of subsequent instructions after the preceding instruction may be invalidated.
It often happens that execution is restarted from an instruction following a preceding instruction. In such a case, the real address corresponding to the logical address due to the read request of the subsequent instruction may not exist in the translation buffer, and the address translation mechanism is forced to perform address translation unrelated to execution. The memory system seen by the execution control unit is closed, and the unnecessary processing of creating a real address corresponding to the logical address of the instruction word of the subsequent instruction causes a delay in execution time.
本発明の目的は変換バツフアに要求論理アドレ
スに対応する実アドレスが存在しない場合に無駄
なアドレス変換の演算による実行時間の遅れを生
じさせないようにした情報処理装置を提供するこ
とにある。 SUMMARY OF THE INVENTION An object of the present invention is to provide an information processing apparatus that does not cause a delay in execution time due to unnecessary address translation operations when a real address corresponding to a requested logical address does not exist in the translation buffer.
前記目的を達成するために本発明による情報処
理装置は論理アドレスに対応した実アドレスを格
納するアドレス変換バツフアと、このアドレス変
換バツフアに変換アドレスが存在しないとき、そ
の論理アドレスを対応の実アドレスに変換するア
ドレス変換機構とを有する情報処理装置におい
て、先行制御を行なうための読出し要求に、分岐
命令の処理結果が確定していないことを示すフラ
グを付加するとともに、前記アドレス変換バツフ
アに前記要求の論理アドレスの実アドレスが存在
しない場合に前記フラグの状態により、分岐命令
の処理結果を確定する実行制御部からのアドレス
変換開始信号またはアドレス変換破棄信号を待ち
合せ、前記アドレス変換開始信号またはアドレス
変換破棄信号を受けて前記アドレス変換機構を制
御するアドレス変換指令解読回路を設け、前記読
出し要求が有効なときは前記アドレス変換機構を
起動し、無効なときは前記アドレス変換機構の起
動を抑止するように構成してある。 In order to achieve the above object, an information processing device according to the present invention includes an address translation buffer that stores a real address corresponding to a logical address, and when a translated address does not exist in this address translation buffer, converts the logical address into the corresponding real address. In an information processing device having an address translation mechanism that performs translation, a flag indicating that the processing result of a branch instruction is not determined is added to a read request for performing advance control, and a flag is added to the read request for performing advance control, and the address translation buffer When the real address of the logical address does not exist, depending on the state of the flag, wait for an address translation start signal or address translation discard signal from the execution control unit that determines the processing result of the branch instruction, and perform the address translation start signal or address translation discard signal. An address conversion command decoding circuit is provided for receiving a signal and controlling the address conversion mechanism, and activates the address conversion mechanism when the read request is valid, and inhibits activation of the address conversion mechanism when the read request is invalid. It is configured.
前記構成によれば無駄になるアドレス変換機構
の動作を取り除き、実行時間の短縮化が実現で
き、本発明の目的は完全に達成される。 According to the above configuration, wasted operations of the address translation mechanism can be eliminated, execution time can be shortened, and the object of the present invention can be completely achieved.
以下、図面を参照して本発明をさらに詳しく説
明する。 Hereinafter, the present invention will be explained in more detail with reference to the drawings.
第1図は本発明による情報処理装置の一実施例
を示すブロツク図である。多くの大形情報処理装
置では過去アクセスしたことのある論理アドレス
について、その時にアドレス変換により求めた実
アドレスを論理アドレスとの対でアドレス変換バ
ツフア21に格納しておきそれ以後、同一の論理
アドレスへのアクセスの際に、このアドレス変換
バツフア21より論理アドレスに対応する実アド
レスを引出し、アドレス変換による時間的遅れを
極力生じさせないような構成にしている。しかし
この変換バツフア21も制御方式および金物量の
制限などにより十分大きな容量を持つことは困難
であり、プログラムの性質上要求論理アドレスに
対応する実アドレスがアドレス変換バツフア21
に存在しない場合も多く、このときは先行制御部
3および実行制御部4より発行したデータ転送要
求アドレス変換のためのいくつかの情報をキヤツ
シユメモリ24または主記憶1よりデータバス1
21,211を介して読出し、多くの時間を費し
て実アドレスを求め、その実アドレスによりキヤ
ツシユメモリ24内または主記憶1内のデータを
読出し要求元にデータを転送しなければならな
い。したがつて不要なアドレス変換処理をさける
ことは性能向上に大いに寄与することになる。 FIG. 1 is a block diagram showing an embodiment of an information processing apparatus according to the present invention. In many large-scale information processing devices, for logical addresses that have been accessed in the past, the real address obtained by address conversion at that time is stored in the address conversion buffer 21 as a pair with the logical address. When accessing, the real address corresponding to the logical address is extracted from the address translation buffer 21, and the structure is designed to minimize time delay due to address translation. However, it is difficult for this conversion buffer 21 to have a sufficiently large capacity due to the control system and restrictions on the amount of hardware, and due to the nature of the program, the real address corresponding to the requested logical address is limited to the address conversion buffer 21.
In many cases, the information for converting the data transfer request address issued by the advance control unit 3 and the execution control unit 4 is transferred from the cache memory 24 or the main memory 1 to the data bus 1.
21, 211, spend a lot of time to find the real address, read the data in the cache memory 24 or the main memory 1 using the real address, and transfer the data to the source of the read request. Therefore, avoiding unnecessary address translation processing will greatly contribute to improving performance.
本発明はこの点に着目し、先行制御を行なつて
いるために実際は処理に使われるか否かが不確か
である。分岐命令に続く命令の読出し要求信号と
アドレス信号と、本要求の使用が不確かであるこ
とを指示するフラグ、すなわち変換指令待ち信号
を先行制御部3より、各信号線322,321,
323を介してバツフアメモリ制御回路26に対
して送出する。 The present invention focuses on this point, and since advance control is performed, it is uncertain whether or not it will actually be used for processing. The read request signal and address signal of the instruction following the branch instruction, and a flag indicating that the use of this request is uncertain, that is, a conversion command wait signal, are sent from the advance control unit 3 to each signal line 322, 321,
323 to the buffer memory control circuit 26.
次にバツフアメモリ制御回路26がアドレス変
換バツフア21に信号線251を介して必要な論
理アドレス部を送ると、それに対応する実アドレ
スが存在するかどうかの検出信号が信号線252
を介してアドレス変換バツフア21より報告され
る。これにより存在しないことが報告されるとバ
ツフアメモリ制御回路26は変換待ち信号323
が付加されてきたことを付して制御線261から
アドレス変換要求をアドレス変換指令解読回路2
7へ送出し、アドレス変換指令解読回路27はこ
のとき実行制御4からのアドレス変換開始または
破棄指令信号441を待ち合せる。 Next, when the buffer memory control circuit 26 sends a necessary logical address part to the address conversion buffer 21 via the signal line 251, a detection signal indicating whether a corresponding real address exists is sent to the signal line 251.
It is reported from the address translation buffer 21 via the address conversion buffer 21. When it is reported that the buffer memory does not exist, the buffer memory control circuit 26 sends a conversion wait signal 323.
The address conversion command decoding circuit 2 sends an address conversion request from the control line 261 with the information that the address conversion command has been added.
At this time, the address conversion command decoding circuit 27 waits for an address conversion start or discard command signal 441 from the execution control 4.
そしてアドレス変換開始が指令される(分岐不
成功の場合)と、アドレス変換機構28に対し信
号線271を介して変換指示を、一方、バツフア
メモリ制御回路26はアドレス線262を介して
論理アドレスを送る。アドレス変換機構28では
アドレス変換制御回路22の制御の下でアドレス
変換器25が起動され、変換に必要なテーブル情
報の索引が通常のデータ読出しと同様にデータ線
263を通してバツフアメモリ制御回路26との
間で行なわれる。 When the start of address conversion is commanded (in case of branch failure), a conversion instruction is sent to the address conversion mechanism 28 via the signal line 271, while the buffer memory control circuit 26 sends the logical address via the address line 262. . In the address conversion mechanism 28, the address converter 25 is activated under the control of the address conversion control circuit 22, and the index of the table information necessary for conversion is transferred to the buffer memory control circuit 26 through the data line 263 in the same way as in normal data reading. It will be held in
またアドレス変換破棄が指令される(分岐成功
の場合)と、バツフアメモリ制御回路26のアド
レス変換要求を無視し、アドレス変換機構28へ
の動作指示を破棄する。このようにして不要なア
ドレス変換機構の起動を抑止することにより、ア
ドレス変換バツフアから有効なアドレス変換デー
タを追出すのを防ぐとともに、後続の有効な読出
し要求に対するメモリバツフア制御部の早期使用
を可能にする。 Further, when a command to discard address translation is issued (in case of a successful branch), the address translation request from the buffer memory control circuit 26 is ignored and the operation instruction to the address translation mechanism 28 is discarded. In this way, by suppressing the activation of unnecessary address translation mechanisms, it is possible to prevent valid address translation data from being evicted from the address translation buffer, and to enable early use of the memory buffer control unit for subsequent valid read requests. do.
本実施例では分岐命令に続く命令の予測先取り
の場合の例を説明したが、同様にして分岐先命令
の予測先取りをした場合に、分岐の不成功を検出
してアドレス変換機構の起動を抑止するときの不
要条件検出時のアドレス変換機構の起動を抑止す
る装置も可能である。 In this embodiment, an example of predictive pre-fetching of the instruction following a branch instruction has been explained, but if a branch destination instruction is similarly predictively pre-fetched, failure of the branch will be detected and activation of the address translation mechanism will be inhibited. It is also possible to provide a device that suppresses activation of the address translation mechanism when an unnecessary condition is detected.
以上の構成も含めて本発明は特許請求の範囲の
すべてに及ぶものである。 The present invention, including the above configuration, extends to the entire scope of the claims.
本発明は以上詳しく説明したように、データ読
出し要求に対してその論理アドレスの変換アドレ
スが存在しない場合アドレス変換指令待ち指示に
よりアドレス変換機構の起動を制御することによ
り情報処理装置の実行時間の高速化を計れる効果
がある。 As described in detail above, the present invention speeds up the execution time of an information processing device by controlling activation of an address translation mechanism by an address translation command wait instruction when there is no translation address for a logical address in response to a data read request. It has the effect of measuring the
第1図は本発明による情報処理装置の一実施例
の回路構成を示すブロツク図である。
1…主記憶、2…バツフアメモリ制御部、3…
先行制御部、4…実行制御部、21…アドレス変
換制御部、22…アドレス変換制御回路、25…
アドレス変換器、26…バツフアメモリ制御回
路、27…アドレス変換指令解読回路、28…ア
ドレス変換機構、24…キヤツシユメモリ、12
1…主記憶読出しデータパス、211…主記憶ア
ドレス及び主記憶書込みデータパス、231…命
令語及びオペランド語パス、241…オペランド
語パス、261…アドレス変換指示待ちアドレス
変換機構による変換要求信号線、271…アドレ
ス変換機構起動要求信号線、321…メモリ読
出、書込要求データパス、323…アドレス変換
指示待ち信号線、341…命令供給パス、431
…オペランド語要求パス、441…アドレス変換
機構起動開始、抑止指示信号線。
FIG. 1 is a block diagram showing a circuit configuration of an embodiment of an information processing apparatus according to the present invention. 1... Main memory, 2... Buffer memory control section, 3...
Advance control unit, 4... Execution control unit, 21... Address translation control unit, 22... Address translation control circuit, 25...
Address converter, 26... Buffer memory control circuit, 27... Address conversion command decoding circuit, 28... Address conversion mechanism, 24... Cache memory, 12
1...Main memory read data path, 211...Main memory address and main memory write data path, 231...Instruction word and operand word path, 241...Operand word path, 261...Conversion request signal line by address conversion mechanism waiting for address conversion instruction, 271...Address conversion mechanism activation request signal line, 321...Memory read/write request data path, 323...Address conversion instruction wait signal line, 341...Instruction supply path, 431
...Operand word request path, 441...Address translation mechanism activation start, inhibition instruction signal line.
Claims (1)
るアドレス変換バツフアと、このアドレス変換バ
ツフアに変換アドレスが存在しないとき、その論
理アドレスを対応の実アドレスに変換するアドレ
ス変換機構とを有する情報処理装置において、先
行制御を行うための読出し要求に、分岐命令の処
理結果が確定していないことを示すフラグを付加
するとともに、前記アドレス変換バツフアに前記
要求の論理アドレスの実アドレスが存在しない場
合に前記フラグの状態により、分岐命令の処理結
果を確定する実行制御部からのアドレス変換開始
信号またはアドレス変換破棄信号を待ち合わせ、
前記アドレス変換開始信号またはアドレス変換破
棄信号を受けて前記アドレス変換機構を制御する
アドレス変換指令解読回路を設け、前記読出し要
求が有効なときは前記アドレス変換機構を起動
し、無効なときは前記アドレス変換機構の起動を
抑止するように構成したことを特徴とする情報処
理装置。1. In an information processing device having an address translation buffer that stores a real address corresponding to a logical address, and an address translation mechanism that translates the logical address into the corresponding real address when the translated address does not exist in this address translation buffer, A flag indicating that the processing result of a branch instruction is not determined is added to a read request for advance control, and the flag is added when the real address of the logical address of the request does not exist in the address translation buffer. Depending on the state, it waits for an address translation start signal or an address translation discard signal from the execution control unit that determines the processing result of the branch instruction.
An address translation command decoding circuit is provided to control the address translation mechanism in response to the address translation start signal or the address translation discard signal, and activates the address translation mechanism when the read request is valid and reads the address when the read request is invalid. An information processing device characterized by being configured to suppress activation of a conversion mechanism.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP56214134A JPS58115680A (en) | 1981-12-28 | 1981-12-28 | Information processor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP56214134A JPS58115680A (en) | 1981-12-28 | 1981-12-28 | Information processor |
Publications (2)
Publication Number | Publication Date |
---|---|
JPS58115680A JPS58115680A (en) | 1983-07-09 |
JPH0432415B2 true JPH0432415B2 (en) | 1992-05-29 |
Family
ID=16650785
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP56214134A Granted JPS58115680A (en) | 1981-12-28 | 1981-12-28 | Information processor |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPS58115680A (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0827718B2 (en) * | 1983-10-05 | 1996-03-21 | 株式会社日立製作所 | Information processing device |
JPS63197232A (en) * | 1987-02-12 | 1988-08-16 | Toshiba Corp | Microprocessor |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS56134383A (en) * | 1980-03-24 | 1981-10-21 | Fujitsu Ltd | Data processor |
-
1981
- 1981-12-28 JP JP56214134A patent/JPS58115680A/en active Granted
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS56134383A (en) * | 1980-03-24 | 1981-10-21 | Fujitsu Ltd | Data processor |
Also Published As
Publication number | Publication date |
---|---|
JPS58115680A (en) | 1983-07-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5265236A (en) | Method and apparatus for increasing the speed of memory access in a virtual memory system having fast page mode | |
JPH0137773B2 (en) | ||
GB2242294A (en) | Memory architecture using page mode writes and single level write buffering | |
JPH0564813B2 (en) | ||
US6044447A (en) | Method and apparatus for communicating translation command information in a multithreaded environment | |
JPH0432415B2 (en) | ||
EP0488566A2 (en) | Method and apparatus for fast page mode selection | |
JPS61217834A (en) | Data processor | |
JP3013996B2 (en) | Information processing device | |
JP2845754B2 (en) | Multiprocessor system | |
JP2000347941A (en) | Cache memory device | |
JPH0210450A (en) | Prefetch control system for cache memory | |
JP3171289B2 (en) | Information processing device | |
JPH041373B2 (en) | ||
JPH09212465A (en) | Memory allocation device | |
JPS63214852A (en) | Address converting system | |
JP2917915B2 (en) | Multiprocessor system | |
JPH057740B2 (en) | ||
JPH03127146A (en) | Information processor | |
JPS61221845A (en) | Producing system for invalidated address of buffer memory | |
JPH05289940A (en) | Cache memory control system | |
JPH01230152A (en) | Cache memory | |
JPS6222165A (en) | Control system for access to main storage device | |
JPH0713865A (en) | Cache memory control system | |
JPH0228737A (en) | Buffer storage control system |