[go: up one dir, main page]

TW583583B - Apparatus and method for selective control of condition code write back - Google Patents

Apparatus and method for selective control of condition code write back Download PDF

Info

Publication number
TW583583B
TW583583B TW091116956A TW91116956A TW583583B TW 583583 B TW583583 B TW 583583B TW 091116956 A TW091116956 A TW 091116956A TW 91116956 A TW91116956 A TW 91116956A TW 583583 B TW583583 B TW 583583B
Authority
TW
Taiwan
Prior art keywords
extended
instruction
item
code
microprocessor
Prior art date
Application number
TW091116956A
Other languages
Chinese (zh)
Inventor
G Glenn Henry
Rodney E Hooker
Terry Parks
Original Assignee
Ip First Llc
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
Priority claimed from US10/144,593 external-priority patent/US7185180B2/en
Application filed by Ip First Llc filed Critical Ip First Llc
Application granted granted Critical
Publication of TW583583B publication Critical patent/TW583583B/en

Links

Landscapes

  • Executing Machine-Instructions (AREA)

Abstract

A microprocessor apparatus and method are provided, for selectively controlling write back of condition codes. The microprocessor apparatus has translation logic and extended execution logic. The translation logic translates an extended instruction into corresponding micro instructions. The extended instruction includes an extended prefix and an extended prefix tag. The extended prefix disables write back of the condition codes, where the condition codes correspond to a result of a prescribed operation. The extended prefix tag indicates the extended prefix, where the extended prefix tag is an otherwise architecturally specified opcode within an instruction set for a microprocessor. The extended execution logic is coupled to the translation logic. The extended execution logic receives the corresponding micro instructions, and generates the result, and disables write back of the condition codes.

Description

583583583583

-J · A7 B7_ 五、發明說明(/) ~ ^ 與相關申請案之對照 [0001]本申睛案依據以下美國中請案主張優 號10/144,593,申請曰為2〇〇2年5月9曰,專利名稱為「選 擇性地控制條件碼回寫之裝置及方法」。 、 [〇〇〇2]本帽案與下制在t射之美國專利申請案 有關’其申請曰與本案相同,且具有相同的申請人與^明 人0 TW SERIAL NUMBER DOCKET NUMBER 專利名稱 91116957 CNTR:2176 爸使微處理器指令集&奘置及方法 91116958 CNTR:2186 執行條件指令之裝置及方法 91116959 91116672 CNTR:2189 CNTR:2198 沒土聲處理器之暫查的機制 選擇性地控制結果回寫之裝置及方 法 (一) 發明技術領域: [0003] 本發明係有關微電子的領域,尤指一種能將選 擇性的條件碼回寫控制特徵納入一既有之微處理器指令集 架構的技術。 (二) 發明技術背景: [0004] 自1970年代初發韌以來,微處理器之使用即呈 2 本紙張尺度適用中國國家標準(CNS)A4規格(210 X 2·97公复) IV-----^--- (請先閱讀背面之注意事項再填寫本頁) 經濟部智慧財產局員工消費合作社印製 583583 經濟部智慧財產局員工消費合作社印製 A7 五、發明說明(> ) 指數般成長。從最早應用於科學與技術的領域,到如今已 從那些特殊領域引進商業的消費者領域,如桌上型與膝上 型(laptop)電腦、視訊遊戲控制器以及許多其他常見的家 用與商用裝置等產品。 [0005] 隨著過去三十年來使用上的爆炸性成長,在技 術上也歷經一相對應之提昇,其特徵在於對下列項目有著 曰益昇高之要求:更快的速度、更強的定址能力、更快的 圮憶體存取、更大的運算元、更多種運算(如浮點運算、 單一指令多重資料(SIMD)、條件移動等)以及附加的特 殊運算(如多媒體運算)。如此造就了該領域中驚人的技 術進展,且都已應用於微處理器之設計,像擴充管線化 (extensive pipelining )、超純量架構(犯阿髮hr architecture )、快取結構、亂序處理(祕〇f-〇rder pr〇cessmg )、 爆發式存取(burst access )、分支預測(branch predicati〇n) 以及假想執行(speculative execution)。直言之,比起30 年剷剛出現時,現在的微處理器呈現出驚人的複雜度,且 具備了強大的能力。 [0006] 但與許多其他產品不同的是,有另一非常重要 的因素已限制了,並持續限制著微處理器架構之演進。現 今微處理器會如此複雜,一大部分得歸因於這項因素,即 舊有軟體之相容性。在市場考量下,所多製造商選擇將新 的架構特徵納入最新的微處理器設計中,但同時在這些最 7的產ΠΠ中,又保留了所有為確保相容於較舊的、即所謂 舊有」(legacy)應用程式所必需之能力。 -----V---V-----裝--- (請先閱讀背面之注意事項再填寫本頁) 訂:-J · A7 B7_ V. Description of the invention (/) ~ ^ Contrast with related applications [0001] This application claims the superior number 10 / 144,593 based on the following US applications, and the application is May 2002 On the 9th, the patent was entitled "Apparatus and Method for Selectively Controlling Condition Code Writeback". [00〇2] This hat case is related to the U.S. patent application filed below: 'The application is the same as this case, and has the same applicant and ^ Mingren 0 TW SERIAL NUMBER DOCKET NUMBER Patent name 91116957 CNTR : 2176 Dad's microprocessor instruction set & setup and method 91116958 CNTR: 2186 Device and method for executing conditional instructions 91116959 91116672 CNTR: 2189 CNTR: 2198 The mechanism of the temporary check of the unvoiced processor selectively controls the result back Device and method for writing (1) Technical field of the invention: [0003] The present invention relates to the field of microelectronics, in particular to a device capable of incorporating selective condition code write-back control features into an existing microprocessor instruction set architecture. technology. (II) Technical background of the invention: [0004] Since its development in the early 1970s, the use of microprocessors has been 2 paper sizes applicable to the Chinese National Standard (CNS) A4 specification (210 X 2.97 public copy) IV-- --- ^ --- (Please read the precautions on the back before filling out this page) Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 583583 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs A7 5. Description of the invention (>) Exponential growth. From the earliest applications in science and technology, to today's consumer domains that have introduced commerce from special areas such as desktop and laptop computers, video game controllers, and many other common home and business devices And other products. [0005] With the explosive growth in use over the past thirty years, it has also undergone a corresponding improvement in technology, which is characterized by the following requirements for increased benefits: faster speed, stronger addressing ability , Faster memory access, larger operands, more operations (such as floating-point operations, single instruction multiple data (SIMD), conditional movement, etc.), and additional special operations (such as multimedia operations). This has created amazing technological progress in this field, and has been applied to the design of microprocessors, such as extensive pipelining, ultra-scalar architecture, cache structure, and out-of-order processing. (Of-〇rder prcesscess), burst access, branch predication, and speculative execution. To put it bluntly, today's microprocessors are surprisingly more complex and powerful than they were when the shovel first appeared in 30 years. [0006] But unlike many other products, another very important factor has been limited and continues to limit the evolution of microprocessor architectures. Today's microprocessors can be so complicated, and a large part can be attributed to the compatibility of legacy software. In consideration of the market, most manufacturers choose to incorporate new architectural features into the latest microprocessor designs, but at the same time, in these most recent productions, all of them are retained to ensure compatibility with older, so-called Required for legacy applications. ----- V --- V ----- install --- (Please read the precautions on the back before filling this page) Order:

X 297公釐) 583583 A7X 297 mm) 583583 A7

經濟部智慧財產局員工消費合作社印製 [〇〇〇7]這種舊有軟體相容性的負擔,沒有其他地方, 會比在x86-相容之微處理器的發展史中更加顯而易見。大 家都知道’現在的32/16位元之虛擬模式(virtual_m〇de)x86 微處理器’仍可執行测年代所撰寫之8位元真實模式 (real-mode)的應用程式。而熟習此領域技術者也承認,有 不少相_架構「包被」堆在χ86 _中,只是為了支援 與舊有應録式及運倾式_。軸在過去,研發 者可將新開發的架構特徵加錢有的齡餘構,但如今 使用這些特徵所憑藉之卫具,即可程式化的指令,卻變得 ,當稀少。更簡單地說,在某些重要的指令集中,已沒有 「多餘」的指令,讓設計者可藉以將更_特徵納入一既 有的架構中。 ^[0008]例如’在χ86指令集架構中,已經沒有任何一 未定義的-位元組大小的運算碼狀態,是尚未被使用的。 ,主要的-位元組大小之Μ6運算碼圖中,全部說個運 算碼狀態都已被既有的指令佔用了。結果是,χ86微處理器 的設計者現在必須在提供新特徵與放棄舊有軟體相容性兩 者間作抉擇。若要提供新的可財化特徵,貞彳必須分派運 ΐ碼狀f給這些特徵。若既有的指令餘構沒有多餘的運 异碼狀態’則某些既存的運算碼狀態必須重新定義, 供給新的特徵。ϋ此’為了提供新的特徵,麟犧牲舊 軟體相容性了。 ^ [〇〇〇9卜個持續發展但在料指令絲射仍未 的領域’即是條件碼回寫(conditioncodes讀eback)之選 {請先閱讀背面之注意事項再填寫本頁} 裝 4 五 、發明說明(+ ) 3= 二在!多現代的指令集中,有些指令會指示-符 二舊有ΐ之麟理11中的執行邏輯,對所提供之運算元 將兩運算元相加,或對兩運算元執行;輯 =生^Γ輯娜之類型為何,運算執行完後 二 ”果(Insult)。而且情況經常是,應用程式可 =:=產;結果之特定屬性,而採行不同路徑 1 p ’右結果是—正數’職程可能進行至 二果點’而若結果是—負數,則流程將進行至B點。 右、、、口果為零,則流程將進行至C點。 由於_程柄流程鱗蚊於舰生之結果 定的邊界條件(如負數、正數、零、溢位、奇 泣 π、、、ϋ果之邊界屬性,並依此改變程式的 ^用66二’用來儲存及檢驗結果條件之卫具中,最廣 .....卩m系列之條件碼位元連同-4b條件分支 存器所存之—個或更多位元的狀態來 運算域生—對舰果,微處理器内之條 每個邊界條件來評估該對應結果,並設定該 3條件竭暫存器中之對應位元,以供條件分支指令後續 置洶之用。 [丨]在許多現代的微處理器中,結果之評估與條件 自動進行的。每計算出一個新的結果,條件 石”、,❻(也稱為旗標暫存器)態就被更新,以反映 583583 經濟部智慧財產局員工消費合作社印製 A7 五、發明說明(f ) 該新的結果之邊界屬性。雖然這項特徵在許多情況下相當 有用,但仍有些時候是不希望在一結果產生後,便將條; 碼暫存器予以更新。例如,熟悉此領域技術者都知道,評 估所,邊界條件以及更新每-新產生結果之對應條件碼位 元,疋相當複雜且花時間的工作,以致於對大部分微處理 器而σ,每计异出一結果,就必須將其執行管線停頓一個 或更夕週期’以兀成邊界條件之評估與相關條件碼位元之 «又疋。這個現象已被微處理器設計者與編譯器設計者注音 ^。所以,當—條件分支指令(條件碼使用者)之前緊鄰 著一建立分支條件的指令(條件碼製造者)時,在這種情 況下,大部分的編譯器與-些微處理器將在指令流中向^ 找尋-個或更多不影響條件碼之後續指令,並將這些後續 指令插入製造條件碼之指令與條件分支指令間,以使程式 瑪的執行工作能安排得更有效率。 陶2]但熟悉此領域技術者都知道 指 是_利插人,因此,管線運作之停_實常常=非: 問題在執行重複的程式迴圈時,會更加嚴重。而且,若八 支條件已經建立,此時再將不需要的虛擬操作指令(η〇·ς instruCtlGn)插场式餘巾,對朗程式之整體執行速度 而言,已證明是非常不利的。 新的[=17 ’我倾_是,-觀觸條件碼更 新的k擇性控制納入既有微處理器指令集架構的技術,立 中該微處理器指令集係被已定義之運算竭完全佔用,且納 入該技術並不影響-符合舊有規格之微處縣執行舊有應 ___ 6 本紙張尺度適用中_家標準(CN_S)A4規格咖χ四 I I lew— --- (請先閱讀背面之注意事項再填寫本頁) 訂- 583583 583583 五、發明說明(“) 用程式的能力。 _4]在另一情況下,程式員可能想 此 1牛:以作為後續運算是否執行之判斷依據… 二“ 因此 經濟部智慧財產局員工消費合作社印製 A7 B7 2令的執行過程中’能維持一組條件;是需= =條:r侧繼咐鄕恤先前已設 (三)發明簡要說明: _5]本發明如同前述其他申請案,係針對上述及其 他習知技狀問題與缺點純克服。本發明提供_種更ς 的技術,肋擴充微處㈣之指令集,使其超越現有的能 力,對於所產生結果之職條件碼的更新,進行可程式化 的控制。在-具體實施财,提供了—種微處理器裝置, 用以選擇性地控制條件碼之回寫。該微處理器裝置具有一 轉譯邏輯(translation logic)與—延伸執行轉、(咖顧 executum l〇glC)。該轉譯邏輯將一延伸指令轉譯成對應之 微指令(micro instruction)。該延伸指令具一延伸前^碼 (extended prefix)與一延伸前置碼標記(extended ρΓ^χPrinted by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs [0007] This old software compatibility burden is nowhere more apparent than in the history of the development of x86-compatible microprocessors. Everyone knows that 'the current 32 / 16-bit virtual_mode x86 microprocessor' can still execute 8-bit real-mode applications written in chronology. And those who are familiar with this field also acknowledge that there are a lot of similar _architectural “coatings” stacked in χ86 _, just to support the old recordable and operational style_. In the past, developers could add newly developed architectural features to some old-fashioned structures, but nowadays, using the guards that these features rely on, they can be programmed instructions, but they are scarce. To put it simply, in some important instruction sets, there are no “redundant” instructions, allowing designers to incorporate more features into an existing architecture. ^ [0008] For example, 'In the x86 instruction set architecture, there is no one undefined-byte-sized opcode state, and it has not been used yet. In the main -byte size M6 opcode diagram, all said that the status of the opcode has been occupied by the existing instructions. As a result, designers of x86 microprocessors must now choose between providing new features and giving up legacy software compatibility. To provide new financial characteristics, Chastity must assign shipping codes to these characteristics. If the existing instruction coform does not have any extra code state ', then some existing opcode states must be redefined to provide new features. So, to provide new features, Lin sacrifices compatibility of old software. ^ [〇〇〇〇09 Continued development, but in the field of material instruction has not yet shot 'is the condition code readback (conditioncodes read eback) choice {Please read the precautions on the back before filling this page} Pack 4 5 2. Description of the invention (+) 3 = 2 In the multi-modern instruction set, some instructions will instruct the execution logic in the old-fashioned linguistic theory 11 to add the two operands to the provided operands, or Execute on two operands; what is the type of 辑 = 生 ^ Γ 娜 Na, and the result is “Insult” after the operation is performed. And often, the application can be =: = product; the specific attributes of the result, but different Path 1 p 'Right result is-positive number' The job may proceed to the second fruit point 'and if the result is-negative number, the process will proceed to point B. If the right,, and fruit are zero, the process will proceed to point C Because of the boundary conditions (such as negative, positive, zero, overflow, odd π, ,, and capsule) boundary conditions determined by the _ Cheng handle process of Lepidoptera mosquitoes, and change the program's ^ use 66 two 'The most widely used ... And the -4b conditional branch register to store the state of one or more bits to calculate the domain generation-for ship results, each boundary condition in the microprocessor evaluates the corresponding result, and sets the 3 condition exhaustion The corresponding bit in the register is used for subsequent use of conditional branch instructions. [丨] In many modern microprocessors, the result evaluation and condition are performed automatically. Each time a new result is calculated, the condition The status of "stone", "❻" (also known as the flag register) was updated to reflect the printing of A7 by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs, 583583. V. Invention Description (f) The boundary attributes of this new result. Although This feature is quite useful in many cases, but there are still times when you don't want to update the bar code register once a result is produced. For example, those skilled in the art know, evaluate, boundary conditions, and Update the corresponding condition code bits of each-newly generated result, which is quite complicated and time-consuming work, so that for most microprocessors, σ, each time a result is different, the execution pipeline must be stopped This cycle is based on the evaluation of boundary conditions and related condition code bits. This phenomenon has been pronounced by microprocessor designers and compiler designers ^. Therefore, when-conditional branch instructions (conditions Code user) immediately before an instruction (condition code maker) that establishes a branch condition. In this case, most compilers and some microprocessors will look for ^ or more in the instruction stream. Does not affect the subsequent instructions of the condition code, and inserts these subsequent instructions between the instruction that makes the condition code and the conditional branch instruction, so that the execution of the program can be arranged more efficiently. Tao 2] But those skilled in the art will know Refers to _ inserting people, so the stop of the pipeline operation is often = no: the problem will be more serious when performing repeated program loops. Moreover, if eight conditions have been established, then inserting unnecessary virtual operation instructions (η〇 · ς instruCtlGn) into the field-type wiper has proved to be very unfavorable to the overall execution speed of the Lang program. The new [= 17 '我 倾 _ 是 , -watch the conditional code update k selective control technology incorporated into the existing microprocessor instruction set architecture, the microprocessor instruction set is completely completed by the defined operations Occupation, and the inclusion of this technology does not affect-the implementation of the old application should be carried out in micro-county counties that meet the old specifications ___ 6 This paper size is applicable _ home standard (CN_S) A4 size coffee χ four II lew — --- (please first Read the notes on the back and fill in this page) Order-583583 583583 V. Inventive (") The ability to use programs. _4] In another case, the programmer may want this 1 cow: as a judgment of whether the subsequent calculation is performed Based on ... "Therefore, during the implementation of the A7 B7 2 order printed by the Consumer Property Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs," a set of conditions can be maintained; it is necessary to ==: the r side continues to command the shirt to have previously been set. (3) invention brief Explanation: _5] The present invention, like the other applications mentioned above, addresses the problems and shortcomings of the above and other conventional techniques. The present invention provides a more advanced technology, which expands the micro-processing instruction set beyond the existing capabilities, and carries out programmable control over the update of the job condition code of the resulting result. The present invention provides a microprocessor device for selectively controlling the write-back of condition codes. The microprocessor device has a translation logic and an extended execution translation. The translation logic translates an extended instruction into a corresponding micro instruction. The extended instruction has an extended prefix and an extended prefix code (extended ρΓ ^ χ

tag)。該以申前置碼除能對應-指定運算結果之該條件碼 的回寫。該延伸刖置碼標記則指出該延伸前置碼,其^中延 伸前置碼標記係一微處理器指令集内另一依據架構^斤指定 之運算碼。該延伸執行邏輯耦接至轉譯邏輯,用以接^該 對應之微指令,產生該結果,並除能該條件碼之回寫。Xtag). This preamble removal can correspond to the write-back of the condition code of the specified operation result. The extended preamble flag indicates the extended preamble. Among them, the extended preamble flag is another operation code specified in the microprocessor instruction set according to the architecture. The extended execution logic is coupled to the translation logic for receiving the corresponding microinstruction, generating the result, and disabling the write-back of the condition code. X

[0016]本發明的-個目的,係提出一種為既有微處理[0016] An object of the present invention is to provide an existing microprocessing

本紙張尺度適用中國國家標準(CNS)A4規格(210 x 297公釐) 583583This paper size applies to China National Standard (CNS) A4 (210 x 297 mm) 583583

器指令集增添條件旗標之回寫控娜徵 伸機制具有-延伸指令與一轉譯器(一 〇=延= 令指不-微處理n餘止—概舞件簡之子集合的回 寫。該些條件旗標反映出—絲之邊界條件,其中該结果 算n該延伸指令包含該既有微處 理盗^集射—選取之碼,其制歸1位元之 延伸控—置碼。該選取之運算碼指出該延伸指令,而該η 位;ΪΓ控财置碼難出該子集合。該轉譯器接收該 = 並產生""微指令序列,以指示微處理器執行該 運异’並指示―回寫控制邏輯於產生該結果後,排除 該子集合的回寫。 [0017] 本發明的另—目的,在於提出—種為既有微處 理器指令集增添選擇性之條件碼回寫能力的指令集延伸模 組。该指令集延伸模組具有—逸出標記(酿坪吨)、一 條件碼回寫指定元(eGdeswritebaekspedfier)及—條件碼 回寫控制器(conditioncodes讀ebackc〇咖Uer)。該逸出 經 濟 部 智 慧 財 產 局 員 工 消 費 合 作 社 印 製 才不。己由轉澤邏輯接收,並指出一對應指令之附隨部分係 指定了-微處理器所要執行之—延伸運算,其中該逸出標 記為該既有微處理器指令集内之—第—運算碼項目。該條 件碼回寫&疋元柄接至該逸出標記’且為該附隨部分其中 之一’㈣指定複數侧聯魏延料算絲之條件碼。 該條件碼回寫㈣ϋ·至該轉譯邏輯,用赚能該些條 件碼之回寫,並致能其餘條件碼的回寫。 [0018] 本發明的再-目的’在於提供一種擴充微處理 8The write-back control mechanism for adding conditional flags to the device instruction set has an -extend instruction and a translator (10 = delay = order finger not-microprocessing n leftover-writeback of a subset of the sketches). These conditional flags reflect the boundary conditions of the silk, where the result is calculated. The extension instruction includes the code of the existing micro-processing stealth ^ set shot-selection code, which is controlled by the 1-bit extension control-set code. The selection The operation code indicates the extended instruction, and the n-bit control code is difficult to generate the subset. The translator receives the = and generates a " " microinstruction sequence to instruct the microprocessor to execute the operation difference. It also instructs the write-back control logic to exclude the write-back of the sub-set after the result is generated. [0017] Another object of the present invention is to propose a condition code return for adding selectivity to an existing microprocessor instruction set. Instruction set extension module for writing ability. The instruction set extension module has-escape mark (bundling tons), a condition code write-back designated element (eGdeswritebaekspedfier), and-condition code write-back controller (conditioncodes read ebackc〇Ca Uer). The Ministry of Economy It is not printed by the Consumer Cooperatives of the Property Bureau. It has been received by Zhunze Logic and pointed out that the accompanying part of a corresponding instruction specifies-the microprocessor to perform-extended operations, where the escape mark is the existing micro The first-opcode item in the processor's instruction set. The condition code is written back & the element handle is connected to the escape tag 'and is one of the accompanying parts', the number of the specified plural side-associated Weiyan materials is calculated. Condition code. The condition code writeback ㈣ϋ to the translation logic uses the writeback of the condition codes to enable the writeback of the remaining condition codes. [0018] The re-objective of the present invention is to provide an extension Microprocessing 8

583583583583

發明說明(y 器指令集的方法,以提供可 a r該方法包括提供—延伸指令,該延wr包 卞延=二你運騎,透職條件碼回寫前置碼與 親机令之其餘部分指定所要執行之—運算,立中關聯 取條_回寫將被禁止;以i執行該 運开以產生u,並禁止該選取條件碼之回寫。 (四)發明圖示說明: [0019] 本發明之前述與其它目的、特徵及優點,在配 合下列說明及所關示後,將可獲得肢的理解: [0020] 圖-係為—蝴技狀微處理難令格式的方 塊圖; [0021] 圖二係、為_表格,其描述_指令集架構中之指 令,如何對應至圖一指令格式内一 8位元運算碼位元組之 位元邏輯狀態; [0022] 圖三係為本發明之延伸指令格式的方塊圖; [0023] 圖四係為一表格,其顯示依據本發明,延伸架 經濟部智慧財產局員工消費合作社印製 構特徵如何對應至一 8位元延伸前置碼實施例中位元的邏 輯狀態; [0024] 圖五係為解說本發明用以控制關聯於所執行運 算結果之條件碼更新之一管線化微處理器的方塊圖; [0025] 圖六係為本發明用於選擇性地控制條件碼回寫 之延伸前置碼之一具體實施例的方塊圖; 9 本紙張尺度適用中國國家標準(CNS)A4規格(210 χ 297公釐) 583583 A7Description of the Invention (The method of the device instruction set to provide can be provided, the method includes providing-extended instructions, the extension wr package extension = two you Yunqi, write the preamble and the rest of the pro-machine order code Specifying the operation to be performed, the associated association fetch_writeback will be forbidden; execute the operation with i to generate u, and prohibit writeback of the selected condition code. (IV) Illustration of the invention: [0019] The foregoing and other objects, features, and advantages of the present invention can be understood by following the descriptions and descriptions below: [0020] FIG.-Is a block diagram of a butterfly-like microprocessing difficult-to-order format; 0021] Figure 2 is a _ table describing how the instructions in the instruction set architecture correspond to the bit logic states of an 8-bit opcode byte in the instruction format of Figure 1. [0022] Figure 3 is [0023] FIG. 4 is a table showing how the printed features of the consumer cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs correspond to an 8-bit extension front according to the present invention. The logical state of the bits in the code embodiment; [0024 Figure 5 is a block diagram illustrating a pipelined microprocessor used to control condition code updates associated with the results of operations performed by the present invention; [0025] Figure 6 is used to selectively control condition code back according to the present invention A block diagram of a specific embodiment of the extended preamble written; 9 This paper size applies to China National Standard (CNS) A4 (210 x 297 mm) 583583 A7

發明說明( B7 观糊細麟輯之纟1 塊圖Hi圖人係為圖五之微處·内延伸執行邏輯的方 此八U圖九係為描述本發明對提供條件碼回寫控制之心令進行轉譯與執行的方法之運作流程圖。 圖號說明: 100指令格式 102運算碼 200 8位元運算碼圖 202運算碼ΠΗ 300延伸指令格式 302運算碼 304延伸指令標記 400 8位元前置碼圖 5〇〇管線化微處理器 502指令快取記憶體/外部記憶體 503指令彳宁列 505延伸轉譯邏輯 507執行邏輯 600延伸前置碼 602回寫控制欄位 700轉譯階段邏輯 101 103 201 301 303 305 401 501 504 506 508 601 前置碼 位址指定元 運算碼值 前置瑪 位址指定元 延伸前置碼 架構特徵 提取邏輯 轉譯邏輯 微指令佇列 延伸執行邏輯 備用欄位 701啟動狀態訊號 10 *·、氏張尺度適用中國國家標準(CNS)A4規格(21〇 X 997公釐 583583Description of the Invention (B7 Observation of the details of the series: 1 block diagram Hi figure is the nuance of Figure 5; the internal extension of the execution logic is shown in Figure 8 and Figure 9 is to describe the present invention to provide the condition code write-back control of the heart Operation flow chart of the method for ordering translation and execution. Explanation of drawing number: 100 instruction format 102 operation code 200 8-bit operation code figure 202 operation code ΠΗ 300 extended instruction format 302 operation code 304 extended instruction mark 400 8-bit prefix Code map 500 pipelined microprocessor 502 instruction cache memory / external memory 503 instruction Ning column 505 extended translation logic 507 execution logic 600 extended preamble 602 write back control field 700 translation stage logic 101 103 201 301 303 305 401 501 504 506 508 601 601 Preamble address designation meta opcode value Prema address designation meta extension preamble architecture feature extraction logic translation logic microinstruction queue extension execution logic standby field 701 activation status signal 10 * ·, the Zhang scale is applicable to the Chinese National Standard (CNS) A4 specification (21 × 997 mm 583583

702機器特定暫存琴 704指令緩衝器 706轉譯控制器 708逸出指令偵測器 710指令解碼器 712微指令緩衝器 714微運算碼欄位 716來源襴位 800延伸執行邏輯 802微指令暫存器 8〇4其餘襴位 806條件旗標暫存器 8〇8結果匯流排 81〇結果暫存器 703延伸特徵襴位 705轉譯邏輯 707除能訊號延伸前置碼解石馬器 711.控制唯讀記憶體713運算礓延伸項欄位 715目的攔位 717位移攔位 801運算元暫存器803運算碼延伸項襴位 805運算元暫存器條件竭回寫控制器 809算術邏輯單元 Γ方〜9法件碼喊㈣之齡進行轉譯與執行 -------------裝--- (請先閱讀背面之注意事項再填寫本頁} 經濟部智慧財產局員工消費合作社印製 (五)發明詳細說明: [29]以下的說明,係在一特定實施例及其必要條件 的脈絡下而提供,可使-般熟習此微術者能夠利用本發 明。然而,各種對該較佳實施例所作的修改,對熟習此項 技術者而言乃係顯而易見,並且,在此所討論的一般原理, 亦可應用至其他實施例。因此,本發明並不限於此處所展 示與敘述之特定實施例,而是具有與此處所揭露之原理與 11 本紙張尺度適用中國國家標準(CNS)A4規格(210 X 297公釐) 583583 A7702 machine-specific temporary storage 704 instruction buffer 706 translation controller 708 escape instruction detector 710 instruction decoder 712 micro instruction buffer 714 micro op code field 716 source bit 800 extended execution logic 802 micro instruction register 804 Other bits 806 Condition flag register 808 Result bus 81. Result register 703 Extended feature bit 705 Translation logic 707 Disabling signal extension Preamble calculus horse 711. Control read only Memory 713 operation 礓 extension field 715 purpose 717 displacement 801 operation element register 803 operation code extension 襕 bit 805 operand register condition exhaustion write back controller 809 arithmetic logic unit Γ square ~ 9 The legal code shouts ㈣ 之 LING for translation and execution ------------- install --- (Please read the precautions on the back before filling out this page} Printed by the Consumers ’Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs (5) Detailed description of the invention: [29] The following description is provided in the context of a specific embodiment and its necessary conditions, so that a person skilled in the art can use the invention. However, various The modification of the preferred embodiment is familiar with this technology. This is obvious, and the general principles discussed herein can also be applied to other embodiments. Therefore, the present invention is not limited to the specific embodiments shown and described herein, but has the same features as those disclosed herein. Principle and 11 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) 583583 A7

583583 A7 五 經濟部智慧財產局員工消費合作社印製 發明說明(丨> ) 運算,而選用(〇Pti〇nal)之位址指定元項目103位於運算 碼102之後,以指定關於該特定運算之附加資訊,像是如 何執行該運算,運算元位於何處等等。指令格式刚並允 許程式員在-運算碼102前加上前置碼項目1〇1。在運算碼 102所指定之特定運算執行時,前置碼1〇1肢指示是否使 用特定的架構特徵。-般來說,這些架構特徵能應用於指 令集中任何運算碼102所指定運算的大部分。例如,現今 前置碼101存在於一些能使用不同大小運算元(如8位元、 16位元、32位元)執行運算的微處理器中。而當許多此類 處理器被程式化為一預設的運算元大小時(比如32位元),' 在其個別指令集中所提供之前置碼1〇1,仍能使程式員依據 各個指令,選擇性地取代(override)該預設的運算元大小 (如為了執行16位元運算)。可選擇之運算元大小僅是架 構特徵之一例,在許多現代的微處理器中,這些架構特徵 能應用於眾多可由運算碼1〇2加以指定的運算(如加、減、 乘、布林邏輯等)。 [0032]圖一所示之指令格式1〇〇,有一為業界所熟知的 範例,此即x86指令格式100,其為所有現代之劝6_相容微 處理器所採用。更具體地說,x86指令格式1〇〇(也稱為χ86 指令集架構100)使用了 8位元前置碼101、8位元運算碼 102以及8位元位址指定元1〇3。χ86架構100亦具有數個 前置碼101,其中兩個取代了 x86微處理器所預設的位址/ 資料大小(即運算碼狀態66H與67H),另一個則指示微 處理器依據不同的轉譯規則來解譯其後之運算碼位元組 13 本紙張尺度適用中國國家標準(CNS)A4規格(210 x 297公釐) {請先閱讀背面之注意事項再填寫本頁) 裝 訂: 583583 經濟部智慧財產局員工消費合作社印製 A7 發明說明(i)) 102 (即前置碼值隨,其使得轉譯動作是依據所謂的二位 ^且運算碼規則來進行)’其他的前置碼101則使特殊運 算重複執行’直至重複條件滿足為止(即卿運算碼:腿、 F2H 及 F3H)。 [0033]現請參閱圖二,其顯示一表格2〇〇,用以描述一 指令集架構之指令2〇1如何對應至圖一指令格式内一 8位 元運算碼位元組102之位元邏輯狀態。表格2〇〇呈現了一 示範性的8位元運算碼圖200,其將一 8位元運算碼項目 102所具有之最多256個值,關聯到對應之微處理器運算碼 才曰々201。表格2〇〇將運异碼項目IQ)之一特定值,譬如 02H,映射至一對應之運算碼指令2〇1 (即指令ι〇22〇ι)。 在x86運算碼圖的例子巾,為此領域中人所熟知的是,運 算碼值14H係映射至x86之進位累加(續评池Cany, ADC)指令,此指令將一 8位元之直接(丨麵地扯)運算 兀加至架構暫存器AL之内含值。熟習此領域技術者也將發 覺,上文提及之x86前置碼1〇1 (亦即66h、67H、〇FH、 F0H、F2H及F3H)係實際的運算碼值2〇1,其在不同脈絡 下,指定要將特定的架構延伸項應用於隨後之運算碼項 1〇2所指定的運算。例如,在運算碼14H (正常情況下,你 刖述之ADC運算碼)前加上前置碼〇FH,會使得χ86處理 器執行一「解壓縮與插入低壓縮之單精度浮點值」(Unpack and Interleave Low Packed Single-Precision Floating-Point Values)運算,而非原本的ADC運算。諸如此X86例子所 述之特徵,在現代之微處理器中係部分地致能,此因微處 冬戒、又適用中國國家標準(CNS)A4規格(21〇 X 297公釐 (請先閱讀背面之注意事項再填寫本頁) 裝 583583 A7 五、發明說明(I 4) 項 理器内之指令轉譯/解碼邏輯是依序解譯一指令1〇〇的項目 κη-κΒ。所以在過去’於指令集架構中使用特定運算石馬值 作為前置碼ιοί ’可允許微處理器設計者將不少先進的 特徵納人相容舊有軟體之微處職的設計巾,而不合 使用那些特定運算碼狀態的舊有程式,帶來執行上二負面 衝擊。例如’-未曾使用χ86運算碼〇FH的舊有程式,仍 可在今日的X86微處理器上執行。而一較新的應用, 藉著運用滿運算碼0FH作為前置碼1〇1,就能使用 新進納入之X86架構特徵,如單一指令多重資料 運算,條件移動運算等等。 [〇〇34]儘管過去已藉由指定可用/多餘的運算碼值加 作為前置碼101 (也稱核構特徵標記/指標應或逸出指 令=1) ’來提供架構特徵,但許多指令絲構刚在提供 功能上的強化時,仍會因為一非常直接的理由,而碰且 礙:所有可用7多餘的運算碼值已觀完,也就是,運算碼 圖200令的全部運算碼值已被架構化地指定。當所有可用 =值被分派為運算綱目1〇2或前置碼項目韻時, 碼值可作為納入新特徵之用。這個嚴重的問 子在於現在鱗多微處理器架射,因而迫使設計 在與保留_式之相容性兩者_/ 性的方式表意的是,圖二所示之指令201係以一般 (如進;立二’、、即124、186),而非具體指涉實際的運算 已 «理器架槿Φ減互斥或)。這是因為,在一些不同的 中,完全佔用之運算碼圍200在架構上 [____ 15 本紙涵適用 583583 A7 B7 經 濟 部 智 慧 財 產 局 消 費 合 作 社 印 製 五、發明說明(丨/ ) f納入較新進展的可能性排除。雖然圖二例子所提到的, 是=位元的運算碼項目102,熟習此領域技術者仍將發覺, 運碼1〇2的特疋大小,除了作為一特殊情況來討論完全 佔用之運算碼結構所造成關題外,其他方面與問題 本身並不相干。因此,一完全佔用之6位元運算碼圖將有 64個可架構化地指定之運算碼/前置碼2〇卜並將無法提供 可用/多餘的運算碼值作為擴充之用。 [0036]另一種做法,則並非將原有指令集廢棄,以一 新的格=100與運算碼圖200取代,而是只針對一部份既 有的運算碼2〇卜以新的指令意含取代,如圖二之運算碼 撕至4FH。以這種混合的技術,符合舊有規格之微處理器 就可以相容舊有軟體模式運作,其中運算碼撕_伽係依 舊有規則來解澤,或者以加強模式(触姐⑽咖如)運作, 其中運算碼4〇H_4FH係依加強之架構規則來解譯。此項技 術確能允許設計者將新特徵納入設計,然而,當符人舊有 規格之賊理砂加強赋時,缺點仍舊存在^為 微處理料能齡任何伽碼·侧的細程式。 因此’站在保留舊有軟體相容性的立場,相容舊有軟體/加 強模式的技術,還是無法接受的。 [〇〇37]然而,對於運算碼空間已完全佔用之指令 勘,且該空間涵蓋所有於符合舊有規格之微處·上執; 之應用程式的情形’本案發明人已注意到其中運算碼2〇1 的使用狀況’且他們亦觀察出,雖然有些指令202是架構 化地指^ ’但未用於能被微處理馳行之應用程式令。圖 16 度·中國國家標準(CNS)A4規^^^^ (請先閲讀背面之注意事項再填寫本頁) 丨籲丨 裝 -·訂· · 583583 A7 五、發明說明(丨ί ) I —所述之指令IF1 202即為此現象之一例。事實上,相同的 運算碼值202 (亦即F1H)係映射至未用於χ86指令集架構 之一有效指令202。雖然該未使用之χ86指令202是有效的 χ86指令202,其指示要在χ86微處理器上執行一架構化地 指定之運算,但它卻未使用於任何能在現代χ86微處理器 上執行之應用程式。這個特殊的Χ86指令202被稱為電路 内模擬中斷點(InCircuitEmulationBreakpoint)(亦即 ICE BKPT,運算碼值為F1H),之前都是專門使用於一種現在 已不存在之微處理器模擬設備中。ICE BKPT 202從未用於 電路内模擬器之外的應用程式中,並且先前使用ICEBKpT 202之電路内模擬设備已不復存在。因此,在的情形下, 本發明人已在一完全佔用之指令集架構2〇〇内發現一樣工 具,藉著利用一有效但未使用之運算碼2〇2,以允許在微處 理器的設計中納入先進的架構特徵,而不需犧牲舊有軟體 之相容性。在一完全佔用之指令集架構2〇〇中,本發明利 用一架構化地指定但未使用之運算碼2〇2,作為一指標標 記,以指出其後之一 η位元前置碼,因此允許微處理器設 緩 計者可將最多2η個最新發展之架獅徵,納人微處理器的 | ☆种’同時保留與所有舊有軟體完全的相容性。 | >[〇〇38]本發明藉提供一 η位元之條件碼回寫控制指定 | 7L前置碼,錢时置補記/㈣前㈣的概念,因而可 | 允許程式員將-習用之供微處理器執行的運算(如加、減、 ! 布林運算、運算元操作等)程式化,並在相同指令内,對 | 於執行該習用運算所產生的結果,致能/除能其相關之指定 製 L_ 17 本紙張尺度適財關家鮮---- (請先閱讀背面之注意事項再填寫本頁) K-裝 ·- 583583 經 濟 部 智 慧 財 k 局 員 工 消 費 合 作 社 印 製 A7 五、發明說明(。) 條件碼的回寫(亦即更新)。在-具體實施例中,微處理 器之條件碼狀態’存於一條件碼(或旗標)暫存器中,其 依據該η位元延伸前置碼内之欄位狀態來更新,而該n位 元延伸^置碼係將關聯於該結果之一條件碼子集合的更新 予以除能。另一實施例則排除了關聯於該結果之所有條件 碼的回寫。本發明現將參照圖三至九進行討論。 [0039] 現請參閱圖三,其為本發明之延伸指令格式3〇〇 的方塊圖。與圖一所討論之格式1〇〇非常近似,該延伸指 令格式300具有數量可變之指令項目301-305,每一項目設 定為:特定值’集合起來便組成微處理器之—特定指令 300 β亥特疋心令3〇〇指示微處理器執行一特定運算,像是 將兩運算元相加,或是將-運算元從記紐搬移至微處理 器之暫存器内。-般而言’指令3〇〇之運算碼項目3〇2指 定了所要執行之特定運算,而選用之位址指定元項目3〇3 則位於運算碼3〇2後,以指定該特定運算之相關附加資訊, 像是如何執行該運算,運算元位於域科。指令格式· 亦允許私式員在-運算碼302前加上前置碼項目則。在運 所指定之特定運算執行時,石馬項目3〇ι係用 來指示是否要使用既有的架構特徵。 [0040] ,然而,本發明的延伸指令300係前述圖一指令 =式100之-超集合(superset),其具有兩個附加項目綱 與305,可被選擇性作為指令延伸項,並置於 指令300中所有其餘項目忽3Q3之前。這兩細加項目 304與305係用於致能/除能關聯於一指定運算結果之複數 18 氏張尺度適用中國國家標準(CNS)A4規格(21G χ 297 -------------^--------^---------. <請先閱讀背面之注意事項再填寫本頁) 583583 A7583583 A7 The Intellectual Property Bureau of the Ministry of Economy ’s Employees ’Cooperative publishes a description of the operation (丨 &) operation, and selects (〇Pti〇nal) to specify meta-item 103 after the operation code 102 to specify the specific operation. Additional information, such as how to perform the operation, where the operands are located, and so on. The instruction format just allows the programmer to add a preamble item 101 before-opcode 102. When a specific operation specified by the operation code 102 is performed, the preamble 101 indicates whether to use a specific architectural feature. -In general, these architectural features can be applied to most of the operations specified by any opcode 102 in the instruction set. For example, preamble 101 today exists in some microprocessors that can perform operations using different sized operands (such as 8-bit, 16-bit, 32-bit). And when many of these processors are programmed to a preset operand size (such as 32-bit), 'pre-coded to 101 in their individual instruction set still enables the programmer to follow the instructions To selectively override the preset operand size (for example, to perform a 16-bit operation). The selectable operand size is just one example of architectural features. In many modern microprocessors, these architectural features can be applied to many operations that can be specified by opcode 102 (such as addition, subtraction, multiplication, and Bollinger logic). Wait). [0032] The instruction format 100 shown in FIG. 1 has a well-known example in the industry, this is the x86 instruction format 100, which is used by all modern 6-compatible microprocessors. More specifically, the x86 instruction format 100 (also referred to as the x86 instruction set architecture 100) uses an 8-bit preamble 101, an 8-bit opcode 102, and an 8-bit address designator 103. The χ86 architecture 100 also has several preambles 101, two of which replace the preset address / data size of the x86 microprocessor (that is, opcode states 66H and 67H), and the other instructs the microprocessor to use different Translation rules to interpret the subsequent operation code bytes 13 This paper size applies the Chinese National Standard (CNS) A4 specification (210 x 297 mm) {Please read the precautions on the back before filling this page) Binding: 583583 Economy Printed by the Intellectual Property Bureau Employee Consumer Cooperative A7 Invention Description (i)) 102 (that is, the preamble value follows, which makes the translation action based on the so-called two-digit ^ and opcode rules) 'Other preambles 101 Then the special operation is repeated until the repetition condition is satisfied (ie the opcodes: legs, F2H and F3H). [0033] Please refer to FIG. 2, which shows a table 200, which is used to describe how the instruction 201 of an instruction set architecture corresponds to the bit of an 8-bit operation code byte 102 in the instruction format of FIG. 1. Logical state. Table 2000 presents an exemplary 8-bit opcode map 200, which associates up to 256 values of an 8-bit opcode item 102 with the corresponding microprocessor opcode 々201. The table 200 maps a specific value of the different code item IQ), such as 02H, to a corresponding operation code instruction 201 (ie, instruction ι0222). In the example of the x86 opcode diagram, it is well known in the art that the opcode value 14H is mapped to the x86 carry and accumulate (continued evaluation pool Cany, ADC) instruction. This instruction will be an 8-bit direct (丨 from the surface) operation Wu added to the internal value of the structure register AL. Those skilled in this field will also find that the x86 preamble 101 (ie, 66h, 67H, 0FH, F0H, F2H, and F3H) mentioned above is the actual operation code value 201, which is different in different In the context, it is specified that a specific architectural extension is to be applied to the operation specified by the subsequent operation code item 102. For example, adding the preamble 0FH before the opcode 14H (normally, the ADC opcode you describe) will cause the χ86 processor to perform a "decompression and insert low-compression single-precision floating-point value" ( Unpack and Interleave Low Packed Single-Precision Floating-Point Values) operations instead of the original ADC operations. The features described in this X86 example are partially enabled in modern microprocessors. This is due to the small winter ring and also applicable to the Chinese National Standard (CNS) A4 specification (21〇X 297 mm (please read first Note on the back, please fill out this page again) Install 583583 A7 5. Inventory (I 4) The instruction translation / decoding logic in the item processor is to sequentially interpret the items κη-κΒ of an instruction 100. So in the past ' The use of specific math values as a preamble in the instruction set architecture allows the microprocessor designer to incorporate many advanced features into the micro-service design towels of the old software without using those specific The old programs of the state of the operation code bring the negative impact of the execution of the last two. For example, '-the old programs that have not used the χ86 operation code 0FH can still be executed on today's X86 microprocessors. A newer application, By using the full operation code 0FH as the preamble 101, it is possible to use newly incorporated X86 architecture features, such as single instruction multiple data operations, conditional movement operations, etc. [〇〇34] Although it has been available by designation in the past / Excessive opcodes The value plus is used as the preamble 101 (also known as the core feature flag / indicator or escape command = 1) to provide architectural features, but when many instruction fabrics just provide functional enhancements, they will still be very direct The reason for this is to get in the way: all available 7 extra opcode values have been viewed, that is, all opcode values of the opcode graph 200 have been architecturally specified. When all available = values are assigned as the operation outline The code value can be used to incorporate new features when it comes to 10 or preamble items. This serious problem lies in the fact that many microprocessors are now shooting, which forces the design to be compatible with the reserved form. _ / Means that the instruction 201 shown in Figure 2 is general (such as Jin; Li Er ', 124, 186), rather than specifically referring to the actual operation has been reduced Mutually exclusive or). This is because, in some different cases, the fully occupied operation code is around 200 in the structure [____ 15 This paper is applicable to 583583 A7 B7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 5. Inventory (丨 /) f Includes newer The possibility of progress is ruled out. Although the example in Figure 2 refers to the = bit operation code item 102, those skilled in the art will still notice that the special size of the operation code 102 is a special case to discuss the fully occupied operation code. Beyond the problem caused by the structure, other aspects have nothing to do with the problem itself. Therefore, a fully occupied 6-bit opcode map will have 64 opcodes / preambles that can be architecturally specified and will not provide usable / excessive opcode values for expansion. [0036] Another method is not to discard the original instruction set and replace it with a new grid = 100 and opcode map 200, but only to a part of the existing opcodes. With substitution, the operation code shown in Figure 2 is torn to 4FH. With this hybrid technology, microprocessors that meet the old specifications can be compatible with the old software mode operation, in which the operation code tearing_Ga system still has rules to solve the problem, or to strengthen the mode (touch sister ⑽Karu) Operation, where the operation code 40H_4FH is interpreted in accordance with the enhanced architectural rules. This technology does allow the designer to incorporate new features into the design. However, when the old man's old specifications are enhanced, the shortcomings still exist. ^ It is a fine program for the micro-processing materials to age any side. Therefore, from the standpoint of preserving legacy software compatibility, technology that is compatible with legacy software / enhancement mode is still unacceptable. [0037] However, the instruction code for the operation code space has been completely occupied, and the space covers all applications in accordance with the old specifications of the micro-description; the case of the application 'The inventor of this case has noticed that the operation code The use status of 001 'and they also observed that although some instructions 202 refer to the structured ^', they are not used for application programs that can be processed by microprocessing. Figure 16 Degree · Chinese National Standard (CNS) A4 Regulation ^^^^ (Please read the precautions on the back before filling out this page) 丨 Call 丨 Installation- · Order · 583583 A7 V. Description of the Invention (丨 ί) I — The instruction IF1 202 is an example of this phenomenon. In fact, the same opcode value 202 (ie F1H) is mapped to a valid instruction 202 that is not used in the x86 instruction set architecture. Although the unused χ86 instruction 202 is a valid χ86 instruction 202, which instructs a architecturally specified operation to be performed on a χ86 microprocessor, it is not used in any implementation that can be performed on a modern χ86 microprocessor. application. This special X86 instruction 202 is called InCircuitEmulationBreakpoint (InCircuitEmulationBreakpoint) (that is, ICE BKPT, the operation code value is F1H). It was previously used exclusively in a microprocessor simulation device that no longer exists. ICE BKPT 202 has never been used in applications other than in-circuit simulators, and in-circuit simulation devices that previously used ICEBKpT 202 no longer exist. Therefore, in the circumstances, the present inventor has found a tool within a completely occupied instruction set architecture 200, by using an effective but unused operation code 200 to allow the design of the microprocessor Incorporates advanced architectural features without sacrificing compatibility with legacy software. In a completely occupied instruction set architecture 200, the present invention uses a structurally specified but unused operation code 200 as an index mark to indicate the next n-bit preamble, so Allows the microprocessor to set up a delayer to incorporate up to 2η of the latest developments, including the microprocessor's | ☆ species' while retaining full compatibility with all older software. > [〇〇38] The present invention provides a η-bit condition code write-back control designation | 7L preamble, the concept of replenishment / ㈣ before 钱 when money, so | programmers can-use it Operations performed by the microprocessor (such as addition, subtraction,! Brin operations, operand operations, etc.) are stylized, and within the same instruction, enable / disable the results of performing the custom operation Relevant designated system L_ 17 This paper size is suitable for financial and family care. (Please read the precautions on the back before filling out this page) K-pack ·-583583 Printed by A7, Employees ’Cooperatives, Bureau of Wisdom and Finance, Ministry of Economic Affairs 2. Description of the invention (.) Write-back of condition code (that is, update). In a specific embodiment, the condition code status of the microprocessor is stored in a condition code (or flag) register, which is updated according to the status of the field in the n-bit extended preamble, and the The n-bit extension code disables the update of a conditional code sub-set associated with the result. Another embodiment excludes the write-back of all condition codes associated with the result. The invention will now be discussed with reference to Figs. [0039] Please refer to FIG. 3, which is a block diagram of the extended instruction format 300 of the present invention. It is very similar to the format 100 discussed in Figure 1. The extended instruction format 300 has a variable number of instruction items 301-305, each of which is set to: a specific value 'collected to form a microprocessor-a specific instruction 300 β Hite's heart order 300 instructs the microprocessor to perform a specific operation, such as adding two operands, or moving the -operator from the register to the microprocessor's register. -In general, the operation code item 30 of the 'instruction 300' specifies the specific operation to be performed, and the selected address designation meta-item 3 is located after the operation code 3002 to specify the specific operation. For additional information, such as how to perform the operation, the operand is in the domain section. Instruction format · It also allows the private operator to add a preamble item before -operation code 302. The Shima Project 30m is used to indicate whether to use the existing architectural features when the specific operation specified by the operation is performed. [0040] However, the extended instruction 300 of the present invention is the above-mentioned instruction of Fig. 1 = the superset of Equation 100, which has two additional items and 305, which can be selectively used as instruction extension items and placed in the instruction. All remaining items in the 300 before 3Q3. The two fine-added items 304 and 305 are used to enable / disable the complex 18-degree scale associated with a specified operation result. The Chinese National Standard (CNS) A4 specification (21G χ 297 --------- ---- ^ -------- ^ ---------. < Please read the notes on the back before filling this page) 583583 A7

裝------- (請先閱讀背面之注意事項再填寫本頁) ,· 項再A -n n n n , 583583 Α7 Β7 五 、發明說明) --------------裝· (請先閱讀背面之注意事項再填寫本頁) 除執行。在一具體實施例中,延伸前置碼305具八位元的 大小,最多可指定256種不同的該些條件碼之子集合。 元前置碼的實施例,則最多可指定2n種不同的條件碼組合。 [0042]現请參閱圖四,一表格400顯示依據本發明, 條件碼更新控制延伸項如何映射至一 8位元延伸前置碼實 施例之位元邏輯狀態。類似於圖二所討論之運算碼圖2qq, 圖四之表格400呈現一 8位元條件碼指定元之前置碼圖4〇〇 的範例,其將一 8位元延伸前置碼項目305之最多256個 值,關聯到一符合舊有規格之微處理器的對應條件碼狀態 401 (如E34、E4D等)。在一 χ86的具體實施例中,本發 明之8位元延伸特徵前置碼305係提供給指定元4〇1 (亦即 E00-EFF)使用,該些指定元401乃現行χ86指令集架構所 未能提供的。 經濟部智慧財產局員工消費合作社印製 [0043] 圖四所示之延伸特徵4〇ι係以一般性的方式表 示,而非具體指涉實際的特徵,此因本發明之技術可應用 於各種不同的架構延伸項401與特定的指令集架構。熟習 此領域技術者將發覺,許多不同的架構特徵4〇1,其中一些 已於上文提及,可依此處所述之逸出標記3〇4/延伸前置碼 3〇5技術將其納入-既有之指令集。圖四之8位元前置瑪實 施例提供了最多256個不同的特徵4Q1,而_η位元前置碼 實施例則具有最多2η個不同特徵4〇1的程式化選擇。 [0044] 在對應不同_之條件碼贿絲示方式的實 施例中,這些類型對許多現代之微處理器而言是很常見 的,條件碼之組合可被指定為禁止回寫,這些植合包括- 本紙張尺度適用中國國家標準(CNS)A4規格(210 297公餐) 583583 Γ; 經濟部智慧財產局員工消費合作社印製 Α7 -----—2Ζ________ 、發明說明(γ) 結果之邊界屬性,如等於零、不等於零、偶同位(even parity)、可同位(0(jd parity)、帶負號(以职)、不帶負 號、溢位(overflow)、未溢位、進位(carry)、未進位等 等。在許多此種微處理器中,一條件碼狀態項目(亦即暫 存器)以複數個條件碼位元(或旗標)加以組態,其中每 一位元代表一最近產生之結果是否已超過某一結果邊界條 件,像是產生一進位位元,或有一符號位元顯示該結果係 一負數。然而,前述之特定條件碼並未將本發明之範圍限 制於一特定之結果條件碼集合。前述實施例可作為範例, 用來解說依本發明,一條件碼指定元前置碼3〇5如何被編 碼,以便在一結果產生後,排除特定條件碼子集合之更新。 熟習此領域技術者將可察覺,一特殊條件碼指定元前置碼 305之組態,係依據一對應微處理器中條件碼如何表示與儲 存而定。 [0045] 現請參閱圖五,其為解說本發明用以控制關聯 於所執行運算結果的條件碼更新之管線化微處理器5〇〇的 方塊圖。微處理器500具有三個明顯的階段類型:提取、 轉譯及執行。提取階段具有提取邏輯50卜可從指令快取記 憶體502或外部記憶體502提取指令。所提取之指令經由 指令佇列503送至轉譯階段。轉譯階段具有轉譯邏輯5〇4, 耦接至一微指令佇列506。轉譯邏輯504包括延伸轉譯邏輯 505。執行階段則有執行邏輯5〇7,其内具有延伸執行邏輯 508 〇 [0046] 依據本發明,於運作時,提取邏輯501從指令 21 本紙張尺度適用中國國家標準(CNS)A4規格(210 X 297公爱) --------,-----裝--- (請先閲讀背面之注意事項再填寫本頁) _ 583583 五、發明說明) $取A憶體/外部記憶體5G2提取格式化指令,並將這些指 :依^執行順序放入指令符列5〇3中。接著從指令仔列Μ3 提取逆些指令,送至轉譯邏輯5〇4。轉譯邏輯5〇4將每-送 ^的&令轉澤/解碼為—對應之微指令序列,以指示微處理 益5㈨去執行這些指令所指定的運算。依本發明,延伸轉 澤邏輯5〇5偵測那些具有延伸前置碼標記之指令,以進行 ,應條件伽寫指定元前置碼之轉譯/解碼。在-x86的實 $例中,延伸轉譯邏輯5〇5組態為翻其值為随之延伸 前置碼標記,其係、χ8ό之ICE Βκρτ運算碼。微指令欄位則 提供於微指令符列5〇6中,以指定要被排除更新之條件碼。 經濟部智慧財產局員工消費合作社印製 [0047] 微指令從微指令佇列5〇6被送至執行邏輯 5〇7,由延伸執行邏輯508偵測具有選擇性條件碼回寫特徵 的微指令,此處之回寫特徵係依微指令襴位的指示來致 月b。延伸執行邏輯508執行微指令所指定之運算,並產生 對應的結果。在對應結果產生後,延伸執行邏輯5〇8即依 微指令攔位的指示除能邊界條件的評估以及條件碼的更 新。在一具體實施例中,只有被延伸指令指定為要更新之 條件碼,才會被寫回條件碼暫存器。在另一實施例中,只 有未被延伸指令指定為不更新之條件碼,才會被寫回條件 碼暫存器。 " [0048] 熟習此領域技術者將發現,圖五所示之微處理 器500係現代之管線化微處理器50經過簡化的結果。^實 上,現代的管線化微處理器500最多可包含有2〇至3〇個 不同的管線階段。然而,這些階段可概括地歸類為方塊^ 22 本紙張尺度適用中國國家標準(CNS)A4規格(210 X 297公釐)" --—-- 經濟部智慧財產局員工消費合作社印製 583583 A7 ------ - B7 五、發明說明(>>) 所示之三個階段,因此,圖五之方塊圖500可用以點明前 述本發明實施例所需之必要元件。為了簡明起見,微處理 器500中無關的元件並未顯示出來。 [0049] 現請參閱圖六,其為本發明用於選擇性地控制 條件碼回寫之延伸前置碼之一具體實施例的方塊圖。該延 伸條件碼指定元前置碼600係一 8位元之延伸前置碼6〇〇, 且具有一包含複數個位元(起始於位元〇)之回寫控制(呢池 back contrd ’ wbc)欄位6〇2與一包含其餘位元(結束於 位元7)之備用欄位601。在一具體實施例中,wbc攔位 602包含位兀〇,並依本發明指示延伸執行邏輯將所有條件 碼之回寫予以除能。另一具體實施侧使用前置碼6〇〇較 低的五個位元作為WBC襴位6〇2,其中每一位元分別對應 至禁止負號條件、進位條件、溢位條件、零值條件以及同 位條件之回寫。 [0050] 現請參閱圖七,其為圖五之微處理器内轉譯階 段邏輯700之細部的方塊圖。轉譯階段邏輯7〇〇具有一指 令緩衝器704,依本發明,其提供延伸指令至轉譯邏輯7〇5。 轉,邏輯7〇5係柄接至一具有一延伸特徵棚位7〇3之機器 特定暫存器(machine S_flc register) 7〇2。轉譯邏輯 7〇5 具-轉譯控制器7〇6,其提供一除能訊號7〇7至一逸出指令 偵測器708及一延伸前置碼解碼器7〇9。逸出指令债測器 7〇8麵接至延伸前置碼解碼$ 7〇9及一指令解石馬器训。延 伸前置碼解碼器709與指令解碼邏輯71〇存取一控制唯讀 記憶體(ROM) 711,其中儲存了對應至某些延伸指令之樣 --II — I*— I I I ---- ---I ^ — — — — — — — I I ^^1 . (請先閲讀背面之注意事項再填寫本頁) 23 583583 A7 、發明說明(4) 板(te_ate)微指令序列。轉譯邏輯705 φ包含一微指令 緩衝器m ’其具有一運算碼延伸項襴位爪、一微運算碼 欄位7U、-目的攔位715、一來源襴位716以及一 位Ή7。 [〇岡運作上,在微處理器通電啟動期間,機器特定 暫存器观内之延伸欄位703的狀態係藉由訊號啟動狀態 a 哪敝)7(31決定,以指出該特定微處理 疋否犯轉n行本發明之延伸條件執行指令。在一具毙 m訊號701從一特徵控制暫存器(圖上未顯;) *出’該特徵控制暫存關讀取—於 絲陣列加―(未顯示)。機器特定暫 伸特徵攔位703之狀態送至轉譯控制器鄕。=制^ ==令:衝器7。4所提取· 擇f生條件物寫控制轉譯規職既有轉 提供這樣的控制特徵,可允許監督_程式(、如BI⑹ 延伸_徵。若延伸特徵被除能,則 具有被選為延伸特徵標記之運算 只】 臟則進行轉譯。在一 x86的具體有 ==不:,指令異常(exeeptlGn)。_ 」 來。逸出指令_器7〇8因而於偵测出 譯/解碼標記之後的延伸條件碼指定解碼器轉 解·。的運作’並於轉譯/解顯二二之 度適用中國 請 先 閲 讀 背 Φ 器 體 輯 選 致 則 算 F1H 譯 意 事 項Equipment ------- (Please read the precautions on the back before filling out this page), and item A-nnnn, 583583 Α7 Β7 V. Description of the invention) ------------- -Install · (Please read the precautions on the back before filling this page) In a specific embodiment, the extended preamble 305 has an octet size, and a maximum of 256 different subsets of the condition codes can be specified. In the embodiment of the meta-preamble, a maximum of 2n different condition code combinations can be specified. [0042] Referring now to FIG. 4, a table 400 shows how a condition code update control extension is mapped to a bit logic state of an 8-bit extended preamble implementation according to the present invention. Similar to the operation code diagram 2qq discussed in FIG. 2, the table 400 in FIG. 4 presents an example of an 8-bit condition code designator before the code diagram 400, which extends an 8-bit preamble item 305 A maximum of 256 values are associated with the corresponding condition code status 401 (such as E34, E4D, etc.) of a microprocessor that meets the old specifications. In a specific embodiment of χ86, the 8-bit extended feature preamble 305 of the present invention is provided for use by designated elements 401 (ie, E00-EFF), and these designated elements 401 are used by the current x86 instruction set architecture. Failed to provide. Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs [0043] The extended feature 40m shown in Figure 4 is expressed in a general way, and does not specifically refer to actual features. This is because the technology of the present invention can be applied to various Different architecture extensions 401 and specific instruction set architectures. Those skilled in the art will find that there are many different architectural features 401, some of which have been mentioned above, which can be converted according to the escape mark 3 04 / extended preamble 305 technology described here. Inclusion-Existing instruction set. The 8-bit preamble embodiment of FIG. 4 provides a maximum of 256 different features 4Q1, and the _η-bit preamble embodiment has a stylized selection of up to 2η different features 401. [0044] In the embodiments corresponding to different condition code bridging methods, these types are common to many modern microprocessors, and the combination of condition codes can be designated as prohibiting write-back. Including-This paper size applies the Chinese National Standard (CNS) A4 specification (210 297 meals) 583583 Γ; printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs A7 ------2 ________, the boundary of the results of the invention (γ) Attributes, such as equal to zero, not equal to zero, even parity, even parity (0 (jd parity), negative sign (employment), no negative sign, overflow, non-overflow, carry ), No carry, etc. In many such microprocessors, a condition code status item (ie, a register) is configured with a plurality of condition code bits (or flags), where each bit represents Whether a recently produced result has exceeded a certain result boundary condition, such as generating a carry bit, or a sign bit indicating that the result is a negative number. However, the foregoing specific condition code does not limit the scope of the present invention to A specific A condition code set. The foregoing embodiment can be used as an example to explain how a condition code specifies a meta-preamble 305 to be encoded in order to exclude an update of a specific condition code sub-set after a result is generated. Those skilled in the art will perceive that the configuration of a special condition code designation element preamble 305 depends on how a condition code is represented and stored in a corresponding microprocessor. [0045] Please refer to FIG. To illustrate the block diagram of a pipelined microprocessor 500 for controlling condition code updates associated with the results of operations performed by the present invention. The microprocessor 500 has three distinct types of stages: extraction, translation, and execution. Extraction stages With extraction logic 50, instructions can be retrieved from the instruction cache memory 502 or external memory 502. The extracted instructions are sent to the translation stage via the instruction queue 503. The translation stage has translation logic 504, which is coupled to a micro instruction Queue 506. Translation logic 504 includes extended translation logic 505. During the execution phase, there is execution logic 507, which has extended execution logic 508. [0046] According to the present invention, When working, extract logic 501 from instruction 21 This paper size applies Chinese National Standard (CNS) A4 specification (210 X 297 public love) --------, ----- install --- (Please read first Note on the back, please fill out this page again) _ 583583 V. Description of the invention $ Take A memory / external memory 5G2 to extract the formatting instructions, and put these instructions: in the order of ^ execution into the instruction symbol 503. Then, the inverse instructions are extracted from the instruction array M3 and sent to the translation logic 504. The translation logic 504 translates each & order & order into a corresponding micro instruction sequence to instruct the micro processor 5 to execute the operations specified by these instructions. According to the present invention, the extended translation logic 505 detects those instructions with extended preamble flags to perform translation / decoding of the specified meta-prefixes in accordance with conditional writing. In the actual example of -x86, the extended translation logic 505 is configured to translate its value to the corresponding extended preamble mark, which is the ICE κκρτ operation code. The microinstruction field is provided in the microinstruction column 5506 to specify the condition code to be excluded from updating. Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs [0047] Microinstructions are sent from the microinstruction queue 506 to the execution logic 507, and the extended execution logic 508 detects microinstructions with selective condition code write-back features The write-back feature here refers to the month b according to the instruction of the micro instruction niches. The extended execution logic 508 executes the operation specified by the microinstruction and produces a corresponding result. After the corresponding result is generated, the extended execution logic 508, which is the evaluation of the disabling boundary condition according to the instruction of the micro instruction block, and the update of the condition code. In a specific embodiment, only the condition code specified by the extended instruction to be updated will be written back to the condition code register. In another embodiment, only condition codes that are not specified by the extended instruction as not to be updated are written back to the condition code register. [0048] Those skilled in the art will find that the microprocessor 500 shown in FIG. 5 is a simplified result of the modern pipelined microprocessor 50. In fact, a modern pipelined microprocessor 500 may contain up to 20 to 30 different pipeline stages. However, these stages can be broadly categorized as squares. ^ 22 This paper size is applicable to the Chinese National Standard (CNS) A4 (210 X 297 mm) " ----- Printed by the Intellectual Property Bureau of the Ministry of Economic Affairs and Consumer Cooperatives 583583 A7 -------B7 Fifth, the three stages shown in the description of the invention (> >), therefore, the block diagram 500 of FIG. 5 can be used to point out the necessary elements required for the embodiment of the present invention. For brevity, unrelated components in the microprocessor 500 are not shown. [0049] Please refer to FIG. 6, which is a block diagram of a specific embodiment of an extended preamble for selectively controlling writeback of a condition code according to the present invention. The extension condition code designation element preamble 600 is an 8-bit extended preamble 600, and has a write-back control (a pool back contrd 'including a plurality of bits (starting at bit 0)). wbc) field 602 and a spare field 601 containing the remaining bits (ending at bit 7). In a specific embodiment, the wbc block 602 includes bit 0, and according to the instruction of the present invention, the extended execution logic disables the write-back of all condition codes. Another specific implementation side uses the lower five bits of the preamble 600 as the WBC bit 602, where each bit corresponds to a negative sign prohibition condition, a carry condition, an overflow condition, and a zero value condition, respectively. And write-back of parity conditions. [0050] Please refer to FIG. 7, which is a detailed block diagram of the translation stage logic 700 in the microprocessor of FIG. 5. The translation stage logic 700 has an instruction buffer 704, which according to the present invention provides extended instructions to the translation logic 705. Then, the logic 705 handle is connected to a machine-specific register 702 with an extended feature booth 703. The translation logic 705 has a translation controller 706, which provides a disabling signal 707 to an escape instruction detector 708 and an extended preamble decoder 709. The escape instruction debt tester 708 is connected to the extended preamble decoding $ 709 and an instruction calculus horse training. Extended preamble decoder 709 and instruction decoding logic 71. Access to a control read-only memory (ROM) 711, which stores a pattern corresponding to some extended instructions --II — I * — III ----- --I ^ — — — — — — — II ^^ 1. (Please read the precautions on the back before filling out this page) 23 583583 A7 、 Instructions of the invention (4) Micro instruction sequence of the board (te_ate). The translation logic 705 φ includes a microinstruction buffer m ′ which has an opcode extension term claw, a microop field 7U, a destination block 715, a source bit 716, and a bit 7. [〇 Gang operation, during the start-up of the microprocessor, the state of the extended field 703 in the specific register of the machine is determined by the signal activation state a (n)) 7 (31 to indicate the specific micro-processing) If no, transfer to n lines of the extended condition execution instruction of the present invention. In a 701m signal 701 from a feature control register (not shown on the picture;) * out 'the feature control register is read-added to the silk array ― (Not shown). The status of the machine-specific temporary extension feature stop 703 is sent to the translation controller 鄕. = 制 ^ == 令: Puncher 7.4 Extraction and selection of raw materials and writing conditions control translation Providing such control features can allow supervised programs (such as BI⑹ extension_signs. If the extension feature is disabled, the operation with the mark selected as the extension feature is only used.) Dirty, the translation is performed. A specific x86 == No: The instruction is abnormal (exeeptlGn). _ ”Come. The escape instruction _7708 therefore specifies the extension condition code after the translation / decoding mark is detected to specify the decoder's operation. / Explanation 22 is applicable to China Please read the selection of the back body Then calculate the F1H translation items

I I I I I 訂 罈 583583 經濟部智慧財產局員工消費合作社印製 A7 B7 五、發明說明(χψ) 時’致能指令解碼器710。某些特定指令將導致對控制R〇M 711的存取,以獲取對應之微指令序列樣板。微指令緩衝器 712之運异碼延伸項欄位713由前置碼解碼器7〇9進行組 癌,以指定複數個條件碼之一子集合,其中該些條件碼之 更新係在一對應運算之結果產生後被致能/除能。其他緩衝 器欄位714-717則指定該對應運算,並由指令解碼器71〇進 行組悲。經過組態之微指令712被送至一微指令符列(未 顯示於圖中),由處理器進行後續執行。 [0052] 現請參閱圖八,其為圖五微處理器内之延伸執 行邏輯800的方塊圖。該延伸執行邏輯8〇〇具一算術邏輯 單兀(anthematic logic unit,ALU) 809,其經由一結果匯流 排808耦接至條件碼回寫控制器8〇7。條件碼回寫控制器 807則耦接至一條件旗標儲存機制8〇6,或稱條件旗標暫存 器 806。兩運算元 0PERAND 丨與 〇pERAND 2,由 ALU 8〇9 從運算70暫存器801與8〇5提取出來。一微指令暫存器8〇2 提供一微指令給ALU 809與條件碼回寫控制邏輯8〇7。微 指令暫存器802具有一運算碼延伸項攔位8〇3與一其餘欄 位804。ALU809並另外麵接至一結果暫存器81Q。 [0053] 運作上,當一使用選擇性條件碼回寫控制特徵 之延伸指令,依本發明被轉譯成一微指令序列時,延伸微 指令以及暫存器801與8〇5内之可用運算元皆經由微指令 暫存器802,被送至延伸執行邏輯_。運算碼延伸項搁位 803指定了複數個條件碼之一子集合,其中該些條件碼之更 新係在一對應至由其餘欄位804所指定、使用運算元g〇i --------------— (請先閱讀背面之注意事項再填寫本頁) 訂·. - — II. 25 583583 A7I I I I I Order 583583 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs A7 B7 V. When the description of the invention (χψ) 'enables the instruction decoder 710. Certain specific instructions will cause access to control ROM 711 to obtain the corresponding microinstruction sequence template. The different code extension item field 713 of the microinstruction buffer 712 is grouped by the preamble decoder 709 to specify a subset of the plurality of condition codes, wherein the update of the condition codes is a corresponding operation. The result is enabled / disabled after the result. The other buffer fields 714-717 specify the corresponding operation, and the instruction decoder 710 performs group operations. The configured micro-instruction 712 is sent to a micro-instruction string (not shown in the figure) for subsequent execution by the processor. [0052] Please refer to FIG. 8, which is a block diagram of the extended execution logic 800 in the microprocessor of FIG. 5. The extended execution logic 800 has an arithmetic logic unit (ALU) 809, which is coupled to a condition code write-back controller 807 via a result bus 808. The condition code write-back controller 807 is coupled to a condition flag storage mechanism 806, or a condition flag temporary register 806. The two operands 0PERAND 丨 and 〇pERAND 2 are extracted by the ALU 809 from the operation 70 registers 801 and 805. A micro-instruction register 802 provides a micro-instruction to the ALU 809 and the condition code write-back control logic 807. The micro-instruction register 802 has an operation code extension entry block 803 and a remaining field 804. ALU809 is also connected to a result register 81Q. [0053] In operation, when an extended instruction using the selective condition code write-back control feature is translated into a micro instruction sequence according to the present invention, the extended micro instruction and the available operands in the registers 801 and 805 are both Via the micro-instruction register 802, it is sent to the extended execution logic_. The operation code extension item slot 803 specifies a subset of a plurality of condition codes, and the update of the condition codes corresponds to that specified by the remaining field 804 using the operand g0i ------ --------— (Please read the notes on the back before filling out this page) Order ·.--II. 25 583583 A7

583583 經濟部智慧財產局員工消費合作社印製 A7 五、發明說明(必)583583 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs A7 V. Invention Description (Required)

態為指出該延伸前置碼所指定該些條件碼之該古 程接著進行至方塊912 〇 / I 口 H _ί於方塊912中,該指令之其餘部分(如前置碼 項目、運异碼、位址指定疋)被轉譯/解碼,以靖所要 行的運异及關聯運算兀的屬性。流程接著進行至方塊則。 =〇]於方塊9匕中’一微指令序列的其餘搁位被組 怨為才曰疋所指定的運算及其運算S屬性。流程 方塊910 〇 隣]於=塊训中,該微指令序列,其包含方塊_ 中所組態之運算碼延伸項欄位以及方塊9Μ中所組声之其 餘欄位,被送至-微指令件列,由微處理器執行^接 著進行至方塊918。 [0062]於方塊918中,該微指令序列由本發明之延伸 條件執行邏輯進行提取。流程接著進行至方塊92〇。 、_3]於方塊㈣中’㈣執行邏輯執行該指定運算, 並產生該結果。流程接著進行至判斷方塊 _於判斷方㈣中,條件碼二懈 發明排除了該子集合的邊界條件評估與條件碼更新,1中 該子集合係由微指令延伸賴位於方塊⑽中所指定對 於被排除之該子集合,流料行至方塊926。對於未 評估/更新之其餘條件碼,流程則進行至方塊924。 [0065]於方塊924中’評估該結果,並依據這些其餘 的條件碼來更新條件碼暫存器之對應位元^ 進行 至方塊926。 τ 27 本纸i尺没顧中國國家標準(cns)A4規格(2ig χ 297公爱「 (請先閱讀背面之注意事項再填寫本頁)The state is to point out that the ancient process of the condition codes specified by the extended preamble then proceeds to block 912 〇 / 口 H _ί in block 912, the rest of the instruction (such as the preamble item, transport different code, Address designation 疋) is translated / decoded to the properties of the operations and associated operations required by Jing. The flow then proceeds to block rule. = 〇] In block 9, the remaining positions of a sequence of microinstructions are grouped into the operation specified by Cai Yue and its operation S attribute. Flow block 910 〇 Neighbor] In = block training, the micro instruction sequence, which contains the opcode extension field configured in block _ and the remaining fields of the group of sounds set in block 9M, is sent to the -micro instruction The list of items is executed by the microprocessor and then proceeds to block 918. [0062] In block 918, the microinstruction sequence is extracted by the extended conditional execution logic of the present invention. The flow then proceeds to block 92. , _3] Execute the logic in the box 运算 to execute the specified operation and produce the result. The flow then proceeds to the judgment block. In the judgment block, the condition code two innovations exclude the boundary condition evaluation and condition code update of the sub-set. The sub-set in 1 is extended by microinstructions and is located in the block. The excluded subset flows to block 926. For the remaining condition codes that are not evaluated / updated, the flow proceeds to block 924. [0065] In block 924, the result is evaluated, and the corresponding bits of the condition code register are updated according to the remaining condition codes, and proceed to block 926. τ 27 The paper ruler does not take into account the Chinese national standard (cns) A4 specification (2ig χ 297 public love "(Please read the precautions on the back before filling this page)

LSJ 583583 A7 五、發明說明(>7) [0066]於方塊926申,本方法完成。 [_7]㈣本發明及其目的、特徵與優點已詳細敛 述,其它實補亦可包含在本發明之範_。例如,本發 明已就如下的技術加以敘述··利用已完全佔用之指a集架 構内-單-、未使用之運算碼狀態作為標記,以指:其後 之延伸特徵前置碼6但本發明的範圍就任一方面來看,並 不限於已完全個之指令餘構,絲使_指令,或是 單-標記。相反地’本發明涵蓋了未完全映射之指令集、 具已使用運算碼之實施例以及使甩—個以上之指令標記的 實施例。例如’考慮-沒有未使崎算碼狀態之指令集架 構。本發明之-具體實施例包含了選取一作為逸出標記之 運算碼狀態’其中選取標準係依市場因素而決定。另一且 體實施_包含使用運算歇—特殊組合作為標記,如運 算碼狀態簡的_歧。,本伽之本㈣在於使 用-標記序列’其後則為一 n位元之延伸前置碼,可允許 程式員對於-延伸指令之其餘部分所指定運算的執行結 果,致能/除能其相關條件碼的評估與更新。 [0068]此外,本發明已藉由—具有—組條件碼或旗標 之微處理器來作為範例,該組條件碼係指出一所產生結果 之邊界條件,包括同位、溢位、正負號及零。雖然這些類 的條件沾在今日仍廣為使用,但本發明並不僅限於應 用在這些類型的條件。例如,本發明另外之實施例即包含 其他的指定條件,像是一特殊暫存器内容的狀態、一通訊 崞或其他I/O裝置是否使用巾、是否有可狀記憶體或快取 28 本紙張尺度巾關家標準格咖χ挪公^__-^ —----------Aw> Μ — (請先閲讀背面之注意事項再填寫本頁) 訂: 經濟部智慧財產局員工消費合作社印製 583583 A7 五、發明說明(β) 記憶體空間等等。 [0069]再者,雖然上文係利用微處理器為例來解說本 發明及其目^的、特徵和優點,熟習此領域技術者仍可察覺, 本發明的範H並不限於微處理器的_,而可涵蓋所有形 式之可料鱗置,如訊魏理H驗繼(industrial controller)、陣列處理器及其他同類裝置。 处:之以上所述者,僅為本發明之較佳實施例而已,當 不=以之限疋本發明所實施之範圍。大凡依本發明申請專 ^範f所作之均特化與料,皆應仍屬於本發明專利涵 蓋鼙圍内,謹5月貴審查委員明鐘,並祈惠准,是所至禱。 . f -------I----訂· ----I--- (請先閲讀背面之注意事項再填寫本頁) 經濟部智慧財產局員工消費合作社印製 29LSJ 583583 A7 V. Description of the invention (> 7) [0066] At block 926, the method is completed. [_7] The present invention and its objects, features, and advantages have been summarized in detail, and other practical supplements may also be included in the scope of the present invention. For example, the present invention has been described in terms of the following techniques: • The use of the fully-occupied finger a set of framework-single-, unused opcode status as a mark to refer to the following extended feature preamble 6 but this In any aspect, the scope of the invention is not limited to the complete instruction structure, the instruction, or the single-marker. On the contrary, the present invention encompasses incompletely mapped instruction sets, embodiments with used opcodes, and embodiments that flag more than one instruction. For example, 'consider-there is no instruction set architecture that does not render the state of Saki. A specific embodiment of the present invention includes selecting an operation code state as an escape mark ', wherein the selection criterion is determined according to market factors. Another implementation includes the use of arithmetic breaks—special combinations as markers, such as ambiguity in the state of arithmetic codes. The essence of Benjamin lies in the use of the -tag sequence 'followed by an n-bit extended preamble, which allows the programmer to enable / disable the execution result of the operation specified by the rest of the -extended instruction. Evaluation and update of related condition codes. [0068] Furthermore, the present invention has been exemplified by a microprocessor having a set of condition codes or flags, the set of condition codes indicating a boundary condition of a result, including parity, overflow, sign and zero. Although these types of conditions are still widely used today, the invention is not limited to application to these types of conditions. For example, other embodiments of the present invention include other specified conditions, such as the status of a special register content, whether a communication card or other I / O device uses a towel, whether there is a shapeable memory or a cache. Paper scale towels, family standard, standard coffee, χ Norwegian, ^ __- ^ —---------- Aw > Μ — (Please read the notes on the back before filling this page) Order: Intellectual Property Bureau, Ministry of Economic Affairs Printed by employee consumer cooperatives 583583 A7 V. Description of invention (β) Memory space and so on. [0069] Furthermore, although the above uses the microprocessor as an example to explain the present invention and its objects, features, and advantages, those skilled in the art can still perceive that the scope of the present invention is not limited to the microprocessor _, And can cover all forms of expected scales, such as Xunweili H's industrial controller, array processor and other similar devices. Where: The above is only a preferred embodiment of the present invention, and the scope of implementation of the present invention is not limited. All the specialization and materials made by the applicant for the application of the present invention should still fall within the scope of the patent cover of the present invention. I would like to ask your review committee members in May and pray for your approval. f ------- I ---- Order · ---- I --- (Please read the notes on the back before filling out this page) Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 29

Claims (1)

583583 A8 B8 C8 D8 2. 3. 六、申請專利範圍 一樘用以選擇性地控制複數個條件碼回寫之微處理器裝 置,包含: 一轉譯邏輯,用以將一延伸指令轉譯成對應之微指令, 其中該延伸指令包含: 延伸如置碼,用以將該些條件碼之回寫除能,該 些條件碼係對應於一指定運算之結果;以及 一延伸前置碼標記,用以指出該延伸前置碼,其中 該延伸前置碼標記係一微處理器指令集内另一依 據架構所指定之運算碼;以及 -延伸執行邏輯,接至該轉譯邏輯’用以接收該對應 之微指令’產生該、结果’並除能該些條件碼之回寫二 如申请專利範圍第1項所述之微處理器裝置, 伸指令更包含該指令集根據架構所指定之複數個項目/。 如申清專利範圍第2項所述之微處理II裝置,1 娜架構所措定之項目包含一運算碼項目置用== 指定運算。 μ 4.如申請專利範圍第〗項所述之微處理器裝置,其中該延 伸f置碼包含複數個位元,且其令該些位元之每一邏輯 狀態指不該微處理器將複數個結果條件之一子集合的回 寫予以除能,其巾條件碼包含雌結果條件之 殊子集合。 5·如申W專利補第4項所述之微處顧裝置,其中該此 結果條件包含溢位、進位、雜零、帶負肋及同位^ ____ _ JU 本紙張尺度適用中國國家標準(“Μ4規格(21G x 297公£ --------1-----AWI --------^---------. (請先閱讀背面之注意事項再填寫本頁) A8 B8 C8 D8 、申請專利範圍 6. 如申請專利範圍第5項所述之微處理器裝置,其中該些 結果條件係儲存於該微處理器之一旗標暫存器中。 7. 如申請專利範圍第1項所述之微處理器裝置,其中該延 伸前置碼包含8個位元。 8·如申凊專利朗第1項所述之微處職裝置,其中該指 令集包含x86指令集。 •如申凊專利範圍第8項所述之微處理器裝置,其中該延 伸如置碼彳示圮包含X86指令集之運算碼F1(ICEBKpT)。 10·如申請專利範圍第i項所述之微處理器裝置,其中該對 應之微指令包含一微運算碼欄位與一微運算碼延伸項欄 位。 如申π專利範圍第1〇項所述之微處理器裝置,其中該延 伸執1 亍邏輯使用該微運算碼延伸項欄位,以決定要將回 之該絲件碼,且其中該延伸執行邏輯使用該微 f异竭攔仙決定所魏行之該指定運算,藉以產生該 結果。 =申明專利|&圍第11項所述之微處釋裝置,其中該延 伸執行邏輯包含: 條件碼回寫控制H,域為隨著複數個運算結果之產 ^ ’更新—條件碼暫存器,並組態為隨著該結果之 產生,排除該些條件碼之回寫。 13,=ffm圍第1項所述之微處理器裝置,其中該轉 #邏輯包含: —逸出指令偵測邏輯,用於偵測該延伸前置碼標記;以 (請先閱讀背面之注意事項再填寫本頁) 裝 經濟部智慧財產局員工消費合作社印製583583 A8 B8 C8 D8 2. 3. Scope of patent application-A microprocessor device for selectively controlling the writing back of a plurality of condition codes, including: a translation logic for translating an extended instruction into the corresponding A micro instruction, wherein the extended instruction includes: an extension such as a code, which is used to disable write-back of the condition codes, the condition codes corresponding to a result of a specified operation; and an extended preamble mark, which is used to: Point out the extended preamble, where the extended preamble mark is another operation code specified by the architecture in a microprocessor instruction set; and-extended execution logic, connected to the translation logic 'to receive the corresponding The micro-instruction 'generates the result' and disables the write-back of the condition codes. The microprocessor device described in item 1 of the scope of the patent application, the extended instruction further includes a plurality of items specified by the instruction set according to the architecture / . As described in claim 2 of the microprocessing II device of the scope of the patent, the items determined by the 1-na architecture include an opcode item using == the specified operation. μ 4. The microprocessor device as described in the scope of the patent application, wherein the extended f code includes a plurality of bits, and it causes each logical state of the bits to indicate that the microprocessor will The write-back of one of the sub-sets of the result condition is disabled, and the condition code includes a special sub-set of the female result condition. 5. The micro-processing device as described in the fourth item of the patent application of the W patent, wherein the result conditions include overflow, carry, miscellaneous, negative ribs and parity ^ ____ _ JU This paper standard applies to Chinese national standards (" Μ4 specification (21G x 297) £ -------- 1 ----- AWI -------- ^ ---------. (Please read the precautions on the back first (Fill in this page again) A8 B8 C8 D8, patent application scope 6. The microprocessor device described in item 5 of the patent application scope, wherein the result conditions are stored in a flag register of the microprocessor 7. The microprocessor device according to item 1 of the scope of the patent application, wherein the extended preamble contains 8 bits. 8. The micro-service device according to item 1 of the patent claim, wherein the The instruction set contains the x86 instruction set. • The microprocessor device as described in item 8 of the patent application scope, wherein the extension is shown as a code, and contains the operation code F1 (ICEBKpT) of the X86 instruction set. 10 · If applying for a patent The microprocessor device described in item i of the scope, wherein the corresponding microinstruction includes a micro-op field and a micro-op extension field. The microprocessor device described in Item 10, wherein the extension execution logic uses the micro-operation code extension field to determine the silk code to be returned, and wherein the extension execution logic uses the The micro-f exhaustion determines the specified operation performed by Wei Xian to produce the result. = Declaration of the patent | & The micro-processing device described in item 11, wherein the extended execution logic includes: condition code write-back control H, The field is produced as a result of a plurality of operation results ^ 'Update-condition code register, and is configured to exclude the write-back of those condition codes as the result is generated. 13, = ffm described in item 1 The microprocessor device, wherein the transfer # logic includes: — escape instruction detection logic for detecting the extended preamble mark; (please read the precautions on the back before filling this page) to install the wisdom of the Ministry of Economic Affairs Printed by the Property Agency Staff Consumer Cooperative 583583583583 583583 經濟部智慧財產局員工消費合作社印製 A8 B8 C8 D8 六、申請專利範圍 • 其他延伸指令項目,組態為指定該指定運算。 17.如申請專利範圍第16項所述之延伸機制,其中該其他延 仙令項目魏照該既有微纽雜令赫構加以格式 化。 18· =請專利細第u斯述之延伸鋪,其中該些條件 旗標包含溢位、進位、等於零、帶負號以及同位。 19.如申請專利範圍第15項所述之延伸機制,其中該^位元 之延伸控制前置碼包含8個位元。 處理裔指令集係x86微處理器指令集。 儿如申請專利範圍第20項所述之延 運算碼包括撕微處理器指令集中之ICEB^m (即運算碼F1)。 ^ ^ 22tr專利範圍第15項所述之延伸機制,其中該轉譯器 一 肋姻觀料令狀該選取之 m置碼解碼器’_至該逸出指令偵測器,用以 隼^㈣’並產生指定該子 集口之該勸"序列内—微運算碼延伸項 23. —種為一既有微處理器指令集增添娌 、。 能力的指令集延伸模組,該條件碼回^條:碼回寫 逸出標記,用以除能該些條件喝之敢曰^70輕接至該 件竭的回寫。包含: 寫’並致能其餘條 33 私紙張尺Μ財國國家標準(CNS)A4規格(210 X 297公f ^ ^ IIAWI --- (請先閱讀背面之注意事項再填寫本頁) 訂: 4 583583583583 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs A8 B8 C8 D8 6. Scope of Patent Application • Other extended instruction items are configured to specify the specified operation. 17. The extension mechanism as described in item 16 of the scope of patent application, wherein the other Yanxianling project Wei Zhaoxian has an existing micro-news order structure. 18 · = Please refer to the extension of the patent, where the condition flags include overflow, carry, equal to zero, negative sign, and parity. 19. The extension mechanism according to item 15 of the scope of patent application, wherein the ^ bit extension control preamble includes 8 bits. The processing instruction set is the x86 microprocessor instruction set. The extended operation code described in item 20 of the scope of patent application includes ICEB ^ m (that is, operation code F1) in the microprocessor instruction set. ^ ^ The extension mechanism described in item 15 of the scope of 22tr patent, wherein the translator is a data watch, the selected m-coded decoder '_ to the escape instruction detector, and is used for 隼 ^ ㈣' And generate the persuasion "in-sequence—micro-operation code extension term 23.” that specifies the subset port—a kind of addition to an existing microprocessor instruction set. Capability instruction set extension module, the condition code is back to ^ bar: the code is written back. The escape mark is used to eliminate those conditions that can be dared to write ^ 70 to write back to the exhausted. Contains: Write 'and enable the remaining 33 private paper ruler MUSIC National Standard (CNS) A4 specifications (210 X 297 male f ^ ^ IIAWI --- (Please read the precautions on the back before filling this page) Order: 4 583583 經濟部智慧財產局員工消費合作社印製 -逸出標記,由—轉譯邏輯接收,並指出—制指令之 =隨部分係指定了—微處理騎要執行之一延伸運 异,其中該逸出標記為該既有微處理器指令集内之 一第一運算碼項目; 八 一= 碼回寫指定元,纖該逸出標記,且為該附隨 4刀其中之肋指定複數個條件碼巾關聯於該 延伸運算之結杲者;以及 條件馬回寫控制器,耦接至該轉譯邏輯,用以除能關 聯於該延伸運异結果之條件碼的回寫,並致能其餘 該些條件碼的回寫。 4.如申π專利範圍第23項所述之指令集延伸模組,其中該 附Ik部分之其餘部分包含-第二運算碼項目與選用之複 數個位址指定元項目。 25·如申請專利範圍第23項所述之指令集延伸模級,盆中該 條件碼回寫指定元包含一8位元的資料項目。^ " 26·如申請專纖㈣23項賴之齡祕倾組,其中該 既有微處理器指令集係x86微處理器指令集。 27·如申請專利範圍第26項所述之指令集延伸模組,其中該 第一運算碼項目包含x86微處理器指令集中之IC£ BKPT運算碼項目(即運算碼π)。 2 8 ·如申請專利範圍第2 3項所述之指令集延伸模組,其中該 轉譯邏輯將該逸出標記與該附隨部分轉譯成對應的微指 令,該對應的微指令係指示一延伸執行邏輯去執行誃 伸運算。 ---hill·----··-裝--- (請先閱讀背面之注意事項再填寫本頁} · 34 88899 ABCD V申請專利範圍 29.如申請專利範圍第23項所述之指令集延伸模組,其中該 轉譯邏輯包含: -逸出標記偵測邏輯,肋偵測該逸出標記,並指示該 附隨部分的轉譯動作需依據延伸轉譯常規 (conventions);以及 頁 一解碼邏輯,減至該逸出標記_邏輯,用以依據該 无有微處理器々曰令集之常規,執行微處理器指令的 轉譯動作,並依據該延伸轉譯常規執行該對應指令 之轉譯’以允許該些條件碼之麵性回寫。 2=—微處㈣齡躺妓,吨财程式化之 …果條件碼回寫能力,該方法包含·· 提供一延t指令,該延伸指令包含-延伸標記及-條件 别置碼,其中該延伸標記係該微處理器指令 集其中一第一運算碼; I m 透3斤=碼回寫前置碼與該延伸指令之其餘部分指 運算,其令_賤運算結果之選 取條件碼的回寫將被禁止;以及 、 執2運异以產生該結果,且禁止該選取條件竭之回 31打如=算專=包第r項所述之方法,其令該指定所要執 第定之動作使用了該微處理器 32·如申請她_G項所叙奴,㈣提供延伸指 本紙張尺賴财 (210x 297 公 J7" 制 '申請專利範圍 令之動作包含使用-8位元大小之 回寫前置碼進行組態。 崎該條相 33.如申請專利範圍第3〇項所述之方法,其巾該接租 =之動作包含從x86微處理器指令集選取該第一 3《^申請翻翻第%項所述之方法,其中該選取第一運 =碼的動作包含選取_ ICE Βκρτ運算碼(即運算碼 F〇作為該延伸標記。 ” 35. 如申請專利範圍第3〇項所述之方法,更包含: 將該延,令_成微指令, J 4示―延伸執行邏輯麵止該選取條件碼之回 36. 如申請專利範圍第35項 兮 令的動作包含·· 11之料,其巾譯延伸指 於一轉譯邏輯内,偵測該延伸標記;以及 依伸轉譯規則解碼該條件指定元前置碼與該延伸 二刀’該延伸轉譯規則為—既有之微處 理4構k供依轉譯規則所無法提供之可程式 化回寫控制能力。Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs, the escape mark is received by the translation logic, and it is stated that the manufacturing instruction = specified by the system is one of the extended operations to be performed by the micro processor, and the escape mark One of the first opcode items in the existing microprocessor instruction set; Bayi = code write-back designation element, the escape tag, and multiple condition code towel associations for the ribs attached At the end of the extended operation; and a conditional horseback write controller, coupled to the translation logic, to disable the writeback of condition codes associated with the extended result, and enable the remaining condition codes Write back. 4. The instruction set extension module as described in item 23 of the patent scope of claim π, wherein the rest of the attached Ik part includes a second opcode item and a plurality of optional address designation meta items. 25. According to the instruction set extension module described in item 23 of the scope of patent application, the condition code in the basin writes back the specified element including an 8-bit data item. ^ " 26. If you apply for 23 items of Lai Zhiling's Secret Group, the existing microprocessor instruction set is the x86 microprocessor instruction set. 27. The instruction set extension module described in item 26 of the scope of patent application, wherein the first opcode item includes an IC £ BKPT opcode item (ie, opcode π) in the x86 microprocessor instruction set. 2 8 · The instruction set extension module described in item 23 of the scope of patent application, wherein the translation logic translates the escape mark and the accompanying part into corresponding micro instructions, and the corresponding micro instruction indicates an extension Execute logic to perform stretching operations. --- hill · ---- ·· -install --- (Please read the precautions on the back before filling out this page} · 34 88899 ABCD V Application for Patent Scope 29. Instructions as described in Item 23 of Patent Application Scope An extension module, wherein the translation logic includes:-escape tag detection logic, which detects the escape tag and instructs the accompanying part to perform translation operations in accordance with extended translation conventions; and page-one decoding logic , Reduced to the escape tag _ logic to execute the translation action of the microprocessor instruction according to the routine of the non-microprocessor command set, and execute the translation of the corresponding instruction according to the extended translation routine to allow the The conditional write-back of these condition codes. 2 = —the prostitutes with a small amount of age, stylized tons of fruit ... The condition code write-back capability, the method includes ... providing an extended t instruction, the extended instruction includes-extended Tag and conditional code, where the extended tag is one of the first operation codes of the microprocessor instruction set; I m through 3 pounds = code write-back preamble and the rest of the extended instruction refers to the operation, which makes _Selection condition code for base operation result Write back will be prohibited; and, execute 2 different to produce the result, and prohibit the exhaustion of the selection conditions. 31 hit the method described in = calculation = package item r, which causes the specified action to be performed. Using the microprocessor 32. As described in the application of her slaves in item G, I provided an extension of the paper rule Lai Choi (210x 297 J7 " system of patent application order actions including the use of 8-bit size return Write the preamble to configure. Saki this article 33. The method described in item 30 of the scope of patent application, the action of taking the lease = includes selecting the first 3 "^ from the x86 microprocessor instruction set The application refers to the method described in item%, wherein the action of selecting the first operation code includes selecting the _ICE Βκρτ operation code (that is, operation code F0 as the extension mark.) The method further includes: extending the delay to make micro instructions, J 4 shows-extending the execution logic to stop the return of the selection condition code 36. If the action of the 35th order in the scope of the patent application includes the ... The extension of translation is to detect in a translation logic. Extension tag; and decoding the condition-specified meta-prefix and the extension two-knife according to the extension translation rule. The extension translation rule is-the existing microprocessing structure is provided for programmable write-back control that cannot be provided by the translation rule. ability.
TW091116956A 2002-05-09 2002-07-30 Apparatus and method for selective control of condition code write back TW583583B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US10/144,593 US7185180B2 (en) 2002-04-02 2002-05-09 Apparatus and method for selective control of condition code write back

Publications (1)

Publication Number Publication Date
TW583583B true TW583583B (en) 2004-04-11

Family

ID=22509266

Family Applications (1)

Application Number Title Priority Date Filing Date
TW091116956A TW583583B (en) 2002-05-09 2002-07-30 Apparatus and method for selective control of condition code write back

Country Status (2)

Country Link
CN (1) CN100590591C (en)
TW (1) TW583583B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8504807B2 (en) 2009-12-26 2013-08-06 Intel Corporation Rotate instructions that complete execution without reading carry flag

Also Published As

Publication number Publication date
CN1414469A (en) 2003-04-30
CN100590591C (en) 2010-02-17

Similar Documents

Publication Publication Date Title
TW591527B (en) Apparatus and method for extending a microprocessor instruction set
TWI314284B (en) Apparatus and method for instruction-level specification of floating point format
JP5646656B2 (en) Mapping between registers used by multiple instruction sets
Seal ARM architecture reference manual
Sites et al. Alpha AXP architecture reference manual
Sloss et al. ARM system developer's guide: designing and optimizing system software
KR100327777B1 (en) Data processing device using multiple instruction sets
TWI242158B (en) Apparatus and method for data processing using multiple instruction sets and computer readable recording medium for controlling the apparatus
TWI302263B (en) Method and apparatus to improve execution of a stored program
TW525087B (en) Processor architecture for executing two different fixed-length instruction sets
CN104636116B (en) Complete its recursion instruction executed without reading carry flag
CN105164650A (en) Systems, methods and apparatus for improving throughput of contiguous transactional memory regions
TW201224740A (en) Debugging of a data processing apparatus (2)
TW201241747A (en) Efficient conditional ALU instruction in read-port limited register file microprocessor
TW201346715A (en) Method and apparatus for decimal floating-point data logical extraction
CN104067224B (en) Instruction execution that broadcasts and masks data values at different levels of granularity
TWI559219B (en) Apparatus and method for selecting elements of a vector computation
TW444179B (en) A data processing system having an apparatus for exception tracking during out-of-order operation and method therefor
TW583583B (en) Apparatus and method for selective control of condition code write back
TW569140B (en) A method and system for selecting and using source operands in computer system instructions
TW569136B (en) Apparatus and method for conditional instruction execution
TWI220042B (en) Non-temporal memory reference control mechanism
TWI230356B (en) Apparatus and method for extending address modes in a microprocessor
TW561406B (en) Apparatus and method for selective control of results write back
TWI245221B (en) Apparatus and method for selective memory attribute control

Legal Events

Date Code Title Description
MK4A Expiration of patent term of an invention patent