200921388 九、發明說明： 【电明所屬之技術領域】 特別涉及一種儲存 本毛明涉及—種靜態記憶體領域 媒體管理方法。 【先前技術】 現有對儲存媒體的管理方法一 輯位址的斟昭本^ 瓜疋建立物理位址與邏 第一圖所示的邏輯塊與物理物 …二： 查找與邏輯塊對應的物理塊， 田有物理塊成為壞塊時可將 ^,, ^ θ +ju + 了j村/、射除不用。然而隨著儲存婵 體：谷置越來越大，對照表的資料量也越來越大，要在： 入^系統的記憶體中保存整個儲存媒體的對 〜 不實際的。 戍卞疋 —現有-種分段管理技術，對儲存媒體的物理塊分段進 仃g理&第一圖所不的邏輯塊與物理塊對照關係示意 圖’設儲存媒體有01個物理塊，分為L個區域，每個^域 N塊刀別對每個物理塊區域做物理位址與邏輯位址的對 照表。只有使用到相應邏輯位址區域的時候才在記憶體中 建立與該邏輯位址所在區域相應的對照表，這樣就在記情 體空間和效率之間取得了-個折衷。但這也引起了另—個 問題’因為儲存空間被物理分割，各區域之間就被隔絕。 «儲存媒體的實際使用特點，某些區域的物理塊經常被 使用，另外一些區域的物理塊卻很少使用，於是會導致使 用的不均衡，某些物理塊已經達到壽命的極限，而某些物 200921388 =卻還很年輕。當一些物理塊已經損壞時，儲存資料不 可罪必須重新對整個儲存 、 除出來才能再次❹。4進仃Μ ’把壞物理塊剔 針對以上問題，也有另一 管理技術，當—個區域二現!提出基於上述分段 域與其他區域中的資料進行交換將整個區 上述問題，但彦味丄計 技術。该技術可以解決 時間兩要拷貝女旦域的效率低下的問題，會有一段 ’ 貝大1的資料，導致儲存媒體性能不穩定。 【發明内容】 理大2:二的在於提供-種儲存媒體管理方法，用於管 理大谷1儲存媒體。 HW、吕 本發明提供一種儲存媒 理邏輯位址，建立P、—「t 法’包括：分區域管 關係的步驟；將至少:;區車耳位址與物理位址的對應 係載入在記憶體中的步；與物理位址的對應關 理位址㈣應關係執行訪問命乂令的㈣£域邏輯位址與物 ’上述分區域管理邏輯位址，建立至少一區域 與物理位址的對應關係的步驟包括：將邏輯位址 區域;對至少-邏輯區域中的邏輯位址， t-f m m ^ ^ ^ a 0任—物理位址之間的對應關係；將 對應關係形成至少兩個區域對照表。 入在址與物理位址的對應關係載 乂驟包括：根據訪問命令計算指定邏輯塊 200921388 屬於哪一區i或；判冑記憶體中的區域對照表是否對應指定 这輯塊所屬區域，將指定邏輯塊所屬區域的區域對照表载 入到記憶體中，替換記憶體中的區域對照表。 上述將至少一區域邏輯位址與物理位址的對應關係载 入在記憶體中的步驟還包括：將其中—區域對照表保存在 記憶體中的步驟。 優選地，分區域管理邏輯位址，建立至少一區域邏輯 〇位址與物理位址的對應關係的步驟包括：將邏輯位址劃分 為至v兩個區域，對各區域中的邏輯位址，建立與儲存媒 體任一物理位址之間對應關係，形成完整對照表。’、 上述將至少一區域邏輯位址與物理位址的對應關係載 入在兄憶體中的步驟包括：將完整對照表保存在外部儲存 媒體^將完整對照表的位址保存在記憶體中；根據訪問 命令計算指定邏輯塊屬於哪一區域；根據完整對照表的位 址找到兀整對照表；將完整對照表中，指定邏輯塊所屬區 I 域的部分對照表載入到記憶體中。 上述將對照表保存在外部儲存媒體中是指保存在至少 物理塊中，或是保存在至少一物理塊的冗余區中。 優選地，上述儲存媒體管理方法還包括定義至少一交 換塊建立义換塊的邏輯位址與物理位址對應關係的步 驟。 上述根據至少一區域邏輯位址與物理位址的對應關係 執灯訪問命令的步驟包括：根據至少一區域邏輯位址與物 理位址的對應關係’查找與寫命令指定邏輯位址對應的物 8 200921388 理位址’將寫命令要寫入的新資料寫到與交換塊對應的物 理位址中’將與寫命令指定邏輯位址對應的物理位址中不 被寫命令覆蓋的原有資料搬遷到與交換塊對應的物理位址 中’建立父換塊與寫命令指定邏輯位址對應的物理位址之 間的對應關係’建立寫命令指定邏輯位址與交換塊對應的 物理位址之間的對應關係；擦除與交換塊對應的物理位 址。 上述根據至少一區域邏輯位址與物理位址的對應關係 執行訪問命令的步驟還包括：更新至少一區域邏輯位址與 物理位址的對應關係，更新對照表的步驟。 ^ A上述將與寫命令指定邏輯位址對應的物理位址中不被 寫〒令覆盍的原有資料搬遷到與交換塊對應的物理位址中 ^.將與寫命令指^邏輯位址對應的物理位址與交換塊 2 =物理位址比較’得到不被寫命令^寫人新資料的 ^位^ ;將未指定位址中的資料複製到與交 的物理位址中。 卞述儲存媒體管理方法單獨使用一交換塊 用至少一交換塊。 ^ 树明通過將儲存媒體的邏輯位址劃分多個區域，各 =別組織獅’實現佔用極少量的記憶體完成大容 里=子媒體的官理；本發明還利用交 間的物理塊相互交換，從而消除了£戏門j各㈣域之 題，同時避免了整個區域資料嶋而=不均衡的問 問題，安全穩定，易於實：祕遷而導致的性能不穩定 200921388 【實施方式] ☆本發明目的的實現、功能特點及優點將結合實施例， 麥照附圖做進—步說明。 地對日:-目▽出本發明第—實施例的邏輯塊與物理 “ “糸不意圖，本實施例應用於快閃記憶體介質 )，在整個儲存媒體巾作對照表，但 對儲存媒體劃分多個區域，而在物 ：： =區域都可以對照整個儲存媒體的物理塊。當二= 立址的時候’才把該邏輯位址所在區域的對照二 到記憶體中。 4\ 本實施例還設置交換塊（柳），通過共用交換 現物理塊的交換，從而、;肖除物理塊磨損不均衡的問題。1 具體來說，設儲存媒體共有⑽物理塊，將邏輯位址 刀為L個區域，分別是區域Q、區域1......㊣域L，— 個區域N個邏輯塊，每個邏輯塊分別與一個物理塊建立： '日錢，例如區域〇的邏輯塊〇對應物理塊i，區 :LX對應物理塊”再設置一邏輯交換塊，也對應二 1成如表1所示^邏輯塊與物理嬙對昭喪: 區域 邏輯位 「區域 0 ~ 1；——：~~~―1— 200921388 表 1 接收到對儲存媒體的訪問命〜 的處理過程參照第五圖包括如下;驟貫施例對寫命令 ，收到對邏輯塊χ寫入新 . 步驟S12，計算邏輯塊χ 貝㈣冩 牛㈣Q 科尾X屬於哪個邏輯區域； =⑽，將該邏輯區域的 步驟SI4，查對照表，栌储、溫/戰入到。己隐體， 到對應的物理心； 輯抑塊X的邏輯位址查找 步驟S15 ’取預先定義的交換塊； 換神rrsi6 ’將寫命令要寫入邏輯塊X的新資料寫到交 換塊對應的物理塊2中； I貝丁寸舄到乂 步驟S17，將物理塊乂中不 搬遷到物理塊z中； 1 p 7覆孤的原有貝枓 步驟S18 ’將交換塊改為盥物 X改為與物理塊z對應； ”對應，將邏輯塊 步驟⑽’擦除交換塊對應的物理塊y，寫過程结 末。 、° 塊對：::：：圖所示的經過上述寫過程後’邏輯塊與物理 要1糸=化不意圖，邏輯塊χ中的原有資料和寫命令 :::新資料都寫入物理塊2中，而將交換塊從物理塊 文為物理塊”同理，對每一個寫命令，都僅载入一個 區域的對照表’使用交換塊寫入資料，既實現了分區域管 理储存媒體’減少佔用記憶體，又實現了物理塊的使用平 衡，還無需對整個區域資料進行搬遷。 200921388 上述步驟S12，計算邏輯塊X屬於哪個邏輯區域，具 體來說在對邏輯位址分區時，每個邏輯區域的大小都是預 先设疋的’確定下來一般無需改變，如一個區域包括1Q24 個邏輯塊。當收到訪問命令後’本實施例可以通過邏輯位 址來推异要訪問的塊屬於哪個邏輯區域。一個區域包括 1024個邏輯塊為例’則可算出邏輯塊〇到邏輯塊都屬 於區域0，而邏輯塊1024至邏輯塊2047都屬於區域丨，依 此類推。 上述步驟S13 ，將該邏輯區域的對照表載入到記憶 體，其t載入是指把對照表讀取到記憶體中，具體載入方 式見下文說明。 ' ㈣T H局八邂輯塊X的新資料寫 到交換塊對應的物理塊2中是採用現有的寫操作技術完成 的，完成一個寫操作可以分為兩個步驟，一個是設備從主 機接收資料，存到設備的緩衝區，然後從緩衝區把 =㈣。本實施例從寫命令中得知要對邏輯塊χ寫入資 Γ而邏輯塊物理塊y，物理塊y本身是存有原有 1料的，根據快閃記情體的拉卜 ’、百 b "苔〜 物理塊y是不能寫入資 枓的，必須把物理塊乂擦除才 除物理塊y，原有資料就合J :…科。但-旦擦 的資料寫入物理塊y，而口处 ’妾把讀區 中。 寫到交換塊對應的物理塊2 完成將缓衝區資料寫 和交換塊對應的物理塊z 入交換塊的寫操作後，物理塊又 中各保存了邏輯塊X的一部分資 200921388 料’在以後的操作令是無法使用的’因此還需要把物理塊 二:的原有資料搬遷到物理塊z中。至此，邏輯塊χ的所 :科都保存在物理塊ζ中了，物理塊y中雖然存有原有 f ’但已經無用了’就可以把物理塊y擦除。當然也 二：理塊“察除以後，把物理塊z令的所有資料重新拷 貝回物理塊y，但這樣明顯效率很低。因此本實施例 把物理塊y與物理塊z與邏輯塊的對應關係交換一下 應交換塊，物理塊2對應邏輯塊χ，節省了一 -人貝#搬遷的操作，提高了效率。 、欠上述步驟S17，將物理塊y中不被寫命令覆蓋 貝料搬遷到物理塊z中是採 ’、有 包括步驟： 現有的資龍遷技術，具體 址；步驟S171，從寫命令中取得要寫入新資料的指定位 :驟S172，將指定位址與物理塊y的位址比較，得到 二新資料的未指定位址，這些未指定位 的貝科疋不被新賢料覆蓋的原有資料； z的將未覆蓋位址中的原有資料複製到物理塊 ζ的相應位址中，流程結束。 兄 直接性是要先擦除才能寫資料，如* 丨使用父換塊，以保證原有資料的安全。 才 本實施例對於讀命令的處理過程如下： 13 200921388 步驟S21 ，收到讀取邏輯塊χ資料的讀命令； 步驟S22，計算邏輯塊χ屬於哪個邏輯區域，· 步驟S23 ，將該邏輯區域的對照表載入到記憶體； ν驟S24查對照表，根據邏輯塊X的邏輯位址查找 到對應的物理塊y ; 一 步驟S25 ，讀取物理塊7中的指定資料； 步驟S26，返回讀到的資料，讀過程結束。 本教明提出第二實施例，在第__實施例基礎上對對照 表和交換塊的管理提出改進。 本實轭例至少在儲存媒體量產的時候建立一完整對昭 斜靡通^完整對照表’可使所有邏輯位址都與好的物理塊 j來，，不對應壞物理塊，將壞的物理塊忽略不使 立邏輯位址與所有好塊的物理位址之間的對應關 塊，媒體細個物理塊，其中物理塊2是壞 始建立對法是：從邏輯塊◦和物理塊〇開 到物理換二 找一個物理塊與之對應，檢測 為邏輯二Γ:=把物理塊0對應到邏輯塊°;再順序 對庫關係W 塊與之對應，因為物理塊G已經有 物理塊1、對測理塊1是好塊馳 對應，因為物揮换彳 〜、t輯鬼2找一個物理塊與之 開始，發現物，ί 的塊都操作過了 ’就從物理塊2 邏輯塊❹ ？2是壞塊，不能將物理塊2與任何-個 U對應，於是再錢朗物理塊3 200921388 好塊，就把物理塊3對到邏輯塊2。依次對應 輯塊都與物理塊建立對瘅 心有边 不讓壞塊與任何-個邏=庫而對其令壞塊的處理就是 對應，齡Li 也不與任何-個交換塊 ‘ μ際使用過程中永遠不會被使用到。當 可:其他規則建立完整對照表，將壞的物理 Α 建立邏輯位址與所有好塊的物理位址之η 的對應關係。 主m址之間 建立對照表後，在資料卢τ田4 w ^ 如_，或儲存婢統的外部儲存媒體(例 -敕…：! 身）劃出—儲存空間來保存上述 凡正' …、，貝料處理系統的Mcu記憶 照表所在的物理位址。當需要訪問某區域的； 存的完整對照表所在物理位址，就可以爾整 =表：該區域對應的部分，將這部分對照表讀取到記憶 :的儲整對照表的載入。上述在外部儲存媒體中劃 的冗=以是取幾個物理塊’也可以是多個物理塊 =施例還可為每個區域分別建立區域對邏 車耳㈣劃分為L個區域，分別是區域〇、區域】......- ΐ域：:Γ4Ν個邏輯塊。本實施例為每個區域分別建立 :甘：、#共有“固區域對照表’每個對照表覆蓋Ν 塊；以，區域對照表結合起來即等同於上述完整 留^ 以將所有好的物理塊管理起來。在記憶體中僅 ” “品域對照表（例如區域〇對照 對照表保存在外部錯存媒體(例如咖Μ，或儲= 15 200921388 身）劃出的儲存空間中。記憶體中記錄每個區域表的保存 位址（例如區域〇對照表保存在邏輯塊〇，區域1對昭表 保存在邏輯塊1..··.·)。設接收到的訪問命令要訪問邏輯 ，500、’而目前記憶體中保存的區㈣對照表僅覆蓋邏輯 =。f邏輯塊127 ’因此當前區域對照表無法滿足訪問需 二就而要載入與訪問指令相應的區域對照表。通過計曾 二:邏輯塊5GG屬於區域3，讀取區域3對照表並存入纪 f ’替代記憶體中原有的區域◦對照表完成對照表载 部然後利用區域3㈣表執行訪問命令。上述在外 乂 Γ媒脰中劃出的儲存空間可以是取幾個物理塊，也可 以疋多個物理塊的冗余區。 上述區域對照表管理流程如第六圖所示，包括： 步驟S31，將邏輯位址劃分為至少兩個區域； 步驟S32，分別為至少兩個區域建立區域對照表； =驟S33 ’將其中一區域對照表保存在記憶體中，豆 他區域對照表保存在外部儲存媒體中； ，、 步驟S34，接收訪問命令； 步驟S35，計算兮方鬥人 牛驟娜問〒令指定邏輯塊屬於哪一區域； 乂騍S36 ，如果記憶體中的區域對照表 塊所屬區域’則進行步驟S38，否則進行挪^曰疋匕耳 步驟S37，將指定邏輯塊所屬區域對照表載 體中’替換記憶體中的原區妁斜 1 °己十思 S38 ; 域對照表，然後進行步驟 步驟S38 5利用# ,卜立触| h 用^體中的區域對照表，執行訪問命 16 200921388 令； ^驟S39 ’返回命令執行結果，流程結束。 廯j 7為喂°卩令，訪問不改變邏輯塊與物理塊的 r:M，若訪問命令為寫命令，執行完該訪問命令後邏 二〜用報理塊的對應關係改變’區域對照表也應作相庫更 問命令切換區域對照表時，可直接讀取區域對照表。“ 把心：：撕’將區域對照表載入到記憶體中，是指 域部儲存賴讀取到記憶财。例如—區 二料是“ _54321”，本實施例將其保存在外 外部儲存媒體的哪存在 區域對照表的時候，就從”丄列需要用到該 昭表資料*屮*、夂數°己錄的物理位址中把區域對 二、:媒出來，讀到記憶體中用於保留對照表的區域即 區域：心:個記憶體區域在不同的時刻被载入不同的 例中照表載入完成後，本實施例採用第-實施 甲的寫過程執行訪問命令。 到訪只在要用到區域對照表的時候，例如收 令或其他需要查表的情況下，才將區域對照 二隐體卜平時的記憶體完全可以不存放任何區域對 把—個區域對照表常駐在記憶lit是為了提高訪問 ，本發明也可以不預先生成對照表，而是每次需 17 200921388 建該區域的對照表或者臨時重建儲 松:點二二 然後载人到記憶體中。這樣做可以 影響訪問效率，因此建議保存對二'重建對照表可能會 外部銘存媒體(例如ΕΕ_，以嬅=時保存在 媒體内部’要使用時讀取並載:==身）或在储存 果十…广 了兀成大容量儲存媒體的管理。如 存重建對照表的資訊。 則而要在儲存媒體中保 本實施例提出在物理塊冗余 建對照表的做法如下： 仔對…、表貝成及重 以外快體介質的結構是在每個物理塊中除了資料區 二遇有右干位元組的冗余區。冗余區可以保存—些廠 ft的資訊’因此本實施例可以在這個區域放置—：Μ旱 如冗余區㈣位元組，其中1〇位元組放置虹校 元組中，二二7"^可以放置對照表資訊，在這6位 /只轭例記錄本物理塊對應的邏輯塊號。這樣， 起候’就把邏輯塊號和資料以及ECC校驗碼一 許取/子媒體中。在本實施例重建對照表的時候，通過 ‘輯::物'=的：余區:!訊’就知道每個物理塊對應的 表λ〜這樣無需重新掃描儲存媒體，即可重建對照 把對照表保存到儲存媒體的部分物理塊中的做法 在最初建立料日”主 “疋 對如、表的時候，本實施例把部分的好塊劃出來 18 200921388 用於保存對照表，不與邏輯媸料 m ★ 耳塊對應’也不與交換塊對應， 處理系統的„的記憶體記住這些塊的物理塊 f ：對照表發生變化的時候’把對照表寫入這些物理塊 :伴鋪表和資料’本實施例仍需要標記哪個塊 St湖，哪個塊保存的是對照表資訊，該標記可 放到物理塊的冗余區中。這樣，通過掃描冗余區，本實 :例就知道哪個塊保存的是資料，哪個塊保存的是對照、 在找到保存對照表的塊後，把對照表讀出即可， 進行重建的過程。 …、 在上述實施例基礎上’本發明還提出設置多個交換 塊，維持-交換塊池以管理這些交換塊，在讀寫資料過程 可以輪換使用交換塊以提高訪問效率。 以上所述僅為本發明的優選實施例，並非因此限制本 '月的專利圍’凡是利用本發明說明書及附圖内容所作 的等效結構或等效流程變換，或直接或間接運用在其他相 關的技術領域，均同理包括在本發明的專利保護範圍内。 【圖式簡單說明】 第圖疋本發明一種現有技術的邏輯塊與物理塊對照 關係示意圖； 第一圖疋本發明另一種現有技術的邏輯塊與物理塊對 恥關係示意圖； ’ ^第三圖是本發明第一實施例的邏輯塊與物理塊對照關 係示意圖； 19 200921388 第四圖 是本發明第一實施例的邏輯塊與物理塊對照關 係變化示意圖； 第五圖 是本發明第二實施例寫命令的處理過程示意 圖； 第六圖 是本發明第二實施例的區域對照表管理流程示 意圖。 【主要元件符號說明】 步驟S11〜S19 步驟S31〜S39 20200921388 IX. Invention Description: [Technical field to which Mingming belongs] In particular, it relates to a kind of storage. This is related to the media management method in the field of static memory. [Prior Art] The existing management method of the storage medium, the address of the address of the Zhao Zhaoben ^ 疋 疋 establish the physical address and the logic shown in the first diagram of the logical block and physical objects... Second: find the physical block corresponding to the logical block When Tian has a physical block that becomes a bad block, you can ^, ^ θ +ju + j village /, shoot out. However, with the storage of the corpuscle: the valley is getting larger and larger, the amount of data in the comparison table is getting larger and larger, and it is not practical to save the entire storage medium in the memory of the system.戍卞疋 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有 现有It is divided into L areas, and each of the N-blocks performs a comparison table of physical addresses and logical addresses for each physical block area. A comparison table corresponding to the area of the logical address is established in the memory only when the corresponding logical address area is used, thus achieving a trade-off between the physical space and the efficiency. But this also caused another problem. Because the storage space is physically divided, the areas are isolated. «The actual use characteristics of storage media, physical blocks in some areas are often used, and physical blocks in other areas are rarely used, which leads to uneven use, some physical blocks have reached the limit of life, and some 200921388 = but still very young. When some physical blocks have been damaged, it is not in guilty to store the data and re-storage the entire storage. 4 仃Μ 仃Μ 'But the bad physics block against the above problems, there is another management technology, when the area is two! Presented based on the above segmentation domain and other areas of the exchange of data will be the entire area of the above problems, but Yan Yan Tricks technology. This technology can solve the problem of inefficiency in the time to copy the female domain. There will be a piece of information on the 'Bai Da 1', resulting in unstable storage media performance. SUMMARY OF THE INVENTION The second aspect of the present invention is to provide a storage medium management method for managing the Otani 1 storage medium. HW, Lvben invention provides a storage media logical address, establish P, "t method" includes: the step of sub-regional pipe relationship; at least:; the corresponding part of the car ear address and physical address is loaded in The step in the memory; the corresponding address of the physical address (4) should be related to the execution of the access command (4) the domain logical address and the object 'the above-mentioned sub-region management logical address, establish at least one region and physical address The corresponding relationship step includes: a logical address area; a logical address in at least a logical area, tf mm ^ ^ ^ a 0 - a correspondence between physical addresses; forming a correspondence into at least two regions The correspondence between the incoming address and the physical address includes: calculating, according to the access command, which zone i or the specified logical block 200921388 belongs to; determining whether the zone comparison table in the memory corresponds to the zone to which the tile belongs Loading the area comparison table of the area to which the specified logic block belongs into the memory, and replacing the area comparison table in the memory. The above-mentioned correspondence between at least one area logical address and physical address is loaded in the memory. The step of the step further includes: the step of storing the -region comparison table in the memory. Preferably, the step of managing the logical address of the sub-area, establishing a correspondence between the at least one regional logical location and the physical address comprises: The logical address is divided into two areas to v. For the logical address in each area, a correspondence relationship with any physical address of the storage medium is established, and a complete comparison table is formed. ', the above logical address of at least one area is The step of loading the correspondence of the physical addresses in the sibling body includes: saving the complete comparison table in the external storage medium ^ storing the address of the complete comparison table in the memory; and calculating which of the specified logical blocks belongs according to the access command Area; find the adjustment table according to the address of the complete comparison table; in the complete comparison table, the partial comparison table of the area I domain of the specified logic block is loaded into the memory. The above comparison table is stored in the external storage medium. Means that the storage medium management method further includes at least one physical block or is stored in a redundant area of at least one physical block. Preferably, the foregoing storage medium management method further includes defining at least The step of establishing a correspondence between the logical address and the physical address of the switching block by the switching block. The step of performing the light access command according to the correspondence between the logical address of the at least one area and the physical address includes: according to at least one regional logical address Correspondence with physical address 'Find and write command specifies the logical address corresponding to the object 8 200921388 The address 'writes the new data to be written by the write command to the physical address corresponding to the exchange block' will and write command The original data that is not covered by the write command in the physical address corresponding to the specified logical address is moved to the physical address corresponding to the exchange block, and the physical address corresponding to the logical address corresponding to the logical address of the write command is established. The correspondence relationship 'establishes a write command to specify a correspondence between a logical address and a physical address corresponding to the exchange block; and erases a physical address corresponding to the exchange block. The correspondence between the logical address and the physical address according to the at least one area The step of executing the access command further includes: updating the correspondence between the at least one regional logical address and the physical address, and updating the comparison table. ^ A. The above-mentioned physical address corresponding to the logical address corresponding to the write command specified logical address is not relocated to the physical address corresponding to the exchange block. ^. The write command refers to the logical address. The corresponding physical address is compared with the exchange block 2 = physical address 'get the ^ bit ^ which is not written to the command ^ write new data; the data in the unspecified address is copied into the physical address of the intersection. The storage medium management method uses a single exchange block separately with at least one exchange block. ^ Shuming divides the logical address of the storage medium into multiple areas, each of which does not organize the lion's implementation of occupying a very small amount of memory to complete the official sense of the sub-media; the present invention also utilizes the physical blocks of the inter-communication Exchange, thus eliminating the problem of each (four) domain of the play, while avoiding the entire area of data = and = unbalanced question, security and stability, easy to be real: performance instability caused by revival 200921388 [Embodiment] ☆ The realization, functional features and advantages of the object of the present invention will be described in conjunction with the embodiments. Earth-to-day: - The logical block and physical of the first embodiment of the present invention are shown "not intended, the embodiment is applied to a flash memory medium", and the storage medium is used as a comparison table, but the storage medium is Multiple areas are divided, while objects::= areas can be compared to the physical blocks of the entire storage medium. When the second = address, the control of the area where the logical address is located is transferred to the memory. 4\ In this embodiment, the exchange block (willow) is also set, and the exchange of the physical blocks is performed by the shared exchange, thereby eliminating the problem that the physical blocks are unevenly worn. 1 Specifically, the storage medium has a total of (10) physical blocks, and the logical address is divided into L areas, which are area Q, area 1 ... positive domain L, - area N logical blocks, each The logical blocks are respectively established with a physical block: 'day money, for example, the logical block of the area 〇 corresponds to the physical block i, the area: the physical block of the LX corresponds to the physical block.” Then a logical switching block is set, which also corresponds to two 1% as shown in Table 1. Logic block and physical 嫱 昭 : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Through the example, the write command receives a new write to the logic block. Step S12, calculate the logical block χ (4) yak (4) Q which tail X belongs to which logical region; = (10), the logical region step SI4, check Table, 栌 、, 温 温 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / The new data of the logical block X is written into the physical block 2 corresponding to the exchange block; In step S17, the physical block is not moved to the physical block z; 1 p 7 is overlaid with the original shell step S18 'change the swap block to the object X to correspond to the physical block z; Block step (10) 'Erasing the physical block y corresponding to the swap block, the end of the write process. , ° block pair::::: After the above-mentioned writing process, the logic block and the physics are required to be 1 糸 = not intended, the original data and write command in the logic block :::: new data is written In physical block 2, the exchange block is treated as a physical block from the physical block. Similarly, for each write command, only one area of the comparison table is loaded. The data is written using the exchange block, which realizes the sub-regional management storage. The media 'reduced the memory usage, and realized the balance of the use of the physical block, and there is no need to relocate the entire area data. 200921388 In the above step S12, it is calculated which logical area the logical block X belongs to, specifically when partitioning the logical address, The size of each logical area is pre-set. 'Determined generally does not need to be changed, such as an area including 1Q24 logic blocks. When an access command is received', this embodiment can use a logical address to deduce the block to be accessed. Which logical area belongs to. One area includes 1024 logical blocks as an example, then it can be calculated that the logical block 〇 to the logical block belong to area 0, and the logical block 1024 to the logical block 2047 belong to the area 丨, according to this In the above step S13, the comparison table of the logical area is loaded into the memory, and the t loading means that the comparison table is read into the memory, and the specific loading manner is described below. ' (4) T H Bureau Eight Diagrams Series The new data of block X is written to the physical block 2 corresponding to the exchange block by using the existing write operation technology. Completing a write operation can be divided into two steps. One is that the device receives data from the host and stores it in the buffer of the device. And then from the buffer = (four). This embodiment knows from the write command to write the logic block Γ and the logical block physical block y, the physical block y itself is stored with the original material, according to the flash The physical style of Rab', hundred b " moss ~ physical block y can not be written to the capital, you must erase the physical block 才 in addition to the physical block y, the original data is combined with J: ... section. The erased data is written into the physical block y, and the mouth is in the read area. The physical block 2 corresponding to the swap block is written. After the buffer data is written and the physical block corresponding to the swap block is written into the swap block, the write operation is performed. , in the physical block, each part of the logic block X is saved. 200921388 In the future operation order is unusable 'Therefore, it is necessary to relocate the original data of the physical block 2: to the physical block z. At this point, the logical block is stored in the physical block, and the physical block y Although there is an original f 'but it is useless', you can erase the physical block y. Of course, the second block: "After checking, all the data of the physical block z is copied back to the physical block y, but this way Significantly inefficient. Therefore, in this embodiment, the correspondence between the physical block y and the physical block z and the logical block is exchanged, and the physical block 2 corresponds to the logical block, which saves the operation of the one-person-# relocation and improves the efficiency. In the above step S17, the physical block y is not overwritten by the write command to relocate the bedding material to the physical block z. The process includes: the existing steps: the existing resource migration technology, the specific address; step S171, obtained from the write command To write the specified bit of the new data: Step S172, compare the specified address with the address of the physical block y, and obtain the unspecified address of the new data, and the unspecified bits of the Becco are not covered by the new material. The original data; z copies the original data in the uncovered address to the corresponding address in the physical block, and the process ends. Brother directness is to erase before you can write data, such as * 丨 use the father to change the block to ensure the security of the original data. The processing procedure for the read command in this embodiment is as follows: 13 200921388 Step S21, receiving a read command for reading the logical block χ data; Step S22, calculating which logical region 逻辑 the logical block χ belongs to, Step S23, the logical region The comparison table is loaded into the memory; ν S24 checks the comparison table, and finds the corresponding physical block y according to the logical address of the logical block X; in step S25, the specified data in the physical block 7 is read; Step S26, returns to read The information obtained, the reading process is over. The present teachings propose a second embodiment which proposes an improvement on the management of the look-up table and the exchange block on the basis of the __th embodiment. The real yoke example at least when the storage medium is mass-produced to establish a complete pair of slanting 靡 ^ ^ complete comparison table 'can make all logical addresses with good physical block j, does not correspond to bad physical blocks, will be bad The physical block ignores the corresponding block between the logical address and the physical address of all good blocks. The media is a fine physical block, where physical block 2 is the bad start. The pair is: logical block and physical block. Open to physical change to find a physical block corresponding to it, the detection is logical two: = physical block 0 corresponds to the logical block °; then sequentially to the library relationship W block corresponds to it, because the physical block G already has physical block 1 For the test block 1, it is a good block, because the object is changed, the t-ghost 2 finds a physical block and starts with it, and the discovery, the block of ί have all been operated 'from the physical block 2 logic block❹ ? 2 is a bad block, and the physical block 2 cannot be associated with any - U, so that the physical block 3 is paired to the logical block 2. In turn, the corresponding blocks are all set up with the physical blocks. The edges are not allowed to be dealt with any bad blocks and any bad blocks are processed. The age Li is not used with any one exchange block. It will never be used in the process. When: Other rules establish a complete comparison table, and the bad physical Α establishes the correspondence between the logical address and the η of the physical addresses of all good blocks. After establishing a comparison table between the main m sites, the data is stored in the external storage medium (eg, 敕...:! body) of the data τ, or the storage space to save the above ... , the physical address of the Mcu memory table of the bedding processing system. When you need to access the physical address of a complete comparison table stored in a certain area, you can complete the = table: the corresponding part of the area, and read this part of the comparison table to the memory: the loading of the storage consolidation table. The above-mentioned redundant storage in the external storage medium is to take several physical blocks' or multiple physical blocks. The embodiment can also establish a region for each region to divide the logical car (4) into L regions, respectively. Area 〇, area 】 - ΐ domain:: Γ 4 logical blocks. In this embodiment, each area is separately established: Gan:, #shared "solid area comparison table", each comparison table covers the block; and the area comparison table is combined to be equivalent to the above complete reservation to all good physical blocks. Managed. In the memory only "" domain comparison table (for example, the area 〇 comparison table is stored in the storage space drawn by externally erroneous media (such as curry, or storage = 15 200921388). Recording in memory The storage address of each area table (for example, the area 〇 comparison table is stored in the logic block 区域, the area 1 pair is stored in the logic block 1....). The received access command is to access the logic, 500, 'And the area saved in the current memory (four) comparison table only covers the logic = f logic block 127 'so the current regional comparison table can not meet the access needs to be loaded with the corresponding regional comparison table of access instructions. : The logic block 5GG belongs to the area 3, and the read area 3 is compared with the table and stored in the original area of the memory. The comparison table is completed and the access table is executed by the area 3 (four) table. The storage space drawn in the 可以 may be a physical block or a redundant area of a plurality of physical blocks. The above-mentioned regional comparison table management process is as shown in the sixth figure, including: Step S31, dividing the logical address For at least two regions; Step S32, respectively, establish a regional comparison table for at least two regions; =Step S33 'Save one of the regional comparison tables in the memory, and the Bean and other region comparison table is saved in the external storage medium; Step S34, receiving an access command; Step S35, calculating which area the designated logical block belongs to by the 斗方斗人, 乂骒S36, if the area in the memory matches the area to which the table belongs, then proceeding to step S38, Otherwise, the step S37 is performed, and the area to which the specified logic block belongs is compared with the original area in the replacement memory in the table carrier. The angle is 1°; the domain comparison table is performed, and then the step S38 is utilized. , 卜立触| h Use the area comparison table in the body, execute the access life 16 200921388 order; ^S39 'return the command execution result, the process ends. 廯j 7 is the feed ° command, the access does not change the logic r:M of the block and the physical block, if the access command is a write command, after the execution of the access command, the logic 2 changes with the correspondence relationship of the processing block. The area comparison table should also be used as a phase library to ask the command to switch the area comparison table. You can directly read the regional comparison table. "Put the heart:: tear" to load the regional comparison table into the memory, which means that the domain department stores the memory. For example, the area is "_54321", this When the embodiment saves the existing area comparison table of the external storage medium, the area is paired from the physical address of the display data required to be recorded. The medium comes out and reads the area in the memory for retaining the comparison table, that is, the area: the heart: the memory area is loaded in different cases at different times. After the loading is completed, the embodiment adopts the first implementation. The write process performs an access command. When the visitor only needs to use the regional comparison table, such as the order or other need to look up the table, the memory of the regional control two hidden body can not be stored in any area. Residing in the memory lit is to improve access, the present invention may also not generate a comparison table in advance, but each time 17 200921388 need to build a comparison table of the area or temporarily rebuild the storage: point 22 and then carry people into the memory. This can affect access efficiency, so it is recommended to save the pair 'reconstruction checklist' may externally store the media (eg ΕΕ_, 嬅=when saved inside the media' to read and load:== body) or save Fruit ten... widened into the management of large-capacity storage media. For example, save the information of the reconstruction table. Then, in the storage medium, the embodiment of the present invention proposes to construct a comparison table in the physical block redundancy as follows: The structure of the fast medium other than ..., the table and the weight is in each physical block except the data area. There is a redundant area of the right stem. Redundant areas can be saved - some factory ft information 'so this embodiment can be placed in this area -: drought such as redundant area (four) bytes, one of which is placed in the rainbow school tuple, two two 7 " ;^ can be placed in the comparison table information, in this 6-bit / yoke example record the logical block number corresponding to the physical block. In this way, the logical block number and the data and the ECC check code are given in the sub-media. When the reference table is reconstructed in this embodiment, the table λ~ corresponding to each physical block is known by the 'complex:: object'=: the remaining area: !, so that the control can be reconstructed without rescanning the storage medium. The method of saving the table to a part of the physical block of the storage medium is the first time to establish the material "main", such as the table, the embodiment of the part of the good block is drawn out 18 200921388 for saving the comparison table, not with the logic Material m ★ The ear block corresponds to 'not corresponding to the exchange block, the memory of the processing system remembers the physical block of these blocks f: When the comparison table changes', the comparison table is written to these physical blocks: the accompanying table and The data 'this embodiment still needs to mark which block St Lake, which block holds the comparison table information, the mark can be placed in the redundant area of the physical block. Thus, by scanning the redundant area, the real: the example knows which The block holds the data, and the block holds the control. After the block that holds the comparison table is found, the comparison table can be read out, and the reconstruction process is performed. ... Based on the above embodiment, the present invention also proposes to set more Switching blocks, maintaining-switching block pools to manage these switching blocks, and exchanging switching blocks in the process of reading and writing data to improve access efficiency. The above description is only a preferred embodiment of the present invention, and thus does not limit the patent for this month. 'Equivalent structure or equivalent process transformation made by using the present specification and the contents of the drawings, or directly or indirectly applied to other related technical fields, are included in the scope of patent protection of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram showing a relationship between a logical block and a physical block of the prior art; FIG. 1 is a schematic diagram showing another relationship between a logical block and a physical block of the present invention; A schematic diagram of a logical block and a physical block in an embodiment; 19 200921388 The fourth figure is a schematic diagram of a change relationship between a logical block and a physical block according to the first embodiment of the present invention; and a fifth figure is a process of writing a write command according to the second embodiment of the present invention; The sixth diagram is a schematic diagram of the management process of the area comparison table according to the second embodiment of the present invention. No. DESCRIPTION step S11~S19 step S31~S39 20