Abstract:
PURPOSE: A memory device and a managing method of the memory device are provided to simplify an execution condition of garbage collection of a host or a controller. CONSTITUTION: A nonvolatile memory(110) comprises a plurality of blocks including a plurality of pages. An address conversion memory(120) is nonvolatile, receives an external physical address and converts the external physical address into an internal physical address. The address conversion memory accesses one among the pages using the internal physical address.
Abstract:
스토리지 관리 시스템 및 스토리지 관리 방법이 제공된다. 본 발명의 스토리지 관리 시스템은 호스트, 메모리 버퍼, 스토리지 블록들, 및 상기 호스트, 상기 메모리 버퍼 및 상기 스토리지 블록들 사이에서 인터페이스 역할을 하는 입출력 버스를 포함하고, 상기 스토리지 블록들 각각은 하나의 채널을 경유하여 상기 입출력 버스에 연결되고, 상기 스토리지 블록들은 상기 입출력 버스로 데이터를 병렬적으로 입출력시키고, 상기 스토리지 블록들은 하나 이상의 채널을 그룹핑하여 생성되는 채널 그룹 별로 관리되는 것을 특징으로 하며, 이를 통해 스토리지 블록을 효과적으로 관리할 수 있다. 병렬성, 스토리지, 관리 시스템, 관리 방법
Abstract:
A method of reducing a code size of a program by controlling a control flow of the program using software in a computer system is disclosed. The method includes the steps of storing a first program count of a first instruction in a first buffer when an error occurs while the first instruction having an Operand including Offset and Length is being executed among a plurality of instructions loaded in the code memory, changing a current program count of the code memory to a second program count which is obtained by adding the Offset to the first program count, storing a second instruction, which is located at a position shifted from the second program count by a value of the Length, in a second buffer, replacing the second instruction with a third instruction, which is not recognized by a microprocessor, replacing the third instruction with the second instruction stored in the second buffer when an error occurs while the third instruction is being executed, and changing the current program count of the code memory to a predetermined program count next to the first program count stored in the first buffer.
Abstract:
본 발명은 플래시 메모리 장치의 기억 공간과 쓰기 대역폭을 데이터 압축 기법을 사용하여 효율적으로 확장하는 방법 및 장치에 대한 것이다. 본 발명의 일 실시예에 따른, 플래시 메모리 장치에 데이터를 기록하는 방법은, 호스트로부터 제공된 데이터를 페이지 단위로 압축하여 복수의 압축 페이지를 생성하는 단계와, 상기 복수의 압축 페이지를 소정의 플래시 페이지로부터 페이지 세그먼트 단위로 순서대로 저장하는 단계와, 상기 복수의 압축 페이지에 대한 압축률 정보를 기록하는 단계를 포함한다. 플래시 메모리, 물리 주소, 논리 주소, 블록, 페이지, 압축률
Abstract:
A method and a device for storing/restoring register context for fast context exchange between tasks are provided to reduce time and a memory required for storing and restoring the register context generated in the context exchange, and to improve an execution speed of entire system by increasing the speed of the context exchange. An active register analyzer(520) adds an update code for updating information of an activated register to a task file including a code of the task performing a specific function. A code converter(510) converts the task file including the update code into an execution file. An active register information storing part(530) stores activation information of the register. A CPU(540) executes the task by using the execution file, stores the information of the activated register to the active register information storing part by executing the update code, and stores the active register context according to the information of the stored activated register for the context exchange of the task. The CPU executes again the task by restoring the active register context of the task.
Abstract:
PURPOSE: A method for managing a memory in a multi processor system on chip (SOC) is provided to use restricted memory resource effectively as reflecting locality of the memory, by copying a code of a function or a variable stored previously in a determined storage position. CONSTITUTION: It is judged whether a memory area is able to be allocated to a first memory or not, by judging whether there is a memory area used in the lower part of a fixed function (S504). If the memory area is not able to be allocated to the first memory, performance gain in the case that a fixed amount allocated to a second memory is stored in the first memory is compared with the cost of copying/restoring the amount of a memory area used in the first memory to the second memory (S506). If the performance gain is larger than the cost, the amount as big as the memory area is copied to the second memory, and the amount as big as the memory area is allocated to the first memory (S509). [Reference numerals] (AA) Start; (BB,CC,FF) Yes; (DD,EE) No; (GG) End; (S500) Analyze a compiler at the beginning of a function; (S502) Is there any memory only used at the bottom of the function ?; (S504) Can the memory be assigned to a scratch pad memory ?; (S505) Assign to a scratch pad memory; (S506) Calculate an alpha value and a beta value; (S508) Alpha > Beta ?; (S509) Assign to a DRAM; (S510) Copy, function execution, recovery
Abstract:
A storage unit including a write buffer and a control method for improving write hit rate are provided to improve the writing rate of a storage unit by performing write operation in a write buffer. A storage unit comprises a write buffer(110), a storage media(120) and a locality filter part. The locality filter part determines the locality of data corresponding to the input- output operation request received from the host. According to the locality filter section is the locality of data, the input- output operation request is performed in the write buffer or the storage media. The locality filter section is implemented as the embedded software of the storage unit.
Abstract:
A method and a device for scheduling tasks of an RTOS(Real Time Operating System) are provided to perform a task by preventing priority inversion between the task represented by an expired timer and the task requested by an interrupt without degrading performance of the RTOS even if lots of timers are expired when or before the interrupt is generated. An interrupt responder(120) selects a timer representing a time point corresponding to an interrupt generation time point among the timers representing a task, and the time point and priority assigned to the task. A task executor(150) executes the task represented by the selected timer and the tasks requested by the interrupt in priority order. An attribute checker(140) checks whether the task represented by the selected timer can by executed after a deadline. The task executor executes the tasks not to be executed after the deadline and the tasks requested by the interrupt in the priority order after executing the task to be executed after the deadline based on a checking result. First and second timer list storage units(110,130) stores first and second timer lists, and the first timer list storage unit updates the first timer list when a terminated timer is increased according as current time elapses.