会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 36. 发明专利
    • Memory access control system
    • 存储器访问控制系统
    • JPS59132028A
    • 1984-07-30
    • JP645683
    • 1983-01-18
    • Fujitsu Ltd
    • MATSUZAKI SHIGEHARUTSUNODA HARUHIKO
    • G06F13/12G06F3/00G06F12/00
    • G06F3/00
    • PURPOSE:To give a memory access right to the highest request level when memory access requests are generated simultaneously in plural request levels and to give the memory access right with a table logic when plural access requests are generated in the same level. CONSTITUTION:Request levels from channels 11, 12, 13, and 14 are decoded by decoding circuits 15, 16, 17, and 18. The hightest request level (0) of each of channels 11-14 is inputted to a multiplexer 23 and an OR circuit 19. The second highest request level (1) of each of channels 11-14 is inputted to a multiplexer 24 and an OR circuit 20, and the third hightest request level (2) is inputted to a multiplexer 25 and an OR circuit 21. The output of the OR circuit 19 is selected preferentially from outputs of OR circuits 19, 20, and 21 by a priority selecting circuit 29 and is sent to a multiplexer 30. The multiplexer 30 transmits the channel number, which acquires the memory access right, from a terminal C.
    • 目的:当在多个请求级别中同时生成存储器访问请求时,给予最高请求级别的存储器访问,并且在同一级别生成多个访问请求时,使用表逻辑给存储器访问权限。 构成:通道11,12,13和14的请求电平由解码电路15,16,17和18解码。每个通道11-14的最高请求电平(0)被输入到多路复用器23和 或电路19.每个通道11-14的第二最高请求电平(1)被输入到多路复用器24和“或”电路20,并且第三最高请求电平(2)被输入到多路复用器25和“或”电路 OR电路19的输出优先地由优先选择电路29的OR电路19,20和21的输出选择并被发送到多路复用器30.多路复用器30发送获取存储器访问的通道号 对,从终端C.
    • 37. 发明专利
    • Controlling method of interruption
    • 中断控制方法
    • JPS59105151A
    • 1984-06-18
    • JP21417182
    • 1982-12-07
    • Fujitsu Ltd
    • OOKAWA MASAYUKITSUNODA HARUHIKO
    • G06F13/24G06F9/48
    • G06F9/4812
    • PURPOSE:To reduce hardware and to form a very large scale integrated circuit (VLSI) by collating an interruption priority test code by a microprogram with an output of an interruption factor holding circuit to detect the interruption factor and determine the priority. CONSTITUTION:A microprogram part 5 sends a test code to a register 6 from a factor with higher priority in accordance with the priority of each interruption factor previously fixed in each architecture. The test code inputted from the register 6 to the VLSI 8 is decoded by a decoder 9 and the decoded signal is collated with an interruption factor from a register 11. A coincident interruption factor is added to the address from the register 6 by an adder 7 and the interruption code in the microprogram part 5 is read out and sent to a memory 12.
    • 目的:通过将中断优先级测试代码与中断因子保持电路的输出结合起来,以便检测中断因素并确定优先级,来减少硬件并形成一个非常大规模的集成电路(VLSI)。 构成:微程序部分5根据先前在每个架构中确定的每个中断因子的优先级,从具有较高优先级的因子向寄存器6发送测试代码。 从寄存器6输入到VLSI 8的测试码由解码器9解码,解码信号与来自寄存器11的中断因数进行比较。通过加法器7将一致的中断因子从寄存器6加到地址 并且读出微程序部分5中的中断代码并将其发送到存储器12。
    • 38. 发明专利
    • INSTRUCTION RETRY PROCESSING SYSTEM
    • JPS5858658A
    • 1983-04-07
    • JP15643981
    • 1981-09-30
    • FUJITSU LTD
    • MATSUZAKI SHIGEHARUTSUNODA HARUHIKO
    • G06F11/14
    • PURPOSE:To retry an instruction even in case of an instruction of plural processing flows, by converting the instruction to a dummy and executing it before an suppressed processing flow, and normally executing the instruction after the suppressed processing flow, when retrying the instruction. CONSTITUTION:An instruction retry controlling circuit 3 instructs to retry an instruction, to an instruction controlling circuit 1 through a signal line l34 by use of a result of detection of an instruction processing suppressing flow sent from a signal line l2. When retrying an instruction of plural processing flows, the instruction retry controlling circuit 3 executes the instruction in a state that write to a register and store in a memory by the suppressed instruction, until the processing flow immediately before the suppressed processing, and after the flow by which the suppressed processing, the instruction controlling circuit 1 is set to a state that the instruction can be executed normally, and the remaining part of the instruction is executed, by which the instruction of plural processing flows is retried. When retrying the instruction, a data of a save register circuit 4 is read out as an operand address.
    • 39. 发明专利
    • BUFFER CONTROLLING METHOD
    • JPS589275A
    • 1983-01-19
    • JP10778181
    • 1981-07-10
    • FUJITSU LTD
    • TSUNODA HARUHIKONOJIMA KENICHIMATSUZAKI SHIGEHARU
    • G06F12/08G06F12/02
    • PURPOSE:To eliminate waiting of main storage operation by making operation so that when all blocks of n-column corresponding to mi-row are invalidated, all blocks of n-column corresponding to mi+1 row are invalidated. CONSTITUTION:The contents of a stop address register are made number of sets + L (length of a line), and inputted to an input terminal of a comparator 8. On the other hand, the contents of an effective address register 1 are made all ''0''. The prefetch baud 10 inputs address to an adder 11 which is at input of the address register 1. When there is carry from the line address, page address is increased by 1. By this, set position that becomes an object of operation is sent out to a set register 12 and an object set is designated to address array 3 and a buffer 6 by a decoder circuit 4. When this operation is repeated for the number of stes + L, it is detected by a comparator 8 and buffer invalidation operation comes to an end. Thus, waiting of main storage operation can be eliminated.
    • 40. 发明专利
    • Memory access control system for multiprocessor
    • 用于多处理器的存储器访问控制系统
    • JPS5733471A
    • 1982-02-23
    • JP10550380
    • 1980-07-31
    • Fujitsu Ltd
    • TSUNODA HARUHIKO
    • G06F12/08G06F9/46G06F9/52G06F15/16G06F15/177
    • G06F9/52
    • PURPOSE:To reduce the mean access queuing time for the entire system, by locking only the block region to be originally locked and permitting the memory access request given from another CPU to another storage region. CONSTITUTION:The access address information for another CPU is set to another address register 8; the access address information for own CPU is set to own address register 9; and the lock information for the register 9 is set to a lock register 13 respectively. At the same time, the access request blocks for own CPU and another CPU are compared 11-0 and 11-1. Then the memory access requests given from the own CPU and another CPU have a conflict. And a memory access is inhibited for another CPU as long as the lock-indicated own access request block is idential to another access request block. Otherwise the memory access is permitted.
    • 目的:为了减少整个系统的平均访问排队时间,只通过锁定最初被锁定的块区域,并允许从另一个CPU给另一个存储区域的存储器访问请求。 构成:另一个CPU的访问地址信息设置为另一个地址寄存器8; 自己的CPU的访问地址信息设置为自己的地址寄存器9; 并且用于寄存器9的锁定信息分别设置到锁定寄存器13。 同时,将自己的CPU和另一个CPU的访问请求块比较为11-0和11-1。 那么从自己的CPU和另一个CPU给出的内存访问请求有冲突。 只要锁指示的自己的访问请求块与另一个访问请求块相同,则对另一个CPU禁止存储器访问。 否则允许内存访问。