发明名称 Accelerated execution of target of execute instruction
摘要 As disclosed herein a method, executed by a processor, for accelerated instruction execution includes retrieving an execute instruction including a register reference and a reference to a target instruction, retrieving the target instruction, decoding the execute instruction using an instruction pipeline, decoding the target instruction using the instruction pipeline, associating the register reference to the target instruction, and executing the target instruction using the register reference as a source operand modifier. The instruction pipeline is configured such that it allows the target instruction to continue processing without waiting for the register reference to be resolved. The contents of the referenced register may be retrieved in a later stage of the instruction pipeline, and the target instruction may be modified and executed. An apparatus corresponding to the described method is also disclosed herein.
申请公布号 US9389865(B1) 申请公布日期 2016.07.12
申请号 US201514980476 申请日期 2015.12.28
申请人 International Business Machines Corporation 发明人 Alexander Khary J.;Busaba Fadi Y.;Curran Brian W.;Hutton David S.;Malley Edward T.;Prasky Brian R.;Rell, Jr. John G.
分类号 G06F15/00;G06F9/30;G06F9/40 主分类号 G06F15/00
代理机构 代理人 McDaniel Steven F.
主权项 1. A method, executed by a computer, for accelerated execution of an execute instruction, the method comprising: retrieving the execute instruction comprising a register reference and a reference to a target instruction; retrieving the target instruction, wherein a maximum value is assumed for a length operand of the target instruction, and resources are assigned assuming the maximum value, and wherein the length operand of the target instruction is altered by logically ORing the length operand with contents of a register indicated by the register reference, and the contents comprise one or more bits from the register; decoding the execute instruction using one or more decoding stages in an instruction pipeline; decoding the target instruction using the one or more decoding stages in the instruction pipeline, wherein the instruction pipeline is further configured to continue processing the target instruction without waiting for the register reference to be resolved, and a part of the target instruction is modified in a later stage of the instruction pipeline, using the contents of the register indicated by the register reference, and wherein the part of the target instruction corresponds to one of the length operand, an immediate operand, a mask operand, and the register; associating the register reference to the target instruction, wherein the register reference is appended to the target instruction as an additional source operand modifier, and wherein, if a number of sources for the target instruction exceeds a specified limit, the contents of the register indicated by the register reference are obtained using a double issue operation; and executing the target instruction using the register reference as a source operand modifier, wherein the target instruction is executed without modification if the register reference is zero.
地址 Armonk NY US