Procedure and circuit disposition for carrying out separating discurrent input-output operations in an operating data processing system with virtual addressing. (Machine-translation by Google Translate, not legally binding)

摘要

A procedure for carrying out input/output operations sequentially in an operating data processing system with virtual addressing, with the help of channel programs originally routed at will in a virtual or real way, but always with real address, with a sufficient data flow for a plurality of peripheral equipment connected, even when in a data transfer it is necessary to take hold during the development of a channel instruction of more than one page of segmented physical memory, in which one is decoded physical data address set by a physically addressing channel instruction and stored in a data address memory as the address of a data byte, in order to select a memory space in a part of the data memory that forms the memory of the data work, and the corresponding data bytes are transferred between it and a peri computer. Connected, characterized in that the normal data transfer carried out on the basis of the physical data address stored in the data address memory is suspended at the moment on the basis of a byte number that is part of the physical data address and representing the address of a data byte within a page of the memory, it is verified that the data transfer can not be continued already within the previously selected memory page; because then, with the help of a part of the physical data address that designates the physical number of up to then of a memory page, a chaining list memory is used, which, for each physical page number of the memory of work, possesses in consecutive succession a register row in which the physical number of the following virtual memory page is stored in the development of the program; because this new physical page number is joined with the new calculated byte number to form a new definitive physical data address, and it is stored in the data address memory, and because then the data transfer is carried out for a new page of the memory until the limit of the page limit is exceeded in the working memory, exceeding which must be treated in the same way, or respectively without exceeding the limit of the page, until the end of the program is reached. (Machine-translation by Google Translate, not legally binding)