0000000433f07816-dmarc-requ...@listserv.ua.edu (Paul Gilmartin) writes: > I'm trying to envision how to use WAIT/POST to do that since no > more than one task can be WAITing on a given ECB. I vaguely recall > long ago writing code that used a variant of a CS example in PoOps > to manage a queue of ECBs.
charlie had invented compare-and-swap when he was doing fine-grain multiprocessing locking for cp67 at the science center ... some past posts http://www.garlic.com/~lynn/subtopic.html#545tech and http://www.garlic.com/~lynn/subtopic.html#smp initial attempts to get it included as part of 370 were repulsed, the 370 architecture owners said that the POK favorite son operating system people (os/360 and descendants) were saying that TS (test&set) was more than sufficient for multiprocessor support (they had been coming from a single global system spin-lock that was obtained by first-level interupt handlers). thus was born the science center examples how multi-threaded applications (not necessarily running on multiprocessor) could use compare-and-swap ... that still appear in princples of operation. early application uptake of compare-and-swap were the large multi-threaded DBMS and transaction processing. In the 80s, you start seeing other hardware platforms implementing instructions with comapre-and-swap semantics (especially those selling into transaction and DBMS markets). -- virtualization experience starting Jan1968, online at home since Mar1970 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
