Thanks Ed. I do remember that first discussion now, I think I got busy at work and missed following the other longer one.
Hmm-m-m. Maybe an RCF for PoOp complaining about the lack of examples for many of the more recent complex instructions, including vectors and the several newer "PERFORM . . . " instructions, at least the ones that aren't privileged. I suppose one could use the C compiler with ARCH(Max value for your version of the compiler) and OPT(0) and METAL to generate relatively human-understandable examples, albeit not the very most efficient ones. Peter -----Original Message----- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Ed Jaffe Sent: Monday, January 14, 2019 6:42 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: Unreadable code (Was: Concurrent Server Task Dispatch issue multitasking issue) On 1/14/2019 11:18 AM, Farley, Peter x23353 wrote: > I also participate in ASSEMBLER-LIST and I don't remember such a discussion. > Can you pinpoint the approximate time period for that discussion? > > My biggest complaint/concern with the vector instructions is the woeful lack > of any example code showing how to use them for just the kind of functions > you described. It's hard to imagine how to use such complex instructions > without at least some straight-forward examples. I first mentioned using vector instructions in place of SRST in October 2016 in a thread entitled, "SRST Performance" with eight posts. The group was debating which was faster, TRT or SRST. My comment was: "SRST is much, much faster than TRT but still orders of magnitude slower than the vector instructions." to which you replied as you have here with a desire for good examples from the PoOp owners. No one else seemed interested... Later, in October of last year in another thread entitled, "Count Words?" with nearly 50 posts, I once again recommended use of vector instructions to solve the issue under discussion and that time was met with serious push-back from some list participants (not you). In the end, I smiled to myself content in the knowledge that I could solve the problem with a vector-based routine that would run circles around anything any of those "experts" could write using the alternative techniques being suggested. Of course, such a routine would normally be written using dual-path code with fall back to Plan B if the vector instructions were not available. So discussing the best implementation for Plan B wasn't time wasted... -- This message and any attachments are intended only for the use of the addressee and may contain information that is privileged and confidential. If the reader of the message is not the intended recipient or an authorized representative of the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by e-mail and delete the message and any attachments from your system. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
