Le 19/02/2022 à 08:45, Tom Brennan a écrit :
It'll be interesting to learn the details. For example, I can easily imagine how Hercules (and I assume zPDT) works on x86, running code to simulate each mainframe instruction, register, interrupt, etc. But run that same code under z/Linux and you're still interpreting instructions and hardware that's right there already.
the Hercules docs say that you need about 100 x86_64 instructions to emulate each Z-arch instruction, which makes sense, considering the complexity of the instruction set.
on the actual hardware, things are similar, using micro-instructions in the same manner, emulating the ISA with the hardware resources available to the actual state machine.
(this would make for the performance difference between the various system 360 incarnations for example)
---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
