>PR/SM is responsible for saving the state of that logical processor and >restoring it when a physical processor is reassigned to that logical >processor. This is very similar conceptually to what happens with virtual >processors when running under VM.
It's better than 'similar conceptually'. It is almost identical, if not actually. They both use SIE, and at first VM and LPAR didn't play well under PR/SM. LPAR would 'hi-jack' SIE, and VM performance in an LPAR was abysmal. I was at a conference, in Toronto, where IBM accidentally (or not) pre-announced PR/SM 3-4 years after Amdahl released MDF. (I worked at a beta site for MDF). I have also written an article on PR/SM for the online edition of IBM Systems Magazine: 2006-07: "LPAR & PR/SM" http://tinyurl.com/2ay9ot It's not very detailed; it spends its time dealing with configuration rather than implementation. - Too busy driving to stop for gas! ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
