I can envision a routine storing footprints for debugging. Cruel, but all very well for a spree.
Yes, except for fixed pages the strategy of discarding and marking invalid the damaged unmodified page frame is simple and elegant, and the page-in will likely be immediate. -- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 ________________________________________ From: IBM Mainframe Discussion List [[email protected]] on behalf of Paul Gilmartin [[email protected]] Sent: Monday, August 16, 2021 6:24 PM To: [email protected] Subject: Re: RENT binder option On Mon, 16 Aug 2021 21:38:23 +0000, Seymour J Metz wrote: >IBM originally used REFR in support of the mahine check handler; if you got a >parity error in certain refreshable modules then MCH would read in a fresh >copy. REFR did not reuire that the module be read-only, just that refreshing >it would not cause errors. > That's true of successful operation does not depend on that modified content. I find it hard to envision such a design. > ... On modern processors, when there is a good copy of the page on DASD, > discarding a page frame with an unrecoverable memory failure is a better > solution for the same reliability concern. > IOW, the MCH meed only mark the frame as invalid and let paging handle it? (Almost immediately when the failing instruction is retried.) -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
