shmuel+ibm-m...@patriot.net (Shmuel Metz , Seymour J.) writes: > You don't consider a PSECT to be part of the image?
re: http://www.garlic.com/~lynn/2014d.html#25 Mainframe memories http://www.garlic.com/~lynn/2014d.html#27 Mainframe memories the paradigm allowed the same executable image part could be memory mapped as r/o shared segment concurrently in different virtual address spaces at different addresses and work correctly. os/360 convention had relocatable address constants appearing somewhat randomly throughout an executable image which needed to be swizzled by the loader for the address location that the image was loaded. this tended to prevent the same exact executable copy from working correctly concurrently at different virtual addresses (minus possibly some quantum effects where a relocatable address constant would assume the appropriate correct value whenever it was used). I would tend to use the distinction that for the psect, a private copy was loaded and adjusted for the specific virtual address space location ... separately from (r/o) memory mapping the executable image with no requirement for pre-loading and/or changing ... allowing the same exact (r/o) executable image to concurrently occur simultaneously in different virtual address spaces at different virtual addresses (with just a per virtual address space private copy of the psect having been preloaded and swizzled). this also has the side-effect of precluding things like self-modifying instruction streams. when hardware instruction pre-fetch & decode was first starting out ... one of the complicating, high overhead hardware items was checking if the previous instruction had modified the current instruction. -- 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
