Here is a typical procedure: when a page is first requested, the job’s SMT is searched
to locate its PMT; then the PMT is loaded and searched to determine the page’s location
in memory. If the page isn’t in memory, then a page interrupt is issued, the page is
brought into memory, and the table is updated. (As the example indicates, loading the
PMT can cause a page interrupt, or fault, as well.) This process is just as tedious as it
sounds, but it gets easier. Since this segment’s PMT (or part of it) now resides in memory,
any other requests for pages within this segment can be quickly accommodated
because there is no need to bring the PMT into memory. However, accessing these
tables (SMT and PMT) is time-consuming.