On Aug 18, 2010, at 03:54, robin wrote: >> >> If that routine was originally an OS/360 as opposed to an OS/390 >> routine, the move to/from an aligned HW was needed since you would >> get a program check when using LH/STH if the source/target was not HW >> aligned. > > well, on that machine, a couple of ICs and a shift SR may have > been quicker, but the MVC + LH was neat enough. > I wonder why the buffer wasn't aligned? > Alignment of the buffer is no guarantee that RDWs within the buffer will be aligned.
-- gil
