On the mainframe, you have change and reference bits for each page. This permits sophisticated real storage management algorithms not possible on the typical minicomputer architecture which do not provide a reference bit.
-----Original Message----- From: Linux on 390 Port [mailto:[EMAIL PROTECTED]]On Behalf Of Alan Cox Sent: Monday, November 25, 2002 5:37 PM To: [EMAIL PROTECTED] Subject: Re: Missing telnet output (and a bonus newbie questio n) On Mon, 2002-11-25 at 19:46, Ferguson, Neale wrote: > On another, unrelated but equally newbie topic - while TRACEing around the > Linux guest, I saw what seemed an alarming number of PROG 4's (not the > expected page and segment translation exceptions, which I also see). In > systems with which I'm more familiar, this number of protection exceptions > (48 during a "who") would have meant a serious coding problem. But I see > no indication of any error. Should I assume this is just part of normal > Linux operation? I'm not up on the S/390 architecture code but the standard Linux behaviour to generate page aging data in the absence of dirty flags is to force pages read only and take the protection fault internally.
