home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.sun.hardware
- Path: sparky!uunet!island!coney!hue
- From: hue@island.COM (Pond Scum)
- Subject: Re: Old Sun question - HELP!!!!!
- Message-ID: <hue.726652030@coney>
- Sender: usenet@island.COM (The Usenet mail target)
- Organization: Island Graphics Corp.
- References: <rcaldwel.726296968@ponder> <1993Jan6.115253.19572@elroy.jpl.nasa.gov> <1ier5dINN2uj@armory.centerline.com> <16259@auspex-gw.auspex.com>
- Date: Sun, 10 Jan 1993 07:47:10 GMT
- Lines: 24
-
- guy@Auspex.COM (Guy Harris) writes:
- >>(in fact, SunOS 3.5 didn't swap the kernel
- >Neither does SunOS 4.x....
-
- By "swap", do you mean any sort of paging of the kernel itself? If so,
- this is good to hear - it means that I'm not stupid.
-
- Early last year I went to a developers conference to find out how
- SCSA drivers had changed for Solaris 2.x. During one of the sessions
- the speaker from Sun had said that loadable drivers could be swapped out.
- I was worried about interrupt latency so I asked if there was a way
- to pin the driver into memory. I think the answer was that they would
- not get swapped out if they were in use (god I hope, don't want to
- swap out the disk driver...). So then I said, "Oh, so the kernel is
- pageable now." The speaker from Sun said, "Yes, it always has, at least
- ever since 4.0." This suprised the hell out of me because I had never
- heard that the kernel was paged in 4.x (the only UNIX I knew of at the
- time with a pageable kernel was AIX 3.1), but everone started looking at me
- like I was some kind of idiot. Apparently I was the only one who didn't
- know the kernel was paged. I figured if I was the only one in a room
- of 150 who thought the kernel was pinned in memory, I must be wrong.
-
-
- -Jonathan hue@island.COM
-