Date: Wed, 18 Nov 1998 10:42:50 -0800 (PST) From: Matthew Dillon <dillon@apollo.backplane.com> To: Marius Bendiksen <Marius.Bendiksen@scancall.no> Cc: Terry Lambert <tlambert@primenet.com>, rnordier@nordier.com, freebsd-hackers@FreeBSD.ORG Subject: Re: FreeBSD on i386 memory model Message-ID: <199811181842.KAA06180@apollo.backplane.com> References: <199811171806.LAA03809@usr09.primenet.com> <3.0.5.32.19981118121341.00975ac0@mail.scancall.no>
next in thread | previous in thread | raw e-mail | index | archive | help
:> Interrupt gates are definitely faster.
:
:Okay. I seemed to recall it being the other way around. Obtw; are they only
:faster upon entry, or do they return quicker, too?
On the 386 and 486, call gates are faster. On the pentium, pentium-PRO,
and pentium-II, interrupts are faster.
Argument copying is wasteful and has limited use on systems where the
supervisor has access to the user mode memory map. FreeBSD (and virtually
all other operating systems) uses a two-layer design, not a multi-layer
ring design. About the only thing you might see different between OS's
is that some processors have a separate 'interrupt stack'. On Intel cpu's,
however, the abstraction is useless due to the completely broken ring
design because many supervisor instructions only work in ring 0. ring 1
and ring 2 are almost completely useless.
-Matt
:Marius Bendiksen, IT-Trainee, ScanCall AS <marius@scancall.no>
:
Matthew Dillon Engineering, HiWay Technologies, Inc. & BEST Internet
Communications & God knows what else.
<dillon@backplane.com> (Please include original email in any response)
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199811181842.KAA06180>