[6607] in RedHat Linux List
Re: SMP systems, XING streaming video
daemon@ATHENA.MIT.EDU (Stephen Zedalis)
Thu Nov 28 21:22:35 1996
Date: Thu, 28 Nov 1996 21:24:13 -0500 (EST)
From: Stephen Zedalis <tintype@zedalis.com>
To: "Geoffrey D. Bennett" <g@netcraft.com.au>
cc: redhat-list@redhat.com
In-Reply-To: <199611280558.QAA07890@Mich.netcraft.com.au>
Resent-From: redhat-list@redhat.com
Reply-To: redhat-list@redhat.com
On Thu, 28 Nov 1996, Geoffrey D. Bennett wrote:
> No, I don't think it's true. Only one CPU at a time (any one) can be
> in kernel mode; this is due to the coarse-grained locking currently in
> use and making this finer-grained is one of the current SMP projects
> as far as I understand. I can't remember what the situation is with
> interrupts, but the kernel _is_ SMP.
Alan Cox could answer this better, but if you have SMP profiling turned
on, you can see by viewing /proc/smp that the second processor NEVER
processes a hardware i/o device (that is controlled by irq vice polling
anyway) . And to make the code more robust, it seems the code was modified
in recent kernel patches, so that this first processor is the only one
that goes into kernel mode, leaving the others to do user mode code only.
After all, the first processor up runs init, and between its process and
memory control, and hardware devices, how much needs to run in kernel mode
anyway? Its the coarse grained locking that causes the current
implementation to be unacceptable in terms of interrupts etc. You guys
can fight out what SMP really means, but in the real world if the second
processor can't handle ANY i/o then it doesn't do any good in internet
servers etc. Might as well just get a faster Pentium or RISC processor.
--
PLEASE read the Red Hat FAQ, Tips, Errata and the MAILING LIST ARCHIVES!
________________________________________________________________________
http://www.redhat.com/RedHat-FAQ http://www.redhat.com/RedHat-Errata
http://www.redhat.com/RedHat-Tips http://www.redhat.com/mailing-lists
------------------------------------------------------------------------
To unsubscribe: mail -s unsubscribe redhat-list-request@redhat.com < /dev/null
