Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 03 Feb 2003 18:00:22 -0800
From:      Terry Lambert <tlambert2@mindspring.com>
To:        John David Duncan <jdd@greatschools.net>
Cc:        freebsd-smp@FreeBSD.ORG
Subject:   Re: cvs commit: src/sys/i386/i386 mp_machdep.c
Message-ID:  <3E3F1EB6.CCC1B9BB@mindspring.com>
References:  <XFMail.20030130114117.jhb@FreeBSD.org> <5.1.0.14.2.20030203101241.00a22b20@willow.he.net> <Pine.OSX.4.52.0302031642530.2280@dh8.office.greatschools.net>

next in thread | previous in thread | raw e-mail | index | archive | help
John David Duncan wrote:
> Now I'm confused, and I'm wondering if anyone could clarify:
> 
>   4.6-release does not have the idle_hlt sysctl.  Is it using
>   the hlt instruction, or not?

4.6 halts in the idle loop, if vm_page_zero_idle() returns zero,
in the SMP case, or if do_page_zero_idle  (a sysctl that only
exists in the SMP case) is zero, always.

See all uses of _hlt_vector in /sys/i386/i386/swtch.s for more
enlightenement.


>   And if I install the HTT patch, should I also set idle_hlt=1, or
>   does it become irrelevant?

If the patch you are talking about is the one that spin-loops the
CPU core in question, if the other CPU cores in a hyperthreaded
CPU are already HLT'ed, then the answer is "no, it's not irrelevent".

-- Terry

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3E3F1EB6.CCC1B9BB>