Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 02 Jun 2020 20:39:18 +0000
From:      bugzilla-noreply@freebsd.org
To:        bugs@FreeBSD.org
Subject:   [Bug 246940] [wishlist/enhancement, patch incl.]: idle user tasks should be charged as "nice" or "idle" CPU time
Message-ID:  <bug-246940-227-2w3QWf1dBo@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-246940-227@https.bugs.freebsd.org/bugzilla/>
References:  <bug-246940-227@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D246940

--- Comment #5 from t.eichstaedt@gmx.net ---
(In reply to Conrad Meyer from comment #2)
> Why is it counter-intuitive that user processes are counted in user CPU t=
ime?
Because the charge classes supplied by cp_times (roughly) reflect the cpu
scheduling classes and these are ordered, thus I would assume the cp_times =
are
implicitely ordered as well: irq, sys, user, nice, idle.

The point is that by allowing idle user tasks to be charged as idle in the =
load
values, they are completely "invisible" (like the kernel idle tasks).

To achieve the effect mentioned above w/o this patch (do not scale cpu freq=
 up
on idle user load), a power manager would have to iterate through all idle =
user
tasks, sum up their cpu times, and compute sythetic load values himself.  I
consider this counter-effective: consume more user-cpu cycles and - even wo=
rse
- context switches, to save energy (and noise pollution).

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-246940-227-2w3QWf1dBo>