Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 20 Jul 2014 23:48:17 -0400 (EDT)
From:      Garrett Wollman <wollman@hergotha.csail.mit.edu>
To:        jdc@koitsu.org
Cc:        stable@freebsd.org
Subject:   Re: Consistently "high" CPU load on 10.0-STABLE
Message-ID:  <201407210348.s6L3mHnF020806@hergotha.csail.mit.edu>
In-Reply-To: <20140720235945.GA90603@icarus.home.lan>
References:  <20140720062413.GA56318@icarus.home.lan> <CALCpEUHfC1zebCCFrk224Y=MoFznMMPKLM-jN-pNqWFyByOphg@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
In article <20140720235945.GA90603@icarus.home.lan> you write:

>Yeah, I've given that a try.  I think all that script does, in layman's
>terms, is count the number of occurrences of a kernel function/symbol
>being called.

I've had good results in the past using pmcstat(8) in "top" mode.  Of
course you'll need a kernel with the PMC support compiled in to use
that, and a CPU where hwpmc(4) supports a useful counter -- I'd start
with something that counts unhalted cycles or retired instructions.

-GAWollman




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201407210348.s6L3mHnF020806>