Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 6 Mar 2009 09:25:39 +0000 (GMT)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        "Wilkinson, Alex" <alex.wilkinson@dsto.defence.gov.au>
Cc:        freebsd-current@freebsd.org
Subject:   Re: pflogd eats 100% cpu now, updated -current from Feb-4 to Mar-4
Message-ID:  <alpine.BSF.2.00.0903060924220.40469@fledge.watson.org>
In-Reply-To: <20090306002750.GN94275@stlux503.dsto.defence.gov.au>
References:  <49AFB9DA.7030105@dva.dyndns.org> <20090305230639.GA21057@jnz.sqrt.ca> <20090306002750.GN94275@stlux503.dsto.defence.gov.au>

next in thread | previous in thread | raw e-mail | index | archive | help

On Fri, 6 Mar 2009, Wilkinson, Alex wrote:

>    0n Thu, Mar 05, 2009 at 05:06:39PM -0600, Christian Peron wrote:
>
>    >We recently turned on zerocopy so it can see some exposure. I will
>    >look into this.
>
> Can anyone explain what this actually means ?
>
>  net.bpf.zerocopy_enable: Enable new zero-copy BPF buffer sessions

BPF now supports a shared memory buffering scheme between userspace and 
kernel; when the sysctl is enabled, newly opened /dev/bpf devices will permit 
configuration of the zero-copy scheme.  The sysctl doesn't turn zero-copy on 
and off per se, since existing configured sessions will continue to use it, 
but no new zero-copy sessions will be permitted when the sysctl is disabled.

However, it sounds like there's some interaction between pflogd and the 
zero-copy code that requires some debugging...

Robert N M Watson
Computer Laboratory
University of Cambridge



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.0903060924220.40469>