Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 19 Dec 2014 10:28:02 -0500
From:      John Baldwin <jhb@freebsd.org>
To:        Poul-Henning Kamp <phk@phk.freebsd.dk>
Cc:        "freebsd-arch@freebsd.org" <arch@freebsd.org>, Adrian Chadd <adrian@freebsd.org>, Jilles Tjoelker <jilles@stack.nl>
Subject:   Re: Change default VFS timestamp precision?
Message-ID:  <7567696.mqJ3jgzJgL@ralph.baldwin.cx>
In-Reply-To: <77371.1418933642@critter.freebsd.dk>
References:  <201412161348.41219.jhb@freebsd.org> <77322.1418933100@critter.freebsd.dk> <77371.1418933642@critter.freebsd.dk>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thursday, December 18, 2014 08:14:02 PM Poul-Henning Kamp wrote:
> --------
> 
> In message <77322.1418933100@critter.freebsd.dk>, "Poul-Henning Kamp" 
writes:
> >>I don't understand
> >>why you think TSP_USEC is slower than TSP_NSEC.  microtime() and
> >>nanotime()
> >>both just call bintime() and then convert the result using similar math.
> >
> >Because of the pointless nano->micro conversion which makes TSP_USEC
> >take a division longer to deliver a less precise result than TSP_NSEC.
> 
> Actually, that's the other way around: it converts microseconds to
> nanoseconds with a pointless multiplication.

Yes, and multiplication is cheaper than division.  It's not a power of
two (so more than a single bitshift), but possibly in the noise compared
to the work in bintime() itself.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?7567696.mqJ3jgzJgL>