Subject: Re: Compile speed wars [was Re: FreeBSD Bus DMA]
To: Jonathan Stone <jonathan@DSG.Stanford.EDU>
From: John S. Dyson <>
List: tech-kern
Date: 06/12/1998 02:44:26
Jonathan Stone said:
> John Dyson writes:
> >FreeBSD-current, vs. a recent NetBSD.  Anecdotal
> >evidence using Linux shows similar results.
> What is "recent"?  I think FreeBSD 2.2.5R is recent (its what the tech
> support people have on CDrom at work) but if I posted a comparsion
> between FreeBSD 2.2.5R, and NetBSD-current, I would expect to be
> pilloried for an unfair comparison.    And rightly so.
I suspect NetBSD-current and FreeBSD 2.2.6 are roughly similar
in light loaded conditions, however the cache will have
very positive effects in favor of FreeBSD.  VM performance isn't
as much an issue as disk caching here (obviously.)

> Even so, I grant it may well not make much of a difference to the outcome.
> But without dates and versions, this is not adequate disclosure.
> BTW, when you said that, anecdotally, the results with Linux were
> similar, it was impossible to tell whether you meant FreeBSD was
> 2.5-3x faster, than Linux, or that Linux was also 2.5x-3x faster than
> NetBSD at a specific compilation task.
> Since Linux doesnt thave netbsd emulation, maybe you mean the former,
> using Linux binaries on both FreeBSD and Linux, is that it?
Linux is about the same speed as FreeBSD in this case.  Also, I
didn't go into details on that, because of the native Linux
binaries.  It seems to fit that Linux and FreeBSD perform
similarly, since this is really a lightly loaded situation.
In such cases, it appears that FreeBSD and Linux are almost

John                  | Never try to teach a pig to sing,     | it just makes you look stupid,         | and it irritates the pig.