Date: Tue, 16 Oct 2007 19:03:25 -0400 From: "Josh Carroll" <josh.carroll@gmail.com> To: "Kris Kennaway" <kris@freebsd.org> Cc: Eric Kjeldergaard <kjelderg@gmail.com>, freebsd-current@freebsd.org, Peter Schuller <peter.schuller@infidyne.com> Subject: Re: Status of MySQL on 7 w/o patches applied Message-ID: <8cb6106e0710161603r1440b00en720f355bb30be82b@mail.gmail.com> In-Reply-To: <471541E9.4010706@FreeBSD.org> References: <20071016110642.GB54457@hyperion.scode.org> <20071016212307.GA8850@hyperion.scode.org> <d9175cad0710161434w333b4c4es9f7a3314f54931ab@mail.gmail.com> <8cb6106e0710161534v6c814fb8k2dbf373daedbd8c3@mail.gmail.com> <471541E9.4010706@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> > 1. remove WITNESS/INVARIANTS/etc from the kernel > > 2. add WITHOUT_LIBKSE=yes to /etc/make.conf prior to building world > > This doesn't have performance implications. Thanks Kris! I know that ffmpeg was 50-200% faster (depending on the input/output) with libthr in RELENG_6_2, so I thought building without LIBKSE might help. But I guess libthr is the default, so the above doesn't much matter? > Yes. Also if you happen to have a /etc/malloc.conf symlinked to AJ as a > leftover from a previous configuration, remove it. Ok, thanks. I don't have a malloc.conf, so I'm all set there. I'll stick to the other thread then regarding ULE vs. 4BSD performance in RELENG_7 in my particular case. It must not be a debugging option I'm missing that's causing 4BSD to perform faster than ULE for ffmpeg it seems. Thanks! Josh
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?8cb6106e0710161603r1440b00en720f355bb30be82b>