Date: Thu, 8 Sep 2011 06:42:47 +0000 From: Alexey Dokuchaev <danfe@FreeBSD.org> To: Pietro Cerutti <gahr@FreeBSD.org> Cc: cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, ports-committers@FreeBSD.org Subject: Re: cvs commit: ports/audio/beast Makefile Message-ID: <20110908064247.GA77922@FreeBSD.org> In-Reply-To: <20110908062213.GS98648@gahrfit.gahr.ch> References: <201109071436.p87EaDGo097718@repoman.freebsd.org> <20110907155741.GA69663@FreeBSD.org> <20110907160643.GR98648@gahrfit.gahr.ch> <20110908040041.GA61427@FreeBSD.org> <20110908062213.GS98648@gahrfit.gahr.ch>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Sep 08, 2011 at 08:22:13AM +0200, Pietro Cerutti wrote: > In this particular port, whether SSE is available or not decides wheter > a set of SSE-enabled plugins are built, in addition to the standard > plugins. The most harm will be having a handful of additional files > installed, that the target machine won't be able to use. On the other > hand, reverting the change means not being able to predict which files > will be installed, thus having leftovers, thus having a BROKEN port. If SSE detection is performed during run-time, isn't it simpler to just enable SSE plugins unconditionally? Not only this will simplify the Makefile and pkg-plist logic, but I also believe that hw.instruction_sse is a poorly designed sysctl (ideally, we should have hw.cpu.features or even dev.cpu.0.features perhaps) and should be removed (refactored into features one). ./danfe
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110908064247.GA77922>