Date: Thu, 2 Nov 2000 10:13:49 -0500 (EST) From: Andrew Gallatin <gallatin@cs.duke.edu> To: Terry Lambert <tlambert@primenet.com> Cc: freebsd-alpha@FreeBSD.ORG, obrien@FreeBSD.ORG Subject: Re: linux emulation Message-ID: <14849.33572.825297.699854@grasshopper.cs.duke.edu> In-Reply-To: <200011020724.AAA00778@usr07.primenet.com> References: <14848.44982.354268.277746@grasshopper.cs.duke.edu> <200011020724.AAA00778@usr07.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Terry Lambert writes: > This broaches a subject I've been pondering for a bit now, > which is probably going to provoke some flamage and gnashing > of teeth. > > It seems to me that FreeBSD itself is 4x slower than it needs > to be on Alpha. Its not quite that bad. The Compaq compilers distinguish themselves on floating point; there's not a lot of floating point in the kernel ;) > What are the chances of compiling FreeBSD itself with the > Compaq compilers? This would seem to be the natural next > step, and it also seems to me that this could accelerate > the Alpha developement and increase general Alpha stability > immensely. > Even though I know it won't give a 4x speedup, I'd really like to compile at least the kernel with the Compaq compilers. The major hurdle (for me at least) is that there is a large difference between Compaq's syntax for in-line assembly and gcc's syntax. As for math intensive stuff, as I pointed out a few months ago, linking Compaq's math library with FreeBSD apps is dead simple. I run my X server that way at home. That seems to have increased its stability somewhat. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14849.33572.825297.699854>