Date: Wed, 25 Feb 2004 11:00:52 -0800 From: "David O'Brien" <obrien@FreeBSD.org> To: Kenneth Culver <culverk@sweetdreamsracing.biz> Cc: freebsd-amd64@FreeBSD.org Subject: Re: Performance comparison, ULE vs 4BSD and AMD64 vs i386 Message-ID: <20040225190052.GJ7567@dragon.nuxi.com> In-Reply-To: <20040225135035.v66800cwkgw08wwc@www.sweetdreamsracing.biz> References: <1077658664.92943.15.camel@.rochester.rr.com> <20040225110754.hcogcccokg84k44k@www.sweetdreamsracing.biz> <20040225183234.GG7567@dragon.nuxi.com> <20040225135035.v66800cwkgw08wwc@www.sweetdreamsracing.biz>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Feb 25, 2004 at 01:50:35PM -0500, Kenneth Culver wrote:
> Ok well maybe not with amd64, but I thought that when you add registers
^^^^^^^
This is the key word here -- "thought". No one has done any real analsys
and thus we can't say jack. While at first thought compiling for a large
number of GPR's could be more time consuming; I think there are other
phases of code generation where the process is slower due to lack of a
large number of GPR's.
> Like on PowerPC or Alpha or whatever there are a LOT more GPR's than
> there are on even Athlon64... I guess only having 2x the GPR's doesn't
> make a whole lot of difference...
PowerPC and Alpha are also RISC and have other scheduling issues for the
code generator to handle.
--
-- David (obrien@FreeBSD.org)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040225190052.GJ7567>