Subject: Re: PC164 memory bus speeed (was: pciide performance on alpha)
To: Nathan J. Williams <nathanw@MIT.EDU>
From: Andrew Gillham <>
List: port-alpha
Date: 11/03/1999 15:52:54
Nathan J. Williams writes:
> Remember, 'make build' is a pretty bad benchmark, because the
> toolchain does wildly different amounts of work on different
> platforms. For example, egcs does a lot more instruction scheduling
> (the Haifa stuff) for the alpha, powerpc, and sparc targets than it
> does for other processors, including the i386. The exact code built is
> different as well.

I know, but it is kind of a wallclock "speed" measurement.  I am pretty
sure my Alpha is faster than my VAX, since it compiles a kernel, or does
a make build faster.. :-)  Arguably it is doing things quite different
in the compiler on these two platforms.  So everyone "knows" the Alpha
is faster than the VAX, but how would I "know" when comparing my Celeron
to the Alpha?

> If you're going to be fair about it, you should tweak the build
> options on the i386 side for the processor as well.

Should '-mcpu=21164a' automatically pass the '-m21164a' to the assembler?
Or should the option really be "-mcpu=21164a -Wa,-m21164a" ??

Andrew Gillham                            | This space left blank                     | inadvertently.
I speak for myself, not for my employer.  | Contact the publisher.