Subject: Re: test
To: None <port-mac68k@NetBSD.ORG>
From: Arnaud Blanchard <jblancha@pratique.fr>
List: port-mac68k
Date: 06/03/1996 22:55:15
At 5:31 3/06/96, Brian Hechinger is rumored to have typed:

> i just compiled and ran the copyvalue test on two machines (non-mac) and got
> these results:
>
> Sun SPARCstation20/100M RAM, SunOS 5.4:
>
> Start simple copy:
>  long: 0.402 seconds
>   int: 0.402 seconds
> short: 0.402 seconds
>  char: 0.401 seconds
> Start subroutine calls:
>  long: 0.711 seconds
>   int: 0.682 seconds
> short: 0.853 seconds
>  char: 0.868 seconds
>

On my IIci 8/500 + 64 Kb cache-card with GCC 2.72
I am not very slow, compared to this sun :-)

cc -O                           cc -O2
Start simple copy:              Start simple copy:
 long: 1.300 seconds             long: 0.492 seconds
  int: 1.300 seconds              int: 0.486 seconds
short: 1.301 seconds            short: 0.486 seconds
 char: 1.299 seconds             char: 0.486 seconds
Start subroutine calls:         Start subroutine calls:
 long: 4.739 seconds             long: 4.518 seconds
  int: 4.511 seconds              int: 4.336 seconds
short: 6.220 seconds            short: 6.215 seconds
 char: 5.397 seconds             char: 5.277 seconds

But there is a *Big* difference between -O and -O2
Is this normal ?!?!

/jblancha@pratique.fr