On Friday 09 February 2001 11:39 pm, Jason R Thorpe wrote:
> On Fri, Feb 09, 2001 at 11:12:48PM +0000, Chris Gilbert wrote:
>  > DOH!  Sorry those figures are wrong.  I had a kernel with UVMHIST
>  > enabled (from when playing the kernel vm stuff)  The real figures are:
>  > /usr/bin/time ./a.out
>  >         9.17 real         0.00 user         2.06 sys
>  >
>  > Which is more like the Shark figures.
>
> I'm going to cook up a patch to see if I can improve this -- I'll mail
> it to the list later -- but first, can you point me at the program you
> are using for the benchmark?

In this case nothing special, it's the fork exit loop Richard posted:
#include <sys/types.h>
#include <unistd.h>

main()
{
  int i;

  for (i = 0; i < 1000; i++)
  {
    if (fork() == 0)
      exit(0);
    wait (0);
  }
}

The other one I've used is the make configure part of gmake from pkgsrc (do a 
make clean, a make till it starts to configure, ^C it, and then run 
/usr/bin/time make configure)

Oh updated make configure time is, which is half that it is when UVMHIST is 
enabled (doh):
	106.59 real        20.35 user        62.93 sys
I'm going to try it with the new bootloader in a mo.

Anyway either give push the process creation and destruction.

Cheers,
Chris