[illumos-Developer] "od" i386 binary for benchmark

[Garrett D'Amore]

My results differ from yours, but then I use a shell other than ksh93.
(Using ksh93 as your shell, you wind up seeing benefit that of course I
don't.  I don't think that such a test really demonstrates the merit of
your implementation, since you wind up claiming a benefit from a builtin
that not all uses of od will see.)

Using bash and just execing your program just like mine (for fair apples
to apples comparison), my results for your version are very similar to
my program.  In UTF-8 locale the times are almost indistinguishable.
Your program runs slightly faster than my program (about 6% faster) in
the C locale for this test on the utf-8 data file.

Notably, both programs perform substantially better than the old closed
program, which takes about twice as long.

At this point the performance difference of the code is close enough
that I'm not willing to make decision based solely on that.  (Making a
decision based on performance numbers that unfairly benefit ksh93 users
seems unfair to me, since many users of this program won't use ksh93.
In fact, interactive ksh93 users are probably still in the overall
minority.)

	- Garrett

On Wed, 2010-10-20 at 23:50 +0200, ольга крыжановская wrote:
> Garrett, could you please benchmark the binary attached against your
> od? I tested it against your latest version and measures these times
> on vmware:
> timex ~/bin/ksh -c 'export LC_ALL=C ; for ((i=0 ; i < 40 ; i++ )) ; do
> ./garrettod/od -c </usr/pub/UTF-8 >/dev/null ; done ; true'
> real           8.50
> user           8.04
> sys            0.01
> 
> timex ~/bin/ksh -c 'export LC_ALL=C ; for ((i=0 ; i < 40 ; i++ )) ; do
> gisburn_od -c </usr/pub/UTF-8 >/dev/null ; done ; true'
> real           4.86
> user           4.33
> sys            0.01
> 
> Olga