Back to Operating Systems Home Page
Next: A4 - Socketsreading
Up: 1998 term messages
Previous: Notes on VM
On Mon, 13 Apr 1998, Mark Aiken wrote:
> BTW, I've gotten a number of questions about "what do I hand in". One
> person asked (rightly) whether they should hand in a sched.c with the weight
> function changed to 1/5, another sched.c with the weight changed to 1/2,
> etc, etc, and yet another one that chooses the median goodness function.
>
> Maybe you could send mail with a definitive answer? ;-)
As a general rule: enough that your experiment can be reproduced. This
involves both a description in the report detailed enough and code
which can be used (no one-command-liner perl scripts ;-)
In particular, if the modifications to the kernel code are not absolutely
trivial, you should hand in the modified source files.
For the example you mention, a sched.c with something like
#ifdef ONE_FIFTH
#define wgt_factor (1/5)
#endif
#ifdef ONE_HALF
#define wgt_factor (1/2)
#endif
etc etc
Ciao
Franco
\
Franco Callari