Let me simply quote directly from pmapi.h ...
/*
* pmUnits.scaleCount (e.g. count events, syscalls, interrupts, etc.)
* -- these are simply powers of 10, and not enumerated here,
* e.g. 6 for 10^6, or -3 for 10^-3
*/
#define PM_COUNT_ONE 0 /* 1 */
So they are powers of 10, and negative numbers are acceptable ...
although I do admit the notion of a microevent or nanopacket is not all
that common ... 8^)>
On Fri, 2009-10-23 at 11:13 +1100, Max wrote:
>
> On 20/10/2009, at 6:49, Ken McDonell <kenj@xxxxxxxxxxxxxxxx> wrote:
>
> > This commit is not quite correct. I'll have the fix soon, but this is
> > just a heads up before this one escapes too far.
> >
> > QA 027 (at least) is failing as a result.
> The fix makes me think if 027 is testing the right things- when would
> it make sense to have negative scale counts? I thought all counts were
> supposed to be enums like.
>
> max
> >
>
>
|