vmstat -i and systat -vm interrupt reporting on CAMBRIA and AVILA

[John-Mark Gurney]

Ian Lepore wrote this message on Mon, Jul 07, 2014 at 14:04 -0600:
> On Mon, 2014-07-07 at 21:44 +0200, John Hay wrote:
> > Hi Guys,
> > 
> > While looking for my atheros problems, I noticed that there seems to
> > be some reporting problem with interrupts on CAMBRIA and AVILA or
> > maybe my expectations are wrong. I see:
> > 
> > :~ # vmstat -i
> > interrupt                          total       rate
> > Total                              67723        130
> > :~ #
> > 
> > And for systat -vm on the righthand side I see:
> > 
> > Interrupts
> > 126 total
> > 100
> >   6
> >  20
> > 
> > For vmstat -i I would have expected a breakdown of the interrupts with
> > their names, totoal and rate. Not just a Total. Systat seems to find
> > a little more with the 3 numbers, but also cannot find the names?
> > 
> > Regards
> > 
> > John
> 
> I wonder if this could be related to the changes HPS made recently to
> sysctl?  vmstat -i gets its info through sysctl, and it's a particularly
> perverse conspiracy of agreement between kernel and userland on how to
> interpret a binary blob of sysctl data that contains a variable length
> table of variable length strings, as I vaguely recall it.

I'm pretty sure this isn't because of HPS's changes.. I have a kernel
running from Jun 18th that shows similar behavior:
# vmstat -i
interrupt                          total       rate
Total                          150211606        105
# uname -a
FreeBSD avila.funkthat.com 11.0-CURRENT FreeBSD 11.0-CURRENT #28 r267333:267349M: Wed Jun 18 16:49:41 PDT 2014     jmg at carbon.funkthat.com:/usr/obj/arm.armeb/usr/src.avila/sys/AVILA  arm

systat -vm:
Interrupts
103 total

3
100

I'll try to take a look at it, but it'll be a bit before I can really
dig in as I'm still traveling..

-- 
  John-Mark Gurney				Voice: +1 415 225 5579

     "All that I will do, has been done, All that I have, has not."