> 
> IMHO the output in 80 columns is rather useless/hard to decipher on i386
> too. But then there is no real solution, besides leaving out the interrupt
> details in this display.

<trimmed out some bytes so the mail doesn't wrap>

>tals (in kB)             PAGING   SWAPPING      Interrupts
>tual    free             in  out   in  out       256 total
>0320   57684     ops                             100 softcloc
>7460  323792     pages                            22 softnet
>                                                   6 softseri
>Csw   Trp   Sys  Int  Sof   Flt        forks         pin 14
>195   438  2184  127  560     7        fkppw      21 pin 10
>                                       fksvm       1 pin 12
>s   0.0% Ni   0.0% In  97.4% Id        pwait         pin 9
> |    |    |    |    |    |    |       relck         pin 4
>                                       rlkok       6 pin 3

If you include the group names the above looks like:

>tals (in kB)             PAGING   SWAPPING      Interrupts
>tual    free             in  out   in  out       256 total
>0320   57684     ops                             100 cpu0 sof
>7460  323792     pages                            22 cpu0 sof
>                                                   6 cpu0 sof
>Csw   Trp   Sys  Int  Sof   Flt        forks         pic0 pin
>195   438  2184  127  560     7        fkppw      21 pic0 pin
>                                       fksvm       1 pic0 pin
>s   0.0% Ni   0.0% In  97.4% Id        pwait         pic0 pin
> |    |    |    |    |    |    |       relck         pic0 pin
>                                       rlkok       6 pic0 pin

Which is completely worse than useless!  Which is why I split the
available space equally between the group and event names.
However that didn't really work either since you need 5 or 6 chars
to make sense of some of the names - hence the "don't display <= 3
chars of group name" test.

In actual fact the 'pic0 pin %d' strings could probably be changed
to something more meangful.  After all 'pins' 8-15 are actually on
pic1.  Also on a PC some of them are definitely allocated to specific
hardware.  Maybe it is poddibly to work out which driver attached to
the IRQ?

	David

-- 
David Laight: david@l8s.co.uk