Re: [LINUX-KERNEL] HDD Noise stopped with cat /dev/zero

Ralf Baechle (ralf@julia.de)
Fri, 18 Oct 1996 12:14:55 +0200 (MET DST)


Hi,

> Somewhere back in the '50 or '60 Philips also made mainstream computers.
> They got a patent on a method to monitor the CPU performance: they
> connected an amplifier and a speaker to one of the address lines of the
> CPU. In those days the CPU clock wasn't that fast, so it was possible
> to hear the address lines switch. If the sound stopped, you knew that
> your program crashed and you had to start all over.
>
> Any volunteers to design such hardware for PCs? Not that Linux crashes
> that often... If we also modify gcc to generate nice tunes:

Hacking on the 68k mm I found someting in the 68030 manual - the 68030
generates of the half length of a clock pulse when fetching the first
word of an instruction on a CPU pin even when the actual instruction
comes from the cache. Add some trivial electronics and a display et
voila - a original MIPS-O-Meter (TM).

This is getting off topic ...

Ralf