Re: [rfc] aio-core for 2.5.29 (Re: async-io API registration for

Linus Torvalds (torvalds@transmeta.com)
Fri, 2 Aug 2002 08:56:19 -0700 (PDT)


On 2 Aug 2002, Alan Cox wrote:
>
> 2% is way too much for a lot of applications. Thats 28 minutes a day

Note that _most_ PC clocks are a hell of a lot better than 2% a day, so
that was really meant as the worst case for fairly broken hardware. But it
apparently does happen.

A more realistic schenario is less than 0.1%, but with the caveat that if
the machine goes to sleep, the error goes up to infinity..

(Think of the current "jiffies" update and gettimeofday() _without_ any
ntp or /etc/adjtime. For most people it is good enough to use as a wall
clock. But some people literally lose or gain a minute every hour.
That's the kind of drift I'm talking about).

Linus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/