Re: Linux timekeeping plans

Alan Cox (
Sun, 6 Dec 1998 19:46:57 +0000 (GMT)

> It seems that the steps that take place between the sender reading the
> tsc and the receiver reading it are pretty well defined. How will this
> not work?

They are well defined. Things like DMA and other memory bus activity perturb
it but statistically they will be valid for x86

> This removes the need to spin waiting for signal_sent to be cleared,
> since it's subsumed by the receive_ready spin.


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at