Re: ping fails under 2.1.39

Alan Cox (alan@lxorguk.ukuu.org.uk)
Mon, 26 May 1997 22:32:00 +0100 (BST)


> > bus activity can cause invalid reads/writes during this operation.
> Once again, why is not I/O delay done by usleep( good constant )?

Because I/O operations can be bunched, delayed and otherwise scheduled
around other I/O cycles - including optimistic PCI fetches

> I thought that linux is wonderfull thing because it can make old, slow
> computers do _a lot_ of work again.

Old machines it does work fine on, and Linux does drive the boards to spec
- and the original spec not th relaxed spec that some clones dont cope with