On Tue, Oct 26, 2004 at 03:55:12PM -0400, Dan Williams wrote:
> I will leave that do David/jgarzik since I didn't actually write any of
> this code, I just broke the megadiff down.
What's happened is that the old explicit schedule_timeout() constructs
were replaced in CVS with msleep() (ssleep() didn't exist at the
time). In the meantime, at least some of them were replaced with
ssleep() in mainline.
I'm about to commit a patch to CVS replacing the msleep()s with
ssleep()s. In the for_linus branch, at least, HEAD will take longer
because we'll need to come up with something to maintain compatibility
with pre-ssleep() kernels.
> On Tue, 2004-10-26 at 20:42 +0100, Christoph Hellwig wrote:
> > On Tue, Oct 26, 2004 at 03:35:36PM -0400, Dan Williams wrote:
> > > New description:
> > >
> > > o Use msleep() instead of hardcoded schedule_timeout()s
> > > o Normalize sleep calls to use msleep() everywhere
> > care to explain what's the point of the latter?
David Gibson | For every complex problem there is a
david AT gibson.dropbear.id.au | solution which is simple, neat and