On Fri, 2006-08-04 at 23:49 +0200, Michael Hanselmann wrote: > On Fri, Aug 04, 2006 at 05:39:32PM -0300, Aristeu Sergio Rozanski Filho wrote: > > > This guarantees a minimum wait time of 5msec after writing the value. > > > Looking at the specs[1], this should be done. Your code doesn't do that: > > > I doubt that if it's possible to retrying after a fail the hardware will > > bother if we try to do the first read in less than 5ms... but here it > > goes. two versions, feel free to choose. > > The specs are written based on the Mac OS X driver, and I think Apple > knows their hardware. Don't be so sure :) > After discussing both variants, Ren? Nussbaumer and myself prefer the > first one. > > Acked-by: Michael Hanselmann <linux-kernel at hansmi.ch> > Acked-by: Ren? Nussbaumer <linux-kernel at killerfox.forkbomb.ch> > > Greets, > Michael
