On Fri, 26 Mar 2004, Edgar Toernig wrote:
> Sridhar Samudrala wrote:
> >
> > The following patch to 2.6.5-rc2 consolidates 6 different implementations
> > of msecs to jiffies and 3 different implementation of jiffies to msecs.
> > All of them now use the generic msecs_to_jiffies() and jiffies_to_msecs()
> > that are added to include/linux/time.h
> >[...]
> > -#define MSECS(ms) (((ms)*HZ/1000)+1)
> > -return (((ms)*HZ+999)/1000);
> > +return (msecs / 1000) * HZ + (msecs % 1000) * HZ / 1000;
>
> Did you check that all users of the new version will work correctly
> with your rounding? Explicit round-up of delays is often required,
> especially when talking to hardware...
I don't see any issues with the 2.6 default HZ value of 1000 as they become
no-ops and there is no need for any rounding.
I guess you are referring to cases when HZ < 1000(ex: 100) and msecs is
less than 10. In those cases, the new version returns 0, whereas some of the
older versions return 1.
If i am not mistaken, Jeff Garjik/David Miller are the maintainers for most
of the users of these routines and i have got an OK from them.
drivers/block/carmel.c
drivers/net/tulip/de204x.c
include/linux/libata.h
include/net/irda/irda.h
drivers/atm/fore200e.c
include/net/sctp/sctp.h
The only other place where the older version is different is
drivers/char/watchdot/shwdt.c
Dave, Jeff
Do you see any issues with the new generic versions of these routines?
Thanks
Sridhar
>
> Ciao, ET.
>
|