Hi Domen,
On Fri, 18 May 2007 09:24:27 +0200, Domen Puncer wrote:
> On 17/05/07 12:38 +0200, Jean Delvare wrote:
> > On Wed, 16 May 2007 07:34:40 +0200, Manuel Lauss wrote:
> > > if (!(stat & PSC_SMBSTAT_TE) || !(stat & PSC_SMBSTAT_RE)) {
> > > - sp->psc_smbpcr = PSC_SMBPCR_DC;
> > > + au_writel(PSC_SMBPCR_DC, base + PSC_SMBPCR);
> > > au_sync();
> > > - do {
> > > - stat = sp->psc_smbpcr;
> > > - au_sync();
> > > - } while ((stat & PSC_SMBPCR_DC) != 0);
> > > + while (au_readl(base + PSC_SMBPCR) & PSC_SMBPCR_DC)
> > > + msleep(0);
> >
> > You are changing the behavior here, while this patch is supposed to
> > only convert the driver to the new device driver model.
>
> Well... since msleep(0) is nothing, it's the same, but it does
> look weird.
msleep(0) isn't nothing. It sleeps until the next tick.
--
Jean Delvare
