On Fri, Sep 10, 2010 at 23:12:36, Caglar Akyuz wrote: > On Thursday 09 September 2010 01:40:49 pm Sugumar Natarajan wrote: > > This patch adds generic PWM support where it maintains the list of PWM > > control devices that can be added or removed. > > > > The interface provides a list of functions that can be accessed by the > > PWM control driver module and the generic PWM driver. > > > > The PWM control driver module such as eCAP uses the interface to > > register and add itself to the list as a PWM control device. > > The generic PWM driver uses the interface to search for a PWM control > > device and if present, uses the device for PWM control. > > > > Signed-off-by: Sugumar Natarajan <[email protected]> > > --- > > Hi, > > [...] > > > + > > +int pwm_enable(struct pwm_device *pwm) { > > + if (WARN_ON(!pwm)) > > + return -EINVAL; > > + > > + return clk_enable(pwm->clk); > > +} > > +EXPORT_SYMBOL(pwm_enable); > > + > > I think 'clk_enable' should be controlled for un-matched pwm_enable/disable > operation. Otherwise, disabling PWM won't shut down peripheral clock. If I understand inner workings of DaVinci clock implementation clock use_count is incremented with every clk_enable. > > Regards, > Caglar >
Caglar, I could not understand what you mean by "un-match" .could you please explain? Regards, N.sugumar _______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
