Re: [PATCH v2 01/10] boot-mode-reg: Add core

[Dirk Behme <dirk.behme@xxxxxxxxxxxx>]

Hi Geert,

On 11.05.2016 10:41, Geert Uytterhoeven wrote:
> Hi Dirk,
>
> On Wed, May 11, 2016 at 10:39 AM, Dirk Behme <dirk.behme@xxxxxxxxxxxx> wrote:
> > On 11.05.2016 09:54, Geert Uytterhoeven wrote:
> > > On Wed, May 11, 2016 at 7:29 AM, Dirk Behme <dirk.behme@xxxxxxxxxxxx>
> > > wrote:
> > > > --- /dev/null
> > > > +++ b/drivers/misc/boot-mode-reg/core.c
> > >
> > > > +/**
> > > > + * boot_mode_reg_set() - record boot mode register value
> > > > + * @mode: implementation-dependent boot mode register value
> > > > + *
> > > > + * Records the boot mode register value which may subsequently
> > > > + * be retrieved using boot_mode_reg_get().
> > > > + *
> > > > + * return: 0 on success
> > > > + */
> > > > +int boot_mode_reg_set(u32 mode)
> > > > +{
> > > > +       int err = -EBUSY;
> > > > +
> > > > +       mutex_lock(&boot_mode_mutex);
> > > > +       if (!boot_mode_is_set) {
> > >
> > > You've dropped the check for calling this function a subsequent time with
> > > a different value of mode?
> >
> > Sometimes inverting 'complex' if statements is not that easy ;)
> >
> > You mean
> >
> > if (!boot_mode_is_set || boot_mode != mode)
>
> No, De Morgan says
>
>         if (!boot_mode_is_set || boot_mode == mode)


Hmm, sorry if I don't have enough coffee today ;)

But why should we do

if (... boot_mode == mode)
	boot_mode = mode;

?

Or in other words:

We want to execute the if code if

boot_mode_is_set is false

or

boot_mode is not equal mode, i.e. mode has changed

?


Best regards

Dirk