Re: [PATCH v4 02/13] media: media/test_drivers: Replace open-coded parity calculation with parity_odd()

On Thu, Apr 10, 2025 at 02:56:41AM +0800, Kuan-Wei Chiu wrote:
> On Wed, Apr 09, 2025 at 02:41:03PM -0400, Yury Norov wrote:
> > On Thu, Apr 10, 2025 at 02:23:09AM +0800, Kuan-Wei Chiu wrote:
> > > On Wed, Apr 09, 2025 at 01:03:42PM -0400, Yury Norov wrote:
> > > > On Wed, Apr 09, 2025 at 11:43:45PM +0800, Kuan-Wei Chiu wrote:

> > > > So, if val == 0 than parity_odd(val) is also 0, and this can be
> > > > simplified just to:
> > > >         return parity(val) ? 0 : 0x80;
> > > > Or I miss something?
> > > >
> > > If val == 0x01, the return value of calc_parity() will remain 0x01.
> > > If changed to return parity_odd(val) ? 0 : 0x80;, the return value will
> > > be changed to 0x00.
> > 
> > Sorry, I meant
> >         return val ? 0 : 0x80;
> > 
> > This 'val | (parity_odd(val)' is only false when val == 0, right?
> > When val != 0, compiler will return true immediately, not even calling
> > parity().
> >
> I'm still confused.
> 
> Maybe you're interpreting the code as:
> 
> 	(val | parity(val)) ? 0 : 0x80
> 
> But what we're trying to do is:
> 
> 	val | (parity(val) ? 0 : 0x80)
> 
> So, for example, if val == 0x06, the return value will be 0x86.
> Only returning 0 or 0x80 seems wrong to me.
> Or did I misunderstand something?

I misread the whole construction. Sorry, you're right. Scratch this.