Re: [PATCH v4 1/7] staging: r8188eu: remove unnecessary braces in single statement block

["Fabio M. De Francesco" <fmdefrancesco@gmail.com>]

On giovedì 21 aprile 2022 21:52:14 CEST Fabio M. De Francesco wrote:
> On giovedì 21 aprile 2022 21:20:26 CEST Jaehee Park wrote:
> > Remove braces for single statement block to minimize the number of
> > empty lines, without loss of readability. Issue found with checkpatch.
> > WARNING: braces {} are not necessary for single statement blocks
> > 
> > Signed-off-by: Jaehee Park <jhpark1013@gmail.com>
> > ---
> >  drivers/staging/r8188eu/core/rtw_mlme.c | 3 +--
> >  1 file changed, 1 insertion(+), 2 deletions(-)
> > 
> > diff --git a/drivers/staging/r8188eu/core/rtw_mlme.c b/drivers/staging/
> r8188eu/core/rtw_mlme.c
> > index 3e9882f89f76..d3f4d7cdfa08 100644
> > --- a/drivers/staging/r8188eu/core/rtw_mlme.c
> > +++ b/drivers/staging/r8188eu/core/rtw_mlme.c
> > @@ -112,9 +112,8 @@ void _rtw_free_mlme_priv(struct mlme_priv 
*pmlmepriv)
> >  
> >  	rtw_free_mlme_priv_ie_data(pmlmepriv);
> >  
> > -	if (pmlmepriv) {
> > +	if (pmlmepriv)
> >  		vfree(pmlmepriv->free_bss_buf);
> > -	}
> >  
> >  }
> >  
> > -- 
> > 2.25.1
> > 
> 
> Hi Jaehee,
> 
> If you just want to remove the braces, your patch is good.
> 
> But... Did you ask yourself why this code needs to test if "pmlmepriv" is 
> valid before calling vfree() on it? Actually, the code doesn't need this 
> test, it just needs to call "vfree(pmlmepriv)" unconditionally.

When copy-pasting I overlooked a couple of things... However the argument 
is the same: the code doesn't need that 'if' statement before vfree().

I wanted to write "[] it just needs to call "vfree(pmlmepriv-
>free_bss_buf)" unconditionally".

I apologize for not reading twice before sending my email too soon after 
writing and moving sentences around.

> OK, I know that this is not directly related to your patch or any other 
> that is explicitly made just for removing unnecessary braces... 

"pmlmepriv" is valid because we know it from the previous call to 
rtw_free_mlme_priv_ie_data(). Therefore the code can dereference 
"pmlmepriv".

Furthermore, even if "free_bss_buf" were NULL it wouldn't matter because 
"vfree(pmlmepriv->free_bss_buf)" would be a no-op.

Sorry if this inaccurate wording might have confused you about what I was 
trying to say.

Thanks,

Fabio