Re: [PATCH] Staging: rtl8712: Addressed checkpatch.pl issues related to macro parameter wrapping in parentheses.

From: Greg Kroah-Hartman
Date: Mon Jun 15 2020 - 08:34:54 EST


On Mon, Jun 15, 2020 at 10:28:51AM +0100, Ricardo Ferreira wrote:
> On Sun, 14 Jun 2020 at 15:05, Greg Kroah-Hartman
> <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> >
> > On Sun, Jun 14, 2020 at 02:51:25PM +0100, Ricardo Ferreira wrote:
> > > #define init_h2fwcmd_w_parm_no_rsp(pcmd, pparm, code) \
> > > do {\
> > > - INIT_LIST_HEAD(&pcmd->list);\
> > > - pcmd->cmdcode = code;\
> > > - pcmd->parmbuf = (u8 *)(pparm);\
> > > - pcmd->cmdsz = sizeof(*pparm);\
> > > - pcmd->rsp = NULL;\
> > > - pcmd->rspsz = 0;\
> > > + INIT_LIST_HEAD(&(pcmd)->list);\
> > > + (pcmd)->cmdcode = code;\
> > > + (pcmd)->parmbuf = (u8 *)((pparm));\
> > > + (pcmd)->cmdsz = sizeof(*(pparm));\
> > > + (pcmd)->rsp = NULL;\
> > > + (pcmd)->rspsz = 0;\
> > > } while (0)
> >
> > Does that change really make any sense? checkpatch is a nice hint,
> > sometimes it is not correct...
>
> (Replying again since I mistakenly sent my comments only to Greg...)
>
> Yeah I was over-eager and applied some of checkpatche's patches
> without thinking twice... I guess the parenthesis wrapping only makes
> sense when you have an operator (either binary or unary). I've
> rechecked each macro identified by checkpatch to see if there is a
> need for parenthesis wrapping in their current usage.

Yes, please do that, and also test-build your patches. Sending patches
that break the build are a sure way to make maintainers grumpy :)

thanks,

greg k-h