Re: [Outreachy kernel] [RESEND PATCH] staging: emxx_udc: Ending line with argument

[Beatriz Martins de Carvalho]

Em 06/04/21 20:56, Matthew Wilcox escreveu:
> On Tue, Apr 06, 2021 at 08:34:09PM +0100, Beatriz Martins de Carvalho wrote:
> > Cleans up check of "Lines should not end with a '('"
> > with argument present in next line in file emxx_udc.c
> I appreciate that you've removed the checkpatch warning, but this is
> still harder to read than the original used to be.
Thank you for your review.
> > -				_nbu2ss_writel(
> > -					&preg->EP_REGS[ep->epnum - 1].EP_WRITE,
> > -					p_buf_32->dw);
> > +				_nbu2ss_writel(&preg->EP_REGS[ep->epnum - 1].EP_WRITE,
> > +					       p_buf_32->dw);
> > -		length = _nbu2ss_readl(
> > -			&ep->udc->p_regs->EP_REGS[ep->epnum - 1].EP_LEN_DCNT);
> > +		length = _nbu2ss_readl(&ep->udc->p_regs->EP_REGS[ep->epnum - 1].EP_LEN_DCNT);
> > -			regdata = _nbu2ss_readl(
> > -				&preg->EP_REGS[ep->epnum - 1].EP_STATUS);
> > +			regdata = _nbu2ss_readl(&preg->EP_REGS[ep->epnum - 1].EP_STATUS);
> The real problem with this driver is that their abstraction layer is
> wrong.  For example:
>
>          /* Interrupt Status */
>          status = _nbu2ss_readl(&udc->p_regs->EP_REGS[num].EP_STATUS);
>
>          /* Interrupt Clear */
>          _nbu2ss_writel(&udc->p_regs->EP_REGS[num].EP_STATUS, ~status);
>
> If instead this were:
>
> 	status = nbu2ss_read_ep_status(udc, num);
> 	nbu2ss_write_ep_status(udc, num, ~status);
>
> that would be a lot shorter and clearer.  Cleanups along these lines
> would be a lot more useful, and would fix the 80 column warning.

I can see what you mean and makes sense. Thank you.

Beatriz Martins de Carvalho