Re: [PATCH v5 1/3] gpio: nuvoton: Add Nuvoton NPCM sgpio driver

[Jim Liu <jim.t90615@xxxxxxxxx>]

Hi  Paul

Thanks for your reply.

> If this is paid work, using your company email address should be
> preferred in my opinion.

because I can't use company email to send the patch upstream.
So BMC NPCM all drivers always  use gmail to send the patch.
It's the company mail security rule. I am sorry about that.
After the V3 version I changed to gmail.


> One byte would also fit into `unsigned int`, wouldn’t it?

Yes, so should i change u8 ,u16, u32 to standard `unsigned int` is
better than now?
Could you provide more information?

Best regards,
Jim



On Mon, Jul 24, 2023 at 10:44 PM Paul Menzel <pmenzel@xxxxxxxxxxxxx> wrote:
>
> Dear Jim,
>
>
> Am 24.07.23 um 05:04 schrieb Jim Liu:
>
> > sorry for reply late.
>
> No problem. Thank you for your reply. Some minor comments below.
>
> > First, thanks for your review.
> >
> > the description is as below:
> >
> > The SGPIO module can be programmed to support from zero (none) to
> > eight external output ports ,
>
> No space before the comma.
>
> > each with eight output pins (for a total of 64 output pins). The
> > output ports must be serial-to-parallel devices (such as the HC595 or
> > a faster equivalent).
> >
> > The SGPIO can be programmed to accept from zero to eight external
> > input ports (IXPp), each with eight input pins, supporting a total of
> > 64 input pins. The input ports must be parallel-to-serial devices
> > (such as the HC165 or a faster equivalent).
> >
> > you can add hc595 and hc165 ic to get the serial data from BMC and
> > send serial data to BMC.
> > This driver can expand  extra gpio pins up to 64 input and 64 output.
>
> One space before “extra”. Maybe:
>
> hc595 and c165 ic allow to transmit serial data from and to the BMC.
> This driver can expand extra GPIO pins up to 64 inputs and 64 outputs.
>
> > i will use jim.t90615@xxxxxxxxx this mail to upstream this driver not
> > company mail.
>
> If this is paid work, using your company email address should be
> preferred in my opinion.
>
> > The driver needs to fix the length of the variables, because the reg
> > size is one byte.
>
> One byte would also fit into `unsigned int`, wouldn’t it?
>
> > I will follow your suggestion to modify and upstream again. If you
> > have any questions please let me know.
>
> If you could use Mozilla Thunderbird to reply easily in interleaved
> style, that would great.
>
> Otherwise, I am looking forward to the next revision.
>
>
> Kind regards,
>
> Paul