[lkml]   [2017]   [Mar]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH] serial: Do not treat the IIR register as a bitfield
On Fri, Mar 31, 2017 at 4:54 PM, Olliver Schinagl
<> wrote:
> On 30-03-17 11:56, Andy Shevchenko wrote:
>> On Wed, 2017-03-29 at 20:44 +0200, Olliver Schinagl wrote:

>> Looking to implementation I would rather go with some helper like
>> int serial_in_IIR(port, [additional mask])
>> {
>> return port->serial_in(port, UART_IIR) & (_IIR_MASK [| additional
>> mask]);
>> }

> As I just wrote a simply static inline helper function in serial_core.h, I
> just figured that the helper will only work for some of the calls. All
> interrupt checks in xxx_serial_in() obviously can't rely on this. So do you
> still want this helper function added for the other cases? Or have all
> implementations do the masking manually?

You have still few places (3+ IIRC) where it makes sense.

> And then, is iir = serial_port_in(up, UART_IIR) & UART_IIR_MASK; preferred
> over splitting it over two lines, like I did?

With given indentation it might be long enough to uglify the code.

So, I would still go with one / two helpers (do your own choice), but
if you insist that is not beneficial I would not object in-place

static inline int serial_in_IIR_mask(port, mask)
return ... & mask;

static inline int serial_in_IIR(port)
return serial_in_IIR_mask(port, ..._IIR_MASK);

> Finally, why rename it to _IIR_MASK, I assume a typo here?

I usually do such to minimize characters to type (notice leading _
which means I referred to a suffix) and that's why the work "like" is
used above implying you need to modify to function correctly.

With Best Regards,
Andy Shevchenko

 \ /
  Last update: 2017-03-31 16:45    [W:0.068 / U:1.752 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site