Thread (15 messages) 15 messages, 5 authors, 2025-09-16

Re: [PATCH v5 2/3] iio: adc: Support ROHM BD79112 ADC/GPIO

From: Andy Shevchenko <hidden>
Date: 2025-09-16 07:39:46
Also in: linux-gpio, linux-iio, lkml

On Tue, Sep 16, 2025 at 7:48 AM Matti Vaittinen
[off-list ref] wrote:
On 15/09/2025 17:12, Andy Shevchenko wrote:
quoted
On Mon, Sep 15, 2025 at 10:12:43AM +0300, Matti Vaittinen wrote:
...
quoted
quoted
+    data->vref_mv = ret / 1000;
I still think moving to _mV is the right thing to do.
There is no 'mv' in the physics for Volts.
I can see you think so :) For me it doesn't look good. This is in-kernel
C-code not physics textbook.
Yes, and the science is superior to the users of it, isn't it?
For the kernel C it has been convention to
_not_ use capital letters (or CamelCase) for variables.
That convention is for normal variables, the variables that uses real
physical units can be better recognised as such.
This convention
is strong enough reason for me to avoid mV in a variable name because
the capital letter instantly requires my attention and makes me need to
consider if this is "just a variable". What comes to the vref_mv, there
really are no true downside. It is clear what the _mv suffix denotes and
"there is no 'mv' in physics" is really an artificial problem.
Common sense should be used and not some stubbornness over a
"convention". Convention is not something that is carved in stone and
mustn't ever be changed. And here is the exact example, that sometimes
conventions also suck. And the problem is not artificial, the units in
physics are unique in a representation. If you make them all in
case-insensitive, it might be a room for a collision (I can't provide
one from top of my head, but something like 'k' for 'kilo' is not 'K'
for 'Kelvin degree')..


...
quoted
quoted
+    devm_spi_optimize_message(dev, spi, &data->read_msg);
And if it fails?..
I am not really sure under what conditions this would fail. Without
taking a further look at that - then we just use unoptimized SPI
transfers(?). Could warrant a warning print though.
What is the point of having devm_ variant for it if it never fails, please?

-- 
With Best Regards,
Andy Shevchenko
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help