On Mon, 2022-11-07 at 16:08 +0300, Dan Carpenter wrote: > On Mon, Nov 07, 2022 at 09:40:00AM +0530, Deepak R Varma wrote: > > Simplify code by using recommended min helper macro for logical > > evaluation and value assignment. This issue is identified by > > coccicheck using the minmax.cocci file. > > > > Signed-off-by: Deepak R Varma <drv@xxxxxxxxx> > > --- > > drivers/staging/iio/meter/ade7854-i2c.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/staging/iio/meter/ade7854-i2c.c b/drivers/staging/iio/meter/ade7854-i2c.c > > index a9a06e8dda51..a6ce7b24cc8f 100644 > > --- a/drivers/staging/iio/meter/ade7854-i2c.c > > +++ b/drivers/staging/iio/meter/ade7854-i2c.ck > > @@ -61,7 +61,7 @@ static int ade7854_i2c_write_reg(struct device *dev, > > unlock: > > mutex_unlock(&st->buf_lock); > > > > - return ret < 0 ? ret : 0; > > + return min(ret, 0); > > The original code is better. > > If it's a failure return the error code. If it's not return zero. > > You can only compare apples to apples. min() makes sense if you're > talking about two lengths. But here if ret is negative that's an error > code. If it's positive that's the number of bytes. If the error > code is less than the number of bytes then return that? What??? It > makes no sense. > > In terms of run time, this patch is fine but in terms of reading the > code using min() makes it less readable. It's not a runtime question, either should compile to the same object code. It's definitely a readabiity and standardization issue. In this case, IMO it'd be better to use the much more common if (ret < 0) return ret; return 0;
