lkml.org 
[lkml]   [2020]   [Feb]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 1/3] i2c: convert SMBus alert setup function to return an ERRPTR
On 17.02.20 10:14:52, Luca Ceresoli wrote:
> Hi,
>
> On 17/02/20 09:17, Wolfram Sang wrote:
> >
> >>> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> >>
> >>> -struct i2c_client *i2c_setup_smbus_alert(struct i2c_adapter *adapter,
> >>> +struct i2c_client *i2c_install_smbus_alert(struct i2c_adapter *adapter,
> >>> struct i2c_smbus_alert_setup *setup);
> >>
> >> This function naming is a bit odd. It creates a struct i2c_client.
> >> Then, there is also i2c_new_client_device() and i2c_new_device(). For
> >> i2c_new_client_device() there are no users at all outside of
> >> i2c-core-base.c (except for Falcon NIC), it is only a wrapper.
> >
> > i2c_new_device (and friends) returned NULL on error. I am currently
> > converting all i2c_new_* functions to return an ERRPTR. So,
> > i2c_new_client_device is the new function, i2c_new_device is deprecated.
> > If you check v5.6-rc1, you will find many more users. Similarily,
> > i2c_new_dummy is deprecated (and removed already), i2c_new_dummy_device
> > is the new thing.
> >
> >> So how about reducing the interface to those both only to:?
> >>
> >> i2c_new_device()
> >> i2c_new_device_smbus()
> >
> > Given the above, it would be:
> >
> > i2c_new_client_device()
> > i2c_new_smbus_device()
> >
> > Yet, I think this is too vague. Maybe
> >
> > i2c_new_smbus_alert_device()
>
> I always found the function naming a bit messy in the linux i2c
> implementation. Among the names proposed in this thread,
> i2c_new_smbus_alert_device() is the only one that makes sense to me for
> the new function: it is not vague, and it is coherent with other names
> of recently introduced functions (i2c_new_*_device()). Of course the
> "alert device" is not a real device, but it looks like it is as it has
> its own "slave" address. So I vote for this name...

Yes, better fix the naming now that the i/f is new. As all these
function create a i2c_client struct, the whole set of functions could
be also named i2c_client_create*() with specialized functions such as
i2c_client_create_smbus() or so. It will be clear it is a subset.

Anyway, it's just a function name, but while reading the code it was
not obvious to me that i2c_install_smbus_alert() is actually a subset
of i2c_new_client_device(). That said, I like the i2c_client_create*()
variants.

-Robert

\
 
 \ /
  Last update: 2020-02-17 11:01    [W:0.062 / U:2.648 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site