On 12/9/07, Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
>
> On Sun, 2007-12-09 at 15:57 -0500, Jon Smirl wrote:
> >
> > Are there technical concerns with this series? The white space can be
> > fixed in a few minutes.
> >
> > Adding a tag to differentiate matching types has implications that are
> > broader than just i2c. Shouldn't we do this first with the existing
> > scheme and then change the tagging process with later patches?
>
> No, we should decide on what to do with the tagging process (or not do)
> first, don't you think ? (If we need a tagging process, Scott had a
> concern but it might be moot, let's discuss that first).
Right now the tags are simply strings. The second parameter is driver specific.
+static struct i2c_device_id rs5c372_id[] = {
+ {"rtc-rs5c372", rtc_rs5c372a},
+ {"rs5c372a", rtc_rs5c372a},
+ {"rs5c372b", rtc_rs5c372b},
+ {"rv5c386", rtc_rv5c386},
+ {"rv5c387a", rtc_rv5c387a},
+ {"ricoh,rs5c372a", rtc_rs5c372a},
+ {"ricoh,rs5c372b", rtc_rs5c372b},
+ {"ricoh,rv5c386", rtc_rv5c386},
+ {"ricoh,rv5c387a", rtc_rv5c387a},
+ {},
+};
The current mechanism is simple string matching there are no platform
specific namespaces.
We could wrap the device tree style names in a macro that adds a
non-printable character to the front.
Something like this:
#define DT_NAMESPACE "\1"
#define DT_NAME(x) (DT_NAMESPACE x)
+ {DT_NAME("ricoh,rv5c386"), rtc_rv5c386},
And then modify the mpc i2c driver to insert the DT_NAMESPACE in front
of the string.
Another solution would be to make the names disappear on non-device
tree platforms
in mod_devicetable.h:
#ifdef USING_DEVICE_TREES
#define DT_NAME(x) x
#else
#define DTNAME(x)
#endif
+static struct i2c_device_id rs5c372_id[] = {
+ {"rtc-rs5c372", rtc_rs5c372a},
+ {"rs5c372a", rtc_rs5c372a},
+ {"rs5c372b", rtc_rs5c372b},
+ {"rv5c386", rtc_rv5c386},
+ {"rv5c387a", rtc_rv5c387a},
+ DT_NAME({"ricoh,rs5c372a", rtc_rs5c372a},)
+ DT_NAME({"ricoh,rs5c372b", rtc_rs5c372b},)
+ DT_NAME({"ricoh,rv5c386", rtc_rv5c386},)
+ DT_NAME({"ricoh,rv5c387a", rtc_rv5c387a},)
+ {},
But what's the point in making these names specific to device trees?
They are perfectly valid names for the devices that could be used from
any platform.
--
Jon Smirl
[EMAIL PROTECTED]
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@ozlabs.org
https://ozlabs.org/mailman/listinfo/linuxppc-dev
