dev_printk() is now GPL-only
Does dev_driver_string() really need to be GPL-only? Up to this point, proprietary modules have been entitled to call dev_printk(), but now:
#define dev_printk(level, dev, format, arg...) \
printk(level "%s %s: " format , dev_driver_string(dev) , (dev)->bus_id , ## arg)with
EXPORT_SYMBOL_GPL(dev_driver_string);
means that they're not allowed to.
On a related note, one might wonder if
return dev->driver ? dev->driver->name : (dev->bus ? dev->bus->name : "");
really qualifies for the full weight of copyright enforcement, and query whether using it makes your driver a derived work. Particularly since all one is doing is asking the kernel for a nice printk prefix and know nothing about the innards of the device model.
Quoted 4 lines · Matthew Wilcox
Does
dev_driver_string()really need to be GPL-only? Up to this point, proprietary modules have been entitled to calldev_printk(), but now:
Quoted 8 lines
#define dev_printk(level, dev, format, arg...) \ printk(level "%s %s: " format , dev_driver_string(dev) , (dev)->bus_id , ## arg)with
EXPORT_SYMBOL_GPL(dev_driver_string);means that they're not allowed to.
Oops, good point, I never noticed, as I don't have an closed source drivers here to test with :)
Care to send me a patch to fix it up?
thanks,
greg k-h
Quoted 1 line · Greg KH
Care to send me a patch to fix it up?
Sorry for the delay ...
Make dev_printk usable from non-GPL modules again
dev_printk now calls dev_driver_string. We want even proprietary modules to be calling dev_printk, so the export of dev_driver_string needs to be non-GPL-only.
drivers/base/core.c +1 -1
@@ -42,7 +42,7 @@ const char *dev_driver_string(struct devreturn dev->driver ? dev->driver->name : (dev->bus ? dev->bus->name :"" ); } -EXPORT_SYMBOL_GPL (dev_driver_string); +EXPORT_SYMBOL (dev_driver_string);#define to_dev (obj) container_of(obj, struct device, kobj)#define to_dev_attr (_attr) container_of(_attr, struct device_attribute, attr)