Hi Jiri, Looking at hid-core.c, I see the following: dev_set_name(&hdev->dev, "%04X:%04X:%04X.%04X", hdev->bus, hdev->vendor, hdev->product, atomic_inc_return(&id)); This looks plain wrong to me. The vendor and product IDs can be attributes of the devices, but they should not be part of the device name aka bus_id. The bus_id is about the address of the device, not what the device is. Just look at how the PCI or USB subsystems do, or virtually any other kernel subsystem, as I don't think any other subsystem does what HID is doing right now. On top of this, using an auto-incrementing device ID in the bus_id looks wrong too. For one thing, it will cycle after 65536 requests, so the above does not guarantee uniqueness over time. If anything, you should used an idr instead. But even that would probably be wrong, because this scheme itself doesn't guarantee stability of the bus_id over disconnect/reconnect, reboot, etc. While I agree it's not always possible to guarantee stability especially for externally pluggable devices, we should try to keep names as stable as we can. The above makes me wonder if handling hid as a bus is really the right thing to do, as apparently your devices do not have an address you can use to generate a good-looking bus_id. Shouldn't hid devices be class devices instead? -- Jean Delvare -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
