On Fri, Sep 08, 2017 at 09:14:52PM -0700, Darren Hart wrote: > On Fri, Sep 08, 2017 at 03:15:10PM +0000, Mario.Limonciello@xxxxxxxx wrote: > > > -----Original Message----- > > > From: Mika Westerberg [mailto:mika.westerberg@xxxxxxxxxxxxxxx] > > > Sent: Friday, September 8, 2017 3:07 AM > > > To: Limonciello, Mario <Mario_Limonciello@xxxxxxxx> > > > Cc: dvhart@xxxxxxxxxxxxx; LKML <linux-kernel@xxxxxxxxxxxxxxx>; platform-driver- > > > x86@xxxxxxxxxxxxxxx; Richard Hughes <hughsient@xxxxxxxxx>; Yehezkel Bernat > > > <yehezkelshb@xxxxxxxxx> > > > Subject: Re: [PATCH v2] Add driver to force WMI Thunderbolt controller power > > > status > > > > > > On Thu, Sep 07, 2017 at 01:47:22PM -0500, Mario Limonciello wrote: > > > > > > I have a small comment which you might want to consider. > > > > > > > +static int intel_wmi_thunderbolt_probe(struct wmi_device *wdev) > > > > +{ > > > > + return sysfs_create_group(&wdev->dev.kobj, &tbt_attribute_group); > > > > +} > > > > + > > > > +static int intel_wmi_thunderbolt_remove(struct wmi_device *wdev) > > > > +{ > > > > + sysfs_remove_group(&wdev->dev.kobj, &tbt_attribute_group); > > > > + return 0; > > > > +} > > > > > > In order to notify userspace about this, what if you send KOBJ_CHANGE > > > for the wdev when you add/remove this attribute? > > > > Sure, makes sense to me. I'll send a follow up patch. > > Hrm, what am I missing.... the reason I recommended Mario add the attribute > group was so this would be automatically taken care of prior to the KOBJ_ADD > event, per the device documentation: > > driver-model/device.txt At this point the KOBJ_ADD event has already been send to the userspace when the device was added to the system.
