| Message ID | 20170223235613.5112-2-roderick@gaikai.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, 23 Feb 2017, Roderick Colenbrander wrote: > From: Roderick Colenbrander <roderick.colenbrander@sony.com> > > When a user connects a DS4 twice using USB and BT, we reject the > second device connection after the setup work. We then perform > a cleanup, but during cleanup we are not removing the touchpad > device. This leads to leakage of an input device, which we would > never remove. It can likely result into a kernel oops as well > when the touchpad evdev node is accessed and the underlaying HID > device has been removed from the system. Hi Roderick, thanks for the fix. Could you please ammend and resend with: - proper Signed-off-by: line - ideally 'Fixes:' tag as well Thanks,
diff --git a/drivers/hid/hid-sony.c b/drivers/hid/hid-sony.c index f405b07..740996f 100644 --- a/drivers/hid/hid-sony.c +++ b/drivers/hid/hid-sony.c @@ -2632,6 +2632,8 @@ static int sony_input_configured(struct hid_device *hdev, sony_leds_remove(sc); if (sc->quirks & SONY_BATTERY_SUPPORT) sony_battery_remove(sc); + if (sc->touchpad) + sony_unregister_touchpad(sc); sony_cancel_work_sync(sc); kfree(sc->output_report_dmabuf); sony_remove_dev_list(sc);
From: Roderick Colenbrander <roderick.colenbrander@sony.com> When a user connects a DS4 twice using USB and BT, we reject the second device connection after the setup work. We then perform a cleanup, but during cleanup we are not removing the touchpad device. This leads to leakage of an input device, which we would never remove. It can likely result into a kernel oops as well when the touchpad evdev node is accessed and the underlaying HID device has been removed from the system. --- drivers/hid/hid-sony.c | 2 ++ 1 file changed, 2 insertions(+)