[arch-general] xfce-powermanager issue since kernel 3.19

[Catalin Iacob]

On Wed, Mar 25, 2015 at 2:42 PM, Friedrich Strohmaier
<damokles4-listen at bits-fritz.de> wrote:
> Hi all,
>
> since having updated to kernel 3.19 xfce Power Manager Panel Plugin does not
> show my logitech trackball (m570) and keyboard (k230) both connected through
> wireless usb dongle.
> other logitech device (logitech m235 mouse) does.
>
> all are shown with lsusb:
> $ lsusb | grep Logitech
> Bus 004 Device 005: ID 046d:c52b Logitech, Inc. Unifying Receiver
> Bus 004 Device 004: ID 046d:c52f Logitech, Inc. Unifying Receiver
> Bus 003 Device 007: ID 046d:c52b Logitech, Inc. Unifying Receiver
>
> and all work fine.
>
> downgrading to kernel 3.18 solves this issue.
> Nevertheless I'm a little bit lost where to report what?
> --

Hi Friedrich,

It's a kernel regression an as such you should definitely report it to
LKML. Even if userspace was doing weird things, general kernel policy
is that if a kernel upgrade breaks userspace it's the kernel's fault
and the kernel should be fixed.

It doesn't hurt to narrow down the subsystem which might be causing
the problem and include the appropriate people and mailing lists in
the email.  I would start with the HID (input devices) subsystem
(althouh your problem may be more about power management, but still,
if it's not in their court, the HID people will tell you who to
contact or forward your email to the correct people). From
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/MAINTAINERS
it seems in addition to LKML you should mail:

HID CORE LAYER
M:    Jiri Kosina <jkosina at suse.cz>
L:    linux-input at vger.kernel.org

In my experience, if you report a regression people will work with you
to get it narrowed down and fixed.

Even better, if you are comfortable building kernels and can bisect to
identify the exact commit that broke this, you'll probably get a fix
within a day or two. And even better, it could be easy enough to
figure out what's wrong and fix it yourself and submit the patch. I've
done exactly this before for a bug which made sound stop working for
my particular laptop.

The kernel heavily relies on community testing; reporting and
following thorugh on this to get it fixed will help others as well.