Mainline Linux-libre for Chromebook pixel 2013?

[Iru Cai]

On Sun, Nov 29, 2015 at 11:00 PM, Alexandre Oliva <lxoliva at fsfla.org> wrote:

> On Nov 28, 2015, Alexandre Oliva <lxoliva at fsfla.org> wrote:
>
> > On Nov 28, 2015, Iru Cai <mytbk920423 at gmail.com> wrote:
> >> I bought a Chromebook pixel 2013 recently. When it runs Trisquel with
> >> Linux-libre 3.13, the touchscreen and touchpad work fine. However, in
> later
> >> kernel release, it needs to load non-free firmware in mxt_initialize(),
> >> which is in drivers/input/touchscreen/atmel_mxt_ts.c, and the
> atmel_mxt_ts
> >> driver fails to work. Is there a work around for this?
>
> > I suspect blacklisting the specific driver might get the device to
> > fallback to generic emulation mode, which is presumably what you were
> > get on 3.13 (assuming there isn't a specific driver there).
>

Sorry for not cc the replied mail to the mailing list when I was replying
the previous mails.

>
> I've taken a look at the driver, and it is buggy.  Even though it works
> without the blob, it treats an -EINVAL return from r*t_firmware_nowait
> as a hard error, instead of ignoring as it could/should, to work even
> upstream, when the firmware loader is disabled altogether in the build
> configuration.
>
I forgot to mention one thing. The reject_firmware_nowait sometimes causes
a kernel panic in atmel_mxt_ts driver. I didn't save the kernel error
messages, but I think there may be something wrong with the reject firmware
logic.

>
> We could work around that bug in GNU Linux-libre, using an alternate
> call that fakes a success return status, but...  I'd much rather it be
> fixed upstream, so that we can keep local changes to a minimum.


> Would you report that, saying it doesn't work with the firmware loader
> disabled, or if request_firmware_nowait fails for whatever reason, even
> though you know it could work just fine?
>
I meant that it works if I changed reject_firmware_nowait to
request_firmware_nowait, although the kernel will say it fails to load
firmware. It doesn't give a 'Failed to invoke firmware loader.' error.

>
> In the mean time, I suggest patching the driver so that it doesn't take
> the error status from reject_firmware_nowait as fatal, reporting it but
> ignoring it.
>
> --
> Alexandre Oliva, freedom fighter    http://FSFLA.org/~lxoliva/
> You must be the change you wish to see in the world. -- Gandhi
> Be Free! -- http://FSFLA.org/   FSF Latin America board member
> Free Software Evangelist|Red Hat Brasil GNU Toolchain Engineer
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.fsfla.org/pipermail/linux-libre/attachments/20151129/c582b80d/attachment.html>