APPLIED[M]: [PATCH 1/1][SRU][M][Unstable] UBUNTU: SAUCE: platform/x86: dell-uart-backlight: replace chars_in_buffer() with flush_chars()

[Andrea Righi]

On Wed, Aug 23, 2023 at 02:45:00PM +0800, AceLan Kao wrote:
> Andrea Righi <andrea.righi at canonical.com> 於 2023年8月23日 週三 下午2:05寫道：
> >
> > On Tue, Aug 22, 2023 at 11:57:40AM +0800, AceLan Kao wrote:
> > > From: "Chia-Lin Kao (AceLan)" <acelan.kao at canonical.com>
> > >
> > > BugLink: https://bugs.launchpad.net/bugs/2032174
> > >
> > > After v6.5, calls tty->driver->ops->chars_in_buffer() may lead to a
> > > deadlock and got soft lockup error. Try using tty->driver->ops->flush_chars()
> > > to archive to same scenario.
> > >
> > > There is no functional change, even if uart_flush_chars() doesn't do what
> > > we expected, the driver should be able to keep working, as the backlight
> > > commands is only 3 or 4 bytes short and has little chance to stay in the
> > > queue partially.
> > >
> > > Signed-off-by: Chia-Lin Kao (AceLan) <acelan.kao at canonical.com>
> >
> > Is there an upstream thread for this? If so it would be nice to add a
> > link, reference, or a "cherry picked from" line with the URL of the
> > upstream mail thread.
> 
> Hi Andrea,
> 
> No, currently dell-uart-backlight driver is Ubuntu only, so no upstream link.
> 
> This driver need to be re-written to use serdev to be able to submit
> to upstream,
> and serdev seems to have issues with ACPI ID table, I can't make it be loaded.
> Current drivers that uses serdev are for ARM system, although some of them also
> provide ACPI ID table in the driver, but I think it probably doesn't work.

Ah! I was missing that it's *all* SAUCE. Ok in that case, yes, there's
not an upstream thread of course.

Therefore, applied to mantic/linux-unstable.

Thanks for the clarification,
-Andrea