[PATCH 0/4][SRU][Focal] fix ACPI EC initisation failures

[Colin Ian King]

On 28/09/2020 20:14, Alex Hung wrote:
> BugLink: https://bugs.launchpad.net/bugs/1896482
> 
> [Impact]
> 
> The bug causes the EC driver to fail and ACPI events are no longer
> handled by Linux kernel, i.e lid close no longer triggers suspends.
> 
> [Fix]
> 
> Upstream commit 03e9a0e05739cf reworks ec_install_handlers to avoid
> initialisation failures.
> 
> The other three patches are prerequisite to apply 03e9a0e05739cf on
> focal kernel.
> 
> [Test]
> 
> Tested with Lenovo ThinkPad X1 Carbon Gen 8
> 
> [Regression Potential]
> 
> Low. The patches were cherry-picked from mainline kernel (two since 5.5
> and two since 5.7)
> 
> Daniel Drake (2):
>   ACPI: EC: tweak naming in preparation for GpioInt support
>   ACPI: EC: add support for hardware-reduced systems
> 
> Rafael J. Wysocki (2):
>   ACPI: EC: Avoid passing redundant argument to functions
>   ACPI: EC: Consolidate event handler installation code
> 
>  drivers/acpi/ec.c       | 218 +++++++++++++++++++++++++++++++++---------------
>  drivers/acpi/internal.h |   3 +-
>  2 files changed, 154 insertions(+), 67 deletions(-)
> 

The first two patches have been around for a year now so have bedded in
a bit. I checked and found the following commit fixes an issue in the
2nd commit:

commit c823c17a8ea4db4943152388a0beb9a556715716
Author: Rafael J. Wysocki <rafael.j.wysocki at intel.com>
Date:   Thu Feb 27 22:51:36 2020 +0100

    ACPI: EC: Avoid printing confusing messages in acpi_ec_setup()

..is this required too?


Also I see the comment in acpi_ec_add():

+       /*
+        * Ignore the GPE value on Reduced Hardware platforms.
+        * Some products have this set to an erroneous value.
+        */

I guess that's the usual "firmware is defined incorrectly" problem, so
what happens then?

Finally, as a side note, is this something we have checks for in fwts?

Colin