Re: [PATCH] ACPI: GED: Support GpioInt event resources

[Andy Shevchenko]

On Thu, Oct 01, 2026 at 09:48:56AM +0530, Rahul Pon wrote:
> On Thu, Oct 1, 2026 at 1:52 AM Linus Walleij <linusw@xxxxxxxxxx> wrote:
> > I don't understand the patch at all, but I understand that Andy or Mika must
> > review it otherwise it's not going anywhere, so looping them in.
> 
> Thanks, Linus.
> 
> Andy, Mika: some context that should have been in the commit message.
> 
> ACPI 6.6 section 5.6.9.2 says a GED _CRS contains only Interrupt
> resource descriptors, so this firmware is outside the spec. It is still
> the only source of lid events here: LIGE's _CRS is a single GpioInt on
> the GPIO controller (pin 960), and the controller's _AEI lists one
> other, unrelated pin. ECGE is the same, with pin 768. On Windows, the
> in-box ACPI device driver (acpidev.inf) binds both GED devices, and the
> lid works there.
> 
> For _EVT, section 5.6.9.3 passes the GSI number, which a GpioInt does
> not have. The patch passes the pin number from the GpioInt descriptor,
> as section 5.6.5.3 does for _EVT under _AEI. Both _EVT methods on this
> laptop ignore Arg0.
> 
> If the approach is acceptable, v2 will state the spec deviation in the
> commit message.

Is vendor of the firmware / platform being informed? Do they accept their
mistake? Are they going to fix their crap?

-- 
With Best Regards,
Andy Shevchenko