https://bugzilla.kernel.org/show_bug.cgi?id=214899 --- Comment #39 from Hans de Goede (jwrdegoede@xxxxxxxxxxxxxxxxx) --- (In reply to Andy Shevchenko from comment #38) > (In reply to Hans de Goede from comment #36) > > Created attachment 301177 [details] > > Attempt 2 at fixing this (3 patches in a tarbal) > > I would propose, if it's feasible, to evaluate _HID/_CID (to be PCI host > bridge compatible) of the first level of \_SB and evaluate _INI of all of > them. The listing of the names seems a bit fragile, while I don't believe > there will be many new entries. That said, IIRC, \_SB.PCI is also present in > some DSDTs. That is an interesting proposal, but this happens really early on and for some devices _HID / _CID is not just an ACPI Name() field but an actual Method which does a bunch of things, like e.g. checking GPIOs to see which touchscreen/pad model is installed. Although it is unlikely to encounter such _CID / _HID methods on devices directly under _SB, it cannot be ruled out, so calling _CID + _HID for all devices under _SB seems unwise. Also note that these patches still need to change a bit, as I mentioned they cause trouble on a X1C8. I've a plan for this which I plan to implement today and then I'll send out a RFC series to the linux-acpi list (and also attach it here for Johannes to test the new version). -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
