On Wed, Aug 23, 2017 at 11:47:52PM +0100, Mark Brown wrote:
support for the Chromebook), I think that 36afb0ab648 and 55e59aa0525a are wrong and should be reverted. It seems like they're an attempt to make 89128534f925 be done 'properly' but it also seems like the
Please be more specific. The only obvious issue with the original patch "ASoC: rt5677: Add ACPI support" is that it adds an I2C ID instead of an ACPI ID. I don't have 36afb0ab648 so I've no idea what it is and 55e59aa0525a is "ASoC: rt5677: Move platform code to board file" which is a code motion patch and looks more like stylistic faff around the shambles that is ACPI than anything else.
My best guess if you're using mainline is that this is triggered by a36afb0ab6488ea (ASoC: rt5677: Introduce proper table for ACPI enumeration) causing the ACPI core code to run and explode on whatever you've got in the tables on that system. Someone who knows what ACPI is up to should probably dig into what's going on, even if reverting fixes it that looks worryingly like we might explode on other devices.