https://bugzilla.kernel.org/show_bug.cgi?id=204807 --- Comment #155 from Andy Shevchenko (andy.shevchenko@xxxxxxxxx) --- (In reply to Eugene Shalygin from comment #154) > (In reply to Olli Asikainen from comment #153) > > Yeah, it's old and suffers from the same bug, but I reckon supporting this > > would be a completely different story. Thanks for your answer Eugene and > > sorry for the noise. > > This firmware seem to be using ACPI mutex named \_SB_.PCI0.LPCB.SIO1.MUT0 to > guard access to the nct chip registers. > > I wonder, can we rely on these mutexes, whose names seem to be quite stable, > as well, instead of the WMI functions? Then accessing nct registers would > become simpler: lock ACPI mutex if needed, access registers always in the > regular way, unlock the mutex. There is a _SEM (semaphore) handling example in the upstream: https://elixir.bootlin.com/linux/latest/source/drivers/i2c/busses/i2c-designware-baytrail.c https://elixir.bootlin.com/linux/latest/source/arch/x86/platform/intel/iosf_mbi.c#L275 maybe something similar can be done for ACPI Mutex() object. -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
