Hey Matt, On Tue, 24 Dec 2019, Matt Roper wrote: > > When audio power domain is suspended, the display driver must > > save state of AUD_FREQ_CNTRL on Tiger Lake and Ice Lake > > systems. The initial value of the register is set by BIOS and > > I realize this patch landed several months ago, but I was just glancing > through places in the driver where we call out specific platforms with > IS_FOO() rather than using generation tests and noticed this one. > Should this programming be specific to just ICL and TGL, or should it > also apply to other recent platforms like EHL/JSL? > > Our convention in i915 is to usually just assume that future platforms > will follow the lead of the current latest platform until we find out > otherwise. So we may want to add another patch to change the test to this is a valid point. The current check is very limited as the issue has been only observed on these platforms (but was very severe on these devices so a quick response was needed). I did observe originally that specs would indicate the register should be saved on many other platforms as well. I was hesitant to add this to the patch, as there are many platforms that have been shipping for years with no issues reported on this. But, but, I think we should just proceed and extend the check to all documented platforms, past and future (basicly INTEL_GEN>=9). I can make a patch and let's see how it fares in CI. Br, Kai _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
