Comment 3 for bug 1964992

Doesn't this just invert the problem? Before, the user couldn't set the max to below the default min. But setting the min to above the default max would have worked. After this patch, what if the user tries to set the min to above the default max? Won't that now fail?

I don't see any discussion of this upstream. Surely both min and max need to be considered together for a proper fix? Maybe it's not a problem in practice for some reason I'm missing?

Separately, another possible regression is that a user has bad tunables set that aren't currently having any effect, and by issuing a fix those bad tunables will take effect and regress behaviour. That's something for users to look out for, but I don't think we should hold back a fix for a real bug for this reason.