Hi, After doing some testing on the new refcount_t series, here are some changes to improve refcount_t for the paranoid case. ;)
Since the new LKDTM tests depend on the new API, please add them to your series; it doesn't make sense for it to go through drivers-misc. The first patch (Kconfig fix) could be trivially folded into the initial refcount_t patch, too. Thanks! -Kees