On Thursday, August 25, 2022 6:21 AM, Justin He wrote: > > > @@ -566,3 +549,35 @@ void ghes_edac_unregister(struct ghes *ghes) > > > unlock: > > > mutex_unlock(&ghes_reg_mutex); > > > } > > > + > > > +static int __init ghes_edac_init(void) { > > > + struct ghes *g, *g_tmp; > > > + > > > + if (!IS_ENABLED(CONFIG_X86)) > > > + force_load = true; > > > > No, this is not how this works. > > > > > + ghes_devs = ghes_get_devices(force_load); > > > + if (!ghes_devs) > > > + return -ENODEV; > > > > You simply need to check force_load here. > > > > Okay, hence should I export the *ghes_devs* in ghes? It does not matter. This series then moves the force_load check to ghes_edac_preferred(). It is confusing for reviewers... Please divide patches based on the final design. Toshi