On 05/01/17 17:46, Chris Packham wrote: > On 05/01/17 17:04, Florian Fainelli wrote: >> Le 01/04/17 à 19:36, Chris Packham a écrit : >>> +} >>> + >>> +static int __init mv98dx3236_resume_init(void) >>> +{ >>> + struct device_node *np; >>> + struct resource res; >>> + int ret = 0; >>> + >>> + np = of_find_matching_node(NULL, of_mv98dx3236_resume_table); >>> + if (!np) >>> + return 0; >> >> Can't this function be implemented as a smp_ops::smp_init_cpus instead >> of having this initialization done at arch_initcall time? >> > > I'll look into it. My initial reaction is no because I still need > armada_xp_smp_init_cpus(). > I looked at this. I could write a mv98dx3236_smp_init_cpus() that calls armada_xp_smp_init_cpus() and inits the resume controller address. My original reason for this approach was to parallel mvebu_v7_pmsu_init. I won't do anything just yet but is there any downside to the current approach? -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html