On 31.01.22 10:40, Oscar Salvador wrote:
> On Mon, Jan 31, 2022 at 08:48:54AM +0100, David Hildenbrand wrote:
>> Hi Oscar,
>
> Hi David :-),
>
>> Right, and the idea is that the online state of nodes (+ node/zone
>> ranges) already has to be known at that point in time, because
>> otherwise, we'd be in bigger trouble.
>
> Yeah, I wanted to check where exactly did we mark the nodes online,
> and for the few architectures I checked it happens in setup_arch(),
> which is called very early in start_kernel(), while driver_init()
> gets called through arch_call_rest_init(), which happens at the end
> of the function.
>
> I am not sure whether we want to remark that somehow in the changelog,
> so it is crystal clear that by the time the node_dev_init() gets called,
> we already set the nodes online.
>
> Anyway, just saying, but is fine as is.
I'll adjust the first paragraph to:
... and call node_dev_init() after memory_dev_init() from driver_init(),
so before any of the existing arch/subsys calls. All online nodes should
be known at that point: early during boot, arch code determines node and
zone ranges and sets the relevant nodes online; usually this happens in
setup_arch().
Thanks!
--
Thanks,
David / dhildenb
