On Mon, 2017-08-14 at 18:24 +0200, Borislav Petkov wrote:
> On Mon, Aug 14, 2017 at 03:57:35PM +0000, Kani, Toshimitsu wrote:
> > Hmm... Sorry, I failed to see how your patchset solved it. Would
> > you mind to explain how it is done?
>
> +static int __init ghes_edac_register(void)
> {
> + struct ghes_edac_pvt *pvt = ghes_pvt;
>
> Only one local ghes_pvt structure.
>
> And you handle multiple calls into ghes_edac_register() by exiting
> all those which are not the first one as the first one already did
> all the init.
Right, but the issue is how [ghes_edac_]report_mem_error() protects
from possible concurrent calls from multiple GHES sources when there is
only a single mci.
Thanks,
-Toshi
