David Hess (davidwhess.delete@this.gmail.com) on March 4, 2021 1:44 pm wrote:
> dmcq (dmcq.delete@this.fano.co.uk) on March 4, 2021 6:40 am wrote:
> >
> > The initialisation problem could have been easily solved by simply only initialising a page when
> > it was allocated, or ovcerwriting it if it is set from disk. You'd still get some seconds spent
> > on the job but the user wouldn't notice. I can't really believe they wouldn't think of that never
> > mind that booting took a long time anyway so this would be no sort of great gain.
>
> I have not noticed any delay on my workstation which has 64G.
>
> The entire memory would have to be initialized before memory scrubbing could be used. Without scrubbing,
> memory could be initialized as written but this would return errors if uninitialized memory was accessed.


That actually raises a broader question. I wonder how real scrubbing implementations work in the context of a page with a (hard) error. You really want to stop scrubbing on such pages, at least once the OS has decommissioned the page, or you're just going to pointlessly fill your error logs with duplicate messages.