[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Fri, 8 Oct 2021 03:10:26 +0300
From: Konstantin Belousov <kostikbel@...il.com>
To: libc-coord@...ts.openwall.com
Cc: David Holland <dholland-libc@...bsd.org>
Subject: Re: freezero() and freezeroall()
On Thu, Oct 07, 2021 at 05:00:43PM -0700, Keith Packard wrote:
> Hrm. If this page has ever been written to swap, it sure would be nice
> to be able to erase that copy as well, otherwise there's a persisted
> version of the data which survives power off.
For instance we (FreeBSD) have no way to track all possible locations in
a swap where the page was written sometime in the past. I doubt that other
operating systems can do that.
Or during physical memory defragmentation, the page might have been
copied and replaced by another physical page.
Also imagine a VM running under hypervisor, and hypervisor deciding to page
out some VM memory. We have no way to indicate a desire to treat this
specific page as sensitive, to it.
Overall, I think it is only feasible to clear _the page_ itself, without
even trying to provide a kind of guarantee that its data is completely
vanished from the whole system.
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
