[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <84c3696d-8108-4a2e-90d7-7830ca6cc3b9@amazon.com>
Date: Tue, 17 Sep 2024 22:55:20 +0200
From: Alexander Graf <graf@...zon.com>
To: "Jason A. Donenfeld" <Jason@...c4.com>
CC: Lennart Poettering <mzxreary@...inter.de>, <linux-kernel@...r.kernel.org>,
<stable@...r.kernel.org>, Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Linus Torvalds <torvalds@...ux-foundation.org>, Babis Chalios
<bchalios@...zon.es>, Theodore Ts'o <tytso@....edu>, "Cali, Marco"
<xmarcalx@...zon.co.uk>, Arnd Bergmann <arnd@...db.de>, "rostedt@...dmis.org"
<rostedt@...dmis.org>, Christian Brauner <brauner@...nel.org>,
<linux@...mhuis.info>, <regressions@...ts.linux.dev>, Paolo Bonzini
<pbonzini@...hat.com>, "Michael Kelley (LINUX)" <mikelley@...rosoft.com>,
Sean Christopherson <seanjc@...gle.com>
Subject: Re: [REGRESSION] Re: [PATCH] Revert "vmgenid: emit uevent when
VMGENID updates"
On 17.09.24 20:04, Jason A. Donenfeld wrote:
> On Thu, Jun 13, 2024 at 6:37 PM Alexander Graf <graf@...zon.com> wrote:
>> Friendly ping again. We would really like to have a constructive
>> technical conversation and collaboration on how to make forward progress
>> with VM clone notifications for user space applications that hold unique
>> data and hence need to learn about VM clone events, outside of any
>> randomness semantics.
> With the other work now mostly done, sure, let's pick this up again. I
> think next on the list was getting the virtio rng device delivering VM
> clone events and unique UUIDs. There was a spec change posted a while
> back and a patch for the kernel. Do you want to refresh those? I
> thought that was a promising direction -- and the one we all decided
> together in person as the most viable, race-free way, etc --
> especially because it would make ways of exposing those IDs low cost.
> And, importantly for you, I think that might *also* cover the need
> that you have here, so we'll kill several birds with one stone.
The virtio proposal only addressed consumers that require single atomic
memory updates to learn about any event that is disruptive to their
entropy sources. With vgetrandom and/or rdrand we solved that problem,
so we can close the chapter of that class of use cases.
What is still open are user space applications that require event based
notification on VM clone events - and *only* VM clone events. This
mostly caters for tools like systemd which need to execute policy - such
as generating randomly generated MAC addresses - in the event a VM was
cloned.
That's the use case this patch "vmgenid: emit uevent when VMGENID
updates" is about and I think the best path forward is to just revert
the revert. A uevent from the device driver is a well established, well
fitting Linux mechanism for that type of notification.
Alex
Amazon Web Services Development Center Germany GmbH
Krausenstr. 38
10117 Berlin
Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss
Eingetragen am Amtsgericht Charlottenburg unter HRB 257764 B
Sitz: Berlin
Ust-ID: DE 365 538 597
Powered by blists - more mailing lists