Re: [PATCH 5.4 063/355] powerpc/kvm: Fix kvm_use_magic_page
Andreas Gruenbacher <[email protected]>
| Newsgroups | org.kernel.vger.kvm-ppc,org.kernel.vger.linux-kernel,org.kernel.vger.stable |
|---|---|
| Message-ID | <CAHc6FU74ih7Sk7tWmvMy9OsyO2=f0cO7fUoBREKnFQSKf+zyig@mail.gmail.com> |
On Tue, Nov 16, 2021 at 3:15 PM Greg Kroah-Hartman <[email protected]> wrote: > On Tue, Nov 16, 2021 at 03:00:19PM +0100, Andreas Gruenbacher wrote: > > On Tue, Nov 16, 2021 at 2:57 PM Greg Kroah-Hartman > > <[email protected]> wrote: > > > On Tue, Nov 16, 2021 at 02:54:10PM +0100, Andreas Gruenbacher wrote: > > > > On Tue, Nov 16, 2021 at 1:54 AM Sasha Levin <[email protected]> wrote: > > > > > On Mon, Nov 15, 2021 at 06:47:41PM +0100, Andreas Gruenbacher wrote: > > > > > >Greg, > > > > > > > > > > > >On Mon, Nov 15, 2021 at 6:10 PM Greg Kroah-Hartman > > > > > ><[email protected]> wrote: > > > > > >> From: Andreas Gruenbacher <[email protected]> > > > > > >> > > > > > >> commit 0c8eb2884a42d992c7726539328b7d3568f22143 upstream. > > > > > >> > > > > > >> When switching from __get_user to fault_in_pages_readable, commit > > > > > >> 9f9eae5ce717 broke kvm_use_magic_page: like __get_user, > > > > > >> fault_in_pages_readable returns 0 on success. > > > > > > > > > > > >I've not heard back from the maintainers about this patch so far, so > > > > > >it would probably be safer to leave it out of stable for now. > > > > > > > > > > What do you mean exactly? It's upstream. > > > > > > > > Mathieu Malaterre broke this test in 2018 (commit 9f9eae5ce717) but > > > > that wasn't noticed until now (commit 0c8eb2884a42). This means that > > > > this fix probably isn't critical, so I shouldn't be backported. > > > > > > Then why did you tag it to be explicitly backported to all stable > > > kernels newer than 4.18? > > > > Well, sorry for that. What else do you expect me to do in addition to > > pointing out the mistake? > > Ah, I think we are misunderstanding each other here. I will go drop it, > but in the future maybe "hey, I didn't mean to mark this for stable, can > you please drop it?" might be a bit more direct and to the point. Yes, the stable Cc was an accident on my side (caused by a script that takes a SHA1 and spits out Fixes: and Cc: tags). Sorry and thanks for dropping the patch. Andreas