Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Andy Lutomirski <luto@amacapital.net>
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH v2 0/2] Replace and improve "mcsafe" with copy_safe()
Date:   Thu, 30 Apr 2020 18:20:39 -0700
Message-Id: <D47C71D3-349B-49C4-9945-330C9F42A3E0@amacapital.net>
References: <CAHk-=wh1SPyuGkTkQESsacwKTpjWd=_-KwoCK5o=SuC3yMdf7A@mail.gmail.com>
Cc:     Dan Williams <dan.j.williams@intel.com>,
        "Luck, Tony" <tony.luck@intel.com>,
        Andy Lutomirski <luto@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Borislav Petkov <bp@alien8.de>,
        stable <stable@vger.kernel.org>,
        the arch/x86 maintainers <x86@kernel.org>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Paul Mackerras <paulus@samba.org>,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Erwin Tsaur <erwin.tsaur@intel.com>,
        Michael Ellerman <mpe@ellerman.id.au>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        linux-nvdimm <linux-nvdimm@lists.01.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
In-Reply-To: <CAHk-=wh1SPyuGkTkQESsacwKTpjWd=_-KwoCK5o=SuC3yMdf7A@mail.gmail.com>
To:     Linus Torvalds <torvalds@linux-foundation.org>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Apr 30, 2020, at 5:25 PM, Linus Torvalds <torvalds@linux-foundation.org=
> wrote:
>=20
>=20
> It wasn't clear how "copy_to_mc()" could ever fault. Poisoning
> after-the-fact? Why would that be preferable to just mapping a dummy
> page?

If the kernel gets an async memory error and maps a dummy page, then subsequ=
ent reads will subsequently succeed and return garbage when they should fail=
.  If x86 had write-only pages, we could map a dummy write-only page. But we=
 don=E2=80=99t, so I think we=E2=80=99re stuck.

As for naming the kind of memory we=E2=80=99re taking about, ISTM there are t=
wo classes: DAX and monstrous memory-mapped non-persistent cache devices.  B=
oth could be Optane, I suppose.

But I also think it=E2=80=99s legitimate to use these accessors to increase t=
he chance of surviving a failure of normal memory. If a normal page happens t=
o be page cache when it fails and if page cache access use these fancy acces=
sors, then we might actually survive a failure.

We could be ambitious: declare that all page cache and all get_user_page=E2=80=
=99d memory should use the new accessors.  I doubt we=E2=80=99ll ever really=
 succeed due to magical things like rseq and anything that thinks that users=
 can set up their own memory as a kernel-accessed ring buffer, but I suppose=
 we could try.