Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH 2/2] x86/numa: instance all parsed numa node
From:   Andy Lutomirski <luto@amacapital.net>
In-Reply-To: <CAFgQCTui7D6_FQ_v_ijj6k_=+TQzQ3PaGvzxd6p+XEGjQ2S6jw@mail.gmail.com>
Date:   Tue, 9 Jul 2019 07:34:08 -0600
Cc:     Thomas Gleixner <tglx@linutronix.de>, x86@kernel.org,
        Michal Hocko <mhocko@suse.com>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        Mike Rapoport <rppt@linux.ibm.com>,
        Tony Luck <tony.luck@intel.com>,
        Andy Lutomirski <luto@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Vlastimil Babka <vbabka@suse.cz>,
        Oscar Salvador <osalvador@suse.de>,
        Pavel Tatashin <pavel.tatashin@microsoft.com>,
        Mel Gorman <mgorman@techsingularity.net>,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Michael Ellerman <mpe@ellerman.id.au>,
        Stephen Rothwell <sfr@canb.auug.org.au>, Qian Cai <cai@lca.pw>,
        Barret Rhoden <brho@google.com>,
        Bjorn Helgaas <bhelgaas@google.com>,
        David Rientjes <rientjes@google.com>, linux-mm@kvack.org,
        LKML <linux-kernel@vger.kernel.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <4AF3459B-28F2-425F-8E4B-40311DEF30C6@amacapital.net>
References: <1562300143-11671-1-git-send-email-kernelfans@gmail.com> <1562300143-11671-2-git-send-email-kernelfans@gmail.com> <alpine.DEB.2.21.1907072133310.3648@nanos.tec.linutronix.de> <CAFgQCTvwS+yEkAmCJnsCfnr0JS01OFtBnDg4cr41_GqU79A4Gg@mail.gmail.com> <alpine.DEB.2.21.1907081125300.3648@nanos.tec.linutronix.de> <CAFgQCTvAOeerLHQvgvFXy_kLs=H=CuUFjYE+UAN+vhPCG+s=pQ@mail.gmail.com> <alpine.DEB.2.21.1907090810490.1961@nanos.tec.linutronix.de> <CAFgQCTui7D6_FQ_v_ijj6k_=+TQzQ3PaGvzxd6p+XEGjQ2S6jw@mail.gmail.com>
To:     Pingfan Liu <kernelfans@gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Jul 9, 2019, at 1:24 AM, Pingfan Liu <kernelfans@gmail.com> wrote:
>=20
>> On Tue, Jul 9, 2019 at 2:12 PM Thomas Gleixner <tglx@linutronix.de> wrote=
:
>>=20
>>> On Tue, 9 Jul 2019, Pingfan Liu wrote:
>>>> On Mon, Jul 8, 2019 at 5:35 PM Thomas Gleixner <tglx@linutronix.de> wro=
te:
>>>> It can and it does.
>>>>=20
>>>> That's the whole point why we bring up all CPUs in the 'nosmt' case and=

>>>> shut the siblings down again after setting CR4.MCE. Actually that's in f=
act
>>>> a 'let's hope no MCE hits before that happened' approach, but that's al=
l we
>>>> can do.
>>>>=20
>>>> If we don't do that then the MCE broadcast can hit a CPU which has some=

>>>> firmware initialized state. The result can be a full system lockup, tri=
ple
>>>> fault etc.
>>>>=20
>>>> So when the MCE hits a CPU which is still in the crashed kernel lala st=
ate,
>>>> then all hell breaks lose.
>>> Thank you for the comprehensive explain. With your guide, now, I have
>>> a full understanding of the issue.
>>>=20
>>> But when I tried to add something to enable CR4.MCE in
>>> crash_nmi_callback(), I realized that it is undo-able in some case (if
>>> crashed, we will not ask an offline smt cpu to online), also it is
>>> needless. "kexec -l/-p" takes the advantage of the cpu state in the
>>> first kernel, where all logical cpu has CR4.MCE=3D1.
>>>=20
>>> So kexec is exempt from this bug if the first kernel already do it.
>>=20
>> No. If the MCE broadcast is handled by a CPU which is stuck in the old
>> kernel stop loop, then it will execute on the old kernel and eventually r=
un
>> into the memory corruption which crashed the old one.
>>=20
> Yes, you are right. Stuck cpu may execute the old do_machine_check()
> code. But I just found out that we have
> do_machine_check()->__mc_check_crashing_cpu() to against this case.
>=20
> And I think the MCE issue with nr_cpus is not closely related with
> this series, can
> be a separated issue. I had question whether Andy will take it, if
> not, I am glad to do it.
>=20
>=20

Go for it. I=E2=80=99m not familiar enough with the SMP boot stuff that I wo=
uld be able to do it any faster than you. I=E2=80=99ll gladly help review it=
.=