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: Perf: WARNING: arch/x86/entry/common.c:624 idtentry_exit_cond_rcu+0x92/0xc0
Date:   Fri, 12 Jun 2020 02:18:07 -0700
Message-Id: <8DD3180E-0E69-4FD6-92C3-311AAB3F688F@amacapital.net>
References: <87imfwd5f6.fsf@nanos.tec.linutronix.de>
Cc:     Andy Lutomirski <luto@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        Naresh Kamboju <naresh.kamboju@linaro.org>,
        open list <linux-kernel@vger.kernel.org>,
        X86 ML <x86@kernel.org>, cj.chengjian@huawei.com,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Ingo Molnar <mingo@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>, Borislav Petkov <bp@alien8.de>,
        Minchan Kim <minchan@kernel.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Michel Lespinasse <walken@google.com>,
        lkft-triage@lists.linaro.org,
        Dave Hansen <dave.hansen@linux.intel.com>
In-Reply-To: <87imfwd5f6.fsf@nanos.tec.linutronix.de>
To:     Thomas Gleixner <tglx@linutronix.de>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Jun 12, 2020, at 2:01 AM, Thomas Gleixner <tglx@linutronix.de> wrote:
>=20
> =EF=BB=BFAndy Lutomirski <luto@kernel.org> writes:
>> On Thu, Jun 11, 2020 at 4:22 PM Andy Lutomirski <luto@kernel.org> wrote:
>>=20
>> Two bugs here.
>>=20
>> 1. We had an issue with WARN. Patch sent.
>=20
> Grabbed it
>=20
>> 2. idtentry.h has, for x86_32:
>>=20
>> # define DEFINE_IDTENTRY_IST  DEFINE_IDTENTRY
>>=20
>> This is nonsense.  It's getting late over here and I'd rather focus on
>> the more interesting RCU issue, so that's all from me today.
>=20
> Well, this might be nonsense, but it's exactly matching the current code
> in mainline which, e.g. for #DB does:
>=20
> SYM_CODE_START(debug)
>    /*
>     * Entry from sysenter is now handled in common_exception
>     */
>    ASM_CLAC
>    pushl    $0
>    pushl    $do_debug
>    jmp    common_exception
> SYM_CODE_END(debug)
>=20
> There is no IST on 32bit, never was. We do software stack switching for
> device interrupts, but that's a different story.
>=20

DEFINE_IDTENTRY does the idtentry_enter_cond_rcu() dance, which isn=E2=80=99=
t intended to be safe from NMI context.  It should probably map to DEFINE_ID=
TENTRY_RAW() instead.  The specific issue is that NMI ends up there, and at l=
east DEFINE_IDTENTRY_NMI should be raw.

I haven=E2=80=99t tried this at all, nor have I dug through all the users of=
 these macros to check what they expect.  Perhaps we should not have the _IS=
T one defined at all on 32 bit and rename it to DEFINE_IDTENTRY_IST_RAW on 6=
4 bit to make it more clear what=E2=80=99s going on when reading the C code.=


Or maybe I=E2=80=99m too sleepy and I=E2=80=99m nuts. But I don=E2=80=99t th=
ink I am.


> Thanks,
>=20
>        tglx