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;
Date:   Thu, 16 Jul 2020 12:26:59 +1000
From:   Nicholas Piggin <npiggin@gmail.com>
Subject: Re: [RFC PATCH 7/7] lazy tlb: shoot lazies, a non-refcounting lazy
 tlb option
To:     Andy Lutomirski <luto@amacapital.net>
Cc:     Anton Blanchard <anton@ozlabs.org>, Arnd Bergmann <arnd@arndb.de>,
        linux-arch <linux-arch@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux-MM <linux-mm@kvack.org>,
        linuxppc-dev <linuxppc-dev@lists.ozlabs.org>,
        Andy Lutomirski <luto@kernel.org>,
        Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
        Peter Zijlstra <peterz@infradead.org>, X86 ML <x86@kernel.org>
References: <1594708054.04iuyxuyb5.astroid@bobo.none>
        <6D3D1346-DB1E-43EB-812A-184918CCC16A@amacapital.net>
In-Reply-To: <6D3D1346-DB1E-43EB-812A-184918CCC16A@amacapital.net>
MIME-Version: 1.0
Message-Id: <1594866023.tjwr7nhhg2.astroid@bobo.none>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Excerpts from Andy Lutomirski's message of July 14, 2020 10:46 pm:
>=20
>=20
>> On Jul 13, 2020, at 11:31 PM, Nicholas Piggin <npiggin@gmail.com> wrote:
>>=20
>> =EF=BB=BFExcerpts from Nicholas Piggin's message of July 14, 2020 3:04 p=
m:
>>> Excerpts from Andy Lutomirski's message of July 14, 2020 4:18 am:
>>>>=20
>>>>> On Jul 13, 2020, at 9:48 AM, Nicholas Piggin <npiggin@gmail.com> wrot=
e:
>>>>>=20
>>>>> =EF=BB=BFExcerpts from Andy Lutomirski's message of July 14, 2020 1:5=
9 am:
>>>>>>> On Thu, Jul 9, 2020 at 6:57 PM Nicholas Piggin <npiggin@gmail.com> =
wrote:
>>>>>>>=20
>>>>>>> On big systems, the mm refcount can become highly contented when do=
ing
>>>>>>> a lot of context switching with threaded applications (particularly
>>>>>>> switching between the idle thread and an application thread).
>>>>>>>=20
>>>>>>> Abandoning lazy tlb slows switching down quite a bit in the importa=
nt
>>>>>>> user->idle->user cases, so so instead implement a non-refcounted sc=
heme
>>>>>>> that causes __mmdrop() to IPI all CPUs in the mm_cpumask and shoot =
down
>>>>>>> any remaining lazy ones.
>>>>>>>=20
>>>>>>> On a 16-socket 192-core POWER8 system, a context switching benchmar=
k
>>>>>>> with as many software threads as CPUs (so each switch will go in an=
d
>>>>>>> out of idle), upstream can achieve a rate of about 1 million contex=
t
>>>>>>> switches per second. After this patch it goes up to 118 million.
>>>>>>>=20
>>>>>>=20
>>>>>> I read the patch a couple of times, and I have a suggestion that cou=
ld
>>>>>> be nonsense.  You are, effectively, using mm_cpumask() as a sort of
>>>>>> refcount.  You're saying "hey, this mm has no more references, but i=
t
>>>>>> still has nonempty mm_cpumask(), so let's send an IPI and shoot down
>>>>>> those references too."  I'm wondering whether you actually need the
>>>>>> IPI.  What if, instead, you actually treated mm_cpumask as a refcoun=
t
>>>>>> for real?  Roughly, in __mmdrop(), you would only free the page tabl=
es
>>>>>> if mm_cpumask() is empty.  And, in the code that removes a CPU from
>>>>>> mm_cpumask(), you would check if mm_users =3D=3D 0 and, if so, check=
 if
>>>>>> you just removed the last bit from mm_cpumask and potentially free t=
he
>>>>>> mm.
>>>>>>=20
>>>>>> Getting the locking right here could be a bit tricky -- you need to
>>>>>> avoid two CPUs simultaneously exiting lazy TLB and thinking they
>>>>>> should free the mm, and you also need to avoid an mm with mm_users
>>>>>> hitting zero concurrently with the last remote CPU using it lazily
>>>>>> exiting lazy TLB.  Perhaps this could be resolved by having mm_count
>>>>>> =3D=3D 1 mean "mm_cpumask() is might contain bits and, if so, it own=
s the
>>>>>> mm" and mm_count =3D=3D 0 meaning "now it's dead" and using some car=
eful
>>>>>> cmpxchg or dec_return to make sure that only one CPU frees it.
>>>>>>=20
>>>>>> Or maybe you'd need a lock or RCU for this, but the idea would be to
>>>>>> only ever take the lock after mm_users goes to zero.
>>>>>=20
>>>>> I don't think it's nonsense, it could be a good way to avoid IPIs.
>>>>>=20
>>>>> I haven't seen much problem here that made me too concerned about IPI=
s=20
>>>>> yet, so I think the simple patch may be good enough to start with
>>>>> for powerpc. I'm looking at avoiding/reducing the IPIs by combining t=
he
>>>>> unlazying with the exit TLB flush without doing anything fancy with
>>>>> ref counting, but we'll see.
>>>>=20
>>>> I would be cautious with benchmarking here. I would expect that the
>>>> nasty cases may affect power consumption more than performance =E2=80=
=94 the=20
>>>> specific issue is IPIs hitting idle cores, and the main effects are to=
=20
>>>> slow down exit() a bit but also to kick the idle core out of idle.=20
>>>> Although, if the idle core is in a deep sleep, that IPI could be=20
>>>> *very* slow.
>>>=20
>>> It will tend to be self-limiting to some degree (deeper idle cores
>>> would tend to have less chance of IPI) but we have bigger issues on
>>> powerpc with that, like broadcast IPIs to the mm cpumask for THP
>>> management. Power hasn't really shown up as an issue but powerpc
>>> CPUs may have their own requirements and issues there, shall we say.
>>>=20
>>>> So I think it=E2=80=99s worth at least giving this a try.
>>>=20
>>> To be clear it's not a complete solution itself. The problem is of=20
>>> course that mm cpumask gives you false negatives, so the bits
>>> won't always clean up after themselves as CPUs switch away from their
>>> lazy tlb mms.
>>=20
>> ^^
>>=20
>> False positives: CPU is in the mm_cpumask, but is not using the mm
>> as a lazy tlb. So there can be bits left and never freed.
>>=20
>> If you closed the false positives, you're back to a shared mm cache
>> line on lazy mm context switches.
>=20
> x86 has this exact problem. At least no more than 64*8 CPUs share the cac=
he line :)
>=20
> Can your share your benchmark?

It's just context_switch1_threads from will-it-scale, running with 1/2
the number of CPUs, and core affinity with an SMT processor (so each
thread from the switching pairs gets spread to their own CPU and so you
get the task->idle->task switching.

It's really just about the worst case, so I wouldn't say it's something
to panic about.

Thanks,
Nick