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;
MIME-Version: 1.0
In-Reply-To: <CALCETrWDiMhgiR3f8n0jdWcW31EDJ+Waq0wh5sMDutfigANGnA@mail.gmail.com>
References: <20180716190337.26133-1-riel@surriel.com> <20180716190337.26133-5-riel@surriel.com>
 <CALCETrUMntRe_LYX5wj9YD4xZ+84QExK+ZNb3yxEBDEKa7nePQ@mail.gmail.com>
 <B6566AA4-2E8A-43BD-A224-0F6D4747B8FB@surriel.com> <F93A3335-4830-4E36-9179-D0FA5D640C8A@amacapital.net>
 <B976CC13-D014-433A-83DE-F8DF9AB4F421@surriel.com> <CALCETrWDiMhgiR3f8n0jdWcW31EDJ+Waq0wh5sMDutfigANGnA@mail.gmail.com>
From:   Andy Lutomirski <luto@kernel.org>
Date:   Thu, 19 Jul 2018 10:04:09 -0700
Message-ID: <CALCETrXLMsSBChDvrms-omwYV4LHT30GenDjbnD-+LTg55yPow@mail.gmail.com>
Subject: Re: [PATCH 4/7] x86,tlb: make lazy TLB mode lazier
To:     Rik van Riel <riel@surriel.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Juergen Gross <jgross@suse.com>,
        Boris Ostrovsky <boris.ostrovsky@oracle.com>,
        linux-arch <linux-arch@vger.kernel.org>,
        Will Deacon <will.deacon@arm.com>,
        Catalin Marinas <catalin.marinas@arm.com>,
        linux-s390@vger.kernel.org,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        linuxppc-dev <linuxppc-dev@lists.ozlabs.org>
Cc:     Andy Lutomirski <luto@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>, X86 ML <x86@kernel.org>,
        Mike Galbraith <efault@gmx.de>,
        kernel-team <kernel-team@fb.com>, Ingo Molnar <mingo@kernel.org>,
        Dave Hansen <dave.hansen@intel.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Jul 19, 2018 at 9:45 AM, Andy Lutomirski <luto@amacapital.net> wrot=
e:
> [I added PeterZ and Vitaly -- can you see any way in which this would
> break something obscure?  I don't.]
>
> On Thu, Jul 19, 2018 at 7:14 AM, Rik van Riel <riel@surriel.com> wrote:
>> I guess we can skip both switch_ldt and load_mm_cr4 if real_prev equals
>> next?
>
> Yes, AFAICS.
>
>>
>> On to the lazy TLB mm_struct refcounting stuff :)
>>
>>>
>>> Which refcount?  mm_users shouldn=E2=80=99t be hot, so I assume you=E2=
=80=99re talking about
>>> mm_count. My suggestion is to get rid of mm_count instead of trying to
>>> optimize it.
>>
>>
>> Do you have any suggestions on how? :)
>>
>> The TLB shootdown sent at __exit_mm time does not get rid of the
>> kernelthread->active_mm
>> pointer pointing at the mm that is exiting.
>>
>
> Ah, but that's conceptually very easy to fix.  Add a #define like
> ARCH_NO_TASK_ACTIVE_MM.  Then just get rid of active_mm if that
> #define is set.  After some grepping, there are very few users.  The
> only nontrivial ones are the ones in kernel/ and mm/mmu_context.c that
> are involved in the rather complicated dance of refcounting active_mm.
> If that field goes away, it doesn't need to be refcounted.  Instead, I
> think the refcounting can get replaced with something like:
>
> /*
>  * Release any arch-internal references to mm.  Only called when
> mm_users is zero
>  * and all tasks using mm have either been switch_mm()'d away or have had
>  * enter_lazy_tlb() called.
>  */
> extern void arch_shoot_down_dead_mm(struct mm_struct *mm);
>
> which the kernel calls in __mmput() after tearing down all the page
> tables.  The body can be something like:
>
> if (WARN_ON(cpumask_any_but(mm_cpumask(...), ...)) {
>   /* send an IPI.  Maybe just call tlb_flush_remove_tables() */
> }
>
> (You'll also have to fix up the highly questionable users in
> arch/x86/platform/efi/efi_64.c, but that's easy.)
>
> Does all that make sense?  Basically, as I understand it, the
> expensive atomic ops you're seeing are all pointless because they're
> enabling an optimization that hasn't actually worked for a long time,
> if ever.

Hmm.  Xen PV has a big hack in xen_exit_mmap(), which is called from
arch_exit_mmap(), I think.  It's a heavier weight version of more or
less the same thing that arch_shoot_down_dead_mm() would be, except
that it happens before exit_mmap().  But maybe Xen actually has the
right idea.  In other words, rather doing the big pagetable free in
exit_mmap() while there may still be other CPUs pointing at the page
tables, the other order might make more sense.  So maybe, if
ARCH_NO_TASK_ACTIVE_MM is set, arch_exit_mmap() should be responsible
for getting rid of all secret arch references to the mm.

Hmm.  ARCH_FREE_UNUSED_MM_IMMEDIATELY might be a better name.

I added some more arch maintainers.  The idea here is that, on x86 at
least, task->active_mm and all its refcounting is pure overhead.  When
a process exits, __mmput() gets called, but the core kernel has a
longstanding "optimization" in which other tasks (kernel threads and
idle tasks) may have ->active_mm pointing at this mm.  This is nasty,
complicated, and hurts performance on large systems, since it requires
extra atomic operations whenever a CPU switches between real users
threads and idle/kernel threads.

It's also almost completely worthless on x86 at least, since __mmput()
frees pagetables, and that operation *already* forces a remote TLB
flush, so we might as well zap all the active_mm references at the
same time.

But arm64 has real HW remote flushes.  Does arm64 actually benefit
from the active_mm optimization?  What happens on arm64 when a process
exits?  How about s390?  I suspect that x390 has rather larger systems
than arm64, where the cost of the reference counting can be much
higher.

(Also, Rik, x86 on Hyper-V has remote flushes, too. How does that
interact with your previous patch set?)