From: Andy Lutomirski <luto@kernel.org>
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>
Subject: Re: [PATCH 4/7] x86,tlb: make lazy TLB mode lazier
Date: Thu, 19 Jul 2018 10:04:09 -0700 [thread overview]
Message-ID: <CALCETrXLMsSBChDvrms-omwYV4LHT30GenDjbnD-+LTg55yPow@mail.gmail.com> (raw)
In-Reply-To: <CALCETrWDiMhgiR3f8n0jdWcW31EDJ+Waq0wh5sMDutfigANGnA@mail.gmail.com>
On Thu, Jul 19, 2018 at 9:45 AM, Andy Lutomirski <luto@amacapital.net> wrote:
> [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’t be hot, so I assume you’re 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?)
next prev parent reply other threads:[~2018-07-19 17:04 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-16 19:03 [PATCH v6 0/7] x86,tlb,mm: make lazy TLB mode even lazier Rik van Riel
2018-07-16 19:03 ` [PATCH 1/7] mm: allocate mm_cpumask dynamically based on nr_cpu_ids Rik van Riel
2018-07-17 9:33 ` [tip:x86/mm] mm: Allocate the mm_cpumask (mm->cpu_bitmap[]) " tip-bot for Rik van Riel
2018-08-04 22:28 ` [PATCH 1/7] mm: allocate mm_cpumask " Guenter Roeck
2018-07-16 19:03 ` [PATCH 2/7] x86,tlb: leave lazy TLB mode at page table free time Rik van Riel
2018-07-17 9:34 ` [tip:x86/mm] x86/mm/tlb: Leave " tip-bot for Rik van Riel
2018-07-17 11:46 ` Peter Zijlstra
2018-07-25 1:00 ` Anders Roxell
2018-08-16 1:54 ` [PATCH 2/7] x86,tlb: leave " Andy Lutomirski
2018-08-16 5:31 ` Rik van Riel
2018-07-16 19:03 ` [PATCH 3/7] x86,mm: restructure switch_mm_irqs_off Rik van Riel
2018-07-17 9:34 ` [tip:x86/mm] x86/mm/tlb: Restructure switch_mm_irqs_off() tip-bot for Rik van Riel
2018-10-09 14:58 ` tip-bot for Rik van Riel
2018-07-16 19:03 ` [PATCH 4/7] x86,tlb: make lazy TLB mode lazier Rik van Riel
2018-07-17 9:35 ` [tip:x86/mm] x86/mm/tlb: Make " tip-bot for Rik van Riel
2018-07-17 11:33 ` Peter Zijlstra
2018-07-18 15:33 ` Rik van Riel
2018-07-18 16:00 ` Peter Zijlstra
[not found] ` <081E558D-DB34-4A18-A35C-896BC47F6EBA@surriel.com>
2018-07-18 18:23 ` Peter Zijlstra
2018-07-18 18:51 ` Rik van Riel
2018-07-19 9:13 ` Peter Zijlstra
2018-07-17 20:04 ` [PATCH 4/7] x86,tlb: make " Andy Lutomirski
[not found] ` <FF977B78-140F-4787-AA57-0EA934017D85@surriel.com>
2018-07-17 21:29 ` Andy Lutomirski
2018-07-17 22:05 ` Rik van Riel
2018-07-17 22:27 ` Andy Lutomirski
2018-07-18 20:58 ` Rik van Riel
2018-07-18 23:13 ` Andy Lutomirski
[not found] ` <B976CC13-D014-433A-83DE-F8DF9AB4F421@surriel.com>
2018-07-19 16:45 ` Andy Lutomirski
2018-07-19 17:04 ` Andy Lutomirski [this message]
2018-07-20 4:57 ` Benjamin Herrenschmidt
2018-07-20 8:30 ` Peter Zijlstra
2018-07-23 12:26 ` Rik van Riel
2018-07-24 16:33 ` Will Deacon
[not found] ` <CF849A07-B7CE-4DE9-8246-53AC5A53A705@surriel.com>
2018-07-19 17:18 ` Andy Lutomirski
2018-07-20 8:02 ` Vitaly Kuznetsov
2018-07-20 9:49 ` Peter Zijlstra
2018-07-20 10:18 ` Vitaly Kuznetsov
2018-07-20 9:32 ` Peter Zijlstra
2018-07-20 11:04 ` Peter Zijlstra
2018-07-16 19:03 ` [PATCH 5/7] x86,tlb: only send page table free TLB flush to lazy TLB CPUs Rik van Riel
2018-07-17 9:35 ` [tip:x86/mm] x86/mm/tlb: Only " tip-bot for Rik van Riel
2018-07-17 11:39 ` Peter Zijlstra
[not found] ` <1F8BDD25-864D-4105-B872-2109AA417454@surriel.com>
[not found] ` <24AA4367-22A1-450E-8F6A-3CBF39518384@surriel.com>
2018-07-18 16:19 ` Peter Zijlstra
2018-07-16 19:03 ` [PATCH 6/7] x86,mm: always use lazy TLB mode Rik van Riel
2018-07-17 9:36 ` [tip:x86/mm] x86/mm/tlb: Always " tip-bot for Rik van Riel
2018-10-09 14:58 ` tip-bot for Rik van Riel
2018-07-16 19:03 ` [PATCH 7/7] x86,switch_mm: skip atomic operations for init_mm Rik van Riel
2018-07-17 9:36 ` [tip:x86/mm] x86/mm/tlb: Skip atomic operations for 'init_mm' in switch_mm_irqs_off() tip-bot for Rik van Riel
-- strict thread matches above, loose matches on Subject: below --
2018-07-10 14:28 [PATCH v5 0/7] x86,tlb,mm: make lazy TLB mode even lazier Rik van Riel
2018-07-10 14:28 ` [PATCH 4/7] x86,tlb: make lazy TLB mode lazier Rik van Riel
2018-07-06 21:56 [PATCH v4 0/7] x86,tlb,mm: make lazy TLB mode even lazier Rik van Riel
2018-07-06 21:56 ` [PATCH 4/7] x86,tlb: make lazy TLB mode lazier Rik van Riel
2018-06-29 14:29 [PATCH v3 0/7] x86,tlb,mm: make lazy TLB mode even lazier Rik van Riel
2018-06-29 14:29 ` [PATCH 4/7] x86,tlb: make lazy TLB mode lazier Rik van Riel
2018-06-29 17:05 ` Dave Hansen
2018-06-29 17:29 ` Rik van Riel
2018-06-20 19:56 [PATCH 0/7] x86,tlb,mm: make lazy TLB mode even lazier Rik van Riel
2018-06-20 19:56 ` [PATCH 4/7] x86,tlb: make lazy TLB mode lazier Rik van Riel
2018-06-22 15:04 ` Andy Lutomirski
2018-06-22 15:15 ` Rik van Riel
2018-06-22 15:34 ` Andy Lutomirski
2018-06-22 17:05 ` Dave Hansen
2018-06-22 17:16 ` Rik van Riel
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CALCETrXLMsSBChDvrms-omwYV4LHT30GenDjbnD-+LTg55yPow@mail.gmail.com \
--to=luto@kernel.org \
--cc=benh@kernel.crashing.org \
--cc=boris.ostrovsky@oracle.com \
--cc=catalin.marinas@arm.com \
--cc=dave.hansen@intel.com \
--cc=efault@gmx.de \
--cc=jgross@suse.com \
--cc=kernel-team@fb.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=riel@surriel.com \
--cc=vkuznets@redhat.com \
--cc=will.deacon@arm.com \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).