From: Nicholas Piggin <npiggin@gmail.com>
To: Andy Lutomirski <luto@kernel.org>
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>,
Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
Peter Zijlstra <peterz@infradead.org>, X86 ML <x86@kernel.org>
Subject: Re: [RFC PATCH 7/7] lazy tlb: shoot lazies, a non-refcounting lazy tlb option
Date: Tue, 14 Jul 2020 02:48:15 +1000 [thread overview]
Message-ID: <1594658283.qabzoxga67.astroid@bobo.none> (raw)
In-Reply-To: <CALCETrWbD=3SUOuq9P7Syb+a1DoBjjem8hq9_HCvn7wyqETkpw@mail.gmail.com>
Excerpts from Andy Lutomirski's message of July 14, 2020 1:59 am:
> On Thu, Jul 9, 2020 at 6:57 PM Nicholas Piggin <npiggin@gmail.com> wrote:
>>
>> On big systems, the mm refcount can become highly contented when doing
>> a lot of context switching with threaded applications (particularly
>> switching between the idle thread and an application thread).
>>
>> Abandoning lazy tlb slows switching down quite a bit in the important
>> user->idle->user cases, so so instead implement a non-refcounted scheme
>> that causes __mmdrop() to IPI all CPUs in the mm_cpumask and shoot down
>> any remaining lazy ones.
>>
>> On a 16-socket 192-core POWER8 system, a context switching benchmark
>> with as many software threads as CPUs (so each switch will go in and
>> out of idle), upstream can achieve a rate of about 1 million context
>> switches per second. After this patch it goes up to 118 million.
>>
>
> I read the patch a couple of times, and I have a suggestion that could
> be nonsense. You are, effectively, using mm_cpumask() as a sort of
> refcount. You're saying "hey, this mm has no more references, but it
> 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 refcount
> for real? Roughly, in __mmdrop(), you would only free the page tables
> if mm_cpumask() is empty. And, in the code that removes a CPU from
> mm_cpumask(), you would check if mm_users == 0 and, if so, check if
> you just removed the last bit from mm_cpumask and potentially free the
> mm.
>
> 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
> == 1 mean "mm_cpumask() is might contain bits and, if so, it owns the
> mm" and mm_count == 0 meaning "now it's dead" and using some careful
> cmpxchg or dec_return to make sure that only one CPU frees it.
>
> 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.
I don't think it's nonsense, it could be a good way to avoid IPIs.
I haven't seen much problem here that made me too concerned about IPIs
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 the
unlazying with the exit TLB flush without doing anything fancy with
ref counting, but we'll see.
Thanks,
Nick
next prev parent reply other threads:[~2020-07-13 16:48 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-10 1:56 [RFC PATCH 0/7] mmu context cleanup, lazy tlb cleanup, Nicholas Piggin
2020-07-10 1:56 ` [RFC PATCH 1/7] asm-generic: add generic MMU versions of mmu context functions Nicholas Piggin
2020-07-10 1:56 ` [RFC PATCH 2/7] arch: use asm-generic mmu context for no-op implementations Nicholas Piggin
2020-07-10 1:56 ` [RFC PATCH 3/7] mm: introduce exit_lazy_tlb Nicholas Piggin
2020-07-10 1:56 ` [RFC PATCH 4/7] x86: use exit_lazy_tlb rather than membarrier_mm_sync_core_before_usermode Nicholas Piggin
2020-07-10 9:42 ` Peter Zijlstra
2020-07-10 14:02 ` Mathieu Desnoyers
2020-07-10 17:04 ` Andy Lutomirski
2020-07-13 4:45 ` Nicholas Piggin
2020-07-13 13:47 ` Nicholas Piggin
2020-07-13 14:13 ` Mathieu Desnoyers
2020-07-13 15:48 ` Andy Lutomirski
2020-07-13 16:37 ` Nicholas Piggin
2020-07-16 4:15 ` Nicholas Piggin
2020-07-16 4:42 ` Nicholas Piggin
2020-07-16 15:46 ` Mathieu Desnoyers
2020-07-16 16:03 ` Mathieu Desnoyers
2020-07-16 18:58 ` Mathieu Desnoyers
2020-07-16 21:24 ` Alan Stern
2020-07-17 13:39 ` Mathieu Desnoyers
2020-07-17 14:51 ` Alan Stern
2020-07-17 15:39 ` Mathieu Desnoyers
2020-07-17 16:11 ` Alan Stern
2020-07-17 16:22 ` Mathieu Desnoyers
2020-07-17 17:44 ` Alan Stern
2020-07-17 17:52 ` Mathieu Desnoyers
2020-07-17 0:00 ` Nicholas Piggin
2020-07-16 5:18 ` Andy Lutomirski
2020-07-16 6:06 ` Nicholas Piggin
2020-07-16 8:50 ` Peter Zijlstra
2020-07-16 10:03 ` Nicholas Piggin
2020-07-16 11:00 ` peterz
2020-07-16 15:34 ` Mathieu Desnoyers
2020-07-16 23:26 ` Nicholas Piggin
2020-07-17 13:42 ` Mathieu Desnoyers
2020-07-20 3:03 ` Nicholas Piggin
2020-07-20 16:46 ` Mathieu Desnoyers
2020-07-21 10:04 ` Nicholas Piggin
2020-07-21 13:11 ` Mathieu Desnoyers
2020-07-21 14:30 ` Nicholas Piggin
2020-07-21 15:06 ` peterz
2020-07-21 15:15 ` Mathieu Desnoyers
2020-07-21 15:19 ` Peter Zijlstra
2020-07-21 15:22 ` Mathieu Desnoyers
2020-07-10 1:56 ` [RFC PATCH 5/7] lazy tlb: introduce lazy mm refcount helper functions Nicholas Piggin
2020-07-10 9:48 ` Peter Zijlstra
2020-07-10 1:56 ` [RFC PATCH 6/7] lazy tlb: allow lazy tlb mm switching to be configurable Nicholas Piggin
2020-07-10 1:56 ` [RFC PATCH 7/7] lazy tlb: shoot lazies, a non-refcounting lazy tlb option Nicholas Piggin
2020-07-10 9:35 ` Peter Zijlstra
2020-07-13 4:58 ` Nicholas Piggin
2020-07-13 15:59 ` Andy Lutomirski
2020-07-13 16:48 ` Nicholas Piggin [this message]
2020-07-13 18:18 ` Andy Lutomirski
2020-07-14 5:04 ` Nicholas Piggin
2020-07-14 6:31 ` Nicholas Piggin
2020-07-14 12:46 ` Andy Lutomirski
2020-07-14 13:23 ` Peter Zijlstra
2020-07-16 2:26 ` Nicholas Piggin
2020-07-16 2:35 ` Nicholas Piggin
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=1594658283.qabzoxga67.astroid@bobo.none \
--to=npiggin@gmail.com \
--cc=anton@ozlabs.org \
--cc=arnd@arndb.de \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=luto@kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=peterz@infradead.org \
--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).