From: Nadav Amit <nadav.amit@gmail.com>
To: Linus Torvalds <torvalds@linux-foundation.org>,
Andy Lutomirski <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Linux-MM <linux-mm@kvack.org>,
Nicholas Piggin <npiggin@gmail.com>,
Anton Blanchard <anton@ozlabs.org>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Paul Mackerras <paulus@ozlabs.org>,
Randy Dunlap <rdunlap@infradead.org>,
linux-arch <linux-arch@vger.kernel.org>, X86 ML <x86@kernel.org>,
Rik van Riel <riel@surriel.com>,
Dave Hansen <dave.hansen@intel.com>,
Peter Zijlstra <peterz@infradead.org>,
Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Subject: Re: [PATCH 16/23] sched: Use lightweight hazard pointers to grab lazy mms
Date: Sun, 9 Jan 2022 00:49:06 -0800 [thread overview]
Message-ID: <B728EEE2-1EB9-4ACD-9F4E-423276380C0C@gmail.com> (raw)
In-Reply-To: <CAHk-=wiEZO-uyFELSZgYmxF7eOLHrvh-kMWY5qTKjckOdNQxpA@mail.gmail.com>
> On Jan 8, 2022, at 10:48 PM, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> On Sat, Jan 8, 2022 at 9:56 PM Nadav Amit <nadav.amit@gmail.com> wrote:
>>
>> Just wondering: In a world of ASID/PCID - does the “lazy TLB” really
>> have a worthy advantage?
>>
>> Considering the fact that with PTI anyhow address spaces are switched
>> all the time, can’t we just get rid of it?
>
[snip]
>
> Or maybe it's only worth it on platforms where it's free (UP, possibly
> other situations - like if you have IPI and it's "free").
On UP it might be free, but on x86+IPIs there is a tradeoff.
When you decide which CPUs you want to send the IPI to, in the
common flow (no tables freed) you check whether they use
“lazy TLB” or not in order to filter out the lazy ones. In the
past this was on a cacheline with other frequently-dirtied data so
many times the cacheline bounced from cache to cache. Worse, the
test used an indirect branch so was expensive with Spectre v2
mitigations. I fixed it some time ago, so things are better and
today the cacheline of is_lazy should bounce less between caches,
but there is a tradeoff in maintaining and checking both cpumask
and then is_lazy for each CPU in cpumask.
It is possible for instance to get rid of is_lazy, keep the CPU
on mm_cpumask when switching to kernel thread, and then if/when
an IPI is received remove it from cpumask to avoid further
unnecessary TLB shootdown IPIs.
I do not know whether it is a pure win, but there is a tradeoff.
next prev parent reply other threads:[~2022-01-09 8:49 UTC|newest]
Thread overview: 79+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-08 16:43 [PATCH 00/23] mm, sched: Rework lazy mm handling Andy Lutomirski
2022-01-08 16:43 ` [PATCH 01/23] membarrier: Document why membarrier() works Andy Lutomirski
2022-01-12 15:30 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 02/23] x86/mm: Handle unlazying membarrier core sync in the arch code Andy Lutomirski
2022-01-12 15:40 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 03/23] membarrier: Remove membarrier_arch_switch_mm() prototype in core code Andy Lutomirski
2022-01-08 16:43 ` [PATCH 04/23] membarrier: Make the post-switch-mm barrier explicit Andy Lutomirski
2022-01-12 15:52 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 05/23] membarrier, kthread: Use _ONCE accessors for task->mm Andy Lutomirski
2022-01-12 15:55 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 06/23] powerpc/membarrier: Remove special barrier on mm switch Andy Lutomirski
2022-01-08 16:43 ` Andy Lutomirski
2022-01-10 8:42 ` Christophe Leroy
2022-01-10 8:42 ` Christophe Leroy
2022-01-12 15:57 ` Mathieu Desnoyers
2022-01-12 15:57 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 07/23] membarrier: Rewrite sync_core_before_usermode() and improve documentation Andy Lutomirski
2022-01-08 16:43 ` Andy Lutomirski
2022-01-08 16:43 ` Andy Lutomirski
2022-01-12 16:11 ` Mathieu Desnoyers
2022-01-12 16:11 ` Mathieu Desnoyers
2022-01-12 16:11 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 08/23] membarrier: Remove redundant clear of mm->membarrier_state in exec_mmap() Andy Lutomirski
2022-01-12 16:13 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 09/23] membarrier: Fix incorrect barrier positions during exec and kthread_use_mm() Andy Lutomirski
2022-01-12 16:30 ` Mathieu Desnoyers
2022-01-12 17:08 ` Mathieu Desnoyers
2022-01-08 16:43 ` [PATCH 10/23] x86/events, x86/insn-eval: Remove incorrect active_mm references Andy Lutomirski
2022-01-08 16:43 ` [PATCH 11/23] sched/scs: Initialize shadow stack on idle thread bringup, not shutdown Andy Lutomirski
2022-01-10 22:06 ` Sami Tolvanen
2022-01-08 16:43 ` [PATCH 12/23] Rework "sched/core: Fix illegal RCU from offline CPUs" Andy Lutomirski
2022-01-08 16:43 ` [PATCH 13/23] exec: Remove unnecessary vmacache_seqnum clear in exec_mmap() Andy Lutomirski
2022-01-08 16:43 ` [PATCH 14/23] sched, exec: Factor current mm changes out from exec Andy Lutomirski
2022-01-08 16:44 ` [PATCH 15/23] kthread: Switch to __change_current_mm() Andy Lutomirski
2022-01-08 16:44 ` [PATCH 16/23] sched: Use lightweight hazard pointers to grab lazy mms Andy Lutomirski
2022-01-08 19:22 ` Linus Torvalds
2022-01-08 22:04 ` Andy Lutomirski
2022-01-09 0:27 ` Linus Torvalds
2022-01-09 0:53 ` Linus Torvalds
2022-01-09 3:58 ` Andy Lutomirski
2022-01-09 4:38 ` Linus Torvalds
2022-01-09 20:19 ` Andy Lutomirski
2022-01-09 20:48 ` Linus Torvalds
2022-01-09 21:51 ` Linus Torvalds
2022-01-10 0:52 ` Andy Lutomirski
2022-01-10 2:36 ` Rik van Riel
2022-01-10 3:51 ` Linus Torvalds
2022-01-10 4:56 ` Nicholas Piggin
2022-01-10 5:17 ` Nicholas Piggin
2022-01-10 17:19 ` Linus Torvalds
2022-01-11 2:24 ` Nicholas Piggin
2022-01-10 20:52 ` Andy Lutomirski
2022-01-11 3:10 ` Nicholas Piggin
2022-01-11 15:39 ` Andy Lutomirski
2022-01-11 22:48 ` Nicholas Piggin
2022-01-12 0:42 ` Nicholas Piggin
2022-01-11 10:39 ` Will Deacon
2022-01-11 15:22 ` Andy Lutomirski
2022-01-09 5:56 ` Nadav Amit
2022-01-09 6:48 ` Linus Torvalds
2022-01-09 8:49 ` Nadav Amit [this message]
2022-01-09 19:10 ` Linus Torvalds
2022-01-09 19:52 ` Andy Lutomirski
2022-01-09 20:00 ` Linus Torvalds
2022-01-09 20:34 ` Nadav Amit
2022-01-09 20:48 ` Andy Lutomirski
2022-01-09 19:22 ` Rik van Riel
2022-01-09 19:34 ` Nadav Amit
2022-01-09 19:37 ` Rik van Riel
2022-01-09 19:51 ` Nadav Amit
2022-01-09 19:54 ` Linus Torvalds
2022-01-08 16:44 ` [PATCH 17/23] x86/mm: Make use/unuse_temporary_mm() non-static Andy Lutomirski
2022-01-08 16:44 ` [PATCH 18/23] x86/mm: Allow temporary mms when IRQs are on Andy Lutomirski
2022-01-08 16:44 ` [PATCH 19/23] x86/efi: Make efi_enter/leave_mm use the temporary_mm machinery Andy Lutomirski
2022-01-10 13:13 ` Ard Biesheuvel
2022-01-08 16:44 ` [PATCH 20/23] x86/mm: Remove leave_mm() in favor of unlazy_mm_irqs_off() Andy Lutomirski
2022-01-08 16:44 ` [PATCH 21/23] x86/mm: Use unlazy_mm_irqs_off() in TLB flush IPIs Andy Lutomirski
2022-01-08 16:44 ` [PATCH 22/23] x86/mm: Optimize for_each_possible_lazymm_cpu() Andy Lutomirski
2022-01-08 16:44 ` [PATCH 23/23] x86/mm: Opt in to IRQs-off activate_mm() Andy Lutomirski
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=B728EEE2-1EB9-4ACD-9F4E-423276380C0C@gmail.com \
--to=nadav.amit@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=anton@ozlabs.org \
--cc=benh@kernel.crashing.org \
--cc=dave.hansen@intel.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=luto@kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=npiggin@gmail.com \
--cc=paulus@ozlabs.org \
--cc=peterz@infradead.org \
--cc=rdunlap@infradead.org \
--cc=riel@surriel.com \
--cc=torvalds@linux-foundation.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.