[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1e254a35-d0c2-8d41-f020-21694945911a@joelfernandes.org>
Date: Tue, 25 Jul 2023 13:41:32 -0400
From: Joel Fernandes <joel@...lfernandes.org>
To: Valentin Schneider <vschneid@...hat.com>
Cc: linux-kernel@...r.kernel.org, linux-trace-kernel@...r.kernel.org,
linux-doc@...r.kernel.org, kvm@...r.kernel.org, linux-mm@...ck.org,
bpf@...r.kernel.org, x86@...nel.org, rcu@...r.kernel.org,
linux-kselftest@...r.kernel.org,
Peter Zijlstra <peterz@...radead.org>,
Nicolas Saenz Julienne <nsaenzju@...hat.com>,
Steven Rostedt <rostedt@...dmis.org>,
Masami Hiramatsu <mhiramat@...nel.org>,
Jonathan Corbet <corbet@....net>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>,
"H. Peter Anvin" <hpa@...or.com>,
Paolo Bonzini <pbonzini@...hat.com>,
Wanpeng Li <wanpengli@...cent.com>,
Vitaly Kuznetsov <vkuznets@...hat.com>,
Andy Lutomirski <luto@...nel.org>,
Frederic Weisbecker <frederic@...nel.org>,
"Paul E. McKenney" <paulmck@...nel.org>,
Neeraj Upadhyay <quic_neeraju@...cinc.com>,
Josh Triplett <josh@...htriplett.org>,
Boqun Feng <boqun.feng@...il.com>,
Mathieu Desnoyers <mathieu.desnoyers@...icios.com>,
Lai Jiangshan <jiangshanlai@...il.com>,
Zqiang <qiang.zhang1211@...il.com>,
Andrew Morton <akpm@...ux-foundation.org>,
Uladzislau Rezki <urezki@...il.com>,
Christoph Hellwig <hch@...radead.org>,
Lorenzo Stoakes <lstoakes@...il.com>,
Josh Poimboeuf <jpoimboe@...nel.org>,
Jason Baron <jbaron@...mai.com>,
Kees Cook <keescook@...omium.org>,
Sami Tolvanen <samitolvanen@...gle.com>,
Ard Biesheuvel <ardb@...nel.org>,
Nicholas Piggin <npiggin@...il.com>,
Juerg Haefliger <juerg.haefliger@...onical.com>,
Nicolas Saenz Julienne <nsaenz@...nel.org>,
"Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
Nadav Amit <namit@...are.com>,
Dan Carpenter <error27@...il.com>,
Chuang Wang <nashuiliang@...il.com>,
Yang Jihong <yangjihong1@...wei.com>,
Petr Mladek <pmladek@...e.com>,
"Jason A. Donenfeld" <Jason@...c4.com>, Song Liu <song@...nel.org>,
Julian Pidancet <julian.pidancet@...cle.com>,
Tom Lendacky <thomas.lendacky@....com>,
Dionna Glaze <dionnaglaze@...gle.com>,
Thomas Weißschuh <linux@...ssschuh.net>,
Juri Lelli <juri.lelli@...hat.com>,
Daniel Bristot de Oliveira <bristot@...hat.com>,
Marcelo Tosatti <mtosatti@...hat.com>,
Yair Podemsky <ypodemsk@...hat.com>
Subject: Re: [RFC PATCH v2 18/20] context_tracking,x86: Defer kernel text
patching IPIs
On 7/25/23 09:36, Valentin Schneider wrote:
> On 25/07/23 06:49, Joel Fernandes wrote:
>> Interesting series Valentin. Some high-level question/comments on this one:
>>
>>> On Jul 20, 2023, at 12:34 PM, Valentin Schneider <vschneid@...hat.com> wrote:
>>>
>>> text_poke_bp_batch() sends IPIs to all online CPUs to synchronize
>>> them vs the newly patched instruction. CPUs that are executing in userspace
>>> do not need this synchronization to happen immediately, and this is
>>> actually harmful interference for NOHZ_FULL CPUs.
>>
>> Does the amount of harm not correspond to practical frequency of text_poke?
>> How often does instruction patching really happen? If it is very infrequent
>> then I am not sure if it is that harmful.
>>
>
> Being pushed over a latency threshold *once* is enough to impact the
> latency evaluation of your given system/application.
>
> It's mainly about shielding the isolated, NOHZ_FULL CPUs from whatever the
> housekeeping CPUs may be up to (flipping static keys, loading kprobes,
> using ftrace...) - frequency of the interference isn't such a big part of
> the reasoning.
Makes sense.
>>> As the synchronization IPIs are sent using a blocking call, returning from
>>> text_poke_bp_batch() implies all CPUs will observe the patched
>>> instruction(s), and this should be preserved even if the IPI is deferred.
>>> In other words, to safely defer this synchronization, any kernel
>>> instruction leading to the execution of the deferred instruction
>>> sync (ct_work_flush()) must *not* be mutable (patchable) at runtime.
>>
>> If it is not infrequent, then are you handling the case where userland
>> spends multiple seconds before entering the kernel, and all this while
>> the blocking call waits? Perhaps in such situation you want the real IPI
>> to be sent out instead of the deferred one?
>>
>
> The blocking call only waits for CPUs for which it queued a CSD. Deferred
> calls do not queue a CSD thus do not impact the waiting at all. See
> smp_call_function_many_cond().
Ah I see you are using on_each_cpu_cond(). I should have gone through
the other patch before making noise.
thanks,
- Joel
Powered by blists - more mailing lists