From: Joel Fernandes <joelagnelf@nvidia.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Lyude Paul <lyude@redhat.com>,
rust-for-linux@vger.kernel.org,
Thomas Gleixner <tglx@linutronix.de>,
Boqun Feng <boqun.feng@gmail.com>,
linux-kernel@vger.kernel.org,
Daniel Almeida <daniel.almeida@collabora.com>,
Danilo Krummrich <dakr@kernel.org>,
Lorenzo Stoakes <lorenzo.stoakes@oracle.com>,
Vlastimil Babka <vbabka@suse.cz>,
"Liam R. Howlett" <Liam.Howlett@oracle.com>,
Uladzislau Rezki <urezki@gmail.com>,
Miguel Ojeda <ojeda@kernel.org>,
Alex Gaynor <alex.gaynor@gmail.com>, Gary Guo <gary@garyguo.net>,
Bj??rn Roy Baron <bjorn3_gh@protonmail.com>,
Benno Lossin <lossin@kernel.org>,
Andreas Hindborg <a.hindborg@kernel.org>,
Alice Ryhl <aliceryhl@google.com>,
Trevor Gross <tmgross@umich.edu>,
"Rafael J. Wysocki" <rafael@kernel.org>,
Viresh Kumar <viresh.kumar@linaro.org>,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
Ingo Molnar <mingo@kernel.org>,
Ryo Takakura <ryotkkr98@gmail.com>,
K Prateek Nayak <kprateek.nayak@amd.com>,
"open list:CPU FREQUENCY SCALING FRAMEWORK"
<linux-pm@vger.kernel.org>
Subject: Re: [PATCH v13 01/17] preempt: Track NMI nesting to separate per-CPU counter
Date: Mon, 20 Oct 2025 16:44:21 -0400 [thread overview]
Message-ID: <20251020204421.GA197647@joelbox2> (raw)
In-Reply-To: <20251014194349.GC1206438@noisy.programming.kicks-ass.net>
On Tue, Oct 14, 2025 at 09:43:49PM +0200, Peter Zijlstra wrote:
> On Tue, Oct 14, 2025 at 01:55:47PM -0400, Joel Fernandes wrote:
> >
> >
> > On 10/14/2025 6:48 AM, Peter Zijlstra wrote:
> > > On Mon, Oct 13, 2025 at 11:48:03AM -0400, Lyude Paul wrote:
> > >
> > >> #define __nmi_enter() \
> > >> do { \
> > >> lockdep_off(); \
> > >> arch_nmi_enter(); \
> > >> - BUG_ON(in_nmi() == NMI_MASK); \
> > >> - __preempt_count_add(NMI_OFFSET + HARDIRQ_OFFSET); \
> > >> + BUG_ON(__this_cpu_read(nmi_nesting) == UINT_MAX); \
> > >> + __this_cpu_inc(nmi_nesting); \
> > >
> > > An NMI that nests from here..
> > >
> > >> + __preempt_count_add(HARDIRQ_OFFSET); \
> > >> + if (__this_cpu_read(nmi_nesting) == 1) \
> > >
> > > .. until here, will see nmi_nesting > 1 and not set NMI_OFFSET.
> >
> > This is true, I can cure it by setting NMI_OFFSET unconditionally when
> > nmi_nesting >= 1. Then the outer most NMI will then reset it. I think that will
> > work. Do you see any other issue with doing so?
>
> unconditionally set NMI_FFSET, regardless of nmi_nesting
> and only clear on exit when nmi_nesting == 0.
>
> Notably, when you use u64 __preempt_count, you can limit this to 32bit
> only. The NMI nesting can happen in the single instruction window
> between ADD and ADC. But on 64bit you don't have that gap and so don't
> need to fix it.
Wouldn't this break __preempt_count_dec_and_test though? If we make it
64-bit, then there is no longer a way on x86 32-bit to decrement the preempt
count and zero-test the entire word in the same instruction (decl). And I
feel there might be other races as well. Also this means that every
preempt_disable/enable will be heavier on 32-bit.
If we take the approach of this patch, but move the per-cpu counter to cache
hot area, what are the other drawbacks other than few more instructions on
NMI entry/exit? It feels simpler and less risky. But let me know if I missed
something.
thanks,
- Joel
next prev parent reply other threads:[~2025-10-20 20:44 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-13 15:48 [PATCH v13 00/17] Refcounted interrupts, SpinLockIrq for rust Lyude Paul
2025-10-13 15:48 ` [PATCH v13 01/17] preempt: Track NMI nesting to separate per-CPU counter Lyude Paul
2025-10-13 16:19 ` Lyude Paul
2025-10-13 16:32 ` Miguel Ojeda
2025-10-13 20:00 ` Peter Zijlstra
2025-10-13 21:27 ` Joel Fernandes
2025-10-14 8:25 ` Peter Zijlstra
2025-10-14 17:59 ` Joel Fernandes
2025-10-14 19:37 ` Peter Zijlstra
2025-10-14 10:48 ` Peter Zijlstra
2025-10-14 17:55 ` Joel Fernandes
2025-10-14 19:43 ` Peter Zijlstra
2025-10-14 22:05 ` Joel Fernandes
2025-10-20 20:44 ` Joel Fernandes [this message]
2025-10-30 22:56 ` Joel Fernandes
2025-10-13 15:48 ` [PATCH v13 02/17] preempt: Reduce NMI_MASK to single bit and restore HARDIRQ_BITS Lyude Paul
2025-10-13 15:48 ` [PATCH v13 03/17] preempt: Introduce HARDIRQ_DISABLE_BITS Lyude Paul
2025-10-31 14:59 ` Andreas Hindborg
2025-10-31 19:59 ` Boqun Feng
2025-10-13 15:48 ` [PATCH v13 04/17] preempt: Introduce __preempt_count_{sub, add}_return() Lyude Paul
2025-10-13 15:48 ` [PATCH v13 05/17] irq & spin_lock: Add counted interrupt disabling/enabling Lyude Paul
2025-10-15 20:54 ` Bart Van Assche
2025-10-16 8:15 ` Peter Zijlstra
2025-10-17 6:44 ` Boqun Feng
2025-10-16 21:24 ` David Laight
2025-10-17 6:48 ` Boqun Feng
2025-10-13 15:48 ` [PATCH v13 06/17] irq: Add KUnit test for refcounted interrupt enable/disable Lyude Paul
2025-10-13 15:48 ` [PATCH v13 07/17] rust: Introduce interrupt module Lyude Paul
2025-10-13 15:48 ` [PATCH v13 08/17] rust: helper: Add spin_{un,}lock_irq_{enable,disable}() helpers Lyude Paul
2025-10-13 15:48 ` [PATCH v13 09/17] rust: sync: Add SpinLockIrq Lyude Paul
2025-10-13 15:48 ` [PATCH v13 10/17] rust: sync: Introduce lock::Backend::Context Lyude Paul
2025-10-13 15:48 ` [PATCH v13 11/17] rust: sync: lock: Add `Backend::BackendInContext` Lyude Paul
2025-10-13 15:48 ` [PATCH v13 12/17] rust: sync: lock/global: Rename B to G in trait bounds Lyude Paul
2025-10-13 15:48 ` [PATCH v13 13/17] rust: sync: Add a lifetime parameter to lock::global::GlobalGuard Lyude Paul
2025-10-13 15:48 ` [PATCH v13 14/17] rust: sync: Expose lock::Backend Lyude Paul
2025-10-13 15:48 ` [PATCH v13 15/17] rust: sync: lock/global: Add Backend parameter to GlobalGuard Lyude Paul
2025-10-13 15:48 ` [PATCH v13 16/17] rust: sync: lock/global: Add BackendInContext support to GlobalLock Lyude Paul
2025-10-13 15:48 ` [PATCH v13 17/17] locking: Switch to _irq_{disable,enable}() variants in cleanup guards Lyude Paul
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=20251020204421.GA197647@joelbox2 \
--to=joelagnelf@nvidia.com \
--cc=Liam.Howlett@oracle.com \
--cc=a.hindborg@kernel.org \
--cc=alex.gaynor@gmail.com \
--cc=aliceryhl@google.com \
--cc=bigeasy@linutronix.de \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun.feng@gmail.com \
--cc=dakr@kernel.org \
--cc=daniel.almeida@collabora.com \
--cc=gary@garyguo.net \
--cc=kprateek.nayak@amd.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=lorenzo.stoakes@oracle.com \
--cc=lossin@kernel.org \
--cc=lyude@redhat.com \
--cc=mingo@kernel.org \
--cc=ojeda@kernel.org \
--cc=peterz@infradead.org \
--cc=rafael@kernel.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=ryotkkr98@gmail.com \
--cc=tglx@linutronix.de \
--cc=tmgross@umich.edu \
--cc=urezki@gmail.com \
--cc=vbabka@suse.cz \
--cc=viresh.kumar@linaro.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).