lkmm.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: "Paul E. McKenney" <paulmck@kernel.org>
To: Jonas Oberhauser <jonas.oberhauser@huaweicloud.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
	Boqun Feng <boqun.feng@gmail.com>,
	lkmm@lists.linux.dev
Subject: Re: [RFC PATCH 1/1] hpref: Hazard Pointers with Reference Counter
Date: Tue, 1 Oct 2024 08:24:01 -0700	[thread overview]
Message-ID: <c0c83773-2c9b-4cfa-a28e-dec410f1b35a@paulmck-laptop> (raw)
In-Reply-To: <13bc8b46-d391-4f58-b021-481e57d2ac1f@huaweicloud.com>

On Mon, Sep 30, 2024 at 03:31:02PM +0200, Jonas Oberhauser wrote:
> [dropping a lot of CC]
> 
> Am 9/21/2024 um 6:42 PM schrieb Mathieu Desnoyers:
> > 
> > Porting it to the Linux kernel should be straightforward.
> 
> 
> Btw, what is the relation between uatomics and LKMM?
> 
> Should I assume that they intend to have LKMM semantics modulo whatever is
> explained here:
> 
> https://github.com/urcu/userspace-rcu/blob/master/doc/uatomic-api.md
> 
> Because on C11 model, dependencies are generally not considered

I must defer to Mathieu for recent uatomic directions, but the original
idea was to follow LKMM.

You are right that C11 compilers and memory model don't do much with
dependencies, one exception being the "don't create data races" rule
which requires that non-atomic accesses protected by "if" statements
stay protected, at least in the absence of a nearby unprotected access
to that same variable.

So if uatomic is to rely on dependencies, it must use the same tricks
as LKMM does, including code-style restrictions.

							Thanx, Paul

  reply	other threads:[~2024-10-01 15:24 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-21 16:42 [RFC PATCH 1/1] hpref: Hazard Pointers with Reference Counter Mathieu Desnoyers
2024-09-21 21:07 ` Lai Jiangshan
2024-09-22  7:47   ` Mathieu Desnoyers
2024-09-25  5:57 ` Jonas Oberhauser
2024-09-25  6:35   ` Mathieu Desnoyers
2024-09-25 10:06     ` Jonas Oberhauser
2024-09-25 11:36       ` Mathieu Desnoyers
2024-09-25 12:02         ` Jonas Oberhauser
2024-09-28 11:22 ` Jonas Oberhauser
2024-09-28 11:33   ` Mathieu Desnoyers
2024-09-28 11:50     ` Jonas Oberhauser
2024-09-30 13:31 ` Jonas Oberhauser
2024-10-01 15:24   ` Paul E. McKenney [this message]
2026-01-06 14:52   ` Mathieu Desnoyers
2026-01-07  9:38     ` Jonas Oberhauser
2026-01-07 14:40       ` Mathieu Desnoyers
2026-01-07 17:48         ` Paul E. McKenney
2026-01-07 20:04           ` Mathieu Desnoyers

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=c0c83773-2c9b-4cfa-a28e-dec410f1b35a@paulmck-laptop \
    --to=paulmck@kernel.org \
    --cc=boqun.feng@gmail.com \
    --cc=jonas.oberhauser@huaweicloud.com \
    --cc=lkmm@lists.linux.dev \
    --cc=mathieu.desnoyers@efficios.com \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).