From: "André Almeida" <andrealmeid@igalia.com>
To: Uros Bizjak <ubizjak@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
linux-kernel@vger.kernel.org, Ingo Molnar <mingo@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Darren Hart <dvhart@infradead.org>,
Davidlohr Bueso <dave@stgolabs.net>
Subject: Re: [PATCH v2] futex: Rewrite get_inode_sequence_number() to make code simpler
Date: Wed, 9 Oct 2024 16:43:58 -0300 [thread overview]
Message-ID: <bb91d63d-c61a-4063-bf14-4cbbb62bec12@igalia.com> (raw)
In-Reply-To: <20241004085257.10908-1-ubizjak@gmail.com>
Hi Uros,
Em 04/10/2024 05:52, Uros Bizjak escreveu:
> Rewrite get_inode_sequence_number() to make code simpler:
>
> a) Rewrite FOR loop to a DO-WHILE loop with returns moved
> out of the loop.
>
> b) Use atomic64_inc_return() instead of atomic64_add_return().
>
> c) Use !atomic64_try_cmpxchg_relaxed(*ptr, &old, new) instead of
> atomic64_cmpxchg_relaxed (*ptr, old, new) != old. x86 CMPXCHG
> instruction returns success in ZF flag, so this change also saves
> a compare instruction after CMPXCHG.
Remember, it's easy to see in the diff that you replace the function,
but might be not so clear why you did so. I think it would be better to
understand if you write like:
We are trying to set a value for the i_sequence, that we expect that is
zero, but if we fail to do so, we are happy to use the current non-zero
i_sequence value that we found. Instead of using
atomic64_cmpxchg_relaxed(), use atomic64_try_cmpxchg_relaxed() which
provides a better semantic for this situation.
>
> Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@kernel.org>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Darren Hart <dvhart@infradead.org>
> Cc: Davidlohr Bueso <dave@stgolabs.net>
> Cc: "André Almeida" <andrealmeid@igalia.com>
> ---
> v2: Explicitly initialize "old" to zero before the call to
> atomic64_try_cmpxchg_relaxed(). Rewrite commit message to
> state the motivation for the patch.
> ---
> kernel/futex/core.c | 18 ++++++++----------
> 1 file changed, 8 insertions(+), 10 deletions(-)
>
> diff --git a/kernel/futex/core.c b/kernel/futex/core.c
> index 136768ae2637..ac650f7ed56c 100644
> --- a/kernel/futex/core.c
> +++ b/kernel/futex/core.c
> @@ -173,23 +173,21 @@ futex_setup_timer(ktime_t *time, struct hrtimer_sleeper *timeout,
> static u64 get_inode_sequence_number(struct inode *inode)
> {
> static atomic64_t i_seq;
> - u64 old;
> + u64 old, new;
>
> /* Does the inode already have a sequence number? */
> old = atomic64_read(&inode->i_sequence);
> if (likely(old))
> return old;
>
> - for (;;) {
> - u64 new = atomic64_add_return(1, &i_seq);
> - if (WARN_ON_ONCE(!new))
> - continue;
> + do {
> + new = atomic64_inc_return(&i_seq);
> + } while (WARN_ON_ONCE(!new));
>
> - old = atomic64_cmpxchg_relaxed(&inode->i_sequence, 0, new);
> - if (old)
> - return old;
> - return new;
> - }
> + old = 0;
Please initialize it in the variable declaration.
> + if (!atomic64_try_cmpxchg_relaxed(&inode->i_sequence, &old, new))
> + return old;
> + return new;
> }
>
> /**
next prev parent reply other threads:[~2024-10-09 19:44 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-04 8:52 Uros Bizjak
2024-10-09 19:43 ` André Almeida [this message]
2024-10-10 1:47 ` Thomas Gleixner
2024-10-10 6:20 ` Uros Bizjak
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=bb91d63d-c61a-4063-bf14-4cbbb62bec12@igalia.com \
--to=andrealmeid@igalia.com \
--cc=dave@stgolabs.net \
--cc=dvhart@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=ubizjak@gmail.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
all inboxes | Powered by JetHome®