From: Paolo Bonzini <pbonzini@redhat.com>
To: Hillf Danton <hdanton@sina.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
"Michael S. Tsirkin" <mst@redhat.com>,
linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
Al Viro <viro@zeniv.linux.org.uk>
Subject: Re: 5.13-rt1 + KVM = WARNING: at fs/eventfd.c:74 eventfd_signal()
Date: Fri, 23 Jul 2021 12:56:15 +0200 [thread overview]
Message-ID: <41e2046f-ebbb-9fb2-3b30-1d6edceaaced@redhat.com> (raw)
In-Reply-To: <20210723094830.1375-1-hdanton@sina.com>
On 23/07/21 11:48, Hillf Danton wrote:
> On Fri, 23 Jul 2021 09:59:55 +0200 Paolo Bonzini wrote:
>> First and foremost, I'm not sure what you are trying to fix.
>
> The change proposed builds the return value without assuming that the
> event count is stable across poll_wait(). If it is unstable then we know
> there are concurrent reader and/or writer who both are ingnored currently.
Concurrent reads or writes have their own wake_up_locked_poll calls.
Why are they not enough?
>> Second, the patch is wrong even without taking into account the lockless
>> accesses, because the condition for returning EPOLLOUT is certainly wrong.
>
> Given it is detected that event count was consumed, there is room, though
> as racy as it is, in the event count for writer to make some progress.
For one, you do not return EPOLLIN|EPOLLOUT correctly.
>> Third, barriers very rarely speak for themselves. In particular what
>> do they pair with?
>
> There is no need to consider pair frankly. Barriers are just readded for
> removing the seep in the comment. Then the comment goes with the seep.
Then you would need an smp_mb() to order a spin_unlock() against a
READ_ONCE(). But adding memory barriers randomly is the worst way to
write a lockless algorithm that you can explain to others, and "there is
no need to consider pair frankly" all but convinces me that you've put
enough thought in the change you're proposing.
(Shameless plug: https://lwn.net/Articles/844224/).
> What the comment does not cover is the cases of more-than-two-party race.
But, you still haven't explained what's the bug there.
Paolo
next prev parent reply other threads:[~2021-07-23 10:56 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-14 8:01 5.13-rt1 + KVM = WARNING: at fs/eventfd.c:74 eventfd_signal() Daniel Bristot de Oliveira
2021-07-14 8:10 ` Paolo Bonzini
2021-07-14 9:23 ` Jason Wang
2021-07-14 10:35 ` Paolo Bonzini
2021-07-14 10:41 ` Michael S. Tsirkin
2021-07-14 10:44 ` Paolo Bonzini
2021-07-14 12:20 ` Daniel Bristot de Oliveira
2021-07-15 4:14 ` Jason Wang
2021-07-15 5:58 ` Paolo Bonzini
2021-07-15 6:45 ` Jason Wang
2021-07-15 8:22 ` Daniel Bristot de Oliveira
2021-07-15 8:44 ` He Zhe
2021-07-15 9:51 ` Paolo Bonzini
2021-07-15 10:10 ` He Zhe
2021-07-15 11:05 ` Paolo Bonzini
2021-07-16 2:26 ` Jason Wang
2021-07-16 2:43 ` He Zhe
2021-07-16 2:46 ` Jason Wang
2021-07-15 9:46 ` Paolo Bonzini
2021-07-15 12:34 ` Daniel Bristot de Oliveira
[not found] ` <20210715102249.2205-1-hdanton@sina.com>
2021-07-15 12:31 ` Daniel Bristot de Oliveira
[not found] ` <20210716020611.2288-1-hdanton@sina.com>
2021-07-16 6:54 ` Paolo Bonzini
[not found] ` <20210716075539.2376-1-hdanton@sina.com>
2021-07-16 7:59 ` Paolo Bonzini
[not found] ` <20210716093725.2438-1-hdanton@sina.com>
2021-07-16 11:55 ` Paolo Bonzini
2021-07-18 12:42 ` Hillf Danton
2021-07-19 15:38 ` Paolo Bonzini
2021-07-21 7:04 ` Hillf Danton
2021-07-21 7:25 ` Thomas Gleixner
2021-07-21 10:11 ` Hillf Danton
2021-07-21 10:59 ` Paolo Bonzini
2021-07-22 5:58 ` Hillf Danton
2021-07-23 2:23 ` Hillf Danton
2021-07-23 7:59 ` Paolo Bonzini
2021-07-23 9:48 ` Hillf Danton
2021-07-23 10:56 ` Paolo Bonzini [this message]
2021-07-24 4:33 ` Hillf Danton
2021-07-26 11:03 ` Paolo Bonzini
2021-07-28 8:06 ` Thomas Gleixner
2021-07-28 10:21 ` Paolo Bonzini
2021-07-28 19:07 ` Thomas Gleixner
2021-07-29 11:01 ` [PATCH] eventfd: Make signal recursion protection a task bit Thomas Gleixner
2021-07-29 14:32 ` Daniel Bristot de Oliveira
2021-07-29 19:23 ` Daniel Bristot de Oliveira
2021-08-26 7:03 ` Jason Wang
2021-08-27 23:41 ` [tip: sched/core] " tip-bot2 for Thomas Gleixner
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=41e2046f-ebbb-9fb2-3b30-1d6edceaaced@redhat.com \
--to=pbonzini@redhat.com \
--cc=bigeasy@linutronix.de \
--cc=hdanton@sina.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mst@redhat.com \
--cc=tglx@linutronix.de \
--cc=viro@zeniv.linux.org.uk \
/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.