[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <e20c3d90-db24-7722-3f89-adadb83a9bf7@windriver.com>
Date: Thu, 15 Jul 2021 18:10:47 +0800
From: He Zhe <zhe.he@...driver.com>
To: Paolo Bonzini <pbonzini@...hat.com>,
Daniel Bristot de Oliveira <bristot@...hat.com>,
Jason Wang <jasowang@...hat.com>,
Thomas Gleixner <tglx@...utronix.de>,
Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
"Michael S. Tsirkin" <mst@...hat.com>,
Juri Lelli <jlelli@...hat.com>
Cc: LKML <linux-kernel@...r.kernel.org>,
Al Viro <viro@...iv.linux.org.uk>
Subject: Re: 5.13-rt1 + KVM = WARNING: at fs/eventfd.c:74 eventfd_signal()
On 7/15/21 5:51 PM, Paolo Bonzini wrote:
> On 15/07/21 10:44, He Zhe wrote:
>> It was to increase the recursion depth to 2 so that vhost_worker and
>> kvm_vcpu_ioctl syscall could work in parallel
>
> The count is per-CPU, so parallel operations cannot cause it to become 2. Your patch might fix calls from ioeventfd to vhost_worker to another eventfd, but not *parallel* operation of KVM and vhost (except on PREEMPT_RT).
>
> You should identify the exact callstack that caused the warning for vDUSE, and document that one in the commit message, so that reviewers can understand the issue.
The following was provided in this thread. The commit log contains the call traces that I met and fixed back to Apr. 2020.
https://lore.kernel.org/lkml/20210618084412.18257-1-zhe.he@windriver.com/
And the problem has been reported many times until last month. So when this patch was pointed at in this thread, I thought it was still the same case.
https://lore.kernel.org/lkml/beac2025-2e11-8ed0-61e2-9f6e633482e8@redhat.com/
https://lore.kernel.org/lkml/20210703043039-mutt-send-email-mst@kernel.org/
Zhe
>
> Paolo
>
Powered by blists - more mailing lists