Received-SPF: pass (google.com: domain of linux-kernel+bounces-85794-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
From: Jonathan Corbet <corbet@lwn.net>
To: Zhang Qiao <zhangqiao22@huawei.com>
Cc: mingo@kernel.org, chrubis@suse.cz, zhangqiao22@huawei.com,
 linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] sched/Documentation: Add RT_RUNTIME_SHARE documentation
In-Reply-To: <20240131123238.2332064-1-zhangqiao22@huawei.com>
References: <20240131123238.2332064-1-zhangqiao22@huawei.com>
Date: Wed, 28 Feb 2024 15:44:14 -0700
Message-ID: <875xy8p5tt.fsf@meer.lwn.net>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain

Zhang Qiao <zhangqiao22@huawei.com> writes:

> RT_RUNTIME_SHARE is an important strategy for rt bandwidth, and
> we should document this sched feature.
>
> Signed-off-by: Zhang Qiao <zhangqiao22@huawei.com>
> ---
>  Documentation/scheduler/sched-rt-group.rst | 11 +++++++++++
>  1 file changed, 11 insertions(+)
>
> diff --git a/Documentation/scheduler/sched-rt-group.rst b/Documentation/scheduler/sched-rt-group.rst
> index d685609ed3d7..4d8eceb71f5e 100644
> --- a/Documentation/scheduler/sched-rt-group.rst
> +++ b/Documentation/scheduler/sched-rt-group.rst
> @@ -12,6 +12,7 @@ Real-Time group scheduling
>       2.1 System-wide settings
>       2.2 Default behaviour
>       2.3 Basis for grouping tasks
> +     2.4 RT_RUNTIME_SHARE sched feature
>     3. Future plans
>  
>  
> @@ -146,6 +147,16 @@ For now, this can be simplified to just the following (but see Future plans):
>  
>     \Sum_{i} runtime_{i} <= global_runtime
>  
> +2.4 RT_RUNTIME_SHARE sched feature
> +----------------------------
> +
> +RT_RUNTIME_SHARE allows a cpu borrows rt-runtime from other cpus if it runs
> +out of its own rt-runtime.
> +
> +With this feature enabled, a rt-task probably hits 100% cpu usage and starves
> +per-cpu tasks like kworkers, as a result, it may hang up the whole system.
> +Therefore, in order to avoid such exception, recommand to disable this feature
> +by default unless you really know what you're up to.

So this doesn't appear to have been picked up by anybody...should I
carry it in docs, or is there some other reason why it hasn't gone in?

Thanks,

jon