Message-ID: <201d597e-e0c7-411b-a29c-dfc3111aa863@infradead.org>
Date: Fri, 12 Jul 2024 21:22:13 -0700
From: Randy Dunlap <rdunlap@...radead.org>
To: Carlos Bilbao <carlos.bilbao.osdev@...il.com>, corbet@....net,
 sergio.collado@...il.com, peterz@...radead.org
Cc: bilbao@...edu, jembid@....es, linux-kernel@...r.kernel.org,
 linux-doc@...r.kernel.org
Subject: Re: [PATCH 1/1] docs: scheduler: Start documenting the EEVDF
 scheduler

Hi,

On 7/12/24 5:32 PM, Carlos Bilbao wrote:
> Add some documentation regarding the newly introduced scheduler EEVDF.
> 
> Signed-off-by: Carlos Bilbao <carlos.bilbao.osdev@...il.com>
> ---
>  Documentation/scheduler/index.rst            |  1 +
>  Documentation/scheduler/sched-design-CFS.rst | 10 +++--
>  Documentation/scheduler/sched-eevdf.rst      | 44 ++++++++++++++++++++
>  3 files changed, 51 insertions(+), 4 deletions(-)
>  create mode 100644 Documentation/scheduler/sched-eevdf.rst
> 
> diff --git a/Documentation/scheduler/index.rst b/Documentation/scheduler/index.rst
> index 43bd8a145b7a..444a6fef1464 100644
> --- a/Documentation/scheduler/index.rst
> +++ b/Documentation/scheduler/index.rst
> @@ -11,6 +11,7 @@ Scheduler
>      sched-arch
>      sched-bwc
>      sched-deadline
> +    sched-eevdf

I would have probably put EEVDF just after CFS instead of before it...
whatever.

>      sched-design-CFS
>      sched-domains
>      sched-capacity
> diff --git a/Documentation/scheduler/sched-design-CFS.rst b/Documentation/scheduler/sched-design-CFS.rst
> index bc1e507269c6..b703c6dcb3cd 100644
> --- a/Documentation/scheduler/sched-design-CFS.rst
> +++ b/Documentation/scheduler/sched-design-CFS.rst
> @@ -8,10 +8,12 @@ CFS Scheduler
>  1.  OVERVIEW
>  ============
>  
> -CFS stands for "Completely Fair Scheduler," and is the new "desktop" process
> -scheduler implemented by Ingo Molnar and merged in Linux 2.6.23.  It is the
> -replacement for the previous vanilla scheduler's SCHED_OTHER interactivity
> -code.
> +CFS stands for "Completely Fair Scheduler," and is the "desktop" process
> +scheduler implemented by Ingo Molnar and merged in Linux 2.6.23. When
> +originally merged, it was the replacement for the previous vanilla
> +scheduler's SCHED_OTHER interactivity code. Nowadays, CFS is making room
> +for EEVDF, for which documentation can be found in
> +:ref:`sched_design_EEVDF`.
>  
>  80% of CFS's design can be summed up in a single sentence: CFS basically models
>  an "ideal, precise multi-tasking CPU" on real hardware.
> diff --git a/Documentation/scheduler/sched-eevdf.rst b/Documentation/scheduler/sched-eevdf.rst
> new file mode 100644
> index 000000000000..31ad8f995360
> --- /dev/null
> +++ b/Documentation/scheduler/sched-eevdf.rst
> @@ -0,0 +1,44 @@
> +.. _sched_design_EEVDF:
> +
> +===============
> +EEVDF Scheduler
> +===============
> +
> +The "Earliest Eligible Virtual Deadline First" (EEVDF) was first introduced
> +in a scientific publication in 1995 [1]. The Linux kernel began
> +transitioning to EEVDF in version 6.6 (as a new option in 2024), moving
> +away from the earlier Completely Fair Scheduler (CFS) in favor of a version
> +of EEVDF proposed by Peter Zijlstra in 2023 [2-4]. More information
> +regarding CFS can be found in :ref:`sched_design_CFS`.
> +
> +Similarly to CFS, EEVDF aims to distribute CPU time equally among all
> +runnable tasks with the same priority. To do so, it assigns a virtual run
> +time to each task, creating a "lag" value that can be used to determine
> +whether a task has received its fair share of CPU time. In this way, a task
> +with a positive lag is owed CPU time, while a negative lag means the task
> +has exceeded its portion. EEVDF picks tasks with lag greater or equal to
> +zero and calculates a virtual deadline (VD) for each, selecting the task
> +with the earliest VD to execute next. It's important to note that this
> +allows latency-sensitive tasks with shorter time slices to be prioritized,
> +which helps with their responsiveness.
> +
> +There are ongoing discussions on how to manage lag, especially for sleeping
> +tasks; but at the time of writing EEVDF uses a "decaying" mechanism based
> +on virtual run time (VRT). This prevents tasks from exploiting the system
> +by sleeping briefly to reset their negative lag: when a task sleeps, it
> +remains on the run queue but marked for "deferred dequeue," allowing its
> +lag to decay over VRT. Hence, long-sleeping tasks eventually have their lag
> +reset. Finally, tasks can preempt others if their VD is earlier, and tasks
> +can request specific time slices using the new sched_setattr() system call,
> +which further facilitates the job of latency-sensitive applications.
> +
> +4. REFERENCES
> +=============

Why is this section numbered 4?
No other sections here are numbered.

> +
> +[1] https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=805acf7726282721504c8f00575d91ebfd750564
> +
> +[2] https://lore.kernel.org/lkml/a79014e6-ea83-b316-1e12-2ae056bda6fa@linux.vnet.ibm.com/
> +
> +[3] https://lwn.net/Articles/969062/
> +
> +[4] https://lwn.net/Articles/925371/

Other than those 2 comments:

Reviewed-by: Randy Dunlap <rdunlap@...radead.org>
Tested-by: Randy Dunlap <rdunlap@...radead.org>


Thanks.

-- 
~Randy

Powered by blists - more mailing lists