From: Peter Zijlstra <peterz@infradead.org>
To: Frederic Weisbecker <frederic@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>, LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/2] sched: Add default dynamic preempt mode Kconfig
Date: Tue, 8 Jun 2021 16:04:47 +0200 [thread overview]
Message-ID: <YL94/8cSr2R2wiTa@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <YL92vbcw3ozRLycU@hirez.programming.kicks-ass.net>
On Tue, Jun 08, 2021 at 03:55:09PM +0200, Peter Zijlstra wrote:
> Urgh, would something like this work?
>
> ---
> kernel/Kconfig.preempt | 64 ++++++++++++++++++++++++++++++++++++++++----------
> 1 file changed, 51 insertions(+), 13 deletions(-)
>
> diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt
> index bd7c4147b9a8..43c68a806e4e 100644
> --- a/kernel/Kconfig.preempt
> +++ b/kernel/Kconfig.preempt
> @@ -1,11 +1,25 @@
> # SPDX-License-Identifier: GPL-2.0-only
>
> +config PREEMPT_COUNT
> + bool
> +
> +config PREEMPTION
> + bool
> + select PREEMPT_COUNT
> +
> +config PREEMPT_DYNAMIC_OPTION
> + bool
> + help
> + Symbol that gates availablility of PREEMPT_DYNAMIC, selected
> + by preemption models that can be dynamically selected.
> +
> choice
> prompt "Preemption Model"
> + default PREEMPT_TYPE_NONE
>
> +config PREEMPT_TYPE_NONE
> bool "No Forced Preemption (Server)"
> + select PREEMPT_DYNAMIC_OPTION if HAVE_PREEMPT_DYNAMIC
> help
> This is the traditional Linux preemption model, geared towards
> throughput. It will still provide good latencies most of the
> @@ -17,9 +31,10 @@ config PREEMPT_NONE
> raw processing power of the kernel, irrespective of scheduling
> latencies.
>
> +config PREEMPT_TYPE_VOLUNTARY
> bool "Voluntary Kernel Preemption (Desktop)"
> depends on !ARCH_NO_PREEMPT
> + select PREEMPT_DYNAMIC_OPTION if HAVE_PREEMPT_DYNAMIC
> help
> This option reduces the latency of the kernel by adding more
> "explicit preemption points" to the kernel code. These new
> @@ -35,12 +50,10 @@ config PREEMPT_VOLUNTARY
>
> Select this if you are building a kernel for a desktop system.
>
> -config PREEMPT
> +config PREEMPT_TYPE_FULL
> bool "Preemptible Kernel (Low-Latency Desktop)"
> depends on !ARCH_NO_PREEMPT
> + select PREEMPT_DYNAMIC_OPTION if HAVE_PREEMPT_DYNAMIC
> help
> This option reduces the latency of the kernel by making
> all kernel code (that is not executing in a critical section)
> @@ -75,15 +88,24 @@ config PREEMPT_RT
>
> endchoice
>
> +# default model for PREEMPT_DYNAMIC
>
> +config PREEMPT_DYNAMIC_NONE
> + bool
>
> +config PREEMPT_DYNAMIC_VOLUNTARY
> + bool
> +
> +config PREEMPT_DYNAMIC_FULL
> bool
> +
> +config PREEMPT_DYNAMIC
> + bool "Dynamic Preemption Mode"
> + depends on PREEMPT_DYNAMIC_OPTION
> + select PREEMPT
> + select PREEMPT_DYNAMIC_NONE if PREEMPT_TYPE_NONE
> + select PREEMPT_DYNAMIC_VOLUNTARY if PREEMPT_TYPE_VOLUNTARY
> + select PREEMPT_DYNAMIC_FULL if PREEMPT_TYPE_FULL
Arguably PREEMPT_DYNAMIC_{NONE,VOLUNTARY,FULL} are superfluous and the
default selection can use PREEMPT_TYPE_*.
> help
> This option allows to define the preemption model on the kernel
> command line parameter and thus override the default preemption
> @@ -100,6 +122,22 @@ config PREEMPT_DYNAMIC
> Interesting if you want the same pre-built kernel should be used for
> both Server and Desktop workloads.
>
> +# actual preemption model
> +
> +config PREEMPT_NONE
> + bool
> + default y if !PREEMPT_DYNAMIC && PREEMPT_TYPE_NONE
> +
> +config PREEMPT_VOLUNTARY
> + bool
> + default y if !PREEMPT_DYNAMIC && PREEMPT_TYPE_VOLUNTARY
> +
> +config PREEMPT
> + bool
> + default y if !PREEMPT_DYNAMIC && PREEMPT_TYPE_FULL
> + select PREEMPTION
> + select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK
> +
> config SCHED_CORE
> bool "Core Scheduling for SMT"
> default y
next prev parent reply other threads:[~2021-06-08 14:05 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-08 12:04 [PATCH 0/2] sched: Define default preempt dynamic boot mode on Kconfig Frederic Weisbecker
2021-06-08 12:04 ` [PATCH 1/2] sched: Add default dynamic preempt mode Kconfig Frederic Weisbecker
2021-06-08 13:55 ` Peter Zijlstra
2021-06-08 14:04 ` Peter Zijlstra [this message]
2021-06-09 11:27 ` Frederic Weisbecker
2021-06-09 11:13 ` Frederic Weisbecker
2021-06-08 13:55 ` Peter Zijlstra
2021-06-09 11:25 ` Frederic Weisbecker
2021-06-08 12:04 ` [PATCH 2/2] sched: Always print out preempt dynamic state Frederic Weisbecker
2021-06-08 13:56 ` Peter Zijlstra
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=YL94/8cSr2R2wiTa@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=frederic@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
/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