linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stefan Hajnoczi <stefanha@redhat.com>
To: Marcelo Tosatti <mtosatti@redhat.com>
Cc: linux-kernel@vger.kernel.org,
	Daniel Lezcano <daniel.lezcano@linaro.org>,
	Stefano Garzarella <sgarzare@redhat.com>,
	Ming Lei <ming.lei@redhat.com>,
	"Michael S . Tsirkin" <mst@redhat.com>,
	Jens Axboe <axboe@kernel.dk>, Jason Wang <jasowang@redhat.com>,
	linux-block@vger.kernel.org,
	"Rafael J. Wysocki" <rjw@rjwysocki.net>,
	virtualization@lists.linux-foundation.org,
	linux-pm@vger.kernel.org, Christoph Hellwig <hch@infradead.org>
Subject: Re: [RFC 1/3] cpuidle: add poll_source API
Date: Tue, 20 Jul 2021 15:15:16 +0100	[thread overview]
Message-ID: <YPbadGW2DP0Bd67P@stefanha-x1.localdomain> (raw)
In-Reply-To: <20210719210355.GA46373@fuller.cnet>

[-- Attachment #1: Type: text/plain, Size: 6762 bytes --]

On Mon, Jul 19, 2021 at 06:03:55PM -0300, Marcelo Tosatti wrote:
> Hi Stefan,
> 
> On Tue, Jul 13, 2021 at 05:19:04PM +0100, Stefan Hajnoczi wrote:
> > Introduce an API for adding cpuidle poll callbacks:
> > 
> >   struct poll_source_ops {
> >       void (*start)(struct poll_source *src);
> >       void (*stop)(struct poll_source *src);
> >       void (*poll)(struct poll_source *src);
> >   };
> > 
> >   int poll_source_register(struct poll_source *src);
> >   int poll_source_unregister(struct poll_source *src);
> > 
> > When cpuidle enters the poll state it invokes ->start() and then invokes
> > ->poll() repeatedly from the busy wait loop. Finally ->stop() is invoked
> > when the busy wait loop finishes.
> > 
> > The ->poll() function should check for activity and cause
> > TIF_NEED_RESCHED to be set in order to stop the busy wait loop.
> > 
> > This API is intended to be used by drivers that can cheaply poll for
> > events. Participating in cpuidle polling allows them to avoid interrupt
> > latencies during periods where the CPU is going to poll anyway.
> > 
> > Note that each poll_source is bound to a particular CPU. The API is
> > mainly intended to by used by drivers that have multiple queues with irq
> > affinity.
> > 
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> > ---
> >  drivers/cpuidle/Makefile      |  1 +
> >  include/linux/poll_source.h   | 53 +++++++++++++++++++
> >  drivers/cpuidle/poll_source.c | 99 +++++++++++++++++++++++++++++++++++
> >  drivers/cpuidle/poll_state.c  |  6 +++
> >  4 files changed, 159 insertions(+)
> >  create mode 100644 include/linux/poll_source.h
> >  create mode 100644 drivers/cpuidle/poll_source.c
> > 
> > diff --git a/drivers/cpuidle/Makefile b/drivers/cpuidle/Makefile
> > index 26bbc5e74123..994f72d6fe95 100644
> > --- a/drivers/cpuidle/Makefile
> > +++ b/drivers/cpuidle/Makefile
> > @@ -7,6 +7,7 @@ obj-y += cpuidle.o driver.o governor.o sysfs.o governors/
> >  obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o
> >  obj-$(CONFIG_DT_IDLE_STATES)		  += dt_idle_states.o
> >  obj-$(CONFIG_ARCH_HAS_CPU_RELAX)	  += poll_state.o
> > +obj-$(CONFIG_ARCH_HAS_CPU_RELAX)	  += poll_source.o
> >  obj-$(CONFIG_HALTPOLL_CPUIDLE)		  += cpuidle-haltpoll.o
> >  
> >  ##################################################################################
> > diff --git a/include/linux/poll_source.h b/include/linux/poll_source.h
> > new file mode 100644
> > index 000000000000..ccfb424e170b
> > --- /dev/null
> > +++ b/include/linux/poll_source.h
> > @@ -0,0 +1,53 @@
> > +/* SPDX-License-Identifier: GPL-2.0-or-later */
> > +/*
> > + * poll_source.h - cpuidle busy waiting API
> > + */
> > +#ifndef __LINUX_POLLSOURCE_H__
> > +#define __LINUX_POLLSOURCE_H__
> > +
> > +#include <linux/list.h>
> > +
> > +struct poll_source;
> > +
> > +struct poll_source_ops {
> > +	void (*start)(struct poll_source *src);
> > +	void (*stop)(struct poll_source *src);
> > +	void (*poll)(struct poll_source *src);
> > +};
> > +
> > +struct poll_source {
> > +	const struct poll_source_ops *ops;
> > +	struct list_head node;
> > +	int cpu;
> > +};
> > +
> > +/**
> > + * poll_source_register - Add a poll_source for a CPU
> > + */
> > +#if defined(CONFIG_CPU_IDLE) && defined(CONFIG_ARCH_HAS_CPU_RELAX)
> > +int poll_source_register(struct poll_source *src);
> > +#else
> > +static inline int poll_source_register(struct poll_source *src)
> > +{
> > +	return 0;
> > +}
> > +#endif
> > +
> > +/**
> > + * poll_source_unregister - Remove a previously registered poll_source
> > + */
> > +#if defined(CONFIG_CPU_IDLE) && defined(CONFIG_ARCH_HAS_CPU_RELAX)
> > +int poll_source_unregister(struct poll_source *src);
> > +#else
> > +static inline int poll_source_unregister(struct poll_source *src)
> > +{
> > +	return 0;
> > +}
> > +#endif
> > +
> > +/* Used by the cpuidle driver */
> > +void poll_source_start(void);
> > +void poll_source_run_once(void);
> > +void poll_source_stop(void);
> > +
> > +#endif /* __LINUX_POLLSOURCE_H__ */
> > diff --git a/drivers/cpuidle/poll_source.c b/drivers/cpuidle/poll_source.c
> > new file mode 100644
> > index 000000000000..46100e5a71e4
> > --- /dev/null
> > +++ b/drivers/cpuidle/poll_source.c
> > @@ -0,0 +1,99 @@
> > +// SPDX-License-Identifier: GPL-2.0-or-later
> > +/*
> > + * poll_source.c - cpuidle busy waiting API
> > + */
> > +
> > +#include <linux/lockdep.h>
> > +#include <linux/percpu.h>
> > +#include <linux/poll_source.h>
> > +
> > +/* The per-cpu list of registered poll sources */
> > +DEFINE_PER_CPU(struct list_head, poll_source_list);
> > +
> > +/* Called from idle task with TIF_POLLING_NRFLAG set and irqs enabled */
> > +void poll_source_start(void)
> > +{
> > +	struct poll_source *src;
> > +
> > +	list_for_each_entry(src, this_cpu_ptr(&poll_source_list), node)
> > +		src->ops->start(src);
> > +}
> > +
> > +/* Called from idle task with TIF_POLLING_NRFLAG set and irqs enabled */
> > +void poll_source_run_once(void)
> > +{
> > +	struct poll_source *src;
> > +
> > +	list_for_each_entry(src, this_cpu_ptr(&poll_source_list), node)
> > +		src->ops->poll(src);
> > +}
> > +
> > +/* Called from idle task with TIF_POLLING_NRFLAG set and irqs enabled */
> > +void poll_source_stop(void)
> > +{
> > +	struct poll_source *src;
> > +
> > +	list_for_each_entry(src, this_cpu_ptr(&poll_source_list), node)
> > +		src->ops->stop(src);
> > +}
> > +
> > +static void poll_source_register_this_cpu(void *opaque)
> > +{
> > +	struct poll_source *src = opaque;
> > +
> > +	lockdep_assert_irqs_disabled();
> > +
> > +	list_add_tail(&src->node, this_cpu_ptr(&poll_source_list));
> > +}
> > +
> > +int poll_source_register(struct poll_source *src)
> > +{
> > +	if (!list_empty(&src->node))
> > +		return -EBUSY;
> > +
> > +	/*
> > +	 * There is no race with src->cpu iterating over poll_source_list
> > +	 * because smp_call_function_single() just sets TIF_NEED_RESCHED
> 
> Hum... what about 
> 
> CPU-0				CPU-1
> 
> poll_source_start		poll_source_register IPI'ing CPU-0
> 
> Perhaps a llist can be used?

The simplest solution might be queue_work_on(src->cpu, system_hipri_wq,
&work) plus a completion. That way we know poll_source_start/stop/run()
are not executing and there can be no re-entrancy issues. The nice thing
is it's simple - no low-level tricks.

Those are the semantics I wanted with smp_call_function_single() but I
think I chose the wrong API :).

> >               while (!need_resched()) {
> > +                     poll_source_run_once();
> 
> Probably want to use static_key's for ->start, ->stop and ->poll?

Good idea, thanks.

Stefan

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  reply	other threads:[~2021-07-20 14:18 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-13 16:19 [RFC 0/3] cpuidle: add poll_source API and virtio vq polling Stefan Hajnoczi
2021-07-13 16:19 ` [RFC 1/3] cpuidle: add poll_source API Stefan Hajnoczi
2021-07-19 21:03   ` Marcelo Tosatti
2021-07-20 14:15     ` Stefan Hajnoczi [this message]
2021-07-13 16:19 ` [RFC 2/3] virtio: add poll_source virtqueue polling Stefan Hajnoczi
2021-07-13 16:19 ` [RFC 3/3] softirq: participate in cpuidle polling Stefan Hajnoczi
2021-07-21  3:29 ` [RFC 0/3] cpuidle: add poll_source API and virtio vq polling Jason Wang
2021-07-21  9:41   ` Stefan Hajnoczi
2021-07-22  9:04     ` Jason Wang
2021-07-26 15:17       ` Stefan Hajnoczi
2021-07-26 15:47         ` Rafael J. Wysocki
2021-07-26 16:01           ` Stefan Hajnoczi
2021-07-26 16:37             ` Rafael J. Wysocki
2021-07-27 13:32               ` Stefan Hajnoczi

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=YPbadGW2DP0Bd67P@stefanha-x1.localdomain \
    --to=stefanha@redhat.com \
    --cc=axboe@kernel.dk \
    --cc=daniel.lezcano@linaro.org \
    --cc=hch@infradead.org \
    --cc=jasowang@redhat.com \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=ming.lei@redhat.com \
    --cc=mst@redhat.com \
    --cc=mtosatti@redhat.com \
    --cc=rjw@rjwysocki.net \
    --cc=sgarzare@redhat.com \
    --cc=virtualization@lists.linux-foundation.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;
as well as URLs for NNTP newsgroup(s).