From: Lin Ming <ming.m.lin@intel.com>
To: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Jens Axboe <axboe@kernel.dk>,
Alan Stern <stern@rowland.harvard.edu>,
"Rafael J. Wysocki" <rjw@sisk.pl>, Shaohua Li <shli@kernel.org>,
linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
linux-scsi@vger.kernel.org
Subject: Re: [PATCH v5 3/4] block: implement runtime pm strategy
Date: Fri, 06 Jul 2012 14:07:52 +0800 [thread overview]
Message-ID: <1341554872.3628.30.camel@minggr> (raw)
In-Reply-To: <1341550812.3023.3.camel@dabdike.hilton.com>
On Fri, 2012-07-06 at 09:00 +0400, James Bottomley wrote:
> On Fri, 2012-07-06 at 12:04 +0800, Lin Ming wrote:
> > When a request is added:
> > If device is suspended or is suspending and the request is not a
> > PM request, resume the device.
> >
> > When a request finishes:
> > Call pm_runtime_mark_last_busy().
> >
> > When pick a request:
> > If device is resuming/suspending, then only PM request is allowed to go.
> > Return NULL for other cases.
>
> This is a complete reinvention of the quiesce state, just with new names
> and moved up to block in part ... why do we have to have two separate
> systems for stopping a device and sending special commands when the
> device is suspended, why not just one?
Yes, there are some duplicates with scsi layer quiesce state.
I'd like to do the cleanup.
Add runtime pm support to block layer, so other block device drivers may
also add runtime pm support easily in future.
Some helper functions are provided on block layer which can be called by
block device drivers.
void blk_pm_runtime_init(struct request_queue *q, struct device *dev)
int blk_pre_runtime_suspend(struct request_queue *q)
void blk_post_runtime_suspend(struct request_queue *q, int err)
void blk_pre_runtime_resume(struct request_queue *q)
void blk_post_runtime_resume(struct request_queue *q, int err)
Thanks,
Lin Ming
>
> James
>
> > Signed-off-by: Lin Ming <ming.m.lin@intel.com>
> > ---
> > block/blk-core.c | 7 +++++++
> > block/elevator.c | 4 ++++
> > include/linux/blkdev.h | 37 +++++++++++++++++++++++++++++++++++++
> > 3 files changed, 48 insertions(+)
> >
> > diff --git a/block/blk-core.c b/block/blk-core.c
> > index 1cc80ae..cb93501 100644
> > --- a/block/blk-core.c
> > +++ b/block/blk-core.c
> > @@ -1224,6 +1224,8 @@ void __blk_put_request(struct request_queue *q, struct request *req)
> > if (unlikely(--req->ref_count))
> > return;
> >
> > + blk_pm_put_request(q);
> > +
> > elv_completed_request(q, req);
> >
> > /* this is a bio leak */
> > @@ -2012,6 +2014,11 @@ struct request *blk_peek_request(struct request_queue *q)
> > int ret;
> >
> > while ((rq = __elv_next_request(q)) != NULL) {
> > +
> > + rq = blk_pm_peek_request(q, rq);
> > + if (!rq)
> > + break;
> > +
> > if (!(rq->cmd_flags & REQ_STARTED)) {
> > /*
> > * This is the first time the device driver
> > diff --git a/block/elevator.c b/block/elevator.c
> > index 6a55d41..37c1a2b 100644
> > --- a/block/elevator.c
> > +++ b/block/elevator.c
> > @@ -536,6 +536,8 @@ void elv_requeue_request(struct request_queue *q, struct request *rq)
> >
> > rq->cmd_flags &= ~REQ_STARTED;
> >
> > + blk_pm_requeue_request(q);
> > +
> > __elv_add_request(q, rq, ELEVATOR_INSERT_REQUEUE);
> > }
> >
> > @@ -558,6 +560,8 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where)
> > {
> > trace_block_rq_insert(q, rq);
> >
> > + blk_pm_add_request(q, rq);
> > +
> > rq->q = q;
> >
> > if (rq->cmd_flags & REQ_SOFTBARRIER) {
> > diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> > index 9395d39..c8c5951 100644
> > --- a/include/linux/blkdev.h
> > +++ b/include/linux/blkdev.h
> > @@ -908,6 +908,36 @@ extern int blk_pre_runtime_suspend(struct request_queue *q);
> > extern void blk_post_runtime_suspend(struct request_queue *q, int err);
> > extern void blk_pre_runtime_resume(struct request_queue *q);
> > extern void blk_post_runtime_resume(struct request_queue *q, int err);
> > +
> > +static inline void blk_pm_put_request(struct request_queue *q)
> > +{
> > + if (!(--q->nr_pending) && q->dev)
> > + pm_runtime_mark_last_busy(q->dev);
> > +}
> > +
> > +static inline struct request *blk_pm_peek_request(
> > + struct request_queue *q, struct request *rq)
> > +{
> > + if (q->rpm_status == RPM_SUSPENDED ||
> > + (q->rpm_status != RPM_ACTIVE && !(rq->cmd_flags & REQ_PM)))
> > + return NULL;
> > + else
> > + return rq;
> > +}
> > +
> > +static inline void blk_pm_requeue_request(struct request_queue *q)
> > +{
> > + q->nr_pending--;
> > +}
> > +
> > +static inline void blk_pm_add_request(struct request_queue *q,
> > + struct request *rq)
> > +{
> > + if (q->nr_pending++ == 0 && !(rq->cmd_flags & REQ_PM) &&
> > + (q->rpm_status == RPM_SUSPENDED ||
> > + q->rpm_status == RPM_SUSPENDING) && q->dev)
> > + pm_request_resume(q->dev);
> > +}
> > #else
> > static inline void blk_pm_runtime_init(struct request_queue *q,
> > struct device *dev) {}
> > @@ -918,6 +948,13 @@ static inline int blk_pre_runtime_suspend(struct request_queue *q)
> > static inline void blk_post_runtime_suspend(struct request_queue *q, int err) {}
> > static inline void blk_pre_runtime_resume(struct request_queue *q) {}
> > static inline void blk_post_runtime_resume(struct request_queue *q, int err) {}
> > +
> > +static inline void blk_pm_put_request(struct request_queue *q) {}
> > +static inline struct request *blk_pm_peek_request(
> > + struct request_queue *q, struct request *rq) { return rq; }
> > +static inline void blk_pm_requeue_request(struct request_queue *q) {}
> > +static inline void blk_pm_add_request(struct request_queue *q,
> > + struct request *req) {}
> > #endif
> >
> > /*
>
>
next prev parent reply other threads:[~2012-07-06 6:07 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-07-06 4:04 [PATCH v5 0/4] block layer runtime pm Lin Ming
2012-07-06 4:04 ` [PATCH v5 1/4] block: add a flag to identify PM request Lin Ming
2012-07-06 4:04 ` [PATCH v5 2/4] block: add runtime pm helpers Lin Ming
2012-07-06 4:04 ` [PATCH v5 3/4] block: implement runtime pm strategy Lin Ming
2012-07-06 5:00 ` James Bottomley
2012-07-06 6:07 ` Lin Ming [this message]
2012-07-06 8:05 ` James Bottomley
2012-07-06 14:59 ` Alan Stern
2013-01-14 7:18 ` Aaron Lu
2012-07-06 7:27 ` Matthew Wilcox
2012-07-06 14:21 ` Alan Stern
2012-07-06 14:51 ` Lin Ming
2012-07-06 4:04 ` [PATCH v5 4/4] [SCSI] sd: change to auto suspend mode Lin Ming
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=1341554872.3628.30.camel@minggr \
--to=ming.m.lin@intel.com \
--cc=James.Bottomley@HansenPartnership.com \
--cc=axboe@kernel.dk \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=rjw@sisk.pl \
--cc=shli@kernel.org \
--cc=stern@rowland.harvard.edu \
/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).