From: Juergen Gross <jgross@suse.com>
To: "Roger Pau Monné" <roger.pau@citrix.com>
Cc: linux-kernel@vger.kernel.org, xen-devel@lists.xenproject.org,
linux-block@vger.kernel.org, konrad.wilk@oracle.com,
axboe@kernel.dk, boris.ostrovsky@oracle.com
Subject: Re: [PATCH 2/4] xen/blkfront: cleanup stale persistent grants
Date: Tue, 7 Aug 2018 17:56:38 +0200 [thread overview]
Message-ID: <bba40203-ff97-aba6-cbb6-d3e57387778a@suse.com> (raw)
In-Reply-To: <20180807141404.lzsqtdd2seqgwtgx@mac>
On 07/08/18 16:14, Roger Pau Monné wrote:
> On Tue, Aug 07, 2018 at 08:31:31AM +0200, Juergen Gross wrote:
>> On 06/08/18 18:16, Roger Pau Monné wrote:
>>> On Mon, Aug 06, 2018 at 01:34:01PM +0200, Juergen Gross wrote:
>>>> Add a periodic cleanup function to remove old persistent grants which
>>>> are no longer in use on the backend side. This avoids starvation in
>>>> case there are lots of persistent grants for a device which no longer
>>>> is involved in I/O business.
>>>>
>>>> Signed-off-by: Juergen Gross <jgross@suse.com>
>>>> ---
>>>> drivers/block/xen-blkfront.c | 99 ++++++++++++++++++++++++++++++++++++++++++--
>>>> 1 file changed, 95 insertions(+), 4 deletions(-)
>>>>
>>>> diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
>>>> index b5cedccb5d7d..19feb8835fc4 100644
>>>> --- a/drivers/block/xen-blkfront.c
>>>> +++ b/drivers/block/xen-blkfront.c
>>>> @@ -46,6 +46,7 @@
>>>> #include <linux/scatterlist.h>
>>>> #include <linux/bitmap.h>
>>>> #include <linux/list.h>
>>>> +#include <linux/workqueue.h>
>>>>
>>>> #include <xen/xen.h>
>>>> #include <xen/xenbus.h>
>>>> @@ -121,6 +122,9 @@ static inline struct blkif_req *blkif_req(struct request *rq)
>>>>
>>>> static DEFINE_MUTEX(blkfront_mutex);
>>>> static const struct block_device_operations xlvbd_block_fops;
>>>> +static struct delayed_work blkfront_work;
>>>> +static LIST_HEAD(info_list);
>>>> +static bool blkfront_work_active;
>>>>
>>>> /*
>>>> * Maximum number of segments in indirect requests, the actual value used by
>>>> @@ -216,6 +220,7 @@ struct blkfront_info
>>>> /* Save uncomplete reqs and bios for migration. */
>>>> struct list_head requests;
>>>> struct bio_list bio_list;
>>>> + struct list_head info_list;
>>>> };
>>>>
>>>> static unsigned int nr_minors;
>>>> @@ -1764,6 +1769,12 @@ static int write_per_ring_nodes(struct xenbus_transaction xbt,
>>>> return err;
>>>> }
>>>>
>>>> +static void free_info(struct blkfront_info *info)
>>>> +{
>>>> + list_del(&info->info_list);
>>>> + kfree(info);
>>>> +}
>>>> +
>>>> /* Common code used when first setting up, and when resuming. */
>>>> static int talk_to_blkback(struct xenbus_device *dev,
>>>> struct blkfront_info *info)
>>>> @@ -1885,7 +1896,10 @@ static int talk_to_blkback(struct xenbus_device *dev,
>>>> destroy_blkring:
>>>> blkif_free(info, 0);
>>>>
>>>> - kfree(info);
>>>> + mutex_lock(&blkfront_mutex);
>>>> + free_info(info);
>>>> + mutex_unlock(&blkfront_mutex);
>>>> +
>>>> dev_set_drvdata(&dev->dev, NULL);
>>>>
>>>> return err;
>>>> @@ -1996,6 +2010,10 @@ static int blkfront_probe(struct xenbus_device *dev,
>>>> info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0);
>>>> dev_set_drvdata(&dev->dev, info);
>>>>
>>>> + mutex_lock(&blkfront_mutex);
>>>> + list_add(&info->info_list, &info_list);
>>>> + mutex_unlock(&blkfront_mutex);
>>>> +
>>>> return 0;
>>>> }
>>>>
>>>> @@ -2306,6 +2324,15 @@ static void blkfront_gather_backend_features(struct blkfront_info *info)
>>>> if (indirect_segments <= BLKIF_MAX_SEGMENTS_PER_REQUEST)
>>>> indirect_segments = 0;
>>>> info->max_indirect_segments = indirect_segments;
>>>> +
>>>> + if (info->feature_persistent) {
>>>> + mutex_lock(&blkfront_mutex);
>>>> + if (!blkfront_work_active) {
>>>> + blkfront_work_active = true;
>>>> + schedule_delayed_work(&blkfront_work, HZ * 10);
>>>
>>> Does it make sense to provide a module parameter to rune the schedule
>>> of the cleanup routine?
>>
>> I don't think this is something anyone would like to tune.
>>
>> In case you think it should be tunable I can add a parameter, of course.
>
> We can always add it later if required. I'm fine as-is now.
>
>>>
>>>> + }
>>>> + mutex_unlock(&blkfront_mutex);
>>>
>>> Is it really necessary to have the blkfront_work_active boolean? What
>>> happens if you queue the same delayed work more than once?
>>
>> In case there is already work queued later calls of
>> schedule_delayed_work() will be ignored.
>>
>> So yes, I can drop the global boolean (I still need a local flag in
>> blkfront_delay_work() for controlling the need to call
>> schedule_delayed_work() again).
>
> Can't you just call schedule_delayed_work if info->feature_persistent
> is set, even if that means calling it multiple times if multiple
> blkfront instances are using persistent grants?
I don't like that. With mq we have a high chance for multiple instances
to use persistent grants and a local bool is much cheaper than unneeded
calls of schedule_delayed_work().
Juergen
next prev parent reply other threads:[~2018-08-07 15:56 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-08-06 11:33 [PATCH 0/4] xen/blk: persistent grant rework Juergen Gross
2018-08-06 11:33 ` [PATCH 1/4] xen/blkback: don't keep persistent grants too long Juergen Gross
2018-08-06 15:58 ` Roger Pau Monné
2018-08-07 6:34 ` Juergen Gross
2018-08-06 11:34 ` [PATCH] xen/blkfront: remove unused macros Juergen Gross
2018-08-06 11:36 ` Juergen Gross
2018-08-06 11:34 ` [PATCH 2/4] xen/blkfront: cleanup stale persistent grants Juergen Gross
2018-08-06 16:16 ` Roger Pau Monné
2018-08-07 6:31 ` Juergen Gross
2018-08-07 14:14 ` Roger Pau Monné
2018-08-07 15:56 ` Juergen Gross [this message]
2018-08-08 8:27 ` Roger Pau Monné
2018-08-06 11:34 ` [PATCH 3/4] xen/blkfront: reorder tests in xlblk_init() Juergen Gross
2018-08-06 16:18 ` Roger Pau Monné
2018-08-06 11:34 ` [PATCH 4/4] xen/blkback: remove unused pers_gnts_lock from struct xen_blkif_ring Juergen Gross
2018-08-06 16:20 ` Roger Pau Monné
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=bba40203-ff97-aba6-cbb6-d3e57387778a@suse.com \
--to=jgross@suse.com \
--cc=axboe@kernel.dk \
--cc=boris.ostrovsky@oracle.com \
--cc=konrad.wilk@oracle.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=roger.pau@citrix.com \
--cc=xen-devel@lists.xenproject.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).