mirror of
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git
synced 2024-09-12 05:48:40 +00:00
0ce65797a7
Otherwise, for sequential workloads, the dm_request_fn can allow excessive request merging at the expense of increased service time. Add a per-device sysfs attribute to allow the user to control how long a request, that is a reasonable merge candidate, can be queued on the request queue. The resolution of this request dispatch deadline is in microseconds (ranging from 1 to 100000 usecs), to set a 20us deadline: echo 20 > /sys/block/dm-7/dm/rq_based_seq_io_merge_deadline The dm_request_fn's merge heuristic and associated extra accounting is disabled by default (rq_based_seq_io_merge_deadline is 0). This sysfs attribute is not applicable to bio-based DM devices so it will only ever report 0 for them. By allowing a request to remain on the queue it will block others requests on the queue. But introducing a short dequeue delay has proven very effective at enabling certain sequential IO workloads on really fast, yet IOPS constrained, devices to build up slightly larger IOs -- yielding 90+% throughput improvements. Having precise control over the time taken to wait for larger requests to build affords control beyond that of waiting for certain IO sizes to accumulate (which would require a deadline anyway). This knob will only ever make sense with sequential IO workloads and the particular value used is storage configuration specific. Given the expected niche use-case for when this knob is useful it has been deemed acceptable to expose this relatively crude method for crafting optimal IO on specific storage -- especially given the solution is simple yet effective. In the context of DM multipath, it is advisable to tune this sysfs attribute to a value that offers the best performance for the common case (e.g. if 4 paths are expected active, tune for that; if paths fail then performance may be slightly reduced). Alternatives were explored to have request-based DM autotune this value (e.g. if/when paths fail) but they were quickly deemed too fragile and complex to warrant further design and development time. If this problem proves more common as faster storage emerges we'll have to look at elevating a generic solution into the block core. Tested-by: Shiva Krishna Merla <shivakrishna.merla@netapp.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
135 lines
2.9 KiB
C
135 lines
2.9 KiB
C
/*
|
|
* Copyright (C) 2008 Red Hat, Inc. All rights reserved.
|
|
*
|
|
* This file is released under the GPL.
|
|
*/
|
|
|
|
#include <linux/sysfs.h>
|
|
#include <linux/dm-ioctl.h>
|
|
#include "dm.h"
|
|
|
|
struct dm_sysfs_attr {
|
|
struct attribute attr;
|
|
ssize_t (*show)(struct mapped_device *, char *);
|
|
ssize_t (*store)(struct mapped_device *, const char *, size_t count);
|
|
};
|
|
|
|
#define DM_ATTR_RO(_name) \
|
|
struct dm_sysfs_attr dm_attr_##_name = \
|
|
__ATTR(_name, S_IRUGO, dm_attr_##_name##_show, NULL)
|
|
|
|
static ssize_t dm_attr_show(struct kobject *kobj, struct attribute *attr,
|
|
char *page)
|
|
{
|
|
struct dm_sysfs_attr *dm_attr;
|
|
struct mapped_device *md;
|
|
ssize_t ret;
|
|
|
|
dm_attr = container_of(attr, struct dm_sysfs_attr, attr);
|
|
if (!dm_attr->show)
|
|
return -EIO;
|
|
|
|
md = dm_get_from_kobject(kobj);
|
|
if (!md)
|
|
return -EINVAL;
|
|
|
|
ret = dm_attr->show(md, page);
|
|
dm_put(md);
|
|
|
|
return ret;
|
|
}
|
|
|
|
#define DM_ATTR_RW(_name) \
|
|
struct dm_sysfs_attr dm_attr_##_name = \
|
|
__ATTR(_name, S_IRUGO | S_IWUSR, dm_attr_##_name##_show, dm_attr_##_name##_store)
|
|
|
|
static ssize_t dm_attr_store(struct kobject *kobj, struct attribute *attr,
|
|
const char *page, size_t count)
|
|
{
|
|
struct dm_sysfs_attr *dm_attr;
|
|
struct mapped_device *md;
|
|
ssize_t ret;
|
|
|
|
dm_attr = container_of(attr, struct dm_sysfs_attr, attr);
|
|
if (!dm_attr->store)
|
|
return -EIO;
|
|
|
|
md = dm_get_from_kobject(kobj);
|
|
if (!md)
|
|
return -EINVAL;
|
|
|
|
ret = dm_attr->store(md, page, count);
|
|
dm_put(md);
|
|
|
|
return ret;
|
|
}
|
|
|
|
static ssize_t dm_attr_name_show(struct mapped_device *md, char *buf)
|
|
{
|
|
if (dm_copy_name_and_uuid(md, buf, NULL))
|
|
return -EIO;
|
|
|
|
strcat(buf, "\n");
|
|
return strlen(buf);
|
|
}
|
|
|
|
static ssize_t dm_attr_uuid_show(struct mapped_device *md, char *buf)
|
|
{
|
|
if (dm_copy_name_and_uuid(md, NULL, buf))
|
|
return -EIO;
|
|
|
|
strcat(buf, "\n");
|
|
return strlen(buf);
|
|
}
|
|
|
|
static ssize_t dm_attr_suspended_show(struct mapped_device *md, char *buf)
|
|
{
|
|
sprintf(buf, "%d\n", dm_suspended_md(md));
|
|
|
|
return strlen(buf);
|
|
}
|
|
|
|
static DM_ATTR_RO(name);
|
|
static DM_ATTR_RO(uuid);
|
|
static DM_ATTR_RO(suspended);
|
|
static DM_ATTR_RW(rq_based_seq_io_merge_deadline);
|
|
|
|
static struct attribute *dm_attrs[] = {
|
|
&dm_attr_name.attr,
|
|
&dm_attr_uuid.attr,
|
|
&dm_attr_suspended.attr,
|
|
&dm_attr_rq_based_seq_io_merge_deadline.attr,
|
|
NULL,
|
|
};
|
|
|
|
static const struct sysfs_ops dm_sysfs_ops = {
|
|
.show = dm_attr_show,
|
|
.store = dm_attr_store,
|
|
};
|
|
|
|
static struct kobj_type dm_ktype = {
|
|
.sysfs_ops = &dm_sysfs_ops,
|
|
.default_attrs = dm_attrs,
|
|
.release = dm_kobject_release,
|
|
};
|
|
|
|
/*
|
|
* Initialize kobj
|
|
* because nobody using md yet, no need to call explicit dm_get/put
|
|
*/
|
|
int dm_sysfs_init(struct mapped_device *md)
|
|
{
|
|
return kobject_init_and_add(dm_kobject(md), &dm_ktype,
|
|
&disk_to_dev(dm_disk(md))->kobj,
|
|
"%s", "dm");
|
|
}
|
|
|
|
/*
|
|
* Remove kobj, called after all references removed
|
|
*/
|
|
void dm_sysfs_exit(struct mapped_device *md)
|
|
{
|
|
struct kobject *kobj = dm_kobject(md);
|
|
kobject_put(kobj);
|
|
wait_for_completion(dm_get_completion_from_kobject(kobj));
|
|
}
|