Re: [PATCH 11/24] cfq-iosched: add leaf_weight

2013-01-08 Thread Tejun Heo 
Hello, Vivek.

On Tue, Jan 08, 2013 at 10:34:48AM -0500, Vivek Goyal wrote:
> How does it help to map leaf weight to weight in root group. Old programs
> anyway don't know about leaf_weight. So nobody is going to update it. And
> if they update it, they better know what does it do.

Because what cfq does before this patchset in flat hierarchy is
basically treating root weight as the leaf weight.

   R:500
  / \
 A:250  B:750

Ratio of R is computed as 500 / (500 + 250 + 750).  The root blkcg's
tasks are competinig with the children blkcgs.  This is currently the
only supported config and we don't want to break existing users, so
keeping root's weight and leaf_weight synchronized and ignoring root's
weight, which BTW can't mean anything anyway as it doesn't have a
parent or siblings, keeps the behavior unchanged for flat configs.

Thanks.

-- 
tejun
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 11/24] cfq-iosched: add leaf_weight

2013-01-08 Thread Vivek Goyal 
On Fri, Dec 28, 2012 at 12:35:33PM -0800, Tejun Heo wrote:

[..]
> +
> + /* on root, leaf_weight is mapped to weight */
> + {
> + .name = "leaf_weight_device",
> + .flags = CFTYPE_ONLY_ON_ROOT,
> + .read_seq_string = cfqg_print_weight_device,
> + .write_string = cfqg_set_weight_device,
> + .max_write_len = 256,
> + },
> + {
> + .name = "leaf_weight",
> + .flags = CFTYPE_ONLY_ON_ROOT,
> + .read_seq_string = cfq_print_weight,
> + .write_u64 = cfq_set_weight,
> + },

Hi Tejun,

How does it help to map leaf weight to weight in root group. Old programs
anyway don't know about leaf_weight. So nobody is going to update it. And
if they update it, they better know what does it do.

I think we just need to map "weight" to "leaf_weight" once you switch to
making use of leaf_weight and at that point of time a user will expect
that updating weight retains the old behavior. I think you have done
that in later patch.

But mapping leaf_weight to weight seems unnecessary atleast from backward
compatibility point of view.

Thanks
Vivek
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 11/24] cfq-iosched: add leaf_weight

2013-01-08 Thread Vivek Goyal 
On Fri, Dec 28, 2012 at 12:35:33PM -0800, Tejun Heo wrote:

[..]
 +
 + /* on root, leaf_weight is mapped to weight */
 + {
 + .name = leaf_weight_device,
 + .flags = CFTYPE_ONLY_ON_ROOT,
 + .read_seq_string = cfqg_print_weight_device,
 + .write_string = cfqg_set_weight_device,
 + .max_write_len = 256,
 + },
 + {
 + .name = leaf_weight,
 + .flags = CFTYPE_ONLY_ON_ROOT,
 + .read_seq_string = cfq_print_weight,
 + .write_u64 = cfq_set_weight,
 + },

Hi Tejun,

How does it help to map leaf weight to weight in root group. Old programs
anyway don't know about leaf_weight. So nobody is going to update it. And
if they update it, they better know what does it do.

I think we just need to map weight to leaf_weight once you switch to
making use of leaf_weight and at that point of time a user will expect
that updating weight retains the old behavior. I think you have done
that in later patch.

But mapping leaf_weight to weight seems unnecessary atleast from backward
compatibility point of view.

Thanks
Vivek
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 11/24] cfq-iosched: add leaf_weight

2013-01-08 Thread Tejun Heo 
Hello, Vivek.

On Tue, Jan 08, 2013 at 10:34:48AM -0500, Vivek Goyal wrote:
 How does it help to map leaf weight to weight in root group. Old programs
 anyway don't know about leaf_weight. So nobody is going to update it. And
 if they update it, they better know what does it do.

Because what cfq does before this patchset in flat hierarchy is
basically treating root weight as the leaf weight.

   R:500
  / \
 A:250  B:750

Ratio of R is computed as 500 / (500 + 250 + 750).  The root blkcg's
tasks are competinig with the children blkcgs.  This is currently the
only supported config and we don't want to break existing users, so
keeping root's weight and leaf_weight synchronized and ignoring root's
weight, which BTW can't mean anything anyway as it doesn't have a
parent or siblings, keeps the behavior unchanged for flat configs.

Thanks.

-- 
tejun
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[PATCH 11/24] cfq-iosched: add leaf_weight

2012-12-28 Thread Tejun Heo 
cfq blkcg is about to grow proper hierarchy handling, where a child
blkg's weight would nest inside the parent's.  This makes tasks in a
blkg to compete against both tasks in the sibling blkgs and the tasks
of child blkgs.

We're gonna use the existing weight as the group weight which decides
the blkg's weight against its siblings.  This patch introduces a new
weight - leaf_weight - which decides the weight of a blkg against the
child blkgs.

It's named leaf_weight because another way to look at it is that each
internal blkg nodes have a hidden child leaf node which contains all
its tasks and leaf_weight is the weight of the leaf node and handled
the same as the weight of the child blkgs.

This patch only adds leaf_weight fields and exposes it to userland.
The new weight isn't actually used anywhere yet.  Note that
cfq-iosched currently offcially supports only single level hierarchy
and root blkgs compete with the first level blkgs - ie. root weight is
basically being used as leaf_weight.  For root blkgs, the two weights
are kept in sync for backward compatibility.

v2: cfqd->root_group->leaf_weight initialization was missing from
cfq_init_queue() causing divide by zero when
!CONFIG_CFQ_GROUP_SCHED.  Fix it.  Reported by Fengguang.

Signed-off-by: Tejun Heo 
Cc: Fengguang Wu 
---
 block/blk-cgroup.c  |   4 +-
 block/blk-cgroup.h  |   1 +
 block/cfq-iosched.c | 134 
 3 files changed, 130 insertions(+), 9 deletions(-)

diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c
index 942f344..10e1df9 100644
--- a/block/blk-cgroup.c
+++ b/block/blk-cgroup.c
@@ -26,7 +26,8 @@
 
 static DEFINE_MUTEX(blkcg_pol_mutex);
 
-struct blkcg blkcg_root = { .cfq_weight = 2 * CFQ_WEIGHT_DEFAULT };
+struct blkcg blkcg_root = { .cfq_weight = 2 * CFQ_WEIGHT_DEFAULT,
+   .cfq_leaf_weight = 2 * CFQ_WEIGHT_DEFAULT, };
 EXPORT_SYMBOL_GPL(blkcg_root);
 
 static struct blkcg_policy *blkcg_policy[BLKCG_MAX_POLS];
@@ -710,6 +711,7 @@ static struct cgroup_subsys_state *blkcg_css_alloc(struct 
cgroup *cgroup)
return ERR_PTR(-ENOMEM);
 
blkcg->cfq_weight = CFQ_WEIGHT_DEFAULT;
+   blkcg->cfq_leaf_weight = CFQ_WEIGHT_DEFAULT;
blkcg->id = atomic64_inc_return(_seq); /* root is 0, start from 1 */
 done:
spin_lock_init(>lock);
diff --git a/block/blk-cgroup.h b/block/blk-cgroup.h
index b26ed58..2446225 100644
--- a/block/blk-cgroup.h
+++ b/block/blk-cgroup.h
@@ -54,6 +54,7 @@ struct blkcg {
 
/* TODO: per-policy storage in blkcg */
unsigned intcfq_weight; /* belongs to cfq */
+   unsigned intcfq_leaf_weight;
 };
 
 struct blkg_stat {
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index bc076f4..175218d6 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -223,10 +223,21 @@ struct cfq_group {
 
/* group service_tree key */
u64 vdisktime;
+
+   /*
+* There are two weights - (internal) weight is the weight of this
+* cfqg against the sibling cfqgs.  leaf_weight is the wight of
+* this cfqg against the child cfqgs.  For the root cfqg, both
+* weights are kept in sync for backward compatibility.
+*/
unsigned int weight;
unsigned int new_weight;
unsigned int dev_weight;
 
+   unsigned int leaf_weight;
+   unsigned int new_leaf_weight;
+   unsigned int dev_leaf_weight;
+
/* number of cfqq currently on this group */
int nr_cfqq;
 
@@ -1182,10 +1193,16 @@ static void
 cfq_update_group_weight(struct cfq_group *cfqg)
 {
BUG_ON(!RB_EMPTY_NODE(>rb_node));
+
if (cfqg->new_weight) {
cfqg->weight = cfqg->new_weight;
cfqg->new_weight = 0;
}
+
+   if (cfqg->new_leaf_weight) {
+   cfqg->leaf_weight = cfqg->new_leaf_weight;
+   cfqg->new_leaf_weight = 0;
+   }
 }
 
 static void
@@ -1348,6 +1365,7 @@ static void cfq_pd_init(struct blkcg_gq *blkg)
 
cfq_init_cfqg_base(cfqg);
cfqg->weight = blkg->blkcg->cfq_weight;
+   cfqg->leaf_weight = blkg->blkcg->cfq_leaf_weight;
 }
 
 /*
@@ -1404,6 +1422,26 @@ static int cfqg_print_weight_device(struct cgroup *cgrp, 
struct cftype *cft,
return 0;
 }
 
+static u64 cfqg_prfill_leaf_weight_device(struct seq_file *sf,
+ struct blkg_policy_data *pd, int off)
+{
+   struct cfq_group *cfqg = pd_to_cfqg(pd);
+
+   if (!cfqg->dev_leaf_weight)
+   return 0;
+   return __blkg_prfill_u64(sf, pd, cfqg->dev_leaf_weight);
+}
+
+static int cfqg_print_leaf_weight_device(struct cgroup *cgrp,
+struct cftype *cft,
+struct seq_file *sf)
+{
+   blkcg_print_blkgs(sf, cgroup_to_blkcg(cgrp),
+ cfqg_prfill_leaf_weight_device, _policy_cfq, 0,
+

[PATCH 11/24] cfq-iosched: add leaf_weight

2012-12-28 Thread Tejun Heo 
cfq blkcg is about to grow proper hierarchy handling, where a child
blkg's weight would nest inside the parent's.  This makes tasks in a
blkg to compete against both tasks in the sibling blkgs and the tasks
of child blkgs.

We're gonna use the existing weight as the group weight which decides
the blkg's weight against its siblings.  This patch introduces a new
weight - leaf_weight - which decides the weight of a blkg against the
child blkgs.

It's named leaf_weight because another way to look at it is that each
internal blkg nodes have a hidden child leaf node which contains all
its tasks and leaf_weight is the weight of the leaf node and handled
the same as the weight of the child blkgs.

This patch only adds leaf_weight fields and exposes it to userland.
The new weight isn't actually used anywhere yet.  Note that
cfq-iosched currently offcially supports only single level hierarchy
and root blkgs compete with the first level blkgs - ie. root weight is
basically being used as leaf_weight.  For root blkgs, the two weights
are kept in sync for backward compatibility.

v2: cfqd-root_group-leaf_weight initialization was missing from
cfq_init_queue() causing divide by zero when
!CONFIG_CFQ_GROUP_SCHED.  Fix it.  Reported by Fengguang.

Signed-off-by: Tejun Heo t...@kernel.org
Cc: Fengguang Wu fengguang...@intel.com
---
 block/blk-cgroup.c  |   4 +-
 block/blk-cgroup.h  |   1 +
 block/cfq-iosched.c | 134 
 3 files changed, 130 insertions(+), 9 deletions(-)

diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c
index 942f344..10e1df9 100644
--- a/block/blk-cgroup.c
+++ b/block/blk-cgroup.c
@@ -26,7 +26,8 @@
 
 static DEFINE_MUTEX(blkcg_pol_mutex);
 
-struct blkcg blkcg_root = { .cfq_weight = 2 * CFQ_WEIGHT_DEFAULT };
+struct blkcg blkcg_root = { .cfq_weight = 2 * CFQ_WEIGHT_DEFAULT,
+   .cfq_leaf_weight = 2 * CFQ_WEIGHT_DEFAULT, };
 EXPORT_SYMBOL_GPL(blkcg_root);
 
 static struct blkcg_policy *blkcg_policy[BLKCG_MAX_POLS];
@@ -710,6 +711,7 @@ static struct cgroup_subsys_state *blkcg_css_alloc(struct 
cgroup *cgroup)
return ERR_PTR(-ENOMEM);
 
blkcg-cfq_weight = CFQ_WEIGHT_DEFAULT;
+   blkcg-cfq_leaf_weight = CFQ_WEIGHT_DEFAULT;
blkcg-id = atomic64_inc_return(id_seq); /* root is 0, start from 1 */
 done:
spin_lock_init(blkcg-lock);
diff --git a/block/blk-cgroup.h b/block/blk-cgroup.h
index b26ed58..2446225 100644
--- a/block/blk-cgroup.h
+++ b/block/blk-cgroup.h
@@ -54,6 +54,7 @@ struct blkcg {
 
/* TODO: per-policy storage in blkcg */
unsigned intcfq_weight; /* belongs to cfq */
+   unsigned intcfq_leaf_weight;
 };
 
 struct blkg_stat {
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index bc076f4..175218d6 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -223,10 +223,21 @@ struct cfq_group {
 
/* group service_tree key */
u64 vdisktime;
+
+   /*
+* There are two weights - (internal) weight is the weight of this
+* cfqg against the sibling cfqgs.  leaf_weight is the wight of
+* this cfqg against the child cfqgs.  For the root cfqg, both
+* weights are kept in sync for backward compatibility.
+*/
unsigned int weight;
unsigned int new_weight;
unsigned int dev_weight;
 
+   unsigned int leaf_weight;
+   unsigned int new_leaf_weight;
+   unsigned int dev_leaf_weight;
+
/* number of cfqq currently on this group */
int nr_cfqq;
 
@@ -1182,10 +1193,16 @@ static void
 cfq_update_group_weight(struct cfq_group *cfqg)
 {
BUG_ON(!RB_EMPTY_NODE(cfqg-rb_node));
+
if (cfqg-new_weight) {
cfqg-weight = cfqg-new_weight;
cfqg-new_weight = 0;
}
+
+   if (cfqg-new_leaf_weight) {
+   cfqg-leaf_weight = cfqg-new_leaf_weight;
+   cfqg-new_leaf_weight = 0;
+   }
 }
 
 static void
@@ -1348,6 +1365,7 @@ static void cfq_pd_init(struct blkcg_gq *blkg)
 
cfq_init_cfqg_base(cfqg);
cfqg-weight = blkg-blkcg-cfq_weight;
+   cfqg-leaf_weight = blkg-blkcg-cfq_leaf_weight;
 }
 
 /*
@@ -1404,6 +1422,26 @@ static int cfqg_print_weight_device(struct cgroup *cgrp, 
struct cftype *cft,
return 0;
 }
 
+static u64 cfqg_prfill_leaf_weight_device(struct seq_file *sf,
+ struct blkg_policy_data *pd, int off)
+{
+   struct cfq_group *cfqg = pd_to_cfqg(pd);
+
+   if (!cfqg-dev_leaf_weight)
+   return 0;
+   return __blkg_prfill_u64(sf, pd, cfqg-dev_leaf_weight);
+}
+
+static int cfqg_print_leaf_weight_device(struct cgroup *cgrp,
+struct cftype *cft,
+struct seq_file *sf)
+{
+   blkcg_print_blkgs(sf, cgroup_to_blkcg(cgrp),
+ cfqg_prfill_leaf_weight_device,