[PATCH cgroup/for-4.3 1/3] cgroup: replace error handling in cgroup_init() with WARN_ON()s

[PATCH cgroup/for-4.3 1/3] cgroup: replace error handling in cgroup_init() with WARN_ON()s

[Tejun Heo]

The init sequence shouldn't fail short of bugs and even when it does
it's better to continue with the rest of initialization and we were
silently ignoring /proc/cgroups creation failure.

Drop the explicit error handling and wrap sysfs_create_mount_point(),
register_filesystem() and proc_create() with WARN_ON()s.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
---
 kernel/cgroup.c |   15 ++++-----------
 1 file changed, 4 insertions(+), 11 deletions(-)

--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -5014,7 +5014,7 @@ int __init cgroup_init(void)
 {
 	struct cgroup_subsys *ss;
 	unsigned long key;
-	int ssid, err;
+	int ssid;
 
 	BUG_ON(percpu_init_rwsem(&cgroup_threadgroup_rwsem));
 	BUG_ON(cgroup_init_cftypes(NULL, cgroup_dfl_base_files));
@@ -5072,17 +5072,10 @@ int __init cgroup_init(void)
 			ss->bind(init_css_set.subsys[ssid]);
 	}
 
-	err = sysfs_create_mount_point(fs_kobj, "cgroup");
-	if (err)
-		return err;
-
-	err = register_filesystem(&cgroup_fs_type);
-	if (err < 0) {
-		sysfs_remove_mount_point(fs_kobj, "cgroup");
-		return err;
-	}
+	WARN_ON(sysfs_create_mount_point(fs_kobj, "cgroup"));
+	WARN_ON(register_filesystem(&cgroup_fs_type));
+	WARN_ON(!proc_create("cgroups", 0, NULL, &proc_cgroupstats_operations));
 
-	proc_create("cgroups", 0, NULL, &proc_cgroupstats_operations);
 	return 0;
 }
 

Re: [PATCH cgroup/for-4.3 2/3] cgroup: drop cgroup__DEVEL__legacy_files_on_dfl

[Tejun Heo]

Now that interfaces for the major three controllers - cpu, memory, io
- are shaping up, there's no reason to have an option to force legacy
files to show up on the unified hierarchy for testing.  Drop it.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
---
 Documentation/cgroups/unified-hierarchy.txt |    6 -----
 kernel/cgroup.c                             |   30 +---------------------------
 2 files changed, 2 insertions(+), 34 deletions(-)

--- a/Documentation/cgroups/unified-hierarchy.txt
+++ b/Documentation/cgroups/unified-hierarchy.txt
@@ -107,12 +107,6 @@ root of unified hierarchy can be bound t
 allows mixing unified hierarchy with the traditional multiple
 hierarchies in a fully backward compatible way.
 
-For development purposes, the following boot parameter makes all
-controllers to appear on the unified hierarchy whether supported or
-not.
-
- cgroup__DEVEL__legacy_files_on_dfl
-
 A controller can be moved across hierarchies only after the controller
 is no longer referenced in its current hierarchy.  Because per-cgroup
 controller states are destroyed asynchronously and controllers may
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -152,12 +152,6 @@ struct cgroup_root cgrp_dfl_root;
  */
 static bool cgrp_dfl_root_visible;
 
-/*
- * Set by the boot param of the same name and makes subsystems with NULL
- * ->dfl_files to use ->legacy_files on the default hierarchy.
- */
-static bool cgroup_legacy_files_on_dfl;
-
 /* some controllers are not supported in the default hierarchy */
 static unsigned long cgrp_dfl_root_inhibit_ss_mask;
 
@@ -3340,17 +3334,8 @@ int cgroup_add_legacy_cftypes(struct cgr
 {
 	struct cftype *cft;
 
-	/*
-	 * If legacy_flies_on_dfl, we want to show the legacy files on the
-	 * dfl hierarchy but iff the target subsystem hasn't been updated
-	 * for the dfl hierarchy yet.
-	 */
-	if (!cgroup_legacy_files_on_dfl ||
-	    ss->dfl_cftypes != ss->legacy_cftypes) {
-		for (cft = cfts; cft && cft->name[0] != '\0'; cft++)
-			cft->flags |= __CFTYPE_NOT_ON_DFL;
-	}
-
+	for (cft = cfts; cft && cft->name[0] != '\0'; cft++)
+		cft->flags |= __CFTYPE_NOT_ON_DFL;
 	return cgroup_add_cftypes(ss, cfts);
 }
 
@@ -5055,9 +5040,6 @@ int __init cgroup_init(void)
 
 		cgrp_dfl_root.subsys_mask |= 1 << ss->id;
 
-		if (cgroup_legacy_files_on_dfl && !ss->dfl_cftypes)
-			ss->dfl_cftypes = ss->legacy_cftypes;
-
 		if (!ss->dfl_cftypes)
 			cgrp_dfl_root_inhibit_ss_mask |= 1 << ss->id;
 
@@ -5474,14 +5456,6 @@ static int __init cgroup_disable(char *s
 }
 __setup("cgroup_disable=", cgroup_disable);
 
-static int __init cgroup_set_legacy_files_on_dfl(char *str)
-{
-	printk("cgroup: using legacy files on the default hierarchy\n");
-	cgroup_legacy_files_on_dfl = true;
-	return 0;
-}
-__setup("cgroup__DEVEL__legacy_files_on_dfl", cgroup_set_legacy_files_on_dfl);
-
 /**
  * css_tryget_online_from_dir - get corresponding css from a cgroup dentry
  * @dentry: directory dentry of interest

[PATCH cgroup/for-4.3 3/3] cgroup: replace __DEVEL__sane_behavior with cgroup2 fs type

[Tejun Heo]

With major controllers - cpu, memory and io - shaping up for the
unified hierarchy, cgroup2 is about ready to be, gradually, released
into the wild.  Replace __DEVEL__sane_behavior flag which was used to
select the unified hierarchy with a separate filesystem type "cgroup2"
so that unified hierarchy can be mounted as follows.

  mount -t cgroup2 none $MOUNT_POINT

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
---
 Documentation/cgroups/unified-hierarchy.txt |    6 +--
 include/linux/cgroup-defs.h                 |    1 
 kernel/cgroup.c                             |   43 +++++++++++++---------------
 3 files changed, 22 insertions(+), 28 deletions(-)

--- a/Documentation/cgroups/unified-hierarchy.txt
+++ b/Documentation/cgroups/unified-hierarchy.txt
@@ -94,11 +94,9 @@ the process.
 
 2-1. Mounting
 
-Currently, unified hierarchy can be mounted with the following mount
-command.  Note that this is still under development and scheduled to
-change soon.
+Unified hierarchy can be mounted with the following mount command.
 
- mount -t cgroup -o __DEVEL__sane_behavior cgroup $MOUNT_POINT
+ mount -t cgroup2 none $MOUNT_POINT
 
 All controllers which support the unified hierarchy and are not bound
 to other hierarchies are automatically bound to unified hierarchy and
--- a/include/linux/cgroup-defs.h
+++ b/include/linux/cgroup-defs.h
@@ -66,7 +66,6 @@ enum {
 
 /* cgroup_root->flags */
 enum {
-	CGRP_ROOT_SANE_BEHAVIOR	= (1 << 0), /* __DEVEL__sane_behavior specified */
 	CGRP_ROOT_NOPREFIX	= (1 << 1), /* mounted subsystems have no named prefix */
 	CGRP_ROOT_XATTR		= (1 << 2), /* supports extended attributes */
 };
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -183,6 +183,7 @@ static unsigned long have_exit_callback
 /* Ditto for the can_fork callback. */
 static unsigned long have_canfork_callback __read_mostly;
 
+static struct file_system_type cgroup2_fs_type;
 static struct cftype cgroup_dfl_base_files[];
 static struct cftype cgroup_legacy_base_files[];
 
@@ -1391,10 +1392,6 @@ static int parse_cgroupfs_options(char *
 			all_ss = true;
 			continue;
 		}
-		if (!strcmp(token, "__DEVEL__sane_behavior")) {
-			opts->flags |= CGRP_ROOT_SANE_BEHAVIOR;
-			continue;
-		}
 		if (!strcmp(token, "noprefix")) {
 			opts->flags |= CGRP_ROOT_NOPREFIX;
 			continue;
@@ -1461,15 +1458,6 @@ static int parse_cgroupfs_options(char *
 			return -ENOENT;
 	}
 
-	if (opts->flags & CGRP_ROOT_SANE_BEHAVIOR) {
-		pr_warn("sane_behavior: this is still under development and its behaviors will change, proceed at your own risk\n");
-		if (nr_opts != 1) {
-			pr_err("sane_behavior: no other mount options allowed\n");
-			return -EINVAL;
-		}
-		return 0;
-	}
-
 	/*
 	 * If the 'all' option was specified select all the subsystems,
 	 * otherwise if 'none', 'name=' and a subsystem name options were
@@ -1765,6 +1753,17 @@ static struct dentry *cgroup_mount(struc
 	if (!use_task_css_set_links)
 		cgroup_enable_task_cg_lists();
 
+	if (fs_type == &cgroup2_fs_type) {
+		if (data) {
+			pr_err("cgroup2: unknown option \"%s\"\n", (char *)data);
+			return ERR_PTR(-EINVAL);
+		}
+		cgrp_dfl_root_visible = true;
+		root = &cgrp_dfl_root;
+		cgroup_get(&root->cgrp);
+		goto out_mount;
+	}
+
 	mutex_lock(&cgroup_mutex);
 
 	/* First find the desired set of subsystems */
@@ -1772,15 +1771,6 @@ static struct dentry *cgroup_mount(struc
 	if (ret)
 		goto out_unlock;
 
-	/* look for a matching existing root */
-	if (opts.flags & CGRP_ROOT_SANE_BEHAVIOR) {
-		cgrp_dfl_root_visible = true;
-		root = &cgrp_dfl_root;
-		cgroup_get(&root->cgrp);
-		ret = 0;
-		goto out_unlock;
-	}
-
 	/*
 	 * Destruction of cgroup root is asynchronous, so subsystems may
 	 * still be dying after the previous unmount.  Let's drain the
@@ -1891,7 +1881,7 @@ out_free:
 
 	if (ret)
 		return ERR_PTR(ret);
-
+out_mount:
 	dentry = kernfs_mount(fs_type, flags, root->kf_root,
 				CGROUP_SUPER_MAGIC, &new_sb);
 	if (IS_ERR(dentry) || !new_sb)
@@ -1936,6 +1926,12 @@ static struct file_system_type cgroup_fs
 	.kill_sb = cgroup_kill_sb,
 };
 
+static struct file_system_type cgroup2_fs_type = {
+	.name = "cgroup2",
+	.mount = cgroup_mount,
+	.kill_sb = cgroup_kill_sb,
+};
+
 /**
  * task_cgroup_path - cgroup path of a task in the first cgroup hierarchy
  * @task: target task
@@ -5056,6 +5052,7 @@ int __init cgroup_init(void)
 
 	WARN_ON(sysfs_create_mount_point(fs_kobj, "cgroup"));
 	WARN_ON(register_filesystem(&cgroup_fs_type));
+	WARN_ON(register_filesystem(&cgroup2_fs_type));
 	WARN_ON(!proc_create("cgroups", 0, NULL, &proc_cgroupstats_operations));
 
 	return 0;

Re: [PATCH cgroup/for-4.3 1/3] cgroup: replace error handling in cgroup_init() with WARN_ON()s

[Zefan Li]

On 2015/8/6 1:20, Tejun Heo wrote:
> The init sequence shouldn't fail short of bugs and even when it does
> it's better to continue with the rest of initialization and we were
> silently ignoring /proc/cgroups creation failure.
>
> Drop the explicit error handling and wrap sysfs_create_mount_point(),
> register_filesystem() and proc_create() with WARN_ON()s.
>
> Signed-off-by: Tejun Heo <tj@kernel.org>
> Cc: Li Zefan <lizefan@huawei.com>
> Cc: Johannes Weiner <hannes@cmpxchg.org>

acked

Re: [PATCH cgroup/for-4.3 3/3] cgroup: replace __DEVEL__sane_behavior with cgroup2 fs type

[Zefan Li]

On 2015/8/6 1:21, Tejun Heo wrote:
> With major controllers - cpu, memory and io - shaping up for the
> unified hierarchy, cgroup2 is about ready to be, gradually, released
> into the wild.  Replace __DEVEL__sane_behavior flag which was used to
> select the unified hierarchy with a separate filesystem type "cgroup2"
> so that unified hierarchy can be mounted as follows.
>
>    mount -t cgroup2 none $MOUNT_POINT
>
> Signed-off-by: Tejun Heo <tj@kernel.org>
> Cc: Li Zefan <lizefan@huawei.com>
> Cc: Johannes Weiner <hannes@cmpxchg.org>

acked

Re: [PATCH cgroup/for-4.3 2/3] cgroup: drop cgroup__DEVEL__legacy_files_on_dfl

[Tejun Heo]

On Wed, Aug 05, 2015 at 01:20:43PM -0400, Tejun Heo wrote:
> Now that interfaces for the major three controllers - cpu, memory, io
> - are shaping up, there's no reason to have an option to force legacy
> files to show up on the unified hierarchy for testing.  Drop it.
> 
> Signed-off-by: Tejun Heo <tj@kernel.org>
> Cc: Li Zefan <lizefan@huawei.com>
> Cc: Johannes Weiner <hannes@cmpxchg.org>

Applying 1-2 to cgroup/for-4.4.  Will apply the third one once the cpu
controller discussion reaches some resolution.

Thanks.

-- 
tejun