[PATCH 1/2] cgroup: export list of delegatable control files using sysfs

[PATCH 1/2] cgroup: export list of delegatable control files using sysfs

[Roman Gushchin]

Delegatable cgroup v2 control files may require special handling
(e.g. chowning), and the exact list of such files varies between
kernel versions (and likely to be extended in the future).

To guarantee correctness of this list and simplify the life
of userspace (systemd, first of all), let's export the list
via /sys/kernel/cgroup/delegates pseudo-file.

Format is siple: each control file name is printed on a new line.
Example:
  $ cat /sys/kernel/cgroup/delegates
  cgroup.procs
  cgroup.subtree_control

Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: kernel-team@fb.com
---
 kernel/cgroup/cgroup.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 61 insertions(+)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index d6ed725f36d9..9dff92c17a1f 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5832,3 +5832,64 @@ int cgroup_bpf_update(struct cgroup *cgrp, struct bpf_prog *prog,
 	return ret;
 }
 #endif /* CONFIG_CGROUP_BPF */
+
+#ifdef CONFIG_SYSFS
+static ssize_t show_delegatable_files(struct cftype *files, char *buf,
+				      ssize_t size, const char *prefix)
+{
+	struct cftype *cft;
+	ssize_t ret = 0;
+
+	for (cft = files; cft && cft->name[0] != '\0'; cft++) {
+		if (!(cft->flags & CFTYPE_NS_DELEGATABLE))
+			continue;
+
+		if (prefix)
+			ret += snprintf(buf + ret, size - ret, "%s.", prefix);
+
+		ret += snprintf(buf + ret, size - ret, "%s\n", cft->name);
+
+		if (unlikely(ret >= size)) {
+			WARN_ON(1);
+			break;
+		}
+	}
+
+	return ret;
+}
+
+static ssize_t delegates_show(struct kobject *kobj, struct kobj_attribute *attr,
+			      char *buf)
+{
+	struct cgroup_subsys *ss;
+	int ssid;
+	ssize_t ret = 0;
+
+	ret = show_delegatable_files(cgroup_base_files, buf, PAGE_SIZE - ret,
+				     NULL);
+
+	for_each_subsys(ss, ssid)
+		ret += show_delegatable_files(ss->dfl_cftypes, buf + ret,
+					      PAGE_SIZE - ret,
+					      cgroup_subsys_name[ssid]);
+
+	return ret;
+}
+static struct kobj_attribute cgroup_delegates_attr = __ATTR_RO(delegates);
+
+static struct attribute *cgroup_sysfs_attrs[] = {
+	&cgroup_delegates_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group cgroup_sysfs_attr_group = {
+	.attrs = cgroup_sysfs_attrs,
+	.name = "cgroup",
+};
+
+static int __init cgroup_sysfs_init(void)
+{
+	return sysfs_create_group(kernel_kobj, &cgroup_sysfs_attr_group);
+}
+subsys_initcall(cgroup_sysfs_init);
+#endif /* CONFIG_SYSFS */
-- 
2.13.6

[PATCH 2/2] cgroup: export list of cgroups v2 features using sysfs

[Roman Gushchin]

The active development of cgroups v2 sometimes leads to a creation
of interfaces, which are not turned on by default (to provide
backward compatibility). It's handy to know from userspace, which
cgroup v2 features are supported without calculating it based
on the kernel version. So, let's export the list of such features
using /sys/kernel/cgroup/features pseudo-file.

The list is hardcoded and has to be extended when new functionality
is added. Each feature is printed on a new line.

Example:
  $ cat /sys/kernel/cgroup/features
  nsdelegate

Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: kernel-team@fb.com
---
 kernel/cgroup/cgroup.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 9dff92c17a1f..548b12609943 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5877,8 +5877,16 @@ static ssize_t delegates_show(struct kobject *kobj, struct kobj_attribute *attr,
 }
 static struct kobj_attribute cgroup_delegates_attr = __ATTR_RO(delegates);
 
+static ssize_t features_show(struct kobject *kobj, struct kobj_attribute *attr,
+			     char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "nsdelegate\n");
+}
+static struct kobj_attribute cgroup_features_attr = __ATTR_RO(features);
+
 static struct attribute *cgroup_sysfs_attrs[] = {
 	&cgroup_delegates_attr.attr,
+	&cgroup_features_attr.attr,
 	NULL,
 };
 
-- 
2.13.6

Re: [PATCH 1/2] cgroup: export list of delegatable control files using sysfs

[Tejun Heo]

On Fri, Nov 03, 2017 at 01:46:02PM -0400, Roman Gushchin wrote:
> Delegatable cgroup v2 control files may require special handling
> (e.g. chowning), and the exact list of such files varies between
> kernel versions (and likely to be extended in the future).
> 
> To guarantee correctness of this list and simplify the life
> of userspace (systemd, first of all), let's export the list
> via /sys/kernel/cgroup/delegates pseudo-file.
> 
> Format is siple: each control file name is printed on a new line.
> Example:
>   $ cat /sys/kernel/cgroup/delegates

Generally looks good to me but I'm not sure whether "delegates" is the
right name for the file.  According to dictionary, it doesn't seem to
match too well to what we wanna mean here, which is "these files
belong to the child and should be transferred to the child when
delegating the cgroup".

Thanks.

-- 
tejun

Re: [PATCH 1/2] cgroup: export list of delegatable control files using sysfs

[Tejun Heo]

Hello, Roman.

On Fri, Nov 03, 2017 at 10:56:31AM -0700, Tejun Heo wrote:
> On Fri, Nov 03, 2017 at 01:46:02PM -0400, Roman Gushchin wrote:
> > Delegatable cgroup v2 control files may require special handling
> > (e.g. chowning), and the exact list of such files varies between
> > kernel versions (and likely to be extended in the future).
> > 
> > To guarantee correctness of this list and simplify the life
> > of userspace (systemd, first of all), let's export the list
> > via /sys/kernel/cgroup/delegates pseudo-file.
> > 
> > Format is siple: each control file name is printed on a new line.
> > Example:
> >   $ cat /sys/kernel/cgroup/delegates
> 
> Generally looks good to me but I'm not sure whether "delegates" is the
> right name for the file.  According to dictionary, it doesn't seem to
> match too well to what we wanna mean here, which is "these files
> belong to the child and should be transferred to the child when
> delegating the cgroup".

So, "delegates" as the plural noun means people who represent.  Let's
go with "delegate" which can be read as a verb term meaning "delegate
these files together".

Thanks.

-- 
tejun

[PATCH v2 1/2] cgroup: export list of delegatable control files using sysfs

[Roman Gushchin]

Delegatable cgroup v2 control files may require special handling
(e.g. chowning), and the exact list of such files varies between
kernel versions (and likely to be extended in the future).

To guarantee correctness of this list and simplify the life
of userspace (systemd, first of all), let's export the list
via /sys/kernel/cgroup/delegate pseudo-file.

Format is siple: each control file name is printed on a new line.
Example:
  $ cat /sys/kernel/cgroup/delegate
  cgroup.procs
  cgroup.subtree_control

Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: kernel-team@fb.com
---
 kernel/cgroup/cgroup.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 61 insertions(+)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index d6ed725f36d9..eed92ed624e5 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5832,3 +5832,64 @@ int cgroup_bpf_update(struct cgroup *cgrp, struct bpf_prog *prog,
 	return ret;
 }
 #endif /* CONFIG_CGROUP_BPF */
+
+#ifdef CONFIG_SYSFS
+static ssize_t show_delegatable_files(struct cftype *files, char *buf,
+				      ssize_t size, const char *prefix)
+{
+	struct cftype *cft;
+	ssize_t ret = 0;
+
+	for (cft = files; cft && cft->name[0] != '\0'; cft++) {
+		if (!(cft->flags & CFTYPE_NS_DELEGATABLE))
+			continue;
+
+		if (prefix)
+			ret += snprintf(buf + ret, size - ret, "%s.", prefix);
+
+		ret += snprintf(buf + ret, size - ret, "%s\n", cft->name);
+
+		if (unlikely(ret >= size)) {
+			WARN_ON(1);
+			break;
+		}
+	}
+
+	return ret;
+}
+
+static ssize_t delegate_show(struct kobject *kobj, struct kobj_attribute *attr,
+			      char *buf)
+{
+	struct cgroup_subsys *ss;
+	int ssid;
+	ssize_t ret = 0;
+
+	ret = show_delegatable_files(cgroup_base_files, buf, PAGE_SIZE - ret,
+				     NULL);
+
+	for_each_subsys(ss, ssid)
+		ret += show_delegatable_files(ss->dfl_cftypes, buf + ret,
+					      PAGE_SIZE - ret,
+					      cgroup_subsys_name[ssid]);
+
+	return ret;
+}
+static struct kobj_attribute cgroup_delegate_attr = __ATTR_RO(delegate);
+
+static struct attribute *cgroup_sysfs_attrs[] = {
+	&cgroup_delegate_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group cgroup_sysfs_attr_group = {
+	.attrs = cgroup_sysfs_attrs,
+	.name = "cgroup",
+};
+
+static int __init cgroup_sysfs_init(void)
+{
+	return sysfs_create_group(kernel_kobj, &cgroup_sysfs_attr_group);
+}
+subsys_initcall(cgroup_sysfs_init);
+#endif /* CONFIG_SYSFS */
-- 
2.13.6

[PATCH v2 2/2] cgroup: export list of cgroups v2 features using sysfs

[Roman Gushchin]

The active development of cgroups v2 sometimes leads to a creation
of interfaces, which are not turned on by default (to provide
backward compatibility). It's handy to know from userspace, which
cgroup v2 features are supported without calculating it based
on the kernel version. So, let's export the list of such features
using /sys/kernel/cgroup/features pseudo-file.

The list is hardcoded and has to be extended when new functionality
is added. Each feature is printed on a new line.

Example:
  $ cat /sys/kernel/cgroup/features
  nsdelegate

Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: kernel-team@fb.com
---
 kernel/cgroup/cgroup.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index eed92ed624e5..69e65d28fe98 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5877,8 +5877,16 @@ static ssize_t delegate_show(struct kobject *kobj, struct kobj_attribute *attr,
 }
 static struct kobj_attribute cgroup_delegate_attr = __ATTR_RO(delegate);
 
+static ssize_t features_show(struct kobject *kobj, struct kobj_attribute *attr,
+			     char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "nsdelegate\n");
+}
+static struct kobj_attribute cgroup_features_attr = __ATTR_RO(features);
+
 static struct attribute *cgroup_sysfs_attrs[] = {
 	&cgroup_delegate_attr.attr,
+	&cgroup_features_attr.attr,
 	NULL,
 };
 
-- 
2.13.6

Re: [PATCH v2 2/2] cgroup: export list of cgroups v2 features using sysfs

[Tejun Heo]

On Mon, Nov 06, 2017 at 01:30:29PM -0500, Roman Gushchin wrote:
> The active development of cgroups v2 sometimes leads to a creation
> of interfaces, which are not turned on by default (to provide
> backward compatibility). It's handy to know from userspace, which
> cgroup v2 features are supported without calculating it based
> on the kernel version. So, let's export the list of such features
> using /sys/kernel/cgroup/features pseudo-file.
> 
> The list is hardcoded and has to be extended when new functionality
> is added. Each feature is printed on a new line.
> 
> Example:
>   $ cat /sys/kernel/cgroup/features
>   nsdelegate
> 
> Signed-off-by: Roman Gushchin <guro@fb.com>
> Cc: Tejun Heo <tj@kernel.org>
> Cc: kernel-team@fb.com

Applied 1-2 to cgroup/for-4.15.

Thanks.

-- 
tejun