All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
To: lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org
Cc: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
	cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: [PATCH 06/14] cgroup: add __rcu modifier to cgroup->subsys[]
Date: Thu,  8 Aug 2013 16:13:43 -0400	[thread overview]
Message-ID: <1375992831-4650-7-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1375992831-4650-1-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>

For the planned unified hierarchy, each css (cgroup_subsys_state) will
be RCU protected so that it can be created and destroyed individually
while allowing RCU accesses.  Previous changes ensured that all
cgroup->subsys[] accesses use the cgroup_css() accessor.  This patch
adds __rcu modifier to cgroup->subsys[], add matching RCU dereference
in cgroup_css() and convert all assignments to either
rcu_assign_pointer() or RCU_INIT_POINTER().

This change prepares for the actual RCUfication of css's and doesn't
introduce any visible behavior change.  The conversion is verified
with sparse and all accesses are properly RCU annotated.

Signed-off-by: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
 include/linux/cgroup.h |  2 +-
 kernel/cgroup.c        | 19 ++++++++++++++-----
 2 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index 8a5dc91..eb200b5 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -204,7 +204,7 @@ struct cgroup {
 	struct cgroup_name __rcu *name;
 
 	/* Private pointers for each registered subsystem */
-	struct cgroup_subsys_state *subsys[CGROUP_SUBSYS_COUNT];
+	struct cgroup_subsys_state __rcu *subsys[CGROUP_SUBSYS_COUNT];
 
 	struct cgroupfs_root *root;
 
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index d63beff..c271016 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -229,11 +229,16 @@ static int cgroup_addrm_files(struct cgroup *cgrp, struct cftype cfts[],
  * @subsys_id: the subsystem of interest
  *
  * Return @cgrp's css (cgroup_subsys_state) associated with @subsys_id.
+ * This function must be called either under cgroup_mutex or
+ * rcu_read_lock() and the caller is responsible for pinning the returned
+ * css if it wants to keep accessing it outside the said locks.  This
+ * function may return %NULL if @cgrp doesn't have @subsys_id enabled.
  */
 static struct cgroup_subsys_state *cgroup_css(struct cgroup *cgrp,
 					      int subsys_id)
 {
-	return cgrp->subsys[subsys_id];
+	return rcu_dereference_check(cgrp->subsys[subsys_id],
+				     lockdep_is_held(&cgroup_mutex));
 }
 
 /* convenient tests for these bits */
@@ -1072,8 +1077,10 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 			BUG_ON(!cgroup_css(cgroup_dummy_top, i));
 			BUG_ON(cgroup_css(cgroup_dummy_top, i)->cgroup != cgroup_dummy_top);
 
-			cgrp->subsys[i] = cgroup_dummy_top->subsys[i];
+			rcu_assign_pointer(cgrp->subsys[i],
+					   cgroup_css(cgroup_dummy_top, i));
 			cgroup_css(cgrp, i)->cgroup = cgrp;
+
 			list_move(&ss->sibling, &root->subsys_list);
 			ss->root = root;
 			if (ss->bind)
@@ -1088,8 +1095,10 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 
 			if (ss->bind)
 				ss->bind(cgroup_css(cgroup_dummy_top, i));
+
 			cgroup_css(cgroup_dummy_top, i)->cgroup = cgroup_dummy_top;
-			cgrp->subsys[i] = NULL;
+			RCU_INIT_POINTER(cgrp->subsys[i], NULL);
+
 			cgroup_subsys[i]->root = &cgroup_dummy_root;
 			list_move(&ss->sibling, &cgroup_dummy_root.subsys_list);
 
@@ -4314,7 +4323,7 @@ static void init_cgroup_css(struct cgroup_subsys_state *css,
 		css->flags |= CSS_ROOT;
 
 	BUG_ON(cgroup_css(cgrp, ss->subsys_id));
-	cgrp->subsys[ss->subsys_id] = css;
+	rcu_assign_pointer(cgrp->subsys[ss->subsys_id], css);
 }
 
 /* invoke ->css_online() on a new CSS and mark it online if successful */
@@ -4962,7 +4971,7 @@ void cgroup_unload_subsys(struct cgroup_subsys *ss)
 	 * also takes care of freeing the css_id.
 	 */
 	ss->css_free(cgroup_css(cgroup_dummy_top, ss->subsys_id));
-	cgroup_dummy_top->subsys[ss->subsys_id] = NULL;
+	RCU_INIT_POINTER(cgroup_dummy_top->subsys[ss->subsys_id], NULL);
 
 	mutex_unlock(&cgroup_mutex);
 }
-- 
1.8.3.1

WARNING: multiple messages have this Message-ID (diff)
From: Tejun Heo <tj@kernel.org>
To: lizefan@huawei.com
Cc: containers@lists.linux-foundation.org, cgroups@vger.kernel.org,
	linux-kernel@vger.kernel.org, Tejun Heo <tj@kernel.org>
Subject: [PATCH 06/14] cgroup: add __rcu modifier to cgroup->subsys[]
Date: Thu,  8 Aug 2013 16:13:43 -0400	[thread overview]
Message-ID: <1375992831-4650-7-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1375992831-4650-1-git-send-email-tj@kernel.org>

For the planned unified hierarchy, each css (cgroup_subsys_state) will
be RCU protected so that it can be created and destroyed individually
while allowing RCU accesses.  Previous changes ensured that all
cgroup->subsys[] accesses use the cgroup_css() accessor.  This patch
adds __rcu modifier to cgroup->subsys[], add matching RCU dereference
in cgroup_css() and convert all assignments to either
rcu_assign_pointer() or RCU_INIT_POINTER().

This change prepares for the actual RCUfication of css's and doesn't
introduce any visible behavior change.  The conversion is verified
with sparse and all accesses are properly RCU annotated.

Signed-off-by: Tejun Heo <tj@kernel.org>
---
 include/linux/cgroup.h |  2 +-
 kernel/cgroup.c        | 19 ++++++++++++++-----
 2 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index 8a5dc91..eb200b5 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -204,7 +204,7 @@ struct cgroup {
 	struct cgroup_name __rcu *name;
 
 	/* Private pointers for each registered subsystem */
-	struct cgroup_subsys_state *subsys[CGROUP_SUBSYS_COUNT];
+	struct cgroup_subsys_state __rcu *subsys[CGROUP_SUBSYS_COUNT];
 
 	struct cgroupfs_root *root;
 
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index d63beff..c271016 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -229,11 +229,16 @@ static int cgroup_addrm_files(struct cgroup *cgrp, struct cftype cfts[],
  * @subsys_id: the subsystem of interest
  *
  * Return @cgrp's css (cgroup_subsys_state) associated with @subsys_id.
+ * This function must be called either under cgroup_mutex or
+ * rcu_read_lock() and the caller is responsible for pinning the returned
+ * css if it wants to keep accessing it outside the said locks.  This
+ * function may return %NULL if @cgrp doesn't have @subsys_id enabled.
  */
 static struct cgroup_subsys_state *cgroup_css(struct cgroup *cgrp,
 					      int subsys_id)
 {
-	return cgrp->subsys[subsys_id];
+	return rcu_dereference_check(cgrp->subsys[subsys_id],
+				     lockdep_is_held(&cgroup_mutex));
 }
 
 /* convenient tests for these bits */
@@ -1072,8 +1077,10 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 			BUG_ON(!cgroup_css(cgroup_dummy_top, i));
 			BUG_ON(cgroup_css(cgroup_dummy_top, i)->cgroup != cgroup_dummy_top);
 
-			cgrp->subsys[i] = cgroup_dummy_top->subsys[i];
+			rcu_assign_pointer(cgrp->subsys[i],
+					   cgroup_css(cgroup_dummy_top, i));
 			cgroup_css(cgrp, i)->cgroup = cgrp;
+
 			list_move(&ss->sibling, &root->subsys_list);
 			ss->root = root;
 			if (ss->bind)
@@ -1088,8 +1095,10 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 
 			if (ss->bind)
 				ss->bind(cgroup_css(cgroup_dummy_top, i));
+
 			cgroup_css(cgroup_dummy_top, i)->cgroup = cgroup_dummy_top;
-			cgrp->subsys[i] = NULL;
+			RCU_INIT_POINTER(cgrp->subsys[i], NULL);
+
 			cgroup_subsys[i]->root = &cgroup_dummy_root;
 			list_move(&ss->sibling, &cgroup_dummy_root.subsys_list);
 
@@ -4314,7 +4323,7 @@ static void init_cgroup_css(struct cgroup_subsys_state *css,
 		css->flags |= CSS_ROOT;
 
 	BUG_ON(cgroup_css(cgrp, ss->subsys_id));
-	cgrp->subsys[ss->subsys_id] = css;
+	rcu_assign_pointer(cgrp->subsys[ss->subsys_id], css);
 }
 
 /* invoke ->css_online() on a new CSS and mark it online if successful */
@@ -4962,7 +4971,7 @@ void cgroup_unload_subsys(struct cgroup_subsys *ss)
 	 * also takes care of freeing the css_id.
 	 */
 	ss->css_free(cgroup_css(cgroup_dummy_top, ss->subsys_id));
-	cgroup_dummy_top->subsys[ss->subsys_id] = NULL;
+	RCU_INIT_POINTER(cgroup_dummy_top->subsys[ss->subsys_id], NULL);
 
 	mutex_unlock(&cgroup_mutex);
 }
-- 
1.8.3.1


  parent reply	other threads:[~2013-08-08 20:13 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-08 20:13 [PATCHSET cgroup/for-3.12] cgroup: decouple cgroup_subsys_state lifetime from that of cgroup Tejun Heo
2013-08-08 20:13 ` Tejun Heo
2013-08-08 20:13 ` [PATCH 03/14] cgroup: add cgroup_subsys_state->parent Tejun Heo
2013-08-08 20:13 ` [PATCH 05/14] cgroup: make cgroup_file_open() rcu_read_lock() around cgroup_css() and add cfent->css Tejun Heo
     [not found] ` <1375992831-4650-1-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
2013-08-08 20:13   ` [PATCH 01/14] cgroup: always use cgroup_css() Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 02/14] cgroup: rename cgroup_subsys_state->dput_work and its callback function Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 03/14] cgroup: add cgroup_subsys_state->parent Tejun Heo
2013-08-08 20:13   ` [PATCH 04/14] cgroup: cgroup_css_from_dir() now should be called with RCU read locked Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 05/14] cgroup: make cgroup_file_open() rcu_read_lock() around cgroup_css() and add cfent->css Tejun Heo
2013-08-08 20:13   ` Tejun Heo [this message]
2013-08-08 20:13     ` [PATCH 06/14] cgroup: add __rcu modifier to cgroup->subsys[] Tejun Heo
2013-08-08 20:13   ` [PATCH 07/14] cgroup: reorganize css init / exit paths Tejun Heo
2013-08-08 20:13     ` Tejun Heo
     [not found]     ` <1375992831-4650-8-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
2013-08-12  2:47       ` Li Zefan
2013-08-12  2:47         ` Li Zefan
     [not found]         ` <52084CC5.8050207-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-12 13:39           ` Tejun Heo
2013-08-12 13:39             ` Tejun Heo
2013-08-12 13:40       ` [PATCH v2 " Tejun Heo
2013-08-12 13:40         ` Tejun Heo
2013-08-08 20:13   ` [PATCH 08/14] cgroup: move cgroup->subsys[] assignment to online_css() Tejun Heo
2013-08-08 20:13     ` Tejun Heo
     [not found]     ` <1375992831-4650-9-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
2013-08-14  0:27       ` [PATCH v2 " Tejun Heo
2013-08-14  0:27         ` Tejun Heo
2013-08-14  0:27       ` Tejun Heo
2013-08-08 20:13   ` [PATCH 09/14] cgroup: bounce cgroup_subsys_state ref kill confirmation to a work item Tejun Heo
2013-08-08 20:13   ` Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 10/14] cgroup: replace cgroup->css_kill_cnt with ->nr_css Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 11/14] cgroup: decouple cgroup_subsys_state destruction from cgroup destruction Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 12/14] cgroup: factor out kill_css() Tejun Heo
2013-08-08 20:13   ` [PATCH 13/14] cgroup: move subsys file removal to kill_css() Tejun Heo
2013-08-08 20:13   ` Tejun Heo
2013-08-08 20:13     ` Tejun Heo
2013-08-08 20:13   ` [PATCH 14/14] cgroup: RCU protect each cgroup_subsys_state release Tejun Heo
2013-08-13  1:19   ` [PATCHSET cgroup/for-3.12] cgroup: decouple cgroup_subsys_state lifetime from that of cgroup Li Zefan
2013-08-13  1:19     ` Li Zefan
2013-08-13 15:02   ` Tejun Heo
2013-08-13 15:02   ` Tejun Heo
2013-08-13 15:02     ` Tejun Heo
2013-08-08 20:13 ` [PATCH 12/14] cgroup: factor out kill_css() Tejun Heo
2013-08-08 20:13 ` [PATCH 14/14] cgroup: RCU protect each cgroup_subsys_state release Tejun Heo

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=1375992831-4650-7-git-send-email-tj@kernel.org \
    --to=tj-dgejt+ai2ygdnm+yrofe0a@public.gmane.org \
    --cc=cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
    --cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.