* [PATCH v3 12/14] net: amt: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Andrew Lunn, David S. Miller, Eric Dumazet,
Jakub Kicinski, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Taehee Yoo <ap420073@gmail.com>
Cc: Andrew Lunn <andrew+netdev@lunn.ch>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
drivers/net/amt.c | 8 --------
1 file changed, 8 deletions(-)
diff --git a/drivers/net/amt.c b/drivers/net/amt.c
index ed86537b2f61..aaed9fbc7526 100644
--- a/drivers/net/amt.c
+++ b/drivers/net/amt.c
@@ -295,7 +295,6 @@ static void amt_source_work(struct work_struct *work)
tunnel = gnode->tunnel_list;
spin_lock_bh(&tunnel->lock);
- rcu_read_lock();
if (gnode->filter_mode == MCAST_INCLUDE) {
amt_destroy_source(snode);
if (!gnode->nr_sources)
@@ -306,7 +305,6 @@ static void amt_source_work(struct work_struct *work)
*/
snode->status = AMT_SOURCE_STATUS_D_FWD;
}
- rcu_read_unlock();
spin_unlock_bh(&tunnel->lock);
}
@@ -426,7 +424,6 @@ static void amt_group_work(struct work_struct *work)
goto out;
}
- rcu_read_lock();
for (i = 0; i < buckets; i++) {
hlist_for_each_entry_safe(snode, t,
&gnode->sources[i], node) {
@@ -443,7 +440,6 @@ static void amt_group_work(struct work_struct *work)
amt_del_group(amt, gnode);
else
gnode->filter_mode = MCAST_INCLUDE;
- rcu_read_unlock();
spin_unlock_bh(&tunnel->lock);
out:
dev_put(amt->dev);
@@ -1327,11 +1323,9 @@ static void amt_clear_groups(struct amt_tunnel_list *tunnel)
int i;
spin_lock_bh(&tunnel->lock);
- rcu_read_lock();
for (i = 0; i < amt->hash_buckets; i++)
hlist_for_each_entry_safe(gnode, t, &tunnel->groups[i], node)
amt_del_group(amt, gnode);
- rcu_read_unlock();
spin_unlock_bh(&tunnel->lock);
}
@@ -1343,11 +1337,9 @@ static void amt_tunnel_expire(struct work_struct *work)
struct amt_dev *amt = tunnel->amt;
spin_lock_bh(&amt->lock);
- rcu_read_lock();
list_del_rcu(&tunnel->list);
amt->nr_tunnels--;
amt_clear_groups(tunnel);
- rcu_read_unlock();
spin_unlock_bh(&amt->lock);
kfree_rcu(tunnel, rcu);
}
--
2.34.1
^ permalink raw reply related
* [PATCH v3 11/14] net: ncsi: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Samuel Mendoza-Jonas, Paul Fertser,
pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Samuel Mendoza-Jonas <sam@mendozajonas.com>
Cc: Paul Fertser <fercerpav@gmail.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
net/ncsi/ncsi-manage.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/net/ncsi/ncsi-manage.c b/net/ncsi/ncsi-manage.c
index 446e4e3b9553..6e36cd64a31e 100644
--- a/net/ncsi/ncsi-manage.c
+++ b/net/ncsi/ncsi-manage.c
@@ -650,7 +650,6 @@ static int set_one_vid(struct ncsi_dev_priv *ndp, struct ncsi_channel *nc,
spin_lock_irqsave(&nc->lock, flags);
- rcu_read_lock();
list_for_each_entry_rcu(vlan, &ndp->vlan_vids, list) {
vid = vlan->vid;
for (i = 0; i < ncf->n_vids; i++)
@@ -661,7 +660,6 @@ static int set_one_vid(struct ncsi_dev_priv *ndp, struct ncsi_channel *nc,
if (vid)
break;
}
- rcu_read_unlock();
if (!vid) {
/* No VLAN ID is not set */
--
2.34.1
^ permalink raw reply related
* [PATCH v3 10/14] wifi: mac80211: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Johannes Berg, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
net/mac80211/cfg.c | 2 --
net/mac80211/debugfs.c | 2 --
net/mac80211/debugfs_netdev.c | 2 --
net/mac80211/debugfs_sta.c | 2 --
net/mac80211/sta_info.c | 2 --
5 files changed, 10 deletions(-)
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c
index 2ed07fa121ab..4fe50d4c461d 100644
--- a/net/mac80211/cfg.c
+++ b/net/mac80211/cfg.c
@@ -4825,7 +4825,6 @@ static int ieee80211_get_txq_stats(struct wiphy *wiphy,
int ret = 0;
spin_lock_bh(&local->fq.lock);
- rcu_read_lock();
if (wdev) {
sdata = IEEE80211_WDEV_TO_SUB_IF(wdev);
@@ -4851,7 +4850,6 @@ static int ieee80211_get_txq_stats(struct wiphy *wiphy,
}
out:
- rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
return ret;
diff --git a/net/mac80211/debugfs.c b/net/mac80211/debugfs.c
index e8b78ec682da..82099f4cedbe 100644
--- a/net/mac80211/debugfs.c
+++ b/net/mac80211/debugfs.c
@@ -82,7 +82,6 @@ static ssize_t aqm_read(struct file *file,
int len = 0;
spin_lock_bh(&local->fq.lock);
- rcu_read_lock();
len = scnprintf(buf, sizeof(buf),
"access name value\n"
@@ -105,7 +104,6 @@ static ssize_t aqm_read(struct file *file,
fq->limit,
fq->quantum);
- rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
return simple_read_from_buffer(user_buf, count, ppos,
diff --git a/net/mac80211/debugfs_netdev.c b/net/mac80211/debugfs_netdev.c
index 1dac78271045..30a5a978a678 100644
--- a/net/mac80211/debugfs_netdev.c
+++ b/net/mac80211/debugfs_netdev.c
@@ -625,7 +625,6 @@ static ssize_t ieee80211_if_fmt_aqm(
txqi = to_txq_info(sdata->vif.txq);
spin_lock_bh(&local->fq.lock);
- rcu_read_lock();
len = scnprintf(buf,
buflen,
@@ -642,7 +641,6 @@ static ssize_t ieee80211_if_fmt_aqm(
txqi->tin.tx_bytes,
txqi->tin.tx_packets);
- rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
return len;
diff --git a/net/mac80211/debugfs_sta.c b/net/mac80211/debugfs_sta.c
index 49061bd4151b..ef75255d47d5 100644
--- a/net/mac80211/debugfs_sta.c
+++ b/net/mac80211/debugfs_sta.c
@@ -148,7 +148,6 @@ static ssize_t sta_aqm_read(struct file *file, char __user *userbuf,
return -ENOMEM;
spin_lock_bh(&local->fq.lock);
- rcu_read_lock();
p += scnprintf(p,
bufsz + buf - p,
@@ -178,7 +177,6 @@ static ssize_t sta_aqm_read(struct file *file, char __user *userbuf,
test_bit(IEEE80211_TXQ_DIRTY, &txqi->flags) ? " DIRTY" : "");
}
- rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
rv = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c
index 8c550aab9bdc..663318a75d7f 100644
--- a/net/mac80211/sta_info.c
+++ b/net/mac80211/sta_info.c
@@ -2637,13 +2637,11 @@ static void sta_set_tidstats(struct sta_info *sta,
if (link_id < 0 && tid < IEEE80211_NUM_TIDS) {
spin_lock_bh(&local->fq.lock);
- rcu_read_lock();
tidstats->filled |= BIT(NL80211_TID_STATS_TXQ_STATS);
ieee80211_fill_txq_stats(&tidstats->txq_stats,
to_txq_info(sta->sta.txq[tid]));
- rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
}
}
--
2.34.1
^ permalink raw reply related
* [PATCH v3 09/14] cgroup/cpuset: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Johannes Weiner, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Waiman Long <longman@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Waiman Long <longman@redhat.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
kernel/cgroup/cpuset.c | 6 ------
1 file changed, 6 deletions(-)
diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
index 27adb04df675..9b7e8e8e9411 100644
--- a/kernel/cgroup/cpuset.c
+++ b/kernel/cgroup/cpuset.c
@@ -4073,7 +4073,6 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask)
struct cpuset *cs;
spin_lock_irqsave(&callback_lock, flags);
- rcu_read_lock();
cs = task_cs(tsk);
if (cs != &top_cpuset)
@@ -4095,7 +4094,6 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask)
cpumask_copy(pmask, possible_mask);
}
- rcu_read_unlock();
spin_unlock_irqrestore(&callback_lock, flags);
}
@@ -4168,9 +4166,7 @@ nodemask_t cpuset_mems_allowed(struct task_struct *tsk)
unsigned long flags;
spin_lock_irqsave(&callback_lock, flags);
- rcu_read_lock();
guarantee_online_mems(task_cs(tsk), &mask);
- rcu_read_unlock();
spin_unlock_irqrestore(&callback_lock, flags);
return mask;
@@ -4265,10 +4261,8 @@ bool cpuset_current_node_allowed(int node, gfp_t gfp_mask)
/* Not hardwall and node outside mems_allowed: scan up cpusets */
spin_lock_irqsave(&callback_lock, flags);
- rcu_read_lock();
cs = nearest_hardwall_ancestor(task_cs(current));
allowed = node_isset(node, cs->mems_allowed);
- rcu_read_unlock();
spin_unlock_irqrestore(&callback_lock, flags);
return allowed;
--
2.34.1
^ permalink raw reply related
* [PATCH v3 08/14] cgroup: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Johannes Weiner, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Tejun Heo <tj@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Waiman Long <longman@redhat.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
kernel/cgroup/cgroup.c | 2 --
kernel/cgroup/debug.c | 4 ----
2 files changed, 6 deletions(-)
diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 312c6a8b55bb..db9e00a559df 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -2944,14 +2944,12 @@ int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader,
/* look up all src csets */
spin_lock_irq(&css_set_lock);
- rcu_read_lock();
task = leader;
do {
cgroup_migrate_add_src(task_css_set(task), dst_cgrp, &mgctx);
if (!threadgroup)
break;
} while_each_thread(leader, task);
- rcu_read_unlock();
spin_unlock_irq(&css_set_lock);
/* prepare dst csets and commit */
diff --git a/kernel/cgroup/debug.c b/kernel/cgroup/debug.c
index 80aa3f027ac3..81ea38dd6f9d 100644
--- a/kernel/cgroup/debug.c
+++ b/kernel/cgroup/debug.c
@@ -49,7 +49,6 @@ static int current_css_set_read(struct seq_file *seq, void *v)
return -ENODEV;
spin_lock_irq(&css_set_lock);
- rcu_read_lock();
cset = task_css_set(current);
refcnt = refcount_read(&cset->refcount);
seq_printf(seq, "css_set %pK %d", cset, refcnt);
@@ -67,7 +66,6 @@ static int current_css_set_read(struct seq_file *seq, void *v)
seq_printf(seq, "%2d: %-4s\t- %p[%d]\n", ss->id, ss->name,
css, css->id);
}
- rcu_read_unlock();
spin_unlock_irq(&css_set_lock);
cgroup_kn_unlock(of->kn);
return 0;
@@ -95,7 +93,6 @@ static int current_css_set_cg_links_read(struct seq_file *seq, void *v)
return -ENOMEM;
spin_lock_irq(&css_set_lock);
- rcu_read_lock();
cset = task_css_set(current);
list_for_each_entry(link, &cset->cgrp_links, cgrp_link) {
struct cgroup *c = link->cgrp;
@@ -104,7 +101,6 @@ static int current_css_set_cg_links_read(struct seq_file *seq, void *v)
seq_printf(seq, "Root %d group %s\n",
c->root->hierarchy_id, name_buf);
}
- rcu_read_unlock();
spin_unlock_irq(&css_set_lock);
kfree(name_buf);
return 0;
--
2.34.1
^ permalink raw reply related
* [PATCH v3 07/14] yama: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Paul Moore, James Morris, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Kees Cook <kees@kernel.org>
Cc: Paul Moore <paul@paul-moore.com>
Cc: James Morris <jmorris@namei.org>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
security/yama/yama_lsm.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/security/yama/yama_lsm.c b/security/yama/yama_lsm.c
index 3d064dd4e03f..60d38deb181b 100644
--- a/security/yama/yama_lsm.c
+++ b/security/yama/yama_lsm.c
@@ -117,14 +117,12 @@ static void yama_relation_cleanup(struct work_struct *work)
struct ptrace_relation *relation;
spin_lock(&ptracer_relations_lock);
- rcu_read_lock();
list_for_each_entry_rcu(relation, &ptracer_relations, node) {
if (relation->invalid) {
list_del_rcu(&relation->node);
kfree_rcu(relation, rcu);
}
}
- rcu_read_unlock();
spin_unlock(&ptracer_relations_lock);
}
@@ -152,7 +150,6 @@ static int yama_ptracer_add(struct task_struct *tracer,
added->invalid = false;
spin_lock(&ptracer_relations_lock);
- rcu_read_lock();
list_for_each_entry_rcu(relation, &ptracer_relations, node) {
if (relation->invalid)
continue;
@@ -166,7 +163,6 @@ static int yama_ptracer_add(struct task_struct *tracer,
list_add_rcu(&added->node, &ptracer_relations);
out:
- rcu_read_unlock();
spin_unlock(&ptracer_relations_lock);
return 0;
}
--
2.34.1
^ permalink raw reply related
* [PATCH v3 06/14] ipc: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
ipc/msg.c | 1 -
ipc/sem.c | 1 -
ipc/shm.c | 1 -
ipc/util.c | 2 --
4 files changed, 5 deletions(-)
diff --git a/ipc/msg.c b/ipc/msg.c
index ee6af4fe52bf..1e579b57023f 100644
--- a/ipc/msg.c
+++ b/ipc/msg.c
@@ -179,7 +179,6 @@ static int newque(struct ipc_namespace *ns, struct ipc_params *params)
}
ipc_unlock_object(&msq->q_perm);
- rcu_read_unlock();
return msq->q_perm.id;
}
diff --git a/ipc/sem.c b/ipc/sem.c
index a39cdc7bf88f..38ad57b2b558 100644
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -579,7 +579,6 @@ static int newary(struct ipc_namespace *ns, struct ipc_params *params)
ns->used_sems += nsems;
sem_unlock(sma, -1);
- rcu_read_unlock();
return sma->sem_perm.id;
}
diff --git a/ipc/shm.c b/ipc/shm.c
index a9310b6dbbc3..61fae1b6a18e 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -795,7 +795,6 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
error = shp->shm_perm.id;
ipc_unlock_object(&shp->shm_perm);
- rcu_read_unlock();
return error;
no_id:
diff --git a/ipc/util.c b/ipc/util.c
index cae60f11d9c2..1be691b5dcad 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -293,7 +293,6 @@ int ipc_addid(struct ipc_ids *ids, struct kern_ipc_perm *new, int limit)
idr_preload(GFP_KERNEL);
spin_lock_init(&new->lock);
- rcu_read_lock();
spin_lock(&new->lock);
current_euid_egid(&euid, &egid);
@@ -316,7 +315,6 @@ int ipc_addid(struct ipc_ids *ids, struct kern_ipc_perm *new, int limit)
if (idx < 0) {
new->deleted = true;
spin_unlock(&new->lock);
- rcu_read_unlock();
return idx;
}
--
2.34.1
^ permalink raw reply related
* [PATCH v3 05/14] s390/pkey: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Holger Dengler, Vasily Gorbik,
Alexander Gordeev, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Harald Freudenberger <freude@linux.ibm.com>
Cc: Holger Dengler <dengler@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
drivers/s390/crypto/pkey_base.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/s390/crypto/pkey_base.c b/drivers/s390/crypto/pkey_base.c
index b15741461a63..4c4a9feecccc 100644
--- a/drivers/s390/crypto/pkey_base.c
+++ b/drivers/s390/crypto/pkey_base.c
@@ -48,16 +48,13 @@ int pkey_handler_register(struct pkey_handler *handler)
spin_lock(&handler_list_write_lock);
- rcu_read_lock();
list_for_each_entry_rcu(h, &handler_list, list) {
if (h == handler) {
- rcu_read_unlock();
spin_unlock(&handler_list_write_lock);
module_put(handler->module);
return -EEXIST;
}
}
- rcu_read_unlock();
list_add_rcu(&handler->list, &handler_list);
spin_unlock(&handler_list_write_lock);
--
2.34.1
^ permalink raw reply related
* [PATCH v3 04/14] nfs: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Anna Schumaker, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Trond Myklebust <trondmy@kernel.org>
Cc: Anna Schumaker <anna@kernel.org>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
fs/nfs/callback_proc.c | 2 --
fs/nfs/nfs4state.c | 2 --
fs/nfs/pnfs.c | 12 +-----------
fs/nfs/pnfs_dev.c | 4 ----
4 files changed, 1 insertion(+), 19 deletions(-)
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 8397c43358bd..16144db39335 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -721,7 +721,6 @@ __be32 nfs4_callback_offload(void *data, void *dummy,
return cpu_to_be32(NFS4ERR_DELAY);
spin_lock(&cps->clp->cl_lock);
- rcu_read_lock();
list_for_each_entry_rcu(server, &cps->clp->cl_superblocks,
client_link) {
list_for_each_entry(tmp_copy, &server->ss_copies, copies) {
@@ -736,7 +735,6 @@ __be32 nfs4_callback_offload(void *data, void *dummy,
}
}
out:
- rcu_read_unlock();
if (!found) {
memcpy(©->stateid, &args->coa_stateid, NFS4_STATEID_SIZE);
nfs4_copy_cb_args(copy, args);
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 7612e977e80b..598229fc07ed 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -241,13 +241,11 @@ const struct cred *nfs4_get_renew_cred(struct nfs_client *clp)
goto out;
spin_lock(&clp->cl_lock);
- rcu_read_lock();
list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
cred = nfs4_get_renew_cred_server_locked(server);
if (cred != NULL)
break;
}
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
out:
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index a3135b5af7ee..650f86fa144a 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -838,13 +838,12 @@ pnfs_layout_add_bulk_destroy_list(struct inode *inode,
return ret;
}
-/* Caller must hold rcu_read_lock and clp->cl_lock */
+/* Caller must hold clp->cl_lock (implies rcu_read_lock) */
static int
pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
struct nfs_server *server,
struct list_head *layout_list)
__must_hold(&clp->cl_lock)
- __must_hold(RCU)
{
struct pnfs_layout_hdr *lo, *next;
struct inode *inode;
@@ -862,16 +861,13 @@ pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
if (pnfs_layout_add_bulk_destroy_list(inode,
layout_list))
continue;
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
iput(inode);
} else {
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
}
nfs_sb_deactive(server->super);
spin_lock(&clp->cl_lock);
- rcu_read_lock();
return -EAGAIN;
}
return 0;
@@ -922,7 +918,6 @@ int pnfs_layout_destroy_byfsid(struct nfs_client *clp, struct nfs_fsid *fsid,
LIST_HEAD(layout_list);
spin_lock(&clp->cl_lock);
- rcu_read_lock();
restart:
list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
if (memcmp(&server->fsid, fsid, sizeof(*fsid)) != 0)
@@ -932,7 +927,6 @@ int pnfs_layout_destroy_byfsid(struct nfs_client *clp, struct nfs_fsid *fsid,
&layout_list) != 0)
goto restart;
}
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
return pnfs_layout_free_bulk_destroy_list(&layout_list, mode);
@@ -944,14 +938,12 @@ static void pnfs_layout_build_destroy_list_byclient(struct nfs_client *clp,
struct nfs_server *server;
spin_lock(&clp->cl_lock);
- rcu_read_lock();
restart:
list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
if (pnfs_layout_bulk_destroy_byserver_locked(clp, server,
list) != 0)
goto restart;
}
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
}
@@ -990,7 +982,6 @@ static void pnfs_layout_build_recover_list_byclient(struct nfs_client *clp,
struct nfs_server *server;
spin_lock(&clp->cl_lock);
- rcu_read_lock();
restart:
list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
if (!(server->caps & NFS_CAP_REBOOT_LAYOUTRETURN))
@@ -999,7 +990,6 @@ static void pnfs_layout_build_recover_list_byclient(struct nfs_client *clp,
list) != 0)
goto restart;
}
- rcu_read_unlock();
spin_unlock(&clp->cl_lock);
}
diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index bf0f2d67e96c..d19752ec1a95 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -231,9 +231,7 @@ nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *ld,
struct nfs4_deviceid_node *d;
spin_lock(&nfs4_deviceid_lock);
- rcu_read_lock();
d = _lookup_deviceid(ld, clp, id, nfs4_deviceid_hash(id));
- rcu_read_unlock();
if (!d) {
spin_unlock(&nfs4_deviceid_lock);
return;
@@ -331,14 +329,12 @@ _deviceid_purge_client(const struct nfs_client *clp, long hash)
HLIST_HEAD(tmp);
spin_lock(&nfs4_deviceid_lock);
- rcu_read_lock();
hlist_for_each_entry_rcu(d, &nfs4_deviceid_cache[hash], node)
if (d->nfs_client == clp && atomic_read(&d->ref)) {
hlist_del_init_rcu(&d->node);
hlist_add_head(&d->tmpnode, &tmp);
clear_bit(NFS_DEVICEID_NOCACHE, &d->flags);
}
- rcu_read_unlock();
spin_unlock(&nfs4_deviceid_lock);
if (hlist_empty(&tmp))
--
2.34.1
^ permalink raw reply related
* [PATCH v3 03/14] fs: aio: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
fs/aio.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/fs/aio.c b/fs/aio.c
index 7fc7b6221312..e3f9a5a391b5 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -359,15 +359,14 @@ static int aio_ring_mremap(struct vm_area_struct *vma)
int i, res = -EINVAL;
spin_lock(&mm->ioctx_lock);
- rcu_read_lock();
- table = rcu_dereference(mm->ioctx_table);
+ table = rcu_dereference_check(mm->ioctx_table, lockdep_is_held(&mm->ioctx_lock));
if (!table)
goto out_unlock;
for (i = 0; i < table->nr; i++) {
struct kioctx *ctx;
- ctx = rcu_dereference(table->table[i]);
+ ctx = rcu_dereference_check(table->table[i], lockdep_is_held(&mm->ioctx_lock));
if (ctx && ctx->aio_ring_file == file) {
if (!atomic_read(&ctx->dead)) {
ctx->user_id = ctx->mmap_base = vma->vm_start;
@@ -378,7 +377,6 @@ static int aio_ring_mremap(struct vm_area_struct *vma)
}
out_unlock:
- rcu_read_unlock();
spin_unlock(&mm->ioctx_lock);
return res;
}
--
2.34.1
^ permalink raw reply related
* [PATCH v3 02/14] drm/i915/gt: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Joonas Lahtinen, Tvrtko Ursulin,
pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tursulin@ursulin.net>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
drivers/gpu/drm/i915/gt/intel_ring_submission.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/gpu/drm/i915/gt/intel_ring_submission.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
index 2a6d79abf25b..bf73166a1337 100644
--- a/drivers/gpu/drm/i915/gt/intel_ring_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
@@ -392,14 +392,12 @@ static void reset_rewind(struct intel_engine_cs *engine, bool stalled)
rq = NULL;
spin_lock_irqsave(&engine->sched_engine->lock, flags);
- rcu_read_lock();
list_for_each_entry(pos, &engine->sched_engine->requests, sched.link) {
if (!__i915_request_is_complete(pos)) {
rq = pos;
break;
}
}
- rcu_read_unlock();
/*
* The guilty request will get skipped on a hung engine.
--
2.34.1
^ permalink raw reply related
* [PATCH v3 01/14] ACPI: APEI: Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin, Rafael J. Wysocki, Hanjun Guo, pengdonglin
In-Reply-To: <20250916044735.2316171-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
---
drivers/acpi/apei/ghes.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index a0d54993edb3..97ee19f2cae0 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -1207,12 +1207,10 @@ static int ghes_notify_hed(struct notifier_block *this, unsigned long event,
int ret = NOTIFY_DONE;
spin_lock_irqsave(&ghes_notify_lock_irq, flags);
- rcu_read_lock();
list_for_each_entry_rcu(ghes, &ghes_hed, list) {
if (!ghes_proc(ghes))
ret = NOTIFY_OK;
}
- rcu_read_unlock();
spin_unlock_irqrestore(&ghes_notify_lock_irq, flags);
return ret;
--
2.34.1
^ permalink raw reply related
* [PATCH v3 00/14] Remove redundant rcu_read_lock/unlock() in spin_lock
From: pengdonglin @ 2025-09-16 4:47 UTC (permalink / raw)
To: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees
Cc: bigeasy, hdanton, paulmck, linux-kernel, linux-rt-devel,
linux-nfs, linux-aio, linux-fsdevel, linux-security-module,
netdev, intel-gfx, linux-wireless, linux-acpi, linux-s390,
cgroups, pengdonglin
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side function definitions")
there is no difference between rcu_read_lock(), rcu_read_lock_bh() and
rcu_read_lock_sched() in terms of RCU read section and the relevant grace
period. That means that spin_lock(), which implies rcu_read_lock_sched(),
also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has already
been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
[1] https://elixir.bootlin.com/linux/v6.17-rc5/source/Documentation/RCU/rcu_dereference.rst#L407
[2] https://lore.kernel.org/lkml/20180829222021.GA29944@linux.vnet.ibm.com/
[3] https://lwn.net/Articles/777036/
[4] https://lore.kernel.org/lkml/6435833a-bdcb-4114-b29d-28b7f436d47d@paulmck-laptop/
pengdonglin (14):
ACPI: APEI: Remove redundant rcu_read_lock/unlock() in spin_lock
drm/i915/gt: Remove redundant rcu_read_lock/unlock() in spin_lock
fs: aio: Remove redundant rcu_read_lock/unlock() in spin_lock
nfs: Remove redundant rcu_read_lock/unlock() in spin_lock
s390/pkey: Remove redundant rcu_read_lock/unlock() in spin_lock
ipc: Remove redundant rcu_read_lock/unlock() in spin_lock
yama: Remove redundant rcu_read_lock/unlock() in spin_lock
cgroup: Remove redundant rcu_read_lock/unlock() in spin_lock
cgroup/cpuset: Remove redundant rcu_read_lock/unlock() in spin_lock
wifi: mac80211: Remove redundant rcu_read_lock/unlock() in spin_lock
net: ncsi: Remove redundant rcu_read_lock/unlock() in spin_lock
net: amt: Remove redundant rcu_read_lock/unlock() in spin_lock
net: bonding: Remove redundant rcu_read_lock/unlock() in spin_lock
wifi: ath9k: Remove redundant rcu_read_lock/unlock() in spin_lock
drivers/acpi/apei/ghes.c | 2 --
drivers/gpu/drm/i915/gt/intel_ring_submission.c | 2 --
drivers/net/amt.c | 8 --------
drivers/net/bonding/bond_3ad.c | 2 --
drivers/net/wireless/ath/ath9k/xmit.c | 2 --
drivers/s390/crypto/pkey_base.c | 3 ---
fs/aio.c | 6 ++----
fs/nfs/callback_proc.c | 2 --
fs/nfs/nfs4state.c | 2 --
fs/nfs/pnfs.c | 12 +-----------
fs/nfs/pnfs_dev.c | 4 ----
ipc/msg.c | 1 -
ipc/sem.c | 1 -
ipc/shm.c | 1 -
ipc/util.c | 2 --
kernel/cgroup/cgroup.c | 2 --
kernel/cgroup/cpuset.c | 6 ------
kernel/cgroup/debug.c | 4 ----
net/mac80211/cfg.c | 2 --
net/mac80211/debugfs.c | 2 --
net/mac80211/debugfs_netdev.c | 2 --
net/mac80211/debugfs_sta.c | 2 --
net/mac80211/sta_info.c | 2 --
net/ncsi/ncsi-manage.c | 2 --
security/yama/yama_lsm.c | 4 ----
25 files changed, 3 insertions(+), 75 deletions(-)
--
2.34.1
^ permalink raw reply
* Re: [PATCH v2] rcu: Remove redundant rcu_read_lock/unlock() in spin_lock critical sections
From: Donglin Peng @ 2025-09-16 1:55 UTC (permalink / raw)
To: Sebastian Andrzej Siewior
Cc: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees, linux-kernel, linux-rt-devel, linux-nfs, linux-aio,
linux-fsdevel, linux-security-module, netdev, intel-gfx,
linux-acpi, linux-s390, cgroups, Hillf Danton, Paul E . McKenney,
pengdonglin
In-Reply-To: <20250915144052.VHYlgilw@linutronix.de>
On Mon, Sep 15, 2025 at 10:40 PM Sebastian Andrzej Siewior
<bigeasy@linutronix.de> wrote:
>
> On 2025-09-15 21:47:29 [+0800], pengdonglin wrote:
> > From: pengdonglin <pengdonglin@xiaomi.com>
> >
> > Per Documentation/RCU/rcu_dereference.rst [1], since Linux 4.20's RCU
> > consolidation [2][3], RCU read-side critical sections include:
> > - Explicit rcu_read_lock()
> > - BH/interrupt/preemption-disabling regions
> > - Spinlock critical sections (including CONFIG_PREEMPT_RT kernels [4])
> >
> > Thus, explicit rcu_read_lock()/unlock() calls within spin_lock*() regions are redundant.
> > This patch removes them, simplifying locking semantics while preserving RCU protection.
> >
> > [1] https://elixir.bootlin.com/linux/v6.17-rc5/source/Documentation/RCU/rcu_dereference.rst#L407
> > [2] https://lore.kernel.org/lkml/20180829222021.GA29944@linux.vnet.ibm.com/
> > [3] https://lwn.net/Articles/777036/
> > [4] https://lore.kernel.org/lkml/6435833a-bdcb-4114-b29d-28b7f436d47d@paulmck-laptop/
>
> What about something like this:
>
> Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side
> function definitions") there is no difference between rcu_read_lock(),
> rcu_read_lock_bh() and rcu_read_lock_sched() in terms of RCU read
> section and the relevant grace period. That means that spin_lock(),
> which implies rcu_read_lock_sched(), also implies rcu_read_lock().
>
> There is no need no explicitly start a RCU read section if one has
> already been started implicitly by spin_lock().
>
> Simplify the code and remove the inner rcu_read_lock() invocation.
>
>
> The description above should make it clear what:
> - the intention is
> - the proposed solution to it and why it is correct.
Thanks, that's much clearer. I'll use this commit message in v3.
>
> You can't send a patch like this. You need to split it at the very least
> by subsystem. The networking bits need to follow to follow for instance
> Documentation/process/maintainer-netdev.rst
Thanks, I will split this into a series for v3.
>
> and so on.
>
> Sebastian
^ permalink raw reply
* Re: [PATCH v2 0/7] fs/9p: Reuse inode based on path (in addition to qid)
From: Dominique Martinet @ 2025-09-15 23:31 UTC (permalink / raw)
To: Tingmao Wang
Cc: Christian Schoenebeck, Mickaël Salaün,
Eric Van Hensbergen, Latchesar Ionkov, v9fs, Günther Noack,
linux-security-module, Jan Kara, Amir Goldstein,
Matthew Bobrowski, Al Viro, Christian Brauner, linux-fsdevel
In-Reply-To: <41d761e6-d826-4c29-b673-1fb2b0af77b9@maowtm.org>
(Thanks Christian, replying just once but your reply was helpful)
Tingmao Wang wrote on Mon, Sep 15, 2025 at 02:44:44PM +0100:
> > I'm fuzzy on the details but I don't see how inode pointers would be
> > stable for other filesystems as well, what prevents
> > e.g. vm.drop_caches=3 to drop these inodes on ext4?
>
> Landlock holds a reference to the inode in the ruleset, so they shouldn't
> be dropped. On security_sb_delete Landlock will iput those inodes so they
> won't cause issue with unmounting. There is some special mechanism
> ("landlock objects") to decouple the ruleset themselves from the actual
> inodes, so that previously Landlocked things can keep running even after
> the inode has disappeared as a result of unmounting.
Thank you for the explanation, that makes more sense.
iirc even in cacheless mode 9p should keep inode arounds if there's an
open fd somewhere
> > Although looking at the patches what 9p seems to need isn't a new stable
> > handle, but "just" not allocating new inodes in iget5...
> > This was attempted in 724a08450f74 ("fs/9p: simplify iget to remove
> > unnecessary paths"), but later reverted in be2ca3825372 ("Revert "fs/9p:
> > simplify iget to remove unnecessary paths"") because it broke too many
> > users, but if you're comfortable with a new mount option for the lookup
> > by path I think we could make a new option saying
> > "yes_my_server_has_unique_qids"... Which I assume would work for
> > landlock/fsnotify?
>
> I noticed that, but assumed that simply reverting them without additional
> work (such as tracking the string path) would be a no go given the reason
> why they are reverted.
Yes, just reverting and using that as default broke too much things, so
this is unfortunately not acceptable... And 9p has no "negotiation"
phase on mount to say "okay this is qemu with remap mode so we can do
that" to enable to disable the behaviour automatically; which has been
annoying in the past too.
I understand you'd prefer something that works by default.
> > I'm not sure how much effort we want to spend on 32bit: as far as I
> > know, if we have inode number collision on 32 bit we're already in
> > trouble (tools like tar will consider such files to be hardlink of each
> > other and happily skip reading data, producing corrupted archives);
> > this is not a happy state but I don't know how to do better in any
> > reasonable way, so we can probably keep a similar limitation for 32bit
> > and use inode number directly...
>
> I think if 9pfs export a handle it can be the full 64bit qid.path on any
> platform, right?
yes, file handle can be an arbitrary size.
> > I'm not familiar with the qid remap implementation in qemu, but I'm
> > curious in what case you hit that.
> > Deleting and recreating files? Or as you seem to say below the 'qid' is
> > "freed" when fd is closed qemu-side and re-used by later open of other
> > files?
>
> I tried mounting a qemu-exported 9pfs backed on ext4, with
> multidevs=remap, and created a file, used stat to note its inode number,
> deleted the file, created another file (of the same OR different name),
> and that new file will have the same inode number.
>
> (If I don't delete the file, then a newly created file would of course
> have a different ext4 inode number, and in that case QEMU exposes a
> different qid)
Ok so from Christian's reply this is just ext4 reusing the same inode..
I briefly hinted at this above, but in this case ext4 will give the
inode a different generation number (so the ext4 file handle will be
different, and accessing the old one will get ESTALE); but that's not
something qemu currently tracks and it'd be a bit of an overhaul...
In theory qemu could hash mount_id + file handle to get a properly
unique qid, if we need to improve that, but that'd be limited to root
users (and to filesystems that support name_to_handle_at) so I don't
think it's really appropriate either... hmm..
(I also thought of checking if nlink is 0 when getting a new inode, but
that's technically legimitate from /proc/x/fd opens so I don't think we
can do that either)
And then there's also all the servers that don't give unique qids at
all, so we'll just get weird landlock/fsnotify behaviours for them if we
go that way...
-----------------
Okay, you've convinced me something like path tracking seems more
appropriate; I'll just struggle one last time first with a few more open
questions:
- afaiu this (reusing inodes) work in cached mode because the dentry is
kept around; I don't understand the vfs well enough but could the inodes
hold its dentry and dentries hold their parent dentry alive somehow?
So in cacheless mode, if you have a tree like this:
a
└── b
├── c
└── d
with c 'open' (or a reference held by landlock), then dentries for a/b/c
would be kept, but d could be droppable?
My understanding is that in cacheless mode we're dropping dentries
aggressively so that things like readdir() are refreshed, but I'm
thinking this should work even if we keep some dentries alive when their
inode is held up.
- if that doesn't work (or is too complicated), I'm thinking tracking
path is probably better than qid-based filtering based on what we
discussed as it only affects uncached mode.. I'll need to spend some
time testing but I think we can move forward with the current patchset
rather than try something new.
Thanks!
--
Dominique Martinet | Asmadeus
^ permalink raw reply
* Re: [PATCH] ima: setting security.ima to fix security.evm for a file with IMA signature
From: Mimi Zohar @ 2025-09-15 21:05 UTC (permalink / raw)
To: Coiby Xu
Cc: linux-integrity, Roberto Sassu, Dmitry Kasatkin, Eric Snowberg,
Paul Moore, James Morris, Serge E. Hallyn,
open list:SECURITY SUBSYSTEM, open list
In-Reply-To: <r3himk4z2aiyqsjstlpnda4wafeo7i4oum3n2dbvnasmtep5ex@zqodcpjmyx5b>
On Wed, 2025-09-10 at 09:20 +0800, Coiby Xu wrote:
> On Tue, Sep 09, 2025 at 11:31:20AM -0400, Mimi Zohar wrote:
> > On Tue, 2025-09-09 at 12:19 +0800, Coiby Xu wrote:
> > > When both IMA and EVM fix modes are enabled, accessing a file with IMA
> > > signature won't cause security.evm to be fixed. But this doesn't happen
> > > to a file with correct IMA hash already set because accessing it will
> > > cause setting security.ima again which triggers fixing security.evm
> > > thanks to security_inode_post_setxattr->evm_update_evmxattr.
> > >
> > > Let's use the same mechanism to fix security.evm for a file with IMA
> > > signature.
> > >
> > > Signed-off-by: Coiby Xu <coxu@redhat.com>
> >
> > Agreed, re-writing the file signature stored as security.ima would force
> > security.evm to be updated.
> >
> > Unfortunately, I'm missing something. ima_appraise_measurement() first verifies
> > the existing security.evm xattr, before verifying the security.ima xattr. If
> > the EVM HMAC fails to verify, it immediately exits ima_appraise_measurement().
> > security.ima in this case is never verified.
> >
> > This patch seems to address the case where the existing security.evm is valid,
> > but the file signature stored in security.ima is invalid. (To get to the new
> > code, the "status" flag is not INTEGRITY_PASS.) Re-writing the same invalid
> > file signature would solve an invalid security.evm, but not an invalid IMA file
> > signature. What am I missing?
>
> Hi, Mimi,
>
> Thanks for raising the question! This patch is to address the case where
> IMA signature is already added but security.evm doesn't yet exist. So
> EVM HMAC fails to verify but there is no exiting
> ima_appraise_measurement immediately.
>
> And you are right that re-writing an invalid IMA file won't fix an
> invalid IMA file signature. And even when IMA signature is valid, the
> verification may fail because the key is missing from .ima keyring. This
> happens because we need to turn off secure boot to enable fix mode. As a
> result, CA keys won't be loaded into .machine keyring. Btw, if I'm not
> mistaken, current IMA code assumes we are not supposed to fix IMA file
> signature.
> >
> > > ---
> > > security/integrity/ima/ima_appraise.c | 27 +++++++++++++++++++++------
> > > 1 file changed, 21 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/security/integrity/ima/ima_appraise.c b/security/integrity/ima/ima_appraise.c
> > > index f435eff4667f..18c3907c5e44 100644
> > > --- a/security/integrity/ima/ima_appraise.c
> > > +++ b/security/integrity/ima/ima_appraise.c
> > > @@ -595,12 +595,27 @@ int ima_appraise_measurement(enum ima_hooks func, struct ima_iint_cache *iint,
> > > integrity_audit_msg(audit_msgno, inode, filename,
> > > op, cause, rc, 0);
> > > } else if (status != INTEGRITY_PASS) {
> > > - /* Fix mode, but don't replace file signatures. */
> > > - if ((ima_appraise & IMA_APPRAISE_FIX) && !try_modsig &&
> > > - (!xattr_value ||
> > > - xattr_value->type != EVM_IMA_XATTR_DIGSIG)) {
> > > - if (!ima_fix_xattr(dentry, iint))
> > > - status = INTEGRITY_PASS;
> > > + /*
> > > + * Fix mode, but don't replace file signatures.
> > > + *
> > > + * When EVM fix mode is also enabled, security.evm will be
> > > + * fixed automatically when security.ima is set because of
> > > + * security_inode_post_setxattr->evm_update_evmxattr.
> > > + */
> > > + if ((ima_appraise & IMA_APPRAISE_FIX) && !try_modsig) {
> > > + if (!xattr_value ||
> > > + xattr_value->type != EVM_IMA_XATTR_DIGSIG) {
> > > + if (ima_fix_xattr(dentry, iint))
> > > + status = INTEGRITY_PASS;
> > > + } else if (xattr_value->type == EVM_IMA_XATTR_DIGSIG &&
> > > + evm_revalidate_status(XATTR_NAME_IMA)) {
> > > + if (!__vfs_setxattr_noperm(&nop_mnt_idmap,
> > > + dentry,
> > > + XATTR_NAME_IMA,
> > > + xattr_value,
> > > + xattr_len, 0))
> > > + status = INTEGRITY_PASS;
> > > + }
> > > }
Instead of re-writing the IMA signature without a clear explanation, define a
new EVM function named evm_fix_hmac() and add a call here in IMA. Only in EVM
fix mode would evm_fix_hmac() update the EVM hmac.
} else if (status != INTEGRITY_PASS) {
/*
* IMA fix mode updates the IMA file hash, which triggers EVM
* to update security.evm. ....
*
* Similarly, trigger fixing EVM HMAC for IMA file signatures.
*/
if ((ima_appraise & IMA_APPRAISE_FIX) && !try_modsig) {
if (!xattr_value ||
xattr_value->type != EVM_IMA_XATTR_DIGSIG) {
if (ima_fix_xattr(dentry, iint))
status = INTEGRITY_PASS;
} else if (status == INTEGRITY_NOLABEL) {
evm_fix_hmac(dentry, XATTR_NAME_IMA, ....);
}
}
> > >
> > > /*
> > >
> > > base-commit: b320789d6883cc00ac78ce83bccbfe7ed58afcf0
> >
>
^ permalink raw reply
* Re: [PATCH v4 04/12] libbpf: Support exclusive map creation
From: Andrii Nakryiko @ 2025-09-15 19:49 UTC (permalink / raw)
To: KP Singh
Cc: bpf, linux-security-module, bboscaccy, paul, kys, ast, daniel,
andrii
In-Reply-To: <20250914215141.15144-5-kpsingh@kernel.org>
On Sun, Sep 14, 2025 at 2:52 PM KP Singh <kpsingh@kernel.org> wrote:
>
> Implement setters and getters that allow map to be registered as
> exclusive to the specified program. The registration should be done
> before the exclusive program is loaded.
>
> Signed-off-by: KP Singh <kpsingh@kernel.org>
> ---
> tools/lib/bpf/bpf.c | 4 ++-
> tools/lib/bpf/bpf.h | 5 ++-
> tools/lib/bpf/libbpf.c | 69 ++++++++++++++++++++++++++++++++++++++++
> tools/lib/bpf/libbpf.h | 22 +++++++++++++
> tools/lib/bpf/libbpf.map | 3 ++
> 5 files changed, 101 insertions(+), 2 deletions(-)
>
LGTM.
Acked-by: Andrii Nakryiko <andrii@kernel.org>
[...]
^ permalink raw reply
* Re: [PATCH] pid: use ns_capable_noaudit() when determining net sysctl permissions
From: Serge E. Hallyn @ 2025-09-15 17:13 UTC (permalink / raw)
To: Christian Brauner
Cc: Serge E. Hallyn, Paul Moore, Christian Göttsche,
linux-kernel, Christian Göttsche, linux-security-module,
selinux
In-Reply-To: <20250915-geldquelle-magie-eb9075a9764e@brauner>
On Mon, Sep 15, 2025 at 04:13:16PM +0200, Christian Brauner wrote:
> On Fri, Sep 12, 2025 at 02:35:45PM -0500, Serge Hallyn wrote:
> > On Thu, Sep 11, 2025 at 06:05:05PM -0500, Serge E. Hallyn wrote:
> > > On Thu, Sep 11, 2025 at 04:46:20PM -0400, Paul Moore wrote:
> > > > On Sep 10, 2025 =?UTF-8?q?Christian=20G=C3=B6ttsche?= <cgoettsche@seltendoof.de> wrote:
> > > > >
> > > > > The capability check should not be audited since it is only being used
> > > > > to determine the inode permissions. A failed check does not indicate a
> > > > > violation of security policy but, when an LSM is enabled, a denial audit
> > > > > message was being generated.
> > > > >
> > > > > The denial audit message can either lead to the capability being
> > > > > unnecessarily allowed in a security policy, or being silenced potentially
> > > > > masking a legitimate capability check at a later point in time.
> > > > >
> > > > > Similar to commit d6169b0206db ("net: Use ns_capable_noaudit() when
> > > > > determining net sysctl permissions")
> > > > >
> > > > > Fixes: 7863dcc72d0f ("pid: allow pid_max to be set per pid namespace")
> > > > > CC: Christian Brauner <brauner@kernel.org>
> > > > > CC: linux-security-module@vger.kernel.org
> > > > > CC: selinux@vger.kernel.org
> > > > > Signed-off-by: Christian Göttsche <cgzones@googlemail.com>
> > > > > ---
> > > > > kernel/pid.c | 2 +-
> > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > Reviewed-by: Paul Moore <paul@paul-moore.com>
> > >
> > > Thanks.
> > >
> > > Acked-by: Serge Hallyn <serge@hallyn.com>
> > >
> > > I'll queue this up in the capability tree, unless Christian wanted
> > > it in his.
> > >
> > > -serge
> >
> > I've included this in git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux.git#caps-next
>
> What's easier for you? Just want me to stuff this in the pidfs pile?
Yes, please go ahead, thanks.
^ permalink raw reply
* Re: Missing signoff in the capabilities-next tree
From: Serge E. Hallyn @ 2025-09-15 17:10 UTC (permalink / raw)
To: Mark Brown
Cc: Christian Göttsche, Serge Hallyn, Christian Brauner,
Paul Moore, linux-next, linux-kernel, selinux,
linux-security-module
In-Reply-To: <bb5c5c5f-fa31-4339-aa53-d06596351f86@sirena.org.uk>
On Mon, Sep 15, 2025 at 05:04:40PM +0100, Mark Brown wrote:
> Commit
>
> d87d3c530e0ce ("pid: use ns_capable_noaudit() when determining net sysctl permissions")
>
> in the capabilites-next tree is missing a Signed-off-by from its
> committer.
Hm, I had expected b4 shazam to do that for me, but I see that
that requires -s.
In any case, I think I will be dropping it from this tree, and
Christian will take it through his.
thanks,
-serge
^ permalink raw reply
* Smack: Memory leak in smack_fs_context_submount()
From: Casey Schaufler @ 2025-09-15 15:49 UTC (permalink / raw)
To: LSM List; +Cc: Casey Schaufler, David Howells, Jeff Layton, Christian Brauner
In-Reply-To: <5e8be941-626a-4d8f-82bf-e08f75152f20.ref@schaufler-ca.com>
The mount options for Smack should be pointers into the global Smack label
list. smack_fs_context_submount() allocates copies of the labels. These
will never be freed. Change the code to save the pointers instead.
Fixes: d80a8f1b58c2 (vfs, security: Fix automount superblock LSM init problem, preventing NFS sb sharing)
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
---
security/smack/smack_lsm.c | 31 +++++++++----------------------
1 file changed, 9 insertions(+), 22 deletions(-)
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index adf1c542d213..4c002e55d4f6 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -630,31 +630,18 @@ static int smack_fs_context_submount(struct fs_context *fc,
sbsp = smack_superblock(reference);
isp = smack_inode(reference->s_root->d_inode);
- if (sbsp->smk_default) {
- ctx->fsdefault = kstrdup(sbsp->smk_default->smk_known, GFP_KERNEL);
- if (!ctx->fsdefault)
- return -ENOMEM;
- }
+ if (sbsp->smk_default)
+ ctx->fsdefault = sbsp->smk_default->smk_known;
- if (sbsp->smk_floor) {
- ctx->fsfloor = kstrdup(sbsp->smk_floor->smk_known, GFP_KERNEL);
- if (!ctx->fsfloor)
- return -ENOMEM;
- }
+ if (sbsp->smk_floor)
+ ctx->fsfloor = sbsp->smk_floor->smk_known;
- if (sbsp->smk_hat) {
- ctx->fshat = kstrdup(sbsp->smk_hat->smk_known, GFP_KERNEL);
- if (!ctx->fshat)
- return -ENOMEM;
- }
+ if (sbsp->smk_hat)
+ ctx->fshat = sbsp->smk_hat->smk_known;
+
+ if ((isp->smk_flags & SMK_INODE_TRANSMUTE) && sbsp->smk_root)
+ ctx->fstransmute = sbsp->smk_root->smk_known;
- if (isp->smk_flags & SMK_INODE_TRANSMUTE) {
- if (sbsp->smk_root) {
- ctx->fstransmute = kstrdup(sbsp->smk_root->smk_known, GFP_KERNEL);
- if (!ctx->fstransmute)
- return -ENOMEM;
- }
- }
return 0;
}
^ permalink raw reply related
* Missing signoff in the capabilities-next tree
From: Mark Brown @ 2025-09-15 16:04 UTC (permalink / raw)
To: Christian Göttsche, Serge Hallyn, Christian Brauner,
Paul Moore
Cc: linux-next, linux-kernel, selinux, linux-security-module
[-- Attachment #1: Type: text/plain, Size: 175 bytes --]
Commit
d87d3c530e0ce ("pid: use ns_capable_noaudit() when determining net sysctl permissions")
in the capabilites-next tree is missing a Signed-off-by from its
committer.
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply
* Re: [PATCH v2 0/7] fs/9p: Reuse inode based on path (in addition to qid)
From: Christian Schoenebeck @ 2025-09-15 14:10 UTC (permalink / raw)
To: Tingmao Wang, Dominique Martinet
Cc: Mickaël Salaün, Eric Van Hensbergen, Latchesar Ionkov,
v9fs, Günther Noack, linux-security-module, Jan Kara,
Amir Goldstein, Matthew Bobrowski, Al Viro, Christian Brauner,
linux-fsdevel
In-Reply-To: <aMgMOnrAOrwQyVbp@codewreck.org>
On Monday, September 15, 2025 2:53:14 PM CEST Dominique Martinet wrote:
[...]
> > 1. The qid is 9pfs internal data, and we may need extra API for 9pfs to
> >
> > expose this to Landlock. On 64bit, this is easy as it's just the inode
> > number (offset by 2), which we can already get from the struct inode.
> > But perhaps on 32bit we need a way to expose the full 64bit server-sent
> > qid to Landlock (or other kernel subsystems), if we're going to do
> > this.
>
> I'm not sure how much effort we want to spend on 32bit: as far as I
> know, if we have inode number collision on 32 bit we're already in
> trouble (tools like tar will consider such files to be hardlink of each
> other and happily skip reading data, producing corrupted archives);
> this is not a happy state but I don't know how to do better in any
> reasonable way, so we can probably keep a similar limitation for 32bit
> and use inode number directly...
I agree, on 32-bit the game is lost.
One way that would come to my mind though: exposing the full qid path as xattr
on 32-bit, e.g. via "system.9pfs_qid" or something like that.
> > 2. Even though qids are supposed to be unique across the lifetime of a
> >
> > filesystem (including deleted files), this is not the case even for
> > QEMU in multidevs=remap mode, when running on ext4, as tested on QEMU
> > 10.1.0.
>
> I'm not familiar with the qid remap implementation in qemu, but I'm
> curious in what case you hit that.
> Deleting and recreating files? Or as you seem to say below the 'qid' is
> "freed" when fd is closed qemu-side and re-used by later open of other
> files?
The inode remap algorithm in QEMU's 9p server was designed to prevent inode
number collisions of equally numbered inodes of *different* *devices* on host,
exposed to guest via the same 9p mount (which appears as only one 9pfs device
on guest). Basis for this however is still the underlying filesystem's inode
number on host.
So yes, ext4 re-uses inode numbers of deleted files, and when that happens, a
new file appears with the same qid path as the previously deleted file with
QEMU.
/Christian
^ permalink raw reply
* Re: [PATCH] pid: use ns_capable_noaudit() when determining net sysctl permissions
From: Serge E. Hallyn @ 2025-09-15 14:50 UTC (permalink / raw)
To: Christian Brauner
Cc: Serge E. Hallyn, Paul Moore, Christian Göttsche,
linux-kernel, Christian Göttsche, linux-security-module,
selinux
In-Reply-To: <20250915-geldquelle-magie-eb9075a9764e@brauner>
On Mon, Sep 15, 2025 at 04:13:16PM +0200, Christian Brauner wrote:
> On Fri, Sep 12, 2025 at 02:35:45PM -0500, Serge Hallyn wrote:
> > On Thu, Sep 11, 2025 at 06:05:05PM -0500, Serge E. Hallyn wrote:
> > > On Thu, Sep 11, 2025 at 04:46:20PM -0400, Paul Moore wrote:
> > > > On Sep 10, 2025 =?UTF-8?q?Christian=20G=C3=B6ttsche?= <cgoettsche@seltendoof.de> wrote:
> > > > >
> > > > > The capability check should not be audited since it is only being used
> > > > > to determine the inode permissions. A failed check does not indicate a
> > > > > violation of security policy but, when an LSM is enabled, a denial audit
> > > > > message was being generated.
> > > > >
> > > > > The denial audit message can either lead to the capability being
> > > > > unnecessarily allowed in a security policy, or being silenced potentially
> > > > > masking a legitimate capability check at a later point in time.
> > > > >
> > > > > Similar to commit d6169b0206db ("net: Use ns_capable_noaudit() when
> > > > > determining net sysctl permissions")
> > > > >
> > > > > Fixes: 7863dcc72d0f ("pid: allow pid_max to be set per pid namespace")
> > > > > CC: Christian Brauner <brauner@kernel.org>
> > > > > CC: linux-security-module@vger.kernel.org
> > > > > CC: selinux@vger.kernel.org
> > > > > Signed-off-by: Christian Göttsche <cgzones@googlemail.com>
> > > > > ---
> > > > > kernel/pid.c | 2 +-
> > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > Reviewed-by: Paul Moore <paul@paul-moore.com>
> > >
> > > Thanks.
> > >
> > > Acked-by: Serge Hallyn <serge@hallyn.com>
> > >
> > > I'll queue this up in the capability tree, unless Christian wanted
> > > it in his.
> > >
> > > -serge
> >
> > I've included this in git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux.git#caps-next
>
> What's easier for you? Just want me to stuff this in the pidfs pile?
Oh, sure - I can drop it from my tree, since it's the only thing in there
right now.
^ permalink raw reply
* Re: [PATCH v2] rcu: Remove redundant rcu_read_lock/unlock() in spin_lock critical sections
From: Sebastian Andrzej Siewior @ 2025-09-15 14:40 UTC (permalink / raw)
To: pengdonglin
Cc: tj, tony.luck, jani.nikula, ap420073, jv, freude, bcrl, trondmy,
longman, kees, linux-kernel, linux-rt-devel, linux-nfs, linux-aio,
linux-fsdevel, linux-security-module, netdev, intel-gfx,
linux-acpi, linux-s390, cgroups, Hillf Danton, Paul E . McKenney,
pengdonglin
In-Reply-To: <20250915134729.1801557-1-dolinux.peng@gmail.com>
On 2025-09-15 21:47:29 [+0800], pengdonglin wrote:
> From: pengdonglin <pengdonglin@xiaomi.com>
>
> Per Documentation/RCU/rcu_dereference.rst [1], since Linux 4.20's RCU
> consolidation [2][3], RCU read-side critical sections include:
> - Explicit rcu_read_lock()
> - BH/interrupt/preemption-disabling regions
> - Spinlock critical sections (including CONFIG_PREEMPT_RT kernels [4])
>
> Thus, explicit rcu_read_lock()/unlock() calls within spin_lock*() regions are redundant.
> This patch removes them, simplifying locking semantics while preserving RCU protection.
>
> [1] https://elixir.bootlin.com/linux/v6.17-rc5/source/Documentation/RCU/rcu_dereference.rst#L407
> [2] https://lore.kernel.org/lkml/20180829222021.GA29944@linux.vnet.ibm.com/
> [3] https://lwn.net/Articles/777036/
> [4] https://lore.kernel.org/lkml/6435833a-bdcb-4114-b29d-28b7f436d47d@paulmck-laptop/
What about something like this:
Since commit a8bb74acd8efe ("rcu: Consolidate RCU-sched update-side
function definitions") there is no difference between rcu_read_lock(),
rcu_read_lock_bh() and rcu_read_lock_sched() in terms of RCU read
section and the relevant grace period. That means that spin_lock(),
which implies rcu_read_lock_sched(), also implies rcu_read_lock().
There is no need no explicitly start a RCU read section if one has
already been started implicitly by spin_lock().
Simplify the code and remove the inner rcu_read_lock() invocation.
The description above should make it clear what:
- the intention is
- the proposed solution to it and why it is correct.
You can't send a patch like this. You need to split it at the very least
by subsystem. The networking bits need to follow to follow for instance
Documentation/process/maintainer-netdev.rst
and so on.
Sebastian
^ permalink raw reply
* Re: [PATCH v2] rcu: Remove redundant rcu_read_lock/unlock() in spin_lock critical sections
From: Waiman Long @ 2025-09-15 14:13 UTC (permalink / raw)
To: pengdonglin, tj, tony.luck, jani.nikula, ap420073, jv, freude,
bcrl, trondmy, kees
Cc: bigeasy, linux-kernel, linux-rt-devel, linux-nfs, linux-aio,
linux-fsdevel, linux-security-module, netdev, intel-gfx,
linux-acpi, linux-s390, cgroups, Hillf Danton, Paul E . McKenney,
pengdonglin
In-Reply-To: <20250915134729.1801557-1-dolinux.peng@gmail.com>
On 9/15/25 9:47 AM, pengdonglin wrote:
> From: pengdonglin <pengdonglin@xiaomi.com>
>
> Per Documentation/RCU/rcu_dereference.rst [1], since Linux 4.20's RCU
> consolidation [2][3], RCU read-side critical sections include:
> - Explicit rcu_read_lock()
> - BH/interrupt/preemption-disabling regions
> - Spinlock critical sections (including CONFIG_PREEMPT_RT kernels [4])
>
> Thus, explicit rcu_read_lock()/unlock() calls within spin_lock*() regions are redundant.
> This patch removes them, simplifying locking semantics while preserving RCU protection.
>
> [1] https://elixir.bootlin.com/linux/v6.17-rc5/source/Documentation/RCU/rcu_dereference.rst#L407
> [2] https://lore.kernel.org/lkml/20180829222021.GA29944@linux.vnet.ibm.com/
> [3] https://lwn.net/Articles/777036/
> [4] https://lore.kernel.org/lkml/6435833a-bdcb-4114-b29d-28b7f436d47d@paulmck-laptop/
>
> Cc: Waiman Long <longman@redhat.com>
> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> Cc: Hillf Danton <hdanton@sina.com>
> Cc: Paul E. McKenney <paulmck@kernel.org>
> Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
> Signed-off-by: pengdonglin <dolinux.peng@gmail.com>
> ---
> Changes in v2:
> - Clarified commit message to prevent accidental backport to older kernels
> - Added lockdep_is_held() to avoid false positives
> ---
> drivers/acpi/apei/ghes.c | 2 --
> drivers/gpu/drm/i915/gt/intel_ring_submission.c | 2 --
> drivers/net/amt.c | 8 --------
> drivers/net/bonding/bond_3ad.c | 2 --
> drivers/net/wireless/ath/ath9k/xmit.c | 2 --
> drivers/s390/crypto/pkey_base.c | 3 ---
> fs/aio.c | 6 ++----
> fs/nfs/callback_proc.c | 2 --
> fs/nfs/nfs4state.c | 2 --
> fs/nfs/pnfs.c | 9 ---------
> fs/nfs/pnfs_dev.c | 4 ----
> ipc/msg.c | 1 -
> ipc/sem.c | 1 -
> ipc/shm.c | 1 -
> ipc/util.c | 2 --
> kernel/cgroup/cgroup.c | 2 --
> kernel/cgroup/cpuset.c | 6 ------
> kernel/cgroup/debug.c | 4 ----
> net/mac80211/cfg.c | 2 --
> net/mac80211/debugfs.c | 2 --
> net/mac80211/debugfs_netdev.c | 2 --
> net/mac80211/debugfs_sta.c | 2 --
> net/mac80211/sta_info.c | 2 --
> net/ncsi/ncsi-manage.c | 2 --
> security/yama/yama_lsm.c | 4 ----
> 25 files changed, 2 insertions(+), 73 deletions(-)
>
...
> diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
> index 27adb04df675..9b7e8e8e9411 100644
> --- a/kernel/cgroup/cpuset.c
> +++ b/kernel/cgroup/cpuset.c
> @@ -4073,7 +4073,6 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask)
> struct cpuset *cs;
>
> spin_lock_irqsave(&callback_lock, flags);
> - rcu_read_lock();
>
> cs = task_cs(tsk);
> if (cs != &top_cpuset)
> @@ -4095,7 +4094,6 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask)
> cpumask_copy(pmask, possible_mask);
> }
>
> - rcu_read_unlock();
> spin_unlock_irqrestore(&callback_lock, flags);
> }
>
> @@ -4168,9 +4166,7 @@ nodemask_t cpuset_mems_allowed(struct task_struct *tsk)
> unsigned long flags;
>
> spin_lock_irqsave(&callback_lock, flags);
> - rcu_read_lock();
> guarantee_online_mems(task_cs(tsk), &mask);
> - rcu_read_unlock();
> spin_unlock_irqrestore(&callback_lock, flags);
>
> return mask;
> @@ -4265,10 +4261,8 @@ bool cpuset_current_node_allowed(int node, gfp_t gfp_mask)
> /* Not hardwall and node outside mems_allowed: scan up cpusets */
> spin_lock_irqsave(&callback_lock, flags);
>
> - rcu_read_lock();
> cs = nearest_hardwall_ancestor(task_cs(current));
> allowed = node_isset(node, cs->mems_allowed);
> - rcu_read_unlock();
>
> spin_unlock_irqrestore(&callback_lock, flags);
> return allowed;
For cpuset,
Acked-by: Waiman Long <longman@redhat.com>
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox