[PATCH 1/3] taskstats-fork: Add a new taskstats command to get notification on fork/clone

From: Nikanth Karthikesan
Date: Tue Jul 21 2009 - 01:01:45 EST

Add a new taskstats command to register for notification, whenever a new task
forks in the cpumask specified.

Signed-off-by: Nikanth Karthikesan <knikanth@xxxxxxx>

---

diff --git a/include/linux/taskstats.h b/include/linux/taskstats.h
index 341dddb..1623a1f 100644
--- a/include/linux/taskstats.h
+++ b/include/linux/taskstats.h
@@ -188,6 +188,7 @@ enum {
TASKSTATS_TYPE_STATS, /* taskstats structure */
TASKSTATS_TYPE_AGGR_PID, /* contains pid + stats */
TASKSTATS_TYPE_AGGR_TGID, /* contains tgid + stats */
+ TASKSTATS_TYPE_PID_TGID, /* contains pid + tgid */
__TASKSTATS_TYPE_MAX,
};

@@ -199,6 +200,8 @@ enum {
TASKSTATS_CMD_ATTR_TGID,
TASKSTATS_CMD_ATTR_REGISTER_CPUMASK,
TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK,
+ TASKSTATS_CMD_ATTR_REGISTER_FORK_CPUMASK,
+ TASKSTATS_CMD_ATTR_DEREGISTER_FORK_CPUMASK,
__TASKSTATS_CMD_ATTR_MAX,
};

diff --git a/include/linux/taskstats_kern.h b/include/linux/taskstats_kern.h
index 7e9680f..b727370 100644
--- a/include/linux/taskstats_kern.h
+++ b/include/linux/taskstats_kern.h
@@ -26,9 +26,12 @@ static inline void taskstats_tgid_free(struct signal_struct *sig)
kmem_cache_free(taskstats_cache, sig->stats);
}

+extern void taskstats_fork(struct task_struct *);
extern void taskstats_exit(struct task_struct *, int group_dead);
extern void taskstats_init_early(void);
#else
+static inline void taskstats_fork(struct task_struct *tsk)
+{}
static inline void taskstats_exit(struct task_struct *tsk, int group_dead)
{}
static inline void taskstats_tgid_init(struct signal_struct *sig)
diff --git a/kernel/fork.c b/kernel/fork.c
index 467746b..8ed8c9f 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1447,6 +1447,7 @@ long do_fork(unsigned long clone_flags,
freezer_count();
tracehook_report_vfork_done(p, nr);
}
+ taskstats_fork(p);
} else {
nr = PTR_ERR(p);
}
diff --git a/kernel/taskstats.c b/kernel/taskstats.c
index 888adbc..8fcb7d3 100644
--- a/kernel/taskstats.c
+++ b/kernel/taskstats.c
@@ -51,7 +51,9 @@ __read_mostly = {
[TASKSTATS_CMD_ATTR_PID] = { .type = NLA_U32 },
[TASKSTATS_CMD_ATTR_TGID] = { .type = NLA_U32 },
[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK] = { .type = NLA_STRING },
- [TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },};
+ [TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },
+ [TASKSTATS_CMD_ATTR_REGISTER_FORK_CPUMASK] = { .type = NLA_STRING },
+ [TASKSTATS_CMD_ATTR_DEREGISTER_FORK_CPUMASK] = { .type = NLA_STRING },};

static struct nla_policy
cgroupstats_cmd_get_policy[CGROUPSTATS_CMD_ATTR_MAX+1] __read_mostly = {
@@ -68,7 +70,13 @@ struct listener_list {
struct rw_semaphore sem;
struct list_head list;
};
-static DEFINE_PER_CPU(struct listener_list, listener_array);
+static DEFINE_PER_CPU(struct listener_list, fork_listener_array);
+static DEFINE_PER_CPU(struct listener_list, exit_listener_array);
+
+enum forkexit {
+ FORK,
+ EXIT
+};

enum actions {
REGISTER,
@@ -290,7 +298,8 @@ ret:
return;
}

-static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
+static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd,
+ enum forkexit forkexit)
{
struct listener_list *listeners;
struct listener *s, *tmp;
@@ -309,7 +318,11 @@ static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
INIT_LIST_HEAD(&s->list);
s->valid = 1;

- listeners = &per_cpu(listener_array, cpu);
+ if (forkexit == FORK)
+ listeners = &per_cpu(fork_listener_array, cpu);
+ else /* forkexit == EXIT */
+ listeners = &per_cpu(exit_listener_array, cpu);
+
down_write(&listeners->sem);
list_add(&s->list, &listeners->list);
up_write(&listeners->sem);
@@ -320,7 +333,12 @@ static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
/* Deregister or cleanup */
cleanup:
for_each_cpu(cpu, mask) {
- listeners = &per_cpu(listener_array, cpu);
+
+ if (forkexit == FORK)
+ listeners = &per_cpu(fork_listener_array, cpu);
+ else /* forkexit == EXIT */
+ listeners = &per_cpu(exit_listener_array, cpu);
+
down_write(&listeners->sem);
list_for_each_entry_safe(s, tmp, &listeners->list, list) {
if (s->pid == pid) {
@@ -436,11 +454,27 @@ static int taskstats_user_cmd(struct sk_buff *skb, struct genl_info *info)
if (!alloc_cpumask_var(&mask, GFP_KERNEL))
return -ENOMEM;

+ rc = parse(info->attrs[TASKSTATS_CMD_ATTR_REGISTER_FORK_CPUMASK], mask);
+ if (rc < 0)
+ goto free_return_rc;
+ if (rc == 0) {
+ rc = add_del_listener(info->snd_pid, mask, REGISTER, FORK);
+ goto free_return_rc;
+ }
+
+ rc = parse(info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_FORK_CPUMASK], mask);
+ if (rc < 0)
+ goto free_return_rc;
+ if (rc == 0) {
+ rc = add_del_listener(info->snd_pid, mask, DEREGISTER, FORK);
+ goto free_return_rc;
+ }
+
rc = parse(info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK], mask);
if (rc < 0)
goto free_return_rc;
if (rc == 0) {
- rc = add_del_listener(info->snd_pid, mask, REGISTER);
+ rc = add_del_listener(info->snd_pid, mask, REGISTER, EXIT);
goto free_return_rc;
}

@@ -448,7 +482,7 @@ static int taskstats_user_cmd(struct sk_buff *skb, struct genl_info *info)
if (rc < 0)
goto free_return_rc;
if (rc == 0) {
- rc = add_del_listener(info->snd_pid, mask, DEREGISTER);
+ rc = add_del_listener(info->snd_pid, mask, DEREGISTER, EXIT);
free_return_rc:
free_cpumask_var(mask);
return rc;
@@ -517,6 +551,44 @@ ret:
return sig->stats;
}

+/* Send pid out on fork/clone */
+void taskstats_fork(struct task_struct *tsk)
+{
+ struct sk_buff *rep_skb;
+ size_t size;
+ struct listener_list *listeners;
+ struct nlattr *na;
+
+ if (!family_registered)
+ return;
+
+ size = nla_total_size(sizeof(u32)) + nla_total_size(sizeof(tsk->tgid))
+ + nla_total_size(0);
+
+ listeners = &__raw_get_cpu_var(fork_listener_array);
+ if (list_empty(&listeners->list))
+ return;
+
+ if (prepare_reply(NULL, TASKSTATS_CMD_NEW, &rep_skb, size) < 0)
+ return;
+
+ na = nla_nest_start(rep_skb, TASKSTATS_TYPE_PID_TGID);
+ if (!na)
+ goto err;
+ if (nla_put(rep_skb, TASKSTATS_TYPE_PID, sizeof(tsk->pid), &tsk->pid) < 0)
+ goto err;
+ if (nla_put(rep_skb, TASKSTATS_TYPE_TGID, sizeof(tsk->tgid), &tsk->tgid) < 0)
+ goto err;
+ nla_nest_end(rep_skb, na);
+
+ send_cpu_listeners(rep_skb, listeners);
+ return;
+
+err:
+ nlmsg_free(rep_skb);
+
+}
+
/* Send pid data out on exit */
void taskstats_exit(struct task_struct *tsk, int group_dead)
{
@@ -544,7 +616,7 @@ void taskstats_exit(struct task_struct *tsk, int group_dead)
fill_tgid_exit(tsk);
}

- listeners = &__raw_get_cpu_var(listener_array);
+ listeners = &__raw_get_cpu_var(exit_listener_array);
if (list_empty(&listeners->list))
return;

@@ -598,8 +670,10 @@ void __init taskstats_init_early(void)

taskstats_cache = KMEM_CACHE(taskstats, SLAB_PANIC);
for_each_possible_cpu(i) {
- INIT_LIST_HEAD(&(per_cpu(listener_array, i).list));
- init_rwsem(&(per_cpu(listener_array, i).sem));
+ INIT_LIST_HEAD(&(per_cpu(fork_listener_array, i).list));
+ init_rwsem(&(per_cpu(fork_listener_array, i).sem));
+ INIT_LIST_HEAD(&(per_cpu(exit_listener_array, i).list));
+ init_rwsem(&(per_cpu(exit_listener_array, i).sem));
}
}


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