Merge branch 'perf/urgent' into perf/core

Merge reason: pick up the latest fixes - they won't make v3.0.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

5 files changed, 135 insertions, 30 deletions

diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index c997f7371c65..a0e246e2cee3 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -1732,10 +1732,36 @@ static cycle_t		ftrace_update_time;
 static unsigned long    ftrace_update_cnt;
 unsigned long           ftrace_update_tot_cnt;
 
+static int ops_traces_mod(struct ftrace_ops *ops)
+{
+        struct ftrace_hash *hash;
+
+        hash = ops->filter_hash;
+        return !!(!hash || !hash->count);
+}
+
 static int ftrace_update_code(struct module *mod)
 {
         struct dyn_ftrace *p;
         cycle_t start, stop;
+        unsigned long ref = 0;
+
+        /*
+         * When adding a module, we need to check if tracers are
+         * currently enabled and if they are set to trace all functions.
+         * If they are, we need to enable the module functions as well
+         * as update the reference counts for those function records.
+         */
+        if (mod) {
+                struct ftrace_ops *ops;
+
+                for (ops = ftrace_ops_list;
+                     ops != &ftrace_list_end; ops = ops->next) {
+                        if (ops->flags & FTRACE_OPS_FL_ENABLED &&
+                            ops_traces_mod(ops))
+                                ref++;
+                }
+        }
 
         start = ftrace_now(raw_smp_processor_id());
         ftrace_update_cnt = 0;
@@ -1748,7 +1774,7 @@ static int ftrace_update_code(struct module *mod)
 
                 p = ftrace_new_addrs;
                 ftrace_new_addrs = p->newlist;
-                p->flags = 0L;
+                p->flags = ref;
 
                 /*
                  * Do the initial record conversion from mcount jump
@@ -1771,7 +1797,7 @@ static int ftrace_update_code(struct module *mod)
                  * conversion puts the module to the correct state, thus
                  * passing the ftrace_make_call check.
                  */
-                if (ftrace_start_up) {
+                if (ftrace_start_up && ref) {
                         int failed = __ftrace_replace_code(p, 1);
                         if (failed) {
                                 ftrace_bug(failed, p->ip);
@@ -2395,10 +2421,9 @@ ftrace_match_module_records(struct ftrace_hash *hash, char *buff, char *mod)
  */
 
 static int
-ftrace_mod_callback(char *func, char *cmd, char *param, int enable)
+ftrace_mod_callback(struct ftrace_hash *hash,
+                    char *func, char *cmd, char *param, int enable)
 {
-        struct ftrace_ops *ops = &global_ops;
-        struct ftrace_hash *hash;
         char *mod;
         int ret = -EINVAL;
 
@@ -2418,11 +2443,6 @@ ftrace_mod_callback(char *func, char *cmd, char *param, int enable)
         if (!strlen(mod))
                 return ret;
 
-        if (enable)
-                hash = ops->filter_hash;
-        else
-                hash = ops->notrace_hash;
-
         ret = ftrace_match_module_records(hash, func, mod);
         if (!ret)
                 ret = -EINVAL;
@@ -2748,7 +2768,7 @@ static int ftrace_process_regex(struct ftrace_hash *hash,
         mutex_lock(&ftrace_cmd_mutex);
         list_for_each_entry(p, &ftrace_commands, list) {
                 if (strcmp(p->name, command) == 0) {
-                        ret = p->func(func, command, next, enable);
+                        ret = p->func(hash, func, command, next, enable);
                         goto out_unlock;
                 }
         }
diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index a3e2db708072..30a94c26dcb3 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -687,6 +687,7 @@ struct event_subsystem {
         struct dentry           *entry;
         struct event_filter     *filter;
         int                     nr_events;
+        int                     ref_count;
 };
 
 #define FILTER_PRED_INVALID     ((unsigned short)-1)
diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
index 4d7e1498ae91..581876f9f387 100644
--- a/kernel/trace/trace_events.c
+++ b/kernel/trace/trace_events.c
@@ -244,6 +244,35 @@ static void ftrace_clear_events(void)
         mutex_unlock(&event_mutex);
 }
 
+static void __put_system(struct event_subsystem *system)
+{
+        struct event_filter *filter = system->filter;
+
+        WARN_ON_ONCE(system->ref_count == 0);
+        if (--system->ref_count)
+                return;
+
+        if (filter) {
+                kfree(filter->filter_string);
+                kfree(filter);
+        }
+        kfree(system->name);
+        kfree(system);
+}
+
+static void __get_system(struct event_subsystem *system)
+{
+        WARN_ON_ONCE(system->ref_count == 0);
+        system->ref_count++;
+}
+
+static void put_system(struct event_subsystem *system)
+{
+        mutex_lock(&event_mutex);
+        __put_system(system);
+        mutex_unlock(&event_mutex);
+}
+
 /*
  * __ftrace_set_clr_event(NULL, NULL, NULL, set) will set/unset all events.
  */
@@ -519,7 +548,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt,
                    loff_t *ppos)
 {
         const char set_to_char[4] = { '?', '0', '1', 'X' };
-        const char *system = filp->private_data;
+        struct event_subsystem *system = filp->private_data;
         struct ftrace_event_call *call;
         char buf[2];
         int set = 0;
@@ -530,7 +559,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt,
                 if (!call->name || !call->class || !call->class->reg)
                         continue;
 
-                if (system && strcmp(call->class->system, system) != 0)
+                if (system && strcmp(call->class->system, system->name) != 0)
                         continue;
 
                 /*
@@ -560,7 +589,8 @@ static ssize_t
 system_enable_write(struct file *filp, const char __user *ubuf, size_t cnt,
                     loff_t *ppos)
 {
-        const char *system = filp->private_data;
+        struct event_subsystem *system = filp->private_data;
+        const char *name = NULL;
         unsigned long val;
         ssize_t ret;
 
@@ -575,7 +605,14 @@ system_enable_write(struct file *filp, const char __user *ubuf, size_t cnt,
         if (val != 0 && val != 1)
                 return -EINVAL;
 
-        ret = __ftrace_set_clr_event(NULL, system, NULL, val);
+        /*
+         * Opening of "enable" adds a ref count to system,
+         * so the name is safe to use.
+         */
+        if (system)
+                name = system->name;
+
+        ret = __ftrace_set_clr_event(NULL, name, NULL, val);
         if (ret)
                 goto out;
 
@@ -808,6 +845,52 @@ event_filter_write(struct file *filp, const char __user *ubuf, size_t cnt,
         return cnt;
 }
 
+static LIST_HEAD(event_subsystems);
+
+static int subsystem_open(struct inode *inode, struct file *filp)
+{
+        struct event_subsystem *system = NULL;
+        int ret;
+
+        if (!inode->i_private)
+                goto skip_search;
+
+        /* Make sure the system still exists */
+        mutex_lock(&event_mutex);
+        list_for_each_entry(system, &event_subsystems, list) {
+                if (system == inode->i_private) {
+                        /* Don't open systems with no events */
+                        if (!system->nr_events) {
+                                system = NULL;
+                                break;
+                        }
+                        __get_system(system);
+                        break;
+                }
+        }
+        mutex_unlock(&event_mutex);
+
+        if (system != inode->i_private)
+                return -ENODEV;
+
+ skip_search:
+        ret = tracing_open_generic(inode, filp);
+        if (ret < 0 && system)
+                put_system(system);
+
+        return ret;
+}
+
+static int subsystem_release(struct inode *inode, struct file *file)
+{
+        struct event_subsystem *system = inode->i_private;
+
+        if (system)
+                put_system(system);
+
+        return 0;
+}
+
 static ssize_t
 subsystem_filter_read(struct file *filp, char __user *ubuf, size_t cnt,
                       loff_t *ppos)
@@ -945,17 +1028,19 @@ static const struct file_operations ftrace_event_filter_fops = {
 };
 
 static const struct file_operations ftrace_subsystem_filter_fops = {
-        .open = tracing_open_generic,
+        .open = subsystem_open,
         .read = subsystem_filter_read,
         .write = subsystem_filter_write,
         .llseek = default_llseek,
+        .release = subsystem_release,
 };
 
 static const struct file_operations ftrace_system_enable_fops = {
-        .open = tracing_open_generic,
+        .open = subsystem_open,
         .read = system_enable_read,
         .write = system_enable_write,
         .llseek = default_llseek,
+        .release = subsystem_release,
 };
 
 static const struct file_operations ftrace_show_header_fops = {
@@ -984,8 +1069,6 @@ static struct dentry *event_trace_events_dir(void)
         return d_events;
 }
 
-static LIST_HEAD(event_subsystems);
-
 static struct dentry *
 event_subsystem_dir(const char *name, struct dentry *d_events)
 {
@@ -995,6 +1078,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events)
         /* First see if we did not already create this dir */
         list_for_each_entry(system, &event_subsystems, list) {
                 if (strcmp(system->name, name) == 0) {
+                        __get_system(system);
                         system->nr_events++;
                         return system->entry;
                 }
@@ -1017,6 +1101,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events)
         }
 
         system->nr_events = 1;
+        system->ref_count = 1;
         system->name = kstrdup(name, GFP_KERNEL);
         if (!system->name) {
                 debugfs_remove(system->entry);
@@ -1044,8 +1129,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events)
                            "'%s/filter' entry\n", name);
         }
 
-        trace_create_file("enable", 0644, system->entry,
-                          (void *)system->name,
+        trace_create_file("enable", 0644, system->entry, system,
                           &ftrace_system_enable_fops);
 
         return system->entry;
@@ -1166,16 +1250,9 @@ static void remove_subsystem_dir(const char *name)
         list_for_each_entry(system, &event_subsystems, list) {
                 if (strcmp(system->name, name) == 0) {
                         if (!--system->nr_events) {
-                                struct event_filter *filter = system->filter;
-
                                 debugfs_remove_recursive(system->entry);
                                 list_del(&system->list);
-                                if (filter) {
-                                        kfree(filter->filter_string);
-                                        kfree(filter);
-                                }
-                                kfree(system->name);
-                                kfree(system);
+                                __put_system(system);
                         }
                         break;
                 }
diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c
index 8008ddcfbf20..256764ecccd6 100644
--- a/kernel/trace/trace_events_filter.c
+++ b/kernel/trace/trace_events_filter.c
@@ -1886,6 +1886,12 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 
         mutex_lock(&event_mutex);
 
+        /* Make sure the system still has events */
+        if (!system->nr_events) {
+                err = -ENODEV;
+                goto out_unlock;
+        }
+
         if (!strcmp(strstrip(filter_string), "0")) {
                 filter_free_subsystem_preds(system);
                 remove_filter_string(system->filter);
diff --git a/kernel/trace/trace_functions.c b/kernel/trace/trace_functions.c
index 8d0e1cc4e974..c7b0c6a7db09 100644
--- a/kernel/trace/trace_functions.c
+++ b/kernel/trace/trace_functions.c
@@ -324,7 +324,8 @@ ftrace_trace_onoff_unreg(char *glob, char *cmd, char *param)
 }
 
 static int
-ftrace_trace_onoff_callback(char *glob, char *cmd, char *param, int enable)
+ftrace_trace_onoff_callback(struct ftrace_hash *hash,
+                            char *glob, char *cmd, char *param, int enable)
 {
         struct ftrace_probe_ops *ops;
         void *count = (void *)-1;