tracing/function-return-tracer: store return stack into task_struct and allocate...
[cascardo/linux.git] / kernel / trace / ftrace.c
index 3940c71..90d99fb 100644 (file)
@@ -50,6 +50,9 @@ static int last_ftrace_enabled;
 /* Quick disabling of function tracer. */
 int function_trace_stop;
 
+/* By default, current tracing type is normal tracing. */
+enum ftrace_tracing_type_t ftrace_tracing_type = FTRACE_TYPE_ENTER;
+
 /*
  * ftrace_disabled is set when an anomaly is discovered.
  * ftrace_disabled is much stronger than ftrace_enabled.
@@ -358,9 +361,7 @@ static void print_ip_ins(const char *fmt, unsigned char *p)
                printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]);
 }
 
-static void ftrace_bug(int failed, unsigned long ip,
-                      unsigned char *expected,
-                      unsigned char *replace)
+static void ftrace_bug(int failed, unsigned long ip)
 {
        switch (failed) {
        case -EFAULT:
@@ -372,9 +373,7 @@ static void ftrace_bug(int failed, unsigned long ip,
                FTRACE_WARN_ON_ONCE(1);
                pr_info("ftrace failed to modify ");
                print_ip_sym(ip);
-               print_ip_ins(" expected: ", expected);
                print_ip_ins(" actual: ", (unsigned char *)ip);
-               print_ip_ins(" replace: ", replace);
                printk(KERN_CONT "\n");
                break;
        case -EPERM:
@@ -389,101 +388,91 @@ static void ftrace_bug(int failed, unsigned long ip,
        }
 }
 
-#define FTRACE_ADDR ((long)(ftrace_caller))
 
 static int
-__ftrace_replace_code(struct dyn_ftrace *rec,
-                     unsigned char *old, unsigned char *new, int enable)
+__ftrace_replace_code(struct dyn_ftrace *rec, int enable)
 {
        unsigned long ip, fl;
+       unsigned long ftrace_addr;
+
+#ifdef CONFIG_FUNCTION_RET_TRACER
+       if (ftrace_tracing_type == FTRACE_TYPE_ENTER)
+               ftrace_addr = (unsigned long)ftrace_caller;
+       else
+               ftrace_addr = (unsigned long)ftrace_return_caller;
+#else
+       ftrace_addr = (unsigned long)ftrace_caller;
+#endif
 
        ip = rec->ip;
 
-       if (ftrace_filtered && enable) {
+       /*
+        * If this record is not to be traced and
+        * it is not enabled then do nothing.
+        *
+        * If this record is not to be traced and
+        * it is enabled then disabled it.
+        *
+        */
+       if (rec->flags & FTRACE_FL_NOTRACE) {
+               if (rec->flags & FTRACE_FL_ENABLED)
+                       rec->flags &= ~FTRACE_FL_ENABLED;
+               else
+                       return 0;
+
+       } else if (ftrace_filtered && enable) {
                /*
-                * If filtering is on:
-                *
-                * If this record is set to be filtered and
-                * is enabled then do nothing.
-                *
-                * If this record is set to be filtered and
-                * it is not enabled, enable it.
-                *
-                * If this record is not set to be filtered
-                * and it is not enabled do nothing.
-                *
-                * If this record is set not to trace then
-                * do nothing.
-                *
-                * If this record is set not to trace and
-                * it is enabled then disable it.
-                *
-                * If this record is not set to be filtered and
-                * it is enabled, disable it.
+                * Filtering is on:
                 */
 
-               fl = rec->flags & (FTRACE_FL_FILTER | FTRACE_FL_NOTRACE |
-                                  FTRACE_FL_ENABLED);
+               fl = rec->flags & (FTRACE_FL_FILTER | FTRACE_FL_ENABLED);
 
-               if ((fl ==  (FTRACE_FL_FILTER | FTRACE_FL_ENABLED)) ||
-                   (fl ==  (FTRACE_FL_FILTER | FTRACE_FL_NOTRACE)) ||
-                   !fl || (fl == FTRACE_FL_NOTRACE))
+               /* Record is filtered and enabled, do nothing */
+               if (fl == (FTRACE_FL_FILTER | FTRACE_FL_ENABLED))
                        return 0;
 
-               /*
-                * If it is enabled disable it,
-                * otherwise enable it!
-                */
-               if (fl & FTRACE_FL_ENABLED) {
-                       /* swap new and old */
-                       new = old;
-                       old = ftrace_call_replace(ip, FTRACE_ADDR);
+               /* Record is not filtered and is not enabled do nothing */
+               if (!fl)
+                       return 0;
+
+               /* Record is not filtered but enabled, disable it */
+               if (fl == FTRACE_FL_ENABLED)
                        rec->flags &= ~FTRACE_FL_ENABLED;
-               } else {
-                       new = ftrace_call_replace(ip, FTRACE_ADDR);
+               else
+               /* Otherwise record is filtered but not enabled, enable it */
                        rec->flags |= FTRACE_FL_ENABLED;
-               }
        } else {
+               /* Disable or not filtered */
 
                if (enable) {
-                       /*
-                        * If this record is set not to trace and is
-                        * not enabled, do nothing.
-                        */
-                       fl = rec->flags & (FTRACE_FL_NOTRACE | FTRACE_FL_ENABLED);
-                       if (fl == FTRACE_FL_NOTRACE)
-                               return 0;
-
-                       new = ftrace_call_replace(ip, FTRACE_ADDR);
-               } else
-                       old = ftrace_call_replace(ip, FTRACE_ADDR);
-
-               if (enable) {
+                       /* if record is enabled, do nothing */
                        if (rec->flags & FTRACE_FL_ENABLED)
                                return 0;
+
                        rec->flags |= FTRACE_FL_ENABLED;
+
                } else {
+
+                       /* if record is not enabled do nothing */
                        if (!(rec->flags & FTRACE_FL_ENABLED))
                                return 0;
+
                        rec->flags &= ~FTRACE_FL_ENABLED;
                }
        }
 
-       return ftrace_modify_code(ip, old, new);
+       if (rec->flags & FTRACE_FL_ENABLED)
+               return ftrace_make_call(rec, ftrace_addr);
+       else
+               return ftrace_make_nop(NULL, rec, ftrace_addr);
 }
 
 static void ftrace_replace_code(int enable)
 {
        int i, failed;
-       unsigned char *new = NULL, *old = NULL;
        struct dyn_ftrace *rec;
        struct ftrace_page *pg;
 
-       if (enable)
-               old = ftrace_nop_replace();
-       else
-               new = ftrace_nop_replace();
-
        for (pg = ftrace_pages_start; pg; pg = pg->next) {
                for (i = 0; i < pg->index; i++) {
                        rec = &pg->records[i];
@@ -504,34 +493,30 @@ static void ftrace_replace_code(int enable)
                                unfreeze_record(rec);
                        }
 
-                       failed = __ftrace_replace_code(rec, old, new, enable);
+                       failed = __ftrace_replace_code(rec, enable);
                        if (failed && (rec->flags & FTRACE_FL_CONVERTED)) {
                                rec->flags |= FTRACE_FL_FAILED;
                                if ((system_state == SYSTEM_BOOTING) ||
                                    !core_kernel_text(rec->ip)) {
                                        ftrace_free_rec(rec);
                                } else
-                                       ftrace_bug(failed, rec->ip, old, new);
+                                       ftrace_bug(failed, rec->ip);
                        }
                }
        }
 }
 
 static int
-ftrace_code_disable(struct dyn_ftrace *rec)
+ftrace_code_disable(struct module *mod, struct dyn_ftrace *rec)
 {
        unsigned long ip;
-       unsigned char *nop, *call;
        int ret;
 
        ip = rec->ip;
 
-       nop = ftrace_nop_replace();
-       call = ftrace_call_replace(ip, mcount_addr);
-
-       ret = ftrace_modify_code(ip, call, nop);
+       ret = ftrace_make_nop(mod, rec, mcount_addr);
        if (ret) {
-               ftrace_bug(ret, ip, call, nop);
+               ftrace_bug(ret, ip);
                rec->flags |= FTRACE_FL_FAILED;
                return 0;
        }
@@ -571,8 +556,7 @@ static void ftrace_startup(void)
 
        mutex_lock(&ftrace_start_lock);
        ftrace_start_up++;
-       if (ftrace_start_up == 1)
-               command |= FTRACE_ENABLE_CALLS;
+       command |= FTRACE_ENABLE_CALLS;
 
        if (saved_ftrace_func != ftrace_trace_function) {
                saved_ftrace_func = ftrace_trace_function;
@@ -650,7 +634,7 @@ static cycle_t              ftrace_update_time;
 static unsigned long   ftrace_update_cnt;
 unsigned long          ftrace_update_tot_cnt;
 
-static int ftrace_update_code(void)
+static int ftrace_update_code(struct module *mod)
 {
        struct dyn_ftrace *p, *t;
        cycle_t start, stop;
@@ -667,7 +651,7 @@ static int ftrace_update_code(void)
                list_del_init(&p->list);
 
                /* convert record (i.e, patch mcount-call with NOP) */
-               if (ftrace_code_disable(p)) {
+               if (ftrace_code_disable(mod, p)) {
                        p->flags |= FTRACE_FL_CONVERTED;
                        ftrace_update_cnt++;
                } else
@@ -710,7 +694,7 @@ static int __init ftrace_dyn_table_alloc(unsigned long num_to_init)
 
        cnt = num_to_init / ENTRIES_PER_PAGE;
        pr_info("ftrace: allocating %ld entries in %d pages\n",
-               num_to_init, cnt);
+               num_to_init, cnt + 1);
 
        for (i = 0; i < cnt; i++) {
                pg->next = (void *)get_zeroed_page(GFP_KERNEL);
@@ -793,13 +777,11 @@ static void *t_start(struct seq_file *m, loff_t *pos)
        void *p = NULL;
        loff_t l = -1;
 
-       if (*pos != iter->pos) {
-               for (p = t_next(m, p, &l); p && l < *pos; p = t_next(m, p, &l))
-                       ;
-       } else {
-               l = *pos;
-               p = t_next(m, p, &l);
-       }
+       if (*pos > iter->pos)
+               *pos = iter->pos;
+
+       l = *pos;
+       p = t_next(m, p, &l);
 
        return p;
 }
@@ -810,15 +792,21 @@ static void t_stop(struct seq_file *m, void *p)
 
 static int t_show(struct seq_file *m, void *v)
 {
+       struct ftrace_iterator *iter = m->private;
        struct dyn_ftrace *rec = v;
        char str[KSYM_SYMBOL_LEN];
+       int ret = 0;
 
        if (!rec)
                return 0;
 
        kallsyms_lookup(rec->ip, NULL, NULL, NULL, str);
 
-       seq_printf(m, "%s\n", str);
+       ret = seq_printf(m, "%s\n", str);
+       if (ret < 0) {
+               iter->pos--;
+               iter->idx--;
+       }
 
        return 0;
 }
@@ -844,7 +832,7 @@ ftrace_avail_open(struct inode *inode, struct file *file)
                return -ENOMEM;
 
        iter->pg = ftrace_pages_start;
-       iter->pos = -1;
+       iter->pos = 0;
 
        ret = seq_open(file, &show_ftrace_seq_ops);
        if (!ret) {
@@ -931,7 +919,7 @@ ftrace_regex_open(struct inode *inode, struct file *file, int enable)
 
        if (file->f_mode & FMODE_READ) {
                iter->pg = ftrace_pages_start;
-               iter->pos = -1;
+               iter->pos = 0;
                iter->flags = enable ? FTRACE_ITER_FILTER :
                        FTRACE_ITER_NOTRACE;
 
@@ -1222,7 +1210,7 @@ ftrace_regex_release(struct inode *inode, struct file *file, int enable)
 
        mutex_lock(&ftrace_sysctl_lock);
        mutex_lock(&ftrace_start_lock);
-       if (iter->filtered && ftrace_start_up && ftrace_enabled)
+       if (ftrace_start_up && ftrace_enabled)
                ftrace_run_update_code(FTRACE_ENABLE_CALLS);
        mutex_unlock(&ftrace_start_lock);
        mutex_unlock(&ftrace_sysctl_lock);
@@ -1309,7 +1297,8 @@ static __init int ftrace_init_debugfs(void)
 
 fs_initcall(ftrace_init_debugfs);
 
-static int ftrace_convert_nops(unsigned long *start,
+static int ftrace_convert_nops(struct module *mod,
+                              unsigned long *start,
                               unsigned long *end)
 {
        unsigned long *p;
@@ -1320,23 +1309,32 @@ static int ftrace_convert_nops(unsigned long *start,
        p = start;
        while (p < end) {
                addr = ftrace_call_adjust(*p++);
+               /*
+                * Some architecture linkers will pad between
+                * the different mcount_loc sections of different
+                * object files to satisfy alignments.
+                * Skip any NULL pointers.
+                */
+               if (!addr)
+                       continue;
                ftrace_record_ip(addr);
        }
 
        /* disable interrupts to prevent kstop machine */
        local_irq_save(flags);
-       ftrace_update_code();
+       ftrace_update_code(mod);
        local_irq_restore(flags);
        mutex_unlock(&ftrace_start_lock);
 
        return 0;
 }
 
-void ftrace_init_module(unsigned long *start, unsigned long *end)
+void ftrace_init_module(struct module *mod,
+                       unsigned long *start, unsigned long *end)
 {
        if (ftrace_disabled || start == end)
                return;
-       ftrace_convert_nops(start, end);
+       ftrace_convert_nops(mod, start, end);
 }
 
 extern unsigned long __start_mcount_loc[];
@@ -1366,7 +1364,8 @@ void __init ftrace_init(void)
 
        last_ftrace_enabled = ftrace_enabled = 1;
 
-       ret = ftrace_convert_nops(__start_mcount_loc,
+       ret = ftrace_convert_nops(NULL,
+                                 __start_mcount_loc,
                                  __stop_mcount_loc);
 
        return;
@@ -1422,10 +1421,17 @@ int register_ftrace_function(struct ftrace_ops *ops)
                return -1;
 
        mutex_lock(&ftrace_sysctl_lock);
+
+       if (ftrace_tracing_type == FTRACE_TYPE_RETURN) {
+               ret = -EBUSY;
+               goto out;
+       }
+
        ret = __register_ftrace_function(ops);
        ftrace_startup();
-       mutex_unlock(&ftrace_sysctl_lock);
 
+out:
+       mutex_unlock(&ftrace_sysctl_lock);
        return ret;
 }
 
@@ -1491,16 +1497,139 @@ ftrace_enable_sysctl(struct ctl_table *table, int write,
 }
 
 #ifdef CONFIG_FUNCTION_RET_TRACER
+
+static atomic_t ftrace_retfunc_active;
+
+/* The callback that hooks the return of a function */
 trace_function_return_t ftrace_function_return =
                        (trace_function_return_t)ftrace_stub;
-void register_ftrace_return(trace_function_return_t func)
+
+
+/* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */
+static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list)
 {
+       int i;
+       int ret = 0;
+       unsigned long flags;
+       int start = 0, end = FTRACE_RETSTACK_ALLOC_SIZE;
+       struct task_struct *g, *t;
+
+       for (i = 0; i < FTRACE_RETSTACK_ALLOC_SIZE; i++) {
+               ret_stack_list[i] = kmalloc(FTRACE_RETFUNC_DEPTH
+                                       * sizeof(struct ftrace_ret_stack),
+                                       GFP_KERNEL);
+               if (!ret_stack_list[i]) {
+                       start = 0;
+                       end = i;
+                       ret = -ENOMEM;
+                       goto free;
+               }
+       }
+
+       read_lock_irqsave(&tasklist_lock, flags);
+       do_each_thread(g, t) {
+               if (start == end) {
+                       ret = -EAGAIN;
+                       goto unlock;
+               }
+
+               if (t->ret_stack == NULL) {
+                       t->ret_stack = ret_stack_list[start++];
+                       t->curr_ret_stack = -1;
+                       atomic_set(&t->trace_overrun, 0);
+               }
+       } while_each_thread(g, t);
+
+unlock:
+       read_unlock_irqrestore(&tasklist_lock, flags);
+free:
+       for (i = start; i < end; i++)
+               kfree(ret_stack_list[i]);
+       return ret;
+}
+
+/* Allocate a return stack for each task */
+static int start_return_tracing(void)
+{
+       struct ftrace_ret_stack **ret_stack_list;
+       int ret;
+
+       ret_stack_list = kmalloc(FTRACE_RETSTACK_ALLOC_SIZE *
+                               sizeof(struct ftrace_ret_stack *),
+                               GFP_KERNEL);
+
+       if (!ret_stack_list)
+               return -ENOMEM;
+
+       do {
+               ret = alloc_retstack_tasklist(ret_stack_list);
+       } while (ret == -EAGAIN);
+
+       kfree(ret_stack_list);
+       return ret;
+}
+
+int register_ftrace_return(trace_function_return_t func)
+{
+       int ret = 0;
+
+       mutex_lock(&ftrace_sysctl_lock);
+
+       /*
+        * Don't launch return tracing if normal function
+        * tracing is already running.
+        */
+       if (ftrace_trace_function != ftrace_stub) {
+               ret = -EBUSY;
+               goto out;
+       }
+       atomic_inc(&ftrace_retfunc_active);
+       ret = start_return_tracing();
+       if (ret) {
+               atomic_dec(&ftrace_retfunc_active);
+               goto out;
+       }
+       ftrace_tracing_type = FTRACE_TYPE_RETURN;
        ftrace_function_return = func;
+       ftrace_startup();
+
+out:
+       mutex_unlock(&ftrace_sysctl_lock);
+       return ret;
 }
 
 void unregister_ftrace_return(void)
 {
+       mutex_lock(&ftrace_sysctl_lock);
+
+       atomic_dec(&ftrace_retfunc_active);
        ftrace_function_return = (trace_function_return_t)ftrace_stub;
+       ftrace_shutdown();
+       /* Restore normal tracing type */
+       ftrace_tracing_type = FTRACE_TYPE_ENTER;
+
+       mutex_unlock(&ftrace_sysctl_lock);
+}
+
+/* Allocate a return stack for newly created task */
+void ftrace_retfunc_init_task(struct task_struct *t)
+{
+       if (atomic_read(&ftrace_retfunc_active)) {
+               t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH
+                               * sizeof(struct ftrace_ret_stack),
+                               GFP_KERNEL);
+               if (!t->ret_stack)
+                       return;
+               t->curr_ret_stack = -1;
+               atomic_set(&t->trace_overrun, 0);
+       } else
+               t->ret_stack = NULL;
+}
+
+void ftrace_retfunc_exit_task(struct task_struct *t)
+{
+       kfree(t->ret_stack);
+       t->ret_stack = NULL;
 }
 #endif