perf_counter tools: Prepare a small callchain framework

We plan to display the callchains depending on some user-configurable
parameters.

To gather the callchains stats from the recorded stream in a fast way,
this patch introduces an ad hoc radix tree adapted for callchains and also
a rbtree to sort these callchains once we have gathered every events
from the stream.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1246026481-8314-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index d3887ed51a640dba9193116e04cd23cb4eb2eeed..1c1296d8a64bd13f83991f7b4f24df5638a8d7ab 100644 (file)
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -329,6 +329,7 @@ LIB_OBJS += util/symbol.o
 LIB_OBJS += util/color.o
 LIB_OBJS += util/pager.o
 LIB_OBJS += util/header.o
+LIB_OBJS += util/callchain.o
 
 BUILTIN_OBJS += builtin-annotate.o
 BUILTIN_OBJS += builtin-help.o

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 681c2233f8827a2709caf1e578318467247d7c85..28d1cb2127e9ed3cd48ceeb3de56c19cb9e1ddf7 100644 (file)
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -62,11 +62,6 @@ struct ip_event {
        unsigned char __more_data[];
 };
 
-struct ip_callchain {
-       u64 nr;
-       u64 ips[0];
-};
-
 struct mmap_event {
        struct perf_event_header header;
        u32 pid, tid;

diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 16c84fd73c86bd2c89de7aa8a39fdd761e276466..a49842b69a59ce12d91588b3a66e3a5892049078 100644 (file)
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -66,4 +66,9 @@ sys_perf_counter_open(struct perf_counter_attr *attr,
 #define MAX_COUNTERS                   256
 #define MAX_NR_CPUS                    256
 
+struct ip_callchain {
+       u64 nr;
+       u64 ips[0];
+};
+
 #endif

diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
new file mode 100644 (file)
index 0000000..ad3c285
--- /dev/null
+++ b/tools/perf/util/callchain.c
@@ -0,0 +1,174 @@
+/*
+ * Copyright (C) 2009, Frederic Weisbecker <fweisbec@gmail.com>
+ *
+ * Handle the callchains from the stream in an ad-hoc radix tree and then
+ * sort them in an rbtree.
+ *
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <stdbool.h>
+#include <errno.h>
+
+#include "callchain.h"
+
+
+static void rb_insert_callchain(struct rb_root *root, struct callchain_node *chain)
+{
+       struct rb_node **p = &root->rb_node;
+       struct rb_node *parent = NULL;
+       struct callchain_node *rnode;
+
+       while (*p) {
+               parent = *p;
+               rnode = rb_entry(parent, struct callchain_node, rb_node);
+
+               if (rnode->hit < chain->hit)
+                       p = &(*p)->rb_left;
+               else
+                       p = &(*p)->rb_right;
+       }
+
+       rb_link_node(&chain->rb_node, parent, p);
+       rb_insert_color(&chain->rb_node, root);
+}
+
+/*
+ * Once we get every callchains from the stream, we can now
+ * sort them by hit
+ */
+void sort_chain_to_rbtree(struct rb_root *rb_root, struct callchain_node *node)
+{
+       struct callchain_node *child;
+
+       list_for_each_entry(child, &node->children, brothers)
+               sort_chain_to_rbtree(rb_root, child);
+
+       if (node->hit)
+               rb_insert_callchain(rb_root, node);
+}
+
+static struct callchain_node *create_child(struct callchain_node *parent)
+{
+       struct callchain_node *new;
+
+       new = malloc(sizeof(*new));
+       if (!new) {
+               perror("not enough memory to create child for code path tree");
+               return NULL;
+       }
+       new->parent = parent;
+       INIT_LIST_HEAD(&new->children);
+       INIT_LIST_HEAD(&new->val);
+       list_add_tail(&new->brothers, &parent->children);
+
+       return new;
+}
+
+static void
+fill_node(struct callchain_node *node, struct ip_callchain *chain, int start)
+{
+       int i;
+
+       for (i = start; i < chain->nr; i++) {
+               struct callchain_list *call;
+
+               call = malloc(sizeof(*chain));
+               if (!call) {
+                       perror("not enough memory for the code path tree");
+                       return;
+               }
+               call->ip = chain->ips[i];
+               list_add_tail(&call->list, &node->val);
+       }
+       node->val_nr = i - start;
+}
+
+static void add_child(struct callchain_node *parent, struct ip_callchain *chain)
+{
+       struct callchain_node *new;
+
+       new = create_child(parent);
+       fill_node(new, chain, parent->val_nr);
+
+       new->hit = 1;
+}
+
+static void
+split_add_child(struct callchain_node *parent, struct ip_callchain *chain,
+               struct callchain_list *to_split, int idx)
+{
+       struct callchain_node *new;
+
+       /* split */
+       new = create_child(parent);
+       list_move_tail(&to_split->list, &new->val);
+       new->hit = parent->hit;
+       parent->hit = 0;
+       parent->val_nr = idx;
+
+       /* create the new one */
+       add_child(parent, chain);
+}
+
+static int
+__append_chain(struct callchain_node *root, struct ip_callchain *chain,
+               int start);
+
+static int
+__append_chain_children(struct callchain_node *root, struct ip_callchain *chain)
+{
+       struct callchain_node *rnode;
+
+       /* lookup in childrens */
+       list_for_each_entry(rnode, &root->children, brothers) {
+               int ret = __append_chain(rnode, chain, root->val_nr);
+               if (!ret)
+                       return 0;
+       }
+       return -1;
+}
+
+static int
+__append_chain(struct callchain_node *root, struct ip_callchain *chain,
+               int start)
+{
+       struct callchain_list *cnode;
+       int i = start;
+       bool found = false;
+
+       /* lookup in the current node */
+       list_for_each_entry(cnode, &root->val, list) {
+               if (cnode->ip != chain->ips[i++])
+                       break;
+               if (!found)
+                       found = true;
+               if (i == chain->nr)
+                       break;
+       }
+
+       /* matches not, relay on the parent */
+       if (!found)
+               return -1;
+
+       /* we match only a part of the node. Split it and add the new chain */
+       if (i < root->val_nr) {
+               split_add_child(root, chain, cnode, i);
+               return 0;
+       }
+
+       /* we match 100% of the path, increment the hit */
+       if (i == root->val_nr) {
+               root->hit++;
+               return 0;
+       }
+
+       return __append_chain_children(root, chain);
+}
+
+void append_chain(struct callchain_node *root, struct ip_callchain *chain)
+{
+       if (__append_chain_children(root, chain) == -1)
+               add_child(root, chain);
+}

diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
new file mode 100644 (file)
index 0000000..fa1cd2f
--- /dev/null
+++ b/tools/perf/util/callchain.h
@@ -0,0 +1,33 @@
+#ifndef __PERF_CALLCHAIN_H
+#define __PERF_CALLCHAIN_H
+
+#include "../perf.h"
+#include "list.h"
+#include "rbtree.h"
+
+
+struct callchain_node {
+       struct callchain_node   *parent;
+       struct list_head        brothers;
+       struct list_head        children;
+       struct list_head        val;
+       struct rb_node          rb_node;
+       int                     val_nr;
+       int                     hit;
+};
+
+struct callchain_list {
+       unsigned long           ip;
+       struct list_head        list;
+};
+
+static inline void callchain_init(struct callchain_node *node)
+{
+       INIT_LIST_HEAD(&node->brothers);
+       INIT_LIST_HEAD(&node->children);
+       INIT_LIST_HEAD(&node->val);
+}
+
+void append_chain(struct callchain_node *root, struct ip_callchain *chain);
+void sort_chain_to_rbtree(struct rb_root *rb_root, struct callchain_node *node);
+#endif