Return-Path: <linux-kernel-owner+w=401wt.eu-S1755582AbZFHMWS@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1755582AbZFHMWS (ORCPT <rfc822;w@1wt.eu>);
	Mon, 8 Jun 2009 08:22:18 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755092AbZFHMWI
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Mon, 8 Jun 2009 08:22:08 -0400
Received: from viefep15-int.chello.at ([62.179.121.35]:26514 "EHLO
	viefep15-int.chello.at" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1754953AbZFHMWG (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Mon, 8 Jun 2009 08:22:06 -0400
X-SourceIP: 213.93.53.227
Subject: Re: [PATCH 0/8] kernel:lockdep:replace DFS with BFS
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: tom.leiming@gmail.com
Cc: mingo@elte.hu, linux-kernel@vger.kernel.org, akpm@linux-foundation.org
In-Reply-To: <1243781365-26814-1-git-send-email-tom.leiming@gmail.com>
References: <yes>  <1243781365-26814-1-git-send-email-tom.leiming@gmail.com>
Content-Type: text/plain
Date: Mon, 08 Jun 2009 14:22:07 +0200
Message-Id: <1244463727.13761.8700.camel@twins>
Mime-Version: 1.0
X-Mailer: Evolution 2.26.1 
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 18391
Lines: 586

On Sun, 2009-05-31 at 22:49 +0800, tom.leiming@gmail.com wrote:
> Hi,
> Currently lockdep uses recursion DFS(depth-first search) algorithm to
> search target in checking lock circle(check_noncircular()),irq-safe
> -> irq-unsafe(check_irq_usage()) and irq inversion when adding a new
> lock dependency. This patches replace the current DFS with BFS, based on
> the following consideration:
> 
>      1,no loss of efficiency, no matter DFS or BFS, the running time
>      are O(V+E) (V is vertex count, and E is edge count of one
>      graph);
> 
>      2,BFS may be easily implemented by circular queue and consumes
>      much less kernel stack space than DFS for DFS is implemented by
>      recursion.
> 
>      3, The shortest path can be obtained by BFS if the target is
>      found, but can't be got by DFS. By the shortest path, we can
>      shorten the lock dependency chain and help to troubleshoot lock
>      problem easier than before.
> 

OK, so I applied the patches and the cleanup below.

mostly little style nits and moving the circular queue into lockdep.c
(nobody else uses it, so why share it?).

One thing though, macros with return statements?! that's seriously bad
style, don't do that.

One thing that worries me a little is that we loose DaveM's
lockdep_dependency_visit() optimization. My brain seems unwilling to
co-operate on determining if BFS would make that redundant, so a little
word on that would be appreciated.

Then I booted it and all hell broke loose, so either I wrecked something
or you did :-), would you terribly mind poking at that a little?

Over all I like that patches, but they need a little more work. Could
you send delta patches from now on?

---

[    0.000999] =================================
[    0.000999] [ BUG: bad contention detected! ]
[    0.000999] ---------------------------------
[    0.000999] swapper/0 is trying to contend lock (old_style_seqlock_init) at:
[    0.000999] [<ffffffff8133a795>] _spin_lock+0x6d/0x75
[    0.000999] but there are no locks held!
[    0.000999]
[    0.000999] other info that might help us debug this:
[    0.000999] 1 lock held by swapper/0:
[    0.000999]  #0:  (xtime_lock){-.....}, at: [<ffffffff8106a360>] tick_periodic+0x1d/0x74
[    0.000999]
[    0.000999] stack backtrace:
[    0.000999] Pid: 0, comm: swapper Not tainted 2.6.30-rc8-tip #1049
[    0.000999] Call Trace:
[    0.000999]  <IRQ>  [<ffffffff81071bd8>] print_lock_contention_bug+0x100/0x110
[    0.000999]  [<ffffffff81071ce0>] lock_acquired+0xf8/0x2b4
[    0.000999]  [<ffffffff81010221>] ? update_vsyscall+0x2d/0xd0
[    0.000999]  [<ffffffff8133a795>] _spin_lock+0x6d/0x75
[    0.000999]  [<ffffffff81010221>] ? update_vsyscall+0x2d/0xd0
[    0.000999]  [<ffffffff81010221>] update_vsyscall+0x2d/0xd0
[    0.000999]  [<ffffffff81067469>] update_wall_time+0x4c1/0x4cc
[    0.000999]  [<ffffffff8105274b>] do_timer+0x15/0x1c
[    0.000999]  [<ffffffff8106a37e>] tick_periodic+0x3b/0x74
[    0.000999]  [<ffffffff8106a3db>] tick_handle_periodic+0x24/0x71
[    0.000999]  [<ffffffff8100ea23>] timer_interrupt+0x1f/0x26
[    0.000999]  [<ffffffff8109348b>] handle_IRQ_event+0x8e/0x19c
[    0.000999]  [<ffffffff8109501e>] handle_level_irq+0x9d/0xf3
[    0.000999]  [<ffffffff8100e24b>] handle_irq+0x24/0x2c
[    0.000999]  [<ffffffff8133f55b>] do_IRQ+0x63/0xc2
[    0.000999]  [<ffffffff8100c713>] ret_from_intr+0x0/0xf
[    0.000999]  <EOI>  [<ffffffff8133a538>] ? _spin_unlock_irqrestore+0x47/0x6d
[    0.000999]  [<ffffffff81093eea>] ? __setup_irq+0x1ea/0x277
[    0.000999]  [<ffffffff81094120>] ? setup_irq+0x25/0x2a
[    0.000999]  [<ffffffff8158e63e>] ? hpet_time_init+0x20/0x22
[    0.000999]  [<ffffffff8158bbcc>] ? start_kernel+0x2ee/0x37a
[    0.000999]  [<ffffffff8158b29a>] ? x86_64_start_reservations+0xaa/0xae
[    0.000999]  [<ffffffff8158b37f>] ? x86_64_start_kernel+0xe1/0xe8

And ended in a stuck boot at:

[   65.411804] rmmod         R  running task        0   616      1 0x00000008
[   65.411804]  ffff8800023c41a0 ffffea0003336ba8 ffff88007e3a3c08 ffffffff8106f71a
[   65.411804]  ffff88007e3a3c48 0000000000000202 ffff88007f821600 0000000000001823
[   65.411804]  ffff88007e109d20 ffff88007f8b5c80 0000000000000000 ffff88007e3a3d18
[   65.411804] Call Trace:
[   65.411804]  [<ffffffff8106f71a>] ? trace_hardirqs_on+0xd/0xf
[   65.411804]  [<ffffffff8113f1b6>] ? release_sysfs_dirent+0x91/0xb1
[   65.411804]  [<ffffffff81071af6>] ? print_lock_contention_bug+0x1e/0x110
[   65.411804]  [<ffffffff81071af6>] ? print_lock_contention_bug+0x1e/0x110
[   65.411804]  [<ffffffff8113eff3>] ? sysfs_addrm_start+0x7d/0xaa
[   65.411804]  [<ffffffff81071af6>] ? print_lock_contention_bug+0x1e/0x110
[   65.411804]  [<ffffffff810113ff>] ? alternatives_smp_module_del+0x37/0xd0
[   65.411804]  [<ffffffff810e61ee>] ? free_percpu+0x38/0xf8
[   65.411804]  [<ffffffff8106f71a>] ? trace_hardirqs_on+0xd/0xf
[   65.411804]  [<ffffffff8133a542>] ? _spin_unlock_irqrestore+0x51/0x6d
[   65.411804]  [<ffffffff810e62a5>] ? free_percpu+0xef/0xf8
[   65.411804]  [<ffffffff8107ade5>] ? free_module+0x104/0x118
[   65.411804]  [<ffffffff8107b0a8>] ? sys_delete_module+0x20c/0x22e
[   65.411804]  [<ffffffff81339f88>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[   65.411804]  [<ffffffff8100bcdb>] ? system_call_fastpath+0x16/0x1b

( I can send you my .config if you cannot reproduce, but I don't think
its anything special )

---

Subject: lockdep: clean up after BFS patches
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Mon Jun 08 13:32:31 CEST 2009


LKML-Reference: <new-submission>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
---
 include/linux/lockdep.h    |    7 -
 kernel/lockdep.c           |  232 +++++++++++++++++++++++++++------------------
 kernel/lockdep_internals.h |   97 ------------------
 3 files changed, 147 insertions(+), 189 deletions(-)
===================================================================
===================================================================
--- linux-2.6.orig/include/linux/lockdep.h
+++ linux-2.6/include/linux/lockdep.h
@@ -58,7 +58,6 @@ struct lock_class {
 
 	struct lockdep_subclass_key	*key;
 	unsigned int			subclass;
-	unsigned int			dep_gen_id;
 
 	/*
 	 * IRQ/softirq usage tracking bits:
@@ -150,9 +149,9 @@ struct lock_list {
 	struct stack_trace		trace;
 	int				distance;
 
-	/*The parent field is used to implement breadth-first search,and
-	 *the bit 0 is reused to indicate if the lock has been accessed
-	 *in BFS.
+	/*
+	 * The parent field is used to implement breadth-first search, and the
+	 * bit 0 is reused to indicate if the lock has been accessed in BFS.
 	 */
 	struct lock_list		*parent;
 };
===================================================================
--- linux-2.6.orig/kernel/lockdep.c
+++ linux-2.6/kernel/lockdep.c
@@ -43,6 +43,7 @@
 #include <linux/ftrace.h>
 #include <linux/stringify.h>
 #include <linux/bitops.h>
+
 #include <asm/sections.h>
 
 #include "lockdep_internals.h"
@@ -118,7 +119,7 @@ static inline int debug_locks_off_graph_
 static int lockdep_initialized;
 
 unsigned long nr_list_entries;
-struct lock_list list_entries[MAX_LOCKDEP_ENTRIES];
+static struct lock_list list_entries[MAX_LOCKDEP_ENTRIES];
 
 /*
  * All data structures here are protected by the global debug_lock.
@@ -390,19 +391,6 @@ unsigned int nr_process_chains;
 unsigned int max_lockdep_depth;
 unsigned int max_recursion_depth;
 
-static unsigned int lockdep_dependency_gen_id;
-
-static bool lockdep_dependency_visit(struct lock_class *source,
-				     unsigned int depth)
-{
-	if (!depth)
-		lockdep_dependency_gen_id++;
-	if (source->dep_gen_id == lockdep_dependency_gen_id)
-		return true;
-	source->dep_gen_id = lockdep_dependency_gen_id;
-	return false;
-}
-
 #ifdef CONFIG_DEBUG_LOCKDEP
 /*
  * We cannot printk in early bootup code. Not even early_printk()
@@ -575,64 +563,6 @@ static void print_lock_class_header(stru
 	print_ip_sym((unsigned long)class->key);
 }
 
-/*
- * printk the shortest lock dependencies from @start to @end in reverse order:
- */
-static void __used
-print_shortest_lock_dependencies(struct lock_list *leaf,
-				struct lock_list *root)
-{
-	struct lock_list *entry = leaf;
-	int depth;
-
-	/*compute depth from generated tree by BFS*/
-	depth = get_lock_depth(leaf);
-
-	do {
-		print_lock_class_header(entry->class, depth);
-		printk("%*s ... acquired at:\n", depth, "");
-		print_stack_trace(&entry->trace, 2);
-		printk("\n");
-
-		if (depth == 0 && (entry != root)) {
-			printk("lockdep:%s bad BFS generated tree\n", __func__);
-			break;
-		}
-
-		entry = get_lock_parent(entry);
-		depth--;
-	} while (entry && (depth >= 0));
-
-	return;
-}
-/*
- * printk all lock dependencies starting at <entry>:
- */
-static void __used
-print_lock_dependencies(struct lock_class *class, int depth)
-{
-	struct lock_list *entry;
-
-	if (lockdep_dependency_visit(class, depth))
-		return;
-
-	if (DEBUG_LOCKS_WARN_ON(depth >= 20))
-		return;
-
-	print_lock_class_header(class, depth);
-
-	list_for_each_entry(entry, &class->locks_after, entry) {
-		if (DEBUG_LOCKS_WARN_ON(!entry->class))
-			return;
-
-		print_lock_dependencies(entry->class, depth + 1);
-
-		printk("%*s ... acquired at:\n",depth,"");
-		print_stack_trace(&entry->trace, 2);
-		printk("\n");
-	}
-}
-
 static void print_kernel_version(void)
 {
 	printk("%s %.*s\n", init_utsname()->release,
@@ -927,14 +857,106 @@ static int add_lock_to_list(struct lock_
 	return 1;
 }
 
-unsigned long bfs_accessed[BITS_TO_LONGS(MAX_LOCKDEP_ENTRIES)];
-static struct circular_queue  lock_cq;
+/*For good efficiency of modular, we use power of 2*/
+#define MAX_CIRCULAR_QUEUE_SIZE		4096UL
+#define CQ_MASK				(MAX_CIRCULAR_QUEUE_SIZE-1)
+
+/* The circular_queue and helpers is used to implement the
+ * breadth-first search(BFS)algorithem, by which we can build
+ * the shortest path from the next lock to be acquired to the
+ * previous held lock if there is a circular between them.
+ * */
+struct circular_queue {
+	unsigned long element[MAX_CIRCULAR_QUEUE_SIZE];
+	unsigned int  front, rear;
+};
+
+static struct circular_queue lock_cq;
+static unsigned long bfs_accessed[BITS_TO_LONGS(MAX_LOCKDEP_ENTRIES)];
+
 unsigned int max_bfs_queue_depth;
+
+static inline void __cq_init(struct circular_queue *cq)
+{
+	cq->front = cq->rear = 0;
+	bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES);
+}
+
+static inline int __cq_empty(struct circular_queue *cq)
+{
+	return (cq->front == cq->rear);
+}
+
+static inline int __cq_full(struct circular_queue *cq)
+{
+	return ((cq->rear + 1) & CQ_MASK) == cq->front;
+}
+
+static inline int __cq_enqueue(struct circular_queue *cq, unsigned long elem)
+{
+	if (__cq_full(cq))
+		return -1;
+
+	cq->element[cq->rear] = elem;
+	cq->rear = (cq->rear + 1) & CQ_MASK;
+	return 0;
+}
+
+static inline int __cq_dequeue(struct circular_queue *cq, unsigned long *elem)
+{
+	if (__cq_empty(cq))
+		return -1;
+
+	*elem = cq->element[cq->front];
+	cq->front = (cq->front + 1) & CQ_MASK;
+	return 0;
+}
+
+static inline unsigned int  __cq_get_elem_count(struct circular_queue *cq)
+{
+	return (cq->rear - cq->front) & CQ_MASK;
+}
+
+static inline void mark_lock_accessed(struct lock_list *lock,
+					struct lock_list *parent)
+{
+	unsigned long nr;
+	nr = lock - list_entries;
+	WARN_ON(nr >= nr_list_entries);
+	lock->parent = parent;
+	set_bit(nr, bfs_accessed);
+}
+
+static inline unsigned long lock_accessed(struct lock_list *lock)
+{
+	unsigned long nr;
+	nr = lock - list_entries;
+	WARN_ON(nr >= nr_list_entries);
+	return test_bit(nr, bfs_accessed);
+}
+
+static inline struct lock_list *get_lock_parent(struct lock_list *child)
+{
+	return child->parent;
+}
+
+static inline int get_lock_depth(struct lock_list *child)
+{
+	int depth = 0;
+	struct lock_list *parent;
+
+	while ((parent = get_lock_parent(child))) {
+		child = parent;
+		depth++;
+	}
+	return depth;
+}
+
 static int __bfs(struct lock_list *source_entry,
-			void *data,
-			int (*match)(struct lock_list *entry, void *data),
-			struct lock_list **target_entry,
-			int forward)
+		 void *data,
+		 int (*match)(struct lock_list *entry, void *data),
+		 struct lock_list **target_entry,
+		 int forward)
 {
 	struct lock_list *entry;
 	struct list_head *head;
@@ -1202,14 +1224,6 @@ check_noncircular(struct lock_list *root
  * without creating any illegal irq-safe -> irq-unsafe lock dependency.
  */
 
-
-#define   BFS_PROCESS_RET(ret)	do { \
-					if (ret < 0) \
-						return print_bfs_bug(ret); \
-					if (ret == 1) \
-						return 1; \
-				} while (0)
-
 static inline int usage_match(struct lock_list *entry, void *bit)
 {
 	return entry->class->usage_mask & (1 << (enum lock_usage_bit)bit);
@@ -1236,6 +1250,8 @@ find_usage_forwards(struct lock_list *ro
 	debug_atomic_inc(&nr_find_usage_forwards_checks);
 
 	result = __bfs_forwards(root, (void *)bit, usage_match, target_entry);
+	if (result < 0)
+		return print_bfs_bug(result);
 
 	return result;
 }
@@ -1259,10 +1275,42 @@ find_usage_backwards(struct lock_list *r
 	debug_atomic_inc(&nr_find_usage_backwards_checks);
 
 	result = __bfs_backwards(root, (void *)bit, usage_match, target_entry);
+	if (result < 0)
+		return print_bfs_bug(result);
 
 	return result;
 }
 
+/*
+ * printk the shortest lock dependencies from @start to @end in reverse order:
+ */
+static void __used
+print_shortest_lock_dependencies(struct lock_list *leaf,
+				struct lock_list *root)
+{
+	struct lock_list *entry = leaf;
+	int depth;
+
+	/*compute depth from generated tree by BFS*/
+	depth = get_lock_depth(leaf);
+
+	do {
+		print_lock_class_header(entry->class, depth);
+		printk("%*s ... acquired at:\n", depth, "");
+		print_stack_trace(&entry->trace, 2);
+		printk("\n");
+
+		if (depth == 0 && (entry != root)) {
+			printk("lockdep:%s bad BFS generated tree\n", __func__);
+			break;
+		}
+
+		entry = get_lock_parent(entry);
+		depth--;
+	} while (entry && (depth >= 0));
+
+	return;
+}
 
 static int
 print_bad_irq_dependency(struct task_struct *curr,
@@ -1349,12 +1397,14 @@ check_usage(struct task_struct *curr, st
 
 	this.class = hlock_class(prev);
 	ret = find_usage_backwards(&this, bit_backwards, &target_entry);
-	BFS_PROCESS_RET(ret);
+	if (!ret || ret == 1)
+		return ret;
 
 	that.parent = NULL;
 	that.class = hlock_class(next);
 	ret = find_usage_forwards(&that, bit_forwards, &target_entry1);
-	BFS_PROCESS_RET(ret);
+	if (!ret || ret == 1)
+		return ret;
 
 	return print_bad_irq_dependency(curr, &this, &that,
 			target_entry, target_entry1,
@@ -2038,7 +2088,8 @@ check_usage_forwards(struct task_struct 
 	root.parent = NULL;
 	root.class = hlock_class(this);
 	ret = find_usage_forwards(&root, bit, &target_entry);
-	BFS_PROCESS_RET(ret);
+	if (!ret || ret == 1)
+		return ret;
 
 	return print_irq_inversion_bug(curr, &root, target_entry,
 					this, 1, irqclass);
@@ -2059,7 +2110,8 @@ check_usage_backwards(struct task_struct
 	root.parent = NULL;
 	root.class = hlock_class(this);
 	ret = find_usage_backwards(&root, bit, &target_entry);
-	BFS_PROCESS_RET(ret);
+	if (!ret || ret == 1)
+		return ret;
 
 	return print_irq_inversion_bug(curr, &root, target_entry,
 					this, 1, irqclass);
===================================================================
--- linux-2.6.orig/kernel/lockdep_internals.h
+++ linux-2.6/kernel/lockdep_internals.h
@@ -91,6 +91,8 @@ extern unsigned int nr_process_chains;
 extern unsigned int max_lockdep_depth;
 extern unsigned int max_recursion_depth;
 
+extern unsigned int max_bfs_queue_depth;
+
 #ifdef CONFIG_PROVE_LOCKING
 extern unsigned long lockdep_count_forward_deps(struct lock_class *);
 extern unsigned long lockdep_count_backward_deps(struct lock_class *);
@@ -136,98 +138,3 @@ extern atomic_t nr_find_usage_backwards_
 # define debug_atomic_dec(ptr)		do { } while (0)
 # define debug_atomic_read(ptr)		0
 #endif
-
-
-extern unsigned int max_bfs_queue_depth;
-extern unsigned long nr_list_entries;
-extern struct lock_list list_entries[MAX_LOCKDEP_ENTRIES];
-extern unsigned long bfs_accessed[];
-
-/*For good efficiency of modular, we use power of 2*/
-#define  MAX_CIRCULAR_QUE_SIZE	    4096UL
-
-/* The circular_queue and helpers is used to implement the
- * breadth-first search(BFS)algorithem, by which we can build
- * the shortest path from the next lock to be acquired to the
- * previous held lock if there is a circular between them.
- * */
-struct circular_queue{
-	unsigned long element[MAX_CIRCULAR_QUE_SIZE];
-	unsigned int  front, rear;
-};
-
-static inline void __cq_init(struct circular_queue *cq)
-{
-	cq->front = cq->rear = 0;
-	bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES);
-}
-
-static inline int __cq_empty(struct circular_queue *cq)
-{
-	return (cq->front == cq->rear);
-}
-
-static inline int __cq_full(struct circular_queue *cq)
-{
-	return ((cq->rear + 1)&(MAX_CIRCULAR_QUE_SIZE-1))  == cq->front;
-}
-
-static inline int __cq_enqueue(struct circular_queue *cq, unsigned long elem)
-{
-	if (__cq_full(cq))
-		return -1;
-
-	cq->element[cq->rear] = elem;
-	cq->rear = (cq->rear + 1)&(MAX_CIRCULAR_QUE_SIZE-1);
-	return 0;
-}
-
-static inline int __cq_dequeue(struct circular_queue *cq, unsigned long *elem)
-{
-	if (__cq_empty(cq))
-		return -1;
-
-	*elem = cq->element[cq->front];
-	cq->front = (cq->front + 1)&(MAX_CIRCULAR_QUE_SIZE-1);
-	return 0;
-}
-
-static inline unsigned int  __cq_get_elem_count(struct circular_queue *cq)
-{
-	return (cq->rear - cq->front)&(MAX_CIRCULAR_QUE_SIZE-1);
-}
-
-static inline void mark_lock_accessed(struct lock_list *lock,
-					struct lock_list *parent)
-{
-	unsigned long nr;
-	nr = lock - list_entries;
-	WARN_ON(nr >= nr_list_entries);
-	lock->parent = parent;
-	set_bit(nr, bfs_accessed);
-}
-
-static inline unsigned long lock_accessed(struct lock_list *lock)
-{
-	unsigned long nr;
-	nr = lock - list_entries;
-	WARN_ON(nr >= nr_list_entries);
-	return test_bit(nr, bfs_accessed);
-}
-
-static inline struct lock_list *get_lock_parent(struct lock_list *child)
-{
-	return child->parent;
-}
-
-static inline int get_lock_depth(struct lock_list *child)
-{
-	int depth = 0;
-	struct lock_list *parent;
-
-	while ((parent = get_lock_parent(child))) {
-		child = parent;
-		depth++;
-	}
-	return depth;
-}


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