From: Thomas Gleixner <tglx@linutronix.de>
To: linux-tip-commits@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, hpa@zytor.com, mingo@redhat.com,
tglx@linutronix.de
Subject: [tip:core/debugobjects] debugobjects: replace static objects when slab cache becomes available
Date: Tue, 17 Mar 2009 11:30:23 GMT [thread overview]
Message-ID: <tip-1be1cb7b47f0744141ed61cdb25648819ae1a56f@git.kernel.org> (raw)
In-Reply-To: <200903162049.58058.nickpiggin@yahoo.com.au>
Commit-ID: 1be1cb7b47f0744141ed61cdb25648819ae1a56f
Gitweb: http://git.kernel.org/tip/1be1cb7b47f0744141ed61cdb25648819ae1a56f
Author: Thomas Gleixner <tglx@linutronix.de>
AuthorDate: Mon, 16 Mar 2009 18:53:18 +0100
Commit: Thomas Gleixner <tglx@linutronix.de>
CommitDate: Tue, 17 Mar 2009 12:28:30 +0100
debugobjects: replace static objects when slab cache becomes available
Impact: refactor/consolidate object management, prepare for delayed free
debugobjects allocates static reference objects to track objects which
are initialized or activated before the slab cache becomes
available. These static reference objects have to be handled
seperately in free_object(). The handling of these objects is in the
way of implementing a delayed free functionality. The delayed free is
required to avoid callbacks into the mm code from
debug_check_no_obj_freed().
Replace the static object references with dynamic ones after the slab
cache has been initialized. The static objects are now marked initdata.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
LKML-Reference: <200903162049.58058.nickpiggin@yahoo.com.au>
---
lib/debugobjects.c | 66 +++++++++++++++++++++++++++++++++++++++++++++++++--
1 files changed, 63 insertions(+), 3 deletions(-)
diff --git a/lib/debugobjects.c b/lib/debugobjects.c
index 90e46fa..fdcda3d 100644
--- a/lib/debugobjects.c
+++ b/lib/debugobjects.c
@@ -30,7 +30,7 @@ struct debug_bucket {
static struct debug_bucket obj_hash[ODEBUG_HASH_SIZE];
-static struct debug_obj obj_static_pool[ODEBUG_POOL_SIZE];
+static struct debug_obj obj_static_pool[ODEBUG_POOL_SIZE] __initdata;
static DEFINE_SPINLOCK(pool_lock);
@@ -884,6 +884,63 @@ void __init debug_objects_early_init(void)
}
/*
+ * Convert the statically allocated objects to dynamic ones:
+ */
+static int debug_objects_replace_static_objects(void)
+{
+ struct debug_bucket *db = obj_hash;
+ struct hlist_node *node, *tmp;
+ struct debug_obj *obj, *new;
+ HLIST_HEAD(objects);
+ int i, cnt = 0;
+
+ for (i = 0; i < ODEBUG_POOL_SIZE; i++) {
+ obj = kmem_cache_zalloc(obj_cache, GFP_KERNEL);
+ if (!obj)
+ goto free;
+ hlist_add_head(&obj->node, &objects);
+ }
+
+ /*
+ * When debug_objects_mem_init() is called we know that only
+ * one CPU is up, so disabling interrupts is enough
+ * protection. This avoids the lockdep hell of lock ordering.
+ */
+ local_irq_disable();
+
+ /* Remove the statically allocated objects from the pool */
+ hlist_for_each_entry_safe(obj, node, tmp, &obj_pool, node)
+ hlist_del(&obj->node);
+ /* Move the allocated objects to the pool */
+ hlist_move_list(&objects, &obj_pool);
+
+ /* Replace the active object references */
+ for (i = 0; i < ODEBUG_HASH_SIZE; i++, db++) {
+ hlist_move_list(&db->list, &objects);
+
+ hlist_for_each_entry(obj, node, &objects, node) {
+ new = hlist_entry(obj_pool.first, typeof(*obj), node);
+ hlist_del(&new->node);
+ /* copy object data */
+ *new = *obj;
+ hlist_add_head(&new->node, &db->list);
+ cnt++;
+ }
+ }
+
+ printk(KERN_DEBUG "ODEBUG: %d of %d active objects replaced\n", cnt,
+ obj_pool_used);
+ local_irq_enable();
+ return 0;
+free:
+ hlist_for_each_entry_safe(obj, node, tmp, &objects, node) {
+ hlist_del(&obj->node);
+ kmem_cache_free(obj_cache, obj);
+ }
+ return -ENOMEM;
+}
+
+/*
* Called after the kmem_caches are functional to setup a dedicated
* cache pool, which has the SLAB_DEBUG_OBJECTS flag set. This flag
* prevents that the debug code is called on kmem_cache_free() for the
@@ -898,8 +955,11 @@ void __init debug_objects_mem_init(void)
sizeof (struct debug_obj), 0,
SLAB_DEBUG_OBJECTS, NULL);
- if (!obj_cache)
+ if (!obj_cache || debug_objects_replace_static_objects()) {
debug_objects_enabled = 0;
- else
+ if (obj_cache)
+ kmem_cache_destroy(obj_cache);
+ printk(KERN_WARNING "ODEBUG: out of memory.\n");
+ } else
debug_objects_selftest();
}
next prev parent reply other threads:[~2009-03-17 11:31 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-01-28 13:53 [PATCH 00/21] lockdep: pending queue Peter Zijlstra
2009-01-28 13:53 ` [PATCH 01/21] lockdep: annotate reclaim context (__GFP_NOFS) Peter Zijlstra
[not found] ` <tip-bf722c9d324864b4256edaa330751b77f2a19861@git.kernel.org>
2009-03-15 6:48 ` SLOB lockup (was: Re: [tip:core/locking] lockdep: annotate reclaim context (__GFP_NOFS), fix SLOB) Ingo Molnar
2009-03-15 6:51 ` Ingo Molnar
2009-03-15 9:06 ` Nick Piggin
2009-03-15 9:47 ` Ingo Molnar
2009-03-15 10:04 ` Nick Piggin
2009-03-15 15:16 ` Thomas Gleixner
2009-03-16 9:49 ` Nick Piggin
2009-03-17 11:30 ` Thomas Gleixner [this message]
2009-03-17 11:30 ` [tip:core/debugobjects] debugobjects: delay free of internal objects Thomas Gleixner
2009-03-15 14:56 ` SLOB lockup (was: Re: [tip:core/locking] lockdep: annotate reclaim context (__GFP_NOFS), fix SLOB) Matt Mackall
2009-03-16 10:00 ` Nick Piggin
2009-03-16 14:52 ` Matt Mackall
2009-03-16 15:00 ` Nick Piggin
2009-03-23 8:42 ` Pekka Enberg
2009-03-15 18:56 ` Ingo Molnar
2009-01-28 13:53 ` [PATCH 02/21] lockdep: sanitize bit names Peter Zijlstra
2009-01-28 13:53 ` [PATCH 03/21] lockdep: sanitize reclaim " Peter Zijlstra
2009-01-28 13:53 ` [PATCH 04/21] lockdep: lockdep_states.h Peter Zijlstra
2009-01-28 13:53 ` [PATCH 05/21] lockdep: simplify mark_held_locks Peter Zijlstra
2009-01-28 13:53 ` [PATCH 06/21] lockdep: simplify mark_lock() Peter Zijlstra
2009-01-28 13:53 ` [PATCH 07/21] lockdep: move state bit definitions around Peter Zijlstra
2009-01-28 13:54 ` [PATCH 08/21] lockdep: generate the state bit definitions Peter Zijlstra
2009-01-28 13:54 ` [PATCH 09/21] lockdep: generate usage strings Peter Zijlstra
2009-01-28 13:54 ` [PATCH 10/21] lockdep: split up mark_lock_irq() Peter Zijlstra
2009-01-28 13:54 ` [PATCH 11/21] lockdep: simplify the mark_lock_irq() helpers Peter Zijlstra
2009-01-28 13:54 ` [PATCH 12/21] lockdep: further simplify " Peter Zijlstra
2009-01-28 13:54 ` [PATCH 13/21] lockdep: simplify mark_lock_irq() helpers #3 Peter Zijlstra
2009-01-28 13:54 ` [PATCH 14/21] lockdep: merge the _READ mark_lock_irq() helpers Peter Zijlstra
2009-01-28 13:54 ` [PATCH 15/21] lockdep: merge the !_READ " Peter Zijlstra
2009-01-28 13:54 ` [PATCH 16/21] lockdep: fully reduce mark_lock_irq() Peter Zijlstra
2009-01-28 13:54 ` [PATCH 17/21] lockdep: remove macro usage from mark_held_locks() Peter Zijlstra
2009-01-28 13:54 ` [PATCH 18/21] lockdep: add comments to mark_lock_irq() Peter Zijlstra
2009-01-28 13:54 ` [PATCH 19/21] lockdep: simplify get_user_chars() Peter Zijlstra
2009-01-28 13:54 ` [PATCH 20/21] lockdep: get_user_chars() redo Peter Zijlstra
2009-01-28 13:54 ` [PATCH 21/21] lockdep: simplify check_prev_add_irq() Peter Zijlstra
2009-01-29 13:54 ` [PATCH 22/21] lockdep: use stringify.h Peter Zijlstra
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=tip-1be1cb7b47f0744141ed61cdb25648819ae1a56f@git.kernel.org \
--to=tglx@linutronix.de \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-tip-commits@vger.kernel.org \
--cc=mingo@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome