From: Mark Fasheh <mfasheh@suse.com>
To: linux-kernel@vger.kernel.org
Cc: ocfs2-devel@oss.oracle.com, Joel Becker <joel.becker@oracle.com>,
Tao Ma <tao.ma@oracle.com>
Subject: [PATCH 56/62] ocfs2: Add inode stealing for ocfs2_reserve_new_inode
Date: Wed, 2 Apr 2008 13:15:06 -0700 [thread overview]
Message-ID: <12071674361985-git-send-email-mfasheh@suse.com> (raw)
In-Reply-To: <12071674333467-git-send-email-mfasheh@suse.com>
From: Tao Ma <tao.ma@oracle.com>
Inode allocation is modified to look in other nodes allocators during
extreme out of space situations. We retry our own slot when space is freed
back to the global bitmap, or whenever we've allocated more than 1024 inodes
from another slot.
Signed-off-by: Tao Ma <tao.ma@oracle.com>
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
---
fs/ocfs2/alloc.c | 2 +
fs/ocfs2/localalloc.c | 2 +
fs/ocfs2/namei.c | 2 +-
fs/ocfs2/ocfs2.h | 34 +++++++++++++++++++-
fs/ocfs2/suballoc.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++-
fs/ocfs2/super.c | 1 +
6 files changed, 116 insertions(+), 5 deletions(-)
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index 7d81aa6..a268821 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -5150,6 +5150,8 @@ static void ocfs2_truncate_log_worker(struct work_struct *work)
status = ocfs2_flush_truncate_log(osb);
if (status < 0)
mlog_errno(status);
+ else
+ ocfs2_init_inode_steal_slot(osb);
mlog_exit(status);
}
diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c
index b6d0719..ce0dc14 100644
--- a/fs/ocfs2/localalloc.c
+++ b/fs/ocfs2/localalloc.c
@@ -447,6 +447,8 @@ out_mutex:
iput(main_bm_inode);
out:
+ if (!status)
+ ocfs2_init_inode_steal_slot(osb);
mlog_exit(status);
return status;
}
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index ae9ad95..ab5a227 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -424,7 +424,7 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
fe->i_fs_generation = cpu_to_le32(osb->fs_generation);
fe->i_blkno = cpu_to_le64(fe_blkno);
fe->i_suballoc_bit = cpu_to_le16(suballoc_bit);
- fe->i_suballoc_slot = cpu_to_le16(osb->slot_num);
+ fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot);
fe->i_uid = cpu_to_le32(current->fsuid);
if (dir->i_mode & S_ISGID) {
fe->i_gid = cpu_to_le32(dir->i_gid);
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 9ff5811..3169237 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -208,11 +208,14 @@ struct ocfs2_super
u32 s_feature_incompat;
u32 s_feature_ro_compat;
- /* Protects s_next_generaion, osb_flags. Could protect more on
- * osb as it's very short lived. */
+ /* Protects s_next_generation, osb_flags and s_inode_steal_slot.
+ * Could protect more on osb as it's very short lived.
+ */
spinlock_t osb_lock;
u32 s_next_generation;
unsigned long osb_flags;
+ s16 s_inode_steal_slot;
+ atomic_t s_num_inodes_stolen;
unsigned long s_mount_opt;
unsigned int s_atime_quantum;
@@ -537,6 +540,33 @@ static inline unsigned int ocfs2_pages_per_cluster(struct super_block *sb)
return pages_per_cluster;
}
+static inline void ocfs2_init_inode_steal_slot(struct ocfs2_super *osb)
+{
+ spin_lock(&osb->osb_lock);
+ osb->s_inode_steal_slot = OCFS2_INVALID_SLOT;
+ spin_unlock(&osb->osb_lock);
+ atomic_set(&osb->s_num_inodes_stolen, 0);
+}
+
+static inline void ocfs2_set_inode_steal_slot(struct ocfs2_super *osb,
+ s16 slot)
+{
+ spin_lock(&osb->osb_lock);
+ osb->s_inode_steal_slot = slot;
+ spin_unlock(&osb->osb_lock);
+}
+
+static inline s16 ocfs2_get_inode_steal_slot(struct ocfs2_super *osb)
+{
+ s16 slot;
+
+ spin_lock(&osb->osb_lock);
+ slot = osb->s_inode_steal_slot;
+ spin_unlock(&osb->osb_lock);
+
+ return slot;
+}
+
#define ocfs2_set_bit ext2_set_bit
#define ocfs2_clear_bit ext2_clear_bit
#define ocfs2_test_bit ext2_test_bit
diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
index 33d5573..d2d278f 100644
--- a/fs/ocfs2/suballoc.c
+++ b/fs/ocfs2/suballoc.c
@@ -49,6 +49,8 @@
#define NOT_ALLOC_NEW_GROUP 0
#define ALLOC_NEW_GROUP 1
+#define OCFS2_MAX_INODES_TO_STEAL 1024
+
static inline void ocfs2_debug_bg(struct ocfs2_group_desc *bg);
static inline void ocfs2_debug_suballoc_inode(struct ocfs2_dinode *fe);
static inline u16 ocfs2_find_victim_chain(struct ocfs2_chain_list *cl);
@@ -109,7 +111,7 @@ static inline void ocfs2_block_to_cluster_group(struct inode *inode,
u64 *bg_blkno,
u16 *bg_bit_off);
-void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac)
+static void ocfs2_free_ac_resource(struct ocfs2_alloc_context *ac)
{
struct inode *inode = ac->ac_inode;
@@ -120,9 +122,17 @@ void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac)
mutex_unlock(&inode->i_mutex);
iput(inode);
+ ac->ac_inode = NULL;
}
- if (ac->ac_bh)
+ if (ac->ac_bh) {
brelse(ac->ac_bh);
+ ac->ac_bh = NULL;
+ }
+}
+
+void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac)
+{
+ ocfs2_free_ac_resource(ac);
kfree(ac);
}
@@ -522,10 +532,42 @@ bail:
return status;
}
+static int ocfs2_steal_inode_from_other_nodes(struct ocfs2_super *osb,
+ struct ocfs2_alloc_context *ac)
+{
+ int i, status = -ENOSPC;
+ s16 slot = ocfs2_get_inode_steal_slot(osb);
+
+ /* Start to steal inodes from the first slot after ours. */
+ if (slot == OCFS2_INVALID_SLOT)
+ slot = osb->slot_num + 1;
+
+ for (i = 0; i < osb->max_slots; i++, slot++) {
+ if (slot == osb->max_slots)
+ slot = 0;
+
+ if (slot == osb->slot_num)
+ continue;
+
+ status = ocfs2_reserve_suballoc_bits(osb, ac,
+ INODE_ALLOC_SYSTEM_INODE,
+ slot, NOT_ALLOC_NEW_GROUP);
+ if (status >= 0) {
+ ocfs2_set_inode_steal_slot(osb, slot);
+ break;
+ }
+
+ ocfs2_free_ac_resource(ac);
+ }
+
+ return status;
+}
+
int ocfs2_reserve_new_inode(struct ocfs2_super *osb,
struct ocfs2_alloc_context **ac)
{
int status;
+ s16 slot = ocfs2_get_inode_steal_slot(osb);
*ac = kzalloc(sizeof(struct ocfs2_alloc_context), GFP_KERNEL);
if (!(*ac)) {
@@ -539,9 +581,43 @@ int ocfs2_reserve_new_inode(struct ocfs2_super *osb,
(*ac)->ac_group_search = ocfs2_block_group_search;
+ /*
+ * slot is set when we successfully steal inode from other nodes.
+ * It is reset in 3 places:
+ * 1. when we flush the truncate log
+ * 2. when we complete local alloc recovery.
+ * 3. when we successfully allocate from our own slot.
+ * After it is set, we will go on stealing inodes until we find the
+ * need to check our slots to see whether there is some space for us.
+ */
+ if (slot != OCFS2_INVALID_SLOT &&
+ atomic_read(&osb->s_num_inodes_stolen) < OCFS2_MAX_INODES_TO_STEAL)
+ goto inode_steal;
+
+ atomic_set(&osb->s_num_inodes_stolen, 0);
status = ocfs2_reserve_suballoc_bits(osb, *ac,
INODE_ALLOC_SYSTEM_INODE,
osb->slot_num, ALLOC_NEW_GROUP);
+ if (status >= 0) {
+ status = 0;
+
+ /*
+ * Some inodes must be freed by us, so try to allocate
+ * from our own next time.
+ */
+ if (slot != OCFS2_INVALID_SLOT)
+ ocfs2_init_inode_steal_slot(osb);
+ goto bail;
+ } else if (status < 0 && status != -ENOSPC) {
+ mlog_errno(status);
+ goto bail;
+ }
+
+ ocfs2_free_ac_resource(*ac);
+
+inode_steal:
+ status = ocfs2_steal_inode_from_other_nodes(osb, *ac);
+ atomic_inc(&osb->s_num_inodes_stolen);
if (status < 0) {
if (status != -ENOSPC)
mlog_errno(status);
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 96ebe36..df63ba2 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -1394,6 +1394,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
INIT_LIST_HEAD(&osb->blocked_lock_list);
osb->blocked_lock_count = 0;
spin_lock_init(&osb->osb_lock);
+ ocfs2_init_inode_steal_slot(osb);
atomic_set(&osb->alloc_stats.moves, 0);
atomic_set(&osb->alloc_stats.local_data, 0);
--
1.5.4.1
next prev parent reply other threads:[~2008-04-02 20:39 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-02 20:14 [PATCH 0/62] Ocfs2 updates for 2.6.26-rc1 Mark Fasheh
2008-04-02 20:14 ` [PATCH 01/62] ocfs2: Move slot map access into slot_map.c Mark Fasheh
2008-04-02 20:14 ` [PATCH 02/62] ocfs2: Make ocfs2_slot_info private Mark Fasheh
2008-04-02 20:14 ` [PATCH 03/62] ocfs2: Change the recovery map to an array of node numbers Mark Fasheh
2008-04-02 20:14 ` [PATCH 04/62] ocfs2: slot_map I/O based on max_slots Mark Fasheh
2008-04-02 20:14 ` [PATCH 05/62] ocfs2: De-magic the in-memory slot map Mark Fasheh
2008-04-02 20:14 ` [PATCH 06/62] ocfs2: Define the contents of the slot_map file Mark Fasheh
2008-04-02 20:14 ` [PATCH 07/62] ocfs2: New slot map format Mark Fasheh
2008-04-02 20:14 ` [PATCH 08/62] ocfs2: Separate out dlm lock functions Mark Fasheh
2008-04-02 20:14 ` [PATCH 09/62] ocfs2: Use global DLM_ constants in generic code Mark Fasheh
2008-04-02 20:14 ` [PATCH 10/62] ocfs2: Use -errno instead of dlm_status for ocfs2_dlm_lock/unlock() API Mark Fasheh
2008-04-02 20:14 ` [PATCH 11/62] ocfs2: Create the lock status block union Mark Fasheh
2008-04-02 20:14 ` [PATCH 12/62] ocfs2: Introduce the new ocfs2_cluster_connect/disconnect() API Mark Fasheh
2008-04-02 20:14 ` [PATCH 13/62] ocfs2: Abstract out node number queries Mark Fasheh
2008-04-02 20:14 ` [PATCH 14/62] ocfs2: Move o2hb functionality into the stack glue Mark Fasheh
2008-04-02 20:14 ` [PATCH 15/62] ocfs2: Fill node number during cluster stack init Mark Fasheh
2008-04-02 20:14 ` [PATCH 16/62] ocfs2: Remove CANCELGRANT from the view of dlmglue Mark Fasheh
2008-04-02 20:14 ` [PATCH 17/62] ocfs2: handle async EAGAIN from NOQUEUE request Mark Fasheh
2008-04-02 20:14 ` [PATCH 18/62] ocfs2: Abstract out a debugging function for underlying dlms Mark Fasheh
2008-04-02 20:14 ` [PATCH 19/62] ocfs2: Clean up stackglue initialization Mark Fasheh
2008-04-02 20:14 ` [PATCH 20/62] ocfs2: Split o2cb code from generic stack functions Mark Fasheh
2008-04-02 20:14 ` [PATCH 21/62] ocfs2: Create ocfs2_stack_operations and split out the o2cb stack Mark Fasheh
2008-04-02 20:14 ` [PATCH 22/62] ocfs2: Break out stackglue into modules Mark Fasheh
2008-04-02 20:14 ` [PATCH 23/62] ocfs2: Create stack glue sysfs files Mark Fasheh
2008-04-02 20:14 ` [PATCH 24/62] ocfs2: Add the USERSPACE_STACK incompat bit Mark Fasheh
2008-04-02 20:14 ` [PATCH 25/62] ocfs2: Add the 'cluster_stack' sysfs file Mark Fasheh
2008-04-02 20:14 ` [PATCH 26/62] ocfs2: Add the user stack module Mark Fasheh
2008-04-02 20:14 ` [PATCH 27/62] ocfs2: Add the ocfs2_control misc device Mark Fasheh
2008-04-02 20:14 ` [PATCH 28/62] ocfs2: Start the ocfs2_control handshake Mark Fasheh
2008-04-02 20:14 ` [PATCH 29/62] ocfs2: Introduce the DOWN message to ocfs2_control Mark Fasheh
2008-04-02 20:14 ` [PATCH 30/62] ocfs2: Add the local node id to the handshake Mark Fasheh
2008-04-02 20:14 ` [PATCH 31/62] ocfs2: Add the 'set version' message to the ocfs2_control device Mark Fasheh
2008-04-02 20:14 ` [PATCH 32/62] ocfs2: add fsdlm to stackglue Mark Fasheh
2008-04-02 20:14 ` [PATCH 33/62] ocfs2: Change mlog_bug_on to BUG_ON in ocfs2_lockid.h Mark Fasheh
2008-04-02 20:14 ` [PATCH 34/62] ocfs2: Add kbuild for ocfs2_stack_user.ko Mark Fasheh
2008-04-02 20:14 ` [PATCH 35/62] ocfs2: Allow selection of cluster plug-ins Mark Fasheh
2008-04-02 20:14 ` [PATCH 36/62] ocfs2: Document /sys/fs/ocfs2 Mark Fasheh
2008-04-02 20:14 ` [PATCH 37/62] ocfs2/dlm: Rename slabcache dlm_mle_cache to o2dlm_mle Mark Fasheh
2008-04-02 20:14 ` [PATCH 38/62] ocfs2/dlm: Create slabcaches for lock and lockres Mark Fasheh
2008-04-02 20:14 ` [PATCH 39/62] ocfs2/dlm: Link all lockres' to a tracking list Mark Fasheh
2008-04-02 20:14 ` [PATCH 40/62] ocfs2/dlm: Create debugfs dirs Mark Fasheh
2008-04-02 20:14 ` [PATCH 41/62] ocfs2/dlm: Dump the dlm state in a debugfs file Mark Fasheh
2008-04-02 20:14 ` [PATCH 42/62] ocfs2/dlm: Dumps the lockres' into " Mark Fasheh
2008-04-02 20:14 ` [PATCH 43/62] ocfs2/dlm: Move struct dlm_master_list_entry to dlmcommon.h Mark Fasheh
2008-04-02 20:14 ` [PATCH 44/62] ocfs2/dlm: Dumps the mles into a debugfs file Mark Fasheh
2008-04-02 20:14 ` [PATCH 45/62] ocfs2/dlm: Dumps the purgelist " Mark Fasheh
2008-04-02 20:14 ` [PATCH 46/62] ocfs2/dlm: Move dlm_print_one_mle() from dlmmaster.c to dlmdebug.c Mark Fasheh
2008-04-02 20:14 ` [PATCH 47/62] ocfs2/dlm: Fix lockname in lockres print function Mark Fasheh
2008-04-02 20:14 ` [PATCH 48/62] ocfs2/dlm: Cleanup lockres print Mark Fasheh
2008-04-02 20:14 ` [PATCH 49/62] ocfs2: Reconnect after idle time out Mark Fasheh
2008-04-02 20:15 ` [PATCH 50/62] sysfs: Allow removal of symlinks in the sysfs root Mark Fasheh
2008-04-02 20:15 ` [PATCH 51/62] ocfs2: Move /sys/o2cb to /sys/fs/o2cb Mark Fasheh
2008-04-02 20:15 ` [PATCH 52/62] ocfs2: Add support for cross extent block Mark Fasheh
2008-04-02 20:15 ` [PATCH 53/62] ocfs2: Enable cross extent block merge Mark Fasheh
2008-04-02 20:15 ` [PATCH 54/62] ocfs2: Add a new parameter for ocfs2_reserve_suballoc_bits Mark Fasheh
2008-04-02 20:15 ` [PATCH 55/62] ocfs2: Add ac_alloc_slot in ocfs2_alloc_context Mark Fasheh
2008-04-02 20:15 ` Mark Fasheh [this message]
2008-04-02 20:15 ` [PATCH 57/62] fs/ocfs2/aops.c: test for IS_ERR rather than 0 Mark Fasheh
2008-04-02 20:15 ` [PATCH 58/62] ocfs2: Improve rename locking Mark Fasheh
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=12071674361985-git-send-email-mfasheh@suse.com \
--to=mfasheh@suse.com \
--cc=joel.becker@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=ocfs2-devel@oss.oracle.com \
--cc=tao.ma@oracle.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
all inboxes | Powered by JetHome®