* [PATCH 2/2] Btrfs: use a lock to protect incompat/compat flag of the super block
@ 2013-04-11 10:30 Miao Xie
2013-04-17 22:17 ` David Sterba
0 siblings, 1 reply; 3+ messages in thread
From: Miao Xie @ 2013-04-11 10:30 UTC (permalink / raw)
To: Linux Btrfs
The following case will make the incompat/compat flag of the super block
be recovered.
Task1 |Task2
flags = btrfs_super_incompat_flags(); |
|flags = btrfs_super_incompat_flags();
flags |= new_flag1; |
|flags |= new_flag2;
btrfs_set_super_incompat_flags(flags); |
|btrfs_set_super_incompat_flags(flags);
the new_flag1 is recovered.
In order to avoid this problem, we introduce a lock named super_lock into
the btrfs_fs_info structure. If we want to update incompat/compat flags
of the super block, we must hold it.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
---
fs/btrfs/ctree.h | 22 ++++++++++++++++++++--
fs/btrfs/disk-io.c | 5 +++++
fs/btrfs/volumes.c | 10 +---------
3 files changed, 26 insertions(+), 11 deletions(-)
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 0d82922..a883e47 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1360,6 +1360,17 @@ struct btrfs_fs_info {
wait_queue_head_t transaction_blocked_wait;
wait_queue_head_t async_submit_wait;
+ /*
+ * Used to protect the incompat_flags, compat_flags, compat_ro_flags
+ * when they are updated.
+ *
+ * Because we do not clear the flags for ever, so we needn't use
+ * the lock on the read side.
+ *
+ * We also needn't use the lock when we mount the fs, because
+ * there is no other task which will update the flag.
+ */
+ spinlock_t super_lock;
struct btrfs_super_block *super_copy;
struct btrfs_super_block *super_for_commit;
struct block_device *__bdev;
@@ -3663,8 +3674,15 @@ static inline void __btrfs_set_fs_incompat(struct btrfs_fs_info *fs_info,
disk_super = fs_info->super_copy;
features = btrfs_super_incompat_flags(disk_super);
if (!(features & flag)) {
- features |= flag;
- btrfs_set_super_incompat_flags(disk_super, features);
+ spin_lock(&fs_info->super_lock);
+ features = btrfs_super_incompat_flags(disk_super);
+ if (!(features & flag)) {
+ features |= flag;
+ btrfs_set_super_incompat_flags(disk_super, features);
+ printk(KERN_INFO "btrfs: setting %llu feature flag\n",
+ flag);
+ }
+ spin_unlock(&fs_info->super_lock);
}
}
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 6d19a0a..ab8ef37 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2060,6 +2060,7 @@ int open_ctree(struct super_block *sb,
spin_lock_init(&fs_info->defrag_inodes_lock);
spin_lock_init(&fs_info->free_chunk_lock);
spin_lock_init(&fs_info->tree_mod_seq_lock);
+ spin_lock_init(&fs_info->super_lock);
rwlock_init(&fs_info->tree_mod_log_lock);
mutex_init(&fs_info->reloc_mutex);
seqlock_init(&fs_info->profiles_lock);
@@ -2319,6 +2320,10 @@ int open_ctree(struct super_block *sb,
goto fail_alloc;
}
+ /*
+ * Needn't use the lock because there is no other task which will
+ * update the flag.
+ */
btrfs_set_super_incompat_flags(disk_super, features);
features = btrfs_super_compat_ro_flags(disk_super) &
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 2854c82..e710db4 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -3674,18 +3674,10 @@ static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
static void check_raid56_incompat_flag(struct btrfs_fs_info *info, u64 type)
{
- u64 features;
-
if (!(type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)))
return;
- features = btrfs_super_incompat_flags(info->super_copy);
- if (features & BTRFS_FEATURE_INCOMPAT_RAID56)
- return;
-
- features |= BTRFS_FEATURE_INCOMPAT_RAID56;
- btrfs_set_super_incompat_flags(info->super_copy, features);
- printk(KERN_INFO "btrfs: setting RAID5/6 feature flag\n");
+ btrfs_set_fs_incompat(info, RAID56);
}
static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
--
1.8.0.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH 2/2] Btrfs: use a lock to protect incompat/compat flag of the super block
2013-04-11 10:30 [PATCH 2/2] Btrfs: use a lock to protect incompat/compat flag of the super block Miao Xie
@ 2013-04-17 22:17 ` David Sterba
2013-04-22 2:38 ` Miao Xie
0 siblings, 1 reply; 3+ messages in thread
From: David Sterba @ 2013-04-17 22:17 UTC (permalink / raw)
To: Miao Xie; +Cc: Linux Btrfs
On Thu, Apr 11, 2013 at 06:30:16PM +0800, Miao Xie wrote:
> In order to avoid this problem, we introduce a lock named super_lock into
> the btrfs_fs_info structure. If we want to update incompat/compat flags
> of the super block, we must hold it.
>
> + /*
> + * Used to protect the incompat_flags, compat_flags, compat_ro_flags
> + * when they are updated.
> + spinlock_t super_lock;
The lock name is too general for protecting just *_flags, do you have
plans to add more items from superblock under this lock? If no, I
suggest to pick a different name.
> @@ -3663,8 +3674,15 @@ static inline void __btrfs_set_fs_incompat(struct btrfs_fs_info *fs_info,
> disk_super = fs_info->super_copy;
> features = btrfs_super_incompat_flags(disk_super);
> if (!(features & flag)) {
> - features |= flag;
> - btrfs_set_super_incompat_flags(disk_super, features);
> + spin_lock(&fs_info->super_lock);
> + features = btrfs_super_incompat_flags(disk_super);
> + if (!(features & flag)) {
> + features |= flag;
> + btrfs_set_super_incompat_flags(disk_super, features);
> + printk(KERN_INFO "btrfs: setting %llu feature flag\n",
> + flag);
flag is u64, please use (unsigned long long)flag and possibly the new
btrfs_info replacement of printks.
> + }
> + spin_unlock(&fs_info->super_lock);
> }
> }
otherwise ok.
Reviewed-by: David Sterba <dsterba@suse.cz>
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH 2/2] Btrfs: use a lock to protect incompat/compat flag of the super block
2013-04-17 22:17 ` David Sterba
@ 2013-04-22 2:38 ` Miao Xie
0 siblings, 0 replies; 3+ messages in thread
From: Miao Xie @ 2013-04-22 2:38 UTC (permalink / raw)
To: dsterba, Linux Btrfs
On thu, 18 Apr 2013 00:17:11 +0200, David Sterba wrote:
> On Thu, Apr 11, 2013 at 06:30:16PM +0800, Miao Xie wrote:
>> In order to avoid this problem, we introduce a lock named super_lock into
>> the btrfs_fs_info structure. If we want to update incompat/compat flags
>> of the super block, we must hold it.
>>
>> + /*
>> + * Used to protect the incompat_flags, compat_flags, compat_ro_flags
>> + * when they are updated.
>
>> + spinlock_t super_lock;
>
> The lock name is too general for protecting just *_flags, do you have
> plans to add more items from superblock under this lock? If no, I
> suggest to pick a different name.
Yes, I want to add more items from super block under this lock.
>
>> @@ -3663,8 +3674,15 @@ static inline void __btrfs_set_fs_incompat(struct btrfs_fs_info *fs_info,
>> disk_super = fs_info->super_copy;
>> features = btrfs_super_incompat_flags(disk_super);
>> if (!(features & flag)) {
>> - features |= flag;
>> - btrfs_set_super_incompat_flags(disk_super, features);
>> + spin_lock(&fs_info->super_lock);
>> + features = btrfs_super_incompat_flags(disk_super);
>> + if (!(features & flag)) {
>> + features |= flag;
>> + btrfs_set_super_incompat_flags(disk_super, features);
>> + printk(KERN_INFO "btrfs: setting %llu feature flag\n",
>> + flag);
>
> flag is u64, please use (unsigned long long)flag and possibly the new
> btrfs_info replacement of printks.
OK, I'll modify my patch.
Thanks for your view.
Miao
>
>> + }
>> + spin_unlock(&fs_info->super_lock);
>> }
>> }
>
> otherwise ok.
>
> Reviewed-by: David Sterba <dsterba@suse.cz>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2013-04-22 2:37 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-04-11 10:30 [PATCH 2/2] Btrfs: use a lock to protect incompat/compat flag of the super block Miao Xie
2013-04-17 22:17 ` David Sterba
2013-04-22 2:38 ` Miao Xie
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox