[PATCH v5 0/8] Balance mangement

[PATCH v5 0/8] Balance mangement

[Hugo Mills]

   Hi, Chris,

   Another iteration of the balance management patches. This includes
suggestions from David Sterba, who did a fairly comprehensive review
of the patch series on IRC. It also includes fixes for comments from
Josef, and Li Zefan. The only thing I've not included here is Li's
suggestion of keeping track of the size of the chunks and the data in
them -- I'm working on the patch for that, and it'll be along once
I've had a chance to test it sensibly.

   Hugo.

---

Hugo Mills (8):
  btrfs: Balance progress monitoring
  btrfs: Cancel filesystem balance
  btrfs: Factor out enumeration of chunks to a separate function
  btrfs: Implement filtered balance ioctl
  btrfs: Balance filter for device ID
  btrfs: Balance filter for virtual address ranges
  btrfs: Replication-type information
  btrfs: Balance filter for physical device address

 fs/btrfs/ctree.h   |   10 ++
 fs/btrfs/disk-io.c |    2 +
 fs/btrfs/ioctl.c   |  104 +++++++++++++++-
 fs/btrfs/ioctl.h   |   49 +++++++
 fs/btrfs/super.c   |   16 +--
 fs/btrfs/volumes.c |  354 +++++++++++++++++++++++++++++++++++++++++-----------
 fs/btrfs/volumes.h |   21 +++-
 7 files changed, 468 insertions(+), 88 deletions(-)

-- 
1.7.2.5

[PATCH v5 1/8] btrfs: Balance progress monitoring

[Hugo Mills]

This patch introduces a basic form of progress monitoring for balance
operations, by counting the number of block groups remaining. The
information is exposed to userspace by an ioctl.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ctree.h   |    9 ++++++++
 fs/btrfs/disk-io.c |    2 +
 fs/btrfs/ioctl.c   |   34 +++++++++++++++++++++++++++++++
 fs/btrfs/ioctl.h   |    7 ++++++
 fs/btrfs/volumes.c |   56 ++++++++++++++++++++++++++++++++++++++++++++++++++-
 5 files changed, 106 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 7f78cc7..17c7ecc 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -865,6 +865,11 @@ struct btrfs_block_group_cache {
 	struct list_head cluster_list;
 };
 
+struct btrfs_balance_info {
+	u32 expected;
+	u32 completed;
+};
+
 struct reloc_control;
 struct btrfs_device;
 struct btrfs_fs_devices;
@@ -1078,6 +1083,10 @@ struct btrfs_fs_info {
 
 	/* filesystem state */
 	u64 fs_state;
+
+	/* Keep track of any rebalance operations on this FS */
+	spinlock_t balance_info_lock;
+	struct btrfs_balance_info *balance_info;
 };
 
 /*
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 100b07f..3d690de 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1645,6 +1645,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->ref_cache_lock);
 	spin_lock_init(&fs_info->fs_roots_radix_lock);
 	spin_lock_init(&fs_info->delayed_iput_lock);
+	spin_lock_init(&fs_info->balance_info_lock);
 
 	init_completion(&fs_info->kobj_unregister);
 	fs_info->tree_root = tree_root;
@@ -1670,6 +1671,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->sb = sb;
 	fs_info->max_inline = 8192 * 1024;
 	fs_info->metadata_ratio = 0;
+	fs_info->balance_info = NULL;
 
 	fs_info->thread_pool_size = min_t(unsigned long,
 					  num_online_cpus() + 2, 8);
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 5fdb2ab..a8fbb07 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -2375,6 +2375,38 @@ static noinline long btrfs_ioctl_wait_sync(struct file *file, void __user *argp)
 	return btrfs_wait_for_commit(root, transid);
 }
 
+/*
+ * Return the current status of any balance operation
+ */
+long btrfs_ioctl_balance_progress(
+	struct btrfs_fs_info *fs_info,
+	struct btrfs_ioctl_balance_progress __user *user_dest)
+{
+	int ret = 0;
+	struct btrfs_ioctl_balance_progress dest;
+
+	spin_lock(&fs_info->balance_info_lock);
+	if (!fs_info->balance_info) {
+		ret = -EINVAL;
+		goto error;
+	}
+
+	dest.expected = fs_info->balance_info->expected;
+	dest.completed = fs_info->balance_info->completed;
+
+	spin_unlock(&fs_info->balance_info_lock);
+
+	if (copy_to_user(user_dest, &dest,
+			 sizeof(struct btrfs_ioctl_balance_progress)))
+		return -EFAULT;
+
+	return 0;
+
+error:
+	spin_unlock(&fs_info->balance_info_lock);
+	return ret;
+}
+
 long btrfs_ioctl(struct file *file, unsigned int
 		cmd, unsigned long arg)
 {
@@ -2414,6 +2446,8 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_rm_dev(root, argp);
 	case BTRFS_IOC_BALANCE:
 		return btrfs_balance(root->fs_info->dev_root);
+	case BTRFS_IOC_BALANCE_PROGRESS:
+		return btrfs_ioctl_balance_progress(root->fs_info, argp);
 	case BTRFS_IOC_CLONE:
 		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
 	case BTRFS_IOC_CLONE_RANGE:
diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 8fb3821..7c37c6b 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -157,6 +157,11 @@ struct btrfs_ioctl_space_args {
 	struct btrfs_ioctl_space_info spaces[0];
 };
 
+struct btrfs_ioctl_balance_progress {
+	__u32 expected;
+	__u32 completed;
+};
+
 #define BTRFS_IOC_SNAP_CREATE _IOW(BTRFS_IOCTL_MAGIC, 1, \
 				   struct btrfs_ioctl_vol_args)
 #define BTRFS_IOC_DEFRAG _IOW(BTRFS_IOCTL_MAGIC, 2, \
@@ -203,4 +208,6 @@ struct btrfs_ioctl_space_args {
 				   struct btrfs_ioctl_vol_args_v2)
 #define BTRFS_IOC_SUBVOL_GETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 25, __u64)
 #define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64)
+#define BTRFS_IOC_BALANCE_PROGRESS _IOR(BTRFS_IOCTL_MAGIC, 27, \
+				  struct btrfs_ioctl_balance_progress)
 #endif
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index dd13eb8..bb2ffed 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2041,6 +2041,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	struct btrfs_root *chunk_root = dev_root->fs_info->chunk_root;
 	struct btrfs_trans_handle *trans;
 	struct btrfs_key found_key;
+	struct btrfs_balance_info *bal_info;
 
 	if (dev_root->fs_info->sb->s_flags & MS_RDONLY)
 		return -EROFS;
@@ -2051,6 +2052,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	mutex_lock(&dev_root->fs_info->volume_mutex);
 	dev_root = dev_root->fs_info->dev_root;
 
+	bal_info = kmalloc(
+		sizeof(struct btrfs_balance_info),
+		GFP_NOFS);
+	if (!bal_info) {
+		ret = -ENOMEM;
+		goto error_no_status;
+	}
+	spin_lock(&dev_root->fs_info->balance_info_lock);
+	dev_root->fs_info->balance_info = bal_info;
+	bal_info->expected = -1; /* One less than actually counted,
+				    because chunk 0 is special */
+	bal_info->completed = 0;
+	spin_unlock(&dev_root->fs_info->balance_info_lock);
+
 	/* step one make some room on all the devices */
 	list_for_each_entry(device, devices, dev_list) {
 		old_size = device->total_bytes;
@@ -2074,10 +2089,37 @@ int btrfs_balance(struct btrfs_root *dev_root)
 		btrfs_end_transaction(trans, dev_root);
 	}
 
-	/* step two, relocate all the chunks */
+	/* step two, count the chunks */
 	path = btrfs_alloc_path();
-	BUG_ON(!path);
+	if (!path) {
+		ret = -ENOMEM;
+		goto error;
+	}
+
+	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
+	key.offset = (u64)-1;
+	key.type = BTRFS_CHUNK_ITEM_KEY;
+
+	ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
+	if (ret <= 0) {
+		printk(KERN_ERR "btrfs: Failed to find the last chunk.\n");
+		BUG();
+	}
+
+	while (1) {
+		ret = btrfs_previous_item(chunk_root, path, 0,
+					  BTRFS_CHUNK_ITEM_KEY);
+		if (ret)
+			break;
+
+		spin_lock(&dev_root->fs_info->balance_info_lock);
+		bal_info->expected++;
+		spin_unlock(&dev_root->fs_info->balance_info_lock);
+	}
+
+	btrfs_release_path(chunk_root, path);
 
+	/* step three, relocate all the chunks */
 	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
 	key.offset = (u64)-1;
 	key.type = BTRFS_CHUNK_ITEM_KEY;
@@ -2115,10 +2157,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
 					   found_key.offset);
 		BUG_ON(ret && ret != -ENOSPC);
 		key.offset = found_key.offset - 1;
+		spin_lock(&dev_root->fs_info->balance_info_lock);
+		bal_info->completed++;
+		spin_unlock(&dev_root->fs_info->balance_info_lock);
+		printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
+		       bal_info->completed, bal_info->expected);
 	}
 	ret = 0;
 error:
 	btrfs_free_path(path);
+	spin_lock(&dev_root->fs_info->balance_info_lock);
+	kfree(dev_root->fs_info->balance_info);
+	dev_root->fs_info->balance_info = NULL;
+	spin_unlock(&dev_root->fs_info->balance_info_lock);
+error_no_status:
 	mutex_unlock(&dev_root->fs_info->volume_mutex);
 	return ret;
 }
-- 
1.7.2.5

Re: [PATCH v5 1/8] btrfs: Balance progress monitoring

[Helmut Hullen]

Hallo, Hugo,

Du meintest am 10.04.11:

> This patch introduces a basic form of progress monitoring for balance
> operations, by counting the number of block groups remaining. The
> information is exposed to userspace by an ioctl.

Just for curiosity:

If I remember correct then "btrfs device delete" shows growing and  
shrinking numbers, resp. on the remaining and on the deleting  
partition(s).

Can this patch show the remaining number of block groups too?

Viele Gruesse!
Helmut

Re: [PATCH v5 1/8] btrfs: Balance progress monitoring

[Hugo Mills]

[-- Attachment #1: Type: text/plain, Size: 1056 bytes --]

On Mon, Apr 11, 2011 at 07:34:00AM +0200, Helmut Hullen wrote:
> Hallo, Hugo,
> 
> Du meintest am 10.04.11:
> 
> > This patch introduces a basic form of progress monitoring for balance
> > operations, by counting the number of block groups remaining. The
> > information is exposed to userspace by an ioctl.
> 
> Just for curiosity:
> 
> If I remember correct then "btrfs device delete" shows growing and  
> shrinking numbers, resp. on the remaining and on the deleting  
> partition(s).
> 
> Can this patch show the remaining number of block groups too?

   This patch doesn't touch the code in device delete, but once this
patch series is finished and accepted, I'll take a look. We can
probably share a good chunk of the code between the two.

   Hugo.

-- 
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
   --- "How deep will this sub go?" "Oh,  she'll go all the way to ---   
                    the bottom if we don't stop her."                    

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

Re: [PATCH v5 1/8] btrfs: Balance progress monitoring

[David Sterba]

Hi,

I've noticed that Arne's scrub patches add scrub variables directly
into the fs_info structure, while you have a separate struct.

I was wondering whether it would be better to put items of
btrfs_balance_info to fs_info too, balance state is a global info.

Although fs_info is a huge structure now, 9402 bytes on 86_64, there is
no space saving in this case.

On Sun, Apr 10, 2011 at 10:06:04PM +0100, Hugo Mills wrote:
> diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
> index 7f78cc7..17c7ecc 100644
> --- a/fs/btrfs/ctree.h
> +++ b/fs/btrfs/ctree.h
> @@ -865,6 +865,11 @@ struct btrfs_block_group_cache {
>  	struct list_head cluster_list;
>  };
>  
> +struct btrfs_balance_info {
> +	u32 expected;
> +	u32 completed;

two u32 make one u64

> +};
> +
>  struct reloc_control;
>  struct btrfs_device;
>  struct btrfs_fs_devices;
> @@ -1078,6 +1083,10 @@ struct btrfs_fs_info {
>  
>  	/* filesystem state */
>  	u64 fs_state;
> +
> +	/* Keep track of any rebalance operations on this FS */
> +	spinlock_t balance_info_lock;
> +	struct btrfs_balance_info *balance_info;

a pointer is a u64 too

>  };
>  
>  /*
> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> index dd13eb8..bb2ffed 100644
> --- a/fs/btrfs/volumes.c
> +++ b/fs/btrfs/volumes.c
> @@ -2051,6 +2052,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
>  	mutex_lock(&dev_root->fs_info->volume_mutex);
>  	dev_root = dev_root->fs_info->dev_root;
>  
> +	bal_info = kmalloc(
> +		sizeof(struct btrfs_balance_info),
> +		GFP_NOFS);

... drop

> +	if (!bal_info) {
> +		ret = -ENOMEM;
> +		goto error_no_status;
> +	}
> +	spin_lock(&dev_root->fs_info->balance_info_lock);
> +	dev_root->fs_info->balance_info = bal_info;
> +	bal_info->expected = -1; /* One less than actually counted,
> +				    because chunk 0 is special */
> +	bal_info->completed = 0;
> +	spin_unlock(&dev_root->fs_info->balance_info_lock);
> +
>  	/* step one make some room on all the devices */
>  	list_for_each_entry(device, devices, dev_list) {
>  		old_size = device->total_bytes;
> @@ -2115,10 +2157,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
>  					   found_key.offset);
>  		BUG_ON(ret && ret != -ENOSPC);
>  		key.offset = found_key.offset - 1;
> +		spin_lock(&dev_root->fs_info->balance_info_lock);
> +		bal_info->completed++;
> +		spin_unlock(&dev_root->fs_info->balance_info_lock);
> +		printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
> +		       bal_info->completed, bal_info->expected);
>  	}
>  	ret = 0;
>  error:
>  	btrfs_free_path(path);
> +	spin_lock(&dev_root->fs_info->balance_info_lock);
> +	kfree(dev_root->fs_info->balance_info);

... drop

> +	dev_root->fs_info->balance_info = NULL;
> +	spin_unlock(&dev_root->fs_info->balance_info_lock);
> +error_no_status:
>  	mutex_unlock(&dev_root->fs_info->volume_mutex);
>  	return ret;
>  }
> -- 

Re: [PATCH v5 1/8] btrfs: Balance progress monitoring

[Hugo Mills]

[-- Attachment #1: Type: text/plain, Size: 3958 bytes --]

On Tue, Apr 12, 2011 at 07:12:32PM +0200, David Sterba wrote:
> Hi,
> 
> I've noticed that Arne's scrub patches add scrub variables directly
> into the fs_info structure, while you have a separate struct.

   Chris (I think -- might have been Josef) suggested the use of a
struct, back when I was first writing this code.

> I was wondering whether it would be better to put items of
> btrfs_balance_info to fs_info too, balance state is a global info.
> 
> Although fs_info is a huge structure now, 9402 bytes on 86_64, there is
> no space saving in this case.

   There will be savings in the future, however -- when I add Li's
suggestion for tracking the number of bytes (in the block groups as a
whole, and in terms of useful data stored), plus the vaddr of the
last-moved block group, the size of the btrfs_balance_info struct will
go up from its current 8 bytes to 48. I've just not quite finished
that patch yet, and wanted to get the rest of the patches settled
while I work on the new one...

   Hugo.

> On Sun, Apr 10, 2011 at 10:06:04PM +0100, Hugo Mills wrote:
> > diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
> > index 7f78cc7..17c7ecc 100644
> > --- a/fs/btrfs/ctree.h
> > +++ b/fs/btrfs/ctree.h
> > @@ -865,6 +865,11 @@ struct btrfs_block_group_cache {
> >  	struct list_head cluster_list;
> >  };
> >  
> > +struct btrfs_balance_info {
> > +	u32 expected;
> > +	u32 completed;
> 
> two u32 make one u64
> 
> > +};
> > +
> >  struct reloc_control;
> >  struct btrfs_device;
> >  struct btrfs_fs_devices;
> > @@ -1078,6 +1083,10 @@ struct btrfs_fs_info {
> >  
> >  	/* filesystem state */
> >  	u64 fs_state;
> > +
> > +	/* Keep track of any rebalance operations on this FS */
> > +	spinlock_t balance_info_lock;
> > +	struct btrfs_balance_info *balance_info;
> 
> a pointer is a u64 too
> 
> >  };
> >  
> >  /*
> > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> > index dd13eb8..bb2ffed 100644
> > --- a/fs/btrfs/volumes.c
> > +++ b/fs/btrfs/volumes.c
> > @@ -2051,6 +2052,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
> >  	mutex_lock(&dev_root->fs_info->volume_mutex);
> >  	dev_root = dev_root->fs_info->dev_root;
> >  
> > +	bal_info = kmalloc(
> > +		sizeof(struct btrfs_balance_info),
> > +		GFP_NOFS);
> 
> ... drop
> 
> > +	if (!bal_info) {
> > +		ret = -ENOMEM;
> > +		goto error_no_status;
> > +	}
> > +	spin_lock(&dev_root->fs_info->balance_info_lock);
> > +	dev_root->fs_info->balance_info = bal_info;
> > +	bal_info->expected = -1; /* One less than actually counted,
> > +				    because chunk 0 is special */
> > +	bal_info->completed = 0;
> > +	spin_unlock(&dev_root->fs_info->balance_info_lock);
> > +
> >  	/* step one make some room on all the devices */
> >  	list_for_each_entry(device, devices, dev_list) {
> >  		old_size = device->total_bytes;
> > @@ -2115,10 +2157,20 @@ int btrfs_balance(struct btrfs_root *dev_root)
> >  					   found_key.offset);
> >  		BUG_ON(ret && ret != -ENOSPC);
> >  		key.offset = found_key.offset - 1;
> > +		spin_lock(&dev_root->fs_info->balance_info_lock);
> > +		bal_info->completed++;
> > +		spin_unlock(&dev_root->fs_info->balance_info_lock);
> > +		printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
> > +		       bal_info->completed, bal_info->expected);
> >  	}
> >  	ret = 0;
> >  error:
> >  	btrfs_free_path(path);
> > +	spin_lock(&dev_root->fs_info->balance_info_lock);
> > +	kfree(dev_root->fs_info->balance_info);
> 
> ... drop
> 
> > +	dev_root->fs_info->balance_info = NULL;
> > +	spin_unlock(&dev_root->fs_info->balance_info_lock);
> > +error_no_status:
> >  	mutex_unlock(&dev_root->fs_info->volume_mutex);
> >  	return ret;
> >  }

-- 
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
       --- Never underestimate the bandwidth of a Volvo filled ---       
                           with backup tapes.                            

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

Re: [PATCH v5 1/8] btrfs: Balance progress monitoring

[David Sterba]

On Tue, Apr 12, 2011 at 07:42:07PM +0100, Hugo Mills wrote:
>    There will be savings in the future, however -- when I add Li's
> suggestion for tracking the number of bytes (in the block groups as a
> whole, and in terms of useful data stored), plus the vaddr of the
> last-moved block group, the size of the btrfs_balance_info struct will
> go up from its current 8 bytes to 48. I've just not quite finished
> that patch yet, and wanted to get the rest of the patches settled
> while I work on the new one...

makes sense, no objections.

david

[PATCH v5 2/8] btrfs: Cancel filesystem balance

[Hugo Mills]

This patch adds an ioctl for cancelling a btrfs balance operation
mid-flight. The ioctl simply sets a flag, and the operation terminates
after the current block group move has completed.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ctree.h   |    1 +
 fs/btrfs/ioctl.c   |   28 ++++++++++++++++++++++++++++
 fs/btrfs/ioctl.h   |    1 +
 fs/btrfs/volumes.c |    7 ++++++-
 4 files changed, 36 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 17c7ecc..1d26cfd 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -868,6 +868,7 @@ struct btrfs_block_group_cache {
 struct btrfs_balance_info {
 	u32 expected;
 	u32 completed;
+	int cancel_pending;
 };
 
 struct reloc_control;
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index a8fbb07..aef6329 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -2407,6 +2407,32 @@ error:
 	return ret;
 }
 
+/*
+ * Cancel a running balance operation
+ */
+long btrfs_ioctl_balance_cancel(struct btrfs_fs_info *fs_info)
+{
+	int err = 0;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	spin_lock(&fs_info->balance_info_lock);
+	if (!fs_info->balance_info) {
+		err = -EINVAL;
+		goto error;
+	}
+	if (fs_info->balance_info->cancel_pending) {
+		err = -ECANCELED;
+		goto error;
+	}
+	fs_info->balance_info->cancel_pending = 1;
+
+error:
+	spin_unlock(&fs_info->balance_info_lock);
+	return err;
+}
+
 long btrfs_ioctl(struct file *file, unsigned int
 		cmd, unsigned long arg)
 {
@@ -2448,6 +2474,8 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_balance(root->fs_info->dev_root);
 	case BTRFS_IOC_BALANCE_PROGRESS:
 		return btrfs_ioctl_balance_progress(root->fs_info, argp);
+	case BTRFS_IOC_BALANCE_CANCEL:
+		return btrfs_ioctl_balance_cancel(root->fs_info);
 	case BTRFS_IOC_CLONE:
 		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
 	case BTRFS_IOC_CLONE_RANGE:
diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 7c37c6b..2c49add 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -210,4 +210,5 @@ struct btrfs_ioctl_balance_progress {
 #define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64)
 #define BTRFS_IOC_BALANCE_PROGRESS _IOR(BTRFS_IOCTL_MAGIC, 27, \
 				  struct btrfs_ioctl_balance_progress)
+#define BTRFS_IOC_BALANCE_CANCEL _IO(BTRFS_IOCTL_MAGIC, 28)
 #endif
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index bb2ffed..cf019af 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2064,6 +2064,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	bal_info->expected = -1; /* One less than actually counted,
 				    because chunk 0 is special */
 	bal_info->completed = 0;
+	bal_info->cancel_pending = 0;
 	spin_unlock(&dev_root->fs_info->balance_info_lock);
 
 	/* step one make some room on all the devices */
@@ -2124,7 +2125,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	key.offset = (u64)-1;
 	key.type = BTRFS_CHUNK_ITEM_KEY;
 
-	while (1) {
+	while (!bal_info->cancel_pending) {
 		ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
 		if (ret < 0)
 			goto error;
@@ -2164,6 +2165,10 @@ int btrfs_balance(struct btrfs_root *dev_root)
 		       bal_info->completed, bal_info->expected);
 	}
 	ret = 0;
+	if (bal_info->cancel_pending) {
+		printk(KERN_INFO "btrfs: balance cancelled\n");
+		ret = -EINTR;
+	}
 error:
 	btrfs_free_path(path);
 	spin_lock(&dev_root->fs_info->balance_info_lock);
-- 
1.7.2.5

[PATCH v5 3/8] btrfs: Factor out enumeration of chunks to a separate function

[Hugo Mills]

The main balance function has two loops which are functionally
identical in their looping mechanism, but which perform a different
operation on the chunks they loop over. To avoid repeating code more
than necessary, factor this loop out into a separate iterator function
which takes a function parameter for the action to be performed.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/volumes.c |  174 +++++++++++++++++++++++++++++----------------------
 1 files changed, 99 insertions(+), 75 deletions(-)

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index cf019af..20c2772 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2029,6 +2029,97 @@ static u64 div_factor(u64 num, int factor)
 	return num;
 }
 
+/* Define a type, and two functions which can be used for the two
+ * phases of the balance operation: one for counting chunks, and one
+ * for actually moving them. */
+typedef void (*balance_iterator_function)(struct btrfs_root *,
+					  struct btrfs_balance_info *,
+					  struct btrfs_path *,
+					  struct btrfs_key *);
+
+static void balance_count_chunks(struct btrfs_root *chunk_root,
+			  struct btrfs_balance_info *bal_info,
+			  struct btrfs_path *path,
+			  struct btrfs_key *key)
+{
+	spin_lock(&chunk_root->fs_info->balance_info_lock);
+	bal_info->expected++;
+	spin_unlock(&chunk_root->fs_info->balance_info_lock);
+}
+
+static void balance_move_chunks(struct btrfs_root *chunk_root,
+			 struct btrfs_balance_info *bal_info,
+			 struct btrfs_path *path,
+			 struct btrfs_key *key)
+{
+	int ret;
+
+	ret = btrfs_relocate_chunk(chunk_root,
+				   chunk_root->root_key.objectid,
+				   key->objectid,
+				   key->offset);
+	BUG_ON(ret && ret != -ENOSPC);
+	spin_lock(&chunk_root->fs_info->balance_info_lock);
+	bal_info->completed++;
+	spin_unlock(&chunk_root->fs_info->balance_info_lock);
+	printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
+	       bal_info->completed, bal_info->expected);
+}
+
+/* Iterate through all chunks, performing some function on each one. */
+static int balance_iterate_chunks(struct btrfs_root *chunk_root,
+			   struct btrfs_balance_info *bal_info,
+			   balance_iterator_function iterator_fn)
+{
+	int ret = 0;
+	struct btrfs_path *path;
+	struct btrfs_key key;
+	struct btrfs_key found_key;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
+	key.offset = (u64)-1;
+	key.type = BTRFS_CHUNK_ITEM_KEY;
+
+	while (!bal_info->cancel_pending) {
+		ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
+		if (ret < 0)
+			break;
+		/*
+		 * this shouldn't happen, it means the last relocate
+		 * failed
+		 */
+		if (ret == 0)
+			break;
+
+		ret = btrfs_previous_item(chunk_root, path, 0,
+					  BTRFS_CHUNK_ITEM_KEY);
+		if (ret)
+			break;
+
+		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
+				      path->slots[0]);
+		if (found_key.objectid != key.objectid)
+			break;
+
+		/* chunk zero is special */
+		if (found_key.offset == 0)
+			break;
+
+		/* Call the function to do the work for this chunk */
+		btrfs_release_path(chunk_root, path);
+		iterator_fn(chunk_root, bal_info, path, &found_key);
+
+		key.offset = found_key.offset - 1;
+	}
+
+	btrfs_free_path(path);
+	return ret;
+}
+
 int btrfs_balance(struct btrfs_root *dev_root)
 {
 	int ret;
@@ -2036,11 +2127,8 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	struct btrfs_device *device;
 	u64 old_size;
 	u64 size_to_free;
-	struct btrfs_path *path;
-	struct btrfs_key key;
 	struct btrfs_root *chunk_root = dev_root->fs_info->chunk_root;
 	struct btrfs_trans_handle *trans;
-	struct btrfs_key found_key;
 	struct btrfs_balance_info *bal_info;
 
 	if (dev_root->fs_info->sb->s_flags & MS_RDONLY)
@@ -2061,8 +2149,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	}
 	spin_lock(&dev_root->fs_info->balance_info_lock);
 	dev_root->fs_info->balance_info = bal_info;
-	bal_info->expected = -1; /* One less than actually counted,
-				    because chunk 0 is special */
+	bal_info->expected = 0;
 	bal_info->completed = 0;
 	bal_info->cancel_pending = 0;
 	spin_unlock(&dev_root->fs_info->balance_info_lock);
@@ -2091,86 +2178,23 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	}
 
 	/* step two, count the chunks */
-	path = btrfs_alloc_path();
-	if (!path) {
-		ret = -ENOMEM;
+	ret = balance_iterate_chunks(chunk_root, bal_info,
+				     balance_count_chunks);
+	if (ret)
 		goto error;
-	}
-
-	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
-	key.offset = (u64)-1;
-	key.type = BTRFS_CHUNK_ITEM_KEY;
-
-	ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
-	if (ret <= 0) {
-		printk(KERN_ERR "btrfs: Failed to find the last chunk.\n");
-		BUG();
-	}
-
-	while (1) {
-		ret = btrfs_previous_item(chunk_root, path, 0,
-					  BTRFS_CHUNK_ITEM_KEY);
-		if (ret)
-			break;
-
-		spin_lock(&dev_root->fs_info->balance_info_lock);
-		bal_info->expected++;
-		spin_unlock(&dev_root->fs_info->balance_info_lock);
-	}
-
-	btrfs_release_path(chunk_root, path);
 
 	/* step three, relocate all the chunks */
-	key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
-	key.offset = (u64)-1;
-	key.type = BTRFS_CHUNK_ITEM_KEY;
-
-	while (!bal_info->cancel_pending) {
-		ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
-		if (ret < 0)
-			goto error;
-
-		/*
-		 * this shouldn't happen, it means the last relocate
-		 * failed
-		 */
-		if (ret == 0)
-			break;
-
-		ret = btrfs_previous_item(chunk_root, path, 0,
-					  BTRFS_CHUNK_ITEM_KEY);
-		if (ret)
-			break;
-
-		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
-				      path->slots[0]);
-		if (found_key.objectid != key.objectid)
-			break;
-
-		/* chunk zero is special */
-		if (found_key.offset == 0)
-			break;
+	ret = balance_iterate_chunks(chunk_root, bal_info,
+				     balance_move_chunks);
+	if (ret)
+		goto error;
 
-		btrfs_release_path(chunk_root, path);
-		ret = btrfs_relocate_chunk(chunk_root,
-					   chunk_root->root_key.objectid,
-					   found_key.objectid,
-					   found_key.offset);
-		BUG_ON(ret && ret != -ENOSPC);
-		key.offset = found_key.offset - 1;
-		spin_lock(&dev_root->fs_info->balance_info_lock);
-		bal_info->completed++;
-		spin_unlock(&dev_root->fs_info->balance_info_lock);
-		printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
-		       bal_info->completed, bal_info->expected);
-	}
 	ret = 0;
 	if (bal_info->cancel_pending) {
 		printk(KERN_INFO "btrfs: balance cancelled\n");
 		ret = -EINTR;
 	}
 error:
-	btrfs_free_path(path);
 	spin_lock(&dev_root->fs_info->balance_info_lock);
 	kfree(dev_root->fs_info->balance_info);
 	dev_root->fs_info->balance_info = NULL;
-- 
1.7.2.5

Re: [PATCH v5 3/8] btrfs: Factor out enumeration of chunks to a separate function

[David Sterba]

On Sun, Apr 10, 2011 at 10:06:06PM +0100, Hugo Mills wrote:
> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> index cf019af..20c2772 100644
> --- a/fs/btrfs/volumes.c
> +++ b/fs/btrfs/volumes.c
> @@ -2029,6 +2029,97 @@ static u64 div_factor(u64 num, int factor)
> +static void balance_move_chunks(struct btrfs_root *chunk_root,
> +			 struct btrfs_balance_info *bal_info,
> +			 struct btrfs_path *path,
> +			 struct btrfs_key *key)
> +{
> +	int ret;
> +
> +	ret = btrfs_relocate_chunk(chunk_root,
> +				   chunk_root->root_key.objectid,
> +				   key->objectid,
> +				   key->offset);
> +	BUG_ON(ret && ret != -ENOSPC);
> +	spin_lock(&chunk_root->fs_info->balance_info_lock);
> +	bal_info->completed++;
> +	spin_unlock(&chunk_root->fs_info->balance_info_lock);
> +	printk(KERN_INFO "btrfs: balance: %llu/%llu block groups completed\n",
> +	       bal_info->completed, bal_info->expected);

as you've changed 'completed' and 'expected' members to 32bit, change
%llu to %u.

d/

[PATCH v5 4/8] btrfs: Implement filtered balance ioctl

[Hugo Mills]

The filtered balance ioctl provides a facility to perform a balance
operation on a subset of the chunks in the filesystem. This patch
implements the base ioctl for this operation, and one filter type.
The filter in this patch selects chunks on the basis of their chunk
flags field, and can select any combination of bits set or unset.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ioctl.c   |   42 ++++++++++++++++++++++++++++++++-
 fs/btrfs/ioctl.h   |   27 +++++++++++++++++++++
 fs/btrfs/volumes.c |   65 +++++++++++++++++++++++++++++++++++++++++++++------
 fs/btrfs/volumes.h |    4 ++-
 4 files changed, 128 insertions(+), 10 deletions(-)

diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index aef6329..4bc4da2 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -2433,6 +2433,44 @@ error:
 	return err;
 }
 
+long btrfs_ioctl_balance(struct btrfs_root *dev_root,
+			 struct btrfs_ioctl_balance_start __user *user_filters)
+{
+	int ret = 0;
+	struct btrfs_ioctl_balance_start *dest;
+
+	dest = kmalloc(sizeof(struct btrfs_ioctl_balance_start), GFP_KERNEL);
+	if (!dest)
+		return -ENOMEM;
+
+	if (copy_from_user(dest, user_filters,
+			   sizeof(struct btrfs_ioctl_balance_start))) {
+		ret = -EFAULT;
+		goto error;
+	}
+
+	/* Basic sanity checking: has the user requested anything outside
+	 * the range we know about? */
+	if (dest->flags & ~BTRFS_BALANCE_FILTER_MASK) {
+		ret = -ENOTSUPP;
+		goto error;
+	}
+
+	/* Do the balance */
+	ret = btrfs_balance(dev_root, dest);
+	if (ret)
+		goto error;
+
+	if (copy_to_user(user_filters, dest,
+			 sizeof(struct btrfs_ioctl_balance_start))) {
+		ret = -EFAULT;
+	}
+
+error:
+	kfree(dest);
+	return ret;
+}
+
 long btrfs_ioctl(struct file *file, unsigned int
 		cmd, unsigned long arg)
 {
@@ -2471,11 +2509,13 @@ long btrfs_ioctl(struct file *file, unsigned int
 	case BTRFS_IOC_RM_DEV:
 		return btrfs_ioctl_rm_dev(root, argp);
 	case BTRFS_IOC_BALANCE:
-		return btrfs_balance(root->fs_info->dev_root);
+		return btrfs_ioctl_balance(root->fs_info->dev_root, NULL);
 	case BTRFS_IOC_BALANCE_PROGRESS:
 		return btrfs_ioctl_balance_progress(root->fs_info, argp);
 	case BTRFS_IOC_BALANCE_CANCEL:
 		return btrfs_ioctl_balance_cancel(root->fs_info);
+	case BTRFS_IOC_BALANCE_FILTERED:
+		return btrfs_ioctl_balance(root->fs_info->dev_root, argp);
 	case BTRFS_IOC_CLONE:
 		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
 	case BTRFS_IOC_CLONE_RANGE:
diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 2c49add..eb91d20 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -162,6 +162,31 @@ struct btrfs_ioctl_balance_progress {
 	__u32 completed;
 };
 
+/* Types of balance filter */
+#define BTRFS_BALANCE_FILTER_COUNT_ONLY (1 << 0)
+
+#define BTRFS_BALANCE_FILTER_CHUNK_TYPE (1 << 1)
+#define BTRFS_BALANCE_FILTER_MASK ((1 << 2) - 1) /* Logical or of all filter
+				       * flags -- effectively versions
+				       * the filtered balance ioctl */
+
+/* All the possible options for a filter */
+struct btrfs_ioctl_balance_start {
+	__u64 flags; /* Bit field indicating which fields of this struct
+			are filled */
+
+	/* Output values: chunk counts */
+	__u64 examined;
+	__u64 balanced;
+
+	/* For FILTER_CHUNK_TYPE */
+	__u64 chunk_type;      /* Flag bits required */
+	__u64 chunk_type_mask; /* Mask of bits to examine */
+
+	__u64 spare[506]; /* Make up the size of the structure to 4088
+			   * bytes for future expansion */
+};
+
 #define BTRFS_IOC_SNAP_CREATE _IOW(BTRFS_IOCTL_MAGIC, 1, \
 				   struct btrfs_ioctl_vol_args)
 #define BTRFS_IOC_DEFRAG _IOW(BTRFS_IOCTL_MAGIC, 2, \
@@ -211,4 +236,6 @@ struct btrfs_ioctl_balance_progress {
 #define BTRFS_IOC_BALANCE_PROGRESS _IOR(BTRFS_IOCTL_MAGIC, 27, \
 				  struct btrfs_ioctl_balance_progress)
 #define BTRFS_IOC_BALANCE_CANCEL _IO(BTRFS_IOCTL_MAGIC, 28)
+#define BTRFS_IOC_BALANCE_FILTERED _IOWR(BTRFS_IOCTL_MAGIC, 29, \
+				struct btrfs_ioctl_balance_start)
 #endif
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 20c2772..95c603a 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2029,6 +2029,37 @@ static u64 div_factor(u64 num, int factor)
 	return num;
 }
 
+int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
+						 struct btrfs_root *chunk_root,
+						 struct btrfs_path *path,
+						 struct btrfs_key *key)
+{
+	struct extent_buffer *eb;
+	struct btrfs_chunk *chunk;
+
+	/* No filter defined, everything matches */
+	if (!filter)
+		return 1;
+
+	/* No flags set, everything matches */
+	if (filter->flags == 0)
+		return 1;
+
+	eb = path->nodes[0];
+	chunk = btrfs_item_ptr(eb, path->slots[0],
+						   struct btrfs_chunk);
+
+	if (filter->flags & BTRFS_BALANCE_FILTER_CHUNK_TYPE) {
+		if ((btrfs_chunk_type(eb, chunk) & filter->chunk_type_mask)
+			!= filter->chunk_type)
+		{
+			return 0;
+		}
+	}
+
+	return ret;
+}
+
 /* Define a type, and two functions which can be used for the two
  * phases of the balance operation: one for counting chunks, and one
  * for actually moving them. */
@@ -2069,6 +2100,7 @@ static void balance_move_chunks(struct btrfs_root *chunk_root,
 /* Iterate through all chunks, performing some function on each one. */
 static int balance_iterate_chunks(struct btrfs_root *chunk_root,
 			   struct btrfs_balance_info *bal_info,
+			   struct btrfs_ioctl_balance_start *filter,
 			   balance_iterator_function iterator_fn)
 {
 	int ret = 0;
@@ -2084,6 +2116,9 @@ static int balance_iterate_chunks(struct btrfs_root *chunk_root,
 	key.offset = (u64)-1;
 	key.type = BTRFS_CHUNK_ITEM_KEY;
 
+	filter->examined = 0;
+	filter->balanced = 0;
+
 	while (!bal_info->cancel_pending) {
 		ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0);
 		if (ret < 0)
@@ -2110,17 +2145,29 @@ static int balance_iterate_chunks(struct btrfs_root *chunk_root,
 			break;
 
 		/* Call the function to do the work for this chunk */
-		btrfs_release_path(chunk_root, path);
-		iterator_fn(chunk_root, bal_info, path, &found_key);
+		filter->examined += 1;
+
+		if (balance_chunk_filter(filter, chunk_root,
+					 path, &found_key)) {
+			btrfs_release_path(chunk_root, path);
+			iterator_fn(chunk_root, bal_info, path, &found_key);
+			filter->balanced += 1;
+		} else {
+			btrfs_release_path(chunk_root, path);
+		}
 
 		key.offset = found_key.offset - 1;
 	}
 
+	printk(KERN_INFO "btrfs: balance: %llu chunks considered, %llu chunks balanced\n",
+		   filter->examined, filter->balanced);
+
 	btrfs_free_path(path);
 	return ret;
 }
 
-int btrfs_balance(struct btrfs_root *dev_root)
+int btrfs_balance(struct btrfs_root *dev_root,
+				  struct btrfs_ioctl_balance_start *filters)
 {
 	int ret;
 	struct list_head *devices = &dev_root->fs_info->fs_devices->devices;
@@ -2179,15 +2226,17 @@ int btrfs_balance(struct btrfs_root *dev_root)
 
 	/* step two, count the chunks */
 	ret = balance_iterate_chunks(chunk_root, bal_info,
-				     balance_count_chunks);
+				 filters, balance_count_chunks);
 	if (ret)
 		goto error;
 
 	/* step three, relocate all the chunks */
-	ret = balance_iterate_chunks(chunk_root, bal_info,
-				     balance_move_chunks);
-	if (ret)
-		goto error;
+	if (!(filters->flags & BTRFS_BALANCE_FILTER_COUNT_ONLY)) {
+		ret = balance_iterate_chunks(chunk_root, bal_info,
+					     filters, balance_move_chunks);
+		if (ret)
+			goto error;
+	}
 
 	ret = 0;
 	if (bal_info->cancel_pending) {
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index 7fb59d4..168771b 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -22,6 +22,7 @@
 #include <linux/bio.h>
 #include <linux/sort.h>
 #include "async-thread.h"
+#include "ioctl.h"
 
 #define BTRFS_STRIPE_LEN	(64 * 1024)
 
@@ -205,7 +206,8 @@ struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid,
 				       u8 *uuid, u8 *fsid);
 int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
 int btrfs_init_new_device(struct btrfs_root *root, char *path);
-int btrfs_balance(struct btrfs_root *dev_root);
+int btrfs_balance(struct btrfs_root *dev_root,
+		  struct btrfs_ioctl_balance_start *filters);
 void btrfs_unlock_volumes(void);
 void btrfs_lock_volumes(void);
 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
-- 
1.7.2.5

[PATCH v5 5/8] btrfs: Balance filter for device ID

[Hugo Mills]

Balance filter to take only chunks which have (or had) a stripe on the
given device. Useful if a device has been forcibly removed from the
filesystem, and the data from that device needs rebuilding.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ioctl.h   |    8 ++++++--
 fs/btrfs/volumes.c |   16 +++++++++++++++-
 2 files changed, 21 insertions(+), 3 deletions(-)

diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index eb91d20..54523c0 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -166,7 +166,8 @@ struct btrfs_ioctl_balance_progress {
 #define BTRFS_BALANCE_FILTER_COUNT_ONLY (1 << 0)
 
 #define BTRFS_BALANCE_FILTER_CHUNK_TYPE (1 << 1)
-#define BTRFS_BALANCE_FILTER_MASK ((1 << 2) - 1) /* Logical or of all filter
+#define BTRFS_BALANCE_FILTER_DEVID (1 << 2)
+#define BTRFS_BALANCE_FILTER_MASK ((1 << 3) - 1) /* Logical or of all filter
 				       * flags -- effectively versions
 				       * the filtered balance ioctl */
 
@@ -183,7 +184,10 @@ struct btrfs_ioctl_balance_start {
 	__u64 chunk_type;      /* Flag bits required */
 	__u64 chunk_type_mask; /* Mask of bits to examine */
 
-	__u64 spare[506]; /* Make up the size of the structure to 4088
+	/* For FILTER_DEVID */
+	__u64 devid;
+
+	__u64 spare[505]; /* Make up the size of the structure to 4088
 			   * bytes for future expansion */
 };
 
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 95c603a..e7fa2ab 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2036,6 +2036,7 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 {
 	struct extent_buffer *eb;
 	struct btrfs_chunk *chunk;
+	int i;
 
 	/* No filter defined, everything matches */
 	if (!filter)
@@ -2056,8 +2057,21 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 			return 0;
 		}
 	}
+	if (filter->flags & BTRFS_BALANCE_FILTER_DEVID) {
+		int num_stripes = btrfs_chunk_num_stripes(eb, chunk);
+		int res = 0;
+		for (i = 0; i < num_stripes; i++) {
+			struct btrfs_stripe *stripe = btrfs_stripe_nr(chunk, i);
+			if (btrfs_stripe_devid(eb, stripe) == filter->devid) {
+				res = 1;
+				break;
+			}
+		}
+		if (!res)
+			return 0;
+	}
 
-	return ret;
+	return 1;
 }
 
 /* Define a type, and two functions which can be used for the two
-- 
1.7.2.5

[PATCH v5 6/8] btrfs: Balance filter for virtual address ranges

[Hugo Mills]

Allow the balancing of chunks where some part of the chunk lies within
the virtual (i.e. btrfs-internal) address range passed.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ioctl.h   |    9 +++++++--
 fs/btrfs/volumes.c |    6 ++++++
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 54523c0..50d4801 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -167,7 +167,8 @@ struct btrfs_ioctl_balance_progress {
 
 #define BTRFS_BALANCE_FILTER_CHUNK_TYPE (1 << 1)
 #define BTRFS_BALANCE_FILTER_DEVID (1 << 2)
-#define BTRFS_BALANCE_FILTER_MASK ((1 << 3) - 1) /* Logical or of all filter
+#define BTRFS_BALANCE_FILTER_VIRTUAL_ADDRESS_RANGE (1 << 3)
+#define BTRFS_BALANCE_FILTER_MASK ((1 << 4) - 1) /* Logical or of all filter
 				       * flags -- effectively versions
 				       * the filtered balance ioctl */
 
@@ -187,7 +188,11 @@ struct btrfs_ioctl_balance_start {
 	/* For FILTER_DEVID */
 	__u64 devid;
 
-	__u64 spare[505]; /* Make up the size of the structure to 4088
+	/* For FILTER_VIRTUAL_ADDRESS_RANGE */
+	__u64 vrange_start;
+	__u64 vrange_end;
+
+	__u64 spare[503]; /* Make up the size of the structure to 4088
 			   * bytes for future expansion */
 };
 
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index e7fa2ab..230d100 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2070,6 +2070,12 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 		if (!res)
 			return 0;
 	}
+	if (filter->flags & BTRFS_BALANCE_FILTER_VIRTUAL_ADDRESS_RANGE) {
+		u64 start = key->offset;
+		u64 end = start + btrfs_chunk_length(eb, chunk);
+		if (filter->vrange_start >= end || start >= filter->vrange_end)
+			return 0;
+	}
 
 	return 1;
 }
-- 
1.7.2.5

[PATCH v5 7/8] btrfs: Replication-type information

[Hugo Mills]

There are a few places in btrfs where knowledge of the various
parameters of a replication type is needed. Factor this out into a
single function which can supply all the relevant information.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/super.c   |   16 +++-----
 fs/btrfs/volumes.c |   97 +++++++++++++++++++++++++++++++++++-----------------
 fs/btrfs/volumes.h |   17 +++++++++
 3 files changed, 88 insertions(+), 42 deletions(-)

diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index d39a989..4341730 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -879,12 +879,12 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	struct btrfs_device_info *devices_info;
 	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
 	struct btrfs_device *device;
+	struct btrfs_replication_info repl_info;
 	u64 skip_space;
 	u64 type;
 	u64 avail_space;
 	u64 used_space;
 	u64 min_stripe_size;
-	int min_stripes = 1;
 	int i = 0, nr_devices;
 	int ret;
 
@@ -898,12 +898,7 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 
 	/* calc min stripe number for data space alloction */
 	type = btrfs_get_alloc_profile(root, 1);
-	if (type & BTRFS_BLOCK_GROUP_RAID0)
-		min_stripes = 2;
-	else if (type & BTRFS_BLOCK_GROUP_RAID1)
-		min_stripes = 2;
-	else if (type & BTRFS_BLOCK_GROUP_RAID10)
-		min_stripes = 4;
+	btrfs_get_replication_info(&repl_info, type);
 
 	if (type & BTRFS_BLOCK_GROUP_DUP)
 		min_stripe_size = 2 * BTRFS_STRIPE_LEN;
@@ -971,14 +966,15 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 
 	i = nr_devices - 1;
 	avail_space = 0;
-	while (nr_devices >= min_stripes) {
+	while (nr_devices >= repl_info.devs_min) {
 		if (devices_info[i].max_avail >= min_stripe_size) {
 			int j;
 			u64 alloc_size;
 
-			avail_space += devices_info[i].max_avail * min_stripes;
+			avail_space += devices_info[i].max_avail
+			  * repl_info.devs_min;
 			alloc_size = devices_info[i].max_avail;
-			for (j = i + 1 - min_stripes; j <= i; j++)
+			for (j = i + 1 - repl_info.devs_min; j <= i; j++)
 				devices_info[j].max_avail -= alloc_size;
 		}
 		i--;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 230d100..9216ad85 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -141,6 +141,52 @@ static void requeue_list(struct btrfs_pending_bios *pending_bios,
 		pending_bios->tail = tail;
 }
 
+void btrfs_get_replication_info(struct btrfs_replication_info *info,
+								u64 type)
+{
+	info->sub_stripes = 1;
+	info->dev_stripes = 1;
+	info->devs_increment = 1;
+	info->num_copies = 1;
+	info->devs_max = 0;	/* 0 == as many as possible */
+	info->devs_min = 1;
+
+	if (type & BTRFS_BLOCK_GROUP_DUP) {
+		info->dev_stripes = 2;
+		info->num_copies = 2;
+		info->devs_max = 1;
+	} else if (type & BTRFS_BLOCK_GROUP_RAID0) {
+		info->devs_min = 2;
+	} else if (type & BTRFS_BLOCK_GROUP_RAID1) {
+		info->devs_increment = 2;
+		info->num_copies = 2;
+		info->devs_max = 2;
+		info->devs_min = 2;
+	} else if (type & BTRFS_BLOCK_GROUP_RAID10) {
+		info->sub_stripes = 2;
+		info->devs_increment = 2;
+		info->num_copies = 2;
+		info->devs_min = 4;
+	}
+
+	if (type & BTRFS_BLOCK_GROUP_DATA) {
+		info->max_stripe_size = 1024 * 1024 * 1024;
+		info->min_stripe_size = 64 * 1024 * 1024;
+		info->max_chunk_size = 10 * info->max_stripe_size;
+	} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
+		info->max_stripe_size = 256 * 1024 * 1024;
+		info->min_stripe_size = 32 * 1024 * 1024;
+		info->max_chunk_size = info->max_stripe_size;
+	} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
+		info->max_stripe_size = 8 * 1024 * 1024;
+		info->min_stripe_size = 1 * 1024 * 1024;
+		info->max_chunk_size = 2 * info->max_stripe_size;
+	} else {
+		printk(KERN_ERR "Block group is of an unknown usage type: not data, metadata or system.\n");
+		BUG_ON(1);
+	}
+}
+
 /*
  * we try to collect pending bios for a device so we don't get a large
  * number of procs sending bios down to the same device.  This greatly
@@ -1248,6 +1294,7 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	struct block_device *bdev;
 	struct buffer_head *bh = NULL;
 	struct btrfs_super_block *disk_super;
+	struct btrfs_replication_info repl_info;
 	u64 all_avail;
 	u64 devid;
 	u64 num_devices;
@@ -1261,18 +1308,16 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		root->fs_info->avail_system_alloc_bits |
 		root->fs_info->avail_metadata_alloc_bits;
 
-	if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) &&
-	    root->fs_info->fs_devices->num_devices <= 4) {
-		printk(KERN_ERR "btrfs: unable to go below four devices "
-		       "on raid10\n");
-		ret = -EINVAL;
-		goto out;
-	}
+	btrfs_get_replication_info(&repl_info, all_avail);
 
-	if ((all_avail & BTRFS_BLOCK_GROUP_RAID1) &&
-	    root->fs_info->fs_devices->num_devices <= 2) {
-		printk(KERN_ERR "btrfs: unable to go below two "
-		       "devices on raid1\n");
+	if (root->fs_info->fs_devices->num_devices <= repl_info.devs_min) {
+		if (all_avail & BTRFS_BLOCK_GROUP_RAID10) {
+			printk(KERN_ERR "btrfs: unable to go below four devices "
+				   "on raid10\n");
+		} else if (all_avail & BTRFS_BLOCK_GROUP_RAID1) {
+			printk(KERN_ERR "btrfs: unable to go below two "
+				   "devices on raid1\n");
+		}
 		ret = -EINVAL;
 		goto out;
 	}
@@ -2037,6 +2082,7 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 	struct extent_buffer *eb;
 	struct btrfs_chunk *chunk;
 	int i;
+	struct btrfs_replication_info replinfo;
 
 	/* No filter defined, everything matches */
 	if (!filter)
@@ -2050,6 +2096,8 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 	chunk = btrfs_item_ptr(eb, path->slots[0],
 						   struct btrfs_chunk);
 
+	btrfs_get_replication_info(&replinfo, btrfs_chunk_type(eb, chunk));
+
 	if (filter->flags & BTRFS_BALANCE_FILTER_CHUNK_TYPE) {
 		if ((btrfs_chunk_type(eb, chunk) & filter->chunk_type_mask)
 			!= filter->chunk_type)
@@ -2492,34 +2540,19 @@ static u64 __btrfs_calc_stripe_size(struct btrfs_fs_devices *fs_devices,
 				    u64 proposed_size, u64 type,
 				    int num_stripes, int small_stripe)
 {
-	int min_stripe_size = 1 * 1024 * 1024;
+	struct btrfs_replication_info repl_info;
 	u64 calc_size = proposed_size;
 	u64 max_chunk_size = calc_size;
-	int ncopies = 1;
 
-	if (type & (BTRFS_BLOCK_GROUP_RAID1 |
-		    BTRFS_BLOCK_GROUP_DUP |
-		    BTRFS_BLOCK_GROUP_RAID10))
-		ncopies = 2;
-
-	if (type & BTRFS_BLOCK_GROUP_DATA) {
-		max_chunk_size = 10 * calc_size;
-		min_stripe_size = 64 * 1024 * 1024;
-	} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
-		max_chunk_size = 256 * 1024 * 1024;
-		min_stripe_size = 32 * 1024 * 1024;
-	} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
-		calc_size = 8 * 1024 * 1024;
-		max_chunk_size = calc_size * 2;
-		min_stripe_size = 1 * 1024 * 1024;
-	}
+	btrfs_get_replication_info(&repl_info, type);
+	max_chunk_size = repl_info.max_chunk_size;
 
 	/* we don't want a chunk larger than 10% of writeable space */
 	max_chunk_size = min(div_factor(fs_devices->total_rw_bytes, 1),
 			     max_chunk_size);
 
-	if (calc_size * num_stripes > max_chunk_size * ncopies) {
-		calc_size = max_chunk_size * ncopies;
+	if (calc_size * num_stripes > max_chunk_size * repl_info.num_copies) {
+		calc_size = max_chunk_size * repl_info.num_copies;
 		do_div(calc_size, num_stripes);
 		do_div(calc_size, BTRFS_STRIPE_LEN);
 		calc_size *= BTRFS_STRIPE_LEN;
@@ -2527,7 +2560,7 @@ static u64 __btrfs_calc_stripe_size(struct btrfs_fs_devices *fs_devices,
 
 	/* we don't want tiny stripes */
 	if (!small_stripe)
-		calc_size = max_t(u64, min_stripe_size, calc_size);
+		calc_size = max_t(u64, repl_info.min_stripe_size, calc_size);
 
 	/*
 	 * we're about to do_div by the BTRFS_STRIPE_LEN so lets make sure
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index 168771b..a94de82 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -146,6 +146,22 @@ struct btrfs_device_info {
 	u64 max_avail;
 };
 
+/*
+ * Information about a the parameters of a replication strategy (RAID
+ * level)
+ */
+struct btrfs_replication_info {
+	u32 sub_stripes;
+	u32 dev_stripes;
+	u32 devs_increment;
+	u32 num_copies;
+	u32 devs_max;
+	u32 devs_min;
+	u64 max_stripe_size;
+	u64 min_stripe_size;
+	u64 max_chunk_size;
+};
+
 /* Used to sort the devices by max_avail(descending sort) */
 int btrfs_cmp_device_free_bytes(const void *dev_info1, const void *dev_info2);
 
@@ -214,4 +230,5 @@ int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
 int find_free_dev_extent(struct btrfs_trans_handle *trans,
 			 struct btrfs_device *device, u64 num_bytes,
 			 u64 *start, u64 *max_avail);
+void btrfs_get_replication_info(struct btrfs_replication_info *info, u64 type);
 #endif
-- 
1.7.2.5

[PATCH v5 8/8] btrfs: Balance filter for physical device address

[Hugo Mills]

Add a filter for balancing which allows the selection of chunks with
data in the given byte range on any block device in the filesystem. On
its own, this filter is of little use, but when used with the devid
filter, it can be used to rebalance all chunks which lie on a part of
a specific device.

Signed-off-by: Hugo Mills <hugo@carfax.org.uk>
---
 fs/btrfs/ioctl.h   |    9 +++++++--
 fs/btrfs/volumes.c |   19 +++++++++++++++++++
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 50d4801..a037714 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h
@@ -168,7 +168,8 @@ struct btrfs_ioctl_balance_progress {
 #define BTRFS_BALANCE_FILTER_CHUNK_TYPE (1 << 1)
 #define BTRFS_BALANCE_FILTER_DEVID (1 << 2)
 #define BTRFS_BALANCE_FILTER_VIRTUAL_ADDRESS_RANGE (1 << 3)
-#define BTRFS_BALANCE_FILTER_MASK ((1 << 4) - 1) /* Logical or of all filter
+#define BTRFS_BALANCE_FILTER_DEVICE_ADDRESS_RANGE (1 << 4)
+#define BTRFS_BALANCE_FILTER_MASK ((1 << 5) - 1) /* Logical or of all filter
 				       * flags -- effectively versions
 				       * the filtered balance ioctl */
 
@@ -192,7 +193,11 @@ struct btrfs_ioctl_balance_start {
 	__u64 vrange_start;
 	__u64 vrange_end;
 
-	__u64 spare[503]; /* Make up the size of the structure to 4088
+	/* For FILTER_DEVICE_ADDRESS_RANGE */
+	__u64 drange_start;
+	__u64 drange_end;
+
+	__u64 spare[501]; /* Make up the size of the structure to 4088
 			   * bytes for future expansion */
 };
 
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 9216ad85..19dc117 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2124,6 +2124,25 @@ int balance_chunk_filter(struct btrfs_ioctl_balance_start *filter,
 		if (filter->vrange_start >= end || start >= filter->vrange_end)
 			return 0;
 	}
+	if (filter->flags & BTRFS_BALANCE_FILTER_DEVICE_ADDRESS_RANGE) {
+		int num_stripes = btrfs_chunk_num_stripes(eb, chunk);
+		int stripe_length = btrfs_chunk_length(eb, chunk)
+			* num_stripes / replinfo.num_copies;
+		int res = 0;
+
+		for (i = 0; i < num_stripes; i++) {
+			struct btrfs_stripe *stripe = btrfs_stripe_nr(chunk, i);
+			u64 start = btrfs_stripe_offset(eb, stripe);
+			u64 end = start + stripe_length;
+			if (filter->drange_start < end
+			    && start < filter->drange_end) {
+				res = 1;
+				break;
+			}
+		}
+		if (!res)
+			return 0;
+	}
 
 	return 1;
 }
-- 
1.7.2.5