From: Chris Mason <chris.mason@fusionio.com>
Date: Wed, 20 Feb 2013 19:06:05 +0000 (-0500)
Subject: Merge branch 'raid56-experimental' into for-linus-3.9
X-Git-Tag: v3.9-rc1~17^2~28
X-Git-Url: https://www.infradead.org/git/?a=commitdiff_plain;h=e942f883bc6651d50be139477baf6fb0eed3d5bb;p=users%2Fhch%2Fmisc.git
Merge branch 'raid56-experimental' into for-linus-3.9
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Conflicts:
fs/btrfs/ctree.h
fs/btrfs/extent-tree.c
fs/btrfs/inode.c
fs/btrfs/volumes.c
---
e942f883bc6651d50be139477baf6fb0eed3d5bb
diff --cc fs/btrfs/ctree.h
index 1679051f4d39,e3a4fd70f55a..3dcedfe4f759
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@@ -956,16 -954,10 +958,20 @@@ struct btrfs_dev_replace_item
#define BTRFS_BLOCK_GROUP_RAID1 (1ULL << 4)
#define BTRFS_BLOCK_GROUP_DUP (1ULL << 5)
#define BTRFS_BLOCK_GROUP_RAID10 (1ULL << 6)
+ #define BTRFS_BLOCK_GROUP_RAID5 (1 << 7)
+ #define BTRFS_BLOCK_GROUP_RAID6 (1 << 8)
#define BTRFS_BLOCK_GROUP_RESERVED BTRFS_AVAIL_ALLOC_BIT_SINGLE
-#define BTRFS_NR_RAID_TYPES 7
+
+enum btrfs_raid_types {
+ BTRFS_RAID_RAID10,
+ BTRFS_RAID_RAID1,
+ BTRFS_RAID_DUP,
+ BTRFS_RAID_RAID0,
+ BTRFS_RAID_SINGLE,
++ BTRFS_RAID_RAID5,
++ BTRFS_RAID_RAID6,
+ BTRFS_NR_RAID_TYPES
+};
#define BTRFS_BLOCK_GROUP_TYPE_MASK (BTRFS_BLOCK_GROUP_DATA | \
BTRFS_BLOCK_GROUP_SYSTEM | \
@@@ -1237,11 -1235,23 +1249,28 @@@ struct seq_list
u64 seq;
};
+enum btrfs_orphan_cleanup_state {
+ ORPHAN_CLEANUP_STARTED = 1,
+ ORPHAN_CLEANUP_DONE = 2,
+};
+
+ /* used by the raid56 code to lock stripes for read/modify/write */
+ struct btrfs_stripe_hash {
+ struct list_head hash_list;
+ wait_queue_head_t wait;
+ spinlock_t lock;
+ };
+
+ /* used by the raid56 code to lock stripes for read/modify/write */
+ struct btrfs_stripe_hash_table {
+ struct list_head stripe_cache;
+ spinlock_t cache_lock;
+ int cache_size;
+ struct btrfs_stripe_hash table[];
+ };
+
+ #define BTRFS_STRIPE_HASH_TABLE_BITS 11
+
/* fs_info */
struct reloc_control;
struct btrfs_device;
diff --cc fs/btrfs/extent-tree.c
index 5cd44e239595,7e801ada6695..b3ecca447ddf
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@@ -2505,9 -2547,9 +2552,10 @@@ again
ret = run_clustered_refs(trans, root, &cluster);
if (ret < 0) {
+ btrfs_release_ref_cluster(&cluster);
spin_unlock(&delayed_refs->lock);
btrfs_abort_transaction(trans, root, ret);
+ atomic_dec(&delayed_refs->procs_running_refs);
return ret;
}
@@@ -5591,16 -5577,23 +5659,20 @@@ wait_block_group_cache_done(struct btrf
int __get_raid_index(u64 flags)
{
- int index;
-
if (flags & BTRFS_BLOCK_GROUP_RAID10)
- index = 0;
+ return BTRFS_RAID_RAID10;
else if (flags & BTRFS_BLOCK_GROUP_RAID1)
- index = 1;
+ return BTRFS_RAID_RAID1;
else if (flags & BTRFS_BLOCK_GROUP_DUP)
- index = 2;
+ return BTRFS_RAID_DUP;
else if (flags & BTRFS_BLOCK_GROUP_RAID0)
- index = 3;
+ return BTRFS_RAID_RAID0;
- else
- return BTRFS_RAID_SINGLE;
+ else if (flags & BTRFS_BLOCK_GROUP_RAID5)
- index = 5;
++ return BTRFS_RAID_RAID5;
+ else if (flags & BTRFS_BLOCK_GROUP_RAID6)
- index = 6;
- else
- index = 4; /* BTRFS_BLOCK_GROUP_SINGLE */
- return index;
++ return BTRFS_RAID_RAID6;
++
++ return BTRFS_RAID_SINGLE; /* BTRFS_BLOCK_GROUP_SINGLE */
}
static int get_block_group_index(struct btrfs_block_group_cache *cache)
diff --cc fs/btrfs/extent_io.h
index ff182322d112,b14b36a80eba..dc81868d975a
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@@ -72,9 -72,10 +72,9 @@@ struct extent_io_ops
int (*writepage_start_hook)(struct page *page, u64 start, u64 end);
int (*writepage_io_hook)(struct page *page, u64 start, u64 end);
extent_submit_bio_hook_t *submit_bio_hook;
- int (*merge_bio_hook)(struct page *page, unsigned long offset,
+ int (*merge_bio_hook)(int rw, struct page *page, unsigned long offset,
size_t size, struct bio *bio,
unsigned long bio_flags);
- int (*readpage_io_hook)(struct page *page, u64 start, u64 end);
int (*readpage_io_failed_hook)(struct page *page, int failed_mirror);
int (*readpage_end_io_hook)(struct page *page, u64 start, u64 end,
struct extent_state *state, int mirror);
diff --cc fs/btrfs/inode.c
index 1aa98be54ce0,492ee0ee8c64..4e6a11c2cfdd
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@@ -39,7 -39,7 +39,8 @@@
#include <linux/slab.h>
#include <linux/ratelimit.h>
#include <linux/mount.h>
+#include <linux/btrfs.h>
+ #include <linux/blkdev.h>
#include "compat.h"
#include "ctree.h"
#include "disk-io.h"
diff --cc fs/btrfs/transaction.c
index 955204ca0447,c56b9d436204..a83d486cc70c
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@@ -167,9 -156,11 +167,12 @@@ loop
spin_lock_init(&cur_trans->commit_lock);
spin_lock_init(&cur_trans->delayed_refs.lock);
+ atomic_set(&cur_trans->delayed_refs.procs_running_refs, 0);
+ atomic_set(&cur_trans->delayed_refs.ref_seq, 0);
+ init_waitqueue_head(&cur_trans->delayed_refs.wait);
INIT_LIST_HEAD(&cur_trans->pending_snapshots);
+ INIT_LIST_HEAD(&cur_trans->ordered_operations);
list_add_tail(&cur_trans->list, &fs_info->trans_list);
extent_io_tree_init(&cur_trans->dirty_pages,
fs_info->btree_inode->i_mapping);
diff --cc fs/btrfs/volumes.c
index 72b1cf1b2b5e,8818dc34c199..7992dc4ea4cc
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@@ -3130,27 -3092,26 +3154,29 @@@ int btrfs_balance(struct btrfs_balance_
/* allow to reduce meta or sys integrity only if force set */
allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 |
- BTRFS_BLOCK_GROUP_RAID10;
+ BTRFS_BLOCK_GROUP_RAID10 |
+ BTRFS_BLOCK_GROUP_RAID5 |
+ BTRFS_BLOCK_GROUP_RAID6;
-
- if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
- (fs_info->avail_system_alloc_bits & allowed) &&
- !(bctl->sys.target & allowed)) ||
- ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
- (fs_info->avail_metadata_alloc_bits & allowed) &&
- !(bctl->meta.target & allowed))) {
- if (bctl->flags & BTRFS_BALANCE_FORCE) {
- printk(KERN_INFO "btrfs: force reducing metadata "
- "integrity\n");
- } else {
- printk(KERN_ERR "btrfs: balance will reduce metadata "
- "integrity, use force if you want this\n");
- ret = -EINVAL;
- goto out;
+ do {
+ seq = read_seqbegin(&fs_info->profiles_lock);
+
+ if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
+ (fs_info->avail_system_alloc_bits & allowed) &&
+ !(bctl->sys.target & allowed)) ||
+ ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
+ (fs_info->avail_metadata_alloc_bits & allowed) &&
+ !(bctl->meta.target & allowed))) {
+ if (bctl->flags & BTRFS_BALANCE_FORCE) {
+ printk(KERN_INFO "btrfs: force reducing metadata "
+ "integrity\n");
+ } else {
+ printk(KERN_ERR "btrfs: balance will reduce metadata "
+ "integrity, use force if you want this\n");
+ ret = -EINVAL;
+ goto out;
+ }
}
- }
+ } while (read_seqretry(&fs_info->profiles_lock, seq));
if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
int num_tolerated_disk_barrier_failures;
@@@ -3199,16 -3169,9 +3225,11 @@@
btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
}
- if (cancel)
- __cancel_balance(fs_info);
+ if (bargs) {
+ memset(bargs, 0, sizeof(*bargs));
+ update_ioctl_balance_args(fs_info, 0, bargs);
+ }
- if ((ret && ret != -ECANCELED && ret != -ENOSPC) ||
- balance_need_close(fs_info)) {
- __cancel_balance(fs_info);
- }
-
wake_up(&fs_info->balance_wait_q);
return ret;
@@@ -3571,48 -3534,45 +3592,86 @@@ static int btrfs_cmp_device_info(const
}
struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
- /*
- * sub_stripes info for map,
- * dev_stripes -- stripes per dev, 2 for DUP, 1 other wise
- * devs_max -- max devices per stripe, 0 for unlimited
- * devs_min -- min devices per stripe
- * devs_increment -- ndevs must be a multiple of this
- * ncopies -- how many copies of the data we have
- */
- { 2, 1, 0, 4, 2, 2 /* raid10 */ },
- { 1, 1, 2, 2, 2, 2 /* raid1 */ },
- { 1, 2, 1, 1, 1, 2 /* dup */ },
- { 1, 1, 0, 2, 1, 1 /* raid0 */ },
- { 1, 1, 1, 1, 1, 1 /* single */ },
- { 1, 1, 0, 2, 1, 2 /* raid5 */ },
- { 1, 1, 0, 3, 1, 3 /* raid6 */ },
+ [BTRFS_RAID_RAID10] = {
+ .sub_stripes = 2,
+ .dev_stripes = 1,
+ .devs_max = 0, /* 0 == as many as possible */
+ .devs_min = 4,
+ .devs_increment = 2,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_RAID1] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 2,
+ .devs_min = 2,
+ .devs_increment = 2,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_DUP] = {
+ .sub_stripes = 1,
+ .dev_stripes = 2,
+ .devs_max = 1,
+ .devs_min = 1,
+ .devs_increment = 1,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_RAID0] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 0,
+ .devs_min = 2,
+ .devs_increment = 1,
+ .ncopies = 1,
+ },
+ [BTRFS_RAID_SINGLE] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 1,
+ .devs_min = 1,
+ .devs_increment = 1,
+ .ncopies = 1,
+ },
++ [BTRFS_RAID_RAID5] = {
++ .sub_stripes = 1,
++ .dev_stripes = 1,
++ .devs_max = 0,
++ .devs_min = 2,
++ .devs_increment = 1,
++ .ncopies = 2,
++ },
++ [BTRFS_RAID_RAID6] = {
++ .sub_stripes = 1,
++ .dev_stripes = 1,
++ .devs_max = 0,
++ .devs_min = 3,
++ .devs_increment = 1,
++ .ncopies = 3,
++ },
};
-
+
+ static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
+ {
+ /* TODO allow them to set a preferred stripe size */
+ return 64 * 1024;
+ }
+
+ 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");
+ }
+
static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
struct btrfs_root *extent_root,
struct map_lookup **map_ret,
@@@ -3840,19 -3819,14 +3917,21 @@@
info->chunk_root->root_key.objectid,
BTRFS_FIRST_CHUNK_TREE_OBJECTID,
start, dev_offset, stripe_size);
- if (ret) {
- btrfs_abort_transaction(trans, extent_root, ret);
- goto error;
- }
+ if (ret)
+ goto error_dev_extent;
}
+ ret = btrfs_make_block_group(trans, extent_root, 0, type,
+ BTRFS_FIRST_CHUNK_TREE_OBJECTID,
+ start, num_bytes);
+ if (ret) {
+ i = map->num_stripes - 1;
+ goto error_dev_extent;
+ }
+
+ free_extent_map(em);
+ check_raid56_incompat_flag(extent_root->fs_info, type);
+
kfree(devices_info);
return 0;