blob: 6b65fab27a1a2c6daca568c0db675118133c056d [file] [log] [blame]
/*
* Copyright (C) 2007 Oracle. All rights reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public
* License v2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public
* License along with this program; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*/
#include <linux/kernel.h>
#include <linux/bio.h>
#include <linux/buffer_head.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/highmem.h>
#include <linux/time.h>
#include <linux/init.h>
#include <linux/string.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
#include <linux/aio.h>
#include <linux/bit_spinlock.h>
#include <linux/xattr.h>
#include <linux/posix_acl.h>
#include <linux/falloc.h>
#include <linux/slab.h>
#include <linux/ratelimit.h>
#include <linux/mount.h>
#include <linux/btrfs.h>
#include <linux/blkdev.h>
#include <linux/posix_acl_xattr.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "btrfs_inode.h"
#include "print-tree.h"
#include "ordered-data.h"
#include "xattr.h"
#include "tree-log.h"
#include "volumes.h"
#include "compression.h"
#include "locking.h"
#include "free-space-cache.h"
#include "inode-map.h"
#include "backref.h"
#include "hash.h"
#include "props.h"
struct btrfs_iget_args {
struct btrfs_key *location;
struct btrfs_root *root;
};
static const struct inode_operations btrfs_dir_inode_operations;
static const struct inode_operations btrfs_symlink_inode_operations;
static const struct inode_operations btrfs_dir_ro_inode_operations;
static const struct inode_operations btrfs_special_inode_operations;
static const struct inode_operations btrfs_file_inode_operations;
static const struct address_space_operations btrfs_aops;
static const struct address_space_operations btrfs_symlink_aops;
static const struct file_operations btrfs_dir_file_operations;
static struct extent_io_ops btrfs_extent_io_ops;
static struct kmem_cache *btrfs_inode_cachep;
static struct kmem_cache *btrfs_delalloc_work_cachep;
struct kmem_cache *btrfs_trans_handle_cachep;
struct kmem_cache *btrfs_transaction_cachep;
struct kmem_cache *btrfs_path_cachep;
struct kmem_cache *btrfs_free_space_cachep;
#define S_SHIFT 12
static unsigned char btrfs_type_by_mode[S_IFMT >> S_SHIFT] = {
[S_IFREG >> S_SHIFT] = BTRFS_FT_REG_FILE,
[S_IFDIR >> S_SHIFT] = BTRFS_FT_DIR,
[S_IFCHR >> S_SHIFT] = BTRFS_FT_CHRDEV,
[S_IFBLK >> S_SHIFT] = BTRFS_FT_BLKDEV,
[S_IFIFO >> S_SHIFT] = BTRFS_FT_FIFO,
[S_IFSOCK >> S_SHIFT] = BTRFS_FT_SOCK,
[S_IFLNK >> S_SHIFT] = BTRFS_FT_SYMLINK,
};
static int btrfs_setsize(struct inode *inode, struct iattr *attr);
static int btrfs_truncate(struct inode *inode);
static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent);
static noinline int cow_file_range(struct inode *inode,
struct page *locked_page,
u64 start, u64 end, int *page_started,
unsigned long *nr_written, int unlock);
static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
u64 len, u64 orig_start,
u64 block_start, u64 block_len,
u64 orig_block_len, u64 ram_bytes,
int type);
static int btrfs_dirty_inode(struct inode *inode);
static int btrfs_init_inode_security(struct btrfs_trans_handle *trans,
struct inode *inode, struct inode *dir,
const struct qstr *qstr)
{
int err;
err = btrfs_init_acl(trans, inode, dir);
if (!err)
err = btrfs_xattr_security_init(trans, inode, dir, qstr);
return err;
}
/*
* this does all the hard work for inserting an inline extent into
* the btree. The caller should have done a btrfs_drop_extents so that
* no overlapping inline items exist in the btree
*/
static int insert_inline_extent(struct btrfs_trans_handle *trans,
struct btrfs_path *path, int extent_inserted,
struct btrfs_root *root, struct inode *inode,
u64 start, size_t size, size_t compressed_size,
int compress_type,
struct page **compressed_pages)
{
struct extent_buffer *leaf;
struct page *page = NULL;
char *kaddr;
unsigned long ptr;
struct btrfs_file_extent_item *ei;
int err = 0;
int ret;
size_t cur_size = size;
unsigned long offset;
if (compressed_size && compressed_pages)
cur_size = compressed_size;
inode_add_bytes(inode, size);
if (!extent_inserted) {
struct btrfs_key key;
size_t datasize;
key.objectid = btrfs_ino(inode);
key.offset = start;
btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
datasize = btrfs_file_extent_calc_inline_size(cur_size);
path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, root, path, &key,
datasize);
if (ret) {
err = ret;
goto fail;
}
}
leaf = path->nodes[0];
ei = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
btrfs_set_file_extent_generation(leaf, ei, trans->transid);
btrfs_set_file_extent_type(leaf, ei, BTRFS_FILE_EXTENT_INLINE);
btrfs_set_file_extent_encryption(leaf, ei, 0);
btrfs_set_file_extent_other_encoding(leaf, ei, 0);
btrfs_set_file_extent_ram_bytes(leaf, ei, size);
ptr = btrfs_file_extent_inline_start(ei);
if (compress_type != BTRFS_COMPRESS_NONE) {
struct page *cpage;
int i = 0;
while (compressed_size > 0) {
cpage = compressed_pages[i];
cur_size = min_t(unsigned long, compressed_size,
PAGE_CACHE_SIZE);
kaddr = kmap_atomic(cpage);
write_extent_buffer(leaf, kaddr, ptr, cur_size);
kunmap_atomic(kaddr);
i++;
ptr += cur_size;
compressed_size -= cur_size;
}
btrfs_set_file_extent_compression(leaf, ei,
compress_type);
} else {
page = find_get_page(inode->i_mapping,
start >> PAGE_CACHE_SHIFT);
btrfs_set_file_extent_compression(leaf, ei, 0);
kaddr = kmap_atomic(page);
offset = start & (PAGE_CACHE_SIZE - 1);
write_extent_buffer(leaf, kaddr + offset, ptr, size);
kunmap_atomic(kaddr);
page_cache_release(page);
}
btrfs_mark_buffer_dirty(leaf);
btrfs_release_path(path);
/*
* we're an inline extent, so nobody can
* extend the file past i_size without locking
* a page we already have locked.
*
* We must do any isize and inode updates
* before we unlock the pages. Otherwise we
* could end up racing with unlink.
*/
BTRFS_I(inode)->disk_i_size = inode->i_size;
ret = btrfs_update_inode(trans, root, inode);
return ret;
fail:
return err;
}
/*
* conditionally insert an inline extent into the file. This
* does the checks required to make sure the data is small enough
* to fit as an inline extent.
*/
static noinline int cow_file_range_inline(struct btrfs_root *root,
struct inode *inode, u64 start,
u64 end, size_t compressed_size,
int compress_type,
struct page **compressed_pages)
{
struct btrfs_trans_handle *trans;
u64 isize = i_size_read(inode);
u64 actual_end = min(end + 1, isize);
u64 inline_len = actual_end - start;
u64 aligned_end = ALIGN(end, root->sectorsize);
u64 data_len = inline_len;
int ret;
struct btrfs_path *path;
int extent_inserted = 0;
u32 extent_item_size;
if (compressed_size)
data_len = compressed_size;
if (start > 0 ||
actual_end >= PAGE_CACHE_SIZE ||
data_len >= BTRFS_MAX_INLINE_DATA_SIZE(root) ||
(!compressed_size &&
(actual_end & (root->sectorsize - 1)) == 0) ||
end + 1 < isize ||
data_len > root->fs_info->max_inline) {
return 1;
}
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
btrfs_free_path(path);
return PTR_ERR(trans);
}
trans->block_rsv = &root->fs_info->delalloc_block_rsv;
if (compressed_size && compressed_pages)
extent_item_size = btrfs_file_extent_calc_inline_size(
compressed_size);
else
extent_item_size = btrfs_file_extent_calc_inline_size(
inline_len);
ret = __btrfs_drop_extents(trans, root, inode, path,
start, aligned_end, NULL,
1, 1, extent_item_size, &extent_inserted);
if (ret) {
btrfs_abort_transaction(trans, root, ret);
goto out;
}
if (isize > actual_end)
inline_len = min_t(u64, isize, actual_end);
ret = insert_inline_extent(trans, path, extent_inserted,
root, inode, start,
inline_len, compressed_size,
compress_type, compressed_pages);
if (ret && ret != -ENOSPC) {
btrfs_abort_transaction(trans, root, ret);
goto out;
} else if (ret == -ENOSPC) {
ret = 1;
goto out;
}
set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags);
btrfs_delalloc_release_metadata(inode, end + 1 - start);
btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
out:
btrfs_free_path(path);
btrfs_end_transaction(trans, root);
return ret;
}
struct async_extent {
u64 start;
u64 ram_size;
u64 compressed_size;
struct page **pages;
unsigned long nr_pages;
int compress_type;
struct list_head list;
};
struct async_cow {
struct inode *inode;
struct btrfs_root *root;
struct page *locked_page;
u64 start;
u64 end;
struct list_head extents;
struct btrfs_work work;
};
static noinline int add_async_extent(struct async_cow *cow,
u64 start, u64 ram_size,
u64 compressed_size,
struct page **pages,
unsigned long nr_pages,
int compress_type)
{
struct async_extent *async_extent;
async_extent = kmalloc(sizeof(*async_extent), GFP_NOFS);
BUG_ON(!async_extent); /* -ENOMEM */
async_extent->start = start;
async_extent->ram_size = ram_size;
async_extent->compressed_size = compressed_size;
async_extent->pages = pages;
async_extent->nr_pages = nr_pages;
async_extent->compress_type = compress_type;
list_add_tail(&async_extent->list, &cow->extents);
return 0;
}
/*
* we create compressed extents in two phases. The first
* phase compresses a range of pages that have already been
* locked (both pages and state bits are locked).
*
* This is done inside an ordered work queue, and the compression
* is spread across many cpus. The actual IO submission is step
* two, and the ordered work queue takes care of making sure that
* happens in the same order things were put onto the queue by
* writepages and friends.
*
* If this code finds it can't get good compression, it puts an
* entry onto the work queue to write the uncompressed bytes. This
* makes sure that both compressed inodes and uncompressed inodes
* are written in the same order that the flusher thread sent them
* down.
*/
static noinline int compress_file_range(struct inode *inode,
struct page *locked_page,
u64 start, u64 end,
struct async_cow *async_cow,
int *num_added)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 num_bytes;
u64 blocksize = root->sectorsize;
u64 actual_end;
u64 isize = i_size_read(inode);
int ret = 0;
struct page **pages = NULL;
unsigned long nr_pages;
unsigned long nr_pages_ret = 0;
unsigned long total_compressed = 0;
unsigned long total_in = 0;
unsigned long max_compressed = 128 * 1024;
unsigned long max_uncompressed = 128 * 1024;
int i;
int will_compress;
int compress_type = root->fs_info->compress_type;
int redirty = 0;
/* if this is a small write inside eof, kick off a defrag */
if ((end - start + 1) < 16 * 1024 &&
(start > 0 || end + 1 < BTRFS_I(inode)->disk_i_size))
btrfs_add_inode_defrag(NULL, inode);
/*
* skip compression for a small file range(<=blocksize) that
* isn't an inline extent, since it dosen't save disk space at all.
*/
if ((end - start + 1) <= blocksize &&
(start > 0 || end + 1 < BTRFS_I(inode)->disk_i_size))
goto cleanup_and_bail_uncompressed;
actual_end = min_t(u64, isize, end + 1);
again:
will_compress = 0;
nr_pages = (end >> PAGE_CACHE_SHIFT) - (start >> PAGE_CACHE_SHIFT) + 1;
nr_pages = min(nr_pages, (128 * 1024UL) / PAGE_CACHE_SIZE);
/*
* we don't want to send crud past the end of i_size through
* compression, that's just a waste of CPU time. So, if the
* end of the file is before the start of our current
* requested range of bytes, we bail out to the uncompressed
* cleanup code that can deal with all of this.
*
* It isn't really the fastest way to fix things, but this is a
* very uncommon corner.
*/
if (actual_end <= start)
goto cleanup_and_bail_uncompressed;
total_compressed = actual_end - start;
/* we want to make sure that amount of ram required to uncompress
* an extent is reasonable, so we limit the total size in ram
* of a compressed extent to 128k. This is a crucial number
* because it also controls how easily we can spread reads across
* cpus for decompression.
*
* We also want to make sure the amount of IO required to do
* a random read is reasonably small, so we limit the size of
* a compressed extent to 128k.
*/
total_compressed = min(total_compressed, max_uncompressed);
num_bytes = ALIGN(end - start + 1, blocksize);
num_bytes = max(blocksize, num_bytes);
total_in = 0;
ret = 0;
/*
* we do compression for mount -o compress and when the
* inode has not been flagged as nocompress. This flag can
* change at any time if we discover bad compression ratios.
*/
if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS) &&
(btrfs_test_opt(root, COMPRESS) ||
(BTRFS_I(inode)->force_compress) ||
(BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS))) {
WARN_ON(pages);
pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS);
if (!pages) {
/* just bail out to the uncompressed code */
goto cont;
}
if (BTRFS_I(inode)->force_compress)
compress_type = BTRFS_I(inode)->force_compress;
/*
* we need to call clear_page_dirty_for_io on each
* page in the range. Otherwise applications with the file
* mmap'd can wander in and change the page contents while
* we are compressing them.
*
* If the compression fails for any reason, we set the pages
* dirty again later on.
*/
extent_range_clear_dirty_for_io(inode, start, end);
redirty = 1;
ret = btrfs_compress_pages(compress_type,
inode->i_mapping, start,
total_compressed, pages,
nr_pages, &nr_pages_ret,
&total_in,
&total_compressed,
max_compressed);
if (!ret) {
unsigned long offset = total_compressed &
(PAGE_CACHE_SIZE - 1);
struct page *page = pages[nr_pages_ret - 1];
char *kaddr;
/* zero the tail end of the last page, we might be
* sending it down to disk
*/
if (offset) {
kaddr = kmap_atomic(page);
memset(kaddr + offset, 0,
PAGE_CACHE_SIZE - offset);
kunmap_atomic(kaddr);
}
will_compress = 1;
}
}
cont:
if (start == 0) {
/* lets try to make an inline extent */
if (ret || total_in < (actual_end - start)) {
/* we didn't compress the entire range, try
* to make an uncompressed inline extent.
*/
ret = cow_file_range_inline(root, inode, start, end,
0, 0, NULL);
} else {
/* try making a compressed inline extent */
ret = cow_file_range_inline(root, inode, start, end,
total_compressed,
compress_type, pages);
}
if (ret <= 0) {
unsigned long clear_flags = EXTENT_DELALLOC |
EXTENT_DEFRAG;
clear_flags |= (ret < 0) ? EXTENT_DO_ACCOUNTING : 0;
/*
* inline extent creation worked or returned error,
* we don't need to create any more async work items.
* Unlock and free up our temp pages.
*/
extent_clear_unlock_delalloc(inode, start, end, NULL,
clear_flags, PAGE_UNLOCK |
PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK |
PAGE_END_WRITEBACK);
goto free_pages_out;
}
}
if (will_compress) {
/*
* we aren't doing an inline extent round the compressed size
* up to a block size boundary so the allocator does sane
* things
*/
total_compressed = ALIGN(total_compressed, blocksize);
/*
* one last check to make sure the compression is really a
* win, compare the page count read with the blocks on disk
*/
total_in = ALIGN(total_in, PAGE_CACHE_SIZE);
if (total_compressed >= total_in) {
will_compress = 0;
} else {
num_bytes = total_in;
}
}
if (!will_compress && pages) {
/*
* the compression code ran but failed to make things smaller,
* free any pages it allocated and our page pointer array
*/
for (i = 0; i < nr_pages_ret; i++) {
WARN_ON(pages[i]->mapping);
page_cache_release(pages[i]);
}
kfree(pages);
pages = NULL;
total_compressed = 0;
nr_pages_ret = 0;
/* flag the file so we don't compress in the future */
if (!btrfs_test_opt(root, FORCE_COMPRESS) &&
!(BTRFS_I(inode)->force_compress)) {
BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
}
}
if (will_compress) {
*num_added += 1;
/* the async work queues will take care of doing actual
* allocation on disk for these compressed pages,
* and will submit them to the elevator.
*/
add_async_extent(async_cow, start, num_bytes,
total_compressed, pages, nr_pages_ret,
compress_type);
if (start + num_bytes < end) {
start += num_bytes;
pages = NULL;
cond_resched();
goto again;
}
} else {
cleanup_and_bail_uncompressed:
/*
* No compression, but we still need to write the pages in
* the file we've been given so far. redirty the locked
* page if it corresponds to our extent and set things up
* for the async work queue to run cow_file_range to do
* the normal delalloc dance
*/
if (page_offset(locked_page) >= start &&
page_offset(locked_page) <= end) {
__set_page_dirty_nobuffers(locked_page);
/* unlocked later on in the async handlers */
}
if (redirty)
extent_range_redirty_for_io(inode, start, end);
add_async_extent(async_cow, start, end - start + 1,
0, NULL, 0, BTRFS_COMPRESS_NONE);
*num_added += 1;
}
out:
return ret;
free_pages_out:
for (i = 0; i < nr_pages_ret; i++) {
WARN_ON(pages[i]->mapping);
page_cache_release(pages[i]);
}
kfree(pages);
goto out;
}
/*
* phase two of compressed writeback. This is the ordered portion
* of the code, which only gets called in the order the work was
* queued. We walk all the async extents created by compress_file_range
* and send them down to the disk.
*/
static noinline int submit_compressed_extents(struct inode *inode,
struct async_cow *async_cow)
{
struct async_extent *async_extent;
u64 alloc_hint = 0;
struct btrfs_key ins;
struct extent_map *em;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
struct extent_io_tree *io_tree;
int ret = 0;
if (list_empty(&async_cow->extents))
return 0;
again:
while (!list_empty(&async_cow->extents)) {
async_extent = list_entry(async_cow->extents.next,
struct async_extent, list);
list_del(&async_extent->list);
io_tree = &BTRFS_I(inode)->io_tree;
retry:
/* did the compression code fall back to uncompressed IO? */
if (!async_extent->pages) {
int page_started = 0;
unsigned long nr_written = 0;
lock_extent(io_tree, async_extent->start,
async_extent->start +
async_extent->ram_size - 1);
/* allocate blocks */
ret = cow_file_range(inode, async_cow->locked_page,
async_extent->start,
async_extent->start +
async_extent->ram_size - 1,
&page_started, &nr_written, 0);
/* JDM XXX */
/*
* if page_started, cow_file_range inserted an
* inline extent and took care of all the unlocking
* and IO for us. Otherwise, we need to submit
* all those pages down to the drive.
*/
if (!page_started && !ret)
extent_write_locked_range(io_tree,
inode, async_extent->start,
async_extent->start +
async_extent->ram_size - 1,
btrfs_get_extent,
WB_SYNC_ALL);
else if (ret)
unlock_page(async_cow->locked_page);
kfree(async_extent);
cond_resched();
continue;
}
lock_extent(io_tree, async_extent->start,
async_extent->start + async_extent->ram_size - 1);
ret = btrfs_reserve_extent(root,
async_extent->compressed_size,
async_extent->compressed_size,
0, alloc_hint, &ins, 1, 1);
if (ret) {
int i;
for (i = 0; i < async_extent->nr_pages; i++) {
WARN_ON(async_extent->pages[i]->mapping);
page_cache_release(async_extent->pages[i]);
}
kfree(async_extent->pages);
async_extent->nr_pages = 0;
async_extent->pages = NULL;
if (ret == -ENOSPC) {
unlock_extent(io_tree, async_extent->start,
async_extent->start +
async_extent->ram_size - 1);
goto retry;
}
goto out_free;
}
/*
* here we're doing allocation and writeback of the
* compressed pages
*/
btrfs_drop_extent_cache(inode, async_extent->start,
async_extent->start +
async_extent->ram_size - 1, 0);
em = alloc_extent_map();
if (!em) {
ret = -ENOMEM;
goto out_free_reserve;
}
em->start = async_extent->start;
em->len = async_extent->ram_size;
em->orig_start = em->start;
em->mod_start = em->start;
em->mod_len = em->len;
em->block_start = ins.objectid;
em->block_len = ins.offset;
em->orig_block_len = ins.offset;
em->ram_bytes = async_extent->ram_size;
em->bdev = root->fs_info->fs_devices->latest_bdev;
em->compress_type = async_extent->compress_type;
set_bit(EXTENT_FLAG_PINNED, &em->flags);
set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
em->generation = -1;
while (1) {
write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em, 1);
write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
}
btrfs_drop_extent_cache(inode, async_extent->start,
async_extent->start +
async_extent->ram_size - 1, 0);
}
if (ret)
goto out_free_reserve;
ret = btrfs_add_ordered_extent_compress(inode,
async_extent->start,
ins.objectid,
async_extent->ram_size,
ins.offset,
BTRFS_ORDERED_COMPRESSED,
async_extent->compress_type);
if (ret)
goto out_free_reserve;
/*
* clear dirty, set writeback and unlock the pages.
*/
extent_clear_unlock_delalloc(inode, async_extent->start,
async_extent->start +
async_extent->ram_size - 1,
NULL, EXTENT_LOCKED | EXTENT_DELALLOC,
PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK);
ret = btrfs_submit_compressed_write(inode,
async_extent->start,
async_extent->ram_size,
ins.objectid,
ins.offset, async_extent->pages,
async_extent->nr_pages);
alloc_hint = ins.objectid + ins.offset;
kfree(async_extent);
if (ret)
goto out;
cond_resched();
}
ret = 0;
out:
return ret;
out_free_reserve:
btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1);
out_free:
extent_clear_unlock_delalloc(inode, async_extent->start,
async_extent->start +
async_extent->ram_size - 1,
NULL, EXTENT_LOCKED | EXTENT_DELALLOC |
EXTENT_DEFRAG | EXTENT_DO_ACCOUNTING,
PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK);
kfree(async_extent);
goto again;
}
static u64 get_extent_allocation_hint(struct inode *inode, u64 start,
u64 num_bytes)
{
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
struct extent_map *em;
u64 alloc_hint = 0;
read_lock(&em_tree->lock);
em = search_extent_mapping(em_tree, start, num_bytes);
if (em) {
/*
* if block start isn't an actual block number then find the
* first block in this inode and use that as a hint. If that
* block is also bogus then just don't worry about it.
*/
if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
free_extent_map(em);
em = search_extent_mapping(em_tree, 0, 0);
if (em && em->block_start < EXTENT_MAP_LAST_BYTE)
alloc_hint = em->block_start;
if (em)
free_extent_map(em);
} else {
alloc_hint = em->block_start;
free_extent_map(em);
}
}
read_unlock(&em_tree->lock);
return alloc_hint;
}
/*
* when extent_io.c finds a delayed allocation range in the file,
* the call backs end up in this code. The basic idea is to
* allocate extents on disk for the range, and create ordered data structs
* in ram to track those extents.
*
* locked_page is the page that writepage had locked already. We use
* it to make sure we don't do extra locks or unlocks.
*
* *page_started is set to one if we unlock locked_page and do everything
* required to start IO on it. It may be clean and already done with
* IO when we return.
*/
static noinline int cow_file_range(struct inode *inode,
struct page *locked_page,
u64 start, u64 end, int *page_started,
unsigned long *nr_written,
int unlock)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 alloc_hint = 0;
u64 num_bytes;
unsigned long ram_size;
u64 disk_num_bytes;
u64 cur_alloc_size;
u64 blocksize = root->sectorsize;
struct btrfs_key ins;
struct extent_map *em;
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
int ret = 0;
if (btrfs_is_free_space_inode(inode)) {
WARN_ON_ONCE(1);
ret = -EINVAL;
goto out_unlock;
}
num_bytes = ALIGN(end - start + 1, blocksize);
num_bytes = max(blocksize, num_bytes);
disk_num_bytes = num_bytes;
/* if this is a small write inside eof, kick off defrag */
if (num_bytes < 64 * 1024 &&
(start > 0 || end + 1 < BTRFS_I(inode)->disk_i_size))
btrfs_add_inode_defrag(NULL, inode);
if (start == 0) {
/* lets try to make an inline extent */
ret = cow_file_range_inline(root, inode, start, end, 0, 0,
NULL);
if (ret == 0) {
extent_clear_unlock_delalloc(inode, start, end, NULL,
EXTENT_LOCKED | EXTENT_DELALLOC |
EXTENT_DEFRAG, PAGE_UNLOCK |
PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK |
PAGE_END_WRITEBACK);
*nr_written = *nr_written +
(end - start + PAGE_CACHE_SIZE) / PAGE_CACHE_SIZE;
*page_started = 1;
goto out;
} else if (ret < 0) {
goto out_unlock;
}
}
BUG_ON(disk_num_bytes >
btrfs_super_total_bytes(root->fs_info->super_copy));
alloc_hint = get_extent_allocation_hint(inode, start, num_bytes);
btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
while (disk_num_bytes > 0) {
unsigned long op;
cur_alloc_size = disk_num_bytes;
ret = btrfs_reserve_extent(root, cur_alloc_size,
root->sectorsize, 0, alloc_hint,
&ins, 1, 1);
if (ret < 0)
goto out_unlock;
em = alloc_extent_map();
if (!em) {
ret = -ENOMEM;
goto out_reserve;
}
em->start = start;
em->orig_start = em->start;
ram_size = ins.offset;
em->len = ins.offset;
em->mod_start = em->start;
em->mod_len = em->len;
em->block_start = ins.objectid;
em->block_len = ins.offset;
em->orig_block_len = ins.offset;
em->ram_bytes = ram_size;
em->bdev = root->fs_info->fs_devices->latest_bdev;
set_bit(EXTENT_FLAG_PINNED, &em->flags);
em->generation = -1;
while (1) {
write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em, 1);
write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
}
btrfs_drop_extent_cache(inode, start,
start + ram_size - 1, 0);
}
if (ret)
goto out_reserve;
cur_alloc_size = ins.offset;
ret = btrfs_add_ordered_extent(inode, start, ins.objectid,
ram_size, cur_alloc_size, 0);
if (ret)
goto out_reserve;
if (root->root_key.objectid ==
BTRFS_DATA_RELOC_TREE_OBJECTID) {
ret = btrfs_reloc_clone_csums(inode, start,
cur_alloc_size);
if (ret)
goto out_reserve;
}
if (disk_num_bytes < cur_alloc_size)
break;
/* we're not doing compressed IO, don't unlock the first
* page (which the caller expects to stay locked), don't
* clear any dirty bits and don't set any writeback bits
*
* Do set the Private2 bit so we know this page was properly
* setup for writepage
*/
op = unlock ? PAGE_UNLOCK : 0;
op |= PAGE_SET_PRIVATE2;
extent_clear_unlock_delalloc(inode, start,
start + ram_size - 1, locked_page,
EXTENT_LOCKED | EXTENT_DELALLOC,
op);
disk_num_bytes -= cur_alloc_size;
num_bytes -= cur_alloc_size;
alloc_hint = ins.objectid + ins.offset;
start += cur_alloc_size;
}
out:
return ret;
out_reserve:
btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1);
out_unlock:
extent_clear_unlock_delalloc(inode, start, end, locked_page,
EXTENT_LOCKED | EXTENT_DO_ACCOUNTING |
EXTENT_DELALLOC | EXTENT_DEFRAG,
PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK);
goto out;
}
/*
* work queue call back to started compression on a file and pages
*/
static noinline void async_cow_start(struct btrfs_work *work)
{
struct async_cow *async_cow;
int num_added = 0;
async_cow = container_of(work, struct async_cow, work);
compress_file_range(async_cow->inode, async_cow->locked_page,
async_cow->start, async_cow->end, async_cow,
&num_added);
if (num_added == 0) {
btrfs_add_delayed_iput(async_cow->inode);
async_cow->inode = NULL;
}
}
/*
* work queue call back to submit previously compressed pages
*/
static noinline void async_cow_submit(struct btrfs_work *work)
{
struct async_cow *async_cow;
struct btrfs_root *root;
unsigned long nr_pages;
async_cow = container_of(work, struct async_cow, work);
root = async_cow->root;
nr_pages = (async_cow->end - async_cow->start + PAGE_CACHE_SIZE) >>
PAGE_CACHE_SHIFT;
if (atomic_sub_return(nr_pages, &root->fs_info->async_delalloc_pages) <
5 * 1024 * 1024 &&
waitqueue_active(&root->fs_info->async_submit_wait))
wake_up(&root->fs_info->async_submit_wait);
if (async_cow->inode)
submit_compressed_extents(async_cow->inode, async_cow);
}
static noinline void async_cow_free(struct btrfs_work *work)
{
struct async_cow *async_cow;
async_cow = container_of(work, struct async_cow, work);
if (async_cow->inode)
btrfs_add_delayed_iput(async_cow->inode);
kfree(async_cow);
}
static int cow_file_range_async(struct inode *inode, struct page *locked_page,
u64 start, u64 end, int *page_started,
unsigned long *nr_written)
{
struct async_cow *async_cow;
struct btrfs_root *root = BTRFS_I(inode)->root;
unsigned long nr_pages;
u64 cur_end;
int limit = 10 * 1024 * 1024;
clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED,
1, 0, NULL, GFP_NOFS);
while (start < end) {
async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS);
BUG_ON(!async_cow); /* -ENOMEM */
async_cow->inode = igrab(inode);
async_cow->root = root;
async_cow->locked_page = locked_page;
async_cow->start = start;
if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
cur_end = end;
else
cur_end = min(end, start + 512 * 1024 - 1);
async_cow->end = cur_end;
INIT_LIST_HEAD(&async_cow->extents);
btrfs_init_work(&async_cow->work, async_cow_start,
async_cow_submit, async_cow_free);
nr_pages = (cur_end - start + PAGE_CACHE_SIZE) >>
PAGE_CACHE_SHIFT;
atomic_add(nr_pages, &root->fs_info->async_delalloc_pages);
btrfs_queue_work(root->fs_info->delalloc_workers,
&async_cow->work);
if (atomic_read(&root->fs_info->async_delalloc_pages) > limit) {
wait_event(root->fs_info->async_submit_wait,
(atomic_read(&root->fs_info->async_delalloc_pages) <
limit));
}
while (atomic_read(&root->fs_info->async_submit_draining) &&
atomic_read(&root->fs_info->async_delalloc_pages)) {
wait_event(root->fs_info->async_submit_wait,
(atomic_read(&root->fs_info->async_delalloc_pages) ==
0));
}
*nr_written += nr_pages;
start = cur_end + 1;
}
*page_started = 1;
return 0;
}
static noinline int csum_exist_in_range(struct btrfs_root *root,
u64 bytenr, u64 num_bytes)
{
int ret;
struct btrfs_ordered_sum *sums;
LIST_HEAD(list);
ret = btrfs_lookup_csums_range(root->fs_info->csum_root, bytenr,
bytenr + num_bytes - 1, &list, 0);
if (ret == 0 && list_empty(&list))
return 0;
while (!list_empty(&list)) {
sums = list_entry(list.next, struct btrfs_ordered_sum, list);
list_del(&sums->list);
kfree(sums);
}
return 1;
}
/*
* when nowcow writeback call back. This checks for snapshots or COW copies
* of the extents that exist in the file, and COWs the file as required.
*
* If no cow copies or snapshots exist, we write directly to the existing
* blocks on disk
*/
static noinline int run_delalloc_nocow(struct inode *inode,
struct page *locked_page,
u64 start, u64 end, int *page_started, int force,
unsigned long *nr_written)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_trans_handle *trans;
struct extent_buffer *leaf;
struct btrfs_path *path;
struct btrfs_file_extent_item *fi;
struct btrfs_key found_key;
u64 cow_start;
u64 cur_offset;
u64 extent_end;
u64 extent_offset;
u64 disk_bytenr;
u64 num_bytes;
u64 disk_num_bytes;
u64 ram_bytes;
int extent_type;
int ret, err;
int type;
int nocow;
int check_prev = 1;
bool nolock;
u64 ino = btrfs_ino(inode);
path = btrfs_alloc_path();
if (!path) {
extent_clear_unlock_delalloc(inode, start, end, locked_page,
EXTENT_LOCKED | EXTENT_DELALLOC |
EXTENT_DO_ACCOUNTING |
EXTENT_DEFRAG, PAGE_UNLOCK |
PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK |
PAGE_END_WRITEBACK);
return -ENOMEM;
}
nolock = btrfs_is_free_space_inode(inode);
if (nolock)
trans = btrfs_join_transaction_nolock(root);
else
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
extent_clear_unlock_delalloc(inode, start, end, locked_page,
EXTENT_LOCKED | EXTENT_DELALLOC |
EXTENT_DO_ACCOUNTING |
EXTENT_DEFRAG, PAGE_UNLOCK |
PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK |
PAGE_END_WRITEBACK);
btrfs_free_path(path);
return PTR_ERR(trans);
}
trans->block_rsv = &root->fs_info->delalloc_block_rsv;
cow_start = (u64)-1;
cur_offset = start;
while (1) {
ret = btrfs_lookup_file_extent(trans, root, path, ino,
cur_offset, 0);
if (ret < 0)
goto error;
if (ret > 0 && path->slots[0] > 0 && check_prev) {
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key,
path->slots[0] - 1);
if (found_key.objectid == ino &&
found_key.type == BTRFS_EXTENT_DATA_KEY)
path->slots[0]--;
}
check_prev = 0;
next_slot:
leaf = path->nodes[0];
if (path->slots[0] >= btrfs_header_nritems(leaf)) {
ret = btrfs_next_leaf(root, path);
if (ret < 0)
goto error;
if (ret > 0)
break;
leaf = path->nodes[0];
}
nocow = 0;
disk_bytenr = 0;
num_bytes = 0;
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
if (found_key.objectid > ino ||
found_key.type > BTRFS_EXTENT_DATA_KEY ||
found_key.offset > end)
break;
if (found_key.offset > cur_offset) {
extent_end = found_key.offset;
extent_type = 0;
goto out_check;
}
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
extent_type = btrfs_file_extent_type(leaf, fi);
ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
if (extent_type == BTRFS_FILE_EXTENT_REG ||
extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
extent_offset = btrfs_file_extent_offset(leaf, fi);
extent_end = found_key.offset +
btrfs_file_extent_num_bytes(leaf, fi);
disk_num_bytes =
btrfs_file_extent_disk_num_bytes(leaf, fi);
if (extent_end <= start) {
path->slots[0]++;
goto next_slot;
}
if (disk_bytenr == 0)
goto out_check;
if (btrfs_file_extent_compression(leaf, fi) ||
btrfs_file_extent_encryption(leaf, fi) ||
btrfs_file_extent_other_encoding(leaf, fi))
goto out_check;
if (extent_type == BTRFS_FILE_EXTENT_REG && !force)
goto out_check;
if (btrfs_extent_readonly(root, disk_bytenr))
goto out_check;
if (btrfs_cross_ref_exist(trans, root, ino,
found_key.offset -
extent_offset, disk_bytenr))
goto out_check;
disk_bytenr += extent_offset;
disk_bytenr += cur_offset - found_key.offset;
num_bytes = min(end + 1, extent_end) - cur_offset;
/*
* if there are pending snapshots for this root,
* we fall into common COW way.
*/
if (!nolock) {
err = btrfs_start_nocow_write(root);
if (!err)
goto out_check;
}
/*
* force cow if csum exists in the range.
* this ensure that csum for a given extent are
* either valid or do not exist.
*/
if (csum_exist_in_range(root, disk_bytenr, num_bytes))
goto out_check;
nocow = 1;
} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
extent_end = found_key.offset +
btrfs_file_extent_inline_len(leaf,
path->slots[0], fi);
extent_end = ALIGN(extent_end, root->sectorsize);
} else {
BUG_ON(1);
}
out_check:
if (extent_end <= start) {
path->slots[0]++;
if (!nolock && nocow)
btrfs_end_nocow_write(root);
goto next_slot;
}
if (!nocow) {
if (cow_start == (u64)-1)
cow_start = cur_offset;
cur_offset = extent_end;
if (cur_offset > end)
break;
path->slots[0]++;
goto next_slot;
}
btrfs_release_path(path);
if (cow_start != (u64)-1) {
ret = cow_file_range(inode, locked_page,
cow_start, found_key.offset - 1,
page_started, nr_written, 1);
if (ret) {
if (!nolock && nocow)
btrfs_end_nocow_write(root);
goto error;
}
cow_start = (u64)-1;
}
if (extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
struct extent_map *em;
struct extent_map_tree *em_tree;
em_tree = &BTRFS_I(inode)->extent_tree;
em = alloc_extent_map();
BUG_ON(!em); /* -ENOMEM */
em->start = cur_offset;
em->orig_start = found_key.offset - extent_offset;
em->len = num_bytes;
em->block_len = num_bytes;
em->block_start = disk_bytenr;
em->orig_block_len = disk_num_bytes;
em->ram_bytes = ram_bytes;
em->bdev = root->fs_info->fs_devices->latest_bdev;
em->mod_start = em->start;
em->mod_len = em->len;
set_bit(EXTENT_FLAG_PINNED, &em->flags);
set_bit(EXTENT_FLAG_FILLING, &em->flags);
em->generation = -1;
while (1) {
write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em, 1);
write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
}
btrfs_drop_extent_cache(inode, em->start,
em->start + em->len - 1, 0);
}
type = BTRFS_ORDERED_PREALLOC;
} else {
type = BTRFS_ORDERED_NOCOW;
}
ret = btrfs_add_ordered_extent(inode, cur_offset, disk_bytenr,
num_bytes, num_bytes, type);
BUG_ON(ret); /* -ENOMEM */
if (root->root_key.objectid ==
BTRFS_DATA_RELOC_TREE_OBJECTID) {
ret = btrfs_reloc_clone_csums(inode, cur_offset,
num_bytes);
if (ret) {
if (!nolock && nocow)
btrfs_end_nocow_write(root);
goto error;
}
}
extent_clear_unlock_delalloc(inode, cur_offset,
cur_offset + num_bytes - 1,
locked_page, EXTENT_LOCKED |
EXTENT_DELALLOC, PAGE_UNLOCK |
PAGE_SET_PRIVATE2);
if (!nolock && nocow)
btrfs_end_nocow_write(root);
cur_offset = extent_end;
if (cur_offset > end)
break;
}
btrfs_release_path(path);
if (cur_offset <= end && cow_start == (u64)-1) {
cow_start = cur_offset;
cur_offset = end;
}
if (cow_start != (u64)-1) {
ret = cow_file_range(inode, locked_page, cow_start, end,
page_started, nr_written, 1);
if (ret)
goto error;
}
error:
err = btrfs_end_transaction(trans, root);
if (!ret)
ret = err;
if (ret && cur_offset < end)
extent_clear_unlock_delalloc(inode, cur_offset, end,
locked_page, EXTENT_LOCKED |
EXTENT_DELALLOC | EXTENT_DEFRAG |
EXTENT_DO_ACCOUNTING, PAGE_UNLOCK |
PAGE_CLEAR_DIRTY |
PAGE_SET_WRITEBACK |
PAGE_END_WRITEBACK);
btrfs_free_path(path);
return ret;
}
/*
* extent_io.c call back to do delayed allocation processing
*/
static int run_delalloc_range(struct inode *inode, struct page *locked_page,
u64 start, u64 end, int *page_started,
unsigned long *nr_written)
{
int ret;
struct btrfs_root *root = BTRFS_I(inode)->root;
if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW) {
ret = run_delalloc_nocow(inode, locked_page, start, end,
page_started, 1, nr_written);
} else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC) {
ret = run_delalloc_nocow(inode, locked_page, start, end,
page_started, 0, nr_written);
} else if (!btrfs_test_opt(root, COMPRESS) &&
!(BTRFS_I(inode)->force_compress) &&
!(BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS)) {
ret = cow_file_range(inode, locked_page, start, end,
page_started, nr_written, 1);
} else {
set_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
&BTRFS_I(inode)->runtime_flags);
ret = cow_file_range_async(inode, locked_page, start, end,
page_started, nr_written);
}
return ret;
}
static void btrfs_split_extent_hook(struct inode *inode,
struct extent_state *orig, u64 split)
{
/* not delalloc, ignore it */
if (!(orig->state & EXTENT_DELALLOC))
return;
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents++;
spin_unlock(&BTRFS_I(inode)->lock);
}
/*
* extent_io.c merge_extent_hook, used to track merged delayed allocation
* extents so we can keep track of new extents that are just merged onto old
* extents, such as when we are doing sequential writes, so we can properly
* account for the metadata space we'll need.
*/
static void btrfs_merge_extent_hook(struct inode *inode,
struct extent_state *new,
struct extent_state *other)
{
/* not delalloc, ignore it */
if (!(other->state & EXTENT_DELALLOC))
return;
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents--;
spin_unlock(&BTRFS_I(inode)->lock);
}
static void btrfs_add_delalloc_inodes(struct btrfs_root *root,
struct inode *inode)
{
spin_lock(&root->delalloc_lock);
if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
&root->delalloc_inodes);
set_bit(BTRFS_INODE_IN_DELALLOC_LIST,
&BTRFS_I(inode)->runtime_flags);
root->nr_delalloc_inodes++;
if (root->nr_delalloc_inodes == 1) {
spin_lock(&root->fs_info->delalloc_root_lock);
BUG_ON(!list_empty(&root->delalloc_root));
list_add_tail(&root->delalloc_root,
&root->fs_info->delalloc_roots);
spin_unlock(&root->fs_info->delalloc_root_lock);
}
}
spin_unlock(&root->delalloc_lock);
}
static void btrfs_del_delalloc_inode(struct btrfs_root *root,
struct inode *inode)
{
spin_lock(&root->delalloc_lock);
if (!list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
list_del_init(&BTRFS_I(inode)->delalloc_inodes);
clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
&BTRFS_I(inode)->runtime_flags);
root->nr_delalloc_inodes--;
if (!root->nr_delalloc_inodes) {
spin_lock(&root->fs_info->delalloc_root_lock);
BUG_ON(list_empty(&root->delalloc_root));
list_del_init(&root->delalloc_root);
spin_unlock(&root->fs_info->delalloc_root_lock);
}
}
spin_unlock(&root->delalloc_lock);
}
/*
* extent_io.c set_bit_hook, used to track delayed allocation
* bytes in this file, and to maintain the list of inodes that
* have pending delalloc work to be done.
*/
static void btrfs_set_bit_hook(struct inode *inode,
struct extent_state *state, unsigned long *bits)
{
/*
* set_bit and clear bit hooks normally require _irqsave/restore
* but in this case, we are only testing for the DELALLOC
* bit, which is only set or cleared with irqs on
*/
if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 len = state->end + 1 - state->start;
bool do_list = !btrfs_is_free_space_inode(inode);
if (*bits & EXTENT_FIRST_DELALLOC) {
*bits &= ~EXTENT_FIRST_DELALLOC;
} else {
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents++;
spin_unlock(&BTRFS_I(inode)->lock);
}
__percpu_counter_add(&root->fs_info->delalloc_bytes, len,
root->fs_info->delalloc_batch);
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->delalloc_bytes += len;
if (do_list && !test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
&BTRFS_I(inode)->runtime_flags))
btrfs_add_delalloc_inodes(root, inode);
spin_unlock(&BTRFS_I(inode)->lock);
}
}
/*
* extent_io.c clear_bit_hook, see set_bit_hook for why
*/
static void btrfs_clear_bit_hook(struct inode *inode,
struct extent_state *state,
unsigned long *bits)
{
/*
* set_bit and clear bit hooks normally require _irqsave/restore
* but in this case, we are only testing for the DELALLOC
* bit, which is only set or cleared with irqs on
*/
if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 len = state->end + 1 - state->start;
bool do_list = !btrfs_is_free_space_inode(inode);
if (*bits & EXTENT_FIRST_DELALLOC) {
*bits &= ~EXTENT_FIRST_DELALLOC;
} else if (!(*bits & EXTENT_DO_ACCOUNTING)) {
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents--;
spin_unlock(&BTRFS_I(inode)->lock);
}
/*
* We don't reserve metadata space for space cache inodes so we
* don't need to call dellalloc_release_metadata if there is an
* error.
*/
if (*bits & EXTENT_DO_ACCOUNTING &&
root != root->fs_info->tree_root)
btrfs_delalloc_release_metadata(inode, len);
if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
&& do_list && !(state->state & EXTENT_NORESERVE))
btrfs_free_reserved_data_space(inode, len);
__percpu_counter_add(&root->fs_info->delalloc_bytes, -len,
root->fs_info->delalloc_batch);
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->delalloc_bytes -= len;
if (do_list && BTRFS_I(inode)->delalloc_bytes == 0 &&
test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
&BTRFS_I(inode)->runtime_flags))
btrfs_del_delalloc_inode(root, inode);
spin_unlock(&BTRFS_I(inode)->lock);
}
}
/*
* extent_io.c merge_bio_hook, this must check the chunk tree to make sure
* we don't create bios that span stripes or chunks
*/
int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset,
size_t size, struct bio *bio,
unsigned long bio_flags)
{
struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
u64 logical = (u64)bio->bi_iter.bi_sector << 9;
u64 length = 0;
u64 map_length;
int ret;
if (bio_flags & EXTENT_BIO_COMPRESSED)
return 0;
length = bio->bi_iter.bi_size;
map_length = length;
ret = btrfs_map_block(root->fs_info, rw, logical,
&map_length, NULL, 0);
/* Will always return 0 with map_multi == NULL */
BUG_ON(ret < 0);
if (map_length < length + size)
return 1;
return 0;
}
/*
* in order to insert checksums into the metadata in large chunks,
* we wait until bio submission time. All the pages in the bio are
* checksummed and sums are attached onto the ordered extent record.
*
* At IO completion time the cums attached on the ordered extent record
* are inserted into the btree
*/
static int __btrfs_submit_bio_start(struct inode *inode, int rw,
struct bio *bio, int mirror_num,
unsigned long bio_flags,
u64 bio_offset)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret = 0;
ret = btrfs_csum_one_bio(root, inode, bio, 0, 0);
BUG_ON(ret); /* -ENOMEM */
return 0;
}
/*
* in order to insert checksums into the metadata in large chunks,
* we wait until bio submission time. All the pages in the bio are
* checksummed and sums are attached onto the ordered extent record.
*
* At IO completion time the cums attached on the ordered extent record
* are inserted into the btree
*/
static int __btrfs_submit_bio_done(struct inode *inode, int rw, struct bio *bio,
int mirror_num, unsigned long bio_flags,
u64 bio_offset)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret;
ret = btrfs_map_bio(root, rw, bio, mirror_num, 1);
if (ret)
bio_endio(bio, ret);
return ret;
}
/*
* extent_io.c submission hook. This does the right thing for csum calculation
* on write, or reading the csums from the tree before a read
*/
static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
int mirror_num, unsigned long bio_flags,
u64 bio_offset)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret = 0;
int skip_sum;
int metadata = 0;
int async = !atomic_read(&BTRFS_I(inode)->sync_writers);
skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
if (btrfs_is_free_space_inode(inode))
metadata = 2;
if (!(rw & REQ_WRITE)) {
ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
if (ret)
goto out;
if (bio_flags & EXTENT_BIO_COMPRESSED) {
ret = btrfs_submit_compressed_read(inode, bio,
mirror_num,
bio_flags);
goto out;
} else if (!skip_sum) {
ret = btrfs_lookup_bio_sums(root, inode, bio, NULL);
if (ret)
goto out;
}
goto mapit;
} else if (async && !skip_sum) {
/* csum items have already been cloned */
if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
goto mapit;
/* we're doing a write, do the async checksumming */
ret = btrfs_wq_submit_bio(BTRFS_I(inode)->root->fs_info,
inode, rw, bio, mirror_num,
bio_flags, bio_offset,
__btrfs_submit_bio_start,
__btrfs_submit_bio_done);
goto out;
} else if (!skip_sum) {
ret = btrfs_csum_one_bio(root, inode, bio, 0, 0);
if (ret)
goto out;
}
mapit:
ret = btrfs_map_bio(root, rw, bio, mirror_num, 0);
out:
if (ret < 0)
bio_endio(bio, ret);
return ret;
}
/*
* given a list of ordered sums record them in the inode. This happens
* at IO completion time based on sums calculated at bio submission time.
*/
static noinline int add_pending_csums(struct btrfs_trans_handle *trans,
struct inode *inode, u64 file_offset,
struct list_head *list)
{
struct btrfs_ordered_sum *sum;
list_for_each_entry(sum, list, list) {
trans->adding_csums = 1;
btrfs_csum_file_blocks(trans,
BTRFS_I(inode)->root->fs_info->csum_root, sum);
trans->adding_csums = 0;
}
return 0;
}
int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
struct extent_state **cached_state)
{
WARN_ON((end & (PAGE_CACHE_SIZE - 1)) == 0);
return set_extent_delalloc(&BTRFS_I(inode)->io_tree, start, end,
cached_state, GFP_NOFS);
}
/* see btrfs_writepage_start_hook for details on why this is required */
struct btrfs_writepage_fixup {
struct page *page;
struct btrfs_work work;
};
static void btrfs_writepage_fixup_worker(struct btrfs_work *work)
{
struct btrfs_writepage_fixup *fixup;
struct btrfs_ordered_extent *ordered;
struct extent_state *cached_state = NULL;
struct page *page;
struct inode *inode;
u64 page_start;
u64 page_end;
int ret;
fixup = container_of(work, struct btrfs_writepage_fixup, work);
page = fixup->page;
again:
lock_page(page);
if (!page->mapping || !PageDirty(page) || !PageChecked(page)) {
ClearPageChecked(page);
goto out_page;
}
inode = page->mapping->host;
page_start = page_offset(page);
page_end = page_offset(page) + PAGE_CACHE_SIZE - 1;
lock_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end, 0,
&cached_state);
/* already ordered? We're done */
if (PagePrivate2(page))
goto out;
ordered = btrfs_lookup_ordered_extent(inode, page_start);
if (ordered) {
unlock_extent_cached(&BTRFS_I(inode)->io_tree, page_start,
page_end, &cached_state, GFP_NOFS);
unlock_page(page);
btrfs_start_ordered_extent(inode, ordered, 1);
btrfs_put_ordered_extent(ordered);
goto again;
}
ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE);
if (ret) {
mapping_set_error(page->mapping, ret);
end_extent_writepage(page, ret, page_start, page_end);
ClearPageChecked(page);
goto out;
}
btrfs_set_extent_delalloc(inode, page_start, page_end, &cached_state);
ClearPageChecked(page);
set_page_dirty(page);
out:
unlock_extent_cached(&BTRFS_I(inode)->io_tree, page_start, page_end,
&cached_state, GFP_NOFS);
out_page:
unlock_page(page);
page_cache_release(page);
kfree(fixup);
}
/*
* There are a few paths in the higher layers of the kernel that directly
* set the page dirty bit without asking the filesystem if it is a
* good idea. This causes problems because we want to make sure COW
* properly happens and the data=ordered rules are followed.
*
* In our case any range that doesn't have the ORDERED bit set
* hasn't been properly setup for IO. We kick off an async process
* to fix it up. The async helper will wait for ordered extents, set
* the delalloc bit and make it safe to write the page.
*/
static int btrfs_writepage_start_hook(struct page *page, u64 start, u64 end)
{
struct inode *inode = page->mapping->host;
struct btrfs_writepage_fixup *fixup;
struct btrfs_root *root = BTRFS_I(inode)->root;
/* this page is properly in the ordered list */
if (TestClearPagePrivate2(page))
return 0;
if (PageChecked(page))
return -EAGAIN;
fixup = kzalloc(sizeof(*fixup), GFP_NOFS);
if (!fixup)
return -EAGAIN;
SetPageChecked(page);
page_cache_get(page);
btrfs_init_work(&fixup->work, btrfs_writepage_fixup_worker, NULL, NULL);
fixup->page = page;
btrfs_queue_work(root->fs_info->fixup_workers, &fixup->work);
return -EBUSY;
}
static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
struct inode *inode, u64 file_pos,
u64 disk_bytenr, u64 disk_num_bytes,
u64 num_bytes, u64 ram_bytes,
u8 compression, u8 encryption,
u16 other_encoding, int extent_type)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_file_extent_item *fi;
struct btrfs_path *path;
struct extent_buffer *leaf;
struct btrfs_key ins;
int extent_inserted = 0;
int ret;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
/*
* we may be replacing one extent in the tree with another.
* The new extent is pinned in the extent map, and we don't want
* to drop it from the cache until it is completely in the btree.
*
* So, tell btrfs_drop_extents to leave this extent in the cache.
* the caller is expected to unpin it and allow it to be merged
* with the others.
*/
ret = __btrfs_drop_extents(trans, root, inode, path, file_pos,
file_pos + num_bytes, NULL, 0,
1, sizeof(*fi), &extent_inserted);
if (ret)
goto out;
if (!extent_inserted) {
ins.objectid = btrfs_ino(inode);
ins.offset = file_pos;
ins.type = BTRFS_EXTENT_DATA_KEY;
path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, root, path, &ins,
sizeof(*fi));
if (ret)
goto out;
}
leaf = path->nodes[0];
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
btrfs_set_file_extent_generation(leaf, fi, trans->transid);
btrfs_set_file_extent_type(leaf, fi, extent_type);
btrfs_set_file_extent_disk_bytenr(leaf, fi, disk_bytenr);
btrfs_set_file_extent_disk_num_bytes(leaf, fi, disk_num_bytes);
btrfs_set_file_extent_offset(leaf, fi, 0);
btrfs_set_file_extent_num_bytes(leaf, fi, num_bytes);
btrfs_set_file_extent_ram_bytes(leaf, fi, ram_bytes);
btrfs_set_file_extent_compression(leaf, fi, compression);
btrfs_set_file_extent_encryption(leaf, fi, encryption);
btrfs_set_file_extent_other_encoding(leaf, fi, other_encoding);
btrfs_mark_buffer_dirty(leaf);
btrfs_release_path(path);
inode_add_bytes(inode, num_bytes);
ins.objectid = disk_bytenr;
ins.offset = disk_num_bytes;
ins.type = BTRFS_EXTENT_ITEM_KEY;
ret = btrfs_alloc_reserved_file_extent(trans, root,
root->root_key.objectid,
btrfs_ino(inode), file_pos, &ins);
out:
btrfs_free_path(path);
return ret;
}
/* snapshot-aware defrag */
struct sa_defrag_extent_backref {
struct rb_node node;
struct old_sa_defrag_extent *old;
u64 root_id;
u64 inum;
u64 file_pos;
u64 extent_offset;
u64 num_bytes;
u64 generation;
};
struct old_sa_defrag_extent {
struct list_head list;
struct new_sa_defrag_extent *new;
u64 extent_offset;
u64 bytenr;
u64 offset;
u64 len;
int count;
};
struct new_sa_defrag_extent {
struct rb_root root;
struct list_head head;
struct btrfs_path *path;
struct inode *inode;
u64 file_pos;
u64 len;
u64 bytenr;
u64 disk_len;
u8 compress_type;
};
static int backref_comp(struct sa_defrag_extent_backref *b1,
struct sa_defrag_extent_backref *b2)
{
if (b1->root_id < b2->root_id)
return -1;
else if (b1->root_id > b2->root_id)
return 1;
if (b1->inum < b2->inum)
return -1;
else if (b1->inum > b2->inum)
return 1;
if (b1->file_pos < b2->file_pos)
return -1;
else if (b1->file_pos > b2->file_pos)
return 1;
/*
* [------------------------------] ===> (a range of space)
* |<--->| |<---->| =============> (fs/file tree A)
* |<---------------------------->| ===> (fs/file tree B)
*
* A range of space can refer to two file extents in one tree while
* refer to only one file extent in another tree.
*
* So we may process a disk offset more than one time(two extents in A)
* and locate at the same extent(one extent in B), then insert two same
* backrefs(both refer to the extent in B).
*/
return 0;
}
static void backref_insert(struct rb_root *root,
struct sa_defrag_extent_backref *backref)
{
struct rb_node **p = &root->rb_node;
struct rb_node *parent = NULL;
struct sa_defrag_extent_backref *entry;
int ret;
while (*p) {
parent = *p;
entry = rb_entry(parent, struct sa_defrag_extent_backref, node);
ret = backref_comp(backref, entry);
if (ret < 0)
p = &(*p)->rb_left;
else
p = &(*p)->rb_right;
}
rb_link_node(&backref->node, parent, p);
rb_insert_color(&backref->node, root);
}
/*
* Note the backref might has changed, and in this case we just return 0.
*/
static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
void *ctx)
{
struct btrfs_file_extent_item *extent;
struct btrfs_fs_info *fs_info;
struct old_sa_defrag_extent *old = ctx;
struct new_sa_defrag_extent *new = old->new;
struct btrfs_path *path = new->path;
struct btrfs_key key;
struct btrfs_root *root;
struct sa_defrag_extent_backref *backref;
struct extent_buffer *leaf;
struct inode *inode = new->inode;
int slot;
int ret;
u64 extent_offset;
u64 num_bytes;
if (BTRFS_I(inode)->root->root_key.objectid == root_id &&
inum == btrfs_ino(inode))
return 0;
key.objectid = root_id;
key.type = BTRFS_ROOT_ITEM_KEY;
key.offset = (u64)-1;
fs_info = BTRFS_I(inode)->root->fs_info;
root = btrfs_read_fs_root_no_name(fs_info, &key);
if (IS_ERR(root)) {
if (PTR_ERR(root) == -ENOENT)
return 0;
WARN_ON(1);
pr_debug("inum=%llu, offset=%llu, root_id=%llu\n",
inum, offset, root_id);
return PTR_ERR(root);
}
key.objectid = inum;
key.type = BTRFS_EXTENT_DATA_KEY;
if (offset > (u64)-1 << 32)
key.offset = 0;
else
key.offset = offset;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (WARN_ON(ret < 0))
return ret;
ret = 0;
while (1) {
cond_resched();
leaf = path->nodes[0];
slot = path->slots[0];
if (slot >= btrfs_header_nritems(leaf)) {
ret = btrfs_next_leaf(root, path);
if (ret < 0) {
goto out;
} else if (ret > 0) {
ret = 0;
goto out;
}
continue;
}
path->slots[0]++;
btrfs_item_key_to_cpu(leaf, &key, slot);
if (key.objectid > inum)
goto out;
if (key.objectid < inum || key.type != BTRFS_EXTENT_DATA_KEY)
continue;
extent = btrfs_item_ptr(leaf, slot,
struct btrfs_file_extent_item);
if (btrfs_file_extent_disk_bytenr(leaf, extent) != old->bytenr)
continue;
/*
* 'offset' refers to the exact key.offset,
* NOT the 'offset' field in btrfs_extent_data_ref, ie.
* (key.offset - extent_offset).
*/
if (key.offset != offset)
continue;
extent_offset = btrfs_file_extent_offset(leaf, extent);
num_bytes = btrfs_file_extent_num_bytes(leaf, extent);
if (extent_offset >= old->extent_offset + old->offset +
old->len || extent_offset + num_bytes <=
old->extent_offset + old->offset)
continue;
break;
}
backref = kmalloc(sizeof(*backref), GFP_NOFS);
if (!backref) {
ret = -ENOENT;
goto out;
}
backref->root_id = root_id;
backref->inum = inum;
backref->file_pos = offset;
backref->num_bytes = num_bytes;
backref->extent_offset = extent_offset;
backref->generation = btrfs_file_extent_generation(leaf, extent);
backref->old = old;
backref_insert(&new->root, backref);
old->count++;
out:
btrfs_release_path(path);
WARN_ON(ret);
return ret;
}
static noinline bool record_extent_backrefs(struct btrfs_path *path,
struct new_sa_defrag_extent *new)
{
struct btrfs_fs_info *fs_info = BTRFS_I(new->inode)->root->fs_info;
struct old_sa_defrag_extent *old, *tmp;
int ret;
new->path = path;
list_for_each_entry_safe(old, tmp, &new->head, list) {
ret = iterate_inodes_from_logical(old->bytenr +
old->extent_offset, fs_info,
path, record_one_backref,
old);
if (ret < 0 && ret != -ENOENT)
return false;
/* no backref to be processed for this extent */
if (!old->count) {
list_del(&old->list);
kfree(old);
}
}
if (list_empty(&new->head))
return false;
return true;
}
static int relink_is_mergable(struct extent_buffer *leaf,
struct btrfs_file_extent_item *fi,
struct new_sa_defrag_extent *new)
{
if (btrfs_file_extent_disk_bytenr(leaf, fi) != new->bytenr)
return 0;
if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
return 0;
if (btrfs_file_extent_compression(leaf, fi) != new->compress_type)
return 0;
if (btrfs_file_extent_encryption(leaf, fi) ||
btrfs_file_extent_other_encoding(leaf, fi))
return 0;
return 1;
}
/*
* Note the backref might has changed, and in this case we just return 0.
*/
static noinline int relink_extent_backref(struct btrfs_path *path,
struct sa_defrag_extent_backref *prev,
struct sa_defrag_extent_backref *backref)
{
struct btrfs_file_extent_item *extent;
struct btrfs_file_extent_item *item;
struct btrfs_ordered_extent *ordered;
struct btrfs_trans_handle *trans;
struct btrfs_fs_info *fs_info;
struct btrfs_root *root;
struct btrfs_key key;
struct extent_buffer *leaf;
struct old_sa_defrag_extent *old = backref->old;
struct new_sa_defrag_extent *new = old->new;
struct inode *src_inode = new->inode;
struct inode *inode;
struct extent_state *cached = NULL;
int ret = 0;
u64 start;
u64 len;
u64 lock_start;
u64 lock_end;
bool merge = false;
int index;
if (prev && prev->root_id == backref->root_id &&
prev->inum == backref->inum &&
prev->file_pos + prev->num_bytes == backref->file_pos)
merge = true;
/* step 1: get root */
key.objectid = backref->root_id;
key.type = BTRFS_ROOT_ITEM_KEY;
key.offset = (u64)-1;
fs_info = BTRFS_I(src_inode)->root->fs_info;
index = srcu_read_lock(&fs_info->subvol_srcu);
root = btrfs_read_fs_root_no_name(fs_info, &key);
if (IS_ERR(root)) {
srcu_read_unlock(&fs_info->subvol_srcu, index);
if (PTR_ERR(root) == -ENOENT)
return 0;
return PTR_ERR(root);
}
if (btrfs_root_readonly(root)) {
srcu_read_unlock(&fs_info->subvol_srcu, index);
return 0;
}
/* step 2: get inode */
key.objectid = backref->inum;
key.type = BTRFS_INODE_ITEM_KEY;
key.offset = 0;
inode = btrfs_iget(fs_info->sb, &key, root, NULL);
if (IS_ERR(inode)) {
srcu_read_unlock(&fs_info->subvol_srcu, index);
return 0;
}
srcu_read_unlock(&fs_info->subvol_srcu, index);
/* step 3: relink backref */
lock_start = backref->file_pos;
lock_end = backref->file_pos + backref->num_bytes - 1;
lock_extent_bits(&BTRFS_I(inode)->io_tree, lock_start, lock_end,
0, &cached);
ordered = btrfs_lookup_first_ordered_extent(inode, lock_end);
if (ordered) {
btrfs_put_ordered_extent(ordered);
goto out_unlock;
}
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
goto out_unlock;
}
key.objectid = backref->inum;
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = backref->file_pos;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0) {
goto out_free_path;
} else if (ret > 0) {
ret = 0;
goto out_free_path;
}
extent = btrfs_item_ptr(path->nodes[0], path->slots[0],
struct btrfs_file_extent_item);
if (btrfs_file_extent_generation(path->nodes[0], extent) !=
backref->generation)
goto out_free_path;
btrfs_release_path(path);
start = backref->file_pos;
if (backref->extent_offset < old->extent_offset + old->offset)
start += old->extent_offset + old->offset -
backref->extent_offset;
len = min(backref->extent_offset + backref->num_bytes,
old->extent_offset + old->offset + old->len);
len -= max(backref->extent_offset, old->extent_offset + old->offset);
ret = btrfs_drop_extents(trans, root, inode, start,
start + len, 1);
if (ret)
goto out_free_path;
again:
key.objectid = btrfs_ino(inode);
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = start;
path->leave_spinning = 1;
if (merge) {
struct btrfs_file_extent_item *fi;
u64 extent_len;
struct btrfs_key found_key;
ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
if (ret < 0)
goto out_free_path;
path->slots[0]--;
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
extent_len = btrfs_file_extent_num_bytes(leaf, fi);
if (extent_len + found_key.offset == start &&
relink_is_mergable(leaf, fi, new)) {
btrfs_set_file_extent_num_bytes(leaf, fi,
extent_len + len);
btrfs_mark_buffer_dirty(leaf);
inode_add_bytes(inode, len);
ret = 1;
goto out_free_path;
} else {
merge = false;
btrfs_release_path(path);
goto again;
}
}
ret = btrfs_insert_empty_item(trans, root, path, &key,
sizeof(*extent));
if (ret) {
btrfs_abort_transaction(trans, root, ret);
goto out_free_path;
}
leaf = path->nodes[0];
item = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
btrfs_set_file_extent_disk_bytenr(leaf, item, new->bytenr);
btrfs_set_file_extent_disk_num_bytes(leaf, item, new->disk_len);
btrfs_set_file_extent_offset(leaf, item, start - new->file_pos);
btrfs_set_file_extent_num_bytes(leaf, item, len);
btrfs_set_file_extent_ram_bytes(leaf, item, new->len);
btrfs_set_file_extent_generation(leaf, item, trans->transid);
btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
btrfs_set_file_extent_compression(leaf, item, new->compress_type);
btrfs_set_file_extent_encryption(leaf, item, 0);
btrfs_set_file_extent_other_encoding(leaf, item, 0);
btrfs_mark_buffer_dirty(leaf);
inode_add_bytes(inode, len);
btrfs_release_path(path);
ret = btrfs_inc_extent_ref(trans, root, new->bytenr,
new->disk_len, 0,
backref->root_id, backref->inum,
new->file_pos, 0); /* start - extent_offset */
if (ret) {
btrfs_abort_transaction(trans, root, ret);
goto out_free_path;
}
ret = 1;
out_free_path:
btrfs_release_path(path);
path->leave_spinning = 0;
btrfs_end_transaction(trans, root);
out_unlock:
unlock_extent_cached(&BTRFS_I(inode)->io_tree, lock_start, lock_end,
&cached, GFP_NOFS);
iput(inode);
return ret;
}
static void free_sa_defrag_extent(struct new_sa_defrag_extent *new)
{
struct old_sa_defrag_extent *old, *tmp;
if (!new)
return;
list_for_each_entry_safe(old, tmp, &new->head, list) {
list_del(&old->list);
kfree(old);
}
kfree(new);
}
static void relink_file_extents(struct new_sa_defrag_extent *new)
{
struct btrfs_path *path;
struct sa_defrag_extent_backref *backref;
struct sa_defrag_extent_backref *prev = NULL;
struct inode *inode;
struct btrfs_root *root;
struct rb_node *node;
int ret;
inode = new->inode;
root = BTRFS_I(inode)->root;
path = btrfs_alloc_path();
if (!path)
return;
if (!record_extent_backrefs(path, new)) {
btrfs_free_path(path);
goto out;
}
btrfs_release_path(path);
while (1) {
node = rb_first(&new->root);
if (!node)
break;
rb_erase(node, &new->root);
backref = rb_entry(node, struct sa_defrag_extent_backref, node);
ret = relink_extent_backref(path, prev, backref);
WARN_ON(ret < 0);
kfree(prev);
if (ret == 1)
prev = backref;
else
prev = NULL;
cond_resched();
}
kfree(prev);
btrfs_free_path(path);
out:
free_sa_defrag_extent(new);
atomic_dec(&root->fs_info->defrag_running);
wake_up(&root->fs_info->transaction_wait);
}
static struct new_sa_defrag_extent *
record_old_file_extents(struct inode *inode,
struct btrfs_ordered_extent *ordered)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
struct btrfs_key key;
struct old_sa_defrag_extent *old;
struct new_sa_defrag_extent *new;
int ret;
new = kmalloc(sizeof(*new), GFP_NOFS);
if (!new)
return NULL;
new->inode = inode;
new->file_pos = ordered->file_offset;
new->len = ordered->len;
new->bytenr = ordered->start;
new->disk_len = ordered->disk_len;
new->compress_type = ordered->compress_type;
new->root = RB_ROOT;
INIT_LIST_HEAD(&new->head);
path = btrfs_alloc_path();
if (!path)
goto out_kfree;
key.objectid = btrfs_ino(inode);
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = new->file_pos;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
goto out_free_path;
if (ret > 0 && path->slots[0] > 0)
path->slots[0]--;
/* find out all the old extents for the file range */
while (1) {
struct btrfs_file_extent_item *extent;
struct extent_buffer *l;
int slot;
u64 num_bytes;
u64 offset;
u64 end;
u64 disk_bytenr;
u64 extent_offset;
l = path->nodes[0];
slot = path->slots[0];
if (slot >= btrfs_header_nritems(l)) {
ret = btrfs_next_leaf(root, path);
if (ret < 0)
goto out_free_path;
else if (ret > 0)
break;
continue;
}
btrfs_item_key_to_cpu(l, &key, slot);
if (key.objectid != btrfs_ino(inode))
break;
if (key.type != BTRFS_EXTENT_DATA_KEY)
break;
if (key.offset >= new->file_pos + new->len)
break;
extent = btrfs_item_ptr(l, slot, struct btrfs_file_extent_item);
num_bytes = btrfs_file_extent_num_bytes(l, extent);
if (key.offset + num_bytes < new->file_pos)
goto next;
disk_bytenr = btrfs_file_extent_disk_bytenr(l, extent);
if (!disk_bytenr)
goto next;
extent_offset = btrfs_file_extent_offset(l, extent);
old = kmalloc(sizeof(*old), GFP_NOFS);
if (!old)
goto out_free_path;
offset = max(new->file_pos, key.offset);
end = min(new->file_pos + new->len, key.offset + num_bytes);
old->bytenr = disk_bytenr;
old->extent_offset = extent_offset;
old->offset = offset - key.offset;
old->len = end - offset;
old->new = new;
old->count = 0;
list_add_tail(&old->list, &new->head);
next:
path->slots[0]++;
cond_resched();
}
btrfs_free_path(path);
atomic_inc(&root->fs_info->defrag_running);
return new;
out_free_path:
btrfs_free_path(path);
out_kfree:
free_sa_defrag_extent(new);
return NULL;
}
static void btrfs_release_delalloc_bytes(struct btrfs_root *root,
u64 start, u64 len)
{
struct btrfs_block_group_cache *cache;
cache = btrfs_lookup_block_group(root->fs_info, start);
ASSERT(cache);
spin_lock(&cache->lock);
cache->delalloc_bytes -= len;
spin_unlock(&cache->lock);
btrfs_put_block_group(cache);
}
/* as ordered data IO finishes, this gets called so we can finish
* an ordered extent if the range of bytes in the file it covers are
* fully written.
*/
static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
{
struct inode *inode = ordered_extent->inode;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_trans_handle *trans = NULL;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
struct extent_state *cached_state = NULL;
struct new_sa_defrag_extent *new = NULL;
int compress_type = 0;
int ret = 0;
u64 logical_len = ordered_extent->len;
bool nolock;
bool truncated = false;
nolock = btrfs_is_free_space_inode(inode);
if (test_bit(BTRFS_ORDERED_IOERR, &ordered_extent->flags)) {
ret = -EIO;
goto out;
}
if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered_extent->flags)) {
truncated = true;
logical_len = ordered_extent->truncated_len;
/* Truncated the entire extent, don't bother adding */
if (!logical_len)
goto out;
}
if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
BUG_ON(!list_empty(&ordered_extent->list)); /* Logic error */
btrfs_ordered_update_i_size(inode, 0, ordered_extent);
if (nolock)
trans = btrfs_join_transaction_nolock(root);
else
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
trans = NULL;
goto out;
}
trans->block_rsv = &root->fs_info->delalloc_block_rsv;
ret = btrfs_update_inode_fallback(trans, root, inode);
if (ret) /* -ENOMEM or corruption */
btrfs_abort_transaction(trans, root, ret);
goto out;
}
lock_extent_bits(io_tree, ordered_extent->file_offset,
ordered_extent->file_offset + ordered_extent->len - 1,
0, &cached_state);
ret = test_range_bit(io_tree, ordered_extent->file_offset,
ordered_extent->file_offset + ordered_extent->len - 1,
EXTENT_DEFRAG, 1, cached_state);
if (ret) {
u64 last_snapshot = btrfs_root_last_snapshot(&root->root_item);
if (0 && last_snapshot >= BTRFS_I(inode)->generation)
/* the inode is shared */
new = record_old_file_extents(inode, ordered_extent);
clear_extent_bit(io_tree, ordered_extent->file_offset,
ordered_extent->file_offset + ordered_extent->len - 1,
EXTENT_DEFRAG, 0, 0, &cached_state, GFP_NOFS);
}
if (nolock)
trans = btrfs_join_transaction_nolock(root);
else
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
trans = NULL;
goto out_unlock;
}
trans->block_rsv = &root->fs_info->delalloc_block_rsv;
if (test_bit(BTRFS_ORDERED_COMPRESSED, &ordered_extent->flags))
compress_type = ordered_extent->compress_type;
if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) {
BUG_ON(compress_type);
ret = btrfs_mark_extent_written(trans, inode,
ordered_extent->file_offset,
ordered_extent->file_offset +
logical_len);
} else {
BUG_ON(root == root->fs_info->tree_root);
ret = insert_reserved_file_extent(trans, inode,
ordered_extent->file_offset,
ordered_extent->start,
ordered_extent->disk_len,
logical_len, logical_len,
compress_type, 0, 0,
BTRFS_FILE_EXTENT_REG);
if (!ret)
btrfs_release_delalloc_bytes(root,
ordered_extent->start,
ordered_extent->disk_len);
}
unpin_extent_cache(&BTRFS_I(inode)->extent_tree,
ordered_extent->file_offset, ordered_extent->len,
trans->transid);
if (ret < 0) {
btrfs_abort_transaction(trans, root, ret);
goto out_unlock;
}
add_pending_csums(trans, inode, ordered_extent->file_offset,
&ordered_extent->list);
btrfs_ordered_update_i_size(inode, 0, ordered_extent);
ret = btrfs_update_inode_fallback(trans, root, inode);
if (ret) { /* -ENOMEM or corruption */
btrfs_abort_transaction(trans, root, ret);
goto out_unlock;
}
ret = 0;
out_unlock:
unlock_extent_cached(io_tree, ordered_extent->file_offset,
ordered_extent->file_offset +
ordered_extent->len - 1, &cached_state, GFP_NOFS);
out:
if (root != root->fs_info->tree_root)
btrfs_delalloc_release_metadata(inode, ordered_extent->len);
if (trans)
btrfs_end_transaction(trans, root);
if (ret || truncated) {
u64 start, end;
if (truncated)
start = ordered_extent->file_offset + logical_len;
else
start = ordered_extent->file_offset;
end = ordered_extent->file_offset + ordered_extent->len - 1;
clear_extent_uptodate(io_tree, start, end, NULL, GFP_NOFS);
/* Drop the cache for the part of the extent we didn't write. */
btrfs_drop_extent_cache(inode, start, end, 0);
/*
* If the ordered extent had an IOERR or something else went
* wrong we need to return the space for this ordered extent
* back to the allocator. We only free the extent in the
* truncated case if we didn't write out the extent at all.
*/
if ((ret || !logical_len) &&
!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) &&
!test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags))
btrfs_free_reserved_extent(root, ordered_extent->start,
ordered_extent->disk_len, 1);
}
/*
* This needs to be done to make sure anybody waiting knows we are done
* updating everything for this ordered extent.
*/
btrfs_remove_ordered_extent(inode, ordered_extent);
/* for snapshot-aware defrag */
if (new) {
if (ret) {
free_sa_defrag_extent(new);
atomic_dec(&root->fs_info->defrag_running);
} else {
relink_file_extents(new);
}
}
/* once for us */
btrfs_put_ordered_extent(ordered_extent);
/* once for the tree */
btrfs_put_ordered_extent(ordered_extent);
return ret;
}
static void finish_ordered_fn(struct btrfs_work *work)
{
struct btrfs_ordered_extent *ordered_extent;
ordered_extent = container_of(work, struct btrfs_ordered_extent, work);
btrfs_finish_ordered_io(ordered_extent);
}
static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
struct extent_state *state, int uptodate)
{
struct inode *inode = page->mapping->host;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_ordered_extent *ordered_extent = NULL;
struct btrfs_workqueue *workers;
trace_btrfs_writepage_end_io_hook(page, start, end, uptodate);
ClearPagePrivate2(page);
if (!btrfs_dec_test_ordered_pending(inode, &ordered_extent, start,
end - start + 1, uptodate))
return 0;
btrfs_init_work(&ordered_extent->work, finish_ordered_fn, NULL, NULL);
if (btrfs_is_free_space_inode(inode))
workers = root->fs_info->endio_freespace_worker;
else
workers = root->fs_info->endio_write_workers;
btrfs_queue_work(workers, &ordered_extent->work);
return 0;
}
/*
* when reads are done, we need to check csums to verify the data is correct
* if there's a match, we allow the bio to finish. If not, the code in
* extent_io.c will try to find good copies for us.
*/
static int btrfs_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
u64 phy_offset, struct page *page,
u64 start, u64 end, int mirror)
{
size_t offset = start - page_offset(page);
struct inode *inode = page->mapping->host;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
char *kaddr;
struct btrfs_root *root = BTRFS_I(inode)->root;
u32 csum_expected;
u32 csum = ~(u32)0;
static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
DEFAULT_RATELIMIT_BURST);
if (PageChecked(page)) {
ClearPageChecked(page);
goto good;
}
if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
goto good;
if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID &&
test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1, NULL)) {
clear_extent_bits(io_tree, start, end, EXTENT_NODATASUM,
GFP_NOFS);
return 0;
}
phy_offset >>= inode->i_sb->s_blocksize_bits;
csum_expected = *(((u32 *)io_bio->csum) + phy_offset);
kaddr = kmap_atomic(page);
csum = btrfs_csum_data(kaddr + offset, csum, end - start + 1);
btrfs_csum_final(csum, (char *)&csum);
if (csum != csum_expected)
goto zeroit;
kunmap_atomic(kaddr);
good:
return 0;
zeroit:
if (__ratelimit(&_rs))
btrfs_info(root->fs_info, "csum failed ino %llu off %llu csum %u expected csum %u",
btrfs_ino(page->mapping->host), start, csum, csum_expected);
memset(kaddr + offset, 1, end - start + 1);
flush_dcache_page(page);
kunmap_atomic(kaddr);
if (csum_expected == 0)
return 0;
return -EIO;
}
struct delayed_iput {
struct list_head list;
struct inode *inode;
};
/* JDM: If this is fs-wide, why can't we add a pointer to
* btrfs_inode instead and avoid the allocation? */
void btrfs_add_delayed_iput(struct inode *inode)
{
struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
struct delayed_iput *delayed;
if (atomic_add_unless(&inode->i_count, -1, 1))
return;
delayed = kmalloc(sizeof(*delayed), GFP_NOFS | __GFP_NOFAIL);
delayed->inode = inode;
spin_lock(&fs_info->delayed_iput_lock);
list_add_tail(&delayed->list, &fs_info->delayed_iputs);
spin_unlock(&fs_info->delayed_iput_lock);
}
void btrfs_run_delayed_iputs(struct btrfs_root *root)
{
LIST_HEAD(list);
struct btrfs_fs_info *fs_info = root->fs_info;
struct delayed_iput *delayed;
int empty;
spin_lock(&fs_info->delayed_iput_lock);
empty = list_empty(&fs_info->delayed_iputs);
spin_unlock(&fs_info->delayed_iput_lock);
if (empty)
return;
spin_lock(&fs_info->delayed_iput_lock);
list_splice_init(&fs_info->delayed_iputs, &list);
spin_unlock(&fs_info->delayed_iput_lock);
while (!list_empty(&list)) {
delayed = list_entry(list.next, struct delayed_iput, list);
list_del(&delayed->list);
iput(delayed->inode);
kfree(delayed);
}
}
/*
* This is called in transaction commit time. If there are no orphan
* files in the subvolume, it removes orphan item and frees block_rsv
* structure.
*/
void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans,
struct btrfs_root *root)
{
struct btrfs_block_rsv *block_rsv;
int ret;
if (atomic_read(&root->orphan_inodes) ||
root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE)
return;
spin_lock(&root->orphan_lock);
if (atomic_read(&root->orphan_inodes)) {
spin_unlock(&root->orphan_lock);
return;
}
if (root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE) {
spin_unlock(&root->orphan_lock);
return;
}
block_rsv = root->orphan_block_rsv;
root->orphan_block_rsv = NULL;
spin_unlock(&root->orphan_lock);
if (test_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state) &&
btrfs_root_refs(&root->root_item) > 0) {
ret = btrfs_del_orphan_item(trans, root->fs_info->tree_root,
root->root_key.objectid);
if (ret)
btrfs_abort_transaction(trans, root, ret);
else
clear_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED,
&root->state);
}
if (block_rsv) {
WARN_ON(block_rsv->size > 0);
btrfs_free_block_rsv(root, block_rsv);
}
}
/*
* This creates an orphan entry for the given inode in case something goes
* wrong in the middle of an unlink/truncate.
*
* NOTE: caller of this function should reserve 5 units of metadata for
* this function.
*/
int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_block_rsv *block_rsv = NULL;
int reserve = 0;
int insert = 0;
int ret;
if (!root->orphan_block_rsv) {
block_rsv = btrfs_alloc_block_rsv(root, BTRFS_BLOCK_RSV_TEMP);
if (!block_rsv)
return -ENOMEM;
}
spin_lock(&root->orphan_lock);
if (!root->orphan_block_rsv) {
root->orphan_block_rsv = block_rsv;
} else if (block_rsv) {
btrfs_free_block_rsv(root, block_rsv);
block_rsv = NULL;
}
if (!test_and_set_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
&BTRFS_I(inode)->runtime_flags)) {
#if 0
/*
* For proper ENOSPC handling, we should do orphan
* cleanup when mounting. But this introduces backward
* compatibility issue.
*/
if (!xchg(&root->orphan_item_inserted, 1))
insert = 2;
else
insert = 1;
#endif
insert = 1;
atomic_inc(&root->orphan_inodes);
}
if (!test_and_set_bit(BTRFS_INODE_ORPHAN_META_RESERVED,
&BTRFS_I(inode)->runtime_flags))
reserve =