2007-06-12 09:07:21 -04:00
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
|
|
|
|
2008-02-20 12:07:25 -05:00
|
|
|
#include <linux/bio.h>
|
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/highmem.h>
|
2007-03-15 19:03:33 -04:00
|
|
|
#include "ctree.h"
|
2007-03-26 16:00:06 -04:00
|
|
|
#include "disk-io.h"
|
2007-03-20 14:38:32 -04:00
|
|
|
#include "transaction.h"
|
2007-05-29 15:17:08 -04:00
|
|
|
#include "print-tree.h"
|
2007-03-15 19:03:33 -04:00
|
|
|
|
2009-01-05 21:25:51 -05:00
|
|
|
#define MAX_CSUM_ITEMS(r, size) ((((BTRFS_LEAF_DATA_SIZE(r) - \
|
2008-12-02 07:17:45 -05:00
|
|
|
sizeof(struct btrfs_item) * 2) / \
|
|
|
|
size) - 1))
|
2009-01-06 11:42:00 -05:00
|
|
|
|
|
|
|
#define MAX_ORDERED_SUM_BYTES(r) ((PAGE_SIZE - \
|
|
|
|
sizeof(struct btrfs_ordered_sum)) / \
|
|
|
|
sizeof(struct btrfs_sector_sum) * \
|
|
|
|
(r)->sectorsize - (r)->sectorsize)
|
|
|
|
|
2007-04-17 13:26:50 -04:00
|
|
|
int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
|
2008-05-02 14:43:14 -04:00
|
|
|
struct btrfs_root *root,
|
|
|
|
u64 objectid, u64 pos,
|
|
|
|
u64 disk_offset, u64 disk_num_bytes,
|
Btrfs: Add zlib compression support
This is a large change for adding compression on reading and writing,
both for inline and regular extents. It does some fairly large
surgery to the writeback paths.
Compression is off by default and enabled by mount -o compress. Even
when the -o compress mount option is not used, it is possible to read
compressed extents off the disk.
If compression for a given set of pages fails to make them smaller, the
file is flagged to avoid future compression attempts later.
* While finding delalloc extents, the pages are locked before being sent down
to the delalloc handler. This allows the delalloc handler to do complex things
such as cleaning the pages, marking them writeback and starting IO on their
behalf.
* Inline extents are inserted at delalloc time now. This allows us to compress
the data before inserting the inline extent, and it allows us to insert
an inline extent that spans multiple pages.
* All of the in-memory extent representations (extent_map.c, ordered-data.c etc)
are changed to record both an in-memory size and an on disk size, as well
as a flag for compression.
From a disk format point of view, the extent pointers in the file are changed
to record the on disk size of a given extent and some encoding flags.
Space in the disk format is allocated for compression encoding, as well
as encryption and a generic 'other' field. Neither the encryption or the
'other' field are currently used.
In order to limit the amount of data read for a single random read in the
file, the size of a compressed extent is limited to 128k. This is a
software only limit, the disk format supports u64 sized compressed extents.
In order to limit the ram consumed while processing extents, the uncompressed
size of a compressed extent is limited to 256k. This is a software only limit
and will be subject to tuning later.
Checksumming is still done on compressed extents, and it is done on the
uncompressed version of the data. This way additional encodings can be
layered on without having to figure out which encoding to checksum.
Compression happens at delalloc time, which is basically singled threaded because
it is usually done by a single pdflush thread. This makes it tricky to
spread the compression load across all the cpus on the box. We'll have to
look at parallel pdflush walks of dirty inodes at a later time.
Decompression is hooked into readpages and it does spread across CPUs nicely.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-10-29 14:49:59 -04:00
|
|
|
u64 num_bytes, u64 offset, u64 ram_bytes,
|
|
|
|
u8 compression, u8 encryption, u16 other_encoding)
|
2007-03-20 14:38:32 -04:00
|
|
|
{
|
2007-03-26 16:00:06 -04:00
|
|
|
int ret = 0;
|
|
|
|
struct btrfs_file_extent_item *item;
|
|
|
|
struct btrfs_key file_key;
|
2007-04-02 11:20:42 -04:00
|
|
|
struct btrfs_path *path;
|
2007-10-15 16:14:19 -04:00
|
|
|
struct extent_buffer *leaf;
|
2007-03-26 16:00:06 -04:00
|
|
|
|
2007-04-02 11:20:42 -04:00
|
|
|
path = btrfs_alloc_path();
|
|
|
|
BUG_ON(!path);
|
2007-03-26 16:00:06 -04:00
|
|
|
file_key.objectid = objectid;
|
2007-04-17 13:26:50 -04:00
|
|
|
file_key.offset = pos;
|
2007-03-26 16:00:06 -04:00
|
|
|
btrfs_set_key_type(&file_key, BTRFS_EXTENT_DATA_KEY);
|
|
|
|
|
2007-04-02 11:20:42 -04:00
|
|
|
ret = btrfs_insert_empty_item(trans, root, path, &file_key,
|
2007-03-26 16:00:06 -04:00
|
|
|
sizeof(*item));
|
2007-06-22 14:16:25 -04:00
|
|
|
if (ret < 0)
|
|
|
|
goto out;
|
2007-03-27 11:26:26 -04:00
|
|
|
BUG_ON(ret);
|
2007-10-15 16:14:19 -04:00
|
|
|
leaf = path->nodes[0];
|
|
|
|
item = btrfs_item_ptr(leaf, path->slots[0],
|
2007-03-26 16:00:06 -04:00
|
|
|
struct btrfs_file_extent_item);
|
2008-05-02 14:43:14 -04:00
|
|
|
btrfs_set_file_extent_disk_bytenr(leaf, item, disk_offset);
|
2007-10-15 16:15:53 -04:00
|
|
|
btrfs_set_file_extent_disk_num_bytes(leaf, item, disk_num_bytes);
|
2008-05-02 14:43:14 -04:00
|
|
|
btrfs_set_file_extent_offset(leaf, item, offset);
|
2007-10-15 16:15:53 -04:00
|
|
|
btrfs_set_file_extent_num_bytes(leaf, item, num_bytes);
|
Btrfs: Add zlib compression support
This is a large change for adding compression on reading and writing,
both for inline and regular extents. It does some fairly large
surgery to the writeback paths.
Compression is off by default and enabled by mount -o compress. Even
when the -o compress mount option is not used, it is possible to read
compressed extents off the disk.
If compression for a given set of pages fails to make them smaller, the
file is flagged to avoid future compression attempts later.
* While finding delalloc extents, the pages are locked before being sent down
to the delalloc handler. This allows the delalloc handler to do complex things
such as cleaning the pages, marking them writeback and starting IO on their
behalf.
* Inline extents are inserted at delalloc time now. This allows us to compress
the data before inserting the inline extent, and it allows us to insert
an inline extent that spans multiple pages.
* All of the in-memory extent representations (extent_map.c, ordered-data.c etc)
are changed to record both an in-memory size and an on disk size, as well
as a flag for compression.
From a disk format point of view, the extent pointers in the file are changed
to record the on disk size of a given extent and some encoding flags.
Space in the disk format is allocated for compression encoding, as well
as encryption and a generic 'other' field. Neither the encryption or the
'other' field are currently used.
In order to limit the amount of data read for a single random read in the
file, the size of a compressed extent is limited to 128k. This is a
software only limit, the disk format supports u64 sized compressed extents.
In order to limit the ram consumed while processing extents, the uncompressed
size of a compressed extent is limited to 256k. This is a software only limit
and will be subject to tuning later.
Checksumming is still done on compressed extents, and it is done on the
uncompressed version of the data. This way additional encodings can be
layered on without having to figure out which encoding to checksum.
Compression happens at delalloc time, which is basically singled threaded because
it is usually done by a single pdflush thread. This makes it tricky to
spread the compression load across all the cpus on the box. We'll have to
look at parallel pdflush walks of dirty inodes at a later time.
Decompression is hooked into readpages and it does spread across CPUs nicely.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-10-29 14:49:59 -04:00
|
|
|
btrfs_set_file_extent_ram_bytes(leaf, item, ram_bytes);
|
2007-10-15 16:14:19 -04:00
|
|
|
btrfs_set_file_extent_generation(leaf, item, trans->transid);
|
|
|
|
btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
|
Btrfs: Add zlib compression support
This is a large change for adding compression on reading and writing,
both for inline and regular extents. It does some fairly large
surgery to the writeback paths.
Compression is off by default and enabled by mount -o compress. Even
when the -o compress mount option is not used, it is possible to read
compressed extents off the disk.
If compression for a given set of pages fails to make them smaller, the
file is flagged to avoid future compression attempts later.
* While finding delalloc extents, the pages are locked before being sent down
to the delalloc handler. This allows the delalloc handler to do complex things
such as cleaning the pages, marking them writeback and starting IO on their
behalf.
* Inline extents are inserted at delalloc time now. This allows us to compress
the data before inserting the inline extent, and it allows us to insert
an inline extent that spans multiple pages.
* All of the in-memory extent representations (extent_map.c, ordered-data.c etc)
are changed to record both an in-memory size and an on disk size, as well
as a flag for compression.
From a disk format point of view, the extent pointers in the file are changed
to record the on disk size of a given extent and some encoding flags.
Space in the disk format is allocated for compression encoding, as well
as encryption and a generic 'other' field. Neither the encryption or the
'other' field are currently used.
In order to limit the amount of data read for a single random read in the
file, the size of a compressed extent is limited to 128k. This is a
software only limit, the disk format supports u64 sized compressed extents.
In order to limit the ram consumed while processing extents, the uncompressed
size of a compressed extent is limited to 256k. This is a software only limit
and will be subject to tuning later.
Checksumming is still done on compressed extents, and it is done on the
uncompressed version of the data. This way additional encodings can be
layered on without having to figure out which encoding to checksum.
Compression happens at delalloc time, which is basically singled threaded because
it is usually done by a single pdflush thread. This makes it tricky to
spread the compression load across all the cpus on the box. We'll have to
look at parallel pdflush walks of dirty inodes at a later time.
Decompression is hooked into readpages and it does spread across CPUs nicely.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-10-29 14:49:59 -04:00
|
|
|
btrfs_set_file_extent_compression(leaf, item, compression);
|
|
|
|
btrfs_set_file_extent_encryption(leaf, item, encryption);
|
|
|
|
btrfs_set_file_extent_other_encoding(leaf, item, other_encoding);
|
|
|
|
|
2007-10-15 16:14:19 -04:00
|
|
|
btrfs_mark_buffer_dirty(leaf);
|
2007-06-22 14:16:25 -04:00
|
|
|
out:
|
2007-04-02 11:20:42 -04:00
|
|
|
btrfs_free_path(path);
|
2007-06-22 14:16:25 -04:00
|
|
|
return ret;
|
2007-03-20 14:38:32 -04:00
|
|
|
}
|
2007-03-26 16:00:06 -04:00
|
|
|
|
2007-04-17 13:26:50 -04:00
|
|
|
struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans,
|
|
|
|
struct btrfs_root *root,
|
|
|
|
struct btrfs_path *path,
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
u64 bytenr, int cow)
|
2007-04-16 09:22:45 -04:00
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
struct btrfs_key file_key;
|
|
|
|
struct btrfs_key found_key;
|
|
|
|
struct btrfs_csum_item *item;
|
2007-10-15 16:14:19 -04:00
|
|
|
struct extent_buffer *leaf;
|
2007-04-16 09:22:45 -04:00
|
|
|
u64 csum_offset = 0;
|
2008-12-02 07:17:45 -05:00
|
|
|
u16 csum_size =
|
|
|
|
btrfs_super_csum_size(&root->fs_info->super_copy);
|
2007-04-18 16:15:28 -04:00
|
|
|
int csums_in_item;
|
2007-04-16 09:22:45 -04:00
|
|
|
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
|
|
|
|
file_key.offset = bytenr;
|
|
|
|
btrfs_set_key_type(&file_key, BTRFS_EXTENT_CSUM_KEY);
|
2007-04-17 13:26:50 -04:00
|
|
|
ret = btrfs_search_slot(trans, root, &file_key, path, 0, cow);
|
2007-04-16 09:22:45 -04:00
|
|
|
if (ret < 0)
|
|
|
|
goto fail;
|
2007-10-15 16:14:19 -04:00
|
|
|
leaf = path->nodes[0];
|
2007-04-16 09:22:45 -04:00
|
|
|
if (ret > 0) {
|
|
|
|
ret = 1;
|
2007-04-17 15:39:32 -04:00
|
|
|
if (path->slots[0] == 0)
|
2007-04-16 09:22:45 -04:00
|
|
|
goto fail;
|
|
|
|
path->slots[0]--;
|
2007-10-15 16:14:19 -04:00
|
|
|
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (btrfs_key_type(&found_key) != BTRFS_EXTENT_CSUM_KEY)
|
2007-04-16 09:22:45 -04:00
|
|
|
goto fail;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
|
|
|
|
csum_offset = (bytenr - found_key.offset) >>
|
2007-04-16 09:22:45 -04:00
|
|
|
root->fs_info->sb->s_blocksize_bits;
|
2007-10-15 16:14:19 -04:00
|
|
|
csums_in_item = btrfs_item_size_nr(leaf, path->slots[0]);
|
2008-12-02 07:17:45 -05:00
|
|
|
csums_in_item /= csum_size;
|
2007-04-18 16:15:28 -04:00
|
|
|
|
|
|
|
if (csum_offset >= csums_in_item) {
|
|
|
|
ret = -EFBIG;
|
2007-04-16 09:22:45 -04:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
|
2007-05-10 12:36:17 -04:00
|
|
|
item = (struct btrfs_csum_item *)((unsigned char *)item +
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_offset * csum_size);
|
2007-04-16 09:22:45 -04:00
|
|
|
return item;
|
|
|
|
fail:
|
|
|
|
if (ret > 0)
|
2007-04-17 13:26:50 -04:00
|
|
|
ret = -ENOENT;
|
2007-04-16 09:22:45 -04:00
|
|
|
return ERR_PTR(ret);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2007-03-26 16:00:06 -04:00
|
|
|
int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
|
|
|
|
struct btrfs_root *root,
|
|
|
|
struct btrfs_path *path, u64 objectid,
|
2007-03-27 11:26:26 -04:00
|
|
|
u64 offset, int mod)
|
2007-03-26 16:00:06 -04:00
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
struct btrfs_key file_key;
|
|
|
|
int ins_len = mod < 0 ? -1 : 0;
|
|
|
|
int cow = mod != 0;
|
|
|
|
|
|
|
|
file_key.objectid = objectid;
|
2007-04-17 15:39:32 -04:00
|
|
|
file_key.offset = offset;
|
2007-03-26 16:00:06 -04:00
|
|
|
btrfs_set_key_type(&file_key, BTRFS_EXTENT_DATA_KEY);
|
|
|
|
ret = btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
|
|
|
|
return ret;
|
|
|
|
}
|
2007-03-29 15:15:27 -04:00
|
|
|
|
2008-12-12 10:03:38 -05:00
|
|
|
|
2008-07-31 15:42:53 -04:00
|
|
|
int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
struct bio *bio, u32 *dst)
|
2008-07-31 15:42:53 -04:00
|
|
|
{
|
|
|
|
u32 sum;
|
|
|
|
struct bio_vec *bvec = bio->bi_io_vec;
|
|
|
|
int bio_index = 0;
|
|
|
|
u64 offset;
|
|
|
|
u64 item_start_offset = 0;
|
|
|
|
u64 item_last_offset = 0;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
u64 disk_bytenr;
|
2008-07-31 15:42:53 -04:00
|
|
|
u32 diff;
|
2008-12-02 07:17:45 -05:00
|
|
|
u16 csum_size =
|
|
|
|
btrfs_super_csum_size(&root->fs_info->super_copy);
|
2008-07-31 15:42:53 -04:00
|
|
|
int ret;
|
|
|
|
struct btrfs_path *path;
|
|
|
|
struct btrfs_csum_item *item = NULL;
|
|
|
|
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
|
|
|
|
|
|
|
|
path = btrfs_alloc_path();
|
2008-08-20 09:44:52 -04:00
|
|
|
if (bio->bi_size > PAGE_CACHE_SIZE * 8)
|
|
|
|
path->reada = 2;
|
2008-07-31 15:42:53 -04:00
|
|
|
|
|
|
|
WARN_ON(bio->bi_vcnt <= 0);
|
|
|
|
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
disk_bytenr = (u64)bio->bi_sector << 9;
|
2009-01-05 21:25:51 -05:00
|
|
|
while (bio_index < bio->bi_vcnt) {
|
2008-07-31 15:42:53 -04:00
|
|
|
offset = page_offset(bvec->bv_page) + bvec->bv_offset;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
ret = btrfs_find_ordered_sum(inode, offset, disk_bytenr, &sum);
|
2008-07-31 15:42:53 -04:00
|
|
|
if (ret == 0)
|
|
|
|
goto found;
|
|
|
|
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (!item || disk_bytenr < item_start_offset ||
|
|
|
|
disk_bytenr >= item_last_offset) {
|
2008-07-31 15:42:53 -04:00
|
|
|
struct btrfs_key found_key;
|
|
|
|
u32 item_size;
|
|
|
|
|
|
|
|
if (item)
|
|
|
|
btrfs_release_path(root, path);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
item = btrfs_lookup_csum(NULL, root->fs_info->csum_root,
|
|
|
|
path, disk_bytenr, 0);
|
2008-07-31 15:42:53 -04:00
|
|
|
if (IS_ERR(item)) {
|
|
|
|
ret = PTR_ERR(item);
|
|
|
|
if (ret == -ENOENT || ret == -EFBIG)
|
|
|
|
ret = 0;
|
|
|
|
sum = 0;
|
2008-12-12 10:03:38 -05:00
|
|
|
if (BTRFS_I(inode)->root->root_key.objectid ==
|
|
|
|
BTRFS_DATA_RELOC_TREE_OBJECTID) {
|
|
|
|
set_extent_bits(io_tree, offset,
|
|
|
|
offset + bvec->bv_len - 1,
|
|
|
|
EXTENT_NODATASUM, GFP_NOFS);
|
|
|
|
} else {
|
2009-01-05 21:25:51 -05:00
|
|
|
printk(KERN_INFO "btrfs no csum found "
|
|
|
|
"for inode %lu start %llu\n",
|
|
|
|
inode->i_ino,
|
2008-12-12 10:03:38 -05:00
|
|
|
(unsigned long long)offset);
|
|
|
|
}
|
2008-08-04 08:35:53 -04:00
|
|
|
item = NULL;
|
2008-11-10 11:50:50 -05:00
|
|
|
btrfs_release_path(root, path);
|
2008-07-31 15:42:53 -04:00
|
|
|
goto found;
|
|
|
|
}
|
|
|
|
btrfs_item_key_to_cpu(path->nodes[0], &found_key,
|
|
|
|
path->slots[0]);
|
|
|
|
|
|
|
|
item_start_offset = found_key.offset;
|
|
|
|
item_size = btrfs_item_size_nr(path->nodes[0],
|
|
|
|
path->slots[0]);
|
|
|
|
item_last_offset = item_start_offset +
|
2008-12-02 07:17:45 -05:00
|
|
|
(item_size / csum_size) *
|
2008-07-31 15:42:53 -04:00
|
|
|
root->sectorsize;
|
|
|
|
item = btrfs_item_ptr(path->nodes[0], path->slots[0],
|
|
|
|
struct btrfs_csum_item);
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* this byte range must be able to fit inside
|
|
|
|
* a single leaf so it will also fit inside a u32
|
|
|
|
*/
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
diff = disk_bytenr - item_start_offset;
|
2008-07-31 15:42:53 -04:00
|
|
|
diff = diff / root->sectorsize;
|
2008-12-02 07:17:45 -05:00
|
|
|
diff = diff * csum_size;
|
2008-07-31 15:42:53 -04:00
|
|
|
|
|
|
|
read_extent_buffer(path->nodes[0], &sum,
|
2008-08-04 23:17:27 -04:00
|
|
|
((unsigned long)item) + diff,
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_size);
|
2008-07-31 15:42:53 -04:00
|
|
|
found:
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (dst)
|
|
|
|
*dst++ = sum;
|
|
|
|
else
|
|
|
|
set_state_private(io_tree, offset, sum);
|
|
|
|
disk_bytenr += bvec->bv_len;
|
2008-07-31 15:42:53 -04:00
|
|
|
bio_index++;
|
|
|
|
bvec++;
|
|
|
|
}
|
|
|
|
btrfs_free_path(path);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-12-12 10:03:38 -05:00
|
|
|
int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
|
|
|
|
struct list_head *list)
|
|
|
|
{
|
|
|
|
struct btrfs_key key;
|
|
|
|
struct btrfs_path *path;
|
|
|
|
struct extent_buffer *leaf;
|
|
|
|
struct btrfs_ordered_sum *sums;
|
|
|
|
struct btrfs_sector_sum *sector_sum;
|
|
|
|
struct btrfs_csum_item *item;
|
|
|
|
unsigned long offset;
|
|
|
|
int ret;
|
|
|
|
size_t size;
|
|
|
|
u64 csum_end;
|
|
|
|
u16 csum_size = btrfs_super_csum_size(&root->fs_info->super_copy);
|
|
|
|
|
|
|
|
path = btrfs_alloc_path();
|
|
|
|
BUG_ON(!path);
|
|
|
|
|
|
|
|
key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
|
|
|
|
key.offset = start;
|
|
|
|
key.type = BTRFS_EXTENT_CSUM_KEY;
|
|
|
|
|
2009-01-06 11:42:00 -05:00
|
|
|
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
|
2008-12-12 10:03:38 -05:00
|
|
|
if (ret < 0)
|
|
|
|
goto fail;
|
|
|
|
if (ret > 0 && path->slots[0] > 0) {
|
|
|
|
leaf = path->nodes[0];
|
|
|
|
btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
|
|
|
|
if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
|
|
|
|
key.type == BTRFS_EXTENT_CSUM_KEY) {
|
|
|
|
offset = (start - key.offset) >>
|
|
|
|
root->fs_info->sb->s_blocksize_bits;
|
|
|
|
if (offset * csum_size <
|
|
|
|
btrfs_item_size_nr(leaf, path->slots[0] - 1))
|
|
|
|
path->slots[0]--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
while (start <= end) {
|
|
|
|
leaf = path->nodes[0];
|
|
|
|
if (path->slots[0] >= btrfs_header_nritems(leaf)) {
|
2009-01-06 11:42:00 -05:00
|
|
|
ret = btrfs_next_leaf(root, path);
|
2008-12-12 10:03:38 -05:00
|
|
|
if (ret < 0)
|
|
|
|
goto fail;
|
|
|
|
if (ret > 0)
|
|
|
|
break;
|
|
|
|
leaf = path->nodes[0];
|
|
|
|
}
|
|
|
|
|
|
|
|
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
|
|
|
|
if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
|
|
|
|
key.type != BTRFS_EXTENT_CSUM_KEY)
|
|
|
|
break;
|
|
|
|
|
|
|
|
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
|
|
|
|
if (key.offset > end)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (key.offset > start)
|
|
|
|
start = key.offset;
|
|
|
|
|
|
|
|
size = btrfs_item_size_nr(leaf, path->slots[0]);
|
|
|
|
csum_end = key.offset + (size / csum_size) * root->sectorsize;
|
2008-12-17 10:21:48 -05:00
|
|
|
if (csum_end <= start) {
|
|
|
|
path->slots[0]++;
|
|
|
|
continue;
|
|
|
|
}
|
2008-12-12 10:03:38 -05:00
|
|
|
|
2009-01-06 11:42:00 -05:00
|
|
|
csum_end = min(csum_end, end + 1);
|
2008-12-12 10:03:38 -05:00
|
|
|
item = btrfs_item_ptr(path->nodes[0], path->slots[0],
|
|
|
|
struct btrfs_csum_item);
|
2009-01-06 11:42:00 -05:00
|
|
|
while (start < csum_end) {
|
|
|
|
size = min_t(size_t, csum_end - start,
|
|
|
|
MAX_ORDERED_SUM_BYTES(root));
|
|
|
|
sums = kzalloc(btrfs_ordered_sum_size(root, size),
|
|
|
|
GFP_NOFS);
|
|
|
|
BUG_ON(!sums);
|
2008-12-12 10:03:38 -05:00
|
|
|
|
2009-01-06 11:42:00 -05:00
|
|
|
sector_sum = sums->sums;
|
|
|
|
sums->bytenr = start;
|
|
|
|
sums->len = size;
|
|
|
|
|
|
|
|
offset = (start - key.offset) >>
|
|
|
|
root->fs_info->sb->s_blocksize_bits;
|
|
|
|
offset *= csum_size;
|
|
|
|
|
|
|
|
while (size > 0) {
|
|
|
|
read_extent_buffer(path->nodes[0],
|
|
|
|
§or_sum->sum,
|
|
|
|
((unsigned long)item) +
|
|
|
|
offset, csum_size);
|
|
|
|
sector_sum->bytenr = start;
|
|
|
|
|
|
|
|
size -= root->sectorsize;
|
|
|
|
start += root->sectorsize;
|
|
|
|
offset += csum_size;
|
|
|
|
sector_sum++;
|
|
|
|
}
|
|
|
|
list_add_tail(&sums->list, list);
|
|
|
|
}
|
2008-12-12 10:03:38 -05:00
|
|
|
path->slots[0]++;
|
|
|
|
}
|
|
|
|
ret = 0;
|
|
|
|
fail:
|
|
|
|
btrfs_free_path(path);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-07-18 06:17:13 -04:00
|
|
|
int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
struct bio *bio, u64 file_start, int contig)
|
2008-04-16 11:15:20 -04:00
|
|
|
{
|
2008-07-17 12:53:50 -04:00
|
|
|
struct btrfs_ordered_sum *sums;
|
|
|
|
struct btrfs_sector_sum *sector_sum;
|
2008-07-18 06:17:13 -04:00
|
|
|
struct btrfs_ordered_extent *ordered;
|
2008-04-16 11:15:20 -04:00
|
|
|
char *data;
|
|
|
|
struct bio_vec *bvec = bio->bi_io_vec;
|
|
|
|
int bio_index = 0;
|
2008-07-18 06:17:13 -04:00
|
|
|
unsigned long total_bytes = 0;
|
|
|
|
unsigned long this_sum_bytes = 0;
|
|
|
|
u64 offset;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
u64 disk_bytenr;
|
2008-04-16 11:15:20 -04:00
|
|
|
|
2008-07-17 12:53:50 -04:00
|
|
|
WARN_ON(bio->bi_vcnt <= 0);
|
|
|
|
sums = kzalloc(btrfs_ordered_sum_size(root, bio->bi_size), GFP_NOFS);
|
2008-04-16 11:15:20 -04:00
|
|
|
if (!sums)
|
|
|
|
return -ENOMEM;
|
2008-07-18 06:17:13 -04:00
|
|
|
|
2008-07-22 23:06:42 -04:00
|
|
|
sector_sum = sums->sums;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
disk_bytenr = (u64)bio->bi_sector << 9;
|
2008-07-17 12:53:50 -04:00
|
|
|
sums->len = bio->bi_size;
|
|
|
|
INIT_LIST_HEAD(&sums->list);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
|
|
|
|
if (contig)
|
|
|
|
offset = file_start;
|
|
|
|
else
|
|
|
|
offset = page_offset(bvec->bv_page) + bvec->bv_offset;
|
|
|
|
|
|
|
|
ordered = btrfs_lookup_ordered_extent(inode, offset);
|
2008-07-18 06:17:13 -04:00
|
|
|
BUG_ON(!ordered);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
sums->bytenr = ordered->start;
|
2008-04-16 11:15:20 -04:00
|
|
|
|
2009-01-05 21:25:51 -05:00
|
|
|
while (bio_index < bio->bi_vcnt) {
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (!contig)
|
|
|
|
offset = page_offset(bvec->bv_page) + bvec->bv_offset;
|
|
|
|
|
|
|
|
if (!contig && (offset >= ordered->file_offset + ordered->len ||
|
|
|
|
offset < ordered->file_offset)) {
|
2008-07-18 06:17:13 -04:00
|
|
|
unsigned long bytes_left;
|
|
|
|
sums->len = this_sum_bytes;
|
|
|
|
this_sum_bytes = 0;
|
|
|
|
btrfs_add_ordered_sum(inode, ordered, sums);
|
|
|
|
btrfs_put_ordered_extent(ordered);
|
|
|
|
|
|
|
|
bytes_left = bio->bi_size - total_bytes;
|
|
|
|
|
|
|
|
sums = kzalloc(btrfs_ordered_sum_size(root, bytes_left),
|
|
|
|
GFP_NOFS);
|
|
|
|
BUG_ON(!sums);
|
2008-07-22 23:06:42 -04:00
|
|
|
sector_sum = sums->sums;
|
2008-07-18 06:17:13 -04:00
|
|
|
sums->len = bytes_left;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
ordered = btrfs_lookup_ordered_extent(inode, offset);
|
2008-07-18 06:17:13 -04:00
|
|
|
BUG_ON(!ordered);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
sums->bytenr = ordered->start;
|
2008-07-18 06:17:13 -04:00
|
|
|
}
|
|
|
|
|
2008-04-16 11:15:20 -04:00
|
|
|
data = kmap_atomic(bvec->bv_page, KM_USER0);
|
2008-07-17 12:53:50 -04:00
|
|
|
sector_sum->sum = ~(u32)0;
|
|
|
|
sector_sum->sum = btrfs_csum_data(root,
|
|
|
|
data + bvec->bv_offset,
|
|
|
|
sector_sum->sum,
|
|
|
|
bvec->bv_len);
|
2008-04-16 11:15:20 -04:00
|
|
|
kunmap_atomic(data, KM_USER0);
|
2008-07-17 12:53:50 -04:00
|
|
|
btrfs_csum_final(sector_sum->sum,
|
|
|
|
(char *)§or_sum->sum);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
sector_sum->bytenr = disk_bytenr;
|
2008-07-22 23:06:42 -04:00
|
|
|
|
2008-07-17 12:53:50 -04:00
|
|
|
sector_sum++;
|
2008-04-16 11:15:20 -04:00
|
|
|
bio_index++;
|
2008-07-18 06:17:13 -04:00
|
|
|
total_bytes += bvec->bv_len;
|
|
|
|
this_sum_bytes += bvec->bv_len;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
disk_bytenr += bvec->bv_len;
|
|
|
|
offset += bvec->bv_len;
|
2008-04-16 11:15:20 -04:00
|
|
|
bvec++;
|
|
|
|
}
|
2008-07-22 23:06:42 -04:00
|
|
|
this_sum_bytes = 0;
|
2008-07-18 06:17:13 -04:00
|
|
|
btrfs_add_ordered_sum(inode, ordered, sums);
|
|
|
|
btrfs_put_ordered_extent(ordered);
|
2008-04-16 11:15:20 -04:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-12-10 09:10:46 -05:00
|
|
|
/*
|
|
|
|
* helper function for csum removal, this expects the
|
|
|
|
* key to describe the csum pointed to by the path, and it expects
|
|
|
|
* the csum to overlap the range [bytenr, len]
|
|
|
|
*
|
|
|
|
* The csum should not be entirely contained in the range and the
|
|
|
|
* range should not be entirely contained in the csum.
|
|
|
|
*
|
|
|
|
* This calls btrfs_truncate_item with the correct args based on the
|
|
|
|
* overlap, and fixes up the key as required.
|
|
|
|
*/
|
|
|
|
static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
|
|
|
|
struct btrfs_root *root,
|
|
|
|
struct btrfs_path *path,
|
|
|
|
struct btrfs_key *key,
|
|
|
|
u64 bytenr, u64 len)
|
|
|
|
{
|
|
|
|
struct extent_buffer *leaf;
|
|
|
|
u16 csum_size =
|
|
|
|
btrfs_super_csum_size(&root->fs_info->super_copy);
|
|
|
|
u64 csum_end;
|
|
|
|
u64 end_byte = bytenr + len;
|
|
|
|
u32 blocksize_bits = root->fs_info->sb->s_blocksize_bits;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
leaf = path->nodes[0];
|
|
|
|
csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
|
|
|
|
csum_end <<= root->fs_info->sb->s_blocksize_bits;
|
|
|
|
csum_end += key->offset;
|
|
|
|
|
|
|
|
if (key->offset < bytenr && csum_end <= end_byte) {
|
|
|
|
/*
|
|
|
|
* [ bytenr - len ]
|
|
|
|
* [ ]
|
|
|
|
* [csum ]
|
|
|
|
* A simple truncate off the end of the item
|
|
|
|
*/
|
|
|
|
u32 new_size = (bytenr - key->offset) >> blocksize_bits;
|
|
|
|
new_size *= csum_size;
|
|
|
|
ret = btrfs_truncate_item(trans, root, path, new_size, 1);
|
|
|
|
BUG_ON(ret);
|
|
|
|
} else if (key->offset >= bytenr && csum_end > end_byte &&
|
|
|
|
end_byte > key->offset) {
|
|
|
|
/*
|
|
|
|
* [ bytenr - len ]
|
|
|
|
* [ ]
|
|
|
|
* [csum ]
|
|
|
|
* we need to truncate from the beginning of the csum
|
|
|
|
*/
|
|
|
|
u32 new_size = (csum_end - end_byte) >> blocksize_bits;
|
|
|
|
new_size *= csum_size;
|
|
|
|
|
|
|
|
ret = btrfs_truncate_item(trans, root, path, new_size, 0);
|
|
|
|
BUG_ON(ret);
|
|
|
|
|
|
|
|
key->offset = end_byte;
|
|
|
|
ret = btrfs_set_item_key_safe(trans, root, path, key);
|
|
|
|
BUG_ON(ret);
|
|
|
|
} else {
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* deletes the csum items from the csum tree for a given
|
|
|
|
* range of bytes.
|
|
|
|
*/
|
|
|
|
int btrfs_del_csums(struct btrfs_trans_handle *trans,
|
|
|
|
struct btrfs_root *root, u64 bytenr, u64 len)
|
|
|
|
{
|
|
|
|
struct btrfs_path *path;
|
|
|
|
struct btrfs_key key;
|
|
|
|
u64 end_byte = bytenr + len;
|
|
|
|
u64 csum_end;
|
|
|
|
struct extent_buffer *leaf;
|
|
|
|
int ret;
|
|
|
|
u16 csum_size =
|
|
|
|
btrfs_super_csum_size(&root->fs_info->super_copy);
|
|
|
|
int blocksize_bits = root->fs_info->sb->s_blocksize_bits;
|
|
|
|
|
|
|
|
root = root->fs_info->csum_root;
|
|
|
|
|
|
|
|
path = btrfs_alloc_path();
|
|
|
|
|
2009-01-05 21:25:51 -05:00
|
|
|
while (1) {
|
2008-12-10 09:10:46 -05:00
|
|
|
key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
|
|
|
|
key.offset = end_byte - 1;
|
|
|
|
key.type = BTRFS_EXTENT_CSUM_KEY;
|
|
|
|
|
|
|
|
ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
|
|
|
|
if (ret > 0) {
|
|
|
|
if (path->slots[0] == 0)
|
|
|
|
goto out;
|
|
|
|
path->slots[0]--;
|
|
|
|
}
|
|
|
|
leaf = path->nodes[0];
|
|
|
|
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
|
|
|
|
|
|
|
|
if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
|
|
|
|
key.type != BTRFS_EXTENT_CSUM_KEY) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (key.offset >= end_byte)
|
|
|
|
break;
|
|
|
|
|
|
|
|
csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
|
|
|
|
csum_end <<= blocksize_bits;
|
|
|
|
csum_end += key.offset;
|
|
|
|
|
|
|
|
/* this csum ends before we start, we're done */
|
|
|
|
if (csum_end <= bytenr)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* delete the entire item, it is inside our range */
|
|
|
|
if (key.offset >= bytenr && csum_end <= end_byte) {
|
|
|
|
ret = btrfs_del_item(trans, root, path);
|
|
|
|
BUG_ON(ret);
|
2008-12-16 13:51:01 -05:00
|
|
|
if (key.offset == bytenr)
|
|
|
|
break;
|
2008-12-10 09:10:46 -05:00
|
|
|
} else if (key.offset < bytenr && csum_end > end_byte) {
|
|
|
|
unsigned long offset;
|
|
|
|
unsigned long shift_len;
|
|
|
|
unsigned long item_offset;
|
|
|
|
/*
|
|
|
|
* [ bytenr - len ]
|
|
|
|
* [csum ]
|
|
|
|
*
|
|
|
|
* Our bytes are in the middle of the csum,
|
|
|
|
* we need to split this item and insert a new one.
|
|
|
|
*
|
|
|
|
* But we can't drop the path because the
|
|
|
|
* csum could change, get removed, extended etc.
|
|
|
|
*
|
|
|
|
* The trick here is the max size of a csum item leaves
|
|
|
|
* enough room in the tree block for a single
|
|
|
|
* item header. So, we split the item in place,
|
|
|
|
* adding a new header pointing to the existing
|
|
|
|
* bytes. Then we loop around again and we have
|
|
|
|
* a nicely formed csum item that we can neatly
|
|
|
|
* truncate.
|
|
|
|
*/
|
|
|
|
offset = (bytenr - key.offset) >> blocksize_bits;
|
|
|
|
offset *= csum_size;
|
|
|
|
|
|
|
|
shift_len = (len >> blocksize_bits) * csum_size;
|
|
|
|
|
|
|
|
item_offset = btrfs_item_ptr_offset(leaf,
|
|
|
|
path->slots[0]);
|
|
|
|
|
|
|
|
memset_extent_buffer(leaf, 0, item_offset + offset,
|
|
|
|
shift_len);
|
|
|
|
key.offset = bytenr;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* btrfs_split_item returns -EAGAIN when the
|
|
|
|
* item changed size or key
|
|
|
|
*/
|
|
|
|
ret = btrfs_split_item(trans, root, path, &key, offset);
|
|
|
|
BUG_ON(ret && ret != -EAGAIN);
|
|
|
|
|
|
|
|
key.offset = end_byte - 1;
|
|
|
|
} else {
|
|
|
|
ret = truncate_one_csum(trans, root, path,
|
|
|
|
&key, bytenr, len);
|
|
|
|
BUG_ON(ret);
|
2008-12-16 13:51:01 -05:00
|
|
|
if (key.offset < bytenr)
|
|
|
|
break;
|
2008-12-10 09:10:46 -05:00
|
|
|
}
|
|
|
|
btrfs_release_path(root, path);
|
|
|
|
}
|
|
|
|
out:
|
|
|
|
btrfs_free_path(path);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-02-20 12:07:25 -05:00
|
|
|
int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
struct btrfs_root *root,
|
2008-07-17 12:53:50 -04:00
|
|
|
struct btrfs_ordered_sum *sums)
|
2007-03-29 15:15:27 -04:00
|
|
|
{
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
u64 bytenr;
|
2007-03-29 15:15:27 -04:00
|
|
|
int ret;
|
|
|
|
struct btrfs_key file_key;
|
2007-04-16 09:22:45 -04:00
|
|
|
struct btrfs_key found_key;
|
2008-02-20 12:07:25 -05:00
|
|
|
u64 next_offset;
|
2008-07-17 12:53:50 -04:00
|
|
|
u64 total_bytes = 0;
|
2008-02-20 12:07:25 -05:00
|
|
|
int found_next;
|
2007-04-02 11:20:42 -04:00
|
|
|
struct btrfs_path *path;
|
2007-03-29 15:15:27 -04:00
|
|
|
struct btrfs_csum_item *item;
|
2008-02-20 12:07:25 -05:00
|
|
|
struct btrfs_csum_item *item_end;
|
2007-10-15 16:22:25 -04:00
|
|
|
struct extent_buffer *leaf = NULL;
|
2007-04-16 09:22:45 -04:00
|
|
|
u64 csum_offset;
|
2008-07-17 12:53:50 -04:00
|
|
|
struct btrfs_sector_sum *sector_sum;
|
2007-10-25 15:42:56 -04:00
|
|
|
u32 nritems;
|
|
|
|
u32 ins_size;
|
2008-02-20 12:07:25 -05:00
|
|
|
char *eb_map;
|
|
|
|
char *eb_token;
|
|
|
|
unsigned long map_len;
|
|
|
|
unsigned long map_start;
|
2008-12-02 07:17:45 -05:00
|
|
|
u16 csum_size =
|
|
|
|
btrfs_super_csum_size(&root->fs_info->super_copy);
|
2008-02-20 12:07:25 -05:00
|
|
|
|
2007-04-02 11:20:42 -04:00
|
|
|
path = btrfs_alloc_path();
|
|
|
|
BUG_ON(!path);
|
2008-07-22 23:06:42 -04:00
|
|
|
sector_sum = sums->sums;
|
2008-02-20 12:07:25 -05:00
|
|
|
again:
|
|
|
|
next_offset = (u64)-1;
|
|
|
|
found_next = 0;
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
|
|
|
|
file_key.offset = sector_sum->bytenr;
|
|
|
|
bytenr = sector_sum->bytenr;
|
|
|
|
btrfs_set_key_type(&file_key, BTRFS_EXTENT_CSUM_KEY);
|
2007-04-18 16:15:28 -04:00
|
|
|
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
item = btrfs_lookup_csum(trans, root, path, sector_sum->bytenr, 1);
|
2007-10-15 16:22:25 -04:00
|
|
|
if (!IS_ERR(item)) {
|
|
|
|
leaf = path->nodes[0];
|
2008-08-28 06:15:25 -04:00
|
|
|
ret = 0;
|
2007-04-18 16:15:28 -04:00
|
|
|
goto found;
|
2007-10-15 16:22:25 -04:00
|
|
|
}
|
2007-04-18 16:15:28 -04:00
|
|
|
ret = PTR_ERR(item);
|
|
|
|
if (ret == -EFBIG) {
|
|
|
|
u32 item_size;
|
|
|
|
/* we found one, but it isn't big enough yet */
|
2007-10-15 16:14:19 -04:00
|
|
|
leaf = path->nodes[0];
|
|
|
|
item_size = btrfs_item_size_nr(leaf, path->slots[0]);
|
2008-12-02 07:17:45 -05:00
|
|
|
if ((item_size / csum_size) >=
|
|
|
|
MAX_CSUM_ITEMS(root, csum_size)) {
|
2007-04-18 16:15:28 -04:00
|
|
|
/* already at max size, make a new one */
|
|
|
|
goto insert;
|
|
|
|
}
|
|
|
|
} else {
|
2007-10-25 15:42:56 -04:00
|
|
|
int slot = path->slots[0] + 1;
|
2007-04-18 16:15:28 -04:00
|
|
|
/* we didn't find a csum item, insert one */
|
2007-10-25 15:42:56 -04:00
|
|
|
nritems = btrfs_header_nritems(path->nodes[0]);
|
|
|
|
if (path->slots[0] >= nritems - 1) {
|
|
|
|
ret = btrfs_next_leaf(root, path);
|
2007-10-29 12:01:05 -04:00
|
|
|
if (ret == 1)
|
2007-10-25 15:42:56 -04:00
|
|
|
found_next = 1;
|
2007-10-29 12:01:05 -04:00
|
|
|
if (ret != 0)
|
2007-10-25 15:42:56 -04:00
|
|
|
goto insert;
|
2007-10-29 12:01:05 -04:00
|
|
|
slot = 0;
|
2007-10-25 15:42:56 -04:00
|
|
|
}
|
|
|
|
btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
|
|
|
|
found_key.type != BTRFS_EXTENT_CSUM_KEY) {
|
2007-10-25 15:42:56 -04:00
|
|
|
found_next = 1;
|
|
|
|
goto insert;
|
|
|
|
}
|
|
|
|
next_offset = found_key.offset;
|
|
|
|
found_next = 1;
|
2007-04-18 16:15:28 -04:00
|
|
|
goto insert;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* at this point, we know the tree has an item, but it isn't big
|
|
|
|
* enough yet to put our csum in. Grow it
|
|
|
|
*/
|
|
|
|
btrfs_release_path(root, path);
|
2007-04-16 09:22:45 -04:00
|
|
|
ret = btrfs_search_slot(trans, root, &file_key, path,
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_size, 1);
|
2007-04-16 09:22:45 -04:00
|
|
|
if (ret < 0)
|
2008-08-15 15:34:18 -04:00
|
|
|
goto fail_unlock;
|
2008-12-10 09:10:46 -05:00
|
|
|
|
|
|
|
if (ret > 0) {
|
|
|
|
if (path->slots[0] == 0)
|
|
|
|
goto insert;
|
|
|
|
path->slots[0]--;
|
2007-04-16 09:22:45 -04:00
|
|
|
}
|
2008-12-10 09:10:46 -05:00
|
|
|
|
2007-10-15 16:14:19 -04:00
|
|
|
leaf = path->nodes[0];
|
|
|
|
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
csum_offset = (bytenr - found_key.offset) >>
|
2007-04-16 09:22:45 -04:00
|
|
|
root->fs_info->sb->s_blocksize_bits;
|
2008-12-10 09:10:46 -05:00
|
|
|
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (btrfs_key_type(&found_key) != BTRFS_EXTENT_CSUM_KEY ||
|
|
|
|
found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_offset >= MAX_CSUM_ITEMS(root, csum_size)) {
|
2007-04-16 09:22:45 -04:00
|
|
|
goto insert;
|
|
|
|
}
|
2008-12-10 09:10:46 -05:00
|
|
|
|
2007-10-15 16:14:19 -04:00
|
|
|
if (csum_offset >= btrfs_item_size_nr(leaf, path->slots[0]) /
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_size) {
|
|
|
|
u32 diff = (csum_offset + 1) * csum_size;
|
2008-12-10 09:10:46 -05:00
|
|
|
|
|
|
|
/*
|
|
|
|
* is the item big enough already? we dropped our lock
|
|
|
|
* before and need to recheck
|
|
|
|
*/
|
|
|
|
if (diff < btrfs_item_size_nr(leaf, path->slots[0]))
|
|
|
|
goto csum;
|
|
|
|
|
2007-10-15 16:14:19 -04:00
|
|
|
diff = diff - btrfs_item_size_nr(leaf, path->slots[0]);
|
2009-01-05 21:25:51 -05:00
|
|
|
if (diff != csum_size)
|
2007-05-24 13:35:57 -04:00
|
|
|
goto insert;
|
2008-12-10 09:10:46 -05:00
|
|
|
|
2007-04-18 16:15:28 -04:00
|
|
|
ret = btrfs_extend_item(trans, root, path, diff);
|
2007-04-16 09:22:45 -04:00
|
|
|
BUG_ON(ret);
|
|
|
|
goto csum;
|
|
|
|
}
|
|
|
|
|
|
|
|
insert:
|
2007-04-18 16:15:28 -04:00
|
|
|
btrfs_release_path(root, path);
|
2007-04-16 09:22:45 -04:00
|
|
|
csum_offset = 0;
|
2007-10-25 15:42:56 -04:00
|
|
|
if (found_next) {
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
u64 tmp = total_bytes + root->sectorsize;
|
|
|
|
u64 next_sector = sector_sum->bytenr;
|
|
|
|
struct btrfs_sector_sum *next = sector_sum + 1;
|
|
|
|
|
2009-01-05 21:25:51 -05:00
|
|
|
while (tmp < sums->len) {
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (next_sector + root->sectorsize != next->bytenr)
|
|
|
|
break;
|
|
|
|
tmp += root->sectorsize;
|
|
|
|
next_sector = next->bytenr;
|
|
|
|
next++;
|
|
|
|
}
|
|
|
|
tmp = min(tmp, next_offset - file_key.offset);
|
2007-10-25 15:42:56 -04:00
|
|
|
tmp >>= root->fs_info->sb->s_blocksize_bits;
|
|
|
|
tmp = max((u64)1, tmp);
|
2008-12-02 07:17:45 -05:00
|
|
|
tmp = min(tmp, (u64)MAX_CSUM_ITEMS(root, csum_size));
|
|
|
|
ins_size = csum_size * tmp;
|
2007-10-25 15:42:56 -04:00
|
|
|
} else {
|
2008-12-02 07:17:45 -05:00
|
|
|
ins_size = csum_size;
|
2007-10-25 15:42:56 -04:00
|
|
|
}
|
2007-04-02 11:20:42 -04:00
|
|
|
ret = btrfs_insert_empty_item(trans, root, path, &file_key,
|
2007-10-25 15:42:56 -04:00
|
|
|
ins_size);
|
2007-06-22 14:16:25 -04:00
|
|
|
if (ret < 0)
|
2008-08-15 15:34:18 -04:00
|
|
|
goto fail_unlock;
|
2007-04-18 16:15:28 -04:00
|
|
|
if (ret != 0) {
|
|
|
|
WARN_ON(1);
|
2008-08-15 15:34:18 -04:00
|
|
|
goto fail_unlock;
|
2007-04-18 16:15:28 -04:00
|
|
|
}
|
2007-04-16 09:22:45 -04:00
|
|
|
csum:
|
2007-10-15 16:14:19 -04:00
|
|
|
leaf = path->nodes[0];
|
|
|
|
item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
|
2007-03-29 15:15:27 -04:00
|
|
|
ret = 0;
|
2007-05-10 12:36:17 -04:00
|
|
|
item = (struct btrfs_csum_item *)((unsigned char *)item +
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_offset * csum_size);
|
2007-04-17 13:26:50 -04:00
|
|
|
found:
|
2008-02-20 12:07:25 -05:00
|
|
|
item_end = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
|
|
|
|
item_end = (struct btrfs_csum_item *)((unsigned char *)item_end +
|
|
|
|
btrfs_item_size_nr(leaf, path->slots[0]));
|
2008-02-20 12:07:25 -05:00
|
|
|
eb_token = NULL;
|
2008-08-15 15:34:18 -04:00
|
|
|
cond_resched();
|
2008-07-17 12:53:50 -04:00
|
|
|
next_sector:
|
2008-01-29 09:10:27 -05:00
|
|
|
|
2008-02-20 12:07:25 -05:00
|
|
|
if (!eb_token ||
|
2008-12-02 07:17:45 -05:00
|
|
|
(unsigned long)item + csum_size >= map_start + map_len) {
|
2008-02-20 12:07:25 -05:00
|
|
|
int err;
|
|
|
|
|
|
|
|
if (eb_token)
|
2008-02-21 09:30:08 -05:00
|
|
|
unmap_extent_buffer(leaf, eb_token, KM_USER1);
|
2008-02-20 12:07:25 -05:00
|
|
|
eb_token = NULL;
|
|
|
|
err = map_private_extent_buffer(leaf, (unsigned long)item,
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_size,
|
2008-02-20 12:07:25 -05:00
|
|
|
&eb_token, &eb_map,
|
2008-02-21 09:30:08 -05:00
|
|
|
&map_start, &map_len, KM_USER1);
|
2008-02-20 12:07:25 -05:00
|
|
|
if (err)
|
|
|
|
eb_token = NULL;
|
|
|
|
}
|
|
|
|
if (eb_token) {
|
|
|
|
memcpy(eb_token + ((unsigned long)item & (PAGE_CACHE_SIZE - 1)),
|
2008-12-02 07:17:45 -05:00
|
|
|
§or_sum->sum, csum_size);
|
2008-02-20 12:07:25 -05:00
|
|
|
} else {
|
2008-07-17 12:53:50 -04:00
|
|
|
write_extent_buffer(leaf, §or_sum->sum,
|
2008-12-02 07:17:45 -05:00
|
|
|
(unsigned long)item, csum_size);
|
2008-02-20 12:07:25 -05:00
|
|
|
}
|
2008-07-18 12:01:11 -04:00
|
|
|
|
2008-07-17 12:53:50 -04:00
|
|
|
total_bytes += root->sectorsize;
|
|
|
|
sector_sum++;
|
|
|
|
if (total_bytes < sums->len) {
|
2008-02-20 12:07:25 -05:00
|
|
|
item = (struct btrfs_csum_item *)((char *)item +
|
2008-12-02 07:17:45 -05:00
|
|
|
csum_size);
|
Btrfs: move data checksumming into a dedicated tree
Btrfs stores checksums for each data block. Until now, they have
been stored in the subvolume trees, indexed by the inode that is
referencing the data block. This means that when we read the inode,
we've probably read in at least some checksums as well.
But, this has a few problems:
* The checksums are indexed by logical offset in the file. When
compression is on, this means we have to do the expensive checksumming
on the uncompressed data. It would be faster if we could checksum
the compressed data instead.
* If we implement encryption, we'll be checksumming the plain text and
storing that on disk. This is significantly less secure.
* For either compression or encryption, we have to get the plain text
back before we can verify the checksum as correct. This makes the raid
layer balancing and extent moving much more expensive.
* It makes the front end caching code more complex, as we have touch
the subvolume and inodes as we cache extents.
* There is potentitally one copy of the checksum in each subvolume
referencing an extent.
The solution used here is to store the extent checksums in a dedicated
tree. This allows us to index the checksums by phyiscal extent
start and length. It means:
* The checksum is against the data stored on disk, after any compression
or encryption is done.
* The checksum is stored in a central location, and can be verified without
following back references, or reading inodes.
This makes compression significantly faster by reducing the amount of
data that needs to be checksummed. It will also allow much faster
raid management code in general.
The checksums are indexed by a key with a fixed objectid (a magic value
in ctree.h) and offset set to the starting byte of the extent. This
allows us to copy the checksum items into the fsync log tree directly (or
any other tree), without having to invent a second format for them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2008-12-08 16:58:54 -05:00
|
|
|
if (item < item_end && bytenr + PAGE_CACHE_SIZE ==
|
|
|
|
sector_sum->bytenr) {
|
|
|
|
bytenr = sector_sum->bytenr;
|
2008-07-17 12:53:50 -04:00
|
|
|
goto next_sector;
|
2008-02-20 15:44:32 -05:00
|
|
|
}
|
2008-02-20 12:07:25 -05:00
|
|
|
}
|
2008-02-20 12:07:25 -05:00
|
|
|
if (eb_token) {
|
2008-02-21 09:30:08 -05:00
|
|
|
unmap_extent_buffer(leaf, eb_token, KM_USER1);
|
2008-02-20 12:07:25 -05:00
|
|
|
eb_token = NULL;
|
|
|
|
}
|
2007-04-02 11:20:42 -04:00
|
|
|
btrfs_mark_buffer_dirty(path->nodes[0]);
|
2008-08-15 15:34:18 -04:00
|
|
|
cond_resched();
|
2008-07-17 12:53:50 -04:00
|
|
|
if (total_bytes < sums->len) {
|
2008-02-20 12:07:25 -05:00
|
|
|
btrfs_release_path(root, path);
|
|
|
|
goto again;
|
|
|
|
}
|
2008-08-15 15:34:18 -04:00
|
|
|
out:
|
2007-04-02 11:20:42 -04:00
|
|
|
btrfs_free_path(path);
|
2007-03-29 15:15:27 -04:00
|
|
|
return ret;
|
2008-08-15 15:34:18 -04:00
|
|
|
|
|
|
|
fail_unlock:
|
|
|
|
goto out;
|
2007-03-29 15:15:27 -04:00
|
|
|
}
|