f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
|
|
|
/*
|
|
|
|
* f2fs compress support
|
|
|
|
*
|
|
|
|
* Copyright (c) 2019 Chao Yu <chao@kernel.org>
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/f2fs_fs.h>
|
|
|
|
#include <linux/writeback.h>
|
|
|
|
#include <linux/backing-dev.h>
|
|
|
|
#include <linux/lzo.h>
|
|
|
|
#include <linux/lz4.h>
|
2020-03-03 09:46:02 +00:00
|
|
|
#include <linux/zstd.h>
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
#include "f2fs.h"
|
|
|
|
#include "node.h"
|
|
|
|
#include <trace/events/f2fs.h>
|
|
|
|
|
2020-09-14 09:05:14 +00:00
|
|
|
static struct kmem_cache *cic_entry_slab;
|
|
|
|
static struct kmem_cache *dic_entry_slab;
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
static void *page_array_alloc(struct inode *inode, int nr)
|
|
|
|
{
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
|
|
|
|
unsigned int size = sizeof(struct page *) * nr;
|
|
|
|
|
|
|
|
if (likely(size <= sbi->page_array_slab_size))
|
|
|
|
return kmem_cache_zalloc(sbi->page_array_slab, GFP_NOFS);
|
|
|
|
return f2fs_kzalloc(sbi, size, GFP_NOFS);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void page_array_free(struct inode *inode, void *pages, int nr)
|
|
|
|
{
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
|
|
|
|
unsigned int size = sizeof(struct page *) * nr;
|
|
|
|
|
|
|
|
if (!pages)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (likely(size <= sbi->page_array_slab_size))
|
|
|
|
kmem_cache_free(sbi->page_array_slab, pages);
|
|
|
|
else
|
|
|
|
kfree(pages);
|
|
|
|
}
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
struct f2fs_compress_ops {
|
|
|
|
int (*init_compress_ctx)(struct compress_ctx *cc);
|
|
|
|
void (*destroy_compress_ctx)(struct compress_ctx *cc);
|
|
|
|
int (*compress_pages)(struct compress_ctx *cc);
|
2020-03-03 08:57:07 +00:00
|
|
|
int (*init_decompress_ctx)(struct decompress_io_ctx *dic);
|
|
|
|
void (*destroy_decompress_ctx)(struct decompress_io_ctx *dic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
int (*decompress_pages)(struct decompress_io_ctx *dic);
|
|
|
|
};
|
|
|
|
|
|
|
|
static unsigned int offset_in_cluster(struct compress_ctx *cc, pgoff_t index)
|
|
|
|
{
|
|
|
|
return index & (cc->cluster_size - 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static pgoff_t cluster_idx(struct compress_ctx *cc, pgoff_t index)
|
|
|
|
{
|
|
|
|
return index >> cc->log_cluster_size;
|
|
|
|
}
|
|
|
|
|
|
|
|
static pgoff_t start_idx_of_cluster(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
return cc->cluster_idx << cc->log_cluster_size;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool f2fs_is_compressed_page(struct page *page)
|
|
|
|
{
|
|
|
|
if (!PagePrivate(page))
|
|
|
|
return false;
|
|
|
|
if (!page_private(page))
|
|
|
|
return false;
|
|
|
|
if (IS_ATOMIC_WRITTEN_PAGE(page) || IS_DUMMY_WRITTEN_PAGE(page))
|
|
|
|
return false;
|
2020-06-20 07:58:29 +00:00
|
|
|
/*
|
|
|
|
* page->private may be set with pid.
|
|
|
|
* pid_max is enough to check if it is traced.
|
|
|
|
*/
|
|
|
|
if (IS_IO_TRACED_PAGE(page))
|
|
|
|
return false;
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
f2fs_bug_on(F2FS_M_SB(page->mapping),
|
|
|
|
*((u32 *)page_private(page)) != F2FS_COMPRESSED_PAGE_MAGIC);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_set_compressed_page(struct page *page,
|
2020-03-28 09:33:23 +00:00
|
|
|
struct inode *inode, pgoff_t index, void *data)
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
{
|
|
|
|
SetPagePrivate(page);
|
|
|
|
set_page_private(page, (unsigned long)data);
|
|
|
|
|
|
|
|
/* i_crypto_info and iv index */
|
|
|
|
page->index = index;
|
|
|
|
page->mapping = inode->i_mapping;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_drop_rpages(struct compress_ctx *cc, int len, bool unlock)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < len; i++) {
|
|
|
|
if (!cc->rpages[i])
|
|
|
|
continue;
|
|
|
|
if (unlock)
|
|
|
|
unlock_page(cc->rpages[i]);
|
|
|
|
else
|
|
|
|
put_page(cc->rpages[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_put_rpages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
f2fs_drop_rpages(cc, cc->cluster_size, false);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_unlock_rpages(struct compress_ctx *cc, int len)
|
|
|
|
{
|
|
|
|
f2fs_drop_rpages(cc, len, true);
|
|
|
|
}
|
|
|
|
|
2020-06-08 12:03:17 +00:00
|
|
|
static void f2fs_put_rpages_mapping(struct address_space *mapping,
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
pgoff_t start, int len)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < len; i++) {
|
|
|
|
struct page *page = find_get_page(mapping, start + i);
|
|
|
|
|
|
|
|
put_page(page);
|
|
|
|
put_page(page);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_put_rpages_wbc(struct compress_ctx *cc,
|
|
|
|
struct writeback_control *wbc, bool redirty, int unlock)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
if (!cc->rpages[i])
|
|
|
|
continue;
|
|
|
|
if (redirty)
|
|
|
|
redirty_page_for_writepage(wbc, cc->rpages[i]);
|
|
|
|
f2fs_put_page(cc->rpages[i], unlock);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
struct page *f2fs_compress_control_page(struct page *page)
|
|
|
|
{
|
|
|
|
return ((struct compress_io_ctx *)page_private(page))->rpages[0];
|
|
|
|
}
|
|
|
|
|
|
|
|
int f2fs_init_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
2020-09-23 07:54:50 +00:00
|
|
|
if (cc->rpages)
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return 0;
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
cc->rpages = page_array_alloc(cc->inode, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return cc->rpages ? 0 : -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_destroy_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cc->inode, cc->rpages, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
cc->rpages = NULL;
|
|
|
|
cc->nr_rpages = 0;
|
|
|
|
cc->nr_cpages = 0;
|
|
|
|
cc->cluster_idx = NULL_CLUSTER;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_compress_ctx_add_page(struct compress_ctx *cc, struct page *page)
|
|
|
|
{
|
|
|
|
unsigned int cluster_ofs;
|
|
|
|
|
|
|
|
if (!f2fs_cluster_can_merge_page(cc, page->index))
|
|
|
|
f2fs_bug_on(F2FS_I_SB(cc->inode), 1);
|
|
|
|
|
|
|
|
cluster_ofs = offset_in_cluster(cc, page->index);
|
|
|
|
cc->rpages[cluster_ofs] = page;
|
|
|
|
cc->nr_rpages++;
|
|
|
|
cc->cluster_idx = cluster_idx(cc, page->index);
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_F2FS_FS_LZO
|
|
|
|
static int lzo_init_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
cc->private = f2fs_kvmalloc(F2FS_I_SB(cc->inode),
|
|
|
|
LZO1X_MEM_COMPRESS, GFP_NOFS);
|
|
|
|
if (!cc->private)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
cc->clen = lzo1x_worst_compress(PAGE_SIZE << cc->log_cluster_size);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void lzo_destroy_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
kvfree(cc->private);
|
|
|
|
cc->private = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int lzo_compress_pages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = lzo1x_1_compress(cc->rbuf, cc->rlen, cc->cbuf->cdata,
|
|
|
|
&cc->clen, cc->private);
|
|
|
|
if (ret != LZO_E_OK) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lzo compress failed, ret:%d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, ret);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int lzo_decompress_pages(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = lzo1x_decompress_safe(dic->cbuf->cdata, dic->clen,
|
|
|
|
dic->rbuf, &dic->rlen);
|
|
|
|
if (ret != LZO_E_OK) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lzo decompress failed, ret:%d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, ret);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dic->rlen != PAGE_SIZE << dic->log_cluster_size) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lzo invalid rlen:%zu, "
|
|
|
|
"expected:%lu\n", KERN_ERR,
|
|
|
|
F2FS_I_SB(dic->inode)->sb->s_id,
|
|
|
|
dic->rlen,
|
|
|
|
PAGE_SIZE << dic->log_cluster_size);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct f2fs_compress_ops f2fs_lzo_ops = {
|
|
|
|
.init_compress_ctx = lzo_init_compress_ctx,
|
|
|
|
.destroy_compress_ctx = lzo_destroy_compress_ctx,
|
|
|
|
.compress_pages = lzo_compress_pages,
|
|
|
|
.decompress_pages = lzo_decompress_pages,
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_F2FS_FS_LZ4
|
|
|
|
static int lz4_init_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
cc->private = f2fs_kvmalloc(F2FS_I_SB(cc->inode),
|
|
|
|
LZ4_MEM_COMPRESS, GFP_NOFS);
|
|
|
|
if (!cc->private)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2020-05-09 07:01:04 +00:00
|
|
|
/*
|
|
|
|
* we do not change cc->clen to LZ4_compressBound(inputsize) to
|
|
|
|
* adapt worst compress case, because lz4 compressor can handle
|
|
|
|
* output budget properly.
|
|
|
|
*/
|
|
|
|
cc->clen = cc->rlen - PAGE_SIZE - COMPRESS_HEADER_SIZE;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void lz4_destroy_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
kvfree(cc->private);
|
|
|
|
cc->private = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int lz4_compress_pages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
int len;
|
|
|
|
|
|
|
|
len = LZ4_compress_default(cc->rbuf, cc->cbuf->cdata, cc->rlen,
|
|
|
|
cc->clen, cc->private);
|
2020-05-09 07:01:04 +00:00
|
|
|
if (!len)
|
|
|
|
return -EAGAIN;
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
cc->clen = len;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int lz4_decompress_pages(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = LZ4_decompress_safe(dic->cbuf->cdata, dic->rbuf,
|
|
|
|
dic->clen, dic->rlen);
|
|
|
|
if (ret < 0) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lz4 decompress failed, ret:%d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, ret);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ret != PAGE_SIZE << dic->log_cluster_size) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lz4 invalid rlen:%zu, "
|
|
|
|
"expected:%lu\n", KERN_ERR,
|
|
|
|
F2FS_I_SB(dic->inode)->sb->s_id,
|
|
|
|
dic->rlen,
|
|
|
|
PAGE_SIZE << dic->log_cluster_size);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct f2fs_compress_ops f2fs_lz4_ops = {
|
|
|
|
.init_compress_ctx = lz4_init_compress_ctx,
|
|
|
|
.destroy_compress_ctx = lz4_destroy_compress_ctx,
|
|
|
|
.compress_pages = lz4_compress_pages,
|
|
|
|
.decompress_pages = lz4_decompress_pages,
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
2020-03-03 09:46:02 +00:00
|
|
|
#ifdef CONFIG_F2FS_FS_ZSTD
|
|
|
|
#define F2FS_ZSTD_DEFAULT_CLEVEL 1
|
|
|
|
|
|
|
|
static int zstd_init_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
ZSTD_parameters params;
|
|
|
|
ZSTD_CStream *stream;
|
|
|
|
void *workspace;
|
|
|
|
unsigned int workspace_size;
|
|
|
|
|
|
|
|
params = ZSTD_getParams(F2FS_ZSTD_DEFAULT_CLEVEL, cc->rlen, 0);
|
|
|
|
workspace_size = ZSTD_CStreamWorkspaceBound(params.cParams);
|
|
|
|
|
|
|
|
workspace = f2fs_kvmalloc(F2FS_I_SB(cc->inode),
|
|
|
|
workspace_size, GFP_NOFS);
|
|
|
|
if (!workspace)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
stream = ZSTD_initCStream(params, 0, workspace, workspace_size);
|
|
|
|
if (!stream) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initCStream failed\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id,
|
|
|
|
__func__);
|
|
|
|
kvfree(workspace);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
cc->private = workspace;
|
|
|
|
cc->private2 = stream;
|
|
|
|
|
|
|
|
cc->clen = cc->rlen - PAGE_SIZE - COMPRESS_HEADER_SIZE;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zstd_destroy_compress_ctx(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
kvfree(cc->private);
|
|
|
|
cc->private = NULL;
|
|
|
|
cc->private2 = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int zstd_compress_pages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
ZSTD_CStream *stream = cc->private2;
|
|
|
|
ZSTD_inBuffer inbuf;
|
|
|
|
ZSTD_outBuffer outbuf;
|
|
|
|
int src_size = cc->rlen;
|
|
|
|
int dst_size = src_size - PAGE_SIZE - COMPRESS_HEADER_SIZE;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
inbuf.pos = 0;
|
|
|
|
inbuf.src = cc->rbuf;
|
|
|
|
inbuf.size = src_size;
|
|
|
|
|
|
|
|
outbuf.pos = 0;
|
|
|
|
outbuf.dst = cc->cbuf->cdata;
|
|
|
|
outbuf.size = dst_size;
|
|
|
|
|
|
|
|
ret = ZSTD_compressStream(stream, &outbuf, &inbuf);
|
|
|
|
if (ZSTD_isError(ret)) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id,
|
|
|
|
__func__, ZSTD_getErrorCode(ret));
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = ZSTD_endStream(stream, &outbuf);
|
|
|
|
if (ZSTD_isError(ret)) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_endStream returned %d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id,
|
|
|
|
__func__, ZSTD_getErrorCode(ret));
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2020-05-08 01:16:03 +00:00
|
|
|
/*
|
|
|
|
* there is compressed data remained in intermediate buffer due to
|
|
|
|
* no more space in cbuf.cdata
|
|
|
|
*/
|
|
|
|
if (ret)
|
|
|
|
return -EAGAIN;
|
|
|
|
|
2020-03-03 09:46:02 +00:00
|
|
|
cc->clen = outbuf.pos;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int zstd_init_decompress_ctx(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
ZSTD_DStream *stream;
|
|
|
|
void *workspace;
|
|
|
|
unsigned int workspace_size;
|
2020-09-02 07:01:52 +00:00
|
|
|
unsigned int max_window_size =
|
|
|
|
MAX_COMPRESS_WINDOW_SIZE(dic->log_cluster_size);
|
2020-03-03 09:46:02 +00:00
|
|
|
|
2020-09-02 07:01:52 +00:00
|
|
|
workspace_size = ZSTD_DStreamWorkspaceBound(max_window_size);
|
2020-03-03 09:46:02 +00:00
|
|
|
|
|
|
|
workspace = f2fs_kvmalloc(F2FS_I_SB(dic->inode),
|
|
|
|
workspace_size, GFP_NOFS);
|
|
|
|
if (!workspace)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2020-09-02 07:01:52 +00:00
|
|
|
stream = ZSTD_initDStream(max_window_size, workspace, workspace_size);
|
2020-03-03 09:46:02 +00:00
|
|
|
if (!stream) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initDStream failed\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id,
|
|
|
|
__func__);
|
|
|
|
kvfree(workspace);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
dic->private = workspace;
|
|
|
|
dic->private2 = stream;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zstd_destroy_decompress_ctx(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
kvfree(dic->private);
|
|
|
|
dic->private = NULL;
|
|
|
|
dic->private2 = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int zstd_decompress_pages(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
ZSTD_DStream *stream = dic->private2;
|
|
|
|
ZSTD_inBuffer inbuf;
|
|
|
|
ZSTD_outBuffer outbuf;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
inbuf.pos = 0;
|
|
|
|
inbuf.src = dic->cbuf->cdata;
|
|
|
|
inbuf.size = dic->clen;
|
|
|
|
|
|
|
|
outbuf.pos = 0;
|
|
|
|
outbuf.dst = dic->rbuf;
|
|
|
|
outbuf.size = dic->rlen;
|
|
|
|
|
|
|
|
ret = ZSTD_decompressStream(stream, &outbuf, &inbuf);
|
|
|
|
if (ZSTD_isError(ret)) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id,
|
|
|
|
__func__, ZSTD_getErrorCode(ret));
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dic->rlen != outbuf.pos) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): %s ZSTD invalid rlen:%zu, "
|
|
|
|
"expected:%lu\n", KERN_ERR,
|
|
|
|
F2FS_I_SB(dic->inode)->sb->s_id,
|
|
|
|
__func__, dic->rlen,
|
|
|
|
PAGE_SIZE << dic->log_cluster_size);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct f2fs_compress_ops f2fs_zstd_ops = {
|
|
|
|
.init_compress_ctx = zstd_init_compress_ctx,
|
|
|
|
.destroy_compress_ctx = zstd_destroy_compress_ctx,
|
|
|
|
.compress_pages = zstd_compress_pages,
|
|
|
|
.init_decompress_ctx = zstd_init_decompress_ctx,
|
|
|
|
.destroy_decompress_ctx = zstd_destroy_decompress_ctx,
|
|
|
|
.decompress_pages = zstd_decompress_pages,
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
2020-04-08 11:56:32 +00:00
|
|
|
#ifdef CONFIG_F2FS_FS_LZO
|
|
|
|
#ifdef CONFIG_F2FS_FS_LZORLE
|
|
|
|
static int lzorle_compress_pages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = lzorle1x_1_compress(cc->rbuf, cc->rlen, cc->cbuf->cdata,
|
|
|
|
&cc->clen, cc->private);
|
|
|
|
if (ret != LZO_E_OK) {
|
|
|
|
printk_ratelimited("%sF2FS-fs (%s): lzo-rle compress failed, ret:%d\n",
|
|
|
|
KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, ret);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct f2fs_compress_ops f2fs_lzorle_ops = {
|
|
|
|
.init_compress_ctx = lzo_init_compress_ctx,
|
|
|
|
.destroy_compress_ctx = lzo_destroy_compress_ctx,
|
|
|
|
.compress_pages = lzorle_compress_pages,
|
|
|
|
.decompress_pages = lzo_decompress_pages,
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
static const struct f2fs_compress_ops *f2fs_cops[COMPRESS_MAX] = {
|
|
|
|
#ifdef CONFIG_F2FS_FS_LZO
|
|
|
|
&f2fs_lzo_ops,
|
|
|
|
#else
|
|
|
|
NULL,
|
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_F2FS_FS_LZ4
|
|
|
|
&f2fs_lz4_ops,
|
|
|
|
#else
|
|
|
|
NULL,
|
|
|
|
#endif
|
2020-03-03 09:46:02 +00:00
|
|
|
#ifdef CONFIG_F2FS_FS_ZSTD
|
|
|
|
&f2fs_zstd_ops,
|
|
|
|
#else
|
|
|
|
NULL,
|
|
|
|
#endif
|
2020-04-08 11:56:32 +00:00
|
|
|
#if defined(CONFIG_F2FS_FS_LZO) && defined(CONFIG_F2FS_FS_LZORLE)
|
|
|
|
&f2fs_lzorle_ops,
|
|
|
|
#else
|
|
|
|
NULL,
|
|
|
|
#endif
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
bool f2fs_is_compress_backend_ready(struct inode *inode)
|
|
|
|
{
|
|
|
|
if (!f2fs_compressed_file(inode))
|
|
|
|
return true;
|
|
|
|
return f2fs_cops[F2FS_I(inode)->i_compress_algorithm];
|
|
|
|
}
|
|
|
|
|
2020-06-16 16:56:51 +00:00
|
|
|
static mempool_t *compress_page_pool;
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
static int num_compress_pages = 512;
|
|
|
|
module_param(num_compress_pages, uint, 0444);
|
|
|
|
MODULE_PARM_DESC(num_compress_pages,
|
|
|
|
"Number of intermediate compress pages to preallocate");
|
|
|
|
|
|
|
|
int f2fs_init_compress_mempool(void)
|
|
|
|
{
|
|
|
|
compress_page_pool = mempool_create_page_pool(num_compress_pages, 0);
|
|
|
|
if (!compress_page_pool)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_destroy_compress_mempool(void)
|
|
|
|
{
|
|
|
|
mempool_destroy(compress_page_pool);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct page *f2fs_compress_alloc_page(void)
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
{
|
|
|
|
struct page *page;
|
|
|
|
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
page = mempool_alloc(compress_page_pool, GFP_NOFS);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
lock_page(page);
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return page;
|
|
|
|
}
|
|
|
|
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
static void f2fs_compress_free_page(struct page *page)
|
|
|
|
{
|
|
|
|
if (!page)
|
|
|
|
return;
|
|
|
|
set_page_private(page, (unsigned long)NULL);
|
|
|
|
ClearPagePrivate(page);
|
|
|
|
page->mapping = NULL;
|
|
|
|
unlock_page(page);
|
|
|
|
mempool_free(page, compress_page_pool);
|
|
|
|
}
|
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
#define MAX_VMAP_RETRIES 3
|
|
|
|
|
|
|
|
static void *f2fs_vmap(struct page **pages, unsigned int count)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
void *buf = NULL;
|
|
|
|
|
|
|
|
for (i = 0; i < MAX_VMAP_RETRIES; i++) {
|
|
|
|
buf = vm_map_ram(pages, count, -1);
|
|
|
|
if (buf)
|
|
|
|
break;
|
|
|
|
vm_unmap_aliases();
|
|
|
|
}
|
|
|
|
return buf;
|
|
|
|
}
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
static int f2fs_compress_pages(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
struct f2fs_inode_info *fi = F2FS_I(cc->inode);
|
|
|
|
const struct f2fs_compress_ops *cops =
|
|
|
|
f2fs_cops[fi->i_compress_algorithm];
|
2020-09-14 09:05:13 +00:00
|
|
|
unsigned int max_len, new_nr_cpages;
|
|
|
|
struct page **new_cpages;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
int i, ret;
|
|
|
|
|
|
|
|
trace_f2fs_compress_pages_start(cc->inode, cc->cluster_idx,
|
|
|
|
cc->cluster_size, fi->i_compress_algorithm);
|
|
|
|
|
2020-03-03 08:57:07 +00:00
|
|
|
if (cops->init_compress_ctx) {
|
|
|
|
ret = cops->init_compress_ctx(cc);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
}
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
max_len = COMPRESS_HEADER_SIZE + cc->clen;
|
|
|
|
cc->nr_cpages = DIV_ROUND_UP(max_len, PAGE_SIZE);
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
cc->cpages = page_array_alloc(cc->inode, cc->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cc->cpages) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto destroy_compress_ctx;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < cc->nr_cpages; i++) {
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
cc->cpages[i] = f2fs_compress_alloc_page();
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cc->cpages[i]) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_free_cpages;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
cc->rbuf = f2fs_vmap(cc->rpages, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cc->rbuf) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_free_cpages;
|
|
|
|
}
|
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
cc->cbuf = f2fs_vmap(cc->cpages, cc->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cc->cbuf) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_vunmap_rbuf;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = cops->compress_pages(cc);
|
|
|
|
if (ret)
|
|
|
|
goto out_vunmap_cbuf;
|
|
|
|
|
|
|
|
max_len = PAGE_SIZE * (cc->cluster_size - 1) - COMPRESS_HEADER_SIZE;
|
|
|
|
|
|
|
|
if (cc->clen > max_len) {
|
|
|
|
ret = -EAGAIN;
|
|
|
|
goto out_vunmap_cbuf;
|
|
|
|
}
|
|
|
|
|
|
|
|
cc->cbuf->clen = cpu_to_le32(cc->clen);
|
|
|
|
|
|
|
|
for (i = 0; i < COMPRESS_DATA_RESERVED_SIZE; i++)
|
|
|
|
cc->cbuf->reserved[i] = cpu_to_le32(0);
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
new_nr_cpages = DIV_ROUND_UP(cc->clen + COMPRESS_HEADER_SIZE, PAGE_SIZE);
|
|
|
|
|
|
|
|
/* Now we're going to cut unnecessary tail pages */
|
|
|
|
new_cpages = page_array_alloc(cc->inode, new_nr_cpages);
|
|
|
|
if (!new_cpages) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_vunmap_cbuf;
|
|
|
|
}
|
2020-02-21 04:50:37 +00:00
|
|
|
|
|
|
|
/* zero out any unused part of the last page */
|
|
|
|
memset(&cc->cbuf->cdata[cc->clen], 0,
|
2020-09-14 09:05:13 +00:00
|
|
|
(new_nr_cpages * PAGE_SIZE) -
|
|
|
|
(cc->clen + COMPRESS_HEADER_SIZE));
|
2020-02-21 04:50:37 +00:00
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
vm_unmap_ram(cc->cbuf, cc->nr_cpages);
|
|
|
|
vm_unmap_ram(cc->rbuf, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
for (i = 0; i < cc->nr_cpages; i++) {
|
|
|
|
if (i < new_nr_cpages) {
|
|
|
|
new_cpages[i] = cc->cpages[i];
|
|
|
|
continue;
|
|
|
|
}
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
f2fs_compress_free_page(cc->cpages[i]);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
cc->cpages[i] = NULL;
|
|
|
|
}
|
|
|
|
|
2020-03-03 08:57:07 +00:00
|
|
|
if (cops->destroy_compress_ctx)
|
|
|
|
cops->destroy_compress_ctx(cc);
|
2020-03-03 08:57:06 +00:00
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cc->inode, cc->cpages, cc->nr_cpages);
|
|
|
|
cc->cpages = new_cpages;
|
|
|
|
cc->nr_cpages = new_nr_cpages;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx,
|
|
|
|
cc->clen, ret);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_vunmap_cbuf:
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
vm_unmap_ram(cc->cbuf, cc->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out_vunmap_rbuf:
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
vm_unmap_ram(cc->rbuf, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out_free_cpages:
|
|
|
|
for (i = 0; i < cc->nr_cpages; i++) {
|
|
|
|
if (cc->cpages[i])
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
f2fs_compress_free_page(cc->cpages[i]);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cc->inode, cc->cpages, cc->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
cc->cpages = NULL;
|
|
|
|
destroy_compress_ctx:
|
2020-03-03 08:57:07 +00:00
|
|
|
if (cops->destroy_compress_ctx)
|
|
|
|
cops->destroy_compress_ctx(cc);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out:
|
|
|
|
trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx,
|
|
|
|
cc->clen, ret);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_decompress_pages(struct bio *bio, struct page *page, bool verity)
|
|
|
|
{
|
|
|
|
struct decompress_io_ctx *dic =
|
|
|
|
(struct decompress_io_ctx *)page_private(page);
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(dic->inode);
|
|
|
|
struct f2fs_inode_info *fi= F2FS_I(dic->inode);
|
|
|
|
const struct f2fs_compress_ops *cops =
|
|
|
|
f2fs_cops[fi->i_compress_algorithm];
|
|
|
|
int ret;
|
2020-07-25 01:17:48 +00:00
|
|
|
int i;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
dec_page_count(sbi, F2FS_RD_DATA);
|
|
|
|
|
|
|
|
if (bio->bi_status || PageError(page))
|
|
|
|
dic->failed = true;
|
|
|
|
|
2020-08-10 10:39:30 +00:00
|
|
|
if (atomic_dec_return(&dic->pending_pages))
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return;
|
|
|
|
|
|
|
|
trace_f2fs_decompress_pages_start(dic->inode, dic->cluster_idx,
|
|
|
|
dic->cluster_size, fi->i_compress_algorithm);
|
|
|
|
|
|
|
|
/* submit partial compressed pages */
|
|
|
|
if (dic->failed) {
|
|
|
|
ret = -EIO;
|
|
|
|
goto out_free_dic;
|
|
|
|
}
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
dic->tpages = page_array_alloc(dic->inode, dic->cluster_size);
|
2020-07-25 01:17:48 +00:00
|
|
|
if (!dic->tpages) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_free_dic;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < dic->cluster_size; i++) {
|
|
|
|
if (dic->rpages[i]) {
|
|
|
|
dic->tpages[i] = dic->rpages[i];
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
dic->tpages[i] = f2fs_compress_alloc_page();
|
|
|
|
if (!dic->tpages[i]) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_free_dic;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2020-03-03 08:57:07 +00:00
|
|
|
if (cops->init_decompress_ctx) {
|
|
|
|
ret = cops->init_decompress_ctx(dic);
|
|
|
|
if (ret)
|
|
|
|
goto out_free_dic;
|
|
|
|
}
|
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
dic->rbuf = f2fs_vmap(dic->tpages, dic->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!dic->rbuf) {
|
|
|
|
ret = -ENOMEM;
|
2020-03-03 08:57:07 +00:00
|
|
|
goto destroy_decompress_ctx;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
dic->cbuf = f2fs_vmap(dic->cpages, dic->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!dic->cbuf) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out_vunmap_rbuf;
|
|
|
|
}
|
|
|
|
|
|
|
|
dic->clen = le32_to_cpu(dic->cbuf->clen);
|
|
|
|
dic->rlen = PAGE_SIZE << dic->log_cluster_size;
|
|
|
|
|
|
|
|
if (dic->clen > PAGE_SIZE * dic->nr_cpages - COMPRESS_HEADER_SIZE) {
|
|
|
|
ret = -EFSCORRUPTED;
|
|
|
|
goto out_vunmap_cbuf;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = cops->decompress_pages(dic);
|
|
|
|
|
|
|
|
out_vunmap_cbuf:
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
vm_unmap_ram(dic->cbuf, dic->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out_vunmap_rbuf:
|
f2fs: change virtual mapping way for compression pages
By profiling f2fs compression works, I've found vmap() callings have
unexpected hikes in the execution time in our test environment and
those are bottlenecks of f2fs decompression path. Changing these with
vm_map_ram(), we can enhance f2fs decompression speed pretty much.
[Verification]
Android Pixel 3(ARM64, 6GB RAM, 128GB UFS)
Turned on only 0-3 little cores(at 1.785GHz)
dd if=/dev/zero of=dummy bs=1m count=1000
echo 3 > /proc/sys/vm/drop_caches
dd if=dummy of=/dev/zero bs=512k
- w/o compression -
1048576000 bytes (0.9 G) copied, 2.082554 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.081634 s, 480 M/s
1048576000 bytes (0.9 G) copied, 2.090861 s, 478 M/s
- before patch -
1048576000 bytes (0.9 G) copied, 7.407527 s, 135 M/s
1048576000 bytes (0.9 G) copied, 7.283734 s, 137 M/s
1048576000 bytes (0.9 G) copied, 7.291508 s, 137 M/s
- after patch -
1048576000 bytes (0.9 G) copied, 1.998959 s, 500 M/s
1048576000 bytes (0.9 G) copied, 1.987554 s, 503 M/s
1048576000 bytes (0.9 G) copied, 1.986380 s, 503 M/s
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-08-12 05:17:11 +00:00
|
|
|
vm_unmap_ram(dic->rbuf, dic->cluster_size);
|
2020-03-03 08:57:07 +00:00
|
|
|
destroy_decompress_ctx:
|
|
|
|
if (cops->destroy_decompress_ctx)
|
|
|
|
cops->destroy_decompress_ctx(dic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out_free_dic:
|
2020-03-23 09:43:04 +00:00
|
|
|
if (verity)
|
2020-08-10 10:39:30 +00:00
|
|
|
atomic_set(&dic->pending_pages, dic->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!verity)
|
|
|
|
f2fs_decompress_end_io(dic->rpages, dic->cluster_size,
|
|
|
|
ret, false);
|
|
|
|
|
|
|
|
trace_f2fs_decompress_pages_end(dic->inode, dic->cluster_idx,
|
|
|
|
dic->clen, ret);
|
|
|
|
if (!verity)
|
|
|
|
f2fs_free_dic(dic);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool is_page_in_cluster(struct compress_ctx *cc, pgoff_t index)
|
|
|
|
{
|
|
|
|
if (cc->cluster_idx == NULL_CLUSTER)
|
|
|
|
return true;
|
|
|
|
return cc->cluster_idx == cluster_idx(cc, index);
|
|
|
|
}
|
|
|
|
|
|
|
|
bool f2fs_cluster_is_empty(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
return cc->nr_rpages == 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool f2fs_cluster_is_full(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
return cc->cluster_size == cc->nr_rpages;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool f2fs_cluster_can_merge_page(struct compress_ctx *cc, pgoff_t index)
|
|
|
|
{
|
|
|
|
if (f2fs_cluster_is_empty(cc))
|
|
|
|
return true;
|
|
|
|
return is_page_in_cluster(cc, index);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool __cluster_may_compress(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(cc->inode);
|
|
|
|
loff_t i_size = i_size_read(cc->inode);
|
|
|
|
unsigned nr_pages = DIV_ROUND_UP(i_size, PAGE_SIZE);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
struct page *page = cc->rpages[i];
|
|
|
|
|
|
|
|
f2fs_bug_on(sbi, !page);
|
|
|
|
|
|
|
|
if (unlikely(f2fs_cp_error(sbi)))
|
|
|
|
return false;
|
|
|
|
if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* beyond EOF */
|
|
|
|
if (page->index >= nr_pages)
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2020-03-12 02:45:29 +00:00
|
|
|
static int __f2fs_cluster_blocks(struct compress_ctx *cc, bool compr)
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
{
|
|
|
|
struct dnode_of_data dn;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
set_new_dnode(&dn, cc->inode, NULL, NULL, 0);
|
|
|
|
ret = f2fs_get_dnode_of_data(&dn, start_idx_of_cluster(cc),
|
|
|
|
LOOKUP_NODE);
|
|
|
|
if (ret) {
|
|
|
|
if (ret == -ENOENT)
|
|
|
|
ret = 0;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dn.data_blkaddr == COMPRESS_ADDR) {
|
|
|
|
int i;
|
|
|
|
|
|
|
|
ret = 1;
|
|
|
|
for (i = 1; i < cc->cluster_size; i++) {
|
|
|
|
block_t blkaddr;
|
|
|
|
|
2020-02-14 09:44:10 +00:00
|
|
|
blkaddr = data_blkaddr(dn.inode,
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
dn.node_page, dn.ofs_in_node + i);
|
2020-03-12 02:45:29 +00:00
|
|
|
if (compr) {
|
|
|
|
if (__is_valid_data_blkaddr(blkaddr))
|
|
|
|
ret++;
|
|
|
|
} else {
|
|
|
|
if (blkaddr != NULL_ADDR)
|
|
|
|
ret++;
|
|
|
|
}
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
fail:
|
|
|
|
f2fs_put_dnode(&dn);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2020-03-12 02:45:29 +00:00
|
|
|
/* return # of compressed blocks in compressed cluster */
|
|
|
|
static int f2fs_compressed_blocks(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
return __f2fs_cluster_blocks(cc, true);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* return # of valid blocks in compressed cluster */
|
2020-06-16 02:39:24 +00:00
|
|
|
static int f2fs_cluster_blocks(struct compress_ctx *cc)
|
2020-03-12 02:45:29 +00:00
|
|
|
{
|
|
|
|
return __f2fs_cluster_blocks(cc, false);
|
|
|
|
}
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index)
|
|
|
|
{
|
|
|
|
struct compress_ctx cc = {
|
|
|
|
.inode = inode,
|
|
|
|
.log_cluster_size = F2FS_I(inode)->i_log_cluster_size,
|
|
|
|
.cluster_size = F2FS_I(inode)->i_cluster_size,
|
|
|
|
.cluster_idx = index >> F2FS_I(inode)->i_log_cluster_size,
|
|
|
|
};
|
|
|
|
|
2020-06-16 02:39:24 +00:00
|
|
|
return f2fs_cluster_blocks(&cc);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static bool cluster_may_compress(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
if (!f2fs_compressed_file(cc->inode))
|
|
|
|
return false;
|
|
|
|
if (f2fs_is_atomic_file(cc->inode))
|
|
|
|
return false;
|
|
|
|
if (f2fs_is_mmap_file(cc->inode))
|
|
|
|
return false;
|
|
|
|
if (!f2fs_cluster_is_full(cc))
|
|
|
|
return false;
|
2020-05-26 01:55:02 +00:00
|
|
|
if (unlikely(f2fs_cp_error(F2FS_I_SB(cc->inode))))
|
|
|
|
return false;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return __cluster_may_compress(cc);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void set_cluster_writeback(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
if (cc->rpages[i])
|
|
|
|
set_page_writeback(cc->rpages[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void set_cluster_dirty(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++)
|
|
|
|
if (cc->rpages[i])
|
|
|
|
set_page_dirty(cc->rpages[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int prepare_compress_overwrite(struct compress_ctx *cc,
|
|
|
|
struct page **pagep, pgoff_t index, void **fsdata)
|
|
|
|
{
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(cc->inode);
|
|
|
|
struct address_space *mapping = cc->inode->i_mapping;
|
|
|
|
struct page *page;
|
|
|
|
struct dnode_of_data dn;
|
|
|
|
sector_t last_block_in_bio;
|
|
|
|
unsigned fgp_flag = FGP_LOCK | FGP_WRITE | FGP_CREAT;
|
|
|
|
pgoff_t start_idx = start_idx_of_cluster(cc);
|
|
|
|
int i, ret;
|
|
|
|
bool prealloc;
|
|
|
|
|
|
|
|
retry:
|
2020-06-16 02:39:24 +00:00
|
|
|
ret = f2fs_cluster_blocks(cc);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (ret <= 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
/* compressed case */
|
|
|
|
prealloc = (ret < cc->cluster_size);
|
|
|
|
|
|
|
|
ret = f2fs_init_compress_ctx(cc);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
/* keep page reference to avoid page reclaim */
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
page = f2fs_pagecache_get_page(mapping, start_idx + i,
|
|
|
|
fgp_flag, GFP_NOFS);
|
|
|
|
if (!page) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto unlock_pages;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (PageUptodate(page))
|
|
|
|
unlock_page(page);
|
|
|
|
else
|
|
|
|
f2fs_compress_ctx_add_page(cc, page);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!f2fs_cluster_is_empty(cc)) {
|
|
|
|
struct bio *bio = NULL;
|
|
|
|
|
|
|
|
ret = f2fs_read_multi_pages(cc, &bio, cc->cluster_size,
|
2020-02-18 10:21:35 +00:00
|
|
|
&last_block_in_bio, false, true);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
f2fs_destroy_compress_ctx(cc);
|
|
|
|
if (ret)
|
|
|
|
goto release_pages;
|
|
|
|
if (bio)
|
|
|
|
f2fs_submit_bio(sbi, bio, DATA);
|
|
|
|
|
|
|
|
ret = f2fs_init_compress_ctx(cc);
|
|
|
|
if (ret)
|
|
|
|
goto release_pages;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
f2fs_bug_on(sbi, cc->rpages[i]);
|
|
|
|
|
|
|
|
page = find_lock_page(mapping, start_idx + i);
|
|
|
|
f2fs_bug_on(sbi, !page);
|
|
|
|
|
|
|
|
f2fs_wait_on_page_writeback(page, DATA, true, true);
|
|
|
|
|
|
|
|
f2fs_compress_ctx_add_page(cc, page);
|
|
|
|
f2fs_put_page(page, 0);
|
|
|
|
|
|
|
|
if (!PageUptodate(page)) {
|
|
|
|
f2fs_unlock_rpages(cc, i + 1);
|
2020-06-08 12:03:17 +00:00
|
|
|
f2fs_put_rpages_mapping(mapping, start_idx,
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
cc->cluster_size);
|
|
|
|
f2fs_destroy_compress_ctx(cc);
|
|
|
|
goto retry;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (prealloc) {
|
2020-06-18 06:36:22 +00:00
|
|
|
f2fs_do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, true);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
set_new_dnode(&dn, cc->inode, NULL, NULL, 0);
|
|
|
|
|
|
|
|
for (i = cc->cluster_size - 1; i > 0; i--) {
|
|
|
|
ret = f2fs_get_block(&dn, start_idx + i);
|
|
|
|
if (ret) {
|
|
|
|
i = cc->cluster_size;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dn.data_blkaddr != NEW_ADDR)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2020-06-18 06:36:22 +00:00
|
|
|
f2fs_do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, false);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (likely(!ret)) {
|
|
|
|
*fsdata = cc->rpages;
|
|
|
|
*pagep = cc->rpages[offset_in_cluster(cc, index)];
|
|
|
|
return cc->cluster_size;
|
|
|
|
}
|
|
|
|
|
|
|
|
unlock_pages:
|
|
|
|
f2fs_unlock_rpages(cc, i);
|
|
|
|
release_pages:
|
2020-06-08 12:03:17 +00:00
|
|
|
f2fs_put_rpages_mapping(mapping, start_idx, i);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
f2fs_destroy_compress_ctx(cc);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
int f2fs_prepare_compress_overwrite(struct inode *inode,
|
|
|
|
struct page **pagep, pgoff_t index, void **fsdata)
|
|
|
|
{
|
|
|
|
struct compress_ctx cc = {
|
|
|
|
.inode = inode,
|
|
|
|
.log_cluster_size = F2FS_I(inode)->i_log_cluster_size,
|
|
|
|
.cluster_size = F2FS_I(inode)->i_cluster_size,
|
|
|
|
.cluster_idx = index >> F2FS_I(inode)->i_log_cluster_size,
|
|
|
|
.rpages = NULL,
|
|
|
|
.nr_rpages = 0,
|
|
|
|
};
|
|
|
|
|
|
|
|
return prepare_compress_overwrite(&cc, pagep, index, fsdata);
|
|
|
|
}
|
|
|
|
|
|
|
|
bool f2fs_compress_write_end(struct inode *inode, void *fsdata,
|
|
|
|
pgoff_t index, unsigned copied)
|
|
|
|
|
|
|
|
{
|
|
|
|
struct compress_ctx cc = {
|
2020-09-14 09:05:13 +00:00
|
|
|
.inode = inode,
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
.log_cluster_size = F2FS_I(inode)->i_log_cluster_size,
|
|
|
|
.cluster_size = F2FS_I(inode)->i_cluster_size,
|
|
|
|
.rpages = fsdata,
|
|
|
|
};
|
|
|
|
bool first_index = (index == cc.rpages[0]->index);
|
|
|
|
|
|
|
|
if (copied)
|
|
|
|
set_cluster_dirty(&cc);
|
|
|
|
|
|
|
|
f2fs_put_rpages_wbc(&cc, NULL, false, 1);
|
|
|
|
f2fs_destroy_compress_ctx(&cc);
|
|
|
|
|
|
|
|
return first_index;
|
|
|
|
}
|
|
|
|
|
2020-03-18 08:22:59 +00:00
|
|
|
int f2fs_truncate_partial_cluster(struct inode *inode, u64 from, bool lock)
|
|
|
|
{
|
|
|
|
void *fsdata = NULL;
|
|
|
|
struct page *pagep;
|
|
|
|
int log_cluster_size = F2FS_I(inode)->i_log_cluster_size;
|
|
|
|
pgoff_t start_idx = from >> (PAGE_SHIFT + log_cluster_size) <<
|
|
|
|
log_cluster_size;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
err = f2fs_is_compressed_cluster(inode, start_idx);
|
|
|
|
if (err < 0)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
/* truncate normal cluster */
|
|
|
|
if (!err)
|
|
|
|
return f2fs_do_truncate_blocks(inode, from, lock);
|
|
|
|
|
|
|
|
/* truncate compressed cluster */
|
|
|
|
err = f2fs_prepare_compress_overwrite(inode, &pagep,
|
|
|
|
start_idx, &fsdata);
|
|
|
|
|
|
|
|
/* should not be a normal cluster */
|
|
|
|
f2fs_bug_on(F2FS_I_SB(inode), err == 0);
|
|
|
|
|
|
|
|
if (err <= 0)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (err > 0) {
|
|
|
|
struct page **rpages = fsdata;
|
|
|
|
int cluster_size = F2FS_I(inode)->i_cluster_size;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = cluster_size - 1; i >= 0; i--) {
|
|
|
|
loff_t start = rpages[i]->index << PAGE_SHIFT;
|
|
|
|
|
|
|
|
if (from <= start) {
|
|
|
|
zero_user_segment(rpages[i], 0, PAGE_SIZE);
|
|
|
|
} else {
|
|
|
|
zero_user_segment(rpages[i], from - start,
|
|
|
|
PAGE_SIZE);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
f2fs_compress_write_end(inode, fsdata, start_idx, true);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
static int f2fs_write_compressed_pages(struct compress_ctx *cc,
|
|
|
|
int *submitted,
|
|
|
|
struct writeback_control *wbc,
|
|
|
|
enum iostat_type io_type)
|
|
|
|
{
|
|
|
|
struct inode *inode = cc->inode;
|
|
|
|
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
|
|
|
|
struct f2fs_inode_info *fi = F2FS_I(inode);
|
|
|
|
struct f2fs_io_info fio = {
|
|
|
|
.sbi = sbi,
|
|
|
|
.ino = cc->inode->i_ino,
|
|
|
|
.type = DATA,
|
|
|
|
.op = REQ_OP_WRITE,
|
|
|
|
.op_flags = wbc_to_write_flags(wbc),
|
|
|
|
.old_blkaddr = NEW_ADDR,
|
|
|
|
.page = NULL,
|
|
|
|
.encrypted_page = NULL,
|
|
|
|
.compressed_page = NULL,
|
|
|
|
.submitted = false,
|
|
|
|
.io_type = io_type,
|
|
|
|
.io_wbc = wbc,
|
2020-07-02 01:56:06 +00:00
|
|
|
.encrypted = fscrypt_inode_uses_fs_layer_crypto(cc->inode),
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
};
|
|
|
|
struct dnode_of_data dn;
|
|
|
|
struct node_info ni;
|
|
|
|
struct compress_io_ctx *cic;
|
|
|
|
pgoff_t start_idx = start_idx_of_cluster(cc);
|
|
|
|
unsigned int last_index = cc->cluster_size - 1;
|
|
|
|
loff_t psize;
|
|
|
|
int i, err;
|
|
|
|
|
2020-06-18 06:36:23 +00:00
|
|
|
if (IS_NOQUOTA(inode)) {
|
|
|
|
/*
|
|
|
|
* We need to wait for node_write to avoid block allocation during
|
|
|
|
* checkpoint. This can only happen to quota writes which can cause
|
|
|
|
* the below discard race condition.
|
|
|
|
*/
|
|
|
|
down_read(&sbi->node_write);
|
|
|
|
} else if (!f2fs_trylock_op(sbi)) {
|
2020-09-14 09:05:13 +00:00
|
|
|
goto out_free;
|
2020-06-18 06:36:23 +00:00
|
|
|
}
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
2020-02-24 11:20:16 +00:00
|
|
|
set_new_dnode(&dn, cc->inode, NULL, NULL, 0);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
err = f2fs_get_dnode_of_data(&dn, start_idx, LOOKUP_NODE);
|
|
|
|
if (err)
|
|
|
|
goto out_unlock_op;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
2020-02-14 09:44:10 +00:00
|
|
|
if (data_blkaddr(dn.inode, dn.node_page,
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
dn.ofs_in_node + i) == NULL_ADDR)
|
|
|
|
goto out_put_dnode;
|
|
|
|
}
|
|
|
|
|
|
|
|
psize = (loff_t)(cc->rpages[last_index]->index + 1) << PAGE_SHIFT;
|
|
|
|
|
|
|
|
err = f2fs_get_node_info(fio.sbi, dn.nid, &ni);
|
|
|
|
if (err)
|
|
|
|
goto out_put_dnode;
|
|
|
|
|
|
|
|
fio.version = ni.version;
|
|
|
|
|
2020-09-14 09:05:14 +00:00
|
|
|
cic = kmem_cache_zalloc(cic_entry_slab, GFP_NOFS);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cic)
|
|
|
|
goto out_put_dnode;
|
|
|
|
|
|
|
|
cic->magic = F2FS_COMPRESSED_PAGE_MAGIC;
|
|
|
|
cic->inode = inode;
|
2020-08-10 10:39:30 +00:00
|
|
|
atomic_set(&cic->pending_pages, cc->nr_cpages);
|
2020-09-14 09:05:13 +00:00
|
|
|
cic->rpages = page_array_alloc(cc->inode, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!cic->rpages)
|
|
|
|
goto out_put_cic;
|
|
|
|
|
|
|
|
cic->nr_rpages = cc->cluster_size;
|
|
|
|
|
|
|
|
for (i = 0; i < cc->nr_cpages; i++) {
|
|
|
|
f2fs_set_compressed_page(cc->cpages[i], inode,
|
2020-03-28 09:33:23 +00:00
|
|
|
cc->rpages[i + 1]->index, cic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
fio.compressed_page = cc->cpages[i];
|
2020-07-03 08:40:11 +00:00
|
|
|
|
|
|
|
fio.old_blkaddr = data_blkaddr(dn.inode, dn.node_page,
|
|
|
|
dn.ofs_in_node + i + 1);
|
|
|
|
|
|
|
|
/* wait for GCed page writeback via META_MAPPING */
|
|
|
|
f2fs_wait_on_block_writeback(inode, fio.old_blkaddr);
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (fio.encrypted) {
|
|
|
|
fio.page = cc->rpages[i + 1];
|
|
|
|
err = f2fs_encrypt_one_page(&fio);
|
|
|
|
if (err)
|
|
|
|
goto out_destroy_crypt;
|
|
|
|
cc->cpages[i] = fio.encrypted_page;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
set_cluster_writeback(cc);
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++)
|
|
|
|
cic->rpages[i] = cc->rpages[i];
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++, dn.ofs_in_node++) {
|
|
|
|
block_t blkaddr;
|
|
|
|
|
2020-02-14 09:44:10 +00:00
|
|
|
blkaddr = f2fs_data_blkaddr(&dn);
|
2020-02-28 10:08:46 +00:00
|
|
|
fio.page = cc->rpages[i];
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
fio.old_blkaddr = blkaddr;
|
|
|
|
|
|
|
|
/* cluster header */
|
|
|
|
if (i == 0) {
|
|
|
|
if (blkaddr == COMPRESS_ADDR)
|
|
|
|
fio.compr_blocks++;
|
|
|
|
if (__is_valid_data_blkaddr(blkaddr))
|
|
|
|
f2fs_invalidate_blocks(sbi, blkaddr);
|
|
|
|
f2fs_update_data_blkaddr(&dn, COMPRESS_ADDR);
|
|
|
|
goto unlock_continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (fio.compr_blocks && __is_valid_data_blkaddr(blkaddr))
|
|
|
|
fio.compr_blocks++;
|
|
|
|
|
|
|
|
if (i > cc->nr_cpages) {
|
|
|
|
if (__is_valid_data_blkaddr(blkaddr)) {
|
|
|
|
f2fs_invalidate_blocks(sbi, blkaddr);
|
|
|
|
f2fs_update_data_blkaddr(&dn, NEW_ADDR);
|
|
|
|
}
|
|
|
|
goto unlock_continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
f2fs_bug_on(fio.sbi, blkaddr == NULL_ADDR);
|
|
|
|
|
|
|
|
if (fio.encrypted)
|
|
|
|
fio.encrypted_page = cc->cpages[i - 1];
|
|
|
|
else
|
|
|
|
fio.compressed_page = cc->cpages[i - 1];
|
|
|
|
|
|
|
|
cc->cpages[i - 1] = NULL;
|
|
|
|
f2fs_outplace_write_data(&dn, &fio);
|
|
|
|
(*submitted)++;
|
|
|
|
unlock_continue:
|
|
|
|
inode_dec_dirty_pages(cc->inode);
|
|
|
|
unlock_page(fio.page);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (fio.compr_blocks)
|
|
|
|
f2fs_i_compr_blocks_update(inode, fio.compr_blocks - 1, false);
|
|
|
|
f2fs_i_compr_blocks_update(inode, cc->nr_cpages, true);
|
|
|
|
|
|
|
|
set_inode_flag(cc->inode, FI_APPEND_WRITE);
|
|
|
|
if (cc->cluster_idx == 0)
|
|
|
|
set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN);
|
|
|
|
|
|
|
|
f2fs_put_dnode(&dn);
|
2020-06-18 06:36:23 +00:00
|
|
|
if (IS_NOQUOTA(inode))
|
|
|
|
up_read(&sbi->node_write);
|
|
|
|
else
|
2020-04-09 17:25:21 +00:00
|
|
|
f2fs_unlock_op(sbi);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
2020-02-27 11:30:03 +00:00
|
|
|
spin_lock(&fi->i_size_lock);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (fi->last_disk_size < psize)
|
|
|
|
fi->last_disk_size = psize;
|
2020-02-27 11:30:03 +00:00
|
|
|
spin_unlock(&fi->i_size_lock);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
f2fs_put_rpages(cc);
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cc->inode, cc->cpages, cc->nr_cpages);
|
|
|
|
cc->cpages = NULL;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
f2fs_destroy_compress_ctx(cc);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_destroy_crypt:
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cc->inode, cic->rpages, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
for (--i; i >= 0; i--)
|
|
|
|
fscrypt_finalize_bounce_page(&cc->cpages[i]);
|
|
|
|
for (i = 0; i < cc->nr_cpages; i++) {
|
|
|
|
if (!cc->cpages[i])
|
|
|
|
continue;
|
|
|
|
f2fs_put_page(cc->cpages[i], 1);
|
|
|
|
}
|
|
|
|
out_put_cic:
|
2020-09-14 09:05:14 +00:00
|
|
|
kmem_cache_free(cic_entry_slab, cic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
out_put_dnode:
|
|
|
|
f2fs_put_dnode(&dn);
|
|
|
|
out_unlock_op:
|
2020-06-18 06:36:23 +00:00
|
|
|
if (IS_NOQUOTA(inode))
|
|
|
|
up_read(&sbi->node_write);
|
|
|
|
else
|
2020-04-09 17:25:21 +00:00
|
|
|
f2fs_unlock_op(sbi);
|
2020-09-14 09:05:13 +00:00
|
|
|
out_free:
|
|
|
|
page_array_free(cc->inode, cc->cpages, cc->nr_cpages);
|
|
|
|
cc->cpages = NULL;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_compress_write_end_io(struct bio *bio, struct page *page)
|
|
|
|
{
|
|
|
|
struct f2fs_sb_info *sbi = bio->bi_private;
|
|
|
|
struct compress_io_ctx *cic =
|
|
|
|
(struct compress_io_ctx *)page_private(page);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (unlikely(bio->bi_status))
|
|
|
|
mapping_set_error(cic->inode->i_mapping, -EIO);
|
|
|
|
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
f2fs_compress_free_page(page);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
|
|
|
|
dec_page_count(sbi, F2FS_WB_DATA);
|
|
|
|
|
2020-08-10 10:39:30 +00:00
|
|
|
if (atomic_dec_return(&cic->pending_pages))
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return;
|
|
|
|
|
|
|
|
for (i = 0; i < cic->nr_rpages; i++) {
|
|
|
|
WARN_ON(!cic->rpages[i]);
|
|
|
|
clear_cold_data(cic->rpages[i]);
|
|
|
|
end_page_writeback(cic->rpages[i]);
|
|
|
|
}
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(cic->inode, cic->rpages, cic->nr_rpages);
|
2020-09-14 09:05:14 +00:00
|
|
|
kmem_cache_free(cic_entry_slab, cic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int f2fs_write_raw_pages(struct compress_ctx *cc,
|
|
|
|
int *submitted,
|
|
|
|
struct writeback_control *wbc,
|
|
|
|
enum iostat_type io_type)
|
|
|
|
{
|
|
|
|
struct address_space *mapping = cc->inode->i_mapping;
|
|
|
|
int _submitted, compr_blocks, ret;
|
|
|
|
int i = -1, err = 0;
|
|
|
|
|
|
|
|
compr_blocks = f2fs_compressed_blocks(cc);
|
|
|
|
if (compr_blocks < 0) {
|
|
|
|
err = compr_blocks;
|
|
|
|
goto out_err;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < cc->cluster_size; i++) {
|
|
|
|
if (!cc->rpages[i])
|
|
|
|
continue;
|
|
|
|
retry_write:
|
|
|
|
if (cc->rpages[i]->mapping != mapping) {
|
|
|
|
unlock_page(cc->rpages[i]);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
BUG_ON(!PageLocked(cc->rpages[i]));
|
|
|
|
|
|
|
|
ret = f2fs_write_single_data_page(cc->rpages[i], &_submitted,
|
|
|
|
NULL, NULL, wbc, io_type,
|
|
|
|
compr_blocks);
|
|
|
|
if (ret) {
|
|
|
|
if (ret == AOP_WRITEPAGE_ACTIVATE) {
|
|
|
|
unlock_page(cc->rpages[i]);
|
|
|
|
ret = 0;
|
|
|
|
} else if (ret == -EAGAIN) {
|
2020-03-20 10:14:31 +00:00
|
|
|
/*
|
|
|
|
* for quota file, just redirty left pages to
|
|
|
|
* avoid deadlock caused by cluster update race
|
|
|
|
* from foreground operation.
|
|
|
|
*/
|
|
|
|
if (IS_NOQUOTA(cc->inode)) {
|
|
|
|
err = 0;
|
|
|
|
goto out_err;
|
|
|
|
}
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
ret = 0;
|
|
|
|
cond_resched();
|
2020-02-17 09:45:44 +00:00
|
|
|
congestion_wait(BLK_RW_ASYNC,
|
|
|
|
DEFAULT_IO_TIMEOUT);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
lock_page(cc->rpages[i]);
|
2020-06-19 09:14:19 +00:00
|
|
|
|
|
|
|
if (!PageDirty(cc->rpages[i])) {
|
|
|
|
unlock_page(cc->rpages[i]);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
clear_page_dirty_for_io(cc->rpages[i]);
|
|
|
|
goto retry_write;
|
|
|
|
}
|
|
|
|
err = ret;
|
2020-03-20 10:14:31 +00:00
|
|
|
goto out_err;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
*submitted += _submitted;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
out_err:
|
|
|
|
for (++i; i < cc->cluster_size; i++) {
|
|
|
|
if (!cc->rpages[i])
|
|
|
|
continue;
|
|
|
|
redirty_page_for_writepage(wbc, cc->rpages[i]);
|
|
|
|
unlock_page(cc->rpages[i]);
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
int f2fs_write_multi_pages(struct compress_ctx *cc,
|
|
|
|
int *submitted,
|
|
|
|
struct writeback_control *wbc,
|
|
|
|
enum iostat_type io_type)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
|
|
|
*submitted = 0;
|
|
|
|
if (cluster_may_compress(cc)) {
|
|
|
|
err = f2fs_compress_pages(cc);
|
|
|
|
if (err == -EAGAIN) {
|
|
|
|
goto write;
|
|
|
|
} else if (err) {
|
|
|
|
f2fs_put_rpages_wbc(cc, wbc, true, 1);
|
|
|
|
goto destroy_out;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = f2fs_write_compressed_pages(cc, submitted,
|
|
|
|
wbc, io_type);
|
|
|
|
if (!err)
|
|
|
|
return 0;
|
|
|
|
f2fs_bug_on(F2FS_I_SB(cc->inode), err != -EAGAIN);
|
|
|
|
}
|
|
|
|
write:
|
|
|
|
f2fs_bug_on(F2FS_I_SB(cc->inode), *submitted);
|
|
|
|
|
|
|
|
err = f2fs_write_raw_pages(cc, submitted, wbc, io_type);
|
|
|
|
f2fs_put_rpages_wbc(cc, wbc, false, 0);
|
|
|
|
destroy_out:
|
|
|
|
f2fs_destroy_compress_ctx(cc);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc)
|
|
|
|
{
|
|
|
|
struct decompress_io_ctx *dic;
|
|
|
|
pgoff_t start_idx = start_idx_of_cluster(cc);
|
|
|
|
int i;
|
|
|
|
|
2020-09-14 09:05:14 +00:00
|
|
|
dic = kmem_cache_zalloc(dic_entry_slab, GFP_NOFS);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!dic)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
dic->rpages = page_array_alloc(cc->inode, cc->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!dic->rpages) {
|
2020-09-14 09:05:14 +00:00
|
|
|
kmem_cache_free(dic_entry_slab, dic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
dic->magic = F2FS_COMPRESSED_PAGE_MAGIC;
|
|
|
|
dic->inode = cc->inode;
|
2020-08-10 10:39:30 +00:00
|
|
|
atomic_set(&dic->pending_pages, cc->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
dic->cluster_idx = cc->cluster_idx;
|
|
|
|
dic->cluster_size = cc->cluster_size;
|
|
|
|
dic->log_cluster_size = cc->log_cluster_size;
|
|
|
|
dic->nr_cpages = cc->nr_cpages;
|
|
|
|
dic->failed = false;
|
|
|
|
|
|
|
|
for (i = 0; i < dic->cluster_size; i++)
|
|
|
|
dic->rpages[i] = cc->rpages[i];
|
|
|
|
dic->nr_rpages = cc->cluster_size;
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
dic->cpages = page_array_alloc(dic->inode, dic->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!dic->cpages)
|
|
|
|
goto out_free;
|
|
|
|
|
|
|
|
for (i = 0; i < dic->nr_cpages; i++) {
|
|
|
|
struct page *page;
|
|
|
|
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
page = f2fs_compress_alloc_page();
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
if (!page)
|
|
|
|
goto out_free;
|
|
|
|
|
|
|
|
f2fs_set_compressed_page(page, cc->inode,
|
2020-03-28 09:33:23 +00:00
|
|
|
start_idx + i + 1, dic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
dic->cpages[i] = page;
|
|
|
|
}
|
|
|
|
|
|
|
|
return dic;
|
|
|
|
|
|
|
|
out_free:
|
|
|
|
f2fs_free_dic(dic);
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_free_dic(struct decompress_io_ctx *dic)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (dic->tpages) {
|
|
|
|
for (i = 0; i < dic->cluster_size; i++) {
|
|
|
|
if (dic->rpages[i])
|
|
|
|
continue;
|
2020-03-26 09:42:26 +00:00
|
|
|
if (!dic->tpages[i])
|
|
|
|
continue;
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
f2fs_compress_free_page(dic->tpages[i]);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(dic->inode, dic->tpages, dic->cluster_size);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (dic->cpages) {
|
|
|
|
for (i = 0; i < dic->nr_cpages; i++) {
|
|
|
|
if (!dic->cpages[i])
|
|
|
|
continue;
|
f2fs: introduce mempool for {,de}compress intermediate page allocation
If compression feature is on, in scenario of no enough free memory,
page refault ratio is higher than before, the root cause is:
- {,de}compression flow needs to allocate intermediate pages to store
compressed data in cluster, so during their allocation, vm may reclaim
mmaped pages.
- if above reclaimed pages belong to compressed cluster, during its
refault, it may cause more intermediate pages allocation, result in
reclaiming more mmaped pages.
So this patch introduces a mempool for intermediate page allocation,
in order to avoid high refault ratio, by default, number of
preallocated page in pool is 512, user can change the number by
assigning 'num_compress_pages' parameter during module initialization.
Ma Feng found warnings in the original patch and fixed like below.
Fix the following sparse warning:
fs/f2fs/compress.c:501:5: warning: symbol 'num_compress_pages' was not declared.
Should it be static?
fs/f2fs/compress.c:530:6: warning: symbol 'f2fs_compress_free_page' was not
declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Ma Feng <mafeng.ma@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2020-04-08 11:56:05 +00:00
|
|
|
f2fs_compress_free_page(dic->cpages[i]);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(dic->inode, dic->cpages, dic->nr_cpages);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
2020-09-14 09:05:13 +00:00
|
|
|
page_array_free(dic->inode, dic->rpages, dic->nr_rpages);
|
2020-09-14 09:05:14 +00:00
|
|
|
kmem_cache_free(dic_entry_slab, dic);
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_decompress_end_io(struct page **rpages,
|
|
|
|
unsigned int cluster_size, bool err, bool verity)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < cluster_size; i++) {
|
|
|
|
struct page *rpage = rpages[i];
|
|
|
|
|
|
|
|
if (!rpage)
|
|
|
|
continue;
|
|
|
|
|
2020-03-21 12:24:11 +00:00
|
|
|
if (err || PageError(rpage))
|
|
|
|
goto clear_uptodate;
|
|
|
|
|
|
|
|
if (!verity || fsverity_verify_page(rpage)) {
|
|
|
|
SetPageUptodate(rpage);
|
|
|
|
goto unlock;
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
}
|
2020-03-21 12:24:11 +00:00
|
|
|
clear_uptodate:
|
|
|
|
ClearPageUptodate(rpage);
|
|
|
|
ClearPageError(rpage);
|
|
|
|
unlock:
|
f2fs: support data compression
This patch tries to support compression in f2fs.
- New term named cluster is defined as basic unit of compression, file can
be divided into multiple clusters logically. One cluster includes 4 << n
(n >= 0) logical pages, compression size is also cluster size, each of
cluster can be compressed or not.
- In cluster metadata layout, one special flag is used to indicate cluster
is compressed one or normal one, for compressed cluster, following metadata
maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs stores
data including compress header and compressed data.
- In order to eliminate write amplification during overwrite, F2FS only
support compression on write-once file, data can be compressed only when
all logical blocks in file are valid and cluster compress ratio is lower
than specified threshold.
- To enable compression on regular inode, there are three ways:
* chattr +c file
* chattr +c dir; touch dir/file
* mount w/ -o compress_extension=ext; touch file.ext
Compress metadata layout:
[Dnode Structure]
+-----------------------------------------------+
| cluster 1 | cluster 2 | ......... | cluster N |
+-----------------------------------------------+
. . . .
. . . .
. Compressed Cluster . . Normal Cluster .
+----------+---------+---------+---------+ +---------+---------+---------+---------+
|compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 |
+----------+---------+---------+---------+ +---------+---------+---------+---------+
. .
. .
. .
+-------------+-------------+----------+----------------------------+
| data length | data chksum | reserved | compressed data |
+-------------+-------------+----------+----------------------------+
Changelog:
20190326:
- fix error handling of read_end_io().
- remove unneeded comments in f2fs_encrypt_one_page().
20190327:
- fix wrong use of f2fs_cluster_is_full() in f2fs_mpage_readpages().
- don't jump into loop directly to avoid uninitialized variables.
- add TODO tag in error path of f2fs_write_cache_pages().
20190328:
- fix wrong merge condition in f2fs_read_multi_pages().
- check compressed file in f2fs_post_read_required().
20190401
- allow overwrite on non-compressed cluster.
- check cluster meta before writing compressed data.
20190402
- don't preallocate blocks for compressed file.
- add lz4 compress algorithm
- process multiple post read works in one workqueue
Now f2fs supports processing post read work in multiple workqueue,
it shows low performance due to schedule overhead of multiple
workqueue executing orderly.
20190921
- compress: support buffered overwrite
C: compress cluster flag
V: valid block address
N: NEW_ADDR
One cluster contain 4 blocks
before overwrite after overwrite
- VVVV -> CVNN
- CVNN -> VVVV
- CVNN -> CVNN
- CVNN -> CVVV
- CVVV -> CVNN
- CVVV -> CVVV
20191029
- add kconfig F2FS_FS_COMPRESSION to isolate compression related
codes, add kconfig F2FS_FS_{LZO,LZ4} to cover backend algorithm.
note that: will remove lzo backend if Jaegeuk agreed that too.
- update codes according to Eric's comments.
20191101
- apply fixes from Jaegeuk
20191113
- apply fixes from Jaegeuk
- split workqueue for fsverity
20191216
- apply fixes from Jaegeuk
20200117
- fix to avoid NULL pointer dereference
[Jaegeuk Kim]
- add tracepoint for f2fs_{,de}compress_pages()
- fix many bugs and add some compression stats
- fix overwrite/mmap bugs
- address 32bit build error, reported by Geert.
- bug fixes when handling errors and i_compressed_blocks
Reported-by: <noreply@ellerman.id.au>
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2019-11-01 10:07:14 +00:00
|
|
|
unlock_page(rpage);
|
|
|
|
}
|
|
|
|
}
|
2020-09-14 09:05:13 +00:00
|
|
|
|
|
|
|
int f2fs_init_page_array_cache(struct f2fs_sb_info *sbi)
|
|
|
|
{
|
|
|
|
dev_t dev = sbi->sb->s_bdev->bd_dev;
|
|
|
|
char slab_name[32];
|
|
|
|
|
|
|
|
sprintf(slab_name, "f2fs_page_array_entry-%u:%u", MAJOR(dev), MINOR(dev));
|
|
|
|
|
|
|
|
sbi->page_array_slab_size = sizeof(struct page *) <<
|
|
|
|
F2FS_OPTION(sbi).compress_log_size;
|
|
|
|
|
|
|
|
sbi->page_array_slab = f2fs_kmem_cache_create(slab_name,
|
|
|
|
sbi->page_array_slab_size);
|
|
|
|
if (!sbi->page_array_slab)
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_destroy_page_array_cache(struct f2fs_sb_info *sbi)
|
|
|
|
{
|
|
|
|
kmem_cache_destroy(sbi->page_array_slab);
|
|
|
|
}
|
2020-09-14 09:05:14 +00:00
|
|
|
|
|
|
|
static int __init f2fs_init_cic_cache(void)
|
|
|
|
{
|
|
|
|
cic_entry_slab = f2fs_kmem_cache_create("f2fs_cic_entry",
|
|
|
|
sizeof(struct compress_io_ctx));
|
|
|
|
if (!cic_entry_slab)
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_destroy_cic_cache(void)
|
|
|
|
{
|
|
|
|
kmem_cache_destroy(cic_entry_slab);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __init f2fs_init_dic_cache(void)
|
|
|
|
{
|
|
|
|
dic_entry_slab = f2fs_kmem_cache_create("f2fs_dic_entry",
|
|
|
|
sizeof(struct decompress_io_ctx));
|
|
|
|
if (!dic_entry_slab)
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void f2fs_destroy_dic_cache(void)
|
|
|
|
{
|
|
|
|
kmem_cache_destroy(dic_entry_slab);
|
|
|
|
}
|
|
|
|
|
|
|
|
int __init f2fs_init_compress_cache(void)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
|
|
|
err = f2fs_init_cic_cache();
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
err = f2fs_init_dic_cache();
|
|
|
|
if (err)
|
|
|
|
goto free_cic;
|
|
|
|
return 0;
|
|
|
|
free_cic:
|
|
|
|
f2fs_destroy_cic_cache();
|
|
|
|
out:
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
void f2fs_destroy_compress_cache(void)
|
|
|
|
{
|
|
|
|
f2fs_destroy_dic_cache();
|
|
|
|
f2fs_destroy_cic_cache();
|
|
|
|
}
|