License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 22:07:57 +08:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* linux/fs/ext4/inode.c
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* Copyright (C) 1992, 1993, 1994, 1995
|
|
|
|
* Remy Card (card@masi.ibp.fr)
|
|
|
|
* Laboratoire MASI - Institut Blaise Pascal
|
|
|
|
* Universite Pierre et Marie Curie (Paris VI)
|
|
|
|
*
|
|
|
|
* from
|
|
|
|
*
|
|
|
|
* linux/fs/minix/inode.c
|
|
|
|
*
|
|
|
|
* Copyright (C) 1991, 1992 Linus Torvalds
|
|
|
|
*
|
|
|
|
* 64-bit file support on 64-bit platforms by Jakub Jelinek
|
|
|
|
* (jj@sunsite.ms.mff.cuni.cz)
|
|
|
|
*
|
2006-10-11 16:20:53 +08:00
|
|
|
* Assorted race fixes, rewrite of ext4_get_block() by Al Viro, 2000
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/fs.h>
|
2021-01-21 21:19:57 +08:00
|
|
|
#include <linux/mount.h>
|
2006-10-11 16:20:50 +08:00
|
|
|
#include <linux/time.h>
|
|
|
|
#include <linux/highuid.h>
|
|
|
|
#include <linux/pagemap.h>
|
2015-09-09 05:58:40 +08:00
|
|
|
#include <linux/dax.h>
|
2006-10-11 16:20:50 +08:00
|
|
|
#include <linux/quotaops.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/buffer_head.h>
|
|
|
|
#include <linux/writeback.h>
|
2008-07-12 07:27:31 +08:00
|
|
|
#include <linux/pagevec.h>
|
2006-10-11 16:20:50 +08:00
|
|
|
#include <linux/mpage.h>
|
2008-12-20 04:47:15 +08:00
|
|
|
#include <linux/namei.h>
|
2006-10-11 16:20:50 +08:00
|
|
|
#include <linux/uio.h>
|
|
|
|
#include <linux/bio.h>
|
2009-09-29 03:48:41 +08:00
|
|
|
#include <linux/workqueue.h>
|
2010-03-05 05:14:02 +08:00
|
|
|
#include <linux/kernel.h>
|
2011-01-13 08:59:13 +08:00
|
|
|
#include <linux/printk.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
|
|
|
#include <linux/slab.h>
|
2014-03-30 22:20:01 +08:00
|
|
|
#include <linux/bitops.h>
|
2016-11-21 06:36:06 +08:00
|
|
|
#include <linux/iomap.h>
|
2018-01-29 19:41:30 +08:00
|
|
|
#include <linux/iversion.h>
|
2009-06-17 23:48:11 +08:00
|
|
|
|
2008-04-30 06:13:32 +08:00
|
|
|
#include "ext4_jbd2.h"
|
2006-10-11 16:20:50 +08:00
|
|
|
#include "xattr.h"
|
|
|
|
#include "acl.h"
|
2011-06-28 07:16:04 +08:00
|
|
|
#include "truncate.h"
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2009-06-17 23:48:11 +08:00
|
|
|
#include <trace/events/ext4.h>
|
|
|
|
|
2012-04-30 06:31:10 +08:00
|
|
|
static __u32 ext4_inode_csum(struct inode *inode, struct ext4_inode *raw,
|
|
|
|
struct ext4_inode_info *ei)
|
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
__u32 csum;
|
2016-07-04 05:51:39 +08:00
|
|
|
__u16 dummy_csum = 0;
|
|
|
|
int offset = offsetof(struct ext4_inode, i_checksum_lo);
|
|
|
|
unsigned int csum_size = sizeof(dummy_csum);
|
2012-04-30 06:31:10 +08:00
|
|
|
|
2016-07-04 05:51:39 +08:00
|
|
|
csum = ext4_chksum(sbi, ei->i_csum_seed, (__u8 *)raw, offset);
|
|
|
|
csum = ext4_chksum(sbi, csum, (__u8 *)&dummy_csum, csum_size);
|
|
|
|
offset += csum_size;
|
|
|
|
csum = ext4_chksum(sbi, csum, (__u8 *)raw + offset,
|
|
|
|
EXT4_GOOD_OLD_INODE_SIZE - offset);
|
2012-04-30 06:31:10 +08:00
|
|
|
|
2016-07-04 05:51:39 +08:00
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
|
|
|
|
offset = offsetof(struct ext4_inode, i_checksum_hi);
|
|
|
|
csum = ext4_chksum(sbi, csum, (__u8 *)raw +
|
|
|
|
EXT4_GOOD_OLD_INODE_SIZE,
|
|
|
|
offset - EXT4_GOOD_OLD_INODE_SIZE);
|
|
|
|
if (EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi)) {
|
|
|
|
csum = ext4_chksum(sbi, csum, (__u8 *)&dummy_csum,
|
|
|
|
csum_size);
|
|
|
|
offset += csum_size;
|
|
|
|
}
|
2016-12-02 00:49:12 +08:00
|
|
|
csum = ext4_chksum(sbi, csum, (__u8 *)raw + offset,
|
|
|
|
EXT4_INODE_SIZE(inode->i_sb) - offset);
|
2012-04-30 06:31:10 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
return csum;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_inode_csum_verify(struct inode *inode, struct ext4_inode *raw,
|
|
|
|
struct ext4_inode_info *ei)
|
|
|
|
{
|
|
|
|
__u32 provided, calculated;
|
|
|
|
|
|
|
|
if (EXT4_SB(inode->i_sb)->s_es->s_creator_os !=
|
|
|
|
cpu_to_le32(EXT4_OS_LINUX) ||
|
2014-10-13 15:36:16 +08:00
|
|
|
!ext4_has_metadata_csum(inode->i_sb))
|
2012-04-30 06:31:10 +08:00
|
|
|
return 1;
|
|
|
|
|
|
|
|
provided = le16_to_cpu(raw->i_checksum_lo);
|
|
|
|
calculated = ext4_inode_csum(inode, raw, ei);
|
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE &&
|
|
|
|
EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi))
|
|
|
|
provided |= ((__u32)le16_to_cpu(raw->i_checksum_hi)) << 16;
|
|
|
|
else
|
|
|
|
calculated &= 0xFFFF;
|
|
|
|
|
|
|
|
return provided == calculated;
|
|
|
|
}
|
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw,
|
|
|
|
struct ext4_inode_info *ei)
|
2012-04-30 06:31:10 +08:00
|
|
|
{
|
|
|
|
__u32 csum;
|
|
|
|
|
|
|
|
if (EXT4_SB(inode->i_sb)->s_es->s_creator_os !=
|
|
|
|
cpu_to_le32(EXT4_OS_LINUX) ||
|
2014-10-13 15:36:16 +08:00
|
|
|
!ext4_has_metadata_csum(inode->i_sb))
|
2012-04-30 06:31:10 +08:00
|
|
|
return;
|
|
|
|
|
|
|
|
csum = ext4_inode_csum(inode, raw, ei);
|
|
|
|
raw->i_checksum_lo = cpu_to_le16(csum & 0xFFFF);
|
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE &&
|
|
|
|
EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi))
|
|
|
|
raw->i_checksum_hi = cpu_to_le16(csum >> 16);
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
static inline int ext4_begin_ordered_truncate(struct inode *inode,
|
|
|
|
loff_t new_size)
|
|
|
|
{
|
2010-11-09 02:51:33 +08:00
|
|
|
trace_ext4_begin_ordered_truncate(inode, new_size);
|
2011-01-11 01:29:43 +08:00
|
|
|
/*
|
|
|
|
* If jinode is zero, then we never opened the file for
|
|
|
|
* writing, so there's no need to call
|
|
|
|
* jbd2_journal_begin_ordered_truncate() since there's no
|
|
|
|
* outstanding writes we need to flush.
|
|
|
|
*/
|
|
|
|
if (!EXT4_I(inode)->jinode)
|
|
|
|
return 0;
|
|
|
|
return jbd2_journal_begin_ordered_truncate(EXT4_JOURNAL(inode),
|
|
|
|
EXT4_I(inode)->jinode,
|
|
|
|
new_size);
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
|
|
|
|
2010-10-28 09:30:09 +08:00
|
|
|
static int __ext4_journalled_writepage(struct page *page, unsigned int len);
|
2017-06-22 23:44:55 +08:00
|
|
|
static int ext4_meta_trans_blocks(struct inode *inode, int lblocks,
|
|
|
|
int pextents);
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* Test whether an inode is a fast symlink.
|
2017-07-04 12:11:21 +08:00
|
|
|
* A fast symlink has its symlink data stored in ext4_inode_info->i_data.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2015-04-16 13:55:00 +08:00
|
|
|
int ext4_inode_is_fast_symlink(struct inode *inode)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2017-12-04 09:38:01 +08:00
|
|
|
if (!(EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL)) {
|
|
|
|
int ea_blocks = EXT4_I(inode)->i_file_acl ?
|
|
|
|
EXT4_CLUSTER_SIZE(inode->i_sb) >> 9 : 0;
|
|
|
|
|
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0);
|
|
|
|
}
|
2017-07-04 12:11:21 +08:00
|
|
|
return S_ISLNK(inode->i_mode) && inode->i_size &&
|
|
|
|
(inode->i_size < EXT4_N_BLOCKS * 4);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Called at the last iput() if i_nlink is zero.
|
|
|
|
*/
|
2010-06-08 01:16:22 +08:00
|
|
|
void ext4_evict_inode(struct inode *inode)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
handle_t *handle;
|
2008-08-03 09:10:38 +08:00
|
|
|
int err;
|
2019-11-06 00:44:12 +08:00
|
|
|
/*
|
|
|
|
* Credits for final inode cleanup and freeing:
|
|
|
|
* sb + inode (ext4_orphan_del()), block bitmap, group descriptor
|
|
|
|
* (xattr block freeing), bitmap, group descriptor (inode freeing)
|
|
|
|
*/
|
|
|
|
int extra_credits = 6;
|
2017-06-22 22:26:31 +08:00
|
|
|
struct ext4_xattr_inode_array *ea_inode_array = NULL;
|
2020-11-27 19:06:49 +08:00
|
|
|
bool freeze_protected = false;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2010-11-09 02:51:33 +08:00
|
|
|
trace_ext4_evict_inode(inode);
|
2011-08-14 00:17:13 +08:00
|
|
|
|
2022-07-12 18:54:22 +08:00
|
|
|
if (EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL)
|
|
|
|
ext4_evict_ea_inode(inode);
|
2010-06-08 01:16:22 +08:00
|
|
|
if (inode->i_nlink) {
|
2011-07-26 21:07:11 +08:00
|
|
|
/*
|
|
|
|
* When journalling data dirty buffers are tracked only in the
|
|
|
|
* journal. So although mm thinks everything is clean and
|
|
|
|
* ready for reaping the inode might still have some pages to
|
|
|
|
* write in the running transaction or waiting to be
|
2022-02-10 04:21:43 +08:00
|
|
|
* checkpointed. Thus calling jbd2_journal_invalidate_folio()
|
2011-07-26 21:07:11 +08:00
|
|
|
* (via truncate_inode_pages()) to discard these buffers can
|
|
|
|
* cause data loss. Also even if we did not discard these
|
|
|
|
* buffers, we would have no way to find them after the inode
|
|
|
|
* is reaped and thus user could see stale data if he tries to
|
|
|
|
* read them before the transaction is checkpointed. So be
|
|
|
|
* careful and force everything to disk here... We use
|
|
|
|
* ei->i_datasync_tid to store the newest transaction
|
|
|
|
* containing inode's data.
|
|
|
|
*
|
|
|
|
* Note that directories do not have this problem because they
|
|
|
|
* don't use page cache.
|
|
|
|
*/
|
2016-07-04 23:03:00 +08:00
|
|
|
if (inode->i_ino != EXT4_JOURNAL_INO &&
|
|
|
|
ext4_should_journal_data(inode) &&
|
2022-04-24 22:09:36 +08:00
|
|
|
S_ISREG(inode->i_mode) && inode->i_data.nrpages) {
|
2011-07-26 21:07:11 +08:00
|
|
|
journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
|
|
|
|
tid_t commit_tid = EXT4_I(inode)->i_datasync_tid;
|
|
|
|
|
ext4/jbd2: don't wait (forever) for stale tid caused by wraparound
In the case where an inode has a very stale transaction id (tid) in
i_datasync_tid or i_sync_tid, it's possible that after a very large
(2**31) number of transactions, that the tid number space might wrap,
causing tid_geq()'s calculations to fail.
Commit deeeaf13 "jbd2: fix fsync() tid wraparound bug", later modified
by commit e7b04ac0 "jbd2: don't wake kjournald unnecessarily",
attempted to fix this problem, but it only avoided kjournald spinning
forever by fixing the logic in jbd2_log_start_commit().
Unfortunately, in the codepaths in fs/ext4/fsync.c and fs/ext4/inode.c
that might call jbd2_log_start_commit() with a stale tid, those
functions will subsequently call jbd2_log_wait_commit() with the same
stale tid, and then wait for a very long time. To fix this, we
replace the calls to jbd2_log_start_commit() and
jbd2_log_wait_commit() with a call to a new function,
jbd2_complete_transaction(), which will correctly handle stale tid's.
As a bonus, jbd2_complete_transaction() will avoid locking
j_state_lock for writing unless a commit needs to be started. This
should have a small (but probably not measurable) improvement for
ext4's scalability.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reported-by: Ben Hutchings <ben@decadent.org.uk>
Reported-by: George Barnett <gbarnett@atlassian.com>
Cc: stable@vger.kernel.org
2013-04-04 10:02:52 +08:00
|
|
|
jbd2_complete_transaction(journal, commit_tid);
|
2011-07-26 21:07:11 +08:00
|
|
|
filemap_write_and_wait(&inode->i_data);
|
|
|
|
}
|
2014-04-04 05:47:49 +08:00
|
|
|
truncate_inode_pages_final(&inode->i_data);
|
2013-06-05 02:46:12 +08:00
|
|
|
|
2010-06-08 01:16:22 +08:00
|
|
|
goto no_delete;
|
|
|
|
}
|
|
|
|
|
ext4: don't orphan or truncate the boot loader inode
The boot loader inode (inode #5) should never be visible in the
directory hierarchy, but it's possible if the file system is corrupted
that there will be a directory entry that points at inode #5. In
order to avoid accidentally trashing it, when such a directory inode
is opened, the inode will be marked as a bad inode, so that it's not
possible to modify (or read) the inode from userspace.
Unfortunately, when we unlink this (invalid/illegal) directory entry,
we will put the bad inode on the ophan list, and then when try to
unlink the directory, we don't actually remove the bad inode from the
orphan list before freeing in-memory inode structure. This means the
in-memory orphan list is corrupted, leading to a kernel oops.
In addition, avoid truncating a bad inode in ext4_destroy_inode(),
since truncating the boot loader inode is not a smart thing to do.
Reported-by: Sami Liedes <sami.liedes@iki.fi>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2014-10-06 10:47:07 +08:00
|
|
|
if (is_bad_inode(inode))
|
|
|
|
goto no_delete;
|
|
|
|
dquot_initialize(inode);
|
2010-03-03 22:05:06 +08:00
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
if (ext4_should_order_data(inode))
|
|
|
|
ext4_begin_ordered_truncate(inode, 0);
|
2014-04-04 05:47:49 +08:00
|
|
|
truncate_inode_pages_final(&inode->i_data);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-04-21 16:54:45 +08:00
|
|
|
/*
|
|
|
|
* For inodes with journalled data, transaction commit could have
|
|
|
|
* dirtied the inode. Flush worker is ignoring it because of I_FREEING
|
|
|
|
* flag but we still need to remove the inode from the writeback lists.
|
|
|
|
*/
|
|
|
|
if (!list_empty_careful(&inode->i_io_list)) {
|
|
|
|
WARN_ON_ONCE(!ext4_should_journal_data(inode));
|
|
|
|
inode_io_list_del(inode);
|
|
|
|
}
|
|
|
|
|
2012-06-12 22:20:38 +08:00
|
|
|
/*
|
|
|
|
* Protect us against freezing - iput() caller didn't have to have any
|
2020-11-27 19:06:49 +08:00
|
|
|
* protection against it. When we are in a running transaction though,
|
|
|
|
* we are already protected against freezing and we cannot grab further
|
|
|
|
* protection due to lock ordering constraints.
|
2012-06-12 22:20:38 +08:00
|
|
|
*/
|
2020-11-27 19:06:49 +08:00
|
|
|
if (!ext4_journal_current_handle()) {
|
|
|
|
sb_start_intwrite(inode->i_sb);
|
|
|
|
freeze_protected = true;
|
|
|
|
}
|
2017-06-22 09:10:32 +08:00
|
|
|
|
2017-06-22 23:42:09 +08:00
|
|
|
if (!IS_NOQUOTA(inode))
|
|
|
|
extra_credits += EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb);
|
|
|
|
|
2019-11-06 00:44:12 +08:00
|
|
|
/*
|
|
|
|
* Block bitmap, group descriptor, and inode are accounted in both
|
|
|
|
* ext4_blocks_for_truncate() and extra_credits. So subtract 3.
|
|
|
|
*/
|
2017-06-22 23:42:09 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE,
|
2019-11-06 00:44:12 +08:00
|
|
|
ext4_blocks_for_truncate(inode) + extra_credits - 3);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (IS_ERR(handle)) {
|
2008-08-03 09:10:38 +08:00
|
|
|
ext4_std_error(inode->i_sb, PTR_ERR(handle));
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* If we're going to skip the normal cleanup, we still need to
|
|
|
|
* make sure that the in-core orphan linked list is properly
|
|
|
|
* cleaned up.
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_orphan_del(NULL, inode);
|
2020-11-27 19:06:49 +08:00
|
|
|
if (freeze_protected)
|
|
|
|
sb_end_intwrite(inode->i_sb);
|
2006-10-11 16:20:50 +08:00
|
|
|
goto no_delete;
|
|
|
|
}
|
2017-06-22 23:42:09 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
if (IS_SYNC(inode))
|
2009-01-07 13:06:22 +08:00
|
|
|
ext4_handle_sync(handle);
|
2017-07-04 12:11:21 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Set inode->i_size to 0 before calling ext4_truncate(). We need
|
|
|
|
* special handling of symlinks here because i_size is used to
|
|
|
|
* determine whether ext4_inode_info->i_data contains symlink data or
|
|
|
|
* block mappings. Setting i_size to 0 will remove its fast symlink
|
|
|
|
* status. Erase i_data so that it becomes a valid empty block map.
|
|
|
|
*/
|
|
|
|
if (ext4_inode_is_fast_symlink(inode))
|
|
|
|
memset(EXT4_I(inode)->i_data, 0, sizeof(EXT4_I(inode)->i_data));
|
2006-10-11 16:20:50 +08:00
|
|
|
inode->i_size = 0;
|
2008-08-03 09:10:38 +08:00
|
|
|
err = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
if (err) {
|
2010-02-16 03:19:27 +08:00
|
|
|
ext4_warning(inode->i_sb,
|
2008-08-03 09:10:38 +08:00
|
|
|
"couldn't mark inode dirty (err %d)", err);
|
|
|
|
goto stop_handle;
|
|
|
|
}
|
2016-11-14 11:02:26 +08:00
|
|
|
if (inode->i_blocks) {
|
|
|
|
err = ext4_truncate(inode);
|
|
|
|
if (err) {
|
2020-03-29 07:33:43 +08:00
|
|
|
ext4_error_err(inode->i_sb, -err,
|
|
|
|
"couldn't truncate inode %lu (err %d)",
|
|
|
|
inode->i_ino, err);
|
2016-11-14 11:02:26 +08:00
|
|
|
goto stop_handle;
|
|
|
|
}
|
|
|
|
}
|
2008-08-03 09:10:38 +08:00
|
|
|
|
2017-06-22 23:42:09 +08:00
|
|
|
/* Remove xattr references. */
|
|
|
|
err = ext4_xattr_delete_inode(handle, inode, &ea_inode_array,
|
|
|
|
extra_credits);
|
|
|
|
if (err) {
|
|
|
|
ext4_warning(inode->i_sb, "xattr delete (err %d)", err);
|
|
|
|
stop_handle:
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
ext4_orphan_del(NULL, inode);
|
2020-11-27 19:06:49 +08:00
|
|
|
if (freeze_protected)
|
|
|
|
sb_end_intwrite(inode->i_sb);
|
2017-06-22 23:42:09 +08:00
|
|
|
ext4_xattr_inode_array_free(ea_inode_array);
|
|
|
|
goto no_delete;
|
2008-08-03 09:10:38 +08:00
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* Kill off the orphan record which ext4_truncate created.
|
2006-10-11 16:20:50 +08:00
|
|
|
* AKPM: I think this can be inside the above `if'.
|
2006-10-11 16:20:53 +08:00
|
|
|
* Note that ext4_orphan_del() has to be able to cope with the
|
2006-10-11 16:20:50 +08:00
|
|
|
* deletion of a non-existent orphan - this is because we don't
|
2006-10-11 16:20:53 +08:00
|
|
|
* know if ext4_truncate() actually created an orphan record.
|
2006-10-11 16:20:50 +08:00
|
|
|
* (Well, we could do this if we need to, but heck - it works)
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_orphan_del(handle, inode);
|
2018-07-30 03:50:00 +08:00
|
|
|
EXT4_I(inode)->i_dtime = (__u32)ktime_get_real_seconds();
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* One subtle ordering requirement: if anything has gone wrong
|
|
|
|
* (transaction abort, IO errors, whatever), then we can still
|
|
|
|
* do these next steps (the fs will already have been marked as
|
|
|
|
* having errors), but we can't free the inode if the mark_dirty
|
|
|
|
* fails.
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
if (ext4_mark_inode_dirty(handle, inode))
|
2006-10-11 16:20:50 +08:00
|
|
|
/* If that failed, just do the required in-core inode clear. */
|
2010-06-08 01:16:22 +08:00
|
|
|
ext4_clear_inode(inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
else
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_free_inode(handle, inode);
|
|
|
|
ext4_journal_stop(handle);
|
2020-11-27 19:06:49 +08:00
|
|
|
if (freeze_protected)
|
|
|
|
sb_end_intwrite(inode->i_sb);
|
2017-06-22 22:26:31 +08:00
|
|
|
ext4_xattr_inode_array_free(ea_inode_array);
|
2006-10-11 16:20:50 +08:00
|
|
|
return;
|
|
|
|
no_delete:
|
2020-11-06 11:58:51 +08:00
|
|
|
if (!list_empty(&EXT4_I(inode)->i_fc_list))
|
2022-01-17 17:36:54 +08:00
|
|
|
ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_NOMEM, NULL);
|
2010-06-08 01:16:22 +08:00
|
|
|
ext4_clear_inode(inode); /* We must guarantee clearing of inode... */
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2009-12-14 20:21:14 +08:00
|
|
|
#ifdef CONFIG_QUOTA
|
|
|
|
qsize_t *ext4_get_reserved_space(struct inode *inode)
|
2009-01-23 01:13:05 +08:00
|
|
|
{
|
2009-12-14 20:21:14 +08:00
|
|
|
return &EXT4_I(inode)->i_reserved_quota;
|
2009-01-23 01:13:05 +08:00
|
|
|
}
|
2009-12-14 20:21:14 +08:00
|
|
|
#endif
|
2010-01-01 15:41:30 +08:00
|
|
|
|
2009-12-31 03:20:45 +08:00
|
|
|
/*
|
|
|
|
* Called with i_data_sem down, which is important since we can call
|
|
|
|
* ext4_discard_preallocations() from here.
|
|
|
|
*/
|
2010-01-25 17:00:31 +08:00
|
|
|
void ext4_da_update_reserve_space(struct inode *inode,
|
|
|
|
int used, int quota_claim)
|
2008-07-18 04:12:08 +08:00
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2009-12-31 03:20:45 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
|
|
|
|
spin_lock(&ei->i_block_reservation_lock);
|
2011-09-10 07:18:51 +08:00
|
|
|
trace_ext4_da_update_reserve_space(inode, used, quota_claim);
|
2009-12-31 03:20:45 +08:00
|
|
|
if (unlikely(used > ei->i_reserved_data_blocks)) {
|
2013-02-15 04:11:41 +08:00
|
|
|
ext4_warning(inode->i_sb, "%s: ino %lu, used %d "
|
2012-03-20 11:13:43 +08:00
|
|
|
"with only %d reserved data blocks",
|
2009-12-31 03:20:45 +08:00
|
|
|
__func__, inode->i_ino, used,
|
|
|
|
ei->i_reserved_data_blocks);
|
|
|
|
WARN_ON(1);
|
|
|
|
used = ei->i_reserved_data_blocks;
|
|
|
|
}
|
2008-07-18 04:12:08 +08:00
|
|
|
|
2009-12-31 03:20:45 +08:00
|
|
|
/* Update per-inode reservations */
|
|
|
|
ei->i_reserved_data_blocks -= used;
|
2014-07-15 18:02:38 +08:00
|
|
|
percpu_counter_sub(&sbi->s_dirtyclusters_counter, used);
|
2008-10-10 21:39:00 +08:00
|
|
|
|
2021-05-26 13:29:30 +08:00
|
|
|
spin_unlock(&ei->i_block_reservation_lock);
|
2009-01-23 01:13:05 +08:00
|
|
|
|
2010-05-16 23:00:00 +08:00
|
|
|
/* Update quota subsystem for data blocks */
|
|
|
|
if (quota_claim)
|
2011-09-10 07:04:51 +08:00
|
|
|
dquot_claim_block(inode, EXT4_C2B(sbi, used));
|
2010-05-16 23:00:00 +08:00
|
|
|
else {
|
2010-01-25 17:00:31 +08:00
|
|
|
/*
|
|
|
|
* We did fallocate with an offset that is already delayed
|
|
|
|
* allocated. So on delayed allocated writeback we should
|
2010-05-16 23:00:00 +08:00
|
|
|
* not re-claim the quota for fallocated blocks.
|
2010-01-25 17:00:31 +08:00
|
|
|
*/
|
2011-09-10 07:04:51 +08:00
|
|
|
dquot_release_reservation_block(inode, EXT4_C2B(sbi, used));
|
2010-01-25 17:00:31 +08:00
|
|
|
}
|
2009-03-28 10:36:43 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If we have done all the pending block allocations and if
|
|
|
|
* there aren't any writers on the inode, we can discard the
|
|
|
|
* inode's preallocations.
|
|
|
|
*/
|
2009-12-31 03:20:45 +08:00
|
|
|
if ((ei->i_reserved_data_blocks == 0) &&
|
2019-02-11 12:04:16 +08:00
|
|
|
!inode_is_open_for_write(inode))
|
2020-08-17 15:36:15 +08:00
|
|
|
ext4_discard_preallocations(inode, 0);
|
2008-07-18 04:12:08 +08:00
|
|
|
}
|
|
|
|
|
2010-06-30 00:54:28 +08:00
|
|
|
static int __check_block_validity(struct inode *inode, const char *func,
|
2010-07-27 23:56:40 +08:00
|
|
|
unsigned int line,
|
|
|
|
struct ext4_map_blocks *map)
|
2009-05-18 03:38:01 +08:00
|
|
|
{
|
2019-04-10 11:37:08 +08:00
|
|
|
if (ext4_has_feature_journal(inode->i_sb) &&
|
|
|
|
(inode->i_ino ==
|
|
|
|
le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum)))
|
|
|
|
return 0;
|
2020-07-28 21:04:34 +08:00
|
|
|
if (!ext4_inode_block_valid(inode, map->m_pblk, map->m_len)) {
|
2010-07-27 23:56:40 +08:00
|
|
|
ext4_error_inode(inode, func, line, map->m_pblk,
|
2018-06-16 00:27:16 +08:00
|
|
|
"lblock %lu mapped to illegal pblock %llu "
|
2010-07-27 23:56:40 +08:00
|
|
|
"(length %d)", (unsigned long) map->m_lblk,
|
2018-06-16 00:27:16 +08:00
|
|
|
map->m_pblk, map->m_len);
|
2015-10-18 04:16:04 +08:00
|
|
|
return -EFSCORRUPTED;
|
2009-05-18 03:38:01 +08:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-12-08 04:09:35 +08:00
|
|
|
int ext4_issue_zeroout(struct inode *inode, ext4_lblk_t lblk, ext4_fsblk_t pblk,
|
|
|
|
ext4_lblk_t len)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
2019-12-27 00:10:22 +08:00
|
|
|
if (IS_ENCRYPTED(inode) && S_ISREG(inode->i_mode))
|
2016-07-11 02:01:03 +08:00
|
|
|
return fscrypt_zeroout_range(inode, lblk, pblk, len);
|
2015-12-08 04:09:35 +08:00
|
|
|
|
|
|
|
ret = sb_issue_zeroout(inode->i_sb, pblk, len, GFP_NOFS);
|
|
|
|
if (ret > 0)
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2010-06-30 00:54:28 +08:00
|
|
|
#define check_block_validity(inode, map) \
|
2010-07-27 23:56:40 +08:00
|
|
|
__check_block_validity((inode), __func__, __LINE__, (map))
|
2010-06-30 00:54:28 +08:00
|
|
|
|
2013-03-11 09:01:03 +08:00
|
|
|
#ifdef ES_AGGRESSIVE_TEST
|
|
|
|
static void ext4_map_blocks_es_recheck(handle_t *handle,
|
|
|
|
struct inode *inode,
|
|
|
|
struct ext4_map_blocks *es_map,
|
|
|
|
struct ext4_map_blocks *map,
|
|
|
|
int flags)
|
|
|
|
{
|
|
|
|
int retval;
|
|
|
|
|
|
|
|
map->m_flags = 0;
|
|
|
|
/*
|
|
|
|
* There is a race window that the result is not the same.
|
|
|
|
* e.g. xfstests #223 when dioread_nolock enables. The reason
|
|
|
|
* is that we lookup a block mapping in extent status tree with
|
|
|
|
* out taking i_data_sem. So at the time the unwritten extent
|
|
|
|
* could be converted.
|
|
|
|
*/
|
2015-12-08 04:04:57 +08:00
|
|
|
down_read(&EXT4_I(inode)->i_data_sem);
|
2013-03-11 09:01:03 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
|
2020-04-16 04:31:39 +08:00
|
|
|
retval = ext4_ext_map_blocks(handle, inode, map, 0);
|
2013-03-11 09:01:03 +08:00
|
|
|
} else {
|
2020-04-16 04:31:39 +08:00
|
|
|
retval = ext4_ind_map_blocks(handle, inode, map, 0);
|
2013-03-11 09:01:03 +08:00
|
|
|
}
|
2015-12-08 04:04:57 +08:00
|
|
|
up_read((&EXT4_I(inode)->i_data_sem));
|
2013-03-11 09:01:03 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* We don't check m_len because extent will be collpased in status
|
|
|
|
* tree. So the m_len might not equal.
|
|
|
|
*/
|
|
|
|
if (es_map->m_lblk != map->m_lblk ||
|
|
|
|
es_map->m_flags != map->m_flags ||
|
|
|
|
es_map->m_pblk != map->m_pblk) {
|
2013-07-13 12:40:31 +08:00
|
|
|
printk("ES cache assertion failed for inode: %lu "
|
2013-03-11 09:01:03 +08:00
|
|
|
"es_cached ex [%d/%d/%llu/%x] != "
|
|
|
|
"found ex [%d/%d/%llu/%x] retval %d flags %x\n",
|
|
|
|
inode->i_ino, es_map->m_lblk, es_map->m_len,
|
|
|
|
es_map->m_pblk, es_map->m_flags, map->m_lblk,
|
|
|
|
map->m_len, map->m_pblk, map->m_flags,
|
|
|
|
retval, flags);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif /* ES_AGGRESSIVE_TEST */
|
|
|
|
|
2008-02-26 04:29:55 +08:00
|
|
|
/*
|
2010-05-17 07:00:00 +08:00
|
|
|
* The ext4_map_blocks() function tries to look up the requested blocks,
|
2008-07-27 04:15:44 +08:00
|
|
|
* and returns if the blocks are already mapped.
|
2008-02-26 04:29:55 +08:00
|
|
|
*
|
|
|
|
* Otherwise it takes the write lock of the i_data_sem and allocate blocks
|
|
|
|
* and store the allocated blocks in the result buffer head and mark it
|
|
|
|
* mapped.
|
|
|
|
*
|
2010-05-17 07:00:00 +08:00
|
|
|
* If file type is extents based, it will call ext4_ext_map_blocks(),
|
|
|
|
* Otherwise, call with ext4_ind_map_blocks() to handle indirect mapping
|
2008-02-26 04:29:55 +08:00
|
|
|
* based files
|
|
|
|
*
|
2016-03-10 11:54:00 +08:00
|
|
|
* On success, it returns the number of blocks being mapped or allocated. if
|
|
|
|
* create==0 and the blocks are pre-allocated and unwritten, the resulting @map
|
|
|
|
* is marked as unwritten. If the create == 1, it will mark @map as mapped.
|
2008-02-26 04:29:55 +08:00
|
|
|
*
|
|
|
|
* It returns 0 if plain look up failed (blocks have not been allocated), in
|
2016-03-10 11:54:00 +08:00
|
|
|
* that case, @map is returned as unmapped but we still do fill map->m_len to
|
|
|
|
* indicate the length of a hole starting at map->m_lblk.
|
2008-02-26 04:29:55 +08:00
|
|
|
*
|
|
|
|
* It returns the error in case of allocation failure.
|
|
|
|
*/
|
2010-05-17 07:00:00 +08:00
|
|
|
int ext4_map_blocks(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_map_blocks *map, int flags)
|
2008-01-29 12:58:26 +08:00
|
|
|
{
|
2013-02-18 13:29:59 +08:00
|
|
|
struct extent_status es;
|
2008-01-29 12:58:26 +08:00
|
|
|
int retval;
|
2014-03-19 06:05:35 +08:00
|
|
|
int ret = 0;
|
2013-03-11 09:01:03 +08:00
|
|
|
#ifdef ES_AGGRESSIVE_TEST
|
|
|
|
struct ext4_map_blocks orig_map;
|
|
|
|
|
|
|
|
memcpy(&orig_map, map, sizeof(*map));
|
|
|
|
#endif
|
2008-02-26 04:29:55 +08:00
|
|
|
|
2010-05-17 07:00:00 +08:00
|
|
|
map->m_flags = 0;
|
2020-05-10 14:24:55 +08:00
|
|
|
ext_debug(inode, "flag 0x%x, max_blocks %u, logical block %lu\n",
|
|
|
|
flags, map->m_len, (unsigned long) map->m_lblk);
|
2013-02-18 13:29:59 +08:00
|
|
|
|
ext4: avoid possible overflow in ext4_map_blocks()
The ext4_map_blocks() function returns the number of blocks which
satisfying the caller's request. This number of blocks requested by
the caller is specified by an unsigned integer, but the return value
of ext4_map_blocks() is a signed integer (to accomodate error codes
per the kernel's standard error signalling convention).
Historically, overflows could never happen since mballoc() will refuse
to allocate more than 2048 blocks at a time (which is something we
should fix), and if the blocks were already allocated, the fact that
there would be some number of intervening metadata blocks pretty much
guaranteed that there could never be a contiguous region of data
blocks that was greater than 2**31 blocks.
However, this is now possible if there is a file system which is a bit
bigger than 8TB, and is created using the new mke2fs hugeblock
feature, which can create a perfectly contiguous file. In that case,
if a userspace program attempted to call fallocate() on this already
fully allocated file, it's possible that ext4_map_blocks() could
return a number large enough that it would overflow a signed integer,
resulting in a ext4 thinking that the ext4_map_blocks() call had
failed with some strange error code.
Since ext4_map_blocks() is always free to return a smaller number of
blocks than what was requested by the caller, fix this by capping the
number of blocks that ext4_map_blocks() will ever try to map to 2**31
- 1. In practice this should never get hit, except by someone
deliberately trying to provke the above-described bug.
Thanks to the PaX team for asking whethre this could possibly happen
in some off-line discussions about using some static code checking
technology they are developing to find bugs in kernel code.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2014-02-21 01:54:05 +08:00
|
|
|
/*
|
|
|
|
* ext4_map_blocks returns an int, and m_len is an unsigned int
|
|
|
|
*/
|
|
|
|
if (unlikely(map->m_len > INT_MAX))
|
|
|
|
map->m_len = INT_MAX;
|
|
|
|
|
2014-04-07 22:53:28 +08:00
|
|
|
/* We can handle the block number less than EXT_MAX_BLOCKS */
|
|
|
|
if (unlikely(map->m_lblk >= EXT_MAX_BLOCKS))
|
2015-10-18 04:16:04 +08:00
|
|
|
return -EFSCORRUPTED;
|
2014-04-07 22:53:28 +08:00
|
|
|
|
2013-02-18 13:29:59 +08:00
|
|
|
/* Lookup extent status tree firstly */
|
2020-10-16 04:37:59 +08:00
|
|
|
if (!(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY) &&
|
|
|
|
ext4_es_lookup_extent(inode, map->m_lblk, NULL, &es)) {
|
2013-02-18 13:29:59 +08:00
|
|
|
if (ext4_es_is_written(&es) || ext4_es_is_unwritten(&es)) {
|
|
|
|
map->m_pblk = ext4_es_pblock(&es) +
|
|
|
|
map->m_lblk - es.es_lblk;
|
|
|
|
map->m_flags |= ext4_es_is_written(&es) ?
|
|
|
|
EXT4_MAP_MAPPED : EXT4_MAP_UNWRITTEN;
|
|
|
|
retval = es.es_len - (map->m_lblk - es.es_lblk);
|
|
|
|
if (retval > map->m_len)
|
|
|
|
retval = map->m_len;
|
|
|
|
map->m_len = retval;
|
|
|
|
} else if (ext4_es_is_delayed(&es) || ext4_es_is_hole(&es)) {
|
2016-03-10 11:54:00 +08:00
|
|
|
map->m_pblk = 0;
|
|
|
|
retval = es.es_len - (map->m_lblk - es.es_lblk);
|
|
|
|
if (retval > map->m_len)
|
|
|
|
retval = map->m_len;
|
|
|
|
map->m_len = retval;
|
2013-02-18 13:29:59 +08:00
|
|
|
retval = 0;
|
|
|
|
} else {
|
2019-04-08 00:24:43 +08:00
|
|
|
BUG();
|
2013-02-18 13:29:59 +08:00
|
|
|
}
|
2022-04-24 22:09:35 +08:00
|
|
|
|
|
|
|
if (flags & EXT4_GET_BLOCKS_CACHED_NOWAIT)
|
|
|
|
return retval;
|
2013-03-11 09:01:03 +08:00
|
|
|
#ifdef ES_AGGRESSIVE_TEST
|
|
|
|
ext4_map_blocks_es_recheck(handle, inode, map,
|
|
|
|
&orig_map, flags);
|
|
|
|
#endif
|
2013-02-18 13:29:59 +08:00
|
|
|
goto found;
|
|
|
|
}
|
2022-04-24 22:09:35 +08:00
|
|
|
/*
|
|
|
|
* In the query cache no-wait mode, nothing we can do more if we
|
|
|
|
* cannot find extent in the cache.
|
|
|
|
*/
|
|
|
|
if (flags & EXT4_GET_BLOCKS_CACHED_NOWAIT)
|
|
|
|
return 0;
|
2013-02-18 13:29:59 +08:00
|
|
|
|
2008-01-29 12:58:29 +08:00
|
|
|
/*
|
2009-05-14 12:54:29 +08:00
|
|
|
* Try to see if we can get the block without requesting a new
|
|
|
|
* file system block.
|
2008-01-29 12:58:29 +08:00
|
|
|
*/
|
2015-12-08 04:04:57 +08:00
|
|
|
down_read(&EXT4_I(inode)->i_data_sem);
|
2010-05-17 10:00:00 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
|
2020-04-16 04:31:39 +08:00
|
|
|
retval = ext4_ext_map_blocks(handle, inode, map, 0);
|
2008-01-29 12:58:26 +08:00
|
|
|
} else {
|
2020-04-16 04:31:39 +08:00
|
|
|
retval = ext4_ind_map_blocks(handle, inode, map, 0);
|
2008-01-29 12:58:26 +08:00
|
|
|
}
|
2013-02-18 13:28:47 +08:00
|
|
|
if (retval > 0) {
|
2013-08-17 09:22:41 +08:00
|
|
|
unsigned int status;
|
2013-02-18 13:28:47 +08:00
|
|
|
|
2013-07-30 00:51:42 +08:00
|
|
|
if (unlikely(retval != map->m_len)) {
|
|
|
|
ext4_warning(inode->i_sb,
|
|
|
|
"ES len assertion failed for inode "
|
|
|
|
"%lu: retval %d != map->m_len %d",
|
|
|
|
inode->i_ino, retval, map->m_len);
|
|
|
|
WARN_ON(1);
|
2013-03-11 09:01:03 +08:00
|
|
|
}
|
|
|
|
|
2013-02-18 13:28:47 +08:00
|
|
|
status = map->m_flags & EXT4_MAP_UNWRITTEN ?
|
|
|
|
EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN;
|
|
|
|
if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) &&
|
2015-05-03 09:36:55 +08:00
|
|
|
!(status & EXTENT_STATUS_WRITTEN) &&
|
2018-10-02 02:10:39 +08:00
|
|
|
ext4_es_scan_range(inode, &ext4_es_is_delayed, map->m_lblk,
|
|
|
|
map->m_lblk + map->m_len - 1))
|
2013-02-18 13:28:47 +08:00
|
|
|
status |= EXTENT_STATUS_DELAYED;
|
|
|
|
ret = ext4_es_insert_extent(inode, map->m_lblk,
|
|
|
|
map->m_len, map->m_pblk, status);
|
|
|
|
if (ret < 0)
|
|
|
|
retval = ret;
|
|
|
|
}
|
2015-12-08 04:04:57 +08:00
|
|
|
up_read((&EXT4_I(inode)->i_data_sem));
|
2008-02-26 04:29:55 +08:00
|
|
|
|
2013-02-18 13:29:59 +08:00
|
|
|
found:
|
2010-05-17 07:00:00 +08:00
|
|
|
if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) {
|
2014-03-19 06:05:35 +08:00
|
|
|
ret = check_block_validity(inode, map);
|
2009-05-18 03:38:01 +08:00
|
|
|
if (ret != 0)
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-02-26 04:29:55 +08:00
|
|
|
/* If it is only a block(s) look up */
|
2009-05-14 12:58:52 +08:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_CREATE) == 0)
|
2008-02-26 04:29:55 +08:00
|
|
|
return retval;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Returns if the blocks have already allocated
|
|
|
|
*
|
|
|
|
* Note that if blocks have been preallocated
|
2011-10-09 03:53:49 +08:00
|
|
|
* ext4_ext_get_block() returns the create = 0
|
2008-02-26 04:29:55 +08:00
|
|
|
* with buffer head unmapped.
|
|
|
|
*/
|
2010-05-17 07:00:00 +08:00
|
|
|
if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED)
|
2014-03-19 06:05:35 +08:00
|
|
|
/*
|
|
|
|
* If we need to convert extent to unwritten
|
|
|
|
* we continue and do the actual work in
|
|
|
|
* ext4_ext_map_blocks()
|
|
|
|
*/
|
|
|
|
if (!(flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN))
|
|
|
|
return retval;
|
2008-01-29 12:58:29 +08:00
|
|
|
|
2009-05-15 05:05:39 +08:00
|
|
|
/*
|
2013-02-18 13:28:04 +08:00
|
|
|
* Here we clear m_flags because after allocating an new extent,
|
|
|
|
* it will be set again.
|
2009-05-15 05:05:39 +08:00
|
|
|
*/
|
2013-02-18 13:28:04 +08:00
|
|
|
map->m_flags &= ~EXT4_MAP_FLAGS;
|
2009-05-15 05:05:39 +08:00
|
|
|
|
2008-01-29 12:58:29 +08:00
|
|
|
/*
|
2014-04-21 11:45:47 +08:00
|
|
|
* New blocks allocate and/or writing to unwritten extent
|
2008-02-26 04:29:55 +08:00
|
|
|
* will possibly result in updating i_data, so we take
|
2014-09-02 10:15:30 +08:00
|
|
|
* the write lock of i_data_sem, and call get_block()
|
2008-02-26 04:29:55 +08:00
|
|
|
* with create == 1 flag.
|
2008-01-29 12:58:29 +08:00
|
|
|
*/
|
2014-05-13 00:55:07 +08:00
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
2008-07-15 05:52:37 +08:00
|
|
|
|
2008-01-29 12:58:29 +08:00
|
|
|
/*
|
|
|
|
* We need to check for EXT4 here because migrate
|
|
|
|
* could have changed the inode type in between
|
|
|
|
*/
|
2010-05-17 10:00:00 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
|
2010-05-17 07:00:00 +08:00
|
|
|
retval = ext4_ext_map_blocks(handle, inode, map, flags);
|
2008-01-29 12:58:26 +08:00
|
|
|
} else {
|
2010-05-17 07:00:00 +08:00
|
|
|
retval = ext4_ind_map_blocks(handle, inode, map, flags);
|
2008-04-29 20:11:12 +08:00
|
|
|
|
2010-05-17 07:00:00 +08:00
|
|
|
if (retval > 0 && map->m_flags & EXT4_MAP_NEW) {
|
2008-04-29 20:11:12 +08:00
|
|
|
/*
|
|
|
|
* We allocated new blocks which will result in
|
|
|
|
* i_data's format changing. Force the migrate
|
|
|
|
* to fail by clearing migrate flags
|
|
|
|
*/
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
|
2008-04-29 20:11:12 +08:00
|
|
|
}
|
2008-07-15 05:52:37 +08:00
|
|
|
|
2010-01-25 17:00:31 +08:00
|
|
|
/*
|
|
|
|
* Update reserved blocks/metadata blocks after successful
|
|
|
|
* block allocation which had been deferred till now. We don't
|
|
|
|
* support fallocate for non extent files. So we can update
|
|
|
|
* reserve space here.
|
|
|
|
*/
|
|
|
|
if ((retval > 0) &&
|
2010-01-15 14:27:59 +08:00
|
|
|
(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE))
|
2010-01-25 17:00:31 +08:00
|
|
|
ext4_da_update_reserve_space(inode, retval, 1);
|
|
|
|
}
|
2009-05-15 01:57:08 +08:00
|
|
|
|
2013-02-18 13:28:47 +08:00
|
|
|
if (retval > 0) {
|
2013-08-17 09:22:41 +08:00
|
|
|
unsigned int status;
|
2013-02-18 13:28:47 +08:00
|
|
|
|
2013-07-30 00:51:42 +08:00
|
|
|
if (unlikely(retval != map->m_len)) {
|
|
|
|
ext4_warning(inode->i_sb,
|
|
|
|
"ES len assertion failed for inode "
|
|
|
|
"%lu: retval %d != map->m_len %d",
|
|
|
|
inode->i_ino, retval, map->m_len);
|
|
|
|
WARN_ON(1);
|
2013-03-11 09:01:03 +08:00
|
|
|
}
|
|
|
|
|
2015-12-08 04:10:26 +08:00
|
|
|
/*
|
|
|
|
* We have to zeroout blocks before inserting them into extent
|
|
|
|
* status tree. Otherwise someone could look them up there and
|
2016-09-30 14:02:29 +08:00
|
|
|
* use them before they are really zeroed. We also have to
|
|
|
|
* unmap metadata before zeroing as otherwise writeback can
|
|
|
|
* overwrite zeros with stale data from block device.
|
2015-12-08 04:10:26 +08:00
|
|
|
*/
|
|
|
|
if (flags & EXT4_GET_BLOCKS_ZERO &&
|
|
|
|
map->m_flags & EXT4_MAP_MAPPED &&
|
|
|
|
map->m_flags & EXT4_MAP_NEW) {
|
|
|
|
ret = ext4_issue_zeroout(inode, map->m_lblk,
|
|
|
|
map->m_pblk, map->m_len);
|
|
|
|
if (ret) {
|
|
|
|
retval = ret;
|
|
|
|
goto out_sem;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-03-11 09:13:05 +08:00
|
|
|
/*
|
|
|
|
* If the extent has been zeroed out, we don't need to update
|
|
|
|
* extent status tree.
|
|
|
|
*/
|
|
|
|
if ((flags & EXT4_GET_BLOCKS_PRE_IO) &&
|
2019-08-12 04:32:41 +08:00
|
|
|
ext4_es_lookup_extent(inode, map->m_lblk, NULL, &es)) {
|
2013-03-11 09:13:05 +08:00
|
|
|
if (ext4_es_is_written(&es))
|
2015-12-08 04:10:26 +08:00
|
|
|
goto out_sem;
|
2013-03-11 09:13:05 +08:00
|
|
|
}
|
2013-02-18 13:28:47 +08:00
|
|
|
status = map->m_flags & EXT4_MAP_UNWRITTEN ?
|
|
|
|
EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN;
|
|
|
|
if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) &&
|
2015-05-03 09:36:55 +08:00
|
|
|
!(status & EXTENT_STATUS_WRITTEN) &&
|
2018-10-02 02:10:39 +08:00
|
|
|
ext4_es_scan_range(inode, &ext4_es_is_delayed, map->m_lblk,
|
|
|
|
map->m_lblk + map->m_len - 1))
|
2013-02-18 13:28:47 +08:00
|
|
|
status |= EXTENT_STATUS_DELAYED;
|
|
|
|
ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len,
|
|
|
|
map->m_pblk, status);
|
2015-12-08 04:10:26 +08:00
|
|
|
if (ret < 0) {
|
2013-02-18 13:28:47 +08:00
|
|
|
retval = ret;
|
2015-12-08 04:10:26 +08:00
|
|
|
goto out_sem;
|
|
|
|
}
|
2011-09-10 07:20:51 +08:00
|
|
|
}
|
|
|
|
|
2015-12-08 04:10:26 +08:00
|
|
|
out_sem:
|
2008-01-29 12:58:29 +08:00
|
|
|
up_write((&EXT4_I(inode)->i_data_sem));
|
2010-05-17 07:00:00 +08:00
|
|
|
if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) {
|
2014-03-19 06:05:35 +08:00
|
|
|
ret = check_block_validity(inode, map);
|
2009-05-18 03:38:01 +08:00
|
|
|
if (ret != 0)
|
|
|
|
return ret;
|
2016-04-24 12:56:03 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Inodes with freshly allocated blocks where contents will be
|
|
|
|
* visible after transaction commit must be on transaction's
|
|
|
|
* ordered data list.
|
|
|
|
*/
|
|
|
|
if (map->m_flags & EXT4_MAP_NEW &&
|
|
|
|
!(map->m_flags & EXT4_MAP_UNWRITTEN) &&
|
|
|
|
!(flags & EXT4_GET_BLOCKS_ZERO) &&
|
2017-06-22 23:31:25 +08:00
|
|
|
!ext4_is_quota_file(inode) &&
|
2016-04-24 12:56:03 +08:00
|
|
|
ext4_should_order_data(inode)) {
|
2019-06-21 05:26:26 +08:00
|
|
|
loff_t start_byte =
|
|
|
|
(loff_t)map->m_lblk << inode->i_blkbits;
|
|
|
|
loff_t length = (loff_t)map->m_len << inode->i_blkbits;
|
|
|
|
|
2016-04-24 12:56:08 +08:00
|
|
|
if (flags & EXT4_GET_BLOCKS_IO_SUBMIT)
|
2019-06-21 05:26:26 +08:00
|
|
|
ret = ext4_jbd2_inode_add_wait(handle, inode,
|
|
|
|
start_byte, length);
|
2016-04-24 12:56:08 +08:00
|
|
|
else
|
2019-06-21 05:26:26 +08:00
|
|
|
ret = ext4_jbd2_inode_add_write(handle, inode,
|
|
|
|
start_byte, length);
|
2016-04-24 12:56:03 +08:00
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
}
|
2009-05-18 03:38:01 +08:00
|
|
|
}
|
2021-12-21 10:28:39 +08:00
|
|
|
if (retval > 0 && (map->m_flags & EXT4_MAP_UNWRITTEN ||
|
|
|
|
map->m_flags & EXT4_MAP_MAPPED))
|
|
|
|
ext4_fc_track_range(handle, inode, map->m_lblk,
|
|
|
|
map->m_lblk + map->m_len - 1);
|
2020-05-10 14:24:52 +08:00
|
|
|
if (retval < 0)
|
2020-05-10 14:24:55 +08:00
|
|
|
ext_debug(inode, "failed with err %d\n", retval);
|
2008-01-29 12:58:26 +08:00
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
|
2016-02-19 13:18:25 +08:00
|
|
|
/*
|
|
|
|
* Update EXT4_MAP_FLAGS in bh->b_state. For buffer heads attached to pages
|
|
|
|
* we have to be careful as someone else may be manipulating b_state as well.
|
|
|
|
*/
|
|
|
|
static void ext4_update_bh_state(struct buffer_head *bh, unsigned long flags)
|
|
|
|
{
|
|
|
|
unsigned long old_state;
|
|
|
|
unsigned long new_state;
|
|
|
|
|
|
|
|
flags &= EXT4_MAP_FLAGS;
|
|
|
|
|
|
|
|
/* Dummy buffer_head? Set non-atomically. */
|
|
|
|
if (!bh->b_page) {
|
|
|
|
bh->b_state = (bh->b_state & ~EXT4_MAP_FLAGS) | flags;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Someone else may be modifying b_state. Be careful! This is ugly but
|
|
|
|
* once we get rid of using bh as a container for mapping information
|
|
|
|
* to pass to / from get_block functions, this can go away.
|
|
|
|
*/
|
|
|
|
do {
|
|
|
|
old_state = READ_ONCE(bh->b_state);
|
|
|
|
new_state = (old_state & ~EXT4_MAP_FLAGS) | flags;
|
|
|
|
} while (unlikely(
|
|
|
|
cmpxchg(&bh->b_state, old_state, new_state) != old_state));
|
|
|
|
}
|
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
static int _ext4_get_block(struct inode *inode, sector_t iblock,
|
|
|
|
struct buffer_head *bh, int flags)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2010-05-17 08:00:00 +08:00
|
|
|
struct ext4_map_blocks map;
|
2016-03-09 12:35:46 +08:00
|
|
|
int ret = 0;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2012-12-11 03:04:52 +08:00
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
return -ERANGE;
|
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
map.m_lblk = iblock;
|
|
|
|
map.m_len = bh->b_size >> inode->i_blkbits;
|
|
|
|
|
2016-03-09 12:35:46 +08:00
|
|
|
ret = ext4_map_blocks(ext4_journal_current_handle(), inode, &map,
|
|
|
|
flags);
|
2008-02-10 14:08:38 +08:00
|
|
|
if (ret > 0) {
|
2010-05-17 08:00:00 +08:00
|
|
|
map_bh(bh, inode->i_sb, map.m_pblk);
|
2016-02-19 13:18:25 +08:00
|
|
|
ext4_update_bh_state(bh, map.m_flags);
|
2010-05-17 08:00:00 +08:00
|
|
|
bh->b_size = inode->i_sb->s_blocksize * map.m_len;
|
2008-02-10 14:08:38 +08:00
|
|
|
ret = 0;
|
2016-11-08 08:30:58 +08:00
|
|
|
} else if (ret == 0) {
|
|
|
|
/* hole case, need to fill in bh->b_size */
|
|
|
|
bh->b_size = inode->i_sb->s_blocksize * map.m_len;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
int ext4_get_block(struct inode *inode, sector_t iblock,
|
|
|
|
struct buffer_head *bh, int create)
|
|
|
|
{
|
|
|
|
return _ext4_get_block(inode, iblock, bh,
|
|
|
|
create ? EXT4_GET_BLOCKS_CREATE : 0);
|
|
|
|
}
|
|
|
|
|
2016-03-09 12:08:10 +08:00
|
|
|
/*
|
|
|
|
* Get block function used when preparing for buffered write if we require
|
|
|
|
* creating an unwritten extent if blocks haven't been allocated. The extent
|
|
|
|
* will be converted to written after the IO is complete.
|
|
|
|
*/
|
|
|
|
int ext4_get_block_unwritten(struct inode *inode, sector_t iblock,
|
|
|
|
struct buffer_head *bh_result, int create)
|
|
|
|
{
|
|
|
|
ext4_debug("ext4_get_block_unwritten: inode %lu, create flag %d\n",
|
|
|
|
inode->i_ino, create);
|
|
|
|
return _ext4_get_block(inode, iblock, bh_result,
|
2022-05-20 19:14:02 +08:00
|
|
|
EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT);
|
2016-03-09 12:08:10 +08:00
|
|
|
}
|
|
|
|
|
2016-03-09 12:35:46 +08:00
|
|
|
/* Maximum number of blocks we map for direct IO at once. */
|
|
|
|
#define DIO_MAX_BLOCKS 4096
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* `handle' can be NULL if create is zero
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode,
|
2015-06-21 13:25:29 +08:00
|
|
|
ext4_lblk_t block, int map_flags)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2010-05-17 08:00:00 +08:00
|
|
|
struct ext4_map_blocks map;
|
|
|
|
struct buffer_head *bh;
|
2015-06-21 13:25:29 +08:00
|
|
|
int create = map_flags & EXT4_GET_BLOCKS_CREATE;
|
2022-04-24 22:09:35 +08:00
|
|
|
bool nowait = map_flags & EXT4_GET_BLOCKS_CACHED_NOWAIT;
|
2014-08-30 08:51:32 +08:00
|
|
|
int err;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-11-07 23:58:11 +08:00
|
|
|
ASSERT((EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
|
|
|
|
|| handle != NULL || create == 0);
|
2022-04-24 22:09:35 +08:00
|
|
|
ASSERT(create == 0 || !nowait);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
map.m_lblk = block;
|
|
|
|
map.m_len = 1;
|
2015-06-21 13:25:29 +08:00
|
|
|
err = ext4_map_blocks(handle, inode, &map, map_flags);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2014-08-30 08:51:32 +08:00
|
|
|
if (err == 0)
|
|
|
|
return create ? ERR_PTR(-ENOSPC) : NULL;
|
2010-05-17 08:00:00 +08:00
|
|
|
if (err < 0)
|
2014-08-30 08:51:32 +08:00
|
|
|
return ERR_PTR(err);
|
2010-05-17 08:00:00 +08:00
|
|
|
|
2022-04-24 22:09:35 +08:00
|
|
|
if (nowait)
|
|
|
|
return sb_find_get_block(inode->i_sb, map.m_pblk);
|
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
bh = sb_getblk(inode->i_sb, map.m_pblk);
|
2014-08-30 08:51:32 +08:00
|
|
|
if (unlikely(!bh))
|
|
|
|
return ERR_PTR(-ENOMEM);
|
2010-05-17 08:00:00 +08:00
|
|
|
if (map.m_flags & EXT4_MAP_NEW) {
|
2020-11-07 23:58:11 +08:00
|
|
|
ASSERT(create != 0);
|
|
|
|
ASSERT((EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
|
|
|
|
|| (handle != NULL));
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
/*
|
|
|
|
* Now that we do not always journal data, we should
|
|
|
|
* keep in mind whether this should always journal the
|
|
|
|
* new buffer as metadata. For now, regular file
|
|
|
|
* writes use ext4_get_block instead, so it's not a
|
|
|
|
* problem.
|
|
|
|
*/
|
|
|
|
lock_buffer(bh);
|
|
|
|
BUFFER_TRACE(bh, "call get_create_access");
|
2021-08-16 17:57:04 +08:00
|
|
|
err = ext4_journal_get_create_access(handle, inode->i_sb, bh,
|
|
|
|
EXT4_JTR_NONE);
|
2014-08-30 08:51:32 +08:00
|
|
|
if (unlikely(err)) {
|
|
|
|
unlock_buffer(bh);
|
|
|
|
goto errout;
|
|
|
|
}
|
|
|
|
if (!buffer_uptodate(bh)) {
|
2010-05-17 08:00:00 +08:00
|
|
|
memset(bh->b_data, 0, inode->i_sb->s_blocksize);
|
|
|
|
set_buffer_uptodate(bh);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2010-05-17 08:00:00 +08:00
|
|
|
unlock_buffer(bh);
|
|
|
|
BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
|
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, bh);
|
2014-08-30 08:51:32 +08:00
|
|
|
if (unlikely(err))
|
|
|
|
goto errout;
|
|
|
|
} else
|
2010-05-17 08:00:00 +08:00
|
|
|
BUFFER_TRACE(bh, "not a new buffer");
|
|
|
|
return bh;
|
2014-08-30 08:51:32 +08:00
|
|
|
errout:
|
|
|
|
brelse(bh);
|
|
|
|
return ERR_PTR(err);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode,
|
2015-06-21 13:25:29 +08:00
|
|
|
ext4_lblk_t block, int map_flags)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2008-09-09 10:25:24 +08:00
|
|
|
struct buffer_head *bh;
|
2020-09-24 15:33:33 +08:00
|
|
|
int ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2015-06-21 13:25:29 +08:00
|
|
|
bh = ext4_getblk(handle, inode, block, map_flags);
|
2014-08-30 08:52:15 +08:00
|
|
|
if (IS_ERR(bh))
|
2006-10-11 16:20:50 +08:00
|
|
|
return bh;
|
2019-08-23 11:00:32 +08:00
|
|
|
if (!bh || ext4_buffer_uptodate(bh))
|
2006-10-11 16:20:50 +08:00
|
|
|
return bh;
|
2020-09-24 15:33:33 +08:00
|
|
|
|
|
|
|
ret = ext4_read_bh_lock(bh, REQ_META | REQ_PRIO, true);
|
|
|
|
if (ret) {
|
|
|
|
put_bh(bh);
|
|
|
|
return ERR_PTR(ret);
|
|
|
|
}
|
|
|
|
return bh;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2017-08-06 12:07:01 +08:00
|
|
|
/* Read a contiguous batch of blocks. */
|
|
|
|
int ext4_bread_batch(struct inode *inode, ext4_lblk_t block, int bh_count,
|
|
|
|
bool wait, struct buffer_head **bhs)
|
|
|
|
{
|
|
|
|
int i, err;
|
|
|
|
|
|
|
|
for (i = 0; i < bh_count; i++) {
|
|
|
|
bhs[i] = ext4_getblk(NULL, inode, block + i, 0 /* map_flags */);
|
|
|
|
if (IS_ERR(bhs[i])) {
|
|
|
|
err = PTR_ERR(bhs[i]);
|
|
|
|
bh_count = i;
|
|
|
|
goto out_brelse;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < bh_count; i++)
|
|
|
|
/* Note that NULL bhs[i] is valid because of holes. */
|
2020-09-24 15:33:33 +08:00
|
|
|
if (bhs[i] && !ext4_buffer_uptodate(bhs[i]))
|
|
|
|
ext4_read_bh_lock(bhs[i], REQ_META | REQ_PRIO, false);
|
2017-08-06 12:07:01 +08:00
|
|
|
|
|
|
|
if (!wait)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
for (i = 0; i < bh_count; i++)
|
|
|
|
if (bhs[i])
|
|
|
|
wait_on_buffer(bhs[i]);
|
|
|
|
|
|
|
|
for (i = 0; i < bh_count; i++) {
|
|
|
|
if (bhs[i] && !buffer_uptodate(bhs[i])) {
|
|
|
|
err = -EIO;
|
|
|
|
goto out_brelse;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_brelse:
|
|
|
|
for (i = 0; i < bh_count; i++) {
|
|
|
|
brelse(bhs[i]);
|
|
|
|
bhs[i] = NULL;
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2021-08-16 17:57:04 +08:00
|
|
|
int ext4_walk_page_buffers(handle_t *handle, struct inode *inode,
|
2012-12-11 03:05:51 +08:00
|
|
|
struct buffer_head *head,
|
|
|
|
unsigned from,
|
|
|
|
unsigned to,
|
|
|
|
int *partial,
|
2021-08-16 17:57:04 +08:00
|
|
|
int (*fn)(handle_t *handle, struct inode *inode,
|
2012-12-11 03:05:51 +08:00
|
|
|
struct buffer_head *bh))
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
struct buffer_head *bh;
|
|
|
|
unsigned block_start, block_end;
|
|
|
|
unsigned blocksize = head->b_size;
|
|
|
|
int err, ret = 0;
|
|
|
|
struct buffer_head *next;
|
|
|
|
|
2008-09-09 10:25:24 +08:00
|
|
|
for (bh = head, block_start = 0;
|
|
|
|
ret == 0 && (bh != head || !block_start);
|
2009-06-15 05:45:34 +08:00
|
|
|
block_start = block_end, bh = next) {
|
2006-10-11 16:20:50 +08:00
|
|
|
next = bh->b_this_page;
|
|
|
|
block_end = block_start + blocksize;
|
|
|
|
if (block_end <= from || block_start >= to) {
|
|
|
|
if (partial && !buffer_uptodate(bh))
|
|
|
|
*partial = 1;
|
|
|
|
continue;
|
|
|
|
}
|
2021-08-16 17:57:04 +08:00
|
|
|
err = (*fn)(handle, inode, bh);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!ret)
|
|
|
|
ret = err;
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* To preserve ordering, it is essential that the hole instantiation and
|
|
|
|
* the data write be encapsulated in a single transaction. We cannot
|
2006-10-11 16:20:53 +08:00
|
|
|
* close off a transaction and start a new one between the ext4_get_block()
|
2006-10-11 16:21:01 +08:00
|
|
|
* and the commit_write(). So doing the jbd2_journal_start at the start of
|
2006-10-11 16:20:50 +08:00
|
|
|
* prepare_write() is the right place.
|
|
|
|
*
|
2013-01-28 22:30:52 +08:00
|
|
|
* Also, this function can nest inside ext4_writepage(). In that case, we
|
|
|
|
* *know* that ext4_writepage() has generated enough buffer credits to do the
|
|
|
|
* whole page. So we won't block on the journal in that case, which is good,
|
|
|
|
* because the caller may be PF_MEMALLOC.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2006-10-11 16:20:53 +08:00
|
|
|
* By accident, ext4 can be reentered when a transaction is open via
|
2006-10-11 16:20:50 +08:00
|
|
|
* quota file writes. If we were to commit the transaction while thus
|
|
|
|
* reentered, there can be a deadlock - we would be holding a quota
|
|
|
|
* lock, and the commit would never complete if another thread had a
|
|
|
|
* transaction open and was blocking on the quota lock - a ranking
|
|
|
|
* violation.
|
|
|
|
*
|
2006-10-11 16:21:01 +08:00
|
|
|
* So what we do is to rely on the fact that jbd2_journal_stop/journal_start
|
2006-10-11 16:20:50 +08:00
|
|
|
* will _not_ run commit under these circumstances because handle->h_ref
|
|
|
|
* is elevated. We'll still have enough credits for the tiny quotafile
|
|
|
|
* write.
|
|
|
|
*/
|
2021-08-16 17:57:04 +08:00
|
|
|
int do_journal_get_write_access(handle_t *handle, struct inode *inode,
|
2012-12-11 03:05:51 +08:00
|
|
|
struct buffer_head *bh)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2010-08-06 02:41:42 +08:00
|
|
|
int dirty = buffer_dirty(bh);
|
|
|
|
int ret;
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!buffer_mapped(bh) || buffer_freed(bh))
|
|
|
|
return 0;
|
2010-08-06 02:41:42 +08:00
|
|
|
/*
|
2010-10-06 16:47:23 +08:00
|
|
|
* __block_write_begin() could have dirtied some buffers. Clean
|
2010-08-06 02:41:42 +08:00
|
|
|
* the dirty bit as jbd2_journal_get_write_access() could complain
|
|
|
|
* otherwise about fs integrity issues. Setting of the dirty bit
|
2010-10-06 16:47:23 +08:00
|
|
|
* by __block_write_begin() isn't a real problem here as we clear
|
2010-08-06 02:41:42 +08:00
|
|
|
* the bit before releasing a page lock and thus writeback cannot
|
|
|
|
* ever write the buffer.
|
|
|
|
*/
|
|
|
|
if (dirty)
|
|
|
|
clear_buffer_dirty(bh);
|
2014-05-13 10:06:43 +08:00
|
|
|
BUFFER_TRACE(bh, "get write access");
|
2021-08-16 17:57:04 +08:00
|
|
|
ret = ext4_journal_get_write_access(handle, inode->i_sb, bh,
|
|
|
|
EXT4_JTR_NONE);
|
2010-08-06 02:41:42 +08:00
|
|
|
if (!ret && dirty)
|
|
|
|
ret = ext4_handle_dirty_metadata(handle, NULL, bh);
|
|
|
|
return ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2018-12-12 17:50:12 +08:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
2015-04-12 12:55:10 +08:00
|
|
|
static int ext4_block_write_begin(struct page *page, loff_t pos, unsigned len,
|
|
|
|
get_block_t *get_block)
|
|
|
|
{
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
unsigned from = pos & (PAGE_SIZE - 1);
|
2015-04-12 12:55:10 +08:00
|
|
|
unsigned to = from + len;
|
|
|
|
struct inode *inode = page->mapping->host;
|
|
|
|
unsigned block_start, block_end;
|
|
|
|
sector_t block;
|
|
|
|
int err = 0;
|
|
|
|
unsigned blocksize = inode->i_sb->s_blocksize;
|
|
|
|
unsigned bbits;
|
2019-05-21 00:29:50 +08:00
|
|
|
struct buffer_head *bh, *head, *wait[2];
|
|
|
|
int nr_wait = 0;
|
|
|
|
int i;
|
2015-04-12 12:55:10 +08:00
|
|
|
|
|
|
|
BUG_ON(!PageLocked(page));
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
BUG_ON(from > PAGE_SIZE);
|
|
|
|
BUG_ON(to > PAGE_SIZE);
|
2015-04-12 12:55:10 +08:00
|
|
|
BUG_ON(from > to);
|
|
|
|
|
|
|
|
if (!page_has_buffers(page))
|
|
|
|
create_empty_buffers(page, blocksize, 0);
|
|
|
|
head = page_buffers(page);
|
|
|
|
bbits = ilog2(blocksize);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
block = (sector_t)page->index << (PAGE_SHIFT - bbits);
|
2015-04-12 12:55:10 +08:00
|
|
|
|
|
|
|
for (bh = head, block_start = 0; bh != head || !block_start;
|
|
|
|
block++, block_start = block_end, bh = bh->b_this_page) {
|
|
|
|
block_end = block_start + blocksize;
|
|
|
|
if (block_end <= from || block_start >= to) {
|
|
|
|
if (PageUptodate(page)) {
|
2021-04-01 15:03:30 +08:00
|
|
|
set_buffer_uptodate(bh);
|
2015-04-12 12:55:10 +08:00
|
|
|
}
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (buffer_new(bh))
|
|
|
|
clear_buffer_new(bh);
|
|
|
|
if (!buffer_mapped(bh)) {
|
|
|
|
WARN_ON(bh->b_size != blocksize);
|
|
|
|
err = get_block(inode, block, bh, 1);
|
|
|
|
if (err)
|
|
|
|
break;
|
|
|
|
if (buffer_new(bh)) {
|
|
|
|
if (PageUptodate(page)) {
|
|
|
|
clear_buffer_new(bh);
|
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
mark_buffer_dirty(bh);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (block_end > to || block_start < from)
|
|
|
|
zero_user_segments(page, to, block_end,
|
|
|
|
block_start, from);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (PageUptodate(page)) {
|
2021-04-01 15:03:30 +08:00
|
|
|
set_buffer_uptodate(bh);
|
2015-04-12 12:55:10 +08:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (!buffer_uptodate(bh) && !buffer_delay(bh) &&
|
|
|
|
!buffer_unwritten(bh) &&
|
|
|
|
(block_start < from || block_end > to)) {
|
2020-09-24 15:33:33 +08:00
|
|
|
ext4_read_bh_lock(bh, 0, false);
|
2019-05-21 00:29:50 +08:00
|
|
|
wait[nr_wait++] = bh;
|
2015-04-12 12:55:10 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* If we issued read requests, let them complete.
|
|
|
|
*/
|
2019-05-21 00:29:50 +08:00
|
|
|
for (i = 0; i < nr_wait; i++) {
|
|
|
|
wait_on_buffer(wait[i]);
|
|
|
|
if (!buffer_uptodate(wait[i]))
|
2015-04-12 12:55:10 +08:00
|
|
|
err = -EIO;
|
|
|
|
}
|
2019-05-21 00:29:49 +08:00
|
|
|
if (unlikely(err)) {
|
2015-04-12 12:55:10 +08:00
|
|
|
page_zero_new_buffers(page, from, to);
|
2020-07-02 09:56:07 +08:00
|
|
|
} else if (fscrypt_inode_uses_fs_layer_crypto(inode)) {
|
2019-05-21 00:29:50 +08:00
|
|
|
for (i = 0; i < nr_wait; i++) {
|
|
|
|
int err2;
|
|
|
|
|
|
|
|
err2 = fscrypt_decrypt_pagecache_blocks(page, blocksize,
|
|
|
|
bh_offset(wait[i]));
|
|
|
|
if (err2) {
|
|
|
|
clear_buffer_uptodate(wait[i]);
|
|
|
|
err = err2;
|
|
|
|
}
|
|
|
|
}
|
2019-05-21 00:29:49 +08:00
|
|
|
}
|
|
|
|
|
2015-04-12 12:55:10 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2007-10-16 16:25:05 +08:00
|
|
|
static int ext4_write_begin(struct file *file, struct address_space *mapping,
|
2022-02-23 03:31:43 +08:00
|
|
|
loff_t pos, unsigned len,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct page **pagep, void **fsdata)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2008-09-09 10:25:24 +08:00
|
|
|
struct inode *inode = mapping->host;
|
2009-06-05 13:00:26 +08:00
|
|
|
int ret, needed_blocks;
|
2006-10-11 16:20:50 +08:00
|
|
|
handle_t *handle;
|
|
|
|
int retries = 0;
|
2008-09-09 10:25:24 +08:00
|
|
|
struct page *page;
|
2009-06-15 05:45:34 +08:00
|
|
|
pgoff_t index;
|
2008-09-09 10:25:24 +08:00
|
|
|
unsigned from, to;
|
2007-10-16 16:25:05 +08:00
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2022-02-23 03:31:43 +08:00
|
|
|
trace_ext4_write_begin(inode, pos, len);
|
2009-06-05 13:00:26 +08:00
|
|
|
/*
|
|
|
|
* Reserve one block more for addition to orphan list in case
|
|
|
|
* we allocate blocks but write fails for some reason
|
|
|
|
*/
|
|
|
|
needed_blocks = ext4_writepage_trans_blocks(inode) + 1;
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
index = pos >> PAGE_SHIFT;
|
|
|
|
from = pos & (PAGE_SIZE - 1);
|
2008-09-09 10:25:24 +08:00
|
|
|
to = from + len;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2012-12-11 03:05:51 +08:00
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) {
|
|
|
|
ret = ext4_try_to_write_inline_data(mapping, inode, pos, len,
|
2022-02-22 23:36:28 +08:00
|
|
|
pagep);
|
2012-12-11 03:05:51 +08:00
|
|
|
if (ret < 0)
|
2013-02-09 22:24:14 +08:00
|
|
|
return ret;
|
|
|
|
if (ret == 1)
|
|
|
|
return 0;
|
2012-12-11 03:05:51 +08:00
|
|
|
}
|
|
|
|
|
2013-02-09 22:24:14 +08:00
|
|
|
/*
|
|
|
|
* grab_cache_page_write_begin() can take a long time if the
|
|
|
|
* system is thrashing due to memory pressure, or if the page
|
|
|
|
* is being written back. So grab it first before we start
|
|
|
|
* the transaction handle. This also allows us to allocate
|
|
|
|
* the page (if needed) without using GFP_NOFS.
|
|
|
|
*/
|
|
|
|
retry_grab:
|
2022-02-23 00:25:12 +08:00
|
|
|
page = grab_cache_page_write_begin(mapping, index);
|
2013-02-09 22:24:14 +08:00
|
|
|
if (!page)
|
|
|
|
return -ENOMEM;
|
|
|
|
unlock_page(page);
|
|
|
|
|
|
|
|
retry_journal:
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks);
|
2008-09-09 10:25:24 +08:00
|
|
|
if (IS_ERR(handle)) {
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2013-02-09 22:24:14 +08:00
|
|
|
return PTR_ERR(handle);
|
2007-04-02 14:49:44 +08:00
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2013-02-09 22:24:14 +08:00
|
|
|
lock_page(page);
|
|
|
|
if (page->mapping != mapping) {
|
|
|
|
/* The page got truncated from under us */
|
|
|
|
unlock_page(page);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2008-07-12 07:27:31 +08:00
|
|
|
ext4_journal_stop(handle);
|
2013-02-09 22:24:14 +08:00
|
|
|
goto retry_grab;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
2013-08-29 02:30:47 +08:00
|
|
|
/* In case writeback began while the page was unlocked */
|
|
|
|
wait_for_stable_page(page);
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2018-12-12 17:50:12 +08:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
2015-04-12 12:55:10 +08:00
|
|
|
if (ext4_should_dioread_nolock(inode))
|
|
|
|
ret = ext4_block_write_begin(page, pos, len,
|
2016-03-09 12:08:10 +08:00
|
|
|
ext4_get_block_unwritten);
|
2015-04-12 12:55:10 +08:00
|
|
|
else
|
|
|
|
ret = ext4_block_write_begin(page, pos, len,
|
|
|
|
ext4_get_block);
|
|
|
|
#else
|
2010-03-05 05:14:02 +08:00
|
|
|
if (ext4_should_dioread_nolock(inode))
|
2016-03-09 12:08:10 +08:00
|
|
|
ret = __block_write_begin(page, pos, len,
|
|
|
|
ext4_get_block_unwritten);
|
2010-03-05 05:14:02 +08:00
|
|
|
else
|
2010-06-04 17:29:57 +08:00
|
|
|
ret = __block_write_begin(page, pos, len, ext4_get_block);
|
2015-04-12 12:55:10 +08:00
|
|
|
#endif
|
2007-10-16 16:25:05 +08:00
|
|
|
if (!ret && ext4_should_journal_data(inode)) {
|
2021-08-16 17:57:04 +08:00
|
|
|
ret = ext4_walk_page_buffers(handle, inode,
|
|
|
|
page_buffers(page), from, to, NULL,
|
2012-12-11 03:05:51 +08:00
|
|
|
do_journal_get_write_access);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2007-10-16 16:25:05 +08:00
|
|
|
|
|
|
|
if (ret) {
|
2019-07-23 00:26:24 +08:00
|
|
|
bool extended = (pos + len > inode->i_size) &&
|
|
|
|
!ext4_verity_in_progress(inode);
|
|
|
|
|
2008-09-09 10:25:24 +08:00
|
|
|
unlock_page(page);
|
2008-09-14 01:10:25 +08:00
|
|
|
/*
|
2010-06-04 17:29:57 +08:00
|
|
|
* __block_write_begin may have instantiated a few blocks
|
2008-09-14 01:10:25 +08:00
|
|
|
* outside i_size. Trim these off again. Don't need
|
2022-01-21 15:06:11 +08:00
|
|
|
* i_size_read because we hold i_rwsem.
|
2009-06-05 13:00:26 +08:00
|
|
|
*
|
|
|
|
* Add inode to orphan list in case we crash before
|
|
|
|
* truncate finishes
|
2008-09-14 01:10:25 +08:00
|
|
|
*/
|
2019-07-23 00:26:24 +08:00
|
|
|
if (extended && ext4_can_truncate(inode))
|
2009-06-05 13:00:26 +08:00
|
|
|
ext4_orphan_add(handle, inode);
|
|
|
|
|
|
|
|
ext4_journal_stop(handle);
|
2019-07-23 00:26:24 +08:00
|
|
|
if (extended) {
|
2009-12-09 10:24:33 +08:00
|
|
|
ext4_truncate_failed_write(inode);
|
2009-06-15 05:45:34 +08:00
|
|
|
/*
|
2009-07-14 04:22:22 +08:00
|
|
|
* If truncate failed early the inode might
|
2009-06-05 13:00:26 +08:00
|
|
|
* still be on the orphan list; we need to
|
|
|
|
* make sure the inode is removed from the
|
|
|
|
* orphan list in that case.
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink)
|
|
|
|
ext4_orphan_del(NULL, inode);
|
|
|
|
}
|
2007-10-16 16:25:05 +08:00
|
|
|
|
2013-02-09 22:24:14 +08:00
|
|
|
if (ret == -ENOSPC &&
|
|
|
|
ext4_should_retry_alloc(inode->i_sb, &retries))
|
|
|
|
goto retry_journal;
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2013-02-09 22:24:14 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
*pagep = page;
|
2006-10-11 16:20:50 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2007-10-16 16:25:05 +08:00
|
|
|
/* For write_end() in data=journal mode */
|
2021-08-16 17:57:04 +08:00
|
|
|
static int write_end_fn(handle_t *handle, struct inode *inode,
|
|
|
|
struct buffer_head *bh)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2013-04-22 04:45:54 +08:00
|
|
|
int ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!buffer_mapped(bh) || buffer_freed(bh))
|
|
|
|
return 0;
|
|
|
|
set_buffer_uptodate(bh);
|
2013-04-22 04:45:54 +08:00
|
|
|
ret = ext4_handle_dirty_metadata(handle, NULL, bh);
|
|
|
|
clear_buffer_meta(bh);
|
|
|
|
clear_buffer_prio(bh);
|
|
|
|
return ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2013-04-04 00:41:17 +08:00
|
|
|
/*
|
|
|
|
* We need to pick up the new inode size which generic_commit_write gave us
|
|
|
|
* `file' can be NULL - eg, when called from page_symlink().
|
|
|
|
*
|
|
|
|
* ext4 never places buffers on inode->i_mapping->private_list. metadata
|
|
|
|
* buffers are managed internally.
|
|
|
|
*/
|
|
|
|
static int ext4_write_end(struct file *file,
|
|
|
|
struct address_space *mapping,
|
|
|
|
loff_t pos, unsigned len, unsigned copied,
|
|
|
|
struct page *page, void *fsdata)
|
2009-06-05 12:56:49 +08:00
|
|
|
{
|
|
|
|
handle_t *handle = ext4_journal_current_handle();
|
2013-04-04 00:41:17 +08:00
|
|
|
struct inode *inode = mapping->host;
|
2015-02-13 12:00:17 +08:00
|
|
|
loff_t old_size = inode->i_size;
|
2013-04-04 00:41:17 +08:00
|
|
|
int ret = 0, ret2;
|
|
|
|
int i_size_changed = 0;
|
2019-07-23 00:26:24 +08:00
|
|
|
bool verity = ext4_verity_in_progress(inode);
|
2013-04-04 00:41:17 +08:00
|
|
|
|
|
|
|
trace_ext4_write_end(inode, pos, len, copied);
|
2021-07-16 20:20:23 +08:00
|
|
|
|
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
return ext4_write_inline_data_end(inode, pos, len, copied, page);
|
|
|
|
|
|
|
|
copied = block_write_end(file, mapping, pos, len, copied, page, fsdata);
|
2009-06-05 12:56:49 +08:00
|
|
|
/*
|
2014-08-24 05:48:28 +08:00
|
|
|
* it's important to update i_size while still holding page lock:
|
2009-06-05 12:56:49 +08:00
|
|
|
* page writeout could otherwise come in and zero beyond i_size.
|
2019-07-23 00:26:24 +08:00
|
|
|
*
|
|
|
|
* If FS_IOC_ENABLE_VERITY is running on this inode, then Merkle tree
|
|
|
|
* blocks are being written past EOF, so skip the i_size update.
|
2009-06-05 12:56:49 +08:00
|
|
|
*/
|
2019-07-23 00:26:24 +08:00
|
|
|
if (!verity)
|
|
|
|
i_size_changed = ext4_update_inode_size(inode, pos + copied);
|
2009-06-05 12:56:49 +08:00
|
|
|
unlock_page(page);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2009-06-05 12:56:49 +08:00
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (old_size < pos && !verity)
|
2015-02-13 12:00:17 +08:00
|
|
|
pagecache_isize_extended(inode, old_size, pos);
|
2009-06-05 12:56:49 +08:00
|
|
|
/*
|
|
|
|
* Don't mark the inode dirty under page lock. First, it unnecessarily
|
|
|
|
* makes the holding time of page lock longer. Second, it forces lock
|
|
|
|
* ordering of page lock and transaction start for journaling
|
|
|
|
* filesystems.
|
|
|
|
*/
|
2021-07-16 20:20:23 +08:00
|
|
|
if (i_size_changed)
|
2020-04-27 09:34:37 +08:00
|
|
|
ret = ext4_mark_inode_dirty(handle, inode);
|
2009-06-05 12:56:49 +08:00
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (pos + len > inode->i_size && !verity && ext4_can_truncate(inode))
|
2009-06-05 12:56:49 +08:00
|
|
|
/* if we have allocated more blocks and copied
|
|
|
|
* less. We will have blocks allocated outside
|
|
|
|
* inode->i_size. So truncate them
|
|
|
|
*/
|
|
|
|
ext4_orphan_add(handle, inode);
|
2021-07-16 20:20:22 +08:00
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
ret2 = ext4_journal_stop(handle);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!ret)
|
|
|
|
ret = ret2;
|
2007-10-16 16:25:05 +08:00
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (pos + len > inode->i_size && !verity) {
|
2009-12-09 10:24:33 +08:00
|
|
|
ext4_truncate_failed_write(inode);
|
2009-06-15 05:45:34 +08:00
|
|
|
/*
|
2009-07-14 04:22:22 +08:00
|
|
|
* If truncate failed early the inode might still be
|
2009-06-05 12:56:49 +08:00
|
|
|
* on the orphan list; we need to make sure the inode
|
|
|
|
* is removed from the orphan list in that case.
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink)
|
|
|
|
ext4_orphan_del(NULL, inode);
|
|
|
|
}
|
|
|
|
|
2007-10-16 16:25:05 +08:00
|
|
|
return ret ? ret : copied;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2015-10-15 22:29:05 +08:00
|
|
|
/*
|
|
|
|
* This is a private version of page_zero_new_buffers() which doesn't
|
|
|
|
* set the buffer to be dirty, since in data=journalled mode we need
|
|
|
|
* to call ext4_handle_dirty_metadata() instead.
|
|
|
|
*/
|
2017-01-28 03:35:38 +08:00
|
|
|
static void ext4_journalled_zero_new_buffers(handle_t *handle,
|
2021-08-16 17:57:04 +08:00
|
|
|
struct inode *inode,
|
2017-01-28 03:35:38 +08:00
|
|
|
struct page *page,
|
|
|
|
unsigned from, unsigned to)
|
2015-10-15 22:29:05 +08:00
|
|
|
{
|
|
|
|
unsigned int block_start = 0, block_end;
|
|
|
|
struct buffer_head *head, *bh;
|
|
|
|
|
|
|
|
bh = head = page_buffers(page);
|
|
|
|
do {
|
|
|
|
block_end = block_start + bh->b_size;
|
|
|
|
if (buffer_new(bh)) {
|
|
|
|
if (block_end > from && block_start < to) {
|
|
|
|
if (!PageUptodate(page)) {
|
|
|
|
unsigned start, size;
|
|
|
|
|
|
|
|
start = max(from, block_start);
|
|
|
|
size = min(to, block_end) - start;
|
|
|
|
|
|
|
|
zero_user(page, start, size);
|
2021-08-16 17:57:04 +08:00
|
|
|
write_end_fn(handle, inode, bh);
|
2015-10-15 22:29:05 +08:00
|
|
|
}
|
|
|
|
clear_buffer_new(bh);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
block_start = block_end;
|
|
|
|
bh = bh->b_this_page;
|
|
|
|
} while (bh != head);
|
|
|
|
}
|
|
|
|
|
2007-10-16 16:25:05 +08:00
|
|
|
static int ext4_journalled_write_end(struct file *file,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct address_space *mapping,
|
|
|
|
loff_t pos, unsigned len, unsigned copied,
|
|
|
|
struct page *page, void *fsdata)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2006-10-11 16:20:53 +08:00
|
|
|
handle_t *handle = ext4_journal_current_handle();
|
2007-10-16 16:25:05 +08:00
|
|
|
struct inode *inode = mapping->host;
|
2015-02-13 12:00:17 +08:00
|
|
|
loff_t old_size = inode->i_size;
|
2006-10-11 16:20:50 +08:00
|
|
|
int ret = 0, ret2;
|
|
|
|
int partial = 0;
|
2007-10-16 16:25:05 +08:00
|
|
|
unsigned from, to;
|
2014-08-24 05:48:28 +08:00
|
|
|
int size_changed = 0;
|
2019-07-23 00:26:24 +08:00
|
|
|
bool verity = ext4_verity_in_progress(inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2009-06-17 23:48:11 +08:00
|
|
|
trace_ext4_journalled_write_end(inode, pos, len, copied);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
from = pos & (PAGE_SIZE - 1);
|
2007-10-16 16:25:05 +08:00
|
|
|
to = from + len;
|
|
|
|
|
ext4: Fix ext4_should_writeback_data() for no-journal mode
ext4_should_writeback_data() had an incorrect sequence of
tests to determine if it should return 0 or 1: in
particular, even in no-journal mode, 0 was being returned
for a non-regular-file inode.
This meant that, in non-journal mode, we would use
ext4_journalled_aops for directories, symlinks, and other
non-regular files. However, calling journalled aop
callbacks when there is no valid handle, can cause problems.
This would cause a kernel crash with Jan Kara's commit
2d859db3e4 ("ext4: fix data corruption in inodes with
journalled data"), because we now dereference 'handle' in
ext4_journalled_write_end().
I also added BUG_ONs to check for a valid handle in the
obviously journal-only aops callbacks.
I tested this running xfstests with a scratch device in
these modes:
- no-journal
- data=ordered
- data=writeback
- data=journal
All work fine; the data=journal run has many failures and a
crash in xfstests 074, but this is no different from a
vanilla kernel.
Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
2011-08-13 23:25:18 +08:00
|
|
|
BUG_ON(!ext4_handle_valid(handle));
|
|
|
|
|
2021-07-16 20:20:23 +08:00
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
return ext4_write_inline_data_end(inode, pos, len, copied, page);
|
|
|
|
|
|
|
|
if (unlikely(copied < len) && !PageUptodate(page)) {
|
2017-01-28 03:35:38 +08:00
|
|
|
copied = 0;
|
2021-08-16 17:57:04 +08:00
|
|
|
ext4_journalled_zero_new_buffers(handle, inode, page, from, to);
|
2017-01-28 03:35:38 +08:00
|
|
|
} else {
|
|
|
|
if (unlikely(copied < len))
|
2021-08-16 17:57:04 +08:00
|
|
|
ext4_journalled_zero_new_buffers(handle, inode, page,
|
2017-01-28 03:35:38 +08:00
|
|
|
from + copied, to);
|
2021-08-16 17:57:04 +08:00
|
|
|
ret = ext4_walk_page_buffers(handle, inode, page_buffers(page),
|
|
|
|
from, from + copied, &partial,
|
2017-01-28 03:35:38 +08:00
|
|
|
write_end_fn);
|
2012-12-11 03:05:57 +08:00
|
|
|
if (!partial)
|
|
|
|
SetPageUptodate(page);
|
|
|
|
}
|
2019-07-23 00:26:24 +08:00
|
|
|
if (!verity)
|
|
|
|
size_changed = ext4_update_inode_size(inode, pos + copied);
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_JDATA);
|
2011-07-26 21:07:11 +08:00
|
|
|
EXT4_I(inode)->i_datasync_tid = handle->h_transaction->t_tid;
|
2014-08-24 05:48:28 +08:00
|
|
|
unlock_page(page);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2014-08-24 05:48:28 +08:00
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (old_size < pos && !verity)
|
2015-02-13 12:00:17 +08:00
|
|
|
pagecache_isize_extended(inode, old_size, pos);
|
|
|
|
|
2021-07-16 20:20:23 +08:00
|
|
|
if (size_changed) {
|
2006-10-11 16:20:53 +08:00
|
|
|
ret2 = ext4_mark_inode_dirty(handle, inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!ret)
|
|
|
|
ret = ret2;
|
|
|
|
}
|
2007-10-16 16:25:05 +08:00
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (pos + len > inode->i_size && !verity && ext4_can_truncate(inode))
|
2009-06-05 12:56:49 +08:00
|
|
|
/* if we have allocated more blocks and copied
|
|
|
|
* less. We will have blocks allocated outside
|
|
|
|
* inode->i_size. So truncate them
|
|
|
|
*/
|
|
|
|
ext4_orphan_add(handle, inode);
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
ret2 = ext4_journal_stop(handle);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!ret)
|
|
|
|
ret = ret2;
|
2019-07-23 00:26:24 +08:00
|
|
|
if (pos + len > inode->i_size && !verity) {
|
2009-12-09 10:24:33 +08:00
|
|
|
ext4_truncate_failed_write(inode);
|
2009-06-15 05:45:34 +08:00
|
|
|
/*
|
2009-07-14 04:22:22 +08:00
|
|
|
* If truncate failed early the inode might still be
|
2009-06-05 12:56:49 +08:00
|
|
|
* on the orphan list; we need to make sure the inode
|
|
|
|
* is removed from the orphan list in that case.
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink)
|
|
|
|
ext4_orphan_del(NULL, inode);
|
|
|
|
}
|
2007-10-16 16:25:05 +08:00
|
|
|
|
|
|
|
return ret ? ret : copied;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2008-07-15 05:52:37 +08:00
|
|
|
|
2010-01-01 15:41:30 +08:00
|
|
|
/*
|
2015-06-22 09:37:05 +08:00
|
|
|
* Reserve space for a single cluster
|
2010-01-01 15:41:30 +08:00
|
|
|
*/
|
2015-06-22 09:37:05 +08:00
|
|
|
static int ext4_da_reserve_space(struct inode *inode)
|
2008-07-15 05:52:37 +08:00
|
|
|
{
|
2009-01-23 01:13:05 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2009-12-31 03:20:45 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
2010-03-03 22:05:00 +08:00
|
|
|
int ret;
|
2012-07-23 12:00:20 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* We will charge metadata quota at writeout time; this saves
|
|
|
|
* us from metadata over-estimation, though we may go over by
|
|
|
|
* a small amount in the end. Here we just reserve for data.
|
|
|
|
*/
|
|
|
|
ret = dquot_reserve_block(inode, EXT4_C2B(sbi, 1));
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
2008-07-15 05:52:37 +08:00
|
|
|
|
2009-12-31 03:20:45 +08:00
|
|
|
spin_lock(&ei->i_block_reservation_lock);
|
2014-07-15 18:02:38 +08:00
|
|
|
if (ext4_claim_free_clusters(sbi, 1, 0)) {
|
2012-07-23 12:00:20 +08:00
|
|
|
spin_unlock(&ei->i_block_reservation_lock);
|
|
|
|
dquot_release_reservation_block(inode, EXT4_C2B(sbi, 1));
|
2008-07-15 05:52:37 +08:00
|
|
|
return -ENOSPC;
|
|
|
|
}
|
2010-01-01 15:41:30 +08:00
|
|
|
ei->i_reserved_data_blocks++;
|
2015-06-22 09:37:05 +08:00
|
|
|
trace_ext4_da_reserve_space(inode);
|
2009-12-31 03:20:45 +08:00
|
|
|
spin_unlock(&ei->i_block_reservation_lock);
|
2009-12-11 00:36:27 +08:00
|
|
|
|
2008-07-15 05:52:37 +08:00
|
|
|
return 0; /* success */
|
|
|
|
}
|
|
|
|
|
2018-10-02 02:33:24 +08:00
|
|
|
void ext4_da_release_space(struct inode *inode, int to_free)
|
2008-07-15 05:52:37 +08:00
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2009-12-31 03:20:45 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
2008-07-15 05:52:37 +08:00
|
|
|
|
2008-08-20 10:16:59 +08:00
|
|
|
if (!to_free)
|
|
|
|
return; /* Nothing to release, exit */
|
|
|
|
|
2008-07-15 05:52:37 +08:00
|
|
|
spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
|
2008-08-20 10:16:59 +08:00
|
|
|
|
2010-05-17 14:00:00 +08:00
|
|
|
trace_ext4_da_release_space(inode, to_free);
|
2009-12-31 03:20:45 +08:00
|
|
|
if (unlikely(to_free > ei->i_reserved_data_blocks)) {
|
2008-08-20 10:16:59 +08:00
|
|
|
/*
|
2009-12-31 03:20:45 +08:00
|
|
|
* if there aren't enough reserved blocks, then the
|
|
|
|
* counter is messed up somewhere. Since this
|
|
|
|
* function is called from invalidate page, it's
|
|
|
|
* harmless to return without any action.
|
2008-08-20 10:16:59 +08:00
|
|
|
*/
|
2013-02-15 04:11:41 +08:00
|
|
|
ext4_warning(inode->i_sb, "ext4_da_release_space: "
|
2009-12-31 03:20:45 +08:00
|
|
|
"ino %lu, to_free %d with only %d reserved "
|
2012-03-20 11:13:43 +08:00
|
|
|
"data blocks", inode->i_ino, to_free,
|
2009-12-31 03:20:45 +08:00
|
|
|
ei->i_reserved_data_blocks);
|
|
|
|
WARN_ON(1);
|
|
|
|
to_free = ei->i_reserved_data_blocks;
|
2008-08-20 10:16:59 +08:00
|
|
|
}
|
2009-12-31 03:20:45 +08:00
|
|
|
ei->i_reserved_data_blocks -= to_free;
|
2008-08-20 10:16:59 +08:00
|
|
|
|
2010-05-16 23:00:00 +08:00
|
|
|
/* update fs dirty data blocks counter */
|
2011-09-10 06:56:51 +08:00
|
|
|
percpu_counter_sub(&sbi->s_dirtyclusters_counter, to_free);
|
2008-07-15 05:52:37 +08:00
|
|
|
|
|
|
|
spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
|
2009-01-23 01:13:05 +08:00
|
|
|
|
2011-09-10 07:04:51 +08:00
|
|
|
dquot_release_reservation_block(inode, EXT4_C2B(sbi, to_free));
|
2008-07-15 05:52:37 +08:00
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
|
|
|
* Delayed allocation stuff
|
|
|
|
*/
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
struct mpage_da_data {
|
|
|
|
struct inode *inode;
|
|
|
|
struct writeback_control *wbc;
|
2013-06-05 01:21:11 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
pgoff_t first_page; /* The first page to write */
|
|
|
|
pgoff_t next_page; /* Current page to examine */
|
|
|
|
pgoff_t last_page; /* Last page to examine */
|
2009-01-06 10:50:43 +08:00
|
|
|
/*
|
2013-06-05 01:17:40 +08:00
|
|
|
* Extent to map - this can be after first_page because that can be
|
|
|
|
* fully mapped. We somewhat abuse m_flags to store whether the extent
|
|
|
|
* is delalloc or unwritten.
|
2009-01-06 10:50:43 +08:00
|
|
|
*/
|
2013-06-05 01:17:40 +08:00
|
|
|
struct ext4_map_blocks map;
|
|
|
|
struct ext4_io_submit io_submit; /* IO submission data */
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
unsigned int do_map:1;
|
2020-05-25 16:12:15 +08:00
|
|
|
unsigned int scanned_until_end:1;
|
2013-06-05 01:17:40 +08:00
|
|
|
};
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
static void mpage_release_unused_pages(struct mpage_da_data *mpd,
|
|
|
|
bool invalidate)
|
2008-08-20 09:08:18 +08:00
|
|
|
{
|
2022-06-04 03:36:49 +08:00
|
|
|
unsigned nr, i;
|
2008-08-20 09:08:18 +08:00
|
|
|
pgoff_t index, end;
|
2022-06-04 03:36:49 +08:00
|
|
|
struct folio_batch fbatch;
|
2008-08-20 09:08:18 +08:00
|
|
|
struct inode *inode = mpd->inode;
|
|
|
|
struct address_space *mapping = inode->i_mapping;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
/* This is necessary when next_page == 0. */
|
|
|
|
if (mpd->first_page >= mpd->next_page)
|
|
|
|
return;
|
2008-08-20 09:08:18 +08:00
|
|
|
|
2020-05-25 16:12:15 +08:00
|
|
|
mpd->scanned_until_end = 0;
|
2011-02-27 01:27:52 +08:00
|
|
|
index = mpd->first_page;
|
|
|
|
end = mpd->next_page - 1;
|
2013-06-05 01:17:40 +08:00
|
|
|
if (invalidate) {
|
|
|
|
ext4_lblk_t start, last;
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
start = index << (PAGE_SHIFT - inode->i_blkbits);
|
|
|
|
last = end << (PAGE_SHIFT - inode->i_blkbits);
|
2022-06-16 00:05:30 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* avoid racing with extent status tree scans made by
|
|
|
|
* ext4_insert_delayed_block()
|
|
|
|
*/
|
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
2013-06-05 01:17:40 +08:00
|
|
|
ext4_es_remove_extent(inode, start, last - start + 1);
|
2022-06-16 00:05:30 +08:00
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
2012-11-09 10:57:32 +08:00
|
|
|
|
2022-06-04 03:36:49 +08:00
|
|
|
folio_batch_init(&fbatch);
|
2008-08-20 09:08:18 +08:00
|
|
|
while (index <= end) {
|
2022-06-04 03:36:49 +08:00
|
|
|
nr = filemap_get_folios(mapping, &index, end, &fbatch);
|
|
|
|
if (nr == 0)
|
2008-08-20 09:08:18 +08:00
|
|
|
break;
|
2022-06-04 03:36:49 +08:00
|
|
|
for (i = 0; i < nr; i++) {
|
|
|
|
struct folio *folio = fbatch.folios[i];
|
2017-09-07 07:21:30 +08:00
|
|
|
|
2022-06-04 03:36:49 +08:00
|
|
|
if (folio->index < mpd->first_page)
|
|
|
|
continue;
|
|
|
|
if (folio->index + folio_nr_pages(folio) - 1 > end)
|
|
|
|
continue;
|
2022-02-10 04:21:34 +08:00
|
|
|
BUG_ON(!folio_test_locked(folio));
|
|
|
|
BUG_ON(folio_test_writeback(folio));
|
2013-06-05 01:17:40 +08:00
|
|
|
if (invalidate) {
|
2022-02-10 04:21:34 +08:00
|
|
|
if (folio_mapped(folio))
|
|
|
|
folio_clear_dirty_for_io(folio);
|
|
|
|
block_invalidate_folio(folio, 0,
|
|
|
|
folio_size(folio));
|
|
|
|
folio_clear_uptodate(folio);
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
2022-02-10 04:21:34 +08:00
|
|
|
folio_unlock(folio);
|
2008-08-20 09:08:18 +08:00
|
|
|
}
|
2022-06-04 03:36:49 +08:00
|
|
|
folio_batch_release(&fbatch);
|
2008-08-20 09:08:18 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-09-09 11:05:34 +08:00
|
|
|
static void ext4_print_free_blocks(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2012-03-20 11:41:49 +08:00
|
|
|
struct super_block *sb = inode->i_sb;
|
2013-04-04 11:33:30 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
2012-03-20 11:41:49 +08:00
|
|
|
|
|
|
|
ext4_msg(sb, KERN_CRIT, "Total free blocks count %lld",
|
2011-09-10 07:10:51 +08:00
|
|
|
EXT4_C2B(EXT4_SB(inode->i_sb),
|
2013-04-04 11:33:30 +08:00
|
|
|
ext4_count_free_clusters(sb)));
|
2012-03-20 11:41:49 +08:00
|
|
|
ext4_msg(sb, KERN_CRIT, "Free/Dirty block details");
|
|
|
|
ext4_msg(sb, KERN_CRIT, "free_blocks=%lld",
|
2013-04-04 11:33:30 +08:00
|
|
|
(long long) EXT4_C2B(EXT4_SB(sb),
|
2011-09-10 06:56:51 +08:00
|
|
|
percpu_counter_sum(&sbi->s_freeclusters_counter)));
|
2012-03-20 11:41:49 +08:00
|
|
|
ext4_msg(sb, KERN_CRIT, "dirty_blocks=%lld",
|
2013-04-04 11:33:30 +08:00
|
|
|
(long long) EXT4_C2B(EXT4_SB(sb),
|
2011-09-10 07:04:51 +08:00
|
|
|
percpu_counter_sum(&sbi->s_dirtyclusters_counter)));
|
2012-03-20 11:41:49 +08:00
|
|
|
ext4_msg(sb, KERN_CRIT, "Block reservation details");
|
|
|
|
ext4_msg(sb, KERN_CRIT, "i_reserved_data_blocks=%u",
|
2013-04-04 11:33:30 +08:00
|
|
|
ei->i_reserved_data_blocks);
|
2008-09-09 11:05:34 +08:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2021-08-16 17:57:04 +08:00
|
|
|
static int ext4_bh_delay_or_unwritten(handle_t *handle, struct inode *inode,
|
|
|
|
struct buffer_head *bh)
|
2009-05-13 04:30:27 +08:00
|
|
|
{
|
2009-06-15 05:57:10 +08:00
|
|
|
return (buffer_delay(bh) || buffer_unwritten(bh)) && buffer_dirty(bh);
|
2009-05-13 04:30:27 +08:00
|
|
|
}
|
|
|
|
|
2018-10-02 02:19:37 +08:00
|
|
|
/*
|
|
|
|
* ext4_insert_delayed_block - adds a delayed block to the extents status
|
|
|
|
* tree, incrementing the reserved cluster/block
|
|
|
|
* count or making a pending reservation
|
|
|
|
* where needed
|
|
|
|
*
|
|
|
|
* @inode - file containing the newly added block
|
|
|
|
* @lblk - logical block to be added
|
|
|
|
*
|
|
|
|
* Returns 0 on success, negative error code on failure.
|
|
|
|
*/
|
|
|
|
static int ext4_insert_delayed_block(struct inode *inode, ext4_lblk_t lblk)
|
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
int ret;
|
|
|
|
bool allocated = false;
|
2021-08-23 14:13:58 +08:00
|
|
|
bool reserved = false;
|
2018-10-02 02:19:37 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If the cluster containing lblk is shared with a delayed,
|
|
|
|
* written, or unwritten extent in a bigalloc file system, it's
|
|
|
|
* already been accounted for and does not need to be reserved.
|
|
|
|
* A pending reservation must be made for the cluster if it's
|
|
|
|
* shared with a written or unwritten extent and doesn't already
|
|
|
|
* have one. Written and unwritten extents can be purged from the
|
|
|
|
* extents status tree if the system is under memory pressure, so
|
|
|
|
* it's necessary to examine the extent tree if a search of the
|
|
|
|
* extents status tree doesn't get a match.
|
|
|
|
*/
|
|
|
|
if (sbi->s_cluster_ratio == 1) {
|
|
|
|
ret = ext4_da_reserve_space(inode);
|
|
|
|
if (ret != 0) /* ENOSPC */
|
|
|
|
goto errout;
|
2021-08-23 14:13:58 +08:00
|
|
|
reserved = true;
|
2018-10-02 02:19:37 +08:00
|
|
|
} else { /* bigalloc */
|
|
|
|
if (!ext4_es_scan_clu(inode, &ext4_es_is_delonly, lblk)) {
|
|
|
|
if (!ext4_es_scan_clu(inode,
|
|
|
|
&ext4_es_is_mapped, lblk)) {
|
|
|
|
ret = ext4_clu_mapped(inode,
|
|
|
|
EXT4_B2C(sbi, lblk));
|
|
|
|
if (ret < 0)
|
|
|
|
goto errout;
|
|
|
|
if (ret == 0) {
|
|
|
|
ret = ext4_da_reserve_space(inode);
|
|
|
|
if (ret != 0) /* ENOSPC */
|
|
|
|
goto errout;
|
2021-08-23 14:13:58 +08:00
|
|
|
reserved = true;
|
2018-10-02 02:19:37 +08:00
|
|
|
} else {
|
|
|
|
allocated = true;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
allocated = true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = ext4_es_insert_delayed_block(inode, lblk, allocated);
|
2021-08-23 14:13:58 +08:00
|
|
|
if (ret && reserved)
|
|
|
|
ext4_da_release_space(inode, 1);
|
2018-10-02 02:19:37 +08:00
|
|
|
|
|
|
|
errout:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2011-09-10 07:20:51 +08:00
|
|
|
/*
|
|
|
|
* This function is grabs code from the very beginning of
|
|
|
|
* ext4_map_blocks, but assumes that the caller is from delayed write
|
|
|
|
* time. This function looks up the requested blocks and sets the
|
|
|
|
* buffer delay bit under the protection of i_data_sem.
|
|
|
|
*/
|
|
|
|
static int ext4_da_map_blocks(struct inode *inode, sector_t iblock,
|
|
|
|
struct ext4_map_blocks *map,
|
|
|
|
struct buffer_head *bh)
|
|
|
|
{
|
2013-02-18 13:29:59 +08:00
|
|
|
struct extent_status es;
|
2011-09-10 07:20:51 +08:00
|
|
|
int retval;
|
|
|
|
sector_t invalid_block = ~((sector_t) 0xffff);
|
2013-03-11 09:01:03 +08:00
|
|
|
#ifdef ES_AGGRESSIVE_TEST
|
|
|
|
struct ext4_map_blocks orig_map;
|
|
|
|
|
|
|
|
memcpy(&orig_map, map, sizeof(*map));
|
|
|
|
#endif
|
2011-09-10 07:20:51 +08:00
|
|
|
|
|
|
|
if (invalid_block < ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es))
|
|
|
|
invalid_block = ~0;
|
|
|
|
|
|
|
|
map->m_flags = 0;
|
2020-05-10 14:24:55 +08:00
|
|
|
ext_debug(inode, "max_blocks %u, logical block %lu\n", map->m_len,
|
2011-09-10 07:20:51 +08:00
|
|
|
(unsigned long) map->m_lblk);
|
2013-02-18 13:29:59 +08:00
|
|
|
|
|
|
|
/* Lookup extent status tree firstly */
|
2019-08-12 04:32:41 +08:00
|
|
|
if (ext4_es_lookup_extent(inode, iblock, NULL, &es)) {
|
2013-02-18 13:29:59 +08:00
|
|
|
if (ext4_es_is_hole(&es)) {
|
|
|
|
retval = 0;
|
2014-05-13 00:55:07 +08:00
|
|
|
down_read(&EXT4_I(inode)->i_data_sem);
|
2013-02-18 13:29:59 +08:00
|
|
|
goto add_delayed;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2021-10-13 01:19:01 +08:00
|
|
|
* Delayed extent could be allocated by fallocate.
|
|
|
|
* So we need to check it.
|
2013-02-18 13:29:59 +08:00
|
|
|
*/
|
2021-10-13 01:19:01 +08:00
|
|
|
if (ext4_es_is_delayed(&es) && !ext4_es_is_unwritten(&es)) {
|
|
|
|
map_bh(bh, inode->i_sb, invalid_block);
|
|
|
|
set_buffer_new(bh);
|
|
|
|
set_buffer_delay(bh);
|
2013-02-18 13:29:59 +08:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
map->m_pblk = ext4_es_pblock(&es) + iblock - es.es_lblk;
|
|
|
|
retval = es.es_len - (iblock - es.es_lblk);
|
|
|
|
if (retval > map->m_len)
|
|
|
|
retval = map->m_len;
|
|
|
|
map->m_len = retval;
|
|
|
|
if (ext4_es_is_written(&es))
|
|
|
|
map->m_flags |= EXT4_MAP_MAPPED;
|
|
|
|
else if (ext4_es_is_unwritten(&es))
|
|
|
|
map->m_flags |= EXT4_MAP_UNWRITTEN;
|
|
|
|
else
|
2019-04-08 00:24:43 +08:00
|
|
|
BUG();
|
2013-02-18 13:29:59 +08:00
|
|
|
|
2013-03-11 09:01:03 +08:00
|
|
|
#ifdef ES_AGGRESSIVE_TEST
|
|
|
|
ext4_map_blocks_es_recheck(NULL, inode, map, &orig_map, 0);
|
|
|
|
#endif
|
2013-02-18 13:29:59 +08:00
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
|
2011-09-10 07:20:51 +08:00
|
|
|
/*
|
|
|
|
* Try to see if we can get the block without requesting a new
|
|
|
|
* file system block.
|
|
|
|
*/
|
2014-05-13 00:55:07 +08:00
|
|
|
down_read(&EXT4_I(inode)->i_data_sem);
|
2014-11-26 00:41:49 +08:00
|
|
|
if (ext4_has_inline_data(inode))
|
2012-12-11 03:05:57 +08:00
|
|
|
retval = 0;
|
2014-11-26 00:41:49 +08:00
|
|
|
else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
2014-11-26 00:44:37 +08:00
|
|
|
retval = ext4_ext_map_blocks(NULL, inode, map, 0);
|
2011-09-10 07:20:51 +08:00
|
|
|
else
|
2014-11-26 00:44:37 +08:00
|
|
|
retval = ext4_ind_map_blocks(NULL, inode, map, 0);
|
2011-09-10 07:20:51 +08:00
|
|
|
|
2013-02-18 13:29:59 +08:00
|
|
|
add_delayed:
|
2011-09-10 07:20:51 +08:00
|
|
|
if (retval == 0) {
|
2013-02-18 13:28:47 +08:00
|
|
|
int ret;
|
2018-10-02 02:10:39 +08:00
|
|
|
|
2011-09-10 07:20:51 +08:00
|
|
|
/*
|
|
|
|
* XXX: __block_prepare_write() unmaps passed block,
|
|
|
|
* is it OK?
|
|
|
|
*/
|
|
|
|
|
2018-10-02 02:19:37 +08:00
|
|
|
ret = ext4_insert_delayed_block(inode, map->m_lblk);
|
|
|
|
if (ret != 0) {
|
2013-02-18 13:28:47 +08:00
|
|
|
retval = ret;
|
2012-11-09 10:57:32 +08:00
|
|
|
goto out_unlock;
|
2013-02-18 13:28:47 +08:00
|
|
|
}
|
2012-11-09 10:57:32 +08:00
|
|
|
|
2011-09-10 07:20:51 +08:00
|
|
|
map_bh(bh, inode->i_sb, invalid_block);
|
|
|
|
set_buffer_new(bh);
|
|
|
|
set_buffer_delay(bh);
|
2013-02-18 13:28:47 +08:00
|
|
|
} else if (retval > 0) {
|
|
|
|
int ret;
|
2013-08-17 09:22:41 +08:00
|
|
|
unsigned int status;
|
2013-02-18 13:28:47 +08:00
|
|
|
|
2013-07-30 00:51:42 +08:00
|
|
|
if (unlikely(retval != map->m_len)) {
|
|
|
|
ext4_warning(inode->i_sb,
|
|
|
|
"ES len assertion failed for inode "
|
|
|
|
"%lu: retval %d != map->m_len %d",
|
|
|
|
inode->i_ino, retval, map->m_len);
|
|
|
|
WARN_ON(1);
|
2013-03-11 09:01:03 +08:00
|
|
|
}
|
|
|
|
|
2013-02-18 13:28:47 +08:00
|
|
|
status = map->m_flags & EXT4_MAP_UNWRITTEN ?
|
|
|
|
EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN;
|
|
|
|
ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len,
|
|
|
|
map->m_pblk, status);
|
|
|
|
if (ret != 0)
|
|
|
|
retval = ret;
|
2011-09-10 07:20:51 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
out_unlock:
|
|
|
|
up_read((&EXT4_I(inode)->i_data_sem));
|
|
|
|
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
2014-09-02 10:15:30 +08:00
|
|
|
* This is a special get_block_t callback which is used by
|
2009-05-14 12:54:29 +08:00
|
|
|
* ext4_da_write_begin(). It will either return mapped block or
|
|
|
|
* reserve space for a single block.
|
2009-05-13 04:30:27 +08:00
|
|
|
*
|
|
|
|
* For delayed buffer_head we have BH_Mapped, BH_New, BH_Delay set.
|
|
|
|
* We also have b_blocknr = -1 and b_bdev initialized properly
|
|
|
|
*
|
|
|
|
* For unwritten buffer_head we have BH_Mapped, BH_New, BH_Unwritten set.
|
|
|
|
* We also have b_blocknr = physicalblock mapping unwritten extent and b_bdev
|
|
|
|
* initialized properly.
|
2008-07-12 07:27:31 +08:00
|
|
|
*/
|
2012-12-11 03:05:57 +08:00
|
|
|
int ext4_da_get_block_prep(struct inode *inode, sector_t iblock,
|
|
|
|
struct buffer_head *bh, int create)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2010-05-17 08:00:00 +08:00
|
|
|
struct ext4_map_blocks map;
|
2008-07-12 07:27:31 +08:00
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
BUG_ON(create == 0);
|
2010-05-17 08:00:00 +08:00
|
|
|
BUG_ON(bh->b_size != inode->i_sb->s_blocksize);
|
|
|
|
|
|
|
|
map.m_lblk = iblock;
|
|
|
|
map.m_len = 1;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* first, we need to know whether the block is allocated already
|
|
|
|
* preallocated blocks are unmapped but should treated
|
|
|
|
* the same as allocated blocks.
|
|
|
|
*/
|
2011-09-10 07:20:51 +08:00
|
|
|
ret = ext4_da_map_blocks(inode, iblock, &map, bh);
|
|
|
|
if (ret <= 0)
|
2010-05-17 08:00:00 +08:00
|
|
|
return ret;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2010-05-17 08:00:00 +08:00
|
|
|
map_bh(bh, inode->i_sb, map.m_pblk);
|
2016-02-19 13:18:25 +08:00
|
|
|
ext4_update_bh_state(bh, map.m_flags);
|
2010-05-17 08:00:00 +08:00
|
|
|
|
|
|
|
if (buffer_unwritten(bh)) {
|
|
|
|
/* A delayed write to unwritten bh should be marked
|
|
|
|
* new and mapped. Mapped ensures that we don't do
|
|
|
|
* get_block multiple times when we write to the same
|
|
|
|
* offset and new ensures that we do proper zero out
|
|
|
|
* for partial write.
|
|
|
|
*/
|
|
|
|
set_buffer_new(bh);
|
2011-04-11 10:30:07 +08:00
|
|
|
set_buffer_mapped(bh);
|
2010-05-17 08:00:00 +08:00
|
|
|
}
|
|
|
|
return 0;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2009-06-15 05:59:34 +08:00
|
|
|
static int __ext4_journalled_writepage(struct page *page,
|
|
|
|
unsigned int len)
|
|
|
|
{
|
|
|
|
struct address_space *mapping = page->mapping;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
handle_t *handle = NULL;
|
2012-12-11 03:05:57 +08:00
|
|
|
int ret = 0, err = 0;
|
|
|
|
int inline_data = ext4_has_inline_data(inode);
|
|
|
|
struct buffer_head *inode_bh = NULL;
|
2021-12-25 17:09:37 +08:00
|
|
|
loff_t size;
|
2009-06-15 05:59:34 +08:00
|
|
|
|
2010-10-28 09:30:09 +08:00
|
|
|
ClearPageChecked(page);
|
2012-12-11 03:05:57 +08:00
|
|
|
|
|
|
|
if (inline_data) {
|
|
|
|
BUG_ON(page->index != 0);
|
|
|
|
BUG_ON(len > ext4_get_max_inline_size(inode));
|
|
|
|
inode_bh = ext4_journalled_write_inline_data(inode, len, page);
|
|
|
|
if (inode_bh == NULL)
|
|
|
|
goto out;
|
|
|
|
}
|
ext4: fix race between truncate and __ext4_journalled_writepage()
The commit cf108bca465d: "ext4: Invert the locking order of page_lock
and transaction start" caused __ext4_journalled_writepage() to drop
the page lock before the page was written back, as part of changing
the locking order to jbd2_journal_start -> page_lock. However, this
introduced a potential race if there was a truncate racing with the
data=journalled writeback mode.
Fix this by grabbing the page lock after starting the journal handle,
and then checking to see if page had gotten truncated out from under
us.
This fixes a number of different warnings or BUG_ON's when running
xfstests generic/086 in data=journalled mode, including:
jbd2_journal_dirty_metadata: vdc-8: bad jh for block 115643: transaction (ee3fe7
c0, 164), jh->b_transaction ( (null), 0), jh->b_next_transaction ( (null), 0), jlist 0
- and -
kernel BUG at /usr/projects/linux/ext4/fs/jbd2/transaction.c:2200!
...
Call Trace:
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c02b2de5>] __ext4_journalled_invalidatepage+0x10f/0x117
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c027d883>] ? lock_buffer+0x36/0x36
[<c02b2dfa>] ext4_journalled_invalidatepage+0xd/0x22
[<c0229139>] do_invalidatepage+0x22/0x26
[<c0229198>] truncate_inode_page+0x5b/0x85
[<c022934b>] truncate_inode_pages_range+0x156/0x38c
[<c0229592>] truncate_inode_pages+0x11/0x15
[<c022962d>] truncate_pagecache+0x55/0x71
[<c02b913b>] ext4_setattr+0x4a9/0x560
[<c01ca542>] ? current_kernel_time+0x10/0x44
[<c026c4d8>] notify_change+0x1c7/0x2be
[<c0256a00>] do_truncate+0x65/0x85
[<c0226f31>] ? file_ra_state_init+0x12/0x29
- and -
WARNING: CPU: 1 PID: 1331 at /usr/projects/linux/ext4/fs/jbd2/transaction.c:1396
irty_metadata+0x14a/0x1ae()
...
Call Trace:
[<c01b879f>] ? console_unlock+0x3a1/0x3ce
[<c082cbb4>] dump_stack+0x48/0x60
[<c0178b65>] warn_slowpath_common+0x89/0xa0
[<c02ef2cf>] ? jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c0178bef>] warn_slowpath_null+0x14/0x18
[<c02ef2cf>] jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c02d8615>] __ext4_handle_dirty_metadata+0xd4/0x19d
[<c02b2f44>] write_end_fn+0x40/0x53
[<c02b4a16>] ext4_walk_page_buffers+0x4e/0x6a
[<c02b59e7>] ext4_writepage+0x354/0x3b8
[<c02b2f04>] ? mpage_release_unused_pages+0xd4/0xd4
[<c02b1b21>] ? wait_on_buffer+0x2c/0x2c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b5a5b>] __writepage+0x10/0x2e
[<c0225956>] write_cache_pages+0x22d/0x32c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b6ee8>] ext4_writepages+0x102/0x607
[<c019adfe>] ? sched_clock_local+0x10/0x10e
[<c01a8a7c>] ? __lock_is_held+0x2e/0x44
[<c01a8ad5>] ? lock_is_held+0x43/0x51
[<c0226dff>] do_writepages+0x1c/0x29
[<c0276bed>] __writeback_single_inode+0xc3/0x545
[<c0277c07>] writeback_sb_inodes+0x21f/0x36d
...
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2015-06-13 11:45:33 +08:00
|
|
|
/*
|
|
|
|
* We need to release the page lock before we start the
|
|
|
|
* journal, so grab a reference so the page won't disappear
|
|
|
|
* out from under us.
|
|
|
|
*/
|
|
|
|
get_page(page);
|
2009-06-15 05:59:34 +08:00
|
|
|
unlock_page(page);
|
|
|
|
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE,
|
|
|
|
ext4_writepage_trans_blocks(inode));
|
2009-06-15 05:59:34 +08:00
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
ext4: fix race between truncate and __ext4_journalled_writepage()
The commit cf108bca465d: "ext4: Invert the locking order of page_lock
and transaction start" caused __ext4_journalled_writepage() to drop
the page lock before the page was written back, as part of changing
the locking order to jbd2_journal_start -> page_lock. However, this
introduced a potential race if there was a truncate racing with the
data=journalled writeback mode.
Fix this by grabbing the page lock after starting the journal handle,
and then checking to see if page had gotten truncated out from under
us.
This fixes a number of different warnings or BUG_ON's when running
xfstests generic/086 in data=journalled mode, including:
jbd2_journal_dirty_metadata: vdc-8: bad jh for block 115643: transaction (ee3fe7
c0, 164), jh->b_transaction ( (null), 0), jh->b_next_transaction ( (null), 0), jlist 0
- and -
kernel BUG at /usr/projects/linux/ext4/fs/jbd2/transaction.c:2200!
...
Call Trace:
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c02b2de5>] __ext4_journalled_invalidatepage+0x10f/0x117
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c027d883>] ? lock_buffer+0x36/0x36
[<c02b2dfa>] ext4_journalled_invalidatepage+0xd/0x22
[<c0229139>] do_invalidatepage+0x22/0x26
[<c0229198>] truncate_inode_page+0x5b/0x85
[<c022934b>] truncate_inode_pages_range+0x156/0x38c
[<c0229592>] truncate_inode_pages+0x11/0x15
[<c022962d>] truncate_pagecache+0x55/0x71
[<c02b913b>] ext4_setattr+0x4a9/0x560
[<c01ca542>] ? current_kernel_time+0x10/0x44
[<c026c4d8>] notify_change+0x1c7/0x2be
[<c0256a00>] do_truncate+0x65/0x85
[<c0226f31>] ? file_ra_state_init+0x12/0x29
- and -
WARNING: CPU: 1 PID: 1331 at /usr/projects/linux/ext4/fs/jbd2/transaction.c:1396
irty_metadata+0x14a/0x1ae()
...
Call Trace:
[<c01b879f>] ? console_unlock+0x3a1/0x3ce
[<c082cbb4>] dump_stack+0x48/0x60
[<c0178b65>] warn_slowpath_common+0x89/0xa0
[<c02ef2cf>] ? jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c0178bef>] warn_slowpath_null+0x14/0x18
[<c02ef2cf>] jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c02d8615>] __ext4_handle_dirty_metadata+0xd4/0x19d
[<c02b2f44>] write_end_fn+0x40/0x53
[<c02b4a16>] ext4_walk_page_buffers+0x4e/0x6a
[<c02b59e7>] ext4_writepage+0x354/0x3b8
[<c02b2f04>] ? mpage_release_unused_pages+0xd4/0xd4
[<c02b1b21>] ? wait_on_buffer+0x2c/0x2c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b5a5b>] __writepage+0x10/0x2e
[<c0225956>] write_cache_pages+0x22d/0x32c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b6ee8>] ext4_writepages+0x102/0x607
[<c019adfe>] ? sched_clock_local+0x10/0x10e
[<c01a8a7c>] ? __lock_is_held+0x2e/0x44
[<c01a8ad5>] ? lock_is_held+0x43/0x51
[<c0226dff>] do_writepages+0x1c/0x29
[<c0276bed>] __writeback_single_inode+0xc3/0x545
[<c0277c07>] writeback_sb_inodes+0x21f/0x36d
...
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2015-06-13 11:45:33 +08:00
|
|
|
put_page(page);
|
|
|
|
goto out_no_pagelock;
|
2009-06-15 05:59:34 +08:00
|
|
|
}
|
ext4: Fix ext4_should_writeback_data() for no-journal mode
ext4_should_writeback_data() had an incorrect sequence of
tests to determine if it should return 0 or 1: in
particular, even in no-journal mode, 0 was being returned
for a non-regular-file inode.
This meant that, in non-journal mode, we would use
ext4_journalled_aops for directories, symlinks, and other
non-regular files. However, calling journalled aop
callbacks when there is no valid handle, can cause problems.
This would cause a kernel crash with Jan Kara's commit
2d859db3e4 ("ext4: fix data corruption in inodes with
journalled data"), because we now dereference 'handle' in
ext4_journalled_write_end().
I also added BUG_ONs to check for a valid handle in the
obviously journal-only aops callbacks.
I tested this running xfstests with a scratch device in
these modes:
- no-journal
- data=ordered
- data=writeback
- data=journal
All work fine; the data=journal run has many failures and a
crash in xfstests 074, but this is no different from a
vanilla kernel.
Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
2011-08-13 23:25:18 +08:00
|
|
|
BUG_ON(!ext4_handle_valid(handle));
|
|
|
|
|
ext4: fix race between truncate and __ext4_journalled_writepage()
The commit cf108bca465d: "ext4: Invert the locking order of page_lock
and transaction start" caused __ext4_journalled_writepage() to drop
the page lock before the page was written back, as part of changing
the locking order to jbd2_journal_start -> page_lock. However, this
introduced a potential race if there was a truncate racing with the
data=journalled writeback mode.
Fix this by grabbing the page lock after starting the journal handle,
and then checking to see if page had gotten truncated out from under
us.
This fixes a number of different warnings or BUG_ON's when running
xfstests generic/086 in data=journalled mode, including:
jbd2_journal_dirty_metadata: vdc-8: bad jh for block 115643: transaction (ee3fe7
c0, 164), jh->b_transaction ( (null), 0), jh->b_next_transaction ( (null), 0), jlist 0
- and -
kernel BUG at /usr/projects/linux/ext4/fs/jbd2/transaction.c:2200!
...
Call Trace:
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c02b2de5>] __ext4_journalled_invalidatepage+0x10f/0x117
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c027d883>] ? lock_buffer+0x36/0x36
[<c02b2dfa>] ext4_journalled_invalidatepage+0xd/0x22
[<c0229139>] do_invalidatepage+0x22/0x26
[<c0229198>] truncate_inode_page+0x5b/0x85
[<c022934b>] truncate_inode_pages_range+0x156/0x38c
[<c0229592>] truncate_inode_pages+0x11/0x15
[<c022962d>] truncate_pagecache+0x55/0x71
[<c02b913b>] ext4_setattr+0x4a9/0x560
[<c01ca542>] ? current_kernel_time+0x10/0x44
[<c026c4d8>] notify_change+0x1c7/0x2be
[<c0256a00>] do_truncate+0x65/0x85
[<c0226f31>] ? file_ra_state_init+0x12/0x29
- and -
WARNING: CPU: 1 PID: 1331 at /usr/projects/linux/ext4/fs/jbd2/transaction.c:1396
irty_metadata+0x14a/0x1ae()
...
Call Trace:
[<c01b879f>] ? console_unlock+0x3a1/0x3ce
[<c082cbb4>] dump_stack+0x48/0x60
[<c0178b65>] warn_slowpath_common+0x89/0xa0
[<c02ef2cf>] ? jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c0178bef>] warn_slowpath_null+0x14/0x18
[<c02ef2cf>] jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c02d8615>] __ext4_handle_dirty_metadata+0xd4/0x19d
[<c02b2f44>] write_end_fn+0x40/0x53
[<c02b4a16>] ext4_walk_page_buffers+0x4e/0x6a
[<c02b59e7>] ext4_writepage+0x354/0x3b8
[<c02b2f04>] ? mpage_release_unused_pages+0xd4/0xd4
[<c02b1b21>] ? wait_on_buffer+0x2c/0x2c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b5a5b>] __writepage+0x10/0x2e
[<c0225956>] write_cache_pages+0x22d/0x32c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b6ee8>] ext4_writepages+0x102/0x607
[<c019adfe>] ? sched_clock_local+0x10/0x10e
[<c01a8a7c>] ? __lock_is_held+0x2e/0x44
[<c01a8ad5>] ? lock_is_held+0x43/0x51
[<c0226dff>] do_writepages+0x1c/0x29
[<c0276bed>] __writeback_single_inode+0xc3/0x545
[<c0277c07>] writeback_sb_inodes+0x21f/0x36d
...
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2015-06-13 11:45:33 +08:00
|
|
|
lock_page(page);
|
|
|
|
put_page(page);
|
2021-12-25 17:09:37 +08:00
|
|
|
size = i_size_read(inode);
|
|
|
|
if (page->mapping != mapping || page_offset(page) > size) {
|
ext4: fix race between truncate and __ext4_journalled_writepage()
The commit cf108bca465d: "ext4: Invert the locking order of page_lock
and transaction start" caused __ext4_journalled_writepage() to drop
the page lock before the page was written back, as part of changing
the locking order to jbd2_journal_start -> page_lock. However, this
introduced a potential race if there was a truncate racing with the
data=journalled writeback mode.
Fix this by grabbing the page lock after starting the journal handle,
and then checking to see if page had gotten truncated out from under
us.
This fixes a number of different warnings or BUG_ON's when running
xfstests generic/086 in data=journalled mode, including:
jbd2_journal_dirty_metadata: vdc-8: bad jh for block 115643: transaction (ee3fe7
c0, 164), jh->b_transaction ( (null), 0), jh->b_next_transaction ( (null), 0), jlist 0
- and -
kernel BUG at /usr/projects/linux/ext4/fs/jbd2/transaction.c:2200!
...
Call Trace:
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c02b2de5>] __ext4_journalled_invalidatepage+0x10f/0x117
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c027d883>] ? lock_buffer+0x36/0x36
[<c02b2dfa>] ext4_journalled_invalidatepage+0xd/0x22
[<c0229139>] do_invalidatepage+0x22/0x26
[<c0229198>] truncate_inode_page+0x5b/0x85
[<c022934b>] truncate_inode_pages_range+0x156/0x38c
[<c0229592>] truncate_inode_pages+0x11/0x15
[<c022962d>] truncate_pagecache+0x55/0x71
[<c02b913b>] ext4_setattr+0x4a9/0x560
[<c01ca542>] ? current_kernel_time+0x10/0x44
[<c026c4d8>] notify_change+0x1c7/0x2be
[<c0256a00>] do_truncate+0x65/0x85
[<c0226f31>] ? file_ra_state_init+0x12/0x29
- and -
WARNING: CPU: 1 PID: 1331 at /usr/projects/linux/ext4/fs/jbd2/transaction.c:1396
irty_metadata+0x14a/0x1ae()
...
Call Trace:
[<c01b879f>] ? console_unlock+0x3a1/0x3ce
[<c082cbb4>] dump_stack+0x48/0x60
[<c0178b65>] warn_slowpath_common+0x89/0xa0
[<c02ef2cf>] ? jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c0178bef>] warn_slowpath_null+0x14/0x18
[<c02ef2cf>] jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c02d8615>] __ext4_handle_dirty_metadata+0xd4/0x19d
[<c02b2f44>] write_end_fn+0x40/0x53
[<c02b4a16>] ext4_walk_page_buffers+0x4e/0x6a
[<c02b59e7>] ext4_writepage+0x354/0x3b8
[<c02b2f04>] ? mpage_release_unused_pages+0xd4/0xd4
[<c02b1b21>] ? wait_on_buffer+0x2c/0x2c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b5a5b>] __writepage+0x10/0x2e
[<c0225956>] write_cache_pages+0x22d/0x32c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b6ee8>] ext4_writepages+0x102/0x607
[<c019adfe>] ? sched_clock_local+0x10/0x10e
[<c01a8a7c>] ? __lock_is_held+0x2e/0x44
[<c01a8ad5>] ? lock_is_held+0x43/0x51
[<c0226dff>] do_writepages+0x1c/0x29
[<c0276bed>] __writeback_single_inode+0xc3/0x545
[<c0277c07>] writeback_sb_inodes+0x21f/0x36d
...
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2015-06-13 11:45:33 +08:00
|
|
|
/* The page got truncated from under us */
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
ret = 0;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2012-12-11 03:05:57 +08:00
|
|
|
if (inline_data) {
|
2018-07-10 13:07:43 +08:00
|
|
|
ret = ext4_mark_inode_dirty(handle, inode);
|
2012-12-11 03:05:57 +08:00
|
|
|
} else {
|
2021-12-25 17:09:37 +08:00
|
|
|
struct buffer_head *page_bufs = page_buffers(page);
|
|
|
|
|
|
|
|
if (page->index == size >> PAGE_SHIFT)
|
|
|
|
len = size & ~PAGE_MASK;
|
|
|
|
else
|
|
|
|
len = PAGE_SIZE;
|
|
|
|
|
2021-08-16 17:57:04 +08:00
|
|
|
ret = ext4_walk_page_buffers(handle, inode, page_bufs, 0, len,
|
|
|
|
NULL, do_journal_get_write_access);
|
2012-12-11 03:05:57 +08:00
|
|
|
|
2021-08-16 17:57:04 +08:00
|
|
|
err = ext4_walk_page_buffers(handle, inode, page_bufs, 0, len,
|
|
|
|
NULL, write_end_fn);
|
2012-12-11 03:05:57 +08:00
|
|
|
}
|
2020-10-06 08:48:41 +08:00
|
|
|
if (ret == 0)
|
|
|
|
ret = err;
|
2020-10-27 21:27:51 +08:00
|
|
|
err = ext4_jbd2_inode_add_write(handle, inode, page_offset(page), len);
|
2009-06-15 05:59:34 +08:00
|
|
|
if (ret == 0)
|
|
|
|
ret = err;
|
2011-07-26 21:07:11 +08:00
|
|
|
EXT4_I(inode)->i_datasync_tid = handle->h_transaction->t_tid;
|
2009-06-15 05:59:34 +08:00
|
|
|
err = ext4_journal_stop(handle);
|
|
|
|
if (!ret)
|
|
|
|
ret = err;
|
|
|
|
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_JDATA);
|
2009-06-15 05:59:34 +08:00
|
|
|
out:
|
ext4: fix race between truncate and __ext4_journalled_writepage()
The commit cf108bca465d: "ext4: Invert the locking order of page_lock
and transaction start" caused __ext4_journalled_writepage() to drop
the page lock before the page was written back, as part of changing
the locking order to jbd2_journal_start -> page_lock. However, this
introduced a potential race if there was a truncate racing with the
data=journalled writeback mode.
Fix this by grabbing the page lock after starting the journal handle,
and then checking to see if page had gotten truncated out from under
us.
This fixes a number of different warnings or BUG_ON's when running
xfstests generic/086 in data=journalled mode, including:
jbd2_journal_dirty_metadata: vdc-8: bad jh for block 115643: transaction (ee3fe7
c0, 164), jh->b_transaction ( (null), 0), jh->b_next_transaction ( (null), 0), jlist 0
- and -
kernel BUG at /usr/projects/linux/ext4/fs/jbd2/transaction.c:2200!
...
Call Trace:
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c02b2de5>] __ext4_journalled_invalidatepage+0x10f/0x117
[<c02b2ded>] ? __ext4_journalled_invalidatepage+0x117/0x117
[<c027d883>] ? lock_buffer+0x36/0x36
[<c02b2dfa>] ext4_journalled_invalidatepage+0xd/0x22
[<c0229139>] do_invalidatepage+0x22/0x26
[<c0229198>] truncate_inode_page+0x5b/0x85
[<c022934b>] truncate_inode_pages_range+0x156/0x38c
[<c0229592>] truncate_inode_pages+0x11/0x15
[<c022962d>] truncate_pagecache+0x55/0x71
[<c02b913b>] ext4_setattr+0x4a9/0x560
[<c01ca542>] ? current_kernel_time+0x10/0x44
[<c026c4d8>] notify_change+0x1c7/0x2be
[<c0256a00>] do_truncate+0x65/0x85
[<c0226f31>] ? file_ra_state_init+0x12/0x29
- and -
WARNING: CPU: 1 PID: 1331 at /usr/projects/linux/ext4/fs/jbd2/transaction.c:1396
irty_metadata+0x14a/0x1ae()
...
Call Trace:
[<c01b879f>] ? console_unlock+0x3a1/0x3ce
[<c082cbb4>] dump_stack+0x48/0x60
[<c0178b65>] warn_slowpath_common+0x89/0xa0
[<c02ef2cf>] ? jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c0178bef>] warn_slowpath_null+0x14/0x18
[<c02ef2cf>] jbd2_journal_dirty_metadata+0x14a/0x1ae
[<c02d8615>] __ext4_handle_dirty_metadata+0xd4/0x19d
[<c02b2f44>] write_end_fn+0x40/0x53
[<c02b4a16>] ext4_walk_page_buffers+0x4e/0x6a
[<c02b59e7>] ext4_writepage+0x354/0x3b8
[<c02b2f04>] ? mpage_release_unused_pages+0xd4/0xd4
[<c02b1b21>] ? wait_on_buffer+0x2c/0x2c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b5a5b>] __writepage+0x10/0x2e
[<c0225956>] write_cache_pages+0x22d/0x32c
[<c02b5a4b>] ? ext4_writepage+0x3b8/0x3b8
[<c02b6ee8>] ext4_writepages+0x102/0x607
[<c019adfe>] ? sched_clock_local+0x10/0x10e
[<c01a8a7c>] ? __lock_is_held+0x2e/0x44
[<c01a8ad5>] ? lock_is_held+0x43/0x51
[<c0226dff>] do_writepages+0x1c/0x29
[<c0276bed>] __writeback_single_inode+0xc3/0x545
[<c0277c07>] writeback_sb_inodes+0x21f/0x36d
...
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
2015-06-13 11:45:33 +08:00
|
|
|
unlock_page(page);
|
|
|
|
out_no_pagelock:
|
2012-12-11 03:05:57 +08:00
|
|
|
brelse(inode_bh);
|
2009-06-15 05:59:34 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
2009-06-15 05:58:45 +08:00
|
|
|
* Note that we don't need to start a transaction unless we're journaling data
|
|
|
|
* because we should have holes filled from ext4_page_mkwrite(). We even don't
|
|
|
|
* need to file the inode to the transaction's list in ordered mode because if
|
|
|
|
* we are writing back data added by write(), the inode is already there and if
|
2011-03-31 09:57:33 +08:00
|
|
|
* we are writing back data modified via mmap(), no one guarantees in which
|
2009-06-15 05:58:45 +08:00
|
|
|
* transaction the data will hit the disk. In case we are journaling data, we
|
|
|
|
* cannot start transaction directly because transaction start ranks above page
|
|
|
|
* lock so we have to do some magic.
|
|
|
|
*
|
2009-05-14 12:54:29 +08:00
|
|
|
* This function can get called via...
|
2013-06-07 02:00:46 +08:00
|
|
|
* - ext4_writepages after taking page lock (have journal handle)
|
2009-05-14 12:54:29 +08:00
|
|
|
* - journal_submit_inode_data_buffers (no journal handle)
|
2012-07-25 23:12:04 +08:00
|
|
|
* - shrink_page_list via the kswapd/direct reclaim (no journal handle)
|
2009-05-14 12:54:29 +08:00
|
|
|
* - grab_page_cache when doing write_begin (have journal handle)
|
2009-06-15 05:58:45 +08:00
|
|
|
*
|
|
|
|
* We don't do any block allocation in this function. If we have page with
|
|
|
|
* multiple blocks we need to write those buffer_heads that are mapped. This
|
|
|
|
* is important for mmaped based write. So if we do with blocksize 1K
|
|
|
|
* truncate(f, 1024);
|
|
|
|
* a = mmap(f, 0, 4096);
|
|
|
|
* a[0] = 'a';
|
|
|
|
* truncate(f, 4096);
|
|
|
|
* we have in the page first buffer_head mapped via page_mkwrite call back
|
2011-12-05 20:00:34 +08:00
|
|
|
* but other buffer_heads would be unmapped but dirty (dirty done via the
|
2009-06-15 05:58:45 +08:00
|
|
|
* do_wp_page). So writepage should write the first block. If we modify
|
|
|
|
* the mmap area beyond 1024 we will again get a page_fault and the
|
|
|
|
* page_mkwrite callback will do the block allocation and mark the
|
|
|
|
* buffer_heads mapped.
|
|
|
|
*
|
|
|
|
* We redirty the page if we have any buffer_heads that is either delay or
|
|
|
|
* unwritten in the page.
|
|
|
|
*
|
|
|
|
* We can get recursively called as show below.
|
|
|
|
*
|
|
|
|
* ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() ->
|
|
|
|
* ext4_writepage()
|
|
|
|
*
|
|
|
|
* But since we don't do any block allocation we should not deadlock.
|
|
|
|
* Page also have the dirty flag cleared so we don't get recurive page_lock.
|
2008-07-12 07:27:31 +08:00
|
|
|
*/
|
2009-06-15 05:58:45 +08:00
|
|
|
static int ext4_writepage(struct page *page,
|
2009-06-15 05:59:34 +08:00
|
|
|
struct writeback_control *wbc)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2022-02-10 04:21:31 +08:00
|
|
|
struct folio *folio = page_folio(page);
|
2013-01-29 01:55:08 +08:00
|
|
|
int ret = 0;
|
2008-07-12 07:27:31 +08:00
|
|
|
loff_t size;
|
2008-11-05 13:14:04 +08:00
|
|
|
unsigned int len;
|
2010-03-05 05:14:02 +08:00
|
|
|
struct buffer_head *page_bufs = NULL;
|
2008-07-12 07:27:31 +08:00
|
|
|
struct inode *inode = page->mapping->host;
|
2013-01-28 22:30:52 +08:00
|
|
|
struct ext4_io_submit io_submit;
|
2014-05-12 20:12:25 +08:00
|
|
|
bool keep_towrite = false;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) {
|
2022-02-10 04:21:31 +08:00
|
|
|
folio_invalidate(folio, 0, folio_size(folio));
|
|
|
|
folio_unlock(folio);
|
2017-02-05 14:28:48 +08:00
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2011-06-06 21:51:52 +08:00
|
|
|
trace_ext4_writepage(page);
|
2008-07-12 07:27:31 +08:00
|
|
|
size = i_size_read(inode);
|
2019-07-23 00:26:24 +08:00
|
|
|
if (page->index == size >> PAGE_SHIFT &&
|
|
|
|
!ext4_verity_in_progress(inode))
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
len = size & ~PAGE_MASK;
|
2008-07-12 07:27:31 +08:00
|
|
|
else
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
len = PAGE_SIZE;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2022-03-03 22:38:47 +08:00
|
|
|
/* Should never happen but for bugs in other kernel subsystems */
|
|
|
|
if (!page_has_buffers(page)) {
|
|
|
|
ext4_warning_inode(inode,
|
|
|
|
"page %lu does not have buffers attached", page->index);
|
|
|
|
ClearPageDirty(page);
|
|
|
|
unlock_page(page);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-10-28 09:30:09 +08:00
|
|
|
page_bufs = page_buffers(page);
|
|
|
|
/*
|
2013-01-29 10:06:42 +08:00
|
|
|
* We cannot do block allocation or other extent handling in this
|
|
|
|
* function. If there are buffers needing that, we have to redirty
|
|
|
|
* the page. But we may reach here when we do a journal commit via
|
|
|
|
* journal_submit_inode_data_buffers() and in that case we must write
|
|
|
|
* allocated buffers to achieve data=ordered mode guarantees.
|
2015-10-03 22:49:23 +08:00
|
|
|
*
|
|
|
|
* Also, if there is only one buffer per page (the fs block
|
|
|
|
* size == the page size), if one buffer needs block
|
|
|
|
* allocation or needs to modify the extent tree to clear the
|
|
|
|
* unwritten flag, we know that the page can't be written at
|
|
|
|
* all, so we might as well refuse the write immediately.
|
|
|
|
* Unfortunately if the block size != page size, we can't as
|
|
|
|
* easily detect this case using ext4_walk_page_buffers(), but
|
|
|
|
* for the extremely common case, this is an optimization that
|
|
|
|
* skips a useless round trip through ext4_bio_write_page().
|
2010-10-28 09:30:09 +08:00
|
|
|
*/
|
2021-08-16 17:57:04 +08:00
|
|
|
if (ext4_walk_page_buffers(NULL, inode, page_bufs, 0, len, NULL,
|
2012-12-11 03:05:51 +08:00
|
|
|
ext4_bh_delay_or_unwritten)) {
|
2013-01-29 01:55:08 +08:00
|
|
|
redirty_page_for_writepage(wbc, page);
|
2015-10-03 22:49:23 +08:00
|
|
|
if ((current->flags & PF_MEMALLOC) ||
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
(inode->i_sb->s_blocksize == PAGE_SIZE)) {
|
2013-01-29 10:06:42 +08:00
|
|
|
/*
|
|
|
|
* For memory cleaning there's no point in writing only
|
|
|
|
* some buffers. So just bail out. Warn if we came here
|
|
|
|
* from direct reclaim.
|
|
|
|
*/
|
|
|
|
WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD))
|
|
|
|
== PF_MEMALLOC);
|
2008-07-12 07:27:31 +08:00
|
|
|
unlock_page(page);
|
|
|
|
return 0;
|
|
|
|
}
|
2014-05-12 20:12:25 +08:00
|
|
|
keep_towrite = true;
|
2010-10-28 09:30:09 +08:00
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2010-10-28 09:30:09 +08:00
|
|
|
if (PageChecked(page) && ext4_should_journal_data(inode))
|
2009-06-15 05:58:45 +08:00
|
|
|
/*
|
|
|
|
* It's mmapped pagecache. Add buffers and journal it. There
|
|
|
|
* doesn't seem much point in redirtying the page here.
|
|
|
|
*/
|
2009-11-25 00:15:44 +08:00
|
|
|
return __ext4_journalled_writepage(page, len);
|
2009-06-15 05:58:45 +08:00
|
|
|
|
2013-06-04 23:58:58 +08:00
|
|
|
ext4_io_submit_init(&io_submit, wbc);
|
|
|
|
io_submit.io_end = ext4_init_io_end(inode, GFP_NOFS);
|
|
|
|
if (!io_submit.io_end) {
|
|
|
|
redirty_page_for_writepage(wbc, page);
|
|
|
|
unlock_page(page);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2020-12-11 14:54:24 +08:00
|
|
|
ret = ext4_bio_write_page(&io_submit, page, len, keep_towrite);
|
2013-01-28 22:30:52 +08:00
|
|
|
ext4_io_submit(&io_submit);
|
2013-06-04 23:58:58 +08:00
|
|
|
/* Drop io_end reference we got from init */
|
|
|
|
ext4_put_io_end_defer(io_submit.io_end);
|
2008-07-12 07:27:31 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2013-08-17 22:02:33 +08:00
|
|
|
static int mpage_submit_page(struct mpage_da_data *mpd, struct page *page)
|
|
|
|
{
|
|
|
|
int len;
|
2017-05-27 05:45:45 +08:00
|
|
|
loff_t size;
|
2013-08-17 22:02:33 +08:00
|
|
|
int err;
|
|
|
|
|
|
|
|
BUG_ON(page->index != mpd->first_page);
|
2017-05-27 05:45:45 +08:00
|
|
|
clear_page_dirty_for_io(page);
|
|
|
|
/*
|
|
|
|
* We have to be very careful here! Nothing protects writeback path
|
|
|
|
* against i_size changes and the page can be writeably mapped into
|
|
|
|
* page tables. So an application can be growing i_size and writing
|
|
|
|
* data through mmap while writeback runs. clear_page_dirty_for_io()
|
|
|
|
* write-protects our page in page tables and the page cannot get
|
|
|
|
* written to again until we release page lock. So only after
|
|
|
|
* clear_page_dirty_for_io() we are safe to sample i_size for
|
|
|
|
* ext4_bio_write_page() to zero-out tail of the written page. We rely
|
|
|
|
* on the barrier provided by TestClearPageDirty in
|
|
|
|
* clear_page_dirty_for_io() to make sure i_size is really sampled only
|
|
|
|
* after page tables are updated.
|
|
|
|
*/
|
|
|
|
size = i_size_read(mpd->inode);
|
2019-07-23 00:26:24 +08:00
|
|
|
if (page->index == size >> PAGE_SHIFT &&
|
|
|
|
!ext4_verity_in_progress(mpd->inode))
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
len = size & ~PAGE_MASK;
|
2013-08-17 22:02:33 +08:00
|
|
|
else
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
len = PAGE_SIZE;
|
2020-12-11 14:54:24 +08:00
|
|
|
err = ext4_bio_write_page(&mpd->io_submit, page, len, false);
|
2013-08-17 22:02:33 +08:00
|
|
|
if (!err)
|
|
|
|
mpd->wbc->nr_to_write--;
|
|
|
|
mpd->first_page++;
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2020-05-10 14:24:51 +08:00
|
|
|
#define BH_FLAGS (BIT(BH_Unwritten) | BIT(BH_Delay))
|
2013-06-05 01:17:40 +08:00
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
2013-06-05 01:01:11 +08:00
|
|
|
* mballoc gives us at most this number of blocks...
|
|
|
|
* XXX: That seems to be only a limitation of ext4_mb_normalize_request().
|
2013-08-29 02:40:12 +08:00
|
|
|
* The rest of mballoc seems to handle chunks up to full group size.
|
2008-07-12 07:27:31 +08:00
|
|
|
*/
|
2013-06-05 01:01:11 +08:00
|
|
|
#define MAX_WRITEPAGES_EXTENT_LEN 2048
|
2008-08-20 10:15:58 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
|
|
|
* mpage_add_bh_to_extent - try to add bh to extent of blocks to map
|
|
|
|
*
|
|
|
|
* @mpd - extent of blocks
|
|
|
|
* @lblk - logical number of the block in the file
|
2013-08-17 21:57:56 +08:00
|
|
|
* @bh - buffer head we want to add to the extent
|
2013-06-05 01:17:40 +08:00
|
|
|
*
|
2013-08-17 21:57:56 +08:00
|
|
|
* The function is used to collect contig. blocks in the same state. If the
|
|
|
|
* buffer doesn't require mapping for writeback and we haven't started the
|
|
|
|
* extent of buffers to map yet, the function returns 'true' immediately - the
|
|
|
|
* caller can write the buffer right away. Otherwise the function returns true
|
|
|
|
* if the block has been added to the extent, false if the block couldn't be
|
|
|
|
* added.
|
2013-06-05 01:17:40 +08:00
|
|
|
*/
|
2013-08-17 21:57:56 +08:00
|
|
|
static bool mpage_add_bh_to_extent(struct mpage_da_data *mpd, ext4_lblk_t lblk,
|
|
|
|
struct buffer_head *bh)
|
2013-06-05 01:17:40 +08:00
|
|
|
{
|
|
|
|
struct ext4_map_blocks *map = &mpd->map;
|
|
|
|
|
2013-08-17 21:57:56 +08:00
|
|
|
/* Buffer that doesn't need mapping for writeback? */
|
|
|
|
if (!buffer_dirty(bh) || !buffer_mapped(bh) ||
|
|
|
|
(!buffer_delay(bh) && !buffer_unwritten(bh))) {
|
|
|
|
/* So far no extent to map => we write the buffer right away */
|
|
|
|
if (map->m_len == 0)
|
|
|
|
return true;
|
|
|
|
return false;
|
|
|
|
}
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
/* First block in the extent? */
|
|
|
|
if (map->m_len == 0) {
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
/* We cannot map unless handle is started... */
|
|
|
|
if (!mpd->do_map)
|
|
|
|
return false;
|
2013-06-05 01:17:40 +08:00
|
|
|
map->m_lblk = lblk;
|
|
|
|
map->m_len = 1;
|
2013-08-17 21:57:56 +08:00
|
|
|
map->m_flags = bh->b_state & BH_FLAGS;
|
|
|
|
return true;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
|
2013-08-17 21:57:56 +08:00
|
|
|
/* Don't go larger than mballoc is willing to allocate */
|
|
|
|
if (map->m_len >= MAX_WRITEPAGES_EXTENT_LEN)
|
|
|
|
return false;
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/* Can we merge the block to our big extent? */
|
|
|
|
if (lblk == map->m_lblk + map->m_len &&
|
2013-08-17 21:57:56 +08:00
|
|
|
(bh->b_state & BH_FLAGS) == map->m_flags) {
|
2013-06-05 01:17:40 +08:00
|
|
|
map->m_len++;
|
2013-08-17 21:57:56 +08:00
|
|
|
return true;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
2013-08-17 21:57:56 +08:00
|
|
|
return false;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
|
2013-08-17 22:02:33 +08:00
|
|
|
/*
|
|
|
|
* mpage_process_page_bufs - submit page buffers for IO or add them to extent
|
|
|
|
*
|
|
|
|
* @mpd - extent of blocks for mapping
|
|
|
|
* @head - the first buffer in the page
|
|
|
|
* @bh - buffer we should start processing from
|
|
|
|
* @lblk - logical number of the block in the file corresponding to @bh
|
|
|
|
*
|
|
|
|
* Walk through page buffers from @bh upto @head (exclusive) and either submit
|
|
|
|
* the page for IO if all buffers in this page were mapped and there's no
|
|
|
|
* accumulated extent of buffers to map or add buffers in the page to the
|
|
|
|
* extent of buffers to map. The function returns 1 if the caller can continue
|
|
|
|
* by processing the next page, 0 if it should stop adding buffers to the
|
|
|
|
* extent to map because we cannot extend it anymore. It can also return value
|
|
|
|
* < 0 in case of error during IO submission.
|
|
|
|
*/
|
|
|
|
static int mpage_process_page_bufs(struct mpage_da_data *mpd,
|
|
|
|
struct buffer_head *head,
|
|
|
|
struct buffer_head *bh,
|
|
|
|
ext4_lblk_t lblk)
|
2013-06-05 01:17:40 +08:00
|
|
|
{
|
|
|
|
struct inode *inode = mpd->inode;
|
2013-08-17 22:02:33 +08:00
|
|
|
int err;
|
2017-02-28 06:28:32 +08:00
|
|
|
ext4_lblk_t blocks = (i_size_read(inode) + i_blocksize(inode) - 1)
|
2013-06-05 01:17:40 +08:00
|
|
|
>> inode->i_blkbits;
|
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
if (ext4_verity_in_progress(inode))
|
|
|
|
blocks = EXT_MAX_BLOCKS;
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
do {
|
|
|
|
BUG_ON(buffer_locked(bh));
|
|
|
|
|
2013-08-17 21:57:56 +08:00
|
|
|
if (lblk >= blocks || !mpage_add_bh_to_extent(mpd, lblk, bh)) {
|
2013-06-05 01:17:40 +08:00
|
|
|
/* Found extent to map? */
|
|
|
|
if (mpd->map.m_len)
|
2013-08-17 22:02:33 +08:00
|
|
|
return 0;
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
/* Buffer needs mapping and handle is not started? */
|
|
|
|
if (!mpd->do_map)
|
|
|
|
return 0;
|
2013-08-17 21:57:56 +08:00
|
|
|
/* Everything mapped so far and we hit EOF */
|
2013-08-17 22:02:33 +08:00
|
|
|
break;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
} while (lblk++, (bh = bh->b_this_page) != head);
|
2013-08-17 22:02:33 +08:00
|
|
|
/* So far everything mapped? Submit the page for IO. */
|
|
|
|
if (mpd->map.m_len == 0) {
|
|
|
|
err = mpage_submit_page(mpd, head->b_page);
|
|
|
|
if (err < 0)
|
|
|
|
return err;
|
|
|
|
}
|
2020-05-25 16:12:15 +08:00
|
|
|
if (lblk >= blocks) {
|
|
|
|
mpd->scanned_until_end = 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
return 1;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
|
2019-10-16 15:37:09 +08:00
|
|
|
/*
|
|
|
|
* mpage_process_page - update page buffers corresponding to changed extent and
|
|
|
|
* may submit fully mapped page for IO
|
|
|
|
*
|
|
|
|
* @mpd - description of extent to map, on return next extent to map
|
|
|
|
* @m_lblk - logical block mapping.
|
|
|
|
* @m_pblk - corresponding physical mapping.
|
|
|
|
* @map_bh - determines on return whether this page requires any further
|
|
|
|
* mapping or not.
|
|
|
|
* Scan given page buffers corresponding to changed extent and update buffer
|
|
|
|
* state according to new extent state.
|
|
|
|
* We map delalloc buffers to their physical location, clear unwritten bits.
|
|
|
|
* If the given page is not fully mapped, we update @map to the next extent in
|
|
|
|
* the given page that needs mapping & return @map_bh as true.
|
|
|
|
*/
|
|
|
|
static int mpage_process_page(struct mpage_da_data *mpd, struct page *page,
|
|
|
|
ext4_lblk_t *m_lblk, ext4_fsblk_t *m_pblk,
|
|
|
|
bool *map_bh)
|
|
|
|
{
|
|
|
|
struct buffer_head *head, *bh;
|
|
|
|
ext4_io_end_t *io_end = mpd->io_submit.io_end;
|
|
|
|
ext4_lblk_t lblk = *m_lblk;
|
|
|
|
ext4_fsblk_t pblock = *m_pblk;
|
|
|
|
int err = 0;
|
2019-10-16 15:37:10 +08:00
|
|
|
int blkbits = mpd->inode->i_blkbits;
|
|
|
|
ssize_t io_end_size = 0;
|
|
|
|
struct ext4_io_end_vec *io_end_vec = ext4_last_io_end_vec(io_end);
|
2019-10-16 15:37:09 +08:00
|
|
|
|
|
|
|
bh = head = page_buffers(page);
|
|
|
|
do {
|
|
|
|
if (lblk < mpd->map.m_lblk)
|
|
|
|
continue;
|
|
|
|
if (lblk >= mpd->map.m_lblk + mpd->map.m_len) {
|
|
|
|
/*
|
|
|
|
* Buffer after end of mapped extent.
|
|
|
|
* Find next buffer in the page to map.
|
|
|
|
*/
|
|
|
|
mpd->map.m_len = 0;
|
|
|
|
mpd->map.m_flags = 0;
|
2019-10-16 15:37:10 +08:00
|
|
|
io_end_vec->size += io_end_size;
|
2019-10-16 15:37:09 +08:00
|
|
|
|
|
|
|
err = mpage_process_page_bufs(mpd, head, bh, lblk);
|
|
|
|
if (err > 0)
|
|
|
|
err = 0;
|
2019-10-16 15:37:10 +08:00
|
|
|
if (!err && mpd->map.m_len && mpd->map.m_lblk > lblk) {
|
|
|
|
io_end_vec = ext4_alloc_io_end_vec(io_end);
|
2019-11-06 17:38:09 +08:00
|
|
|
if (IS_ERR(io_end_vec)) {
|
|
|
|
err = PTR_ERR(io_end_vec);
|
|
|
|
goto out;
|
|
|
|
}
|
2020-10-08 23:02:48 +08:00
|
|
|
io_end_vec->offset = (loff_t)mpd->map.m_lblk << blkbits;
|
2019-10-16 15:37:10 +08:00
|
|
|
}
|
2019-10-16 15:37:09 +08:00
|
|
|
*map_bh = true;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
if (buffer_delay(bh)) {
|
|
|
|
clear_buffer_delay(bh);
|
|
|
|
bh->b_blocknr = pblock++;
|
|
|
|
}
|
|
|
|
clear_buffer_unwritten(bh);
|
2019-10-16 15:37:10 +08:00
|
|
|
io_end_size += (1 << blkbits);
|
2019-10-16 15:37:09 +08:00
|
|
|
} while (lblk++, (bh = bh->b_this_page) != head);
|
2019-10-16 15:37:10 +08:00
|
|
|
|
|
|
|
io_end_vec->size += io_end_size;
|
2019-10-16 15:37:09 +08:00
|
|
|
*map_bh = false;
|
|
|
|
out:
|
|
|
|
*m_lblk = lblk;
|
|
|
|
*m_pblk = pblock;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
|
|
|
* mpage_map_buffers - update buffers corresponding to changed extent and
|
|
|
|
* submit fully mapped pages for IO
|
|
|
|
*
|
|
|
|
* @mpd - description of extent to map, on return next extent to map
|
|
|
|
*
|
|
|
|
* Scan buffers corresponding to changed extent (we expect corresponding pages
|
|
|
|
* to be already locked) and update buffer state according to new extent state.
|
|
|
|
* We map delalloc buffers to their physical location, clear unwritten bits,
|
2014-04-21 11:45:47 +08:00
|
|
|
* and mark buffers as uninit when we perform writes to unwritten extents
|
2013-06-05 01:17:40 +08:00
|
|
|
* and do extent conversion after IO is finished. If the last page is not fully
|
|
|
|
* mapped, we update @map to the next extent in the last page that needs
|
|
|
|
* mapping. Otherwise we submit the page for IO.
|
|
|
|
*/
|
|
|
|
static int mpage_map_and_submit_buffers(struct mpage_da_data *mpd)
|
|
|
|
{
|
2022-06-05 04:26:01 +08:00
|
|
|
struct folio_batch fbatch;
|
|
|
|
unsigned nr, i;
|
2013-06-05 01:17:40 +08:00
|
|
|
struct inode *inode = mpd->inode;
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
int bpp_bits = PAGE_SHIFT - inode->i_blkbits;
|
2013-06-05 01:17:40 +08:00
|
|
|
pgoff_t start, end;
|
|
|
|
ext4_lblk_t lblk;
|
2019-10-16 15:37:09 +08:00
|
|
|
ext4_fsblk_t pblock;
|
2013-06-05 01:17:40 +08:00
|
|
|
int err;
|
2019-10-16 15:37:09 +08:00
|
|
|
bool map_bh = false;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
start = mpd->map.m_lblk >> bpp_bits;
|
|
|
|
end = (mpd->map.m_lblk + mpd->map.m_len - 1) >> bpp_bits;
|
|
|
|
lblk = start << bpp_bits;
|
|
|
|
pblock = mpd->map.m_pblk;
|
|
|
|
|
2022-06-05 04:26:01 +08:00
|
|
|
folio_batch_init(&fbatch);
|
2013-06-05 01:17:40 +08:00
|
|
|
while (start <= end) {
|
2022-06-05 04:26:01 +08:00
|
|
|
nr = filemap_get_folios(inode->i_mapping, &start, end, &fbatch);
|
|
|
|
if (nr == 0)
|
2013-06-05 01:17:40 +08:00
|
|
|
break;
|
2022-06-05 04:26:01 +08:00
|
|
|
for (i = 0; i < nr; i++) {
|
|
|
|
struct page *page = &fbatch.folios[i]->page;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
2019-10-16 15:37:09 +08:00
|
|
|
err = mpage_process_page(mpd, page, &lblk, &pblock,
|
|
|
|
&map_bh);
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
2019-10-16 15:37:09 +08:00
|
|
|
* If map_bh is true, means page may require further bh
|
|
|
|
* mapping, or maybe the page was submitted for IO.
|
|
|
|
* So we return to call further extent mapping.
|
2013-06-05 01:17:40 +08:00
|
|
|
*/
|
2020-04-20 12:29:18 +08:00
|
|
|
if (err < 0 || map_bh)
|
2019-10-16 15:37:09 +08:00
|
|
|
goto out;
|
2013-06-05 01:17:40 +08:00
|
|
|
/* Page fully mapped - let IO run! */
|
|
|
|
err = mpage_submit_page(mpd, page);
|
2019-10-16 15:37:09 +08:00
|
|
|
if (err < 0)
|
|
|
|
goto out;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
2022-06-05 04:26:01 +08:00
|
|
|
folio_batch_release(&fbatch);
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
/* Extent fully mapped and matches with page boundary. We are done. */
|
|
|
|
mpd->map.m_len = 0;
|
|
|
|
mpd->map.m_flags = 0;
|
|
|
|
return 0;
|
2019-10-16 15:37:09 +08:00
|
|
|
out:
|
2022-06-05 04:26:01 +08:00
|
|
|
folio_batch_release(&fbatch);
|
2019-10-16 15:37:09 +08:00
|
|
|
return err;
|
2013-06-05 01:17:40 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int mpage_map_one_extent(handle_t *handle, struct mpage_da_data *mpd)
|
|
|
|
{
|
|
|
|
struct inode *inode = mpd->inode;
|
|
|
|
struct ext4_map_blocks *map = &mpd->map;
|
|
|
|
int get_blocks_flags;
|
2014-04-21 11:44:47 +08:00
|
|
|
int err, dioread_nolock;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
trace_ext4_da_write_pages_extent(inode, map);
|
|
|
|
/*
|
|
|
|
* Call ext4_map_blocks() to allocate any delayed allocation blocks, or
|
2014-04-21 11:45:47 +08:00
|
|
|
* to convert an unwritten extent to be initialized (in the case
|
2013-06-05 01:17:40 +08:00
|
|
|
* where we have written into one or more preallocated blocks). It is
|
|
|
|
* possible that we're going to need more metadata blocks than
|
|
|
|
* previously reserved. However we must not fail because we're in
|
|
|
|
* writeback and there is nothing we can do about it so it might result
|
|
|
|
* in data loss. So use reserved blocks to allocate metadata if
|
|
|
|
* possible.
|
|
|
|
*
|
2014-09-05 06:08:22 +08:00
|
|
|
* We pass in the magic EXT4_GET_BLOCKS_DELALLOC_RESERVE if
|
|
|
|
* the blocks in question are delalloc blocks. This indicates
|
|
|
|
* that the blocks and quotas has already been checked when
|
|
|
|
* the data was copied into the page cache.
|
2013-06-05 01:17:40 +08:00
|
|
|
*/
|
|
|
|
get_blocks_flags = EXT4_GET_BLOCKS_CREATE |
|
2016-04-24 12:56:08 +08:00
|
|
|
EXT4_GET_BLOCKS_METADATA_NOFAIL |
|
|
|
|
EXT4_GET_BLOCKS_IO_SUBMIT;
|
2014-04-21 11:44:47 +08:00
|
|
|
dioread_nolock = ext4_should_dioread_nolock(inode);
|
|
|
|
if (dioread_nolock)
|
2013-06-05 01:17:40 +08:00
|
|
|
get_blocks_flags |= EXT4_GET_BLOCKS_IO_CREATE_EXT;
|
2020-05-10 14:24:51 +08:00
|
|
|
if (map->m_flags & BIT(BH_Delay))
|
2013-06-05 01:17:40 +08:00
|
|
|
get_blocks_flags |= EXT4_GET_BLOCKS_DELALLOC_RESERVE;
|
|
|
|
|
|
|
|
err = ext4_map_blocks(handle, inode, map, get_blocks_flags);
|
|
|
|
if (err < 0)
|
|
|
|
return err;
|
2014-04-21 11:44:47 +08:00
|
|
|
if (dioread_nolock && (map->m_flags & EXT4_MAP_UNWRITTEN)) {
|
2013-06-05 01:21:11 +08:00
|
|
|
if (!mpd->io_submit.io_end->handle &&
|
|
|
|
ext4_handle_valid(handle)) {
|
|
|
|
mpd->io_submit.io_end->handle = handle->h_rsv_handle;
|
|
|
|
handle->h_rsv_handle = NULL;
|
|
|
|
}
|
2013-06-05 01:19:34 +08:00
|
|
|
ext4_set_io_unwritten_flag(inode, mpd->io_submit.io_end);
|
2013-06-05 01:21:11 +08:00
|
|
|
}
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
BUG_ON(map->m_len == 0);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* mpage_map_and_submit_extent - map extent starting at mpd->lblk of length
|
|
|
|
* mpd->len and submit pages underlying it for IO
|
|
|
|
*
|
|
|
|
* @handle - handle for journal operations
|
|
|
|
* @mpd - extent to map
|
2013-10-16 20:26:08 +08:00
|
|
|
* @give_up_on_write - we set this to true iff there is a fatal error and there
|
|
|
|
* is no hope of writing the data. The caller should discard
|
|
|
|
* dirty pages to avoid infinite loops.
|
2013-06-05 01:17:40 +08:00
|
|
|
*
|
|
|
|
* The function maps extent starting at mpd->lblk of length mpd->len. If it is
|
|
|
|
* delayed, blocks are allocated, if it is unwritten, we may need to convert
|
|
|
|
* them to initialized or split the described range from larger unwritten
|
|
|
|
* extent. Note that we need not map all the described range since allocation
|
|
|
|
* can return less blocks or the range is covered by more unwritten extents. We
|
|
|
|
* cannot map more because we are limited by reserved transaction credits. On
|
|
|
|
* the other hand we always make sure that the last touched page is fully
|
|
|
|
* mapped so that it can be written out (and thus forward progress is
|
|
|
|
* guaranteed). After mapping we submit all mapped pages for IO.
|
|
|
|
*/
|
|
|
|
static int mpage_map_and_submit_extent(handle_t *handle,
|
2013-07-01 20:12:40 +08:00
|
|
|
struct mpage_da_data *mpd,
|
|
|
|
bool *give_up_on_write)
|
2013-06-05 01:17:40 +08:00
|
|
|
{
|
|
|
|
struct inode *inode = mpd->inode;
|
|
|
|
struct ext4_map_blocks *map = &mpd->map;
|
|
|
|
int err;
|
|
|
|
loff_t disksize;
|
2014-08-28 06:40:03 +08:00
|
|
|
int progress = 0;
|
2019-10-16 15:37:10 +08:00
|
|
|
ext4_io_end_t *io_end = mpd->io_submit.io_end;
|
2019-11-06 17:38:09 +08:00
|
|
|
struct ext4_io_end_vec *io_end_vec;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
2019-11-06 17:38:09 +08:00
|
|
|
io_end_vec = ext4_alloc_io_end_vec(io_end);
|
|
|
|
if (IS_ERR(io_end_vec))
|
|
|
|
return PTR_ERR(io_end_vec);
|
2019-10-16 15:37:10 +08:00
|
|
|
io_end_vec->offset = ((loff_t)map->m_lblk) << inode->i_blkbits;
|
2013-07-06 09:57:22 +08:00
|
|
|
do {
|
2013-06-05 01:17:40 +08:00
|
|
|
err = mpage_map_one_extent(handle, mpd);
|
|
|
|
if (err < 0) {
|
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (ext4_forced_shutdown(EXT4_SB(sb)) ||
|
2020-11-06 11:59:09 +08:00
|
|
|
ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED))
|
2013-07-01 20:12:40 +08:00
|
|
|
goto invalidate_dirty_pages;
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
2013-07-01 20:12:40 +08:00
|
|
|
* Let the uper layers retry transient errors.
|
|
|
|
* In the case of ENOSPC, if ext4_count_free_blocks()
|
|
|
|
* is non-zero, a commit should free up blocks.
|
2013-06-05 01:17:40 +08:00
|
|
|
*/
|
2013-07-01 20:12:40 +08:00
|
|
|
if ((err == -ENOMEM) ||
|
2014-08-28 06:40:03 +08:00
|
|
|
(err == -ENOSPC && ext4_count_free_clusters(sb))) {
|
|
|
|
if (progress)
|
|
|
|
goto update_disksize;
|
2013-07-01 20:12:40 +08:00
|
|
|
return err;
|
2014-08-28 06:40:03 +08:00
|
|
|
}
|
2013-07-01 20:12:40 +08:00
|
|
|
ext4_msg(sb, KERN_CRIT,
|
|
|
|
"Delayed block allocation failed for "
|
|
|
|
"inode %lu at logical offset %llu with"
|
|
|
|
" max blocks %u with error %d",
|
|
|
|
inode->i_ino,
|
|
|
|
(unsigned long long)map->m_lblk,
|
|
|
|
(unsigned)map->m_len, -err);
|
|
|
|
ext4_msg(sb, KERN_CRIT,
|
|
|
|
"This should not happen!! Data will "
|
|
|
|
"be lost\n");
|
|
|
|
if (err == -ENOSPC)
|
|
|
|
ext4_print_free_blocks(inode);
|
|
|
|
invalidate_dirty_pages:
|
|
|
|
*give_up_on_write = true;
|
2013-06-05 01:17:40 +08:00
|
|
|
return err;
|
|
|
|
}
|
2014-08-28 06:40:03 +08:00
|
|
|
progress = 1;
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
|
|
|
* Update buffer state, submit mapped pages, and get us new
|
|
|
|
* extent to map
|
|
|
|
*/
|
|
|
|
err = mpage_map_and_submit_buffers(mpd);
|
|
|
|
if (err < 0)
|
2014-08-28 06:40:03 +08:00
|
|
|
goto update_disksize;
|
2013-07-06 09:57:22 +08:00
|
|
|
} while (map->m_len);
|
2013-06-05 01:17:40 +08:00
|
|
|
|
2014-08-28 06:40:03 +08:00
|
|
|
update_disksize:
|
2014-04-11 22:35:17 +08:00
|
|
|
/*
|
|
|
|
* Update on-disk size after IO is submitted. Races with
|
|
|
|
* truncate are avoided by checking i_size under i_data_sem.
|
|
|
|
*/
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
disksize = ((loff_t)mpd->first_page) << PAGE_SHIFT;
|
2020-02-07 22:29:11 +08:00
|
|
|
if (disksize > READ_ONCE(EXT4_I(inode)->i_disksize)) {
|
2013-06-05 01:17:40 +08:00
|
|
|
int err2;
|
2014-04-11 22:35:17 +08:00
|
|
|
loff_t i_size;
|
|
|
|
|
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
|
|
|
i_size = i_size_read(inode);
|
|
|
|
if (disksize > i_size)
|
|
|
|
disksize = i_size;
|
|
|
|
if (disksize > EXT4_I(inode)->i_disksize)
|
|
|
|
EXT4_I(inode)->i_disksize = disksize;
|
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
2017-01-12 11:14:49 +08:00
|
|
|
err2 = ext4_mark_inode_dirty(handle, inode);
|
2019-11-20 10:54:15 +08:00
|
|
|
if (err2) {
|
2020-03-29 07:33:43 +08:00
|
|
|
ext4_error_err(inode->i_sb, -err2,
|
|
|
|
"Failed to mark inode %lu dirty",
|
|
|
|
inode->i_ino);
|
2019-11-20 10:54:15 +08:00
|
|
|
}
|
2013-06-05 01:17:40 +08:00
|
|
|
if (!err)
|
|
|
|
err = err2;
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2013-06-05 01:01:11 +08:00
|
|
|
/*
|
|
|
|
* Calculate the total number of credits to reserve for one writepages
|
2013-06-07 02:00:46 +08:00
|
|
|
* iteration. This is called from ext4_writepages(). We map an extent of
|
2013-08-29 02:40:12 +08:00
|
|
|
* up to MAX_WRITEPAGES_EXTENT_LEN blocks and then we go on and finish mapping
|
2013-06-05 01:01:11 +08:00
|
|
|
* the last partial page. So in total we can map MAX_WRITEPAGES_EXTENT_LEN +
|
|
|
|
* bpp - 1 blocks in bpp different extents.
|
|
|
|
*/
|
2008-08-20 10:15:58 +08:00
|
|
|
static int ext4_da_writepages_trans_blocks(struct inode *inode)
|
|
|
|
{
|
2013-06-05 01:01:11 +08:00
|
|
|
int bpp = ext4_journal_blocks_per_page(inode);
|
2008-08-20 10:15:58 +08:00
|
|
|
|
2013-06-05 01:01:11 +08:00
|
|
|
return ext4_meta_trans_blocks(inode,
|
|
|
|
MAX_WRITEPAGES_EXTENT_LEN + bpp - 1, bpp);
|
2008-08-20 10:15:58 +08:00
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2010-05-17 06:00:00 +08:00
|
|
|
/*
|
2013-06-05 01:17:40 +08:00
|
|
|
* mpage_prepare_extent_to_map - find & lock contiguous range of dirty pages
|
|
|
|
* and underlying extent to map
|
|
|
|
*
|
|
|
|
* @mpd - where to look for pages
|
|
|
|
*
|
|
|
|
* Walk dirty pages in the mapping. If they are fully mapped, submit them for
|
|
|
|
* IO immediately. When we find a page which isn't mapped we start accumulating
|
|
|
|
* extent of buffers underlying these pages that needs mapping (formed by
|
|
|
|
* either delayed or unwritten buffers). We also lock the pages containing
|
|
|
|
* these buffers. The extent found is returned in @mpd structure (starting at
|
|
|
|
* mpd->lblk with length mpd->len blocks).
|
|
|
|
*
|
|
|
|
* Note that this function can attach bios to one io_end structure which are
|
|
|
|
* neither logically nor physically contiguous. Although it may seem as an
|
|
|
|
* unnecessary complication, it is actually inevitable in blocksize < pagesize
|
|
|
|
* case as we need to track IO to all buffers underlying a page in one io_end.
|
2010-05-17 06:00:00 +08:00
|
|
|
*/
|
2013-06-05 01:17:40 +08:00
|
|
|
static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd)
|
2010-05-17 06:00:00 +08:00
|
|
|
{
|
2013-06-05 01:17:40 +08:00
|
|
|
struct address_space *mapping = mpd->inode->i_mapping;
|
|
|
|
struct pagevec pvec;
|
|
|
|
unsigned int nr_pages;
|
2013-10-18 06:56:16 +08:00
|
|
|
long left = mpd->wbc->nr_to_write;
|
2013-06-05 01:17:40 +08:00
|
|
|
pgoff_t index = mpd->first_page;
|
|
|
|
pgoff_t end = mpd->last_page;
|
2017-12-06 06:30:38 +08:00
|
|
|
xa_mark_t tag;
|
2013-06-05 01:17:40 +08:00
|
|
|
int i, err = 0;
|
|
|
|
int blkbits = mpd->inode->i_blkbits;
|
|
|
|
ext4_lblk_t lblk;
|
|
|
|
struct buffer_head *head;
|
2010-05-17 06:00:00 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
if (mpd->wbc->sync_mode == WB_SYNC_ALL || mpd->wbc->tagged_writepages)
|
ext4: implement writeback livelock avoidance using page tagging
This is analogous to Jan Kara's commit,
f446daaea9d4a420d16c606f755f3689dcb2d0ce
mm: implement writeback livelock avoidance using page tagging
but since we forked write_cache_pages, we need to reimplement
it there (and in ext4_da_writepages, since range_cyclic handling
was moved to there)
If you start a large buffered IO to a file, and then set
fsync after it, you'll find that fsync does not complete
until the other IO stops.
If you continue re-dirtying the file (say, putting dd
with conv=notrunc in a loop), when fsync finally completes
(after all IO is done), it reports via tracing that
it has written many more pages than the file contains;
in other words it has synced and re-synced pages in
the file multiple times.
This then leads to problems with our writeback_index
update, since it advances it by pages written, and
essentially sets writeback_index off the end of the
file...
With the following patch, we only sync as much as was
dirty at the time of the sync.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2010-10-28 09:30:13 +08:00
|
|
|
tag = PAGECACHE_TAG_TOWRITE;
|
|
|
|
else
|
|
|
|
tag = PAGECACHE_TAG_DIRTY;
|
|
|
|
|
2017-11-16 09:37:52 +08:00
|
|
|
pagevec_init(&pvec);
|
2013-06-05 01:17:40 +08:00
|
|
|
mpd->map.m_len = 0;
|
|
|
|
mpd->next_page = index;
|
2011-02-27 03:07:37 +08:00
|
|
|
while (index <= end) {
|
2017-11-16 09:34:44 +08:00
|
|
|
nr_pages = pagevec_lookup_range_tag(&pvec, mapping, &index, end,
|
2017-11-16 09:35:19 +08:00
|
|
|
tag);
|
2010-05-17 06:00:00 +08:00
|
|
|
if (nr_pages == 0)
|
2020-05-25 16:12:15 +08:00
|
|
|
break;
|
2010-05-17 06:00:00 +08:00
|
|
|
|
|
|
|
for (i = 0; i < nr_pages; i++) {
|
|
|
|
struct page *page = pvec.pages[i];
|
|
|
|
|
2013-10-18 06:56:16 +08:00
|
|
|
/*
|
|
|
|
* Accumulated enough dirty pages? This doesn't apply
|
|
|
|
* to WB_SYNC_ALL mode. For integrity sync we have to
|
|
|
|
* keep going because someone may be concurrently
|
|
|
|
* dirtying pages, and we might have synced a lot of
|
|
|
|
* newly appeared dirty pages, but have not synced all
|
|
|
|
* of the old dirty pages.
|
|
|
|
*/
|
|
|
|
if (mpd->wbc->sync_mode == WB_SYNC_NONE && left <= 0)
|
|
|
|
goto out;
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/* If we can't merge this page, we are done. */
|
|
|
|
if (mpd->map.m_len > 0 && mpd->next_page != page->index)
|
|
|
|
goto out;
|
2011-02-27 03:09:14 +08:00
|
|
|
|
2010-05-17 06:00:00 +08:00
|
|
|
lock_page(page);
|
|
|
|
/*
|
2013-06-05 01:17:40 +08:00
|
|
|
* If the page is no longer dirty, or its mapping no
|
|
|
|
* longer corresponds to inode we are writing (which
|
|
|
|
* means it has been truncated or invalidated), or the
|
|
|
|
* page is already under writeback and we are not doing
|
|
|
|
* a data integrity writeback, skip the page
|
2010-05-17 06:00:00 +08:00
|
|
|
*/
|
2011-02-27 03:07:37 +08:00
|
|
|
if (!PageDirty(page) ||
|
|
|
|
(PageWriteback(page) &&
|
2013-06-05 01:17:40 +08:00
|
|
|
(mpd->wbc->sync_mode == WB_SYNC_NONE)) ||
|
2011-02-27 03:07:37 +08:00
|
|
|
unlikely(page->mapping != mapping)) {
|
2010-05-17 06:00:00 +08:00
|
|
|
unlock_page(page);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2011-05-19 01:53:20 +08:00
|
|
|
wait_on_page_writeback(page);
|
2010-05-17 06:00:00 +08:00
|
|
|
BUG_ON(PageWriteback(page));
|
|
|
|
|
2022-03-03 22:38:47 +08:00
|
|
|
/*
|
|
|
|
* Should never happen but for buggy code in
|
|
|
|
* other subsystems that call
|
|
|
|
* set_page_dirty() without properly warning
|
|
|
|
* the file system first. See [1] for more
|
|
|
|
* information.
|
|
|
|
*
|
|
|
|
* [1] https://lore.kernel.org/linux-mm/20180103100430.GE4911@quack2.suse.cz
|
|
|
|
*/
|
|
|
|
if (!page_has_buffers(page)) {
|
|
|
|
ext4_warning_inode(mpd->inode, "page %lu does not have buffers attached", page->index);
|
|
|
|
ClearPageDirty(page);
|
|
|
|
unlock_page(page);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
if (mpd->map.m_len == 0)
|
2011-02-27 03:07:31 +08:00
|
|
|
mpd->first_page = page->index;
|
|
|
|
mpd->next_page = page->index + 1;
|
2013-01-29 01:55:08 +08:00
|
|
|
/* Add all dirty buffers to mpd */
|
2013-06-05 01:17:40 +08:00
|
|
|
lblk = ((ext4_lblk_t)page->index) <<
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
(PAGE_SHIFT - blkbits);
|
2013-01-29 01:55:08 +08:00
|
|
|
head = page_buffers(page);
|
2013-08-17 22:02:33 +08:00
|
|
|
err = mpage_process_page_bufs(mpd, head, head, lblk);
|
|
|
|
if (err <= 0)
|
2013-06-05 01:17:40 +08:00
|
|
|
goto out;
|
2013-08-17 22:02:33 +08:00
|
|
|
err = 0;
|
2013-10-18 06:56:16 +08:00
|
|
|
left--;
|
2010-05-17 06:00:00 +08:00
|
|
|
}
|
|
|
|
pagevec_release(&pvec);
|
|
|
|
cond_resched();
|
|
|
|
}
|
2020-05-25 16:12:15 +08:00
|
|
|
mpd->scanned_until_end = 1;
|
2011-02-27 03:07:37 +08:00
|
|
|
return 0;
|
2011-02-27 03:07:31 +08:00
|
|
|
out:
|
|
|
|
pagevec_release(&pvec);
|
2013-06-05 01:17:40 +08:00
|
|
|
return err;
|
2010-05-17 06:00:00 +08:00
|
|
|
}
|
|
|
|
|
2013-06-07 02:00:46 +08:00
|
|
|
static int ext4_writepages(struct address_space *mapping,
|
|
|
|
struct writeback_control *wbc)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2013-06-05 01:17:40 +08:00
|
|
|
pgoff_t writeback_index = 0;
|
|
|
|
long nr_to_write = wbc->nr_to_write;
|
2008-10-16 22:10:36 +08:00
|
|
|
int range_whole = 0;
|
2013-06-05 01:17:40 +08:00
|
|
|
int cycled = 1;
|
2008-07-12 07:27:31 +08:00
|
|
|
handle_t *handle = NULL;
|
2008-09-09 11:05:34 +08:00
|
|
|
struct mpage_da_data mpd;
|
2008-08-19 06:00:57 +08:00
|
|
|
struct inode *inode = mapping->host;
|
2013-06-05 01:21:11 +08:00
|
|
|
int needed_blocks, rsv_blocks = 0, ret = 0;
|
2008-08-19 06:00:57 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb);
|
2011-10-18 22:55:51 +08:00
|
|
|
struct blk_plug plug;
|
2013-07-01 20:12:40 +08:00
|
|
|
bool give_up_on_write = false;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_down_read(&sbi->s_writepages_rwsem);
|
2013-06-07 02:00:46 +08:00
|
|
|
trace_ext4_writepages(inode, wbc);
|
2009-01-04 09:03:21 +08:00
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
|
|
|
* No pages to write? This is mainly a kludge to avoid starting
|
|
|
|
* a transaction for special inodes like journal inode on last iput()
|
|
|
|
* because that could violate lock ordering on umount
|
|
|
|
*/
|
2008-08-20 09:55:02 +08:00
|
|
|
if (!mapping->nrpages || !mapping_tagged(mapping, PAGECACHE_TAG_DIRTY))
|
2013-10-30 19:27:16 +08:00
|
|
|
goto out_writepages;
|
2008-11-05 22:22:24 +08:00
|
|
|
|
2013-06-07 02:00:46 +08:00
|
|
|
if (ext4_should_journal_data(inode)) {
|
2018-03-26 13:32:50 +08:00
|
|
|
ret = generic_writepages(mapping, wbc);
|
2013-10-30 19:27:16 +08:00
|
|
|
goto out_writepages;
|
2013-06-07 02:00:46 +08:00
|
|
|
}
|
|
|
|
|
2008-11-05 22:22:24 +08:00
|
|
|
/*
|
|
|
|
* If the filesystem has aborted, it is read-only, so return
|
|
|
|
* right away instead of dumping stack traces later on that
|
|
|
|
* will obscure the real source of the problem. We test
|
2017-11-28 05:05:09 +08:00
|
|
|
* EXT4_MF_FS_ABORTED instead of sb->s_flag's SB_RDONLY because
|
2008-11-05 22:22:24 +08:00
|
|
|
* the latter could be true if the filesystem is mounted
|
2013-06-07 02:00:46 +08:00
|
|
|
* read-only, and in that case, ext4_writepages should
|
2008-11-05 22:22:24 +08:00
|
|
|
* *never* be called, so if that ever happens, we would want
|
|
|
|
* the stack trace.
|
|
|
|
*/
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(mapping->host->i_sb)) ||
|
2020-11-06 11:59:09 +08:00
|
|
|
ext4_test_mount_flag(inode->i_sb, EXT4_MF_FS_ABORTED))) {
|
2013-10-30 19:27:16 +08:00
|
|
|
ret = -EROFS;
|
|
|
|
goto out_writepages;
|
|
|
|
}
|
2008-11-05 22:22:24 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/*
|
|
|
|
* If we have inline data and arrive here, it means that
|
|
|
|
* we will soon create the block for the 1st page, so
|
|
|
|
* we'd better clear the inline data here.
|
|
|
|
*/
|
|
|
|
if (ext4_has_inline_data(inode)) {
|
|
|
|
/* Just inode will be modified... */
|
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
|
|
|
goto out_writepages;
|
|
|
|
}
|
|
|
|
BUG_ON(ext4_test_inode_state(inode,
|
|
|
|
EXT4_STATE_MAY_INLINE_DATA));
|
|
|
|
ext4_destroy_inline_data(handle, inode);
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
}
|
|
|
|
|
2019-08-12 04:27:41 +08:00
|
|
|
if (ext4_should_dioread_nolock(inode)) {
|
|
|
|
/*
|
|
|
|
* We may need to convert up to one extent per block in
|
|
|
|
* the page and we may dirty the inode.
|
|
|
|
*/
|
|
|
|
rsv_blocks = 1 + ext4_chunk_trans_blocks(inode,
|
|
|
|
PAGE_SIZE >> inode->i_blkbits);
|
|
|
|
}
|
|
|
|
|
2008-10-16 22:10:36 +08:00
|
|
|
if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX)
|
|
|
|
range_whole = 1;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2009-02-14 23:42:58 +08:00
|
|
|
if (wbc->range_cyclic) {
|
2013-06-05 01:17:40 +08:00
|
|
|
writeback_index = mapping->writeback_index;
|
|
|
|
if (writeback_index)
|
2009-02-14 23:42:58 +08:00
|
|
|
cycled = 0;
|
2013-06-05 01:17:40 +08:00
|
|
|
mpd.first_page = writeback_index;
|
|
|
|
mpd.last_page = -1;
|
ext4: implement writeback livelock avoidance using page tagging
This is analogous to Jan Kara's commit,
f446daaea9d4a420d16c606f755f3689dcb2d0ce
mm: implement writeback livelock avoidance using page tagging
but since we forked write_cache_pages, we need to reimplement
it there (and in ext4_da_writepages, since range_cyclic handling
was moved to there)
If you start a large buffered IO to a file, and then set
fsync after it, you'll find that fsync does not complete
until the other IO stops.
If you continue re-dirtying the file (say, putting dd
with conv=notrunc in a loop), when fsync finally completes
(after all IO is done), it reports via tracing that
it has written many more pages than the file contains;
in other words it has synced and re-synced pages in
the file multiple times.
This then leads to problems with our writeback_index
update, since it advances it by pages written, and
essentially sets writeback_index off the end of the
file...
With the following patch, we only sync as much as was
dirty at the time of the sync.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2010-10-28 09:30:13 +08:00
|
|
|
} else {
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
mpd.first_page = wbc->range_start >> PAGE_SHIFT;
|
|
|
|
mpd.last_page = wbc->range_end >> PAGE_SHIFT;
|
ext4: implement writeback livelock avoidance using page tagging
This is analogous to Jan Kara's commit,
f446daaea9d4a420d16c606f755f3689dcb2d0ce
mm: implement writeback livelock avoidance using page tagging
but since we forked write_cache_pages, we need to reimplement
it there (and in ext4_da_writepages, since range_cyclic handling
was moved to there)
If you start a large buffered IO to a file, and then set
fsync after it, you'll find that fsync does not complete
until the other IO stops.
If you continue re-dirtying the file (say, putting dd
with conv=notrunc in a loop), when fsync finally completes
(after all IO is done), it reports via tracing that
it has written many more pages than the file contains;
in other words it has synced and re-synced pages in
the file multiple times.
This then leads to problems with our writeback_index
update, since it advances it by pages written, and
essentially sets writeback_index off the end of the
file...
With the following patch, we only sync as much as was
dirty at the time of the sync.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2010-10-28 09:30:13 +08:00
|
|
|
}
|
2008-08-20 09:55:02 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
mpd.inode = inode;
|
|
|
|
mpd.wbc = wbc;
|
|
|
|
ext4_io_submit_init(&mpd.io_submit, wbc);
|
2009-02-14 23:42:58 +08:00
|
|
|
retry:
|
2010-06-07 00:38:15 +08:00
|
|
|
if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages)
|
2013-06-05 01:17:40 +08:00
|
|
|
tag_pages_for_writeback(mapping, mpd.first_page, mpd.last_page);
|
2011-10-18 22:55:51 +08:00
|
|
|
blk_start_plug(&plug);
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* First writeback pages that don't need mapping - we can avoid
|
|
|
|
* starting a transaction unnecessarily and also avoid being blocked
|
|
|
|
* in the block layer on device congestion while having transaction
|
|
|
|
* started.
|
|
|
|
*/
|
|
|
|
mpd.do_map = 0;
|
2020-05-25 16:12:15 +08:00
|
|
|
mpd.scanned_until_end = 0;
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
mpd.io_submit.io_end = ext4_init_io_end(inode, GFP_KERNEL);
|
|
|
|
if (!mpd.io_submit.io_end) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto unplug;
|
|
|
|
}
|
|
|
|
ret = mpage_prepare_extent_to_map(&mpd);
|
ext4: unlock unused_pages timely when doing writeback
In mpage_add_bh_to_extent(), when accumulated extents length is greater
than MAX_WRITEPAGES_EXTENT_LEN or buffer head's b_stat is not equal, we
will not continue to search unmapped area for this page, but note this
page is locked, and will only be unlocked in mpage_release_unused_pages()
after ext4_io_submit, if io also is throttled by blk-throttle or similar
io qos, we will hold this page locked for a while, it's unnecessary.
I think the best fix is to refactor mpage_add_bh_to_extent() to let it
return some hints whether to unlock this page, but given that we will
improve dioread_nolock later, we can let it done later, so currently
the simple fix would just call mpage_release_unused_pages() before
ext4_io_submit().
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-02-11 12:53:21 +08:00
|
|
|
/* Unlock pages we didn't use */
|
|
|
|
mpage_release_unused_pages(&mpd, false);
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
/* Submit prepared bio */
|
|
|
|
ext4_io_submit(&mpd.io_submit);
|
|
|
|
ext4_put_io_end_defer(mpd.io_submit.io_end);
|
|
|
|
mpd.io_submit.io_end = NULL;
|
|
|
|
if (ret < 0)
|
|
|
|
goto unplug;
|
|
|
|
|
2020-05-25 16:12:15 +08:00
|
|
|
while (!mpd.scanned_until_end && wbc->nr_to_write > 0) {
|
2013-06-05 01:17:40 +08:00
|
|
|
/* For each extent of pages we use new io_end */
|
|
|
|
mpd.io_submit.io_end = ext4_init_io_end(inode, GFP_KERNEL);
|
|
|
|
if (!mpd.io_submit.io_end) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
break;
|
|
|
|
}
|
2008-08-20 09:55:02 +08:00
|
|
|
|
|
|
|
/*
|
2013-06-05 01:17:40 +08:00
|
|
|
* We have two constraints: We find one extent to map and we
|
|
|
|
* must always write out whole page (makes a difference when
|
|
|
|
* blocksize < pagesize) so that we don't block on IO when we
|
|
|
|
* try to write out the rest of the page. Journalled mode is
|
|
|
|
* not supported by delalloc.
|
2008-08-20 09:55:02 +08:00
|
|
|
*/
|
|
|
|
BUG_ON(ext4_should_journal_data(inode));
|
2008-08-20 10:15:58 +08:00
|
|
|
needed_blocks = ext4_da_writepages_trans_blocks(inode);
|
2008-08-20 09:55:02 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
/* start a new transaction */
|
2013-06-05 01:21:11 +08:00
|
|
|
handle = ext4_journal_start_with_reserve(inode,
|
|
|
|
EXT4_HT_WRITE_PAGE, needed_blocks, rsv_blocks);
|
2008-07-12 07:27:31 +08:00
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
2009-09-27 05:43:59 +08:00
|
|
|
ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: "
|
2010-05-17 01:00:00 +08:00
|
|
|
"%ld pages, ino %lu; err %d", __func__,
|
2008-08-20 09:55:02 +08:00
|
|
|
wbc->nr_to_write, inode->i_ino, ret);
|
2013-06-05 01:17:40 +08:00
|
|
|
/* Release allocated io_end */
|
|
|
|
ext4_put_io_end(mpd.io_submit.io_end);
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
mpd.io_submit.io_end = NULL;
|
2013-06-05 01:17:40 +08:00
|
|
|
break;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
mpd.do_map = 1;
|
2009-02-24 05:42:39 +08:00
|
|
|
|
2013-06-05 01:17:40 +08:00
|
|
|
trace_ext4_da_write_pages(inode, mpd.first_page, mpd.wbc);
|
|
|
|
ret = mpage_prepare_extent_to_map(&mpd);
|
2020-05-25 16:12:15 +08:00
|
|
|
if (!ret && mpd.map.m_len)
|
|
|
|
ret = mpage_map_and_submit_extent(handle, &mpd,
|
2013-07-01 20:12:40 +08:00
|
|
|
&give_up_on_write);
|
2016-07-04 22:14:01 +08:00
|
|
|
/*
|
|
|
|
* Caution: If the handle is synchronous,
|
|
|
|
* ext4_journal_stop() can wait for transaction commit
|
|
|
|
* to finish which may depend on writeback of pages to
|
|
|
|
* complete or on page lock to be released. In that
|
2020-08-05 10:48:50 +08:00
|
|
|
* case, we have to wait until after we have
|
2016-07-04 22:14:01 +08:00
|
|
|
* submitted all the IO, released page locks we hold,
|
|
|
|
* and dropped io_end reference (for extent conversion
|
|
|
|
* to be able to complete) before stopping the handle.
|
|
|
|
*/
|
|
|
|
if (!ext4_handle_valid(handle) || handle->h_sync == 0) {
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
handle = NULL;
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
mpd.do_map = 0;
|
2016-07-04 22:14:01 +08:00
|
|
|
}
|
2013-06-05 01:17:40 +08:00
|
|
|
/* Unlock pages we didn't use */
|
2013-07-01 20:12:40 +08:00
|
|
|
mpage_release_unused_pages(&mpd, give_up_on_write);
|
ext4: unlock unused_pages timely when doing writeback
In mpage_add_bh_to_extent(), when accumulated extents length is greater
than MAX_WRITEPAGES_EXTENT_LEN or buffer head's b_stat is not equal, we
will not continue to search unmapped area for this page, but note this
page is locked, and will only be unlocked in mpage_release_unused_pages()
after ext4_io_submit, if io also is throttled by blk-throttle or similar
io qos, we will hold this page locked for a while, it's unnecessary.
I think the best fix is to refactor mpage_add_bh_to_extent() to let it
return some hints whether to unlock this page, but given that we will
improve dioread_nolock later, we can let it done later, so currently
the simple fix would just call mpage_release_unused_pages() before
ext4_io_submit().
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-02-11 12:53:21 +08:00
|
|
|
/* Submit prepared bio */
|
|
|
|
ext4_io_submit(&mpd.io_submit);
|
|
|
|
|
2016-07-04 22:14:01 +08:00
|
|
|
/*
|
|
|
|
* Drop our io_end reference we got from init. We have
|
|
|
|
* to be careful and use deferred io_end finishing if
|
|
|
|
* we are still holding the transaction as we can
|
|
|
|
* release the last reference to io_end which may end
|
|
|
|
* up doing unwritten extent conversion.
|
|
|
|
*/
|
|
|
|
if (handle) {
|
|
|
|
ext4_put_io_end_defer(mpd.io_submit.io_end);
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
} else
|
|
|
|
ext4_put_io_end(mpd.io_submit.io_end);
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
mpd.io_submit.io_end = NULL;
|
2013-06-05 01:17:40 +08:00
|
|
|
|
|
|
|
if (ret == -ENOSPC && sbi->s_journal) {
|
|
|
|
/*
|
|
|
|
* Commit the transaction which would
|
2008-10-16 22:10:36 +08:00
|
|
|
* free blocks released in the transaction
|
|
|
|
* and try again
|
|
|
|
*/
|
2008-09-09 11:05:34 +08:00
|
|
|
jbd2_journal_force_commit_nested(sbi->s_journal);
|
2008-10-16 22:10:36 +08:00
|
|
|
ret = 0;
|
2013-06-05 01:17:40 +08:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* Fatal error - ENOMEM, EIO... */
|
|
|
|
if (ret)
|
2008-07-12 07:27:31 +08:00
|
|
|
break;
|
2008-08-20 09:55:02 +08:00
|
|
|
}
|
ext4: avoid unnecessary transaction stalls during writeback
Currently ext4_writepages() submits all pages with transaction started.
When no page needs block allocation or extent conversion we can submit
all dirty pages in the inode while holding a single transaction handle
and when device is congested this can take significant amount of time.
Thus ext4_writepages() can block transaction commits for extended
periods of time.
Take for example a simple benchmark simulating PostgreSQL database
(pgioperf in mmtest). The benchmark runs 16 processes doing random reads
from a huge file, one process doing random writes to the huge file, and
one process doing sequential writes to a small files and frequently
running fsync. With unpatched kernel transaction commits take on average
~18s with standard deviation of ~41s, top 5 commit times are:
274.466639s, 126.467347s, 86.992429s, 34.351563s, 31.517653s.
After this patch transaction commits take on average 0.1s with standard
deviation of 0.15s, top 5 commit times are:
0.563792s, 0.519980s, 0.509841s, 0.471700s, 0.469899s
[ Modified so we use an explicit do_map flag instead of relying on
io_end not being allocated, the since io_end->inode is needed for I/O
error handling. -- tytso ]
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-05-01 06:29:10 +08:00
|
|
|
unplug:
|
2011-10-18 22:55:51 +08:00
|
|
|
blk_finish_plug(&plug);
|
2013-09-16 20:24:26 +08:00
|
|
|
if (!ret && !cycled && wbc->nr_to_write > 0) {
|
2009-02-14 23:42:58 +08:00
|
|
|
cycled = 1;
|
2013-06-05 01:17:40 +08:00
|
|
|
mpd.last_page = writeback_index - 1;
|
|
|
|
mpd.first_page = 0;
|
2009-02-14 23:42:58 +08:00
|
|
|
goto retry;
|
|
|
|
}
|
2008-10-16 22:10:36 +08:00
|
|
|
|
|
|
|
/* Update index */
|
|
|
|
if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0))
|
|
|
|
/*
|
2013-06-05 01:17:40 +08:00
|
|
|
* Set the writeback_index so that range_cyclic
|
2008-10-16 22:10:36 +08:00
|
|
|
* mode will write it back later
|
|
|
|
*/
|
2013-06-05 01:17:40 +08:00
|
|
|
mapping->writeback_index = mpd.first_page;
|
2008-08-20 09:55:02 +08:00
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
out_writepages:
|
2013-06-07 02:00:46 +08:00
|
|
|
trace_ext4_writepages_result(inode, wbc, ret,
|
|
|
|
nr_to_write - wbc->nr_to_write);
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_up_read(&sbi->s_writepages_rwsem);
|
2008-07-12 07:27:31 +08:00
|
|
|
return ret;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
|
|
|
|
2017-12-22 04:25:11 +08:00
|
|
|
static int ext4_dax_writepages(struct address_space *mapping,
|
|
|
|
struct writeback_control *wbc)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
long nr_to_write = wbc->nr_to_write;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb);
|
|
|
|
|
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_down_read(&sbi->s_writepages_rwsem);
|
2017-12-22 04:25:11 +08:00
|
|
|
trace_ext4_writepages(inode, wbc);
|
|
|
|
|
2020-01-04 02:33:07 +08:00
|
|
|
ret = dax_writeback_mapping_range(mapping, sbi->s_daxdev, wbc);
|
2017-12-22 04:25:11 +08:00
|
|
|
trace_ext4_writepages_result(inode, wbc, ret,
|
|
|
|
nr_to_write - wbc->nr_to_write);
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_up_read(&sbi->s_writepages_rwsem);
|
2017-12-22 04:25:11 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-10-09 11:13:30 +08:00
|
|
|
static int ext4_nonda_switch(struct super_block *sb)
|
|
|
|
{
|
2013-04-09 21:27:31 +08:00
|
|
|
s64 free_clusters, dirty_clusters;
|
2008-10-09 11:13:30 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(sb);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* switch to non delalloc mode if we are running low
|
|
|
|
* on free block. The free block accounting via percpu
|
2009-01-07 06:41:04 +08:00
|
|
|
* counters can get slightly wrong with percpu_counter_batch getting
|
2008-10-09 11:13:30 +08:00
|
|
|
* accumulated on each CPU without updating global counters
|
|
|
|
* Delalloc need an accurate free block accounting. So switch
|
|
|
|
* to non delalloc when we are near to error range.
|
|
|
|
*/
|
2013-04-09 21:27:31 +08:00
|
|
|
free_clusters =
|
|
|
|
percpu_counter_read_positive(&sbi->s_freeclusters_counter);
|
|
|
|
dirty_clusters =
|
|
|
|
percpu_counter_read_positive(&sbi->s_dirtyclusters_counter);
|
ext4: fix potential deadlock in ext4_nonda_switch()
In ext4_nonda_switch(), if the file system is getting full we used to
call writeback_inodes_sb_if_idle(). The problem is that we can be
holding i_mutex already, and this causes a potential deadlock when
writeback_inodes_sb_if_idle() when it tries to take s_umount. (See
lockdep output below).
As it turns out we don't need need to hold s_umount; the fact that we
are in the middle of the write(2) system call will keep the superblock
pinned. Unfortunately writeback_inodes_sb() checks to make sure
s_umount is taken, and the VFS uses a different mechanism for making
sure the file system doesn't get unmounted out from under us. The
simplest way of dealing with this is to just simply grab s_umount
using a trylock, and skip kicking the writeback flusher thread in the
very unlikely case that we can't take a read lock on s_umount without
blocking.
Also, we now check the cirteria for kicking the writeback thread
before we decide to whether to fall back to non-delayed writeback, so
if there are any outstanding delayed allocation writes, we try to get
them resolved as soon as possible.
[ INFO: possible circular locking dependency detected ]
3.6.0-rc1-00042-gce894ca #367 Not tainted
-------------------------------------------------------
dd/8298 is trying to acquire lock:
(&type->s_umount_key#18){++++..}, at: [<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
but task is already holding lock:
(&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
which lock already depends on the new lock.
2 locks held by dd/8298:
#0: (sb_writers#2){.+.+.+}, at: [<c01ddcc5>] generic_file_aio_write+0x56/0xd3
#1: (&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
stack backtrace:
Pid: 8298, comm: dd Not tainted 3.6.0-rc1-00042-gce894ca #367
Call Trace:
[<c015b79c>] ? console_unlock+0x345/0x372
[<c06d62a1>] print_circular_bug+0x190/0x19d
[<c019906c>] __lock_acquire+0x86d/0xb6c
[<c01999db>] ? mark_held_locks+0x5c/0x7b
[<c0199724>] lock_acquire+0x66/0xb9
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c06db935>] down_read+0x28/0x58
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
[<c026f3b2>] ext4_nonda_switch+0xe1/0xf4
[<c0271ece>] ext4_da_write_begin+0x27/0x193
[<c01dcdb0>] generic_file_buffered_write+0xc8/0x1bb
[<c01ddc47>] __generic_file_aio_write+0x1dd/0x205
[<c01ddce7>] generic_file_aio_write+0x78/0xd3
[<c026d336>] ext4_file_write+0x480/0x4a6
[<c0198c1d>] ? __lock_acquire+0x41e/0xb6c
[<c0180944>] ? sched_clock_cpu+0x11a/0x13e
[<c01967e9>] ? trace_hardirqs_off+0xb/0xd
[<c018099f>] ? local_clock+0x37/0x4e
[<c0209f2c>] do_sync_write+0x67/0x9d
[<c0209ec5>] ? wait_on_retry_sync_kiocb+0x44/0x44
[<c020a7b9>] vfs_write+0x7b/0xe6
[<c020a9a6>] sys_write+0x3b/0x64
[<c06dd4bd>] syscall_call+0x7/0xb
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
2012-09-20 10:42:36 +08:00
|
|
|
/*
|
|
|
|
* Start pushing delalloc when 1/2 of free blocks are dirty.
|
|
|
|
*/
|
2013-04-09 21:27:31 +08:00
|
|
|
if (dirty_clusters && (free_clusters < 2 * dirty_clusters))
|
2013-01-10 13:47:57 +08:00
|
|
|
try_to_writeback_inodes_sb(sb, WB_REASON_FS_FREE_SPACE);
|
ext4: fix potential deadlock in ext4_nonda_switch()
In ext4_nonda_switch(), if the file system is getting full we used to
call writeback_inodes_sb_if_idle(). The problem is that we can be
holding i_mutex already, and this causes a potential deadlock when
writeback_inodes_sb_if_idle() when it tries to take s_umount. (See
lockdep output below).
As it turns out we don't need need to hold s_umount; the fact that we
are in the middle of the write(2) system call will keep the superblock
pinned. Unfortunately writeback_inodes_sb() checks to make sure
s_umount is taken, and the VFS uses a different mechanism for making
sure the file system doesn't get unmounted out from under us. The
simplest way of dealing with this is to just simply grab s_umount
using a trylock, and skip kicking the writeback flusher thread in the
very unlikely case that we can't take a read lock on s_umount without
blocking.
Also, we now check the cirteria for kicking the writeback thread
before we decide to whether to fall back to non-delayed writeback, so
if there are any outstanding delayed allocation writes, we try to get
them resolved as soon as possible.
[ INFO: possible circular locking dependency detected ]
3.6.0-rc1-00042-gce894ca #367 Not tainted
-------------------------------------------------------
dd/8298 is trying to acquire lock:
(&type->s_umount_key#18){++++..}, at: [<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
but task is already holding lock:
(&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
which lock already depends on the new lock.
2 locks held by dd/8298:
#0: (sb_writers#2){.+.+.+}, at: [<c01ddcc5>] generic_file_aio_write+0x56/0xd3
#1: (&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
stack backtrace:
Pid: 8298, comm: dd Not tainted 3.6.0-rc1-00042-gce894ca #367
Call Trace:
[<c015b79c>] ? console_unlock+0x345/0x372
[<c06d62a1>] print_circular_bug+0x190/0x19d
[<c019906c>] __lock_acquire+0x86d/0xb6c
[<c01999db>] ? mark_held_locks+0x5c/0x7b
[<c0199724>] lock_acquire+0x66/0xb9
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c06db935>] down_read+0x28/0x58
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
[<c026f3b2>] ext4_nonda_switch+0xe1/0xf4
[<c0271ece>] ext4_da_write_begin+0x27/0x193
[<c01dcdb0>] generic_file_buffered_write+0xc8/0x1bb
[<c01ddc47>] __generic_file_aio_write+0x1dd/0x205
[<c01ddce7>] generic_file_aio_write+0x78/0xd3
[<c026d336>] ext4_file_write+0x480/0x4a6
[<c0198c1d>] ? __lock_acquire+0x41e/0xb6c
[<c0180944>] ? sched_clock_cpu+0x11a/0x13e
[<c01967e9>] ? trace_hardirqs_off+0xb/0xd
[<c018099f>] ? local_clock+0x37/0x4e
[<c0209f2c>] do_sync_write+0x67/0x9d
[<c0209ec5>] ? wait_on_retry_sync_kiocb+0x44/0x44
[<c020a7b9>] vfs_write+0x7b/0xe6
[<c020a9a6>] sys_write+0x3b/0x64
[<c06dd4bd>] syscall_call+0x7/0xb
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
2012-09-20 10:42:36 +08:00
|
|
|
|
2013-04-09 21:27:31 +08:00
|
|
|
if (2 * free_clusters < 3 * dirty_clusters ||
|
|
|
|
free_clusters < (dirty_clusters + EXT4_FREECLUSTERS_WATERMARK)) {
|
2008-10-09 11:13:30 +08:00
|
|
|
/*
|
2009-12-23 20:58:12 +08:00
|
|
|
* free block count is less than 150% of dirty blocks
|
|
|
|
* or free blocks is less than watermark
|
2008-10-09 11:13:30 +08:00
|
|
|
*/
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
|
2022-02-23 03:31:43 +08:00
|
|
|
loff_t pos, unsigned len,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct page **pagep, void **fsdata)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2010-05-16 23:00:00 +08:00
|
|
|
int ret, retries = 0;
|
2008-07-12 07:27:31 +08:00
|
|
|
struct page *page;
|
|
|
|
pgoff_t index;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
index = pos >> PAGE_SHIFT;
|
2008-10-09 11:13:30 +08:00
|
|
|
|
2022-04-24 22:09:36 +08:00
|
|
|
if (ext4_nonda_switch(inode->i_sb) || ext4_verity_in_progress(inode)) {
|
2008-10-09 11:13:30 +08:00
|
|
|
*fsdata = (void *)FALL_BACK_TO_NONDELALLOC;
|
|
|
|
return ext4_write_begin(file, mapping, pos,
|
2022-02-23 03:31:43 +08:00
|
|
|
len, pagep, fsdata);
|
2008-10-09 11:13:30 +08:00
|
|
|
}
|
|
|
|
*fsdata = (void *)0;
|
2022-02-23 03:31:43 +08:00
|
|
|
trace_ext4_da_write_begin(inode, pos, len);
|
2012-12-11 03:05:57 +08:00
|
|
|
|
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) {
|
2022-02-22 23:36:28 +08:00
|
|
|
ret = ext4_da_write_inline_data_begin(mapping, inode, pos, len,
|
2012-12-11 03:05:57 +08:00
|
|
|
pagep, fsdata);
|
|
|
|
if (ret < 0)
|
2013-02-09 22:24:14 +08:00
|
|
|
return ret;
|
|
|
|
if (ret == 1)
|
|
|
|
return 0;
|
2012-12-11 03:05:57 +08:00
|
|
|
}
|
|
|
|
|
2021-07-16 20:20:24 +08:00
|
|
|
retry:
|
2022-02-23 00:25:12 +08:00
|
|
|
page = grab_cache_page_write_begin(mapping, index);
|
2013-02-09 22:24:14 +08:00
|
|
|
if (!page)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
/* In case writeback began while the page was unlocked */
|
2013-08-29 02:30:47 +08:00
|
|
|
wait_for_stable_page(page);
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2018-12-12 17:50:12 +08:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
2015-04-12 12:55:10 +08:00
|
|
|
ret = ext4_block_write_begin(page, pos, len,
|
|
|
|
ext4_da_get_block_prep);
|
|
|
|
#else
|
2010-06-04 17:29:57 +08:00
|
|
|
ret = __block_write_begin(page, pos, len, ext4_da_get_block_prep);
|
2015-04-12 12:55:10 +08:00
|
|
|
#endif
|
2008-07-12 07:27:31 +08:00
|
|
|
if (ret < 0) {
|
|
|
|
unlock_page(page);
|
2021-07-16 20:20:24 +08:00
|
|
|
put_page(page);
|
2008-09-14 01:10:25 +08:00
|
|
|
/*
|
|
|
|
* block_write_begin may have instantiated a few blocks
|
|
|
|
* outside i_size. Trim these off again. Don't need
|
2021-07-16 20:20:24 +08:00
|
|
|
* i_size_read because we hold inode lock.
|
2008-09-14 01:10:25 +08:00
|
|
|
*/
|
|
|
|
if (pos + len > inode->i_size)
|
2009-12-09 10:24:33 +08:00
|
|
|
ext4_truncate_failed_write(inode);
|
2013-02-09 22:24:14 +08:00
|
|
|
|
|
|
|
if (ret == -ENOSPC &&
|
|
|
|
ext4_should_retry_alloc(inode->i_sb, &retries))
|
2021-07-16 20:20:24 +08:00
|
|
|
goto retry;
|
2013-02-09 22:24:14 +08:00
|
|
|
return ret;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
|
|
|
|
2013-02-09 22:24:14 +08:00
|
|
|
*pagep = page;
|
2008-07-12 07:27:31 +08:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
|
|
|
* Check if we should update i_disksize
|
|
|
|
* when write to the end of file but not require block allocation
|
|
|
|
*/
|
|
|
|
static int ext4_da_should_update_i_disksize(struct page *page,
|
2009-06-15 05:45:34 +08:00
|
|
|
unsigned long offset)
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
{
|
|
|
|
struct buffer_head *bh;
|
|
|
|
struct inode *inode = page->mapping->host;
|
|
|
|
unsigned int idx;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
bh = page_buffers(page);
|
|
|
|
idx = offset >> inode->i_blkbits;
|
|
|
|
|
2008-09-09 10:25:24 +08:00
|
|
|
for (i = 0; i < idx; i++)
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
bh = bh->b_this_page;
|
|
|
|
|
2009-05-13 04:30:27 +08:00
|
|
|
if (!buffer_mapped(bh) || (buffer_delay(bh)) || buffer_unwritten(bh))
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
return 0;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
static int ext4_da_write_end(struct file *file,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct address_space *mapping,
|
|
|
|
loff_t pos, unsigned len, unsigned copied,
|
|
|
|
struct page *page, void *fsdata)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
loff_t new_i_size;
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
unsigned long start, end;
|
2008-10-09 11:13:30 +08:00
|
|
|
int write_mode = (int)(unsigned long)fsdata;
|
|
|
|
|
2013-04-04 00:39:17 +08:00
|
|
|
if (write_mode == FALL_BACK_TO_NONDELALLOC)
|
|
|
|
return ext4_write_end(file, mapping, pos,
|
|
|
|
len, copied, page, fsdata);
|
ext4: fix delalloc i_disksize early update issue
Ext4_da_write_end() used walk_page_buffers() with a callback function of
ext4_bh_unmapped_or_delay() to check if it extended the file size
without allocating any blocks (since in this case i_disksize needs to be
updated). However, this is didn't work proprely because the buffer head
has not been marked dirty yet --- this is done later in
block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to
always return false.
In addition, walk_page_buffers() checks all of the buffer heads covering
the page, and the only buffer_head that should be checked is the one
covering the end of the write. Otherwise, given a 1k blocksize
filesystem and a 4k page size, the buffer head covering the first 1k
stripe of the file could be unmapped (because it was a sparse file), and
the second or third buffer_head covering that page could be mapped, and
using walk_page_buffers() would fail in this case since it would stop at
the first unmapped buffer_head and return true.
The core problem is that walk_page_buffers() was intended to do work in
a callback function, and a non-zero return value indicated a failure,
which termined the walk of the buffer heads covering the page. It was
not intended to be used with a boolean function, such as
ext4_bh_unmapped_or_delay().
Add addtional fix from Aneesh to protect i_disksize update rave with truncate.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-07-12 07:27:31 +08:00
|
|
|
|
2009-06-17 23:48:11 +08:00
|
|
|
trace_ext4_da_write_end(inode, pos, len, copied);
|
2021-07-16 20:20:23 +08:00
|
|
|
|
|
|
|
if (write_mode != CONVERT_INLINE_DATA &&
|
|
|
|
ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA) &&
|
|
|
|
ext4_has_inline_data(inode))
|
|
|
|
return ext4_write_inline_data_end(inode, pos, len, copied, page);
|
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
start = pos & (PAGE_SIZE - 1);
|
2008-09-09 10:25:24 +08:00
|
|
|
end = start + copied - 1;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
|
|
|
/*
|
2021-07-16 20:20:21 +08:00
|
|
|
* Since we are holding inode lock, we are sure i_disksize <=
|
|
|
|
* i_size. We also know that if i_disksize < i_size, there are
|
|
|
|
* delalloc writes pending in the range upto i_size. If the end of
|
|
|
|
* the current write is <= i_size, there's no need to touch
|
|
|
|
* i_disksize since writeback will push i_disksize upto i_size
|
|
|
|
* eventually. If the end of the current write is > i_size and
|
|
|
|
* inside an allocated block (ext4_da_should_update_i_disksize()
|
|
|
|
* check), we need to update i_disksize here as neither
|
|
|
|
* ext4_writepage() nor certain ext4_writepages() paths not
|
|
|
|
* allocating blocks update i_disksize.
|
|
|
|
*
|
|
|
|
* Note that we defer inode dirtying to generic_write_end() /
|
|
|
|
* ext4_da_write_inline_data_end().
|
2008-07-12 07:27:31 +08:00
|
|
|
*/
|
|
|
|
new_i_size = pos + copied;
|
2021-07-16 20:20:23 +08:00
|
|
|
if (copied && new_i_size > inode->i_size &&
|
|
|
|
ext4_da_should_update_i_disksize(page, end))
|
|
|
|
ext4_update_i_disksize(inode, new_i_size);
|
2012-12-11 03:05:57 +08:00
|
|
|
|
2021-07-16 20:20:24 +08:00
|
|
|
return generic_write_end(file, mapping, pos, len, copied, page, fsdata);
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
|
|
|
|
2009-02-26 14:04:07 +08:00
|
|
|
/*
|
|
|
|
* Force all delayed allocation blocks to be allocated for a given inode.
|
|
|
|
*/
|
|
|
|
int ext4_alloc_da_blocks(struct inode *inode)
|
|
|
|
{
|
2009-09-17 07:30:40 +08:00
|
|
|
trace_ext4_alloc_da_blocks(inode);
|
|
|
|
|
2014-07-15 18:02:38 +08:00
|
|
|
if (!EXT4_I(inode)->i_reserved_data_blocks)
|
2009-02-26 14:04:07 +08:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We do something simple for now. The filemap_flush() will
|
|
|
|
* also start triggering a write of the data blocks, which is
|
|
|
|
* not strictly speaking necessary (and for users of
|
|
|
|
* laptop_mode, not even desirable). However, to do otherwise
|
|
|
|
* would require replicating code paths in:
|
2009-06-15 05:45:34 +08:00
|
|
|
*
|
2013-06-07 02:00:46 +08:00
|
|
|
* ext4_writepages() ->
|
2009-02-26 14:04:07 +08:00
|
|
|
* write_cache_pages() ---> (via passed in callback function)
|
|
|
|
* __mpage_da_writepage() -->
|
|
|
|
* mpage_add_bh_to_extent()
|
|
|
|
* mpage_da_map_blocks()
|
|
|
|
*
|
|
|
|
* The problem is that write_cache_pages(), located in
|
|
|
|
* mm/page-writeback.c, marks pages clean in preparation for
|
|
|
|
* doing I/O, which is not desirable if we're not planning on
|
|
|
|
* doing I/O at all.
|
|
|
|
*
|
|
|
|
* We could call write_cache_pages(), and then redirty all of
|
2010-11-11 19:23:29 +08:00
|
|
|
* the pages by calling redirty_page_for_writepage() but that
|
2009-02-26 14:04:07 +08:00
|
|
|
* would be ugly in the extreme. So instead we would need to
|
|
|
|
* replicate parts of the code in the above functions,
|
2011-03-31 09:57:33 +08:00
|
|
|
* simplifying them because we wouldn't actually intend to
|
2009-02-26 14:04:07 +08:00
|
|
|
* write out the pages, but rather only collect contiguous
|
|
|
|
* logical block extents, call the multi-block allocator, and
|
|
|
|
* then update the buffer heads with the block allocations.
|
2009-06-15 05:45:34 +08:00
|
|
|
*
|
2009-02-26 14:04:07 +08:00
|
|
|
* For now, though, we'll cheat by calling filemap_flush(),
|
|
|
|
* which will map the blocks, and start the I/O, but not
|
|
|
|
* actually wait for the I/O to complete.
|
|
|
|
*/
|
|
|
|
return filemap_flush(inode->i_mapping);
|
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* bmap() is special. It gets used by applications such as lilo and by
|
|
|
|
* the swapper to find the on-disk block of a specific piece of data.
|
|
|
|
*
|
|
|
|
* Naturally, this is dangerous if the block concerned is still in the
|
2006-10-11 16:20:53 +08:00
|
|
|
* journal. If somebody makes a swapfile on an ext4 data-journaling
|
2006-10-11 16:20:50 +08:00
|
|
|
* filesystem and enables swap, then they may get a nasty shock when the
|
|
|
|
* data getting swapped to that swapfile suddenly gets overwritten by
|
|
|
|
* the original zero's written out previously to the journal and
|
|
|
|
* awaiting writeback in the kernel's buffer cache.
|
|
|
|
*
|
|
|
|
* So, if we see any bmap calls here on a modified, data-journaled file,
|
|
|
|
* take extra steps to flush any blocks which might be in the cache.
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
static sector_t ext4_bmap(struct address_space *mapping, sector_t block)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
journal_t *journal;
|
2022-06-17 09:39:35 +08:00
|
|
|
sector_t ret = 0;
|
2006-10-11 16:20:50 +08:00
|
|
|
int err;
|
|
|
|
|
2022-06-17 09:39:35 +08:00
|
|
|
inode_lock_shared(inode);
|
2012-12-11 03:04:52 +08:00
|
|
|
/*
|
|
|
|
* We can get here for an inline file via the FIBMAP ioctl
|
|
|
|
*/
|
|
|
|
if (ext4_has_inline_data(inode))
|
2022-06-17 09:39:35 +08:00
|
|
|
goto out;
|
2012-12-11 03:04:52 +08:00
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY) &&
|
|
|
|
test_opt(inode->i_sb, DELALLOC)) {
|
|
|
|
/*
|
|
|
|
* With delalloc we want to sync the file
|
|
|
|
* so that we can make sure we allocate
|
|
|
|
* blocks for file
|
|
|
|
*/
|
|
|
|
filemap_write_and_wait(mapping);
|
|
|
|
}
|
|
|
|
|
2010-01-25 03:34:07 +08:00
|
|
|
if (EXT4_JOURNAL(inode) &&
|
|
|
|
ext4_test_inode_state(inode, EXT4_STATE_JDATA)) {
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* This is a REALLY heavyweight approach, but the use of
|
|
|
|
* bmap on dirty files is expected to be extremely rare:
|
|
|
|
* only if we run lilo or swapon on a freshly made file
|
|
|
|
* do we expect this to happen.
|
|
|
|
*
|
|
|
|
* (bmap requires CAP_SYS_RAWIO so this does not
|
|
|
|
* represent an unprivileged user DOS attack --- we'd be
|
|
|
|
* in trouble if mortal users could trigger this path at
|
|
|
|
* will.)
|
|
|
|
*
|
2006-10-11 16:20:53 +08:00
|
|
|
* NB. EXT4_STATE_JDATA is not set on files other than
|
2006-10-11 16:20:50 +08:00
|
|
|
* regular files. If somebody wants to bmap a directory
|
|
|
|
* or symlink and gets confused because the buffer
|
|
|
|
* hasn't yet been flushed to disk, they deserve
|
|
|
|
* everything they get.
|
|
|
|
*/
|
|
|
|
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_clear_inode_state(inode, EXT4_STATE_JDATA);
|
2006-10-11 16:20:53 +08:00
|
|
|
journal = EXT4_JOURNAL(inode);
|
2006-10-11 16:21:01 +08:00
|
|
|
jbd2_journal_lock_updates(journal);
|
2021-05-18 23:13:25 +08:00
|
|
|
err = jbd2_journal_flush(journal, 0);
|
2006-10-11 16:21:01 +08:00
|
|
|
jbd2_journal_unlock_updates(journal);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
if (err)
|
2022-06-17 09:39:35 +08:00
|
|
|
goto out;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2022-06-17 09:39:35 +08:00
|
|
|
ret = iomap_bmap(mapping, block, &ext4_iomap_ops);
|
|
|
|
|
|
|
|
out:
|
|
|
|
inode_unlock_shared(inode);
|
|
|
|
return ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2022-04-29 23:12:16 +08:00
|
|
|
static int ext4_read_folio(struct file *file, struct folio *folio)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2022-04-29 23:12:16 +08:00
|
|
|
struct page *page = &folio->page;
|
2012-12-11 03:04:52 +08:00
|
|
|
int ret = -EAGAIN;
|
|
|
|
struct inode *inode = page->mapping->host;
|
|
|
|
|
2011-03-22 09:38:05 +08:00
|
|
|
trace_ext4_readpage(page);
|
2012-12-11 03:04:52 +08:00
|
|
|
|
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
ret = ext4_readpage_inline(inode, page);
|
|
|
|
|
|
|
|
if (ret == -EAGAIN)
|
2020-06-02 12:47:20 +08:00
|
|
|
return ext4_mpage_readpages(inode, NULL, page);
|
2012-12-11 03:04:52 +08:00
|
|
|
|
|
|
|
return ret;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2020-06-02 12:47:16 +08:00
|
|
|
static void ext4_readahead(struct readahead_control *rac)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2020-06-02 12:47:16 +08:00
|
|
|
struct inode *inode = rac->mapping->host;
|
2012-12-11 03:04:52 +08:00
|
|
|
|
2020-06-02 12:47:16 +08:00
|
|
|
/* If the file has inline data, no need to do readahead. */
|
2012-12-11 03:04:52 +08:00
|
|
|
if (ext4_has_inline_data(inode))
|
2020-06-02 12:47:16 +08:00
|
|
|
return;
|
2012-12-11 03:04:52 +08:00
|
|
|
|
2020-06-02 12:47:20 +08:00
|
|
|
ext4_mpage_readpages(inode, rac, NULL);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2022-02-10 04:21:34 +08:00
|
|
|
static void ext4_invalidate_folio(struct folio *folio, size_t offset,
|
|
|
|
size_t length)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2022-02-10 04:21:43 +08:00
|
|
|
trace_ext4_invalidate_folio(folio, offset, length);
|
2011-03-22 09:38:05 +08:00
|
|
|
|
2012-12-26 02:28:54 +08:00
|
|
|
/* No journalling happens on data buffers when this function is used */
|
2022-02-10 04:21:34 +08:00
|
|
|
WARN_ON(folio_buffers(folio) && buffer_jbd(folio_buffers(folio)));
|
2012-12-26 02:28:54 +08:00
|
|
|
|
2022-02-10 04:21:34 +08:00
|
|
|
block_invalidate_folio(folio, offset, length);
|
2012-12-26 02:28:54 +08:00
|
|
|
}
|
|
|
|
|
2022-02-10 04:21:43 +08:00
|
|
|
static int __ext4_journalled_invalidate_folio(struct folio *folio,
|
|
|
|
size_t offset, size_t length)
|
2012-12-26 02:28:54 +08:00
|
|
|
{
|
2022-02-10 04:21:43 +08:00
|
|
|
journal_t *journal = EXT4_JOURNAL(folio->mapping->host);
|
2012-12-26 02:28:54 +08:00
|
|
|
|
2022-02-10 04:21:43 +08:00
|
|
|
trace_ext4_journalled_invalidate_folio(folio, offset, length);
|
2012-12-26 02:28:54 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* If it's a full truncate we just forget about the pending dirtying
|
|
|
|
*/
|
2022-02-10 04:21:43 +08:00
|
|
|
if (offset == 0 && length == folio_size(folio))
|
|
|
|
folio_clear_checked(folio);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2022-02-10 04:21:43 +08:00
|
|
|
return jbd2_journal_invalidate_folio(journal, folio, offset, length);
|
2012-12-26 02:29:52 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Wrapper for aops... */
|
2022-02-10 04:21:43 +08:00
|
|
|
static void ext4_journalled_invalidate_folio(struct folio *folio,
|
|
|
|
size_t offset,
|
|
|
|
size_t length)
|
2012-12-26 02:29:52 +08:00
|
|
|
{
|
2022-02-10 04:21:43 +08:00
|
|
|
WARN_ON(__ext4_journalled_invalidate_folio(folio, offset, length) < 0);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2022-05-01 11:33:38 +08:00
|
|
|
static bool ext4_release_folio(struct folio *folio, gfp_t wait)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2022-05-01 11:33:38 +08:00
|
|
|
journal_t *journal = EXT4_JOURNAL(folio->mapping->host);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2022-05-01 11:33:38 +08:00
|
|
|
trace_ext4_releasepage(&folio->page);
|
2011-03-22 09:38:05 +08:00
|
|
|
|
2013-03-11 10:19:00 +08:00
|
|
|
/* Page has dirty journalled data -> cannot release */
|
2022-05-01 11:33:38 +08:00
|
|
|
if (folio_test_checked(folio))
|
|
|
|
return false;
|
2009-01-07 13:06:22 +08:00
|
|
|
if (journal)
|
2022-05-01 12:46:03 +08:00
|
|
|
return jbd2_journal_try_to_free_buffers(journal, folio);
|
2009-01-07 13:06:22 +08:00
|
|
|
else
|
2022-05-01 13:08:08 +08:00
|
|
|
return try_to_free_buffers(folio);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2017-11-01 23:36:45 +08:00
|
|
|
static bool ext4_inode_datasync_dirty(struct inode *inode)
|
|
|
|
{
|
|
|
|
journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
|
|
|
|
|
2020-10-16 04:37:57 +08:00
|
|
|
if (journal) {
|
|
|
|
if (jbd2_transaction_committed(journal,
|
2020-10-27 12:49:13 +08:00
|
|
|
EXT4_I(inode)->i_datasync_tid))
|
|
|
|
return false;
|
|
|
|
if (test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT))
|
2020-11-06 11:59:06 +08:00
|
|
|
return !list_empty(&EXT4_I(inode)->i_fc_list);
|
2020-10-27 12:49:13 +08:00
|
|
|
return true;
|
2020-10-16 04:37:57 +08:00
|
|
|
}
|
|
|
|
|
2017-11-01 23:36:45 +08:00
|
|
|
/* Any metadata buffers to write? */
|
|
|
|
if (!list_empty(&inode->i_mapping->private_list))
|
|
|
|
return true;
|
|
|
|
return inode->i_state & I_DIRTY_DATASYNC;
|
|
|
|
}
|
|
|
|
|
2019-11-05 19:59:56 +08:00
|
|
|
static void ext4_set_iomap(struct inode *inode, struct iomap *iomap,
|
|
|
|
struct ext4_map_blocks *map, loff_t offset,
|
2021-11-29 18:22:00 +08:00
|
|
|
loff_t length, unsigned int flags)
|
2016-11-21 06:36:06 +08:00
|
|
|
{
|
2019-11-05 19:59:56 +08:00
|
|
|
u8 blkbits = inode->i_blkbits;
|
2016-11-21 06:36:06 +08:00
|
|
|
|
2019-11-05 19:59:56 +08:00
|
|
|
/*
|
|
|
|
* Writes that span EOF might trigger an I/O size update on completion,
|
|
|
|
* so consider them to be dirty for the purpose of O_DSYNC, even if
|
|
|
|
* there is no other metadata changes being made or are pending.
|
|
|
|
*/
|
2016-11-21 06:36:06 +08:00
|
|
|
iomap->flags = 0;
|
2019-11-05 19:59:56 +08:00
|
|
|
if (ext4_inode_datasync_dirty(inode) ||
|
|
|
|
offset + length > i_size_read(inode))
|
2017-11-01 23:36:45 +08:00
|
|
|
iomap->flags |= IOMAP_F_DIRTY;
|
2019-11-05 19:59:56 +08:00
|
|
|
|
|
|
|
if (map->m_flags & EXT4_MAP_NEW)
|
|
|
|
iomap->flags |= IOMAP_F_NEW;
|
|
|
|
|
2021-11-29 18:22:00 +08:00
|
|
|
if (flags & IOMAP_DAX)
|
|
|
|
iomap->dax_dev = EXT4_SB(inode->i_sb)->s_daxdev;
|
|
|
|
else
|
|
|
|
iomap->bdev = inode->i_sb->s_bdev;
|
2019-11-05 19:59:56 +08:00
|
|
|
iomap->offset = (u64) map->m_lblk << blkbits;
|
|
|
|
iomap->length = (u64) map->m_len << blkbits;
|
2016-11-21 06:36:06 +08:00
|
|
|
|
2020-02-28 17:26:54 +08:00
|
|
|
if ((map->m_flags & EXT4_MAP_MAPPED) &&
|
|
|
|
!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
|
|
|
iomap->flags |= IOMAP_F_MERGED;
|
|
|
|
|
2019-11-05 19:59:56 +08:00
|
|
|
/*
|
|
|
|
* Flags passed to ext4_map_blocks() for direct I/O writes can result
|
|
|
|
* in m_flags having both EXT4_MAP_MAPPED and EXT4_MAP_UNWRITTEN bits
|
|
|
|
* set. In order for any allocated unwritten extents to be converted
|
|
|
|
* into written extents correctly within the ->end_io() handler, we
|
|
|
|
* need to ensure that the iomap->type is set appropriately. Hence, the
|
|
|
|
* reason why we need to check whether the EXT4_MAP_UNWRITTEN bit has
|
|
|
|
* been set first.
|
|
|
|
*/
|
|
|
|
if (map->m_flags & EXT4_MAP_UNWRITTEN) {
|
|
|
|
iomap->type = IOMAP_UNWRITTEN;
|
|
|
|
iomap->addr = (u64) map->m_pblk << blkbits;
|
2021-11-29 18:22:00 +08:00
|
|
|
if (flags & IOMAP_DAX)
|
|
|
|
iomap->addr += EXT4_SB(inode->i_sb)->s_dax_part_off;
|
2019-11-05 19:59:56 +08:00
|
|
|
} else if (map->m_flags & EXT4_MAP_MAPPED) {
|
|
|
|
iomap->type = IOMAP_MAPPED;
|
|
|
|
iomap->addr = (u64) map->m_pblk << blkbits;
|
2021-11-29 18:22:00 +08:00
|
|
|
if (flags & IOMAP_DAX)
|
|
|
|
iomap->addr += EXT4_SB(inode->i_sb)->s_dax_part_off;
|
2016-11-21 06:36:06 +08:00
|
|
|
} else {
|
2019-11-05 19:59:56 +08:00
|
|
|
iomap->type = IOMAP_HOLE;
|
|
|
|
iomap->addr = IOMAP_NULL_ADDR;
|
2016-11-21 06:36:06 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-11-05 20:00:14 +08:00
|
|
|
static int ext4_iomap_alloc(struct inode *inode, struct ext4_map_blocks *map,
|
|
|
|
unsigned int flags)
|
2016-11-21 07:09:11 +08:00
|
|
|
{
|
|
|
|
handle_t *handle;
|
2019-11-05 20:02:39 +08:00
|
|
|
u8 blkbits = inode->i_blkbits;
|
|
|
|
int ret, dio_credits, m_flags = 0, retries = 0;
|
2016-11-21 07:09:11 +08:00
|
|
|
|
|
|
|
/*
|
2019-11-05 20:00:14 +08:00
|
|
|
* Trim the mapping request to the maximum value that we can map at
|
|
|
|
* once for direct I/O.
|
2016-11-21 07:09:11 +08:00
|
|
|
*/
|
2019-11-05 20:00:14 +08:00
|
|
|
if (map->m_len > DIO_MAX_BLOCKS)
|
|
|
|
map->m_len = DIO_MAX_BLOCKS;
|
|
|
|
dio_credits = ext4_chunk_trans_blocks(inode, map->m_len);
|
2016-11-21 07:09:11 +08:00
|
|
|
|
2019-11-05 20:00:14 +08:00
|
|
|
retry:
|
2016-11-21 07:09:11 +08:00
|
|
|
/*
|
2019-11-05 20:00:14 +08:00
|
|
|
* Either we allocate blocks and then don't get an unwritten extent, so
|
|
|
|
* in that case we have reserved enough credits. Or, the blocks are
|
|
|
|
* already allocated and unwritten. In that case, the extent conversion
|
|
|
|
* fits into the credits as well.
|
2016-11-21 07:09:11 +08:00
|
|
|
*/
|
2019-11-05 20:00:14 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, dio_credits);
|
|
|
|
if (IS_ERR(handle))
|
|
|
|
return PTR_ERR(handle);
|
2009-09-29 03:48:41 +08:00
|
|
|
|
2019-11-05 20:02:39 +08:00
|
|
|
/*
|
|
|
|
* DAX and direct I/O are the only two operations that are currently
|
|
|
|
* supported with IOMAP_WRITE.
|
|
|
|
*/
|
2021-11-29 18:21:58 +08:00
|
|
|
WARN_ON(!(flags & (IOMAP_DAX | IOMAP_DIRECT)));
|
|
|
|
if (flags & IOMAP_DAX)
|
2019-11-05 20:02:39 +08:00
|
|
|
m_flags = EXT4_GET_BLOCKS_CREATE_ZERO;
|
|
|
|
/*
|
|
|
|
* We use i_size instead of i_disksize here because delalloc writeback
|
|
|
|
* can complete at any point during the I/O and subsequently push the
|
|
|
|
* i_disksize out to i_size. This could be beyond where direct I/O is
|
|
|
|
* happening and thus expose allocated blocks to direct I/O reads.
|
|
|
|
*/
|
2021-04-12 18:23:33 +08:00
|
|
|
else if (((loff_t)map->m_lblk << blkbits) >= i_size_read(inode))
|
2019-11-05 20:02:39 +08:00
|
|
|
m_flags = EXT4_GET_BLOCKS_CREATE;
|
|
|
|
else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
|
|
|
m_flags = EXT4_GET_BLOCKS_IO_CREATE_EXT;
|
2009-11-04 03:44:54 +08:00
|
|
|
|
2019-11-05 20:02:39 +08:00
|
|
|
ret = ext4_map_blocks(handle, inode, map, m_flags);
|
2009-09-29 03:48:29 +08:00
|
|
|
|
2016-02-29 05:36:38 +08:00
|
|
|
/*
|
2019-11-05 20:02:39 +08:00
|
|
|
* We cannot fill holes in indirect tree based inodes as that could
|
|
|
|
* expose stale data in the case of a crash. Use the magic error code
|
|
|
|
* to fallback to buffered I/O.
|
2016-02-29 05:36:38 +08:00
|
|
|
*/
|
2019-11-05 20:02:39 +08:00
|
|
|
if (!m_flags && !ret)
|
|
|
|
ret = -ENOTBLK;
|
2016-02-08 11:40:51 +08:00
|
|
|
|
2019-11-05 20:00:14 +08:00
|
|
|
ext4_journal_stop(handle);
|
|
|
|
if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
|
|
|
|
goto retry;
|
|
|
|
|
|
|
|
return ret;
|
2009-09-29 03:48:41 +08:00
|
|
|
}
|
2010-03-03 02:28:44 +08:00
|
|
|
|
2019-11-05 20:00:14 +08:00
|
|
|
|
2016-11-21 06:36:06 +08:00
|
|
|
static int ext4_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
|
2019-10-19 07:44:10 +08:00
|
|
|
unsigned flags, struct iomap *iomap, struct iomap *srcmap)
|
2009-09-29 03:48:41 +08:00
|
|
|
{
|
2016-11-21 06:36:06 +08:00
|
|
|
int ret;
|
2019-11-05 20:03:31 +08:00
|
|
|
struct ext4_map_blocks map;
|
|
|
|
u8 blkbits = inode->i_blkbits;
|
2012-07-10 04:29:29 +08:00
|
|
|
|
2018-09-02 00:45:04 +08:00
|
|
|
if ((offset >> blkbits) > EXT4_MAX_LOGICAL_BLOCK)
|
|
|
|
return -EINVAL;
|
2012-07-23 08:19:31 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
if (WARN_ON_ONCE(ext4_has_inline_data(inode)))
|
|
|
|
return -ERANGE;
|
2012-07-23 08:19:31 +08:00
|
|
|
|
2013-06-05 02:27:38 +08:00
|
|
|
/*
|
2019-11-05 20:03:31 +08:00
|
|
|
* Calculate the first and last logical blocks respectively.
|
2013-06-05 02:27:38 +08:00
|
|
|
*/
|
2019-11-05 20:03:31 +08:00
|
|
|
map.m_lblk = offset >> blkbits;
|
|
|
|
map.m_len = min_t(loff_t, (offset + length - 1) >> blkbits,
|
|
|
|
EXT4_MAX_LOGICAL_BLOCK) - map.m_lblk + 1;
|
2013-06-05 02:27:38 +08:00
|
|
|
|
2020-09-18 13:06:35 +08:00
|
|
|
if (flags & IOMAP_WRITE) {
|
|
|
|
/*
|
|
|
|
* We check here if the blocks are already allocated, then we
|
|
|
|
* don't need to start a journal txn and we can directly return
|
|
|
|
* the mapping information. This could boost performance
|
|
|
|
* especially in multi-threaded overwrite requests.
|
|
|
|
*/
|
|
|
|
if (offset + length <= i_size_read(inode)) {
|
|
|
|
ret = ext4_map_blocks(NULL, inode, &map, 0);
|
|
|
|
if (ret > 0 && (map.m_flags & EXT4_MAP_MAPPED))
|
|
|
|
goto out;
|
|
|
|
}
|
2019-11-05 20:00:14 +08:00
|
|
|
ret = ext4_iomap_alloc(inode, &map, flags);
|
2020-09-18 13:06:35 +08:00
|
|
|
} else {
|
2017-10-02 05:58:54 +08:00
|
|
|
ret = ext4_map_blocks(NULL, inode, &map, 0);
|
2020-09-18 13:06:35 +08:00
|
|
|
}
|
2012-07-23 08:19:31 +08:00
|
|
|
|
2019-11-05 20:00:14 +08:00
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
2020-09-18 13:06:35 +08:00
|
|
|
out:
|
ext4: support direct I/O with fscrypt using blk-crypto
Encrypted files traditionally haven't supported DIO, due to the need to
encrypt/decrypt the data. However, when the encryption is implemented
using inline encryption (blk-crypto) instead of the traditional
filesystem-layer encryption, it is straightforward to support DIO.
Therefore, make ext4 support DIO on files that are using inline
encryption. Since ext4 uses iomap for DIO, and fscrypt support was
already added to iomap DIO, this just requires two small changes:
- Let DIO proceed when supported, by checking fscrypt_dio_supported()
instead of assuming that encrypted files never support DIO.
- In ext4_iomap_begin(), use fscrypt_limit_io_blocks() to limit the
length of the mapping in the rare case where a DUN discontiguity
occurs in the middle of an extent. The iomap DIO implementation
requires this, since it assumes that it can submit a bio covering (up
to) the whole mapping, without checking fscrypt constraints itself.
Co-developed-by: Satya Tangirala <satyat@google.com>
Signed-off-by: Satya Tangirala <satyat@google.com>
Acked-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jaegeuk Kim <jaegeuk@kernel.org>
Link: https://lore.kernel.org/r/20220128233940.79464-4-ebiggers@kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
2022-01-29 07:39:38 +08:00
|
|
|
/*
|
|
|
|
* When inline encryption is enabled, sometimes I/O to an encrypted file
|
|
|
|
* has to be broken up to guarantee DUN contiguity. Handle this by
|
|
|
|
* limiting the length of the mapping returned.
|
|
|
|
*/
|
|
|
|
map.m_len = fscrypt_limit_io_blocks(inode, map.m_lblk, map.m_len);
|
|
|
|
|
2021-11-29 18:22:00 +08:00
|
|
|
ext4_set_iomap(inode, iomap, &map, offset, length, flags);
|
2012-07-23 08:19:31 +08:00
|
|
|
|
2016-11-21 06:36:06 +08:00
|
|
|
return 0;
|
|
|
|
}
|
2009-09-29 03:48:29 +08:00
|
|
|
|
2019-12-19 01:44:33 +08:00
|
|
|
static int ext4_iomap_overwrite_begin(struct inode *inode, loff_t offset,
|
|
|
|
loff_t length, unsigned flags, struct iomap *iomap,
|
|
|
|
struct iomap *srcmap)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Even for writes we don't need to allocate blocks, so just pretend
|
|
|
|
* we are reading to save overhead of starting a transaction.
|
|
|
|
*/
|
|
|
|
flags &= ~IOMAP_WRITE;
|
|
|
|
ret = ext4_iomap_begin(inode, offset, length, flags, iomap, srcmap);
|
|
|
|
WARN_ON_ONCE(iomap->type != IOMAP_MAPPED);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-11-21 07:09:11 +08:00
|
|
|
static int ext4_iomap_end(struct inode *inode, loff_t offset, loff_t length,
|
|
|
|
ssize_t written, unsigned flags, struct iomap *iomap)
|
|
|
|
{
|
2012-11-30 10:13:48 +08:00
|
|
|
/*
|
2019-11-05 20:02:39 +08:00
|
|
|
* Check to see whether an error occurred while writing out the data to
|
|
|
|
* the allocated blocks. If so, return the magic error code so that we
|
|
|
|
* fallback to buffered I/O and attempt to complete the remainder of
|
|
|
|
* the I/O. Any blocks that may have been allocated in preparation for
|
|
|
|
* the direct I/O will be reused during buffered I/O.
|
2012-11-30 10:13:48 +08:00
|
|
|
*/
|
2019-11-05 20:02:39 +08:00
|
|
|
if (flags & (IOMAP_WRITE | IOMAP_DIRECT) && written == 0)
|
|
|
|
return -ENOTBLK;
|
2012-11-30 10:13:48 +08:00
|
|
|
|
2019-11-05 20:01:51 +08:00
|
|
|
return 0;
|
2016-11-21 07:09:11 +08:00
|
|
|
}
|
2012-07-23 08:19:31 +08:00
|
|
|
|
2017-01-28 15:20:26 +08:00
|
|
|
const struct iomap_ops ext4_iomap_ops = {
|
2016-11-21 06:36:06 +08:00
|
|
|
.iomap_begin = ext4_iomap_begin,
|
2016-11-21 07:09:11 +08:00
|
|
|
.iomap_end = ext4_iomap_end,
|
2016-11-21 06:36:06 +08:00
|
|
|
};
|
2009-09-29 03:48:29 +08:00
|
|
|
|
2019-12-19 01:44:33 +08:00
|
|
|
const struct iomap_ops ext4_iomap_overwrite_ops = {
|
|
|
|
.iomap_begin = ext4_iomap_overwrite_begin,
|
|
|
|
.iomap_end = ext4_iomap_end,
|
|
|
|
};
|
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
static bool ext4_iomap_is_delalloc(struct inode *inode,
|
|
|
|
struct ext4_map_blocks *map)
|
|
|
|
{
|
|
|
|
struct extent_status es;
|
|
|
|
ext4_lblk_t offset = 0, end = map->m_lblk + map->m_len - 1;
|
2016-05-13 12:44:16 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
ext4_es_find_extent_range(inode, &ext4_es_is_delayed,
|
|
|
|
map->m_lblk, end, &es);
|
2016-05-13 12:44:16 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
if (!es.es_len || es.es_lblk > end)
|
|
|
|
return false;
|
2016-05-13 12:44:16 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
if (es.es_lblk > map->m_lblk) {
|
|
|
|
map->m_len = es.es_lblk - map->m_lblk;
|
|
|
|
return false;
|
2016-05-13 12:44:16 +08:00
|
|
|
}
|
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
offset = map->m_lblk - es.es_lblk;
|
|
|
|
map->m_len = es.es_len - offset;
|
2016-05-13 12:44:16 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
return true;
|
2009-09-29 03:48:41 +08:00
|
|
|
}
|
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
static int ext4_iomap_begin_report(struct inode *inode, loff_t offset,
|
|
|
|
loff_t length, unsigned int flags,
|
|
|
|
struct iomap *iomap, struct iomap *srcmap)
|
2009-09-29 03:48:41 +08:00
|
|
|
{
|
2019-11-05 20:03:31 +08:00
|
|
|
int ret;
|
|
|
|
bool delalloc = false;
|
|
|
|
struct ext4_map_blocks map;
|
|
|
|
u8 blkbits = inode->i_blkbits;
|
2009-09-29 03:48:41 +08:00
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
if ((offset >> blkbits) > EXT4_MAX_LOGICAL_BLOCK)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (ext4_has_inline_data(inode)) {
|
|
|
|
ret = ext4_inline_data_iomap(inode, iomap);
|
|
|
|
if (ret != -EAGAIN) {
|
|
|
|
if (ret == 0 && offset >= iomap->length)
|
|
|
|
ret = -ENOENT;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
2015-04-12 12:55:10 +08:00
|
|
|
|
2011-08-31 23:56:51 +08:00
|
|
|
/*
|
2019-11-05 20:03:31 +08:00
|
|
|
* Calculate the first and last logical block respectively.
|
2011-08-31 23:56:51 +08:00
|
|
|
*/
|
2019-11-05 20:03:31 +08:00
|
|
|
map.m_lblk = offset >> blkbits;
|
|
|
|
map.m_len = min_t(loff_t, (offset + length - 1) >> blkbits,
|
|
|
|
EXT4_MAX_LOGICAL_BLOCK) - map.m_lblk + 1;
|
2011-08-31 23:56:51 +08:00
|
|
|
|
2020-02-28 17:26:57 +08:00
|
|
|
/*
|
|
|
|
* Fiemap callers may call for offset beyond s_bitmap_maxbytes.
|
|
|
|
* So handle it here itself instead of querying ext4_map_blocks().
|
|
|
|
* Since ext4_map_blocks() will warn about it and will return
|
|
|
|
* -EIO error.
|
|
|
|
*/
|
|
|
|
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
|
|
|
|
if (offset >= sbi->s_bitmap_maxbytes) {
|
|
|
|
map.m_flags = 0;
|
|
|
|
goto set_iomap;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
ret = ext4_map_blocks(NULL, inode, &map, 0);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
if (ret == 0)
|
|
|
|
delalloc = ext4_iomap_is_delalloc(inode, &map);
|
2012-12-11 03:04:52 +08:00
|
|
|
|
2020-02-28 17:26:57 +08:00
|
|
|
set_iomap:
|
2021-11-29 18:22:00 +08:00
|
|
|
ext4_set_iomap(inode, iomap, &map, offset, length, flags);
|
2019-11-05 20:03:31 +08:00
|
|
|
if (delalloc && iomap->type == IOMAP_HOLE)
|
|
|
|
iomap->type = IOMAP_DELALLOC;
|
|
|
|
|
|
|
|
return 0;
|
2009-09-29 03:48:41 +08:00
|
|
|
}
|
|
|
|
|
2019-11-05 20:03:31 +08:00
|
|
|
const struct iomap_ops ext4_iomap_report_ops = {
|
|
|
|
.iomap_begin = ext4_iomap_begin_report,
|
|
|
|
};
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2022-03-23 09:26:56 +08:00
|
|
|
* Whenever the folio is being dirtied, corresponding buffers should already
|
|
|
|
* be attached to the transaction (we take care of this in ext4_page_mkwrite()
|
|
|
|
* and ext4_write_begin()). However we cannot move buffers to dirty transaction
|
|
|
|
* lists here because ->dirty_folio is called under VFS locks and the folio
|
2022-03-10 18:18:32 +08:00
|
|
|
* is not necessarily locked.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2022-02-10 04:22:03 +08:00
|
|
|
* We cannot just dirty the folio and leave attached buffers clean, because the
|
2006-10-11 16:20:50 +08:00
|
|
|
* buffers' dirty state is "definitive". We cannot just set the buffers dirty
|
|
|
|
* or jbddirty because all the journalling code will explode.
|
|
|
|
*
|
2022-02-10 04:22:03 +08:00
|
|
|
* So what we do is to mark the folio "pending dirty" and next time writepage
|
2006-10-11 16:20:50 +08:00
|
|
|
* is called, propagate that into the buffers appropriately.
|
|
|
|
*/
|
2022-02-10 04:22:03 +08:00
|
|
|
static bool ext4_journalled_dirty_folio(struct address_space *mapping,
|
|
|
|
struct folio *folio)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2022-03-30 04:21:53 +08:00
|
|
|
WARN_ON_ONCE(!folio_buffers(folio));
|
2022-02-10 04:22:03 +08:00
|
|
|
folio_set_checked(folio);
|
|
|
|
return filemap_dirty_folio(mapping, folio);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2022-02-10 04:22:12 +08:00
|
|
|
static bool ext4_dirty_folio(struct address_space *mapping, struct folio *folio)
|
2016-12-02 00:46:40 +08:00
|
|
|
{
|
2022-02-10 04:22:12 +08:00
|
|
|
WARN_ON_ONCE(!folio_test_locked(folio) && !folio_test_dirty(folio));
|
|
|
|
WARN_ON_ONCE(!folio_buffers(folio));
|
|
|
|
return block_dirty_folio(mapping, folio);
|
2016-12-02 00:46:40 +08:00
|
|
|
}
|
|
|
|
|
2020-09-04 17:16:53 +08:00
|
|
|
static int ext4_iomap_swap_activate(struct swap_info_struct *sis,
|
|
|
|
struct file *file, sector_t *span)
|
|
|
|
{
|
|
|
|
return iomap_swapfile_activate(sis, file, span,
|
|
|
|
&ext4_iomap_report_ops);
|
|
|
|
}
|
|
|
|
|
2013-04-04 00:39:17 +08:00
|
|
|
static const struct address_space_operations ext4_aops = {
|
2022-04-29 23:12:16 +08:00
|
|
|
.read_folio = ext4_read_folio,
|
2020-06-02 12:47:16 +08:00
|
|
|
.readahead = ext4_readahead,
|
2009-06-15 05:58:45 +08:00
|
|
|
.writepage = ext4_writepage,
|
2013-06-07 02:00:46 +08:00
|
|
|
.writepages = ext4_writepages,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.write_begin = ext4_write_begin,
|
2013-04-04 00:39:17 +08:00
|
|
|
.write_end = ext4_write_end,
|
2022-02-10 04:22:12 +08:00
|
|
|
.dirty_folio = ext4_dirty_folio,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.bmap = ext4_bmap,
|
2022-02-10 04:21:34 +08:00
|
|
|
.invalidate_folio = ext4_invalidate_folio,
|
2022-05-01 11:33:38 +08:00
|
|
|
.release_folio = ext4_release_folio,
|
2019-11-05 20:02:39 +08:00
|
|
|
.direct_IO = noop_direct_IO,
|
2022-06-06 22:20:31 +08:00
|
|
|
.migrate_folio = buffer_migrate_folio,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.is_partially_uptodate = block_is_partially_uptodate,
|
2009-09-16 17:50:16 +08:00
|
|
|
.error_remove_page = generic_error_remove_page,
|
2020-09-04 17:16:53 +08:00
|
|
|
.swap_activate = ext4_iomap_swap_activate,
|
2006-10-11 16:20:50 +08:00
|
|
|
};
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
static const struct address_space_operations ext4_journalled_aops = {
|
2022-04-29 23:12:16 +08:00
|
|
|
.read_folio = ext4_read_folio,
|
2020-06-02 12:47:16 +08:00
|
|
|
.readahead = ext4_readahead,
|
2009-06-15 05:58:45 +08:00
|
|
|
.writepage = ext4_writepage,
|
2013-06-07 02:00:46 +08:00
|
|
|
.writepages = ext4_writepages,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.write_begin = ext4_write_begin,
|
|
|
|
.write_end = ext4_journalled_write_end,
|
2022-02-10 04:22:03 +08:00
|
|
|
.dirty_folio = ext4_journalled_dirty_folio,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.bmap = ext4_bmap,
|
2022-02-10 04:21:43 +08:00
|
|
|
.invalidate_folio = ext4_journalled_invalidate_folio,
|
2022-05-01 11:33:38 +08:00
|
|
|
.release_folio = ext4_release_folio,
|
2019-11-05 20:02:39 +08:00
|
|
|
.direct_IO = noop_direct_IO,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.is_partially_uptodate = block_is_partially_uptodate,
|
2009-09-16 17:50:16 +08:00
|
|
|
.error_remove_page = generic_error_remove_page,
|
2020-09-04 17:16:53 +08:00
|
|
|
.swap_activate = ext4_iomap_swap_activate,
|
2006-10-11 16:20:50 +08:00
|
|
|
};
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
static const struct address_space_operations ext4_da_aops = {
|
2022-04-29 23:12:16 +08:00
|
|
|
.read_folio = ext4_read_folio,
|
2020-06-02 12:47:16 +08:00
|
|
|
.readahead = ext4_readahead,
|
2009-06-15 05:58:45 +08:00
|
|
|
.writepage = ext4_writepage,
|
2013-06-07 02:00:46 +08:00
|
|
|
.writepages = ext4_writepages,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.write_begin = ext4_da_write_begin,
|
|
|
|
.write_end = ext4_da_write_end,
|
2022-02-10 04:22:12 +08:00
|
|
|
.dirty_folio = ext4_dirty_folio,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.bmap = ext4_bmap,
|
2022-02-10 04:21:34 +08:00
|
|
|
.invalidate_folio = ext4_invalidate_folio,
|
2022-05-01 11:33:38 +08:00
|
|
|
.release_folio = ext4_release_folio,
|
2019-11-05 20:02:39 +08:00
|
|
|
.direct_IO = noop_direct_IO,
|
2022-06-06 22:20:31 +08:00
|
|
|
.migrate_folio = buffer_migrate_folio,
|
vfs: pagecache usage optimization for pagesize!=blocksize
When we read some part of a file through pagecache, if there is a
pagecache of corresponding index but this page is not uptodate, read IO
is issued and this page will be uptodate.
I think this is good for pagesize == blocksize environment but there is
room for improvement on pagesize != blocksize environment. Because in
this case a page can have multiple buffers and even if a page is not
uptodate, some buffers can be uptodate.
So I suggest that when all buffers which correspond to a part of a file
that we want to read are uptodate, use this pagecache and copy data from
this pagecache to user buffer even if a page is not uptodate. This can
reduce read IO and improve system throughput.
I wrote a benchmark program and got result number with this program.
This benchmark do:
1: mount and open a test file.
2: create a 512MB file.
3: close a file and umount.
4: mount and again open a test file.
5: pwrite randomly 300000 times on a test file. offset is aligned
by IO size(1024bytes).
6: measure time of preading randomly 100000 times on a test file.
The result was:
2.6.26
330 sec
2.6.26-patched
226 sec
Arch:i386
Filesystem:ext3
Blocksize:1024 bytes
Memory: 1GB
On ext3/4, a file is written through buffer/block. So random read/write
mixed workloads or random read after random write workloads are optimized
with this patch under pagesize != blocksize environment. This test result
showed this.
The benchmark program is as follows:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <time.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mount.h>
#define LEN 1024
#define LOOP 1024*512 /* 512MB */
main(void)
{
unsigned long i, offset, filesize;
int fd;
char buf[LEN];
time_t t1, t2;
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
memset(buf, 0, LEN);
fd = open("/root/test1/testfile", O_CREAT|O_RDWR|O_TRUNC);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
for (i = 0; i < LOOP; i++)
write(fd, buf, LEN);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
if (mount("/dev/sda1", "/root/test1/", "ext3", 0, 0) < 0) {
perror("cannot mount\n");
exit(1);
}
fd = open("/root/test1/testfile", O_RDWR);
if (fd < 0) {
perror("cannot open file\n");
exit(1);
}
filesize = LEN * LOOP;
for (i = 0; i < 300000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pwrite(fd, buf, LEN, offset);
}
printf("start test\n");
time(&t1);
for (i = 0; i < 100000; i++){
offset = (random() % filesize) & (~(LEN - 1));
pread(fd, buf, LEN, offset);
}
time(&t2);
printf("%ld sec\n", t2-t1);
close(fd);
if (umount("/root/test1/") < 0) {
perror("cannot umount\n");
exit(1);
}
}
Signed-off-by: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@ucw.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-29 06:46:36 +08:00
|
|
|
.is_partially_uptodate = block_is_partially_uptodate,
|
2009-09-16 17:50:16 +08:00
|
|
|
.error_remove_page = generic_error_remove_page,
|
2020-09-04 17:16:53 +08:00
|
|
|
.swap_activate = ext4_iomap_swap_activate,
|
2008-07-12 07:27:31 +08:00
|
|
|
};
|
|
|
|
|
2017-12-22 04:25:11 +08:00
|
|
|
static const struct address_space_operations ext4_dax_aops = {
|
|
|
|
.writepages = ext4_dax_writepages,
|
|
|
|
.direct_IO = noop_direct_IO,
|
2022-02-10 04:22:13 +08:00
|
|
|
.dirty_folio = noop_dirty_folio,
|
2018-09-16 09:23:41 +08:00
|
|
|
.bmap = ext4_bmap,
|
2020-09-04 17:16:53 +08:00
|
|
|
.swap_activate = ext4_iomap_swap_activate,
|
2017-12-22 04:25:11 +08:00
|
|
|
};
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
void ext4_set_aops(struct inode *inode)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2012-02-21 06:53:00 +08:00
|
|
|
switch (ext4_inode_journal_mode(inode)) {
|
|
|
|
case EXT4_INODE_ORDERED_DATA_MODE:
|
|
|
|
case EXT4_INODE_WRITEBACK_DATA_MODE:
|
|
|
|
break;
|
|
|
|
case EXT4_INODE_JOURNAL_DATA_MODE:
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_mapping->a_ops = &ext4_journalled_aops;
|
2013-04-04 00:39:17 +08:00
|
|
|
return;
|
2012-02-21 06:53:00 +08:00
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
2017-12-22 04:25:11 +08:00
|
|
|
if (IS_DAX(inode))
|
|
|
|
inode->i_mapping->a_ops = &ext4_dax_aops;
|
|
|
|
else if (test_opt(inode->i_sb, DELALLOC))
|
2013-04-04 00:39:17 +08:00
|
|
|
inode->i_mapping->a_ops = &ext4_da_aops;
|
|
|
|
else
|
|
|
|
inode->i_mapping->a_ops = &ext4_aops;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2015-02-17 07:59:38 +08:00
|
|
|
static int __ext4_block_zero_page_range(handle_t *handle,
|
2013-05-28 11:32:35 +08:00
|
|
|
struct address_space *mapping, loff_t from, loff_t length)
|
|
|
|
{
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
ext4_fsblk_t index = from >> PAGE_SHIFT;
|
|
|
|
unsigned offset = from & (PAGE_SIZE-1);
|
2015-02-17 07:59:38 +08:00
|
|
|
unsigned blocksize, pos;
|
2013-05-28 11:32:35 +08:00
|
|
|
ext4_lblk_t iblock;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
struct buffer_head *bh;
|
|
|
|
struct page *page;
|
|
|
|
int err = 0;
|
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
page = find_or_create_page(mapping, from >> PAGE_SHIFT,
|
2015-11-07 08:28:49 +08:00
|
|
|
mapping_gfp_constraint(mapping, ~__GFP_FS));
|
2013-05-28 11:32:35 +08:00
|
|
|
if (!page)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
blocksize = inode->i_sb->s_blocksize;
|
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
iblock = index << (PAGE_SHIFT - inode->i_sb->s_blocksize_bits);
|
2013-05-28 11:32:35 +08:00
|
|
|
|
|
|
|
if (!page_has_buffers(page))
|
|
|
|
create_empty_buffers(page, blocksize, 0);
|
|
|
|
|
|
|
|
/* Find the buffer that contains "offset" */
|
|
|
|
bh = page_buffers(page);
|
|
|
|
pos = blocksize;
|
|
|
|
while (offset >= pos) {
|
|
|
|
bh = bh->b_this_page;
|
|
|
|
iblock++;
|
|
|
|
pos += blocksize;
|
|
|
|
}
|
|
|
|
if (buffer_freed(bh)) {
|
|
|
|
BUFFER_TRACE(bh, "freed: skip");
|
|
|
|
goto unlock;
|
|
|
|
}
|
|
|
|
if (!buffer_mapped(bh)) {
|
|
|
|
BUFFER_TRACE(bh, "unmapped");
|
|
|
|
ext4_get_block(inode, iblock, bh, 0);
|
|
|
|
/* unmapped? It's a hole - nothing to do */
|
|
|
|
if (!buffer_mapped(bh)) {
|
|
|
|
BUFFER_TRACE(bh, "still unmapped");
|
|
|
|
goto unlock;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Ok, it's mapped. Make sure it's up-to-date */
|
|
|
|
if (PageUptodate(page))
|
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
|
|
|
|
if (!buffer_uptodate(bh)) {
|
2020-09-24 15:33:33 +08:00
|
|
|
err = ext4_read_bh_lock(bh, 0, true);
|
|
|
|
if (err)
|
2013-05-28 11:32:35 +08:00
|
|
|
goto unlock;
|
2020-07-02 09:56:07 +08:00
|
|
|
if (fscrypt_inode_uses_fs_layer_crypto(inode)) {
|
2015-04-12 12:56:10 +08:00
|
|
|
/* We expect the key to be set. */
|
2016-07-11 02:01:03 +08:00
|
|
|
BUG_ON(!fscrypt_has_encryption_key(inode));
|
2019-12-26 23:41:05 +08:00
|
|
|
err = fscrypt_decrypt_pagecache_blocks(page, blocksize,
|
|
|
|
bh_offset(bh));
|
|
|
|
if (err) {
|
|
|
|
clear_buffer_uptodate(bh);
|
|
|
|
goto unlock;
|
|
|
|
}
|
2015-04-12 12:56:10 +08:00
|
|
|
}
|
2013-05-28 11:32:35 +08:00
|
|
|
}
|
|
|
|
if (ext4_should_journal_data(inode)) {
|
|
|
|
BUFFER_TRACE(bh, "get write access");
|
2021-08-16 17:57:04 +08:00
|
|
|
err = ext4_journal_get_write_access(handle, inode->i_sb, bh,
|
|
|
|
EXT4_JTR_NONE);
|
2013-05-28 11:32:35 +08:00
|
|
|
if (err)
|
|
|
|
goto unlock;
|
|
|
|
}
|
|
|
|
zero_user(page, offset, length);
|
|
|
|
BUFFER_TRACE(bh, "zeroed end of block");
|
|
|
|
|
|
|
|
if (ext4_should_journal_data(inode)) {
|
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, bh);
|
2013-05-28 11:32:35 +08:00
|
|
|
} else {
|
2013-07-01 20:12:39 +08:00
|
|
|
err = 0;
|
2013-05-28 11:32:35 +08:00
|
|
|
mark_buffer_dirty(bh);
|
2016-04-24 12:56:05 +08:00
|
|
|
if (ext4_should_order_data(inode))
|
2019-06-21 05:26:26 +08:00
|
|
|
err = ext4_jbd2_inode_add_write(handle, inode, from,
|
|
|
|
length);
|
2013-05-28 11:32:35 +08:00
|
|
|
}
|
2013-05-28 11:32:35 +08:00
|
|
|
|
|
|
|
unlock:
|
|
|
|
unlock_page(page);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
put_page(page);
|
2013-05-28 11:32:35 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2015-02-17 07:59:38 +08:00
|
|
|
/*
|
|
|
|
* ext4_block_zero_page_range() zeros out a mapping of length 'length'
|
|
|
|
* starting from file offset 'from'. The range to be zero'd must
|
|
|
|
* be contained with in one block. If the specified range exceeds
|
|
|
|
* the end of the block it will be shortened to end of the block
|
2021-03-27 18:30:05 +08:00
|
|
|
* that corresponds to 'from'
|
2015-02-17 07:59:38 +08:00
|
|
|
*/
|
|
|
|
static int ext4_block_zero_page_range(handle_t *handle,
|
|
|
|
struct address_space *mapping, loff_t from, loff_t length)
|
|
|
|
{
|
|
|
|
struct inode *inode = mapping->host;
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
unsigned offset = from & (PAGE_SIZE-1);
|
2015-02-17 07:59:38 +08:00
|
|
|
unsigned blocksize = inode->i_sb->s_blocksize;
|
|
|
|
unsigned max = blocksize - (offset & (blocksize - 1));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* correct length if it does not fall between
|
|
|
|
* 'from' and the end of the block
|
|
|
|
*/
|
|
|
|
if (length > max || length < 0)
|
|
|
|
length = max;
|
|
|
|
|
2016-11-21 07:08:05 +08:00
|
|
|
if (IS_DAX(inode)) {
|
2021-11-29 18:21:52 +08:00
|
|
|
return dax_zero_range(inode, from, length, NULL,
|
|
|
|
&ext4_iomap_ops);
|
2016-11-21 07:08:05 +08:00
|
|
|
}
|
2015-02-17 07:59:38 +08:00
|
|
|
return __ext4_block_zero_page_range(handle, mapping, from, length);
|
|
|
|
}
|
|
|
|
|
2014-03-25 03:09:16 +08:00
|
|
|
/*
|
|
|
|
* ext4_block_truncate_page() zeroes out a mapping from file offset `from'
|
|
|
|
* up to the end of the block which corresponds to `from'.
|
|
|
|
* This required during truncate. We need to physically zero the tail end
|
|
|
|
* of that block so it doesn't yield old data if the file is later grown.
|
|
|
|
*/
|
2014-05-12 22:50:23 +08:00
|
|
|
static int ext4_block_truncate_page(handle_t *handle,
|
2014-03-25 03:09:16 +08:00
|
|
|
struct address_space *mapping, loff_t from)
|
|
|
|
{
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
unsigned offset = from & (PAGE_SIZE-1);
|
2014-03-25 03:09:16 +08:00
|
|
|
unsigned length;
|
|
|
|
unsigned blocksize;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
|
ext4: don't BUG when truncating encrypted inodes on the orphan list
Fix a BUG when the kernel tries to mount a file system constructed as
follows:
echo foo > foo.txt
mke2fs -Fq -t ext4 -O encrypt foo.img 100
debugfs -w foo.img << EOF
write foo.txt a
set_inode_field a i_flags 0x80800
set_super_value s_last_orphan 12
quit
EOF
root@kvm-xfstests:~# mount -o loop foo.img /mnt
[ 160.238770] ------------[ cut here ]------------
[ 160.240106] kernel BUG at /usr/projects/linux/ext4/fs/ext4/inode.c:3874!
[ 160.240106] invalid opcode: 0000 [#1] SMP
[ 160.240106] Modules linked in:
[ 160.240106] CPU: 0 PID: 2547 Comm: mount Tainted: G W 4.10.0-rc3-00034-gcdd33b941b67 #227
[ 160.240106] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.1-1 04/01/2014
[ 160.240106] task: f4518000 task.stack: f47b6000
[ 160.240106] EIP: ext4_block_zero_page_range+0x1a7/0x2b4
[ 160.240106] EFLAGS: 00010246 CPU: 0
[ 160.240106] EAX: 00000001 EBX: f7be4b50 ECX: f47b7dc0 EDX: 00000007
[ 160.240106] ESI: f43b05a8 EDI: f43babec EBP: f47b7dd0 ESP: f47b7dac
[ 160.240106] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
[ 160.240106] CR0: 80050033 CR2: bfd85b08 CR3: 34a00680 CR4: 000006f0
[ 160.240106] Call Trace:
[ 160.240106] ext4_truncate+0x1e9/0x3e5
[ 160.240106] ext4_fill_super+0x286f/0x2b1e
[ 160.240106] ? set_blocksize+0x2e/0x7e
[ 160.240106] mount_bdev+0x114/0x15f
[ 160.240106] ext4_mount+0x15/0x17
[ 160.240106] ? ext4_calculate_overhead+0x39d/0x39d
[ 160.240106] mount_fs+0x58/0x115
[ 160.240106] vfs_kern_mount+0x4b/0xae
[ 160.240106] do_mount+0x671/0x8c3
[ 160.240106] ? _copy_from_user+0x70/0x83
[ 160.240106] ? strndup_user+0x31/0x46
[ 160.240106] SyS_mount+0x57/0x7b
[ 160.240106] do_int80_syscall_32+0x4f/0x61
[ 160.240106] entry_INT80_32+0x2f/0x2f
[ 160.240106] EIP: 0xb76b919e
[ 160.240106] EFLAGS: 00000246 CPU: 0
[ 160.240106] EAX: ffffffda EBX: 08053838 ECX: 08052188 EDX: 080537e8
[ 160.240106] ESI: c0ed0000 EDI: 00000000 EBP: 080537e8 ESP: bfa13660
[ 160.240106] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b
[ 160.240106] Code: 59 8b 00 a8 01 0f 84 09 01 00 00 8b 07 66 25 00 f0 66 3d 00 80 75 61 89 f8 e8 3e e2 ff ff 84 c0 74 56 83 bf 48 02 00 00 00 75 02 <0f> 0b 81 7d e8 00 10 00 00 74 02 0f 0b 8b 43 04 8b 53 08 31 c9
[ 160.240106] EIP: ext4_block_zero_page_range+0x1a7/0x2b4 SS:ESP: 0068:f47b7dac
[ 160.317241] ---[ end trace d6a773a375c810a5 ]---
The problem is that when the kernel tries to truncate an inode in
ext4_truncate(), it tries to clear any on-disk data beyond i_size.
Without the encryption key, it can't do that, and so it triggers a
BUG.
E2fsck does *not* provide this service, and in practice most file
systems have their orphan list processed by e2fsck, so to avoid
crashing, this patch skips this step if we don't have access to the
encryption key (which is the case when processing the orphan list; in
all other cases, we will have the encryption key, or the kernel
wouldn't have allowed the file to be opened).
An open question is whether the fact that e2fsck isn't clearing the
bytes beyond i_size causing problems --- and if we've lived with it
not doing it for so long, can we drop this from the kernel replay of
the orphan list in all cases (not just when we don't have the key for
encrypted inodes).
Addresses-Google-Bug: #35209576
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-02-15 00:31:15 +08:00
|
|
|
/* If we are processing an encrypted inode during orphan list handling */
|
2018-12-12 17:50:10 +08:00
|
|
|
if (IS_ENCRYPTED(inode) && !fscrypt_has_encryption_key(inode))
|
ext4: don't BUG when truncating encrypted inodes on the orphan list
Fix a BUG when the kernel tries to mount a file system constructed as
follows:
echo foo > foo.txt
mke2fs -Fq -t ext4 -O encrypt foo.img 100
debugfs -w foo.img << EOF
write foo.txt a
set_inode_field a i_flags 0x80800
set_super_value s_last_orphan 12
quit
EOF
root@kvm-xfstests:~# mount -o loop foo.img /mnt
[ 160.238770] ------------[ cut here ]------------
[ 160.240106] kernel BUG at /usr/projects/linux/ext4/fs/ext4/inode.c:3874!
[ 160.240106] invalid opcode: 0000 [#1] SMP
[ 160.240106] Modules linked in:
[ 160.240106] CPU: 0 PID: 2547 Comm: mount Tainted: G W 4.10.0-rc3-00034-gcdd33b941b67 #227
[ 160.240106] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.1-1 04/01/2014
[ 160.240106] task: f4518000 task.stack: f47b6000
[ 160.240106] EIP: ext4_block_zero_page_range+0x1a7/0x2b4
[ 160.240106] EFLAGS: 00010246 CPU: 0
[ 160.240106] EAX: 00000001 EBX: f7be4b50 ECX: f47b7dc0 EDX: 00000007
[ 160.240106] ESI: f43b05a8 EDI: f43babec EBP: f47b7dd0 ESP: f47b7dac
[ 160.240106] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
[ 160.240106] CR0: 80050033 CR2: bfd85b08 CR3: 34a00680 CR4: 000006f0
[ 160.240106] Call Trace:
[ 160.240106] ext4_truncate+0x1e9/0x3e5
[ 160.240106] ext4_fill_super+0x286f/0x2b1e
[ 160.240106] ? set_blocksize+0x2e/0x7e
[ 160.240106] mount_bdev+0x114/0x15f
[ 160.240106] ext4_mount+0x15/0x17
[ 160.240106] ? ext4_calculate_overhead+0x39d/0x39d
[ 160.240106] mount_fs+0x58/0x115
[ 160.240106] vfs_kern_mount+0x4b/0xae
[ 160.240106] do_mount+0x671/0x8c3
[ 160.240106] ? _copy_from_user+0x70/0x83
[ 160.240106] ? strndup_user+0x31/0x46
[ 160.240106] SyS_mount+0x57/0x7b
[ 160.240106] do_int80_syscall_32+0x4f/0x61
[ 160.240106] entry_INT80_32+0x2f/0x2f
[ 160.240106] EIP: 0xb76b919e
[ 160.240106] EFLAGS: 00000246 CPU: 0
[ 160.240106] EAX: ffffffda EBX: 08053838 ECX: 08052188 EDX: 080537e8
[ 160.240106] ESI: c0ed0000 EDI: 00000000 EBP: 080537e8 ESP: bfa13660
[ 160.240106] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b
[ 160.240106] Code: 59 8b 00 a8 01 0f 84 09 01 00 00 8b 07 66 25 00 f0 66 3d 00 80 75 61 89 f8 e8 3e e2 ff ff 84 c0 74 56 83 bf 48 02 00 00 00 75 02 <0f> 0b 81 7d e8 00 10 00 00 74 02 0f 0b 8b 43 04 8b 53 08 31 c9
[ 160.240106] EIP: ext4_block_zero_page_range+0x1a7/0x2b4 SS:ESP: 0068:f47b7dac
[ 160.317241] ---[ end trace d6a773a375c810a5 ]---
The problem is that when the kernel tries to truncate an inode in
ext4_truncate(), it tries to clear any on-disk data beyond i_size.
Without the encryption key, it can't do that, and so it triggers a
BUG.
E2fsck does *not* provide this service, and in practice most file
systems have their orphan list processed by e2fsck, so to avoid
crashing, this patch skips this step if we don't have access to the
encryption key (which is the case when processing the orphan list; in
all other cases, we will have the encryption key, or the kernel
wouldn't have allowed the file to be opened).
An open question is whether the fact that e2fsck isn't clearing the
bytes beyond i_size causing problems --- and if we've lived with it
not doing it for so long, can we drop this from the kernel replay of
the orphan list in all cases (not just when we don't have the key for
encrypted inodes).
Addresses-Google-Bug: #35209576
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-02-15 00:31:15 +08:00
|
|
|
return 0;
|
|
|
|
|
2014-03-25 03:09:16 +08:00
|
|
|
blocksize = inode->i_sb->s_blocksize;
|
|
|
|
length = blocksize - (offset & (blocksize - 1));
|
|
|
|
|
|
|
|
return ext4_block_zero_page_range(handle, mapping, from, length);
|
|
|
|
}
|
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
int ext4_zero_partial_blocks(handle_t *handle, struct inode *inode,
|
|
|
|
loff_t lstart, loff_t length)
|
|
|
|
{
|
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
struct address_space *mapping = inode->i_mapping;
|
2013-07-01 20:12:39 +08:00
|
|
|
unsigned partial_start, partial_end;
|
2013-05-28 11:32:35 +08:00
|
|
|
ext4_fsblk_t start, end;
|
|
|
|
loff_t byte_end = (lstart + length - 1);
|
|
|
|
int err = 0;
|
|
|
|
|
2013-07-01 20:12:39 +08:00
|
|
|
partial_start = lstart & (sb->s_blocksize - 1);
|
|
|
|
partial_end = byte_end & (sb->s_blocksize - 1);
|
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
start = lstart >> sb->s_blocksize_bits;
|
|
|
|
end = byte_end >> sb->s_blocksize_bits;
|
|
|
|
|
|
|
|
/* Handle partial zero within the single block */
|
2013-07-01 20:12:39 +08:00
|
|
|
if (start == end &&
|
|
|
|
(partial_start || (partial_end != sb->s_blocksize - 1))) {
|
2013-05-28 11:32:35 +08:00
|
|
|
err = ext4_block_zero_page_range(handle, mapping,
|
|
|
|
lstart, length);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
/* Handle partial zero out on the start of the range */
|
2013-07-01 20:12:39 +08:00
|
|
|
if (partial_start) {
|
2013-05-28 11:32:35 +08:00
|
|
|
err = ext4_block_zero_page_range(handle, mapping,
|
|
|
|
lstart, sb->s_blocksize);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
/* Handle partial zero out on the end of the range */
|
2013-07-01 20:12:39 +08:00
|
|
|
if (partial_end != sb->s_blocksize - 1)
|
2013-05-28 11:32:35 +08:00
|
|
|
err = ext4_block_zero_page_range(handle, mapping,
|
2013-07-01 20:12:39 +08:00
|
|
|
byte_end - partial_end,
|
|
|
|
partial_end + 1);
|
2013-05-28 11:32:35 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
int ext4_can_truncate(struct inode *inode)
|
|
|
|
{
|
|
|
|
if (S_ISREG(inode->i_mode))
|
|
|
|
return 1;
|
|
|
|
if (S_ISDIR(inode->i_mode))
|
|
|
|
return 1;
|
|
|
|
if (S_ISLNK(inode->i_mode))
|
|
|
|
return !ext4_inode_is_fast_symlink(inode);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-12-08 03:34:49 +08:00
|
|
|
/*
|
|
|
|
* We have to make sure i_disksize gets properly updated before we truncate
|
|
|
|
* page cache due to hole punching or zero range. Otherwise i_disksize update
|
|
|
|
* can get lost as it may have been postponed to submission of writeback but
|
|
|
|
* that will never happen after we truncate page cache.
|
|
|
|
*/
|
|
|
|
int ext4_update_disksize_before_punch(struct inode *inode, loff_t offset,
|
|
|
|
loff_t len)
|
|
|
|
{
|
|
|
|
handle_t *handle;
|
2020-04-27 09:34:37 +08:00
|
|
|
int ret;
|
|
|
|
|
2015-12-08 03:34:49 +08:00
|
|
|
loff_t size = i_size_read(inode);
|
|
|
|
|
2016-01-23 04:40:57 +08:00
|
|
|
WARN_ON(!inode_is_locked(inode));
|
2015-12-08 03:34:49 +08:00
|
|
|
if (offset > size || offset + len < size)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (EXT4_I(inode)->i_disksize >= size)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_MISC, 1);
|
|
|
|
if (IS_ERR(handle))
|
|
|
|
return PTR_ERR(handle);
|
|
|
|
ext4_update_i_disksize(inode, size);
|
2020-04-27 09:34:37 +08:00
|
|
|
ret = ext4_mark_inode_dirty(handle, inode);
|
2015-12-08 03:34:49 +08:00
|
|
|
ext4_journal_stop(handle);
|
|
|
|
|
2020-04-27 09:34:37 +08:00
|
|
|
return ret;
|
2015-12-08 03:34:49 +08:00
|
|
|
}
|
|
|
|
|
2021-02-05 01:05:42 +08:00
|
|
|
static void ext4_wait_dax_page(struct inode *inode)
|
2018-07-30 05:00:22 +08:00
|
|
|
{
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(inode->i_mapping);
|
2018-07-30 05:00:22 +08:00
|
|
|
schedule();
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_lock(inode->i_mapping);
|
2018-07-30 05:00:22 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
int ext4_break_layouts(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct page *page;
|
|
|
|
int error;
|
|
|
|
|
2021-02-05 01:05:42 +08:00
|
|
|
if (WARN_ON_ONCE(!rwsem_is_locked(&inode->i_mapping->invalidate_lock)))
|
2018-07-30 05:00:22 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
do {
|
|
|
|
page = dax_layout_busy_page(inode->i_mapping);
|
|
|
|
if (!page)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
error = ___wait_var_event(&page->_refcount,
|
|
|
|
atomic_read(&page->_refcount) == 1,
|
|
|
|
TASK_INTERRUPTIBLE, 0, 0,
|
2021-02-05 01:05:42 +08:00
|
|
|
ext4_wait_dax_page(inode));
|
2018-09-12 01:31:16 +08:00
|
|
|
} while (error == 0);
|
2018-07-30 05:00:22 +08:00
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2011-05-25 19:41:50 +08:00
|
|
|
/*
|
2016-09-22 23:49:38 +08:00
|
|
|
* ext4_punch_hole: punches a hole in a file by releasing the blocks
|
2011-05-25 19:41:50 +08:00
|
|
|
* associated with the given offset and length
|
|
|
|
*
|
|
|
|
* @inode: File inode
|
|
|
|
* @offset: The offset where the hole will begin
|
|
|
|
* @len: The length of the hole
|
|
|
|
*
|
2012-09-02 01:31:09 +08:00
|
|
|
* Returns: 0 on success or negative on failure
|
2011-05-25 19:41:50 +08:00
|
|
|
*/
|
|
|
|
|
2022-03-09 02:50:43 +08:00
|
|
|
int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
|
2011-05-25 19:41:50 +08:00
|
|
|
{
|
2022-03-09 02:50:43 +08:00
|
|
|
struct inode *inode = file_inode(file);
|
2013-04-04 00:45:17 +08:00
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
ext4_lblk_t first_block, stop_block;
|
|
|
|
struct address_space *mapping = inode->i_mapping;
|
2022-04-01 04:05:15 +08:00
|
|
|
loff_t first_block_offset, last_block_offset, max_length;
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2013-04-04 00:45:17 +08:00
|
|
|
handle_t *handle;
|
|
|
|
unsigned int credits;
|
2020-04-27 09:34:37 +08:00
|
|
|
int ret = 0, ret2 = 0;
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2014-03-19 06:05:35 +08:00
|
|
|
trace_ext4_punch_hole(inode, offset, length, 0);
|
2013-01-17 09:21:26 +08:00
|
|
|
|
2013-04-04 00:45:17 +08:00
|
|
|
/*
|
|
|
|
* Write out all dirty pages to avoid race conditions
|
|
|
|
* Then release them.
|
|
|
|
*/
|
2016-09-22 23:49:38 +08:00
|
|
|
if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
|
2013-04-04 00:45:17 +08:00
|
|
|
ret = filemap_write_and_wait_range(mapping, offset,
|
|
|
|
offset + length - 1);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-01-23 04:40:57 +08:00
|
|
|
inode_lock(inode);
|
2014-04-12 21:47:00 +08:00
|
|
|
|
2013-04-04 00:45:17 +08:00
|
|
|
/* No need to punch hole beyond i_size */
|
|
|
|
if (offset >= inode->i_size)
|
|
|
|
goto out_mutex;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the hole extends beyond i_size, set the hole
|
|
|
|
* to end after the page that contains i_size
|
|
|
|
*/
|
|
|
|
if (offset + length > inode->i_size) {
|
|
|
|
length = inode->i_size +
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
PAGE_SIZE - (inode->i_size & (PAGE_SIZE - 1)) -
|
2013-04-04 00:45:17 +08:00
|
|
|
offset;
|
|
|
|
}
|
|
|
|
|
2022-04-01 04:05:15 +08:00
|
|
|
/*
|
|
|
|
* For punch hole the length + offset needs to be within one block
|
|
|
|
* before last range. Adjust the length if it goes beyond that limit.
|
|
|
|
*/
|
|
|
|
max_length = sbi->s_bitmap_maxbytes - inode->i_sb->s_blocksize;
|
|
|
|
if (offset + length > max_length)
|
|
|
|
length = max_length - offset;
|
|
|
|
|
2013-08-17 09:19:41 +08:00
|
|
|
if (offset & (sb->s_blocksize - 1) ||
|
|
|
|
(offset + length) & (sb->s_blocksize - 1)) {
|
|
|
|
/*
|
|
|
|
* Attach jinode to inode for jbd2 if we do any zeroing of
|
|
|
|
* partial block
|
|
|
|
*/
|
|
|
|
ret = ext4_inode_attach_jinode(inode);
|
|
|
|
if (ret < 0)
|
|
|
|
goto out_mutex;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2022-01-21 15:06:11 +08:00
|
|
|
/* Wait all existing dio workers, newcomers will block on i_rwsem */
|
2015-12-08 03:28:03 +08:00
|
|
|
inode_dio_wait(inode);
|
|
|
|
|
2022-03-09 02:50:43 +08:00
|
|
|
ret = file_modified(file);
|
|
|
|
if (ret)
|
|
|
|
goto out_mutex;
|
|
|
|
|
2015-12-08 03:28:03 +08:00
|
|
|
/*
|
|
|
|
* Prevent page faults from reinstantiating pages we have released from
|
|
|
|
* page cache.
|
|
|
|
*/
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_lock(mapping);
|
2018-07-30 05:00:22 +08:00
|
|
|
|
|
|
|
ret = ext4_break_layouts(inode);
|
|
|
|
if (ret)
|
|
|
|
goto out_dio;
|
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
first_block_offset = round_up(offset, sb->s_blocksize);
|
|
|
|
last_block_offset = round_down((offset + length), sb->s_blocksize) - 1;
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
/* Now release the pages and zero block aligned part of pages*/
|
2015-12-08 03:34:49 +08:00
|
|
|
if (last_block_offset > first_block_offset) {
|
|
|
|
ret = ext4_update_disksize_before_punch(inode, offset, length);
|
|
|
|
if (ret)
|
|
|
|
goto out_dio;
|
2013-05-28 11:32:35 +08:00
|
|
|
truncate_pagecache_range(inode, first_block_offset,
|
|
|
|
last_block_offset);
|
2015-12-08 03:34:49 +08:00
|
|
|
}
|
2013-04-04 00:45:17 +08:00
|
|
|
|
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
|
|
|
credits = ext4_writepage_trans_blocks(inode);
|
|
|
|
else
|
|
|
|
credits = ext4_blocks_for_truncate(inode);
|
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits);
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
|
|
|
ext4_std_error(sb, ret);
|
|
|
|
goto out_dio;
|
|
|
|
}
|
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
ret = ext4_zero_partial_blocks(handle, inode, offset,
|
|
|
|
length);
|
|
|
|
if (ret)
|
|
|
|
goto out_stop;
|
2013-04-04 00:45:17 +08:00
|
|
|
|
|
|
|
first_block = (offset + sb->s_blocksize - 1) >>
|
|
|
|
EXT4_BLOCK_SIZE_BITS(sb);
|
|
|
|
stop_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb);
|
|
|
|
|
2018-05-14 07:28:35 +08:00
|
|
|
/* If there are blocks to remove, do it */
|
|
|
|
if (stop_block > first_block) {
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2018-05-14 07:28:35 +08:00
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
2020-08-17 15:36:15 +08:00
|
|
|
ext4_discard_preallocations(inode, 0);
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2018-05-14 07:28:35 +08:00
|
|
|
ret = ext4_es_remove_extent(inode, first_block,
|
|
|
|
stop_block - first_block);
|
|
|
|
if (ret) {
|
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
|
|
|
goto out_stop;
|
|
|
|
}
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2018-05-14 07:28:35 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
|
|
|
ret = ext4_ext_remove_space(inode, first_block,
|
|
|
|
stop_block - 1);
|
|
|
|
else
|
|
|
|
ret = ext4_ind_remove_space(handle, inode, first_block,
|
|
|
|
stop_block);
|
2013-04-04 00:45:17 +08:00
|
|
|
|
2018-05-14 07:28:35 +08:00
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
|
|
|
}
|
2020-11-06 11:58:53 +08:00
|
|
|
ext4_fc_track_range(handle, inode, first_block, stop_block);
|
2013-04-04 00:45:17 +08:00
|
|
|
if (IS_SYNC(inode))
|
|
|
|
ext4_handle_sync(handle);
|
2014-02-21 05:58:05 +08:00
|
|
|
|
2016-11-15 10:40:10 +08:00
|
|
|
inode->i_mtime = inode->i_ctime = current_time(inode);
|
2020-04-27 09:34:37 +08:00
|
|
|
ret2 = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
if (unlikely(ret2))
|
|
|
|
ret = ret2;
|
2017-05-30 01:24:55 +08:00
|
|
|
if (ret >= 0)
|
|
|
|
ext4_update_inode_fsync_trans(handle, inode, 1);
|
2013-04-04 00:45:17 +08:00
|
|
|
out_stop:
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
out_dio:
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(mapping);
|
2013-04-04 00:45:17 +08:00
|
|
|
out_mutex:
|
2016-01-23 04:40:57 +08:00
|
|
|
inode_unlock(inode);
|
2013-04-04 00:45:17 +08:00
|
|
|
return ret;
|
2011-05-25 19:41:50 +08:00
|
|
|
}
|
|
|
|
|
2013-08-17 09:19:41 +08:00
|
|
|
int ext4_inode_attach_jinode(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
struct jbd2_inode *jinode;
|
|
|
|
|
|
|
|
if (ei->jinode || !EXT4_SB(inode->i_sb)->s_journal)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
jinode = jbd2_alloc_inode(GFP_KERNEL);
|
|
|
|
spin_lock(&inode->i_lock);
|
|
|
|
if (!ei->jinode) {
|
|
|
|
if (!jinode) {
|
|
|
|
spin_unlock(&inode->i_lock);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
ei->jinode = jinode;
|
|
|
|
jbd2_journal_init_jbd_inode(ei->jinode, inode);
|
|
|
|
jinode = NULL;
|
|
|
|
}
|
|
|
|
spin_unlock(&inode->i_lock);
|
|
|
|
if (unlikely(jinode != NULL))
|
|
|
|
jbd2_free_inode(jinode);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_truncate()
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2006-10-11 16:20:53 +08:00
|
|
|
* We block out ext4_get_block() block instantiations across the entire
|
|
|
|
* transaction, and VFS/VM ensures that ext4_truncate() cannot run
|
2006-10-11 16:20:50 +08:00
|
|
|
* simultaneously on behalf of the same inode.
|
|
|
|
*
|
2011-11-29 12:31:00 +08:00
|
|
|
* As we work through the truncate and commit bits of it to the journal there
|
2006-10-11 16:20:50 +08:00
|
|
|
* is one core, guiding principle: the file's tree must always be consistent on
|
|
|
|
* disk. We must be able to restart the truncate after a crash.
|
|
|
|
*
|
|
|
|
* The file's tree may be transiently inconsistent in memory (although it
|
|
|
|
* probably isn't), but whenever we close off and commit a journal transaction,
|
|
|
|
* the contents of (the filesystem + the journal) must be consistent and
|
|
|
|
* restartable. It's pretty simple, really: bottom up, right to left (although
|
|
|
|
* left-to-right works OK too).
|
|
|
|
*
|
|
|
|
* Note that at recovery time, journal replay occurs *before* the restart of
|
|
|
|
* truncate against the orphan inode list.
|
|
|
|
*
|
|
|
|
* The committed inode has the new, desired i_size (which is the same as
|
2006-10-11 16:20:53 +08:00
|
|
|
* i_disksize in this case). After a crash, ext4_orphan_cleanup() will see
|
2006-10-11 16:20:50 +08:00
|
|
|
* that this inode's truncate did not complete and it will again call
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_truncate() to have another go. So there will be instantiated blocks
|
|
|
|
* to the right of the truncation point in a crashed ext4 filesystem. But
|
2006-10-11 16:20:50 +08:00
|
|
|
* that's fine - as long as they are linked from the inode, the post-crash
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_truncate() run will find them and release them.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2016-11-14 11:02:26 +08:00
|
|
|
int ext4_truncate(struct inode *inode)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2013-04-04 00:47:17 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
unsigned int credits;
|
2020-04-27 09:34:37 +08:00
|
|
|
int err = 0, err2;
|
2013-04-04 00:47:17 +08:00
|
|
|
handle_t *handle;
|
|
|
|
struct address_space *mapping = inode->i_mapping;
|
|
|
|
|
2013-04-04 09:58:52 +08:00
|
|
|
/*
|
|
|
|
* There is a possibility that we're either freeing the inode
|
2014-03-25 03:15:07 +08:00
|
|
|
* or it's a completely new inode. In those cases we might not
|
2022-01-21 15:06:11 +08:00
|
|
|
* have i_rwsem locked because it's not necessary.
|
2013-04-04 09:58:52 +08:00
|
|
|
*/
|
|
|
|
if (!(inode->i_state & (I_NEW|I_FREEING)))
|
2016-01-23 04:40:57 +08:00
|
|
|
WARN_ON(!inode_is_locked(inode));
|
2011-03-22 09:38:05 +08:00
|
|
|
trace_ext4_truncate_enter(inode);
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
if (!ext4_can_truncate(inode))
|
2020-07-01 16:30:27 +08:00
|
|
|
goto out_trace;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2009-09-17 21:34:16 +08:00
|
|
|
if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC))
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE);
|
2009-02-24 21:21:14 +08:00
|
|
|
|
2012-12-11 03:06:02 +08:00
|
|
|
if (ext4_has_inline_data(inode)) {
|
|
|
|
int has_inline = 1;
|
|
|
|
|
2017-01-23 08:35:49 +08:00
|
|
|
err = ext4_inline_data_truncate(inode, &has_inline);
|
2020-07-01 16:30:27 +08:00
|
|
|
if (err || has_inline)
|
|
|
|
goto out_trace;
|
2012-12-11 03:06:02 +08:00
|
|
|
}
|
|
|
|
|
2013-08-17 09:19:41 +08:00
|
|
|
/* If we zero-out tail of the page, we have to create jinode for jbd2 */
|
|
|
|
if (inode->i_size & (inode->i_sb->s_blocksize - 1)) {
|
|
|
|
if (ext4_inode_attach_jinode(inode) < 0)
|
2020-07-01 16:30:27 +08:00
|
|
|
goto out_trace;
|
2013-08-17 09:19:41 +08:00
|
|
|
}
|
|
|
|
|
2013-04-04 00:47:17 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
|
|
|
credits = ext4_writepage_trans_blocks(inode);
|
|
|
|
else
|
|
|
|
credits = ext4_blocks_for_truncate(inode);
|
|
|
|
|
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits);
|
2020-07-01 16:30:27 +08:00
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
err = PTR_ERR(handle);
|
|
|
|
goto out_trace;
|
|
|
|
}
|
2013-04-04 00:47:17 +08:00
|
|
|
|
2013-05-28 11:32:35 +08:00
|
|
|
if (inode->i_size & (inode->i_sb->s_blocksize - 1))
|
|
|
|
ext4_block_truncate_page(handle, mapping, inode->i_size);
|
2013-04-04 00:47:17 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* We add the inode to the orphan list, so that if this
|
|
|
|
* truncate spans multiple transactions, and we crash, we will
|
|
|
|
* resume the truncate when the filesystem recovers. It also
|
|
|
|
* marks the inode dirty, to catch the new size.
|
|
|
|
*
|
|
|
|
* Implication: the file must always be in a sane, consistent
|
|
|
|
* truncatable state while each transaction commits.
|
|
|
|
*/
|
2016-11-14 11:02:26 +08:00
|
|
|
err = ext4_orphan_add(handle, inode);
|
|
|
|
if (err)
|
2013-04-04 00:47:17 +08:00
|
|
|
goto out_stop;
|
|
|
|
|
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
|
|
|
|
2020-08-17 15:36:15 +08:00
|
|
|
ext4_discard_preallocations(inode, 0);
|
2013-04-04 00:47:17 +08:00
|
|
|
|
2011-06-28 04:36:31 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
2016-11-14 11:02:28 +08:00
|
|
|
err = ext4_ext_truncate(handle, inode);
|
2011-06-28 04:36:31 +08:00
|
|
|
else
|
2013-04-04 00:47:17 +08:00
|
|
|
ext4_ind_truncate(handle, inode);
|
|
|
|
|
|
|
|
up_write(&ei->i_data_sem);
|
2016-11-14 11:02:28 +08:00
|
|
|
if (err)
|
|
|
|
goto out_stop;
|
2013-04-04 00:47:17 +08:00
|
|
|
|
|
|
|
if (IS_SYNC(inode))
|
|
|
|
ext4_handle_sync(handle);
|
|
|
|
|
|
|
|
out_stop:
|
|
|
|
/*
|
|
|
|
* If this was a simple ftruncate() and the file will remain alive,
|
|
|
|
* then we need to clear up the orphan record which we created above.
|
|
|
|
* However, if this was a real unlink then we were called by
|
2014-11-26 05:17:29 +08:00
|
|
|
* ext4_evict_inode(), and we allow that function to clean up the
|
2013-04-04 00:47:17 +08:00
|
|
|
* orphan info for us.
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink)
|
|
|
|
ext4_orphan_del(handle, inode);
|
|
|
|
|
2016-11-15 10:40:10 +08:00
|
|
|
inode->i_mtime = inode->i_ctime = current_time(inode);
|
2020-04-27 09:34:37 +08:00
|
|
|
err2 = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
if (unlikely(err2 && !err))
|
|
|
|
err = err2;
|
2013-04-04 00:47:17 +08:00
|
|
|
ext4_journal_stop(handle);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-07-01 16:30:27 +08:00
|
|
|
out_trace:
|
2011-03-22 09:38:05 +08:00
|
|
|
trace_ext4_truncate_exit(inode);
|
2016-11-14 11:02:26 +08:00
|
|
|
return err;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2021-09-01 10:09:54 +08:00
|
|
|
static inline u64 ext4_inode_peek_iversion(const struct inode *inode)
|
|
|
|
{
|
|
|
|
if (unlikely(EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL))
|
|
|
|
return inode_peek_iversion_raw(inode);
|
|
|
|
else
|
|
|
|
return inode_peek_iversion(inode);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_inode_blocks_set(struct ext4_inode *raw_inode,
|
|
|
|
struct ext4_inode_info *ei)
|
|
|
|
{
|
|
|
|
struct inode *inode = &(ei->vfs_inode);
|
|
|
|
u64 i_blocks = READ_ONCE(inode->i_blocks);
|
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
|
|
|
|
if (i_blocks <= ~0U) {
|
|
|
|
/*
|
|
|
|
* i_blocks can be represented in a 32 bit variable
|
|
|
|
* as multiple of 512 bytes
|
|
|
|
*/
|
|
|
|
raw_inode->i_blocks_lo = cpu_to_le32(i_blocks);
|
|
|
|
raw_inode->i_blocks_high = 0;
|
|
|
|
ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This should never happen since sb->s_maxbytes should not have
|
|
|
|
* allowed this, sb->s_maxbytes was set according to the huge_file
|
|
|
|
* feature in ext4_fill_super().
|
|
|
|
*/
|
|
|
|
if (!ext4_has_feature_huge_file(sb))
|
|
|
|
return -EFSCORRUPTED;
|
|
|
|
|
|
|
|
if (i_blocks <= 0xffffffffffffULL) {
|
|
|
|
/*
|
|
|
|
* i_blocks can be represented in a 48 bit variable
|
|
|
|
* as multiple of 512 bytes
|
|
|
|
*/
|
|
|
|
raw_inode->i_blocks_lo = cpu_to_le32(i_blocks);
|
|
|
|
raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32);
|
|
|
|
ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE);
|
|
|
|
} else {
|
|
|
|
ext4_set_inode_flag(inode, EXT4_INODE_HUGE_FILE);
|
|
|
|
/* i_block is stored in file system block size */
|
|
|
|
i_blocks = i_blocks >> (inode->i_blkbits - 9);
|
|
|
|
raw_inode->i_blocks_lo = cpu_to_le32(i_blocks);
|
|
|
|
raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_fill_raw_inode(struct inode *inode, struct ext4_inode *raw_inode)
|
|
|
|
{
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
uid_t i_uid;
|
|
|
|
gid_t i_gid;
|
|
|
|
projid_t i_projid;
|
|
|
|
int block;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
err = ext4_inode_blocks_set(raw_inode, ei);
|
|
|
|
|
|
|
|
raw_inode->i_mode = cpu_to_le16(inode->i_mode);
|
|
|
|
i_uid = i_uid_read(inode);
|
|
|
|
i_gid = i_gid_read(inode);
|
|
|
|
i_projid = from_kprojid(&init_user_ns, ei->i_projid);
|
|
|
|
if (!(test_opt(inode->i_sb, NO_UID32))) {
|
|
|
|
raw_inode->i_uid_low = cpu_to_le16(low_16_bits(i_uid));
|
|
|
|
raw_inode->i_gid_low = cpu_to_le16(low_16_bits(i_gid));
|
|
|
|
/*
|
|
|
|
* Fix up interoperability with old kernels. Otherwise,
|
|
|
|
* old inodes get re-used with the upper 16 bits of the
|
|
|
|
* uid/gid intact.
|
|
|
|
*/
|
|
|
|
if (ei->i_dtime && list_empty(&ei->i_orphan)) {
|
|
|
|
raw_inode->i_uid_high = 0;
|
|
|
|
raw_inode->i_gid_high = 0;
|
|
|
|
} else {
|
|
|
|
raw_inode->i_uid_high =
|
|
|
|
cpu_to_le16(high_16_bits(i_uid));
|
|
|
|
raw_inode->i_gid_high =
|
|
|
|
cpu_to_le16(high_16_bits(i_gid));
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
raw_inode->i_uid_low = cpu_to_le16(fs_high2lowuid(i_uid));
|
|
|
|
raw_inode->i_gid_low = cpu_to_le16(fs_high2lowgid(i_gid));
|
|
|
|
raw_inode->i_uid_high = 0;
|
|
|
|
raw_inode->i_gid_high = 0;
|
|
|
|
}
|
|
|
|
raw_inode->i_links_count = cpu_to_le16(inode->i_nlink);
|
|
|
|
|
|
|
|
EXT4_INODE_SET_XTIME(i_ctime, inode, raw_inode);
|
|
|
|
EXT4_INODE_SET_XTIME(i_mtime, inode, raw_inode);
|
|
|
|
EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
|
|
|
|
EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
|
|
|
|
|
|
|
|
raw_inode->i_dtime = cpu_to_le32(ei->i_dtime);
|
|
|
|
raw_inode->i_flags = cpu_to_le32(ei->i_flags & 0xFFFFFFFF);
|
|
|
|
if (likely(!test_opt2(inode->i_sb, HURD_COMPAT)))
|
|
|
|
raw_inode->i_file_acl_high =
|
|
|
|
cpu_to_le16(ei->i_file_acl >> 32);
|
|
|
|
raw_inode->i_file_acl_lo = cpu_to_le32(ei->i_file_acl);
|
|
|
|
ext4_isize_set(raw_inode, ei->i_disksize);
|
|
|
|
|
|
|
|
raw_inode->i_generation = cpu_to_le32(inode->i_generation);
|
|
|
|
if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) {
|
|
|
|
if (old_valid_dev(inode->i_rdev)) {
|
|
|
|
raw_inode->i_block[0] =
|
|
|
|
cpu_to_le32(old_encode_dev(inode->i_rdev));
|
|
|
|
raw_inode->i_block[1] = 0;
|
|
|
|
} else {
|
|
|
|
raw_inode->i_block[0] = 0;
|
|
|
|
raw_inode->i_block[1] =
|
|
|
|
cpu_to_le32(new_encode_dev(inode->i_rdev));
|
|
|
|
raw_inode->i_block[2] = 0;
|
|
|
|
}
|
|
|
|
} else if (!ext4_has_inline_data(inode)) {
|
|
|
|
for (block = 0; block < EXT4_N_BLOCKS; block++)
|
|
|
|
raw_inode->i_block[block] = ei->i_data[block];
|
|
|
|
}
|
|
|
|
|
|
|
|
if (likely(!test_opt2(inode->i_sb, HURD_COMPAT))) {
|
|
|
|
u64 ivers = ext4_inode_peek_iversion(inode);
|
|
|
|
|
|
|
|
raw_inode->i_disk_version = cpu_to_le32(ivers);
|
|
|
|
if (ei->i_extra_isize) {
|
|
|
|
if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi))
|
|
|
|
raw_inode->i_version_hi =
|
|
|
|
cpu_to_le32(ivers >> 32);
|
|
|
|
raw_inode->i_extra_isize =
|
|
|
|
cpu_to_le16(ei->i_extra_isize);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (i_projid != EXT4_DEF_PROJID &&
|
|
|
|
!ext4_has_feature_project(inode->i_sb))
|
|
|
|
err = err ?: -EFSCORRUPTED;
|
|
|
|
|
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE &&
|
|
|
|
EXT4_FITS_IN_INODE(raw_inode, ei, i_projid))
|
|
|
|
raw_inode->i_projid = cpu_to_le32(i_projid);
|
|
|
|
|
|
|
|
ext4_inode_csum_set(inode, raw_inode, ei);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_get_inode_loc returns with an extra refcount against the inode's
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
* underlying buffer_head on success. If we pass 'inode' and it does not
|
|
|
|
* have in-inode xattr, we have all inode data in memory that is needed
|
|
|
|
* to recreate the on-disk version of this inode.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2020-10-16 04:37:59 +08:00
|
|
|
static int __ext4_get_inode_loc(struct super_block *sb, unsigned long ino,
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
struct inode *inode, struct ext4_iloc *iloc,
|
2020-10-16 04:37:59 +08:00
|
|
|
ext4_fsblk_t *ret_block)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2008-10-10 11:53:47 +08:00
|
|
|
struct ext4_group_desc *gdp;
|
|
|
|
struct buffer_head *bh;
|
|
|
|
ext4_fsblk_t block;
|
2019-09-30 08:59:23 +08:00
|
|
|
struct blk_plug plug;
|
2008-10-10 11:53:47 +08:00
|
|
|
int inodes_per_block, inode_offset;
|
|
|
|
|
2008-11-23 04:04:59 +08:00
|
|
|
iloc->bh = NULL;
|
2020-10-16 04:37:59 +08:00
|
|
|
if (ino < EXT4_ROOT_INO ||
|
|
|
|
ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
|
2015-10-18 04:16:04 +08:00
|
|
|
return -EFSCORRUPTED;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
iloc->block_group = (ino - 1) / EXT4_INODES_PER_GROUP(sb);
|
2008-10-10 11:53:47 +08:00
|
|
|
gdp = ext4_get_group_desc(sb, iloc->block_group, NULL);
|
|
|
|
if (!gdp)
|
2006-10-11 16:20:50 +08:00
|
|
|
return -EIO;
|
|
|
|
|
2008-10-10 11:53:47 +08:00
|
|
|
/*
|
|
|
|
* Figure out the offset within the block group inode table
|
|
|
|
*/
|
2011-05-09 22:26:41 +08:00
|
|
|
inodes_per_block = EXT4_SB(sb)->s_inodes_per_block;
|
2020-10-16 04:37:59 +08:00
|
|
|
inode_offset = ((ino - 1) %
|
2008-10-10 11:53:47 +08:00
|
|
|
EXT4_INODES_PER_GROUP(sb));
|
|
|
|
block = ext4_inode_table(sb, gdp) + (inode_offset / inodes_per_block);
|
|
|
|
iloc->offset = (inode_offset % inodes_per_block) * EXT4_INODE_SIZE(sb);
|
|
|
|
|
|
|
|
bh = sb_getblk(sb, block);
|
2013-01-13 05:28:47 +08:00
|
|
|
if (unlikely(!bh))
|
2013-01-13 05:19:36 +08:00
|
|
|
return -ENOMEM;
|
2021-08-26 21:04:08 +08:00
|
|
|
if (ext4_buffer_uptodate(bh))
|
|
|
|
goto has_buffer;
|
2008-07-27 04:39:26 +08:00
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
lock_buffer(bh);
|
2021-09-10 16:03:16 +08:00
|
|
|
if (ext4_buffer_uptodate(bh)) {
|
|
|
|
/* Someone brought it uptodate while we waited */
|
|
|
|
unlock_buffer(bh);
|
|
|
|
goto has_buffer;
|
|
|
|
}
|
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
/*
|
|
|
|
* If we have all information of the inode in memory and this
|
|
|
|
* is the only valid inode in the block, we need not read the
|
|
|
|
* block.
|
|
|
|
*/
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
if (inode && !ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
|
2021-08-26 21:04:08 +08:00
|
|
|
struct buffer_head *bitmap_bh;
|
|
|
|
int i, start;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
start = inode_offset & ~(inodes_per_block - 1);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
/* Is the inode bitmap in cache? */
|
|
|
|
bitmap_bh = sb_getblk(sb, ext4_inode_bitmap(sb, gdp));
|
|
|
|
if (unlikely(!bitmap_bh))
|
|
|
|
goto make_io;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
/*
|
|
|
|
* If the inode bitmap isn't in cache then the
|
|
|
|
* optimisation may end up performing two reads instead
|
|
|
|
* of one, so skip it.
|
|
|
|
*/
|
|
|
|
if (!buffer_uptodate(bitmap_bh)) {
|
2006-10-11 16:20:50 +08:00
|
|
|
brelse(bitmap_bh);
|
2021-08-26 21:04:08 +08:00
|
|
|
goto make_io;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2021-08-26 21:04:08 +08:00
|
|
|
for (i = start; i < start + inodes_per_block; i++) {
|
|
|
|
if (i == inode_offset)
|
|
|
|
continue;
|
|
|
|
if (ext4_test_bit(i, bitmap_bh->b_data))
|
|
|
|
break;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2021-08-26 21:04:08 +08:00
|
|
|
brelse(bitmap_bh);
|
|
|
|
if (i == start + inodes_per_block) {
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
struct ext4_inode *raw_inode =
|
|
|
|
(struct ext4_inode *) (bh->b_data + iloc->offset);
|
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
/* all other inodes are free, so skip I/O */
|
|
|
|
memset(bh->b_data, 0, bh->b_size);
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
if (!ext4_test_inode_state(inode, EXT4_STATE_NEW))
|
|
|
|
ext4_fill_raw_inode(inode, raw_inode);
|
2021-08-26 21:04:08 +08:00
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
unlock_buffer(bh);
|
|
|
|
goto has_buffer;
|
|
|
|
}
|
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
make_io:
|
2021-08-26 21:04:08 +08:00
|
|
|
/*
|
|
|
|
* If we need to do any I/O, try to pre-readahead extra
|
|
|
|
* blocks from the inode table.
|
|
|
|
*/
|
|
|
|
blk_start_plug(&plug);
|
|
|
|
if (EXT4_SB(sb)->s_inode_readahead_blks) {
|
|
|
|
ext4_fsblk_t b, end, table;
|
|
|
|
unsigned num;
|
|
|
|
__u32 ra_blks = EXT4_SB(sb)->s_inode_readahead_blks;
|
|
|
|
|
|
|
|
table = ext4_inode_table(sb, gdp);
|
|
|
|
/* s_inode_readahead_blks is always a power of 2 */
|
|
|
|
b = block & ~((ext4_fsblk_t) ra_blks - 1);
|
|
|
|
if (table > b)
|
|
|
|
b = table;
|
|
|
|
end = b + ra_blks;
|
|
|
|
num = EXT4_INODES_PER_GROUP(sb);
|
|
|
|
if (ext4_has_group_desc_csum(sb))
|
|
|
|
num -= ext4_itable_unused_count(sb, gdp);
|
|
|
|
table += num / inodes_per_block;
|
|
|
|
if (end > table)
|
|
|
|
end = table;
|
|
|
|
while (b <= end)
|
|
|
|
ext4_sb_breadahead_unmovable(sb, b++);
|
|
|
|
}
|
2008-10-10 11:53:47 +08:00
|
|
|
|
2021-08-26 21:04:08 +08:00
|
|
|
/*
|
|
|
|
* There are other valid inodes in the buffer, this inode
|
|
|
|
* has in-inode xattrs, or we don't have this inode in memory.
|
|
|
|
* Read the block from disk.
|
|
|
|
*/
|
|
|
|
trace_ext4_load_inode(sb, ino);
|
|
|
|
ext4_read_bh_nowait(bh, REQ_META | REQ_PRIO, NULL);
|
|
|
|
blk_finish_plug(&plug);
|
|
|
|
wait_on_buffer(bh);
|
|
|
|
ext4_simulate_fail_bh(sb, bh, EXT4_SIM_INODE_EIO);
|
|
|
|
if (!buffer_uptodate(bh)) {
|
|
|
|
if (ret_block)
|
|
|
|
*ret_block = block;
|
|
|
|
brelse(bh);
|
|
|
|
return -EIO;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
has_buffer:
|
|
|
|
iloc->bh = bh;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
static int __ext4_get_inode_loc_noinmem(struct inode *inode,
|
|
|
|
struct ext4_iloc *iloc)
|
|
|
|
{
|
2021-12-02 00:34:21 +08:00
|
|
|
ext4_fsblk_t err_blk = 0;
|
2020-10-16 04:37:59 +08:00
|
|
|
int ret;
|
|
|
|
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
ret = __ext4_get_inode_loc(inode->i_sb, inode->i_ino, NULL, iloc,
|
2020-10-16 04:37:59 +08:00
|
|
|
&err_blk);
|
|
|
|
|
|
|
|
if (ret == -EIO)
|
|
|
|
ext4_error_inode_block(inode, err_blk, EIO,
|
|
|
|
"unable to read itable block");
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
int ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2021-12-02 00:34:21 +08:00
|
|
|
ext4_fsblk_t err_blk = 0;
|
2020-10-16 04:37:59 +08:00
|
|
|
int ret;
|
|
|
|
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
ret = __ext4_get_inode_loc(inode->i_sb, inode->i_ino, inode, iloc,
|
|
|
|
&err_blk);
|
2020-10-16 04:37:59 +08:00
|
|
|
|
|
|
|
if (ret == -EIO)
|
|
|
|
ext4_error_inode_block(inode, err_blk, EIO,
|
|
|
|
"unable to read itable block");
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
int ext4_get_fc_inode_loc(struct super_block *sb, unsigned long ino,
|
|
|
|
struct ext4_iloc *iloc)
|
|
|
|
{
|
ext4: prevent getting empty inode buffer
In ext4_get_inode_loc(), we may skip IO and get an zero && uptodate
inode buffer when the inode monopolize an inode block for performance
reason. For most cases, ext4_mark_iloc_dirty() will fill the inode
buffer to make it fine, but we could miss this call if something bad
happened. Finally, __ext4_get_inode_loc_noinmem() may probably get an
empty inode buffer and trigger ext4 error.
For example, if we remove a nonexistent xattr on inode A,
ext4_xattr_set_handle() will return ENODATA before invoking
ext4_mark_iloc_dirty(), it will left an uptodate but zero buffer. We
will get checksum error message in ext4_iget() when getting inode again.
EXT4-fs error (device sda): ext4_lookup:1784: inode #131074: comm cat: iget: checksum invalid
Even worse, if we allocate another inode B at the same inode block, it
will corrupt the inode A on disk when write back inode B.
So this patch initialize the inode buffer by filling the in-mem inode
contents if we skip read I/O, ensure that the buffer is really uptodate.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20210901020955.1657340-4-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2021-09-01 10:09:55 +08:00
|
|
|
return __ext4_get_inode_loc(sb, ino, NULL, iloc, NULL);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2020-05-28 22:59:58 +08:00
|
|
|
static bool ext4_should_enable_dax(struct inode *inode)
|
2017-10-13 00:00:59 +08:00
|
|
|
{
|
2020-05-28 22:59:58 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
|
2020-05-28 23:00:00 +08:00
|
|
|
if (test_opt2(inode->i_sb, DAX_NEVER))
|
2017-10-13 00:00:59 +08:00
|
|
|
return false;
|
|
|
|
if (!S_ISREG(inode->i_mode))
|
|
|
|
return false;
|
|
|
|
if (ext4_should_journal_data(inode))
|
|
|
|
return false;
|
|
|
|
if (ext4_has_inline_data(inode))
|
|
|
|
return false;
|
2018-12-12 17:50:10 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_ENCRYPT))
|
2017-10-13 00:00:59 +08:00
|
|
|
return false;
|
2019-07-23 00:26:24 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_VERITY))
|
|
|
|
return false;
|
2020-05-28 22:59:58 +08:00
|
|
|
if (!test_bit(EXT4_FLAGS_BDEV_IS_DAX, &sbi->s_ext4_flags))
|
|
|
|
return false;
|
|
|
|
if (test_opt(inode->i_sb, DAX_ALWAYS))
|
|
|
|
return true;
|
|
|
|
|
2020-05-28 23:00:02 +08:00
|
|
|
return ext4_test_inode_flag(inode, EXT4_INODE_DAX);
|
2017-10-13 00:00:59 +08:00
|
|
|
}
|
|
|
|
|
2020-05-28 22:59:59 +08:00
|
|
|
void ext4_set_inode_flags(struct inode *inode, bool init)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2006-10-11 16:20:53 +08:00
|
|
|
unsigned int flags = EXT4_I(inode)->i_flags;
|
2014-03-30 22:20:01 +08:00
|
|
|
unsigned int new_fl = 0;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-05-28 22:59:59 +08:00
|
|
|
WARN_ON_ONCE(IS_DAX(inode) && init);
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
if (flags & EXT4_SYNC_FL)
|
2014-03-30 22:20:01 +08:00
|
|
|
new_fl |= S_SYNC;
|
2006-10-11 16:20:53 +08:00
|
|
|
if (flags & EXT4_APPEND_FL)
|
2014-03-30 22:20:01 +08:00
|
|
|
new_fl |= S_APPEND;
|
2006-10-11 16:20:53 +08:00
|
|
|
if (flags & EXT4_IMMUTABLE_FL)
|
2014-03-30 22:20:01 +08:00
|
|
|
new_fl |= S_IMMUTABLE;
|
2006-10-11 16:20:53 +08:00
|
|
|
if (flags & EXT4_NOATIME_FL)
|
2014-03-30 22:20:01 +08:00
|
|
|
new_fl |= S_NOATIME;
|
2006-10-11 16:20:53 +08:00
|
|
|
if (flags & EXT4_DIRSYNC_FL)
|
2014-03-30 22:20:01 +08:00
|
|
|
new_fl |= S_DIRSYNC;
|
2020-05-28 22:59:59 +08:00
|
|
|
|
|
|
|
/* Because of the way inode_set_flags() works we must preserve S_DAX
|
|
|
|
* here if already set. */
|
|
|
|
new_fl |= (inode->i_flags & S_DAX);
|
|
|
|
if (init && ext4_should_enable_dax(inode))
|
2015-02-17 07:59:38 +08:00
|
|
|
new_fl |= S_DAX;
|
2020-05-28 22:59:59 +08:00
|
|
|
|
2017-10-10 03:15:35 +08:00
|
|
|
if (flags & EXT4_ENCRYPT_FL)
|
|
|
|
new_fl |= S_ENCRYPTED;
|
ext4: Support case-insensitive file name lookups
This patch implements the actual support for case-insensitive file name
lookups in ext4, based on the feature bit and the encoding stored in the
superblock.
A filesystem that has the casefold feature set is able to configure
directories with the +F (EXT4_CASEFOLD_FL) attribute, enabling lookups
to succeed in that directory in a case-insensitive fashion, i.e: match
a directory entry even if the name used by userspace is not a byte per
byte match with the disk name, but is an equivalent case-insensitive
version of the Unicode string. This operation is called a
case-insensitive file name lookup.
The feature is configured as an inode attribute applied to directories
and inherited by its children. This attribute can only be enabled on
empty directories for filesystems that support the encoding feature,
thus preventing collision of file names that only differ by case.
* dcache handling:
For a +F directory, Ext4 only stores the first equivalent name dentry
used in the dcache. This is done to prevent unintentional duplication of
dentries in the dcache, while also allowing the VFS code to quickly find
the right entry in the cache despite which equivalent string was used in
a previous lookup, without having to resort to ->lookup().
d_hash() of casefolded directories is implemented as the hash of the
casefolded string, such that we always have a well-known bucket for all
the equivalencies of the same string. d_compare() uses the
utf8_strncasecmp() infrastructure, which handles the comparison of
equivalent, same case, names as well.
For now, negative lookups are not inserted in the dcache, since they
would need to be invalidated anyway, because we can't trust missing file
dentries. This is bad for performance but requires some leveraging of
the vfs layer to fix. We can live without that for now, and so does
everyone else.
* on-disk data:
Despite using a specific version of the name as the internal
representation within the dcache, the name stored and fetched from the
disk is a byte-per-byte match with what the user requested, making this
implementation 'name-preserving'. i.e. no actual information is lost
when writing to storage.
DX is supported by modifying the hashes used in +F directories to make
them case/encoding-aware. The new disk hashes are calculated as the
hash of the full casefolded string, instead of the string directly.
This allows us to efficiently search for file names in the htree without
requiring the user to provide an exact name.
* Dealing with invalid sequences:
By default, when a invalid UTF-8 sequence is identified, ext4 will treat
it as an opaque byte sequence, ignoring the encoding and reverting to
the old behavior for that unique file. This means that case-insensitive
file name lookup will not work only for that file. An optional bit can
be set in the superblock telling the filesystem code and userspace tools
to enforce the encoding. When that optional bit is set, any attempt to
create a file name using an invalid UTF-8 sequence will fail and return
an error to userspace.
* Normalization algorithm:
The UTF-8 algorithms used to compare strings in ext4 is implemented
lives in fs/unicode, and is based on a previous version developed by
SGI. It implements the Canonical decomposition (NFD) algorithm
described by the Unicode specification 12.1, or higher, combined with
the elimination of ignorable code points (NFDi) and full
case-folding (CF) as documented in fs/unicode/utf8_norm.c.
NFD seems to be the best normalization method for EXT4 because:
- It has a lower cost than NFC/NFKC (which requires
decomposing to NFD as an intermediary step)
- It doesn't eliminate important semantic meaning like
compatibility decompositions.
Although:
- This implementation is not completely linguistic accurate, because
different languages have conflicting rules, which would require the
specialization of the filesystem to a given locale, which brings all
sorts of problems for removable media and for users who use more than
one language.
Signed-off-by: Gabriel Krisman Bertazi <krisman@collabora.co.uk>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-04-26 02:12:08 +08:00
|
|
|
if (flags & EXT4_CASEFOLD_FL)
|
|
|
|
new_fl |= S_CASEFOLD;
|
2019-07-23 00:26:24 +08:00
|
|
|
if (flags & EXT4_VERITY_FL)
|
|
|
|
new_fl |= S_VERITY;
|
2014-03-25 02:43:12 +08:00
|
|
|
inode_set_flags(inode, new_fl,
|
2017-10-10 03:15:35 +08:00
|
|
|
S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_DAX|
|
2019-07-23 00:26:24 +08:00
|
|
|
S_ENCRYPTED|S_CASEFOLD|S_VERITY);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2008-01-29 12:58:26 +08:00
|
|
|
static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct ext4_inode_info *ei)
|
2008-01-29 12:58:26 +08:00
|
|
|
{
|
|
|
|
blkcnt_t i_blocks ;
|
2008-01-29 12:58:27 +08:00
|
|
|
struct inode *inode = &(ei->vfs_inode);
|
|
|
|
struct super_block *sb = inode->i_sb;
|
2008-01-29 12:58:26 +08:00
|
|
|
|
2015-10-18 04:18:43 +08:00
|
|
|
if (ext4_has_feature_huge_file(sb)) {
|
2008-01-29 12:58:26 +08:00
|
|
|
/* we are using combined 48 bit field */
|
|
|
|
i_blocks = ((u64)le16_to_cpu(raw_inode->i_blocks_high)) << 32 |
|
|
|
|
le32_to_cpu(raw_inode->i_blocks_lo);
|
2010-06-14 21:54:48 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_HUGE_FILE)) {
|
2008-01-29 12:58:27 +08:00
|
|
|
/* i_blocks represent file system block size */
|
|
|
|
return i_blocks << (inode->i_blkbits - 9);
|
|
|
|
} else {
|
|
|
|
return i_blocks;
|
|
|
|
}
|
2008-01-29 12:58:26 +08:00
|
|
|
} else {
|
|
|
|
return le32_to_cpu(raw_inode->i_blocks_lo);
|
|
|
|
}
|
|
|
|
}
|
2007-07-18 21:24:20 +08:00
|
|
|
|
2018-05-23 05:14:07 +08:00
|
|
|
static inline int ext4_iget_extra_inode(struct inode *inode,
|
2012-12-03 00:13:24 +08:00
|
|
|
struct ext4_inode *raw_inode,
|
|
|
|
struct ext4_inode_info *ei)
|
|
|
|
{
|
|
|
|
__le32 *magic = (void *)raw_inode +
|
|
|
|
EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize;
|
2018-05-23 05:14:07 +08:00
|
|
|
|
2022-06-16 10:13:58 +08:00
|
|
|
if (EXT4_INODE_HAS_XATTR_SPACE(inode) &&
|
2016-12-02 03:51:58 +08:00
|
|
|
*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) {
|
2012-12-03 00:13:24 +08:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_XATTR);
|
2018-05-23 05:14:07 +08:00
|
|
|
return ext4_find_inline_data_nolock(inode);
|
2012-12-11 03:05:51 +08:00
|
|
|
} else
|
|
|
|
EXT4_I(inode)->i_inline_off = 0;
|
2018-05-23 05:14:07 +08:00
|
|
|
return 0;
|
2012-12-03 00:13:24 +08:00
|
|
|
}
|
|
|
|
|
2016-01-09 05:01:21 +08:00
|
|
|
int ext4_get_projid(struct inode *inode, kprojid_t *projid)
|
|
|
|
{
|
2016-09-06 11:11:58 +08:00
|
|
|
if (!ext4_has_feature_project(inode->i_sb))
|
2016-01-09 05:01:21 +08:00
|
|
|
return -EOPNOTSUPP;
|
|
|
|
*projid = EXT4_I(inode)->i_projid;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-05-10 23:55:31 +08:00
|
|
|
/*
|
|
|
|
* ext4 has self-managed i_version for ea inodes, it stores the lower 32bit of
|
|
|
|
* refcount in i_version, so use raw values if inode has EXT4_EA_INODE_FL flag
|
|
|
|
* set.
|
|
|
|
*/
|
|
|
|
static inline void ext4_inode_set_iversion_queried(struct inode *inode, u64 val)
|
|
|
|
{
|
|
|
|
if (unlikely(EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL))
|
|
|
|
inode_set_iversion_raw(inode, val);
|
|
|
|
else
|
|
|
|
inode_set_iversion_queried(inode, val);
|
|
|
|
}
|
|
|
|
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
|
|
|
|
ext4_iget_flags flags, const char *function,
|
|
|
|
unsigned int line)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2006-10-11 16:20:53 +08:00
|
|
|
struct ext4_iloc iloc;
|
|
|
|
struct ext4_inode *raw_inode;
|
2008-02-07 16:15:37 +08:00
|
|
|
struct ext4_inode_info *ei;
|
2021-08-12 21:31:22 +08:00
|
|
|
struct ext4_super_block *es = EXT4_SB(sb)->s_es;
|
2008-02-07 16:15:37 +08:00
|
|
|
struct inode *inode;
|
2009-12-09 12:51:10 +08:00
|
|
|
journal_t *journal = EXT4_SB(sb)->s_journal;
|
2008-02-07 16:15:37 +08:00
|
|
|
long ret;
|
2016-12-10 22:55:01 +08:00
|
|
|
loff_t size;
|
2006-10-11 16:20:50 +08:00
|
|
|
int block;
|
2012-02-08 07:41:49 +08:00
|
|
|
uid_t i_uid;
|
|
|
|
gid_t i_gid;
|
2016-01-09 05:01:21 +08:00
|
|
|
projid_t i_projid;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2019-01-01 11:34:31 +08:00
|
|
|
if ((!(flags & EXT4_IGET_SPECIAL) &&
|
2021-08-12 21:31:22 +08:00
|
|
|
((ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO) ||
|
|
|
|
ino == le32_to_cpu(es->s_usr_quota_inum) ||
|
|
|
|
ino == le32_to_cpu(es->s_grp_quota_inum) ||
|
2021-08-16 17:57:06 +08:00
|
|
|
ino == le32_to_cpu(es->s_prj_quota_inum) ||
|
|
|
|
ino == le32_to_cpu(es->s_orphan_file_inum))) ||
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
(ino < EXT4_ROOT_INO) ||
|
2021-08-12 21:31:22 +08:00
|
|
|
(ino > le32_to_cpu(es->s_inodes_count))) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
if (flags & EXT4_IGET_HANDLE)
|
|
|
|
return ERR_PTR(-ESTALE);
|
2020-11-27 19:33:57 +08:00
|
|
|
__ext4_error(sb, function, line, false, EFSCORRUPTED, 0,
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
"inode #%lu: comm %s: iget: illegal inode #",
|
|
|
|
ino, current->comm);
|
|
|
|
return ERR_PTR(-EFSCORRUPTED);
|
|
|
|
}
|
|
|
|
|
2008-02-07 16:15:37 +08:00
|
|
|
inode = iget_locked(sb, ino);
|
|
|
|
if (!inode)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
if (!(inode->i_state & I_NEW))
|
|
|
|
return inode;
|
|
|
|
|
|
|
|
ei = EXT4_I(inode);
|
2011-02-22 10:01:42 +08:00
|
|
|
iloc.bh = NULL;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
ret = __ext4_get_inode_loc_noinmem(inode, &iloc);
|
2008-02-07 16:15:37 +08:00
|
|
|
if (ret < 0)
|
2006-10-11 16:20:50 +08:00
|
|
|
goto bad_inode;
|
2006-10-11 16:20:53 +08:00
|
|
|
raw_inode = ext4_raw_inode(&iloc);
|
2012-04-30 06:31:10 +08:00
|
|
|
|
2018-03-30 09:56:09 +08:00
|
|
|
if ((ino == EXT4_ROOT_INO) && (raw_inode->i_links_count == 0)) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: root inode unallocated");
|
2018-03-30 09:56:09 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
|
|
|
goto bad_inode;
|
|
|
|
}
|
|
|
|
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
if ((flags & EXT4_IGET_HANDLE) &&
|
|
|
|
(raw_inode->i_links_count == 0) && (raw_inode->i_mode == 0)) {
|
|
|
|
ret = -ESTALE;
|
|
|
|
goto bad_inode;
|
|
|
|
}
|
|
|
|
|
2012-04-30 06:31:10 +08:00
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
|
|
|
|
ei->i_extra_isize = le16_to_cpu(raw_inode->i_extra_isize);
|
|
|
|
if (EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize >
|
2016-12-02 03:43:33 +08:00
|
|
|
EXT4_INODE_SIZE(inode->i_sb) ||
|
|
|
|
(ei->i_extra_isize & 3)) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: bad extra_isize %u "
|
|
|
|
"(inode size %u)",
|
2016-12-02 03:43:33 +08:00
|
|
|
ei->i_extra_isize,
|
|
|
|
EXT4_INODE_SIZE(inode->i_sb));
|
2015-10-18 04:16:04 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
2012-04-30 06:31:10 +08:00
|
|
|
goto bad_inode;
|
|
|
|
}
|
|
|
|
} else
|
|
|
|
ei->i_extra_isize = 0;
|
|
|
|
|
|
|
|
/* Precompute checksum seed for inode metadata */
|
2014-10-13 15:36:16 +08:00
|
|
|
if (ext4_has_metadata_csum(sb)) {
|
2012-04-30 06:31:10 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
__u32 csum;
|
|
|
|
__le32 inum = cpu_to_le32(inode->i_ino);
|
|
|
|
__le32 gen = raw_inode->i_generation;
|
|
|
|
csum = ext4_chksum(sbi, sbi->s_csum_seed, (__u8 *)&inum,
|
|
|
|
sizeof(inum));
|
|
|
|
ei->i_csum_seed = ext4_chksum(sbi, csum, (__u8 *)&gen,
|
|
|
|
sizeof(gen));
|
|
|
|
}
|
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
if ((!ext4_inode_csum_verify(inode, raw_inode, ei) ||
|
|
|
|
ext4_simulate_fail(sb, EXT4_SIM_INODE_CRC)) &&
|
|
|
|
(!(EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY))) {
|
|
|
|
ext4_error_inode_err(inode, function, line, 0,
|
|
|
|
EFSBADCRC, "iget: checksum invalid");
|
2015-10-18 04:16:04 +08:00
|
|
|
ret = -EFSBADCRC;
|
2012-04-30 06:31:10 +08:00
|
|
|
goto bad_inode;
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
inode->i_mode = le16_to_cpu(raw_inode->i_mode);
|
2012-02-08 07:41:49 +08:00
|
|
|
i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
|
|
|
|
i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
|
2016-09-06 11:11:58 +08:00
|
|
|
if (ext4_has_feature_project(sb) &&
|
2016-01-09 05:01:21 +08:00
|
|
|
EXT4_INODE_SIZE(sb) > EXT4_GOOD_OLD_INODE_SIZE &&
|
|
|
|
EXT4_FITS_IN_INODE(raw_inode, ei, i_projid))
|
|
|
|
i_projid = (projid_t)le32_to_cpu(raw_inode->i_projid);
|
|
|
|
else
|
|
|
|
i_projid = EXT4_DEF_PROJID;
|
|
|
|
|
2008-09-09 10:25:24 +08:00
|
|
|
if (!(test_opt(inode->i_sb, NO_UID32))) {
|
2012-02-08 07:41:49 +08:00
|
|
|
i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
|
|
|
|
i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2012-02-08 07:41:49 +08:00
|
|
|
i_uid_write(inode, i_uid);
|
|
|
|
i_gid_write(inode, i_gid);
|
2016-01-09 05:01:21 +08:00
|
|
|
ei->i_projid = make_kprojid(&init_user_ns, i_projid);
|
2011-10-28 20:13:29 +08:00
|
|
|
set_nlink(inode, le16_to_cpu(raw_inode->i_links_count));
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2011-01-11 01:18:25 +08:00
|
|
|
ext4_clear_state_flags(ei); /* Only relevant on 32-bit archs */
|
2012-12-11 03:04:46 +08:00
|
|
|
ei->i_inline_off = 0;
|
2006-10-11 16:20:50 +08:00
|
|
|
ei->i_dir_start_lookup = 0;
|
|
|
|
ei->i_dtime = le32_to_cpu(raw_inode->i_dtime);
|
|
|
|
/* We now have enough fields to check if the inode was active or not.
|
|
|
|
* This is needed because nfsd might try to access dead inodes
|
|
|
|
* the test is that same one that e2fsck uses
|
|
|
|
* NeilBrown 1999oct15
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink == 0) {
|
2013-04-09 00:54:05 +08:00
|
|
|
if ((inode->i_mode == 0 ||
|
|
|
|
!(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ORPHAN_FS)) &&
|
|
|
|
ino != EXT4_BOOT_LOADER_INO) {
|
2006-10-11 16:20:50 +08:00
|
|
|
/* this inode is deleted */
|
2008-02-07 16:15:37 +08:00
|
|
|
ret = -ESTALE;
|
2006-10-11 16:20:50 +08:00
|
|
|
goto bad_inode;
|
|
|
|
}
|
|
|
|
/* The only unlinked inodes we let through here have
|
|
|
|
* valid i_mode and are being read by the orphan
|
|
|
|
* recovery code: that's fine, we're about to complete
|
2013-04-09 00:54:05 +08:00
|
|
|
* the process of deleting those.
|
|
|
|
* OR it is the EXT4_BOOT_LOADER_INO which is
|
|
|
|
* not initialized on a new filesystem. */
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
ei->i_flags = le32_to_cpu(raw_inode->i_flags);
|
2020-05-28 22:59:59 +08:00
|
|
|
ext4_set_inode_flags(inode, true);
|
2008-01-29 12:58:26 +08:00
|
|
|
inode->i_blocks = ext4_inode_blocks(raw_inode, ei);
|
2008-01-29 12:58:27 +08:00
|
|
|
ei->i_file_acl = le32_to_cpu(raw_inode->i_file_acl_lo);
|
2015-10-18 04:18:43 +08:00
|
|
|
if (ext4_has_feature_64bit(sb))
|
2006-10-11 16:21:09 +08:00
|
|
|
ei->i_file_acl |=
|
|
|
|
((__u64)le16_to_cpu(raw_inode->i_file_acl_high)) << 32;
|
2017-06-22 09:09:57 +08:00
|
|
|
inode->i_size = ext4_isize(sb, raw_inode);
|
2016-12-10 22:55:01 +08:00
|
|
|
if ((size = i_size_read(inode)) < 0) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: bad i_size value: %lld", size);
|
2016-12-10 22:55:01 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
|
|
|
goto bad_inode;
|
|
|
|
}
|
2020-02-10 22:43:16 +08:00
|
|
|
/*
|
|
|
|
* If dir_index is not enabled but there's dir with INDEX flag set,
|
|
|
|
* we'd normally treat htree data as empty space. But with metadata
|
|
|
|
* checksumming that corrupts checksums so forbid that.
|
|
|
|
*/
|
|
|
|
if (!ext4_has_feature_dir_index(sb) && ext4_has_metadata_csum(sb) &&
|
|
|
|
ext4_test_inode_flag(inode, EXT4_INODE_INDEX)) {
|
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: Dir with htree data on filesystem without dir_index feature.");
|
|
|
|
ret = -EFSCORRUPTED;
|
|
|
|
goto bad_inode;
|
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
ei->i_disksize = inode->i_size;
|
2009-12-14 20:21:14 +08:00
|
|
|
#ifdef CONFIG_QUOTA
|
|
|
|
ei->i_reserved_quota = 0;
|
|
|
|
#endif
|
2006-10-11 16:20:50 +08:00
|
|
|
inode->i_generation = le32_to_cpu(raw_inode->i_generation);
|
|
|
|
ei->i_block_group = iloc.block_group;
|
2009-03-13 00:18:34 +08:00
|
|
|
ei->i_last_alloc_group = ~0;
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* NOTE! The in-memory inode i_data array is in little-endian order
|
|
|
|
* even on big-endian machines: we do NOT byteswap the block numbers!
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
for (block = 0; block < EXT4_N_BLOCKS; block++)
|
2006-10-11 16:20:50 +08:00
|
|
|
ei->i_data[block] = raw_inode->i_block[block];
|
|
|
|
INIT_LIST_HEAD(&ei->i_orphan);
|
2020-10-16 04:37:57 +08:00
|
|
|
ext4_fc_init_inode(&ei->vfs_inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2009-12-09 12:51:10 +08:00
|
|
|
/*
|
|
|
|
* Set transaction id's of transactions that have to be committed
|
|
|
|
* to finish f[data]sync. We set them to currently running transaction
|
|
|
|
* as we cannot be sure that the inode or some of its metadata isn't
|
|
|
|
* part of the transaction - the inode could have been reclaimed and
|
|
|
|
* now it is reread from disk.
|
|
|
|
*/
|
|
|
|
if (journal) {
|
|
|
|
transaction_t *transaction;
|
|
|
|
tid_t tid;
|
|
|
|
|
2010-08-04 09:35:12 +08:00
|
|
|
read_lock(&journal->j_state_lock);
|
2009-12-09 12:51:10 +08:00
|
|
|
if (journal->j_running_transaction)
|
|
|
|
transaction = journal->j_running_transaction;
|
|
|
|
else
|
|
|
|
transaction = journal->j_committing_transaction;
|
|
|
|
if (transaction)
|
|
|
|
tid = transaction->t_tid;
|
|
|
|
else
|
|
|
|
tid = journal->j_commit_sequence;
|
2010-08-04 09:35:12 +08:00
|
|
|
read_unlock(&journal->j_state_lock);
|
2009-12-09 12:51:10 +08:00
|
|
|
ei->i_sync_tid = tid;
|
|
|
|
ei->i_datasync_tid = tid;
|
|
|
|
}
|
|
|
|
|
2008-02-06 11:36:43 +08:00
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
|
2006-10-11 16:20:50 +08:00
|
|
|
if (ei->i_extra_isize == 0) {
|
|
|
|
/* The extra space is currently unused. Use it. */
|
2016-12-02 03:43:33 +08:00
|
|
|
BUILD_BUG_ON(sizeof(struct ext4_inode) & 3);
|
2006-10-11 16:20:53 +08:00
|
|
|
ei->i_extra_isize = sizeof(struct ext4_inode) -
|
|
|
|
EXT4_GOOD_OLD_INODE_SIZE;
|
2006-10-11 16:20:50 +08:00
|
|
|
} else {
|
2018-05-23 05:14:07 +08:00
|
|
|
ret = ext4_iget_extra_inode(inode, raw_inode, ei);
|
|
|
|
if (ret)
|
|
|
|
goto bad_inode;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2012-04-30 06:31:10 +08:00
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2007-07-18 21:15:20 +08:00
|
|
|
EXT4_INODE_GET_XTIME(i_ctime, inode, raw_inode);
|
|
|
|
EXT4_INODE_GET_XTIME(i_mtime, inode, raw_inode);
|
|
|
|
EXT4_INODE_GET_XTIME(i_atime, inode, raw_inode);
|
|
|
|
EXT4_EINODE_GET_XTIME(i_crtime, ei, raw_inode);
|
|
|
|
|
2014-03-25 02:09:06 +08:00
|
|
|
if (likely(!test_opt2(inode->i_sb, HURD_COMPAT))) {
|
2018-01-09 21:21:39 +08:00
|
|
|
u64 ivers = le32_to_cpu(raw_inode->i_disk_version);
|
|
|
|
|
2014-03-20 12:32:57 +08:00
|
|
|
if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
|
|
|
|
if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi))
|
2018-01-09 21:21:39 +08:00
|
|
|
ivers |=
|
2014-03-20 12:32:57 +08:00
|
|
|
(__u64)(le32_to_cpu(raw_inode->i_version_hi)) << 32;
|
|
|
|
}
|
2018-05-10 23:55:31 +08:00
|
|
|
ext4_inode_set_iversion_queried(inode, ivers);
|
2008-01-29 12:58:27 +08:00
|
|
|
}
|
|
|
|
|
ext4: Do not try to validate extents on special files
The EXTENTS_FL flag should never be set on special files, but if it
is, don't bother trying to validate that the extents tree is valid,
since only files, directories, and non-fast symlinks will ever have an
extent data structure. We perhaps should flag the filesystem as being
corrupted if we see a special file (named pipes, device nodes, Unix
domain sockets, etc.) with the EXTENTS_FL flag, but e2fsck doesn't
currently check this case, so we'll just ignore this for now, since
it's harmless.
Without this fix, a special device with the extents flag is flagged as
an error by the kernel, so it is impossible to access or delete the
inode, but e2fsck doesn't see it as a problem, leading to
confused/frustrated users.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2009-04-25 06:45:35 +08:00
|
|
|
ret = 0;
|
2009-04-25 01:43:20 +08:00
|
|
|
if (ei->i_file_acl &&
|
2020-07-28 21:04:34 +08:00
|
|
|
!ext4_inode_block_valid(inode, ei->i_file_acl, 1)) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: bad extended attribute block %llu",
|
2010-05-17 09:00:00 +08:00
|
|
|
ei->i_file_acl);
|
2015-10-18 04:16:04 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
2009-04-25 01:43:20 +08:00
|
|
|
goto bad_inode;
|
2012-12-11 03:05:51 +08:00
|
|
|
} else if (!ext4_has_inline_data(inode)) {
|
2018-08-02 12:11:16 +08:00
|
|
|
/* validate the block references in the inode */
|
2020-10-16 04:37:59 +08:00
|
|
|
if (!(EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY) &&
|
|
|
|
(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) ||
|
|
|
|
(S_ISLNK(inode->i_mode) &&
|
|
|
|
!ext4_inode_is_fast_symlink(inode)))) {
|
2018-08-02 12:11:16 +08:00
|
|
|
if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
|
2012-12-11 03:05:51 +08:00
|
|
|
ret = ext4_ext_check_inode(inode);
|
2018-08-02 12:11:16 +08:00
|
|
|
else
|
|
|
|
ret = ext4_ind_check_inode(inode);
|
2012-12-11 03:05:51 +08:00
|
|
|
}
|
2009-03-31 20:36:10 +08:00
|
|
|
}
|
2009-11-14 21:19:05 +08:00
|
|
|
if (ret)
|
2009-06-15 05:45:34 +08:00
|
|
|
goto bad_inode;
|
2009-03-28 04:39:58 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
if (S_ISREG(inode->i_mode)) {
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_op = &ext4_file_inode_operations;
|
2015-04-16 07:15:17 +08:00
|
|
|
inode->i_fop = &ext4_file_operations;
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_set_aops(inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
} else if (S_ISDIR(inode->i_mode)) {
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_op = &ext4_dir_inode_operations;
|
|
|
|
inode->i_fop = &ext4_dir_operations;
|
2006-10-11 16:20:50 +08:00
|
|
|
} else if (S_ISLNK(inode->i_mode)) {
|
2018-05-14 04:45:56 +08:00
|
|
|
/* VFS does not allow setting these so must be corruption */
|
|
|
|
if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) {
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: immutable or append flags "
|
|
|
|
"not allowed on symlinks");
|
2018-05-14 04:45:56 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
|
|
|
goto bad_inode;
|
|
|
|
}
|
2018-12-12 17:50:10 +08:00
|
|
|
if (IS_ENCRYPTED(inode)) {
|
2015-04-28 05:51:30 +08:00
|
|
|
inode->i_op = &ext4_encrypted_symlink_inode_operations;
|
|
|
|
} else if (ext4_inode_is_fast_symlink(inode)) {
|
2015-05-02 22:13:58 +08:00
|
|
|
inode->i_link = (char *)ei->i_data;
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_op = &ext4_fast_symlink_inode_operations;
|
2008-12-20 04:47:15 +08:00
|
|
|
nd_terminate_link(ei->i_data, inode->i_size,
|
|
|
|
sizeof(ei->i_data) - 1);
|
|
|
|
} else {
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_op = &ext4_symlink_inode_operations;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2009-03-26 12:06:19 +08:00
|
|
|
} else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) ||
|
|
|
|
S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) {
|
2006-10-11 16:20:53 +08:00
|
|
|
inode->i_op = &ext4_special_inode_operations;
|
2006-10-11 16:20:50 +08:00
|
|
|
if (raw_inode->i_block[0])
|
|
|
|
init_special_inode(inode, inode->i_mode,
|
|
|
|
old_decode_dev(le32_to_cpu(raw_inode->i_block[0])));
|
|
|
|
else
|
|
|
|
init_special_inode(inode, inode->i_mode,
|
|
|
|
new_decode_dev(le32_to_cpu(raw_inode->i_block[1])));
|
2013-04-09 00:54:05 +08:00
|
|
|
} else if (ino == EXT4_BOOT_LOADER_INO) {
|
|
|
|
make_bad_inode(inode);
|
2009-03-26 12:06:19 +08:00
|
|
|
} else {
|
2015-10-18 04:16:04 +08:00
|
|
|
ret = -EFSCORRUPTED;
|
ext4: avoid declaring fs inconsistent due to invalid file handles
If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.
This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.
#define _GNU_SOURCE
#include <fcntl.h>
struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};
int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};
open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}
Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
2018-12-20 01:29:13 +08:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"iget: bogus i_mode (%o)", inode->i_mode);
|
2009-03-26 12:06:19 +08:00
|
|
|
goto bad_inode;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
2019-09-03 13:43:17 +08:00
|
|
|
if (IS_CASEFOLDED(inode) && !ext4_has_feature_casefold(inode->i_sb))
|
|
|
|
ext4_error_inode(inode, function, line, 0,
|
|
|
|
"casefold flag without casefold feature");
|
2008-09-09 10:25:24 +08:00
|
|
|
brelse(iloc.bh);
|
2017-06-22 23:44:55 +08:00
|
|
|
|
2008-02-07 16:15:37 +08:00
|
|
|
unlock_new_inode(inode);
|
|
|
|
return inode;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
bad_inode:
|
2009-11-14 21:19:05 +08:00
|
|
|
brelse(iloc.bh);
|
2008-02-07 16:15:37 +08:00
|
|
|
iget_failed(inode);
|
|
|
|
return ERR_PTR(ret);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2017-12-01 19:40:16 +08:00
|
|
|
static void __ext4_update_other_inode_time(struct super_block *sb,
|
|
|
|
unsigned long orig_ino,
|
|
|
|
unsigned long ino,
|
|
|
|
struct ext4_inode *raw_inode)
|
2015-02-02 13:37:02 +08:00
|
|
|
{
|
2017-12-01 19:40:16 +08:00
|
|
|
struct inode *inode;
|
|
|
|
|
|
|
|
inode = find_inode_by_ino_rcu(sb, ino);
|
|
|
|
if (!inode)
|
|
|
|
return;
|
2015-02-02 13:37:02 +08:00
|
|
|
|
2021-01-13 03:02:53 +08:00
|
|
|
if (!inode_is_dirtytime_only(inode))
|
2017-12-01 19:40:16 +08:00
|
|
|
return;
|
|
|
|
|
2015-02-02 13:37:02 +08:00
|
|
|
spin_lock(&inode->i_lock);
|
2021-01-13 03:02:53 +08:00
|
|
|
if (inode_is_dirtytime_only(inode)) {
|
2015-02-02 13:37:02 +08:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
|
2020-05-29 22:24:43 +08:00
|
|
|
inode->i_state &= ~I_DIRTY_TIME;
|
2015-02-02 13:37:02 +08:00
|
|
|
spin_unlock(&inode->i_lock);
|
|
|
|
|
|
|
|
spin_lock(&ei->i_raw_lock);
|
2017-12-01 19:40:16 +08:00
|
|
|
EXT4_INODE_SET_XTIME(i_ctime, inode, raw_inode);
|
|
|
|
EXT4_INODE_SET_XTIME(i_mtime, inode, raw_inode);
|
|
|
|
EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
|
|
|
|
ext4_inode_csum_set(inode, raw_inode, ei);
|
2015-02-02 13:37:02 +08:00
|
|
|
spin_unlock(&ei->i_raw_lock);
|
2017-12-01 19:40:16 +08:00
|
|
|
trace_ext4_other_inode_update_time(inode, orig_ino);
|
|
|
|
return;
|
2015-02-02 13:37:02 +08:00
|
|
|
}
|
|
|
|
spin_unlock(&inode->i_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Opportunistically update the other time fields for other inodes in
|
|
|
|
* the same inode table block.
|
|
|
|
*/
|
|
|
|
static void ext4_update_other_inodes_time(struct super_block *sb,
|
|
|
|
unsigned long orig_ino, char *buf)
|
|
|
|
{
|
|
|
|
unsigned long ino;
|
|
|
|
int i, inodes_per_block = EXT4_SB(sb)->s_inodes_per_block;
|
|
|
|
int inode_size = EXT4_INODE_SIZE(sb);
|
|
|
|
|
2015-07-02 11:37:46 +08:00
|
|
|
/*
|
|
|
|
* Calculate the first inode in the inode table block. Inode
|
|
|
|
* numbers are one-based. That is, the first inode in a block
|
|
|
|
* (assuming 4k blocks and 256 byte inodes) is (n*16 + 1).
|
|
|
|
*/
|
|
|
|
ino = ((orig_ino - 1) & ~(inodes_per_block - 1)) + 1;
|
2017-12-01 19:40:16 +08:00
|
|
|
rcu_read_lock();
|
2015-02-02 13:37:02 +08:00
|
|
|
for (i = 0; i < inodes_per_block; i++, ino++, buf += inode_size) {
|
|
|
|
if (ino == orig_ino)
|
|
|
|
continue;
|
2017-12-01 19:40:16 +08:00
|
|
|
__ext4_update_other_inode_time(sb, orig_ino, ino,
|
|
|
|
(struct ext4_inode *)buf);
|
2015-02-02 13:37:02 +08:00
|
|
|
}
|
2017-12-01 19:40:16 +08:00
|
|
|
rcu_read_unlock();
|
2015-02-02 13:37:02 +08:00
|
|
|
}
|
|
|
|
|
2021-09-01 10:09:53 +08:00
|
|
|
/*
|
|
|
|
* Post the struct inode info into an on-disk inode location in the
|
|
|
|
* buffer-cache. This gobbles the caller's reference to the
|
|
|
|
* buffer_head in the inode location struct.
|
|
|
|
*
|
|
|
|
* The caller must have write access to iloc->bh.
|
|
|
|
*/
|
|
|
|
static int ext4_do_update_inode(handle_t *handle,
|
|
|
|
struct inode *inode,
|
|
|
|
struct ext4_iloc *iloc)
|
|
|
|
{
|
|
|
|
struct ext4_inode *raw_inode = ext4_raw_inode(iloc);
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
struct buffer_head *bh = iloc->bh;
|
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
int err;
|
|
|
|
int need_datasync = 0, set_large_file = 0;
|
|
|
|
|
|
|
|
spin_lock(&ei->i_raw_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* For fields not tracked in the in-memory inode, initialise them
|
|
|
|
* to zero for new inodes.
|
|
|
|
*/
|
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_NEW))
|
|
|
|
memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
|
|
|
|
|
|
|
|
if (READ_ONCE(ei->i_disksize) != ext4_isize(inode->i_sb, raw_inode))
|
|
|
|
need_datasync = 1;
|
|
|
|
if (ei->i_disksize > 0x7fffffffULL) {
|
|
|
|
if (!ext4_has_feature_large_file(sb) ||
|
|
|
|
EXT4_SB(sb)->s_es->s_rev_level == cpu_to_le32(EXT4_GOOD_OLD_REV))
|
|
|
|
set_large_file = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = ext4_fill_raw_inode(inode, raw_inode);
|
2014-04-22 02:37:55 +08:00
|
|
|
spin_unlock(&ei->i_raw_lock);
|
2021-08-26 21:04:09 +08:00
|
|
|
if (err) {
|
|
|
|
EXT4_ERROR_INODE(inode, "corrupted inode contents");
|
|
|
|
goto out_brelse;
|
|
|
|
}
|
|
|
|
|
2017-11-28 05:05:09 +08:00
|
|
|
if (inode->i_sb->s_flags & SB_LAZYTIME)
|
2015-02-02 13:37:02 +08:00
|
|
|
ext4_update_other_inodes_time(inode->i_sb, inode->i_ino,
|
|
|
|
bh->b_data);
|
2014-04-22 02:37:55 +08:00
|
|
|
|
2009-09-29 22:07:47 +08:00
|
|
|
BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
|
2021-03-12 14:50:51 +08:00
|
|
|
err = ext4_handle_dirty_metadata(handle, NULL, bh);
|
|
|
|
if (err)
|
2021-08-26 21:04:09 +08:00
|
|
|
goto out_error;
|
2010-01-25 03:34:07 +08:00
|
|
|
ext4_clear_inode_state(inode, EXT4_STATE_NEW);
|
2014-04-22 02:37:55 +08:00
|
|
|
if (set_large_file) {
|
2014-05-13 10:06:43 +08:00
|
|
|
BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access");
|
2021-08-16 17:57:04 +08:00
|
|
|
err = ext4_journal_get_write_access(handle, sb,
|
|
|
|
EXT4_SB(sb)->s_sbh,
|
|
|
|
EXT4_JTR_NONE);
|
2014-04-22 02:37:55 +08:00
|
|
|
if (err)
|
2021-08-26 21:04:09 +08:00
|
|
|
goto out_error;
|
2020-12-16 18:18:39 +08:00
|
|
|
lock_buffer(EXT4_SB(sb)->s_sbh);
|
2015-10-18 04:18:43 +08:00
|
|
|
ext4_set_feature_large_file(sb);
|
2020-12-16 18:18:39 +08:00
|
|
|
ext4_superblock_csum_set(sb);
|
|
|
|
unlock_buffer(EXT4_SB(sb)->s_sbh);
|
2014-04-22 02:37:55 +08:00
|
|
|
ext4_handle_sync(handle);
|
2020-12-16 18:18:44 +08:00
|
|
|
err = ext4_handle_dirty_metadata(handle, NULL,
|
|
|
|
EXT4_SB(sb)->s_sbh);
|
2014-04-22 02:37:55 +08:00
|
|
|
}
|
2012-09-27 09:52:20 +08:00
|
|
|
ext4_update_inode_fsync_trans(handle, inode, need_datasync);
|
2021-08-26 21:04:09 +08:00
|
|
|
out_error:
|
|
|
|
ext4_std_error(inode->i_sb, err);
|
2006-10-11 16:20:50 +08:00
|
|
|
out_brelse:
|
2008-09-09 10:25:24 +08:00
|
|
|
brelse(bh);
|
2006-10-11 16:20:50 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_write_inode()
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* We are called from a few places:
|
|
|
|
*
|
2014-04-08 23:38:28 +08:00
|
|
|
* - Within generic_file_aio_write() -> generic_write_sync() for O_SYNC files.
|
2006-10-11 16:20:50 +08:00
|
|
|
* Here, there will be no transaction running. We wait for any running
|
2012-09-02 01:31:09 +08:00
|
|
|
* transaction to commit.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2014-04-08 23:38:28 +08:00
|
|
|
* - Within flush work (sys_sync(), kupdate and such).
|
|
|
|
* We wait on commit, if told to.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2014-04-08 23:38:28 +08:00
|
|
|
* - Within iput_final() -> write_inode_now()
|
|
|
|
* We wait on commit, if told to.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* In all cases it is actually safe for us to return without doing anything,
|
|
|
|
* because the inode has been copied into a raw inode buffer in
|
2014-04-08 23:38:28 +08:00
|
|
|
* ext4_mark_inode_dirty(). This is a correctness thing for WB_SYNC_ALL
|
|
|
|
* writeback.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* Note that we are absolutely dependent upon all inode dirtiers doing the
|
|
|
|
* right thing: they *must* call mark_inode_dirty() after dirtying info in
|
|
|
|
* which we are interested.
|
|
|
|
*
|
|
|
|
* It would be a bug for them to not do this. The code:
|
|
|
|
*
|
|
|
|
* mark_inode_dirty(inode)
|
|
|
|
* stuff();
|
|
|
|
* inode->i_size = expr;
|
|
|
|
*
|
2014-04-08 23:38:28 +08:00
|
|
|
* is in error because write_inode() could occur while `stuff()' is running,
|
|
|
|
* and the new i_size will be lost. Plus the inode will no longer be on the
|
|
|
|
* superblock's dirty inode list.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2010-03-05 16:21:37 +08:00
|
|
|
int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2009-09-10 10:33:47 +08:00
|
|
|
int err;
|
|
|
|
|
2018-12-20 03:36:58 +08:00
|
|
|
if (WARN_ON_ONCE(current->flags & PF_MEMALLOC) ||
|
|
|
|
sb_rdonly(inode->i_sb))
|
2006-10-11 16:20:50 +08:00
|
|
|
return 0;
|
|
|
|
|
2018-12-20 03:36:58 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2009-09-10 10:33:47 +08:00
|
|
|
if (EXT4_SB(inode->i_sb)->s_journal) {
|
|
|
|
if (ext4_journal_current_handle()) {
|
2022-06-08 19:23:47 +08:00
|
|
|
ext4_debug("called recursively, non-PF_MEMALLOC!\n");
|
2009-09-10 10:33:47 +08:00
|
|
|
dump_stack();
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2014-03-04 23:50:50 +08:00
|
|
|
/*
|
|
|
|
* No need to force transaction in WB_SYNC_NONE mode. Also
|
|
|
|
* ext4_sync_fs() will force the commit after everything is
|
|
|
|
* written.
|
|
|
|
*/
|
|
|
|
if (wbc->sync_mode != WB_SYNC_ALL || wbc->for_sync)
|
2009-09-10 10:33:47 +08:00
|
|
|
return 0;
|
|
|
|
|
2020-10-16 04:37:57 +08:00
|
|
|
err = ext4_fc_commit(EXT4_SB(inode->i_sb)->s_journal,
|
2018-12-20 03:36:58 +08:00
|
|
|
EXT4_I(inode)->i_sync_tid);
|
2009-09-10 10:33:47 +08:00
|
|
|
} else {
|
|
|
|
struct ext4_iloc iloc;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2020-10-16 04:37:59 +08:00
|
|
|
err = __ext4_get_inode_loc_noinmem(inode, &iloc);
|
2009-09-10 10:33:47 +08:00
|
|
|
if (err)
|
|
|
|
return err;
|
2014-03-04 23:50:50 +08:00
|
|
|
/*
|
|
|
|
* sync(2) will flush the whole buffer cache. No need to do
|
|
|
|
* it here separately for each inode.
|
|
|
|
*/
|
|
|
|
if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync)
|
2009-09-29 22:07:47 +08:00
|
|
|
sync_dirty_buffer(iloc.bh);
|
|
|
|
if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) {
|
2020-03-29 07:33:43 +08:00
|
|
|
ext4_error_inode_block(inode, iloc.bh->b_blocknr, EIO,
|
|
|
|
"IO error syncing inode");
|
2009-09-29 22:07:47 +08:00
|
|
|
err = -EIO;
|
|
|
|
}
|
2010-04-04 05:44:16 +08:00
|
|
|
brelse(iloc.bh);
|
2009-09-10 10:33:47 +08:00
|
|
|
}
|
|
|
|
return err;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2012-12-26 02:29:52 +08:00
|
|
|
/*
|
2022-02-10 04:21:43 +08:00
|
|
|
* In data=journal mode ext4_journalled_invalidate_folio() may fail to invalidate
|
|
|
|
* buffers that are attached to a folio straddling i_size and are undergoing
|
2012-12-26 02:29:52 +08:00
|
|
|
* commit. In that case we have to wait for commit to finish and try again.
|
|
|
|
*/
|
|
|
|
static void ext4_wait_for_tail_page_commit(struct inode *inode)
|
|
|
|
{
|
|
|
|
unsigned offset;
|
|
|
|
journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
|
|
|
|
tid_t commit_tid = 0;
|
|
|
|
int ret;
|
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
offset = inode->i_size & (PAGE_SIZE - 1);
|
2012-12-26 02:29:52 +08:00
|
|
|
/*
|
2022-02-10 04:21:43 +08:00
|
|
|
* If the folio is fully truncated, we don't need to wait for any commit
|
|
|
|
* (and we even should not as __ext4_journalled_invalidate_folio() may
|
|
|
|
* strip all buffers from the folio but keep the folio dirty which can then
|
|
|
|
* confuse e.g. concurrent ext4_writepage() seeing dirty folio without
|
ext4: fix a bug in ext4_wait_for_tail_page_commit
No need to wait for any commit once the page is fully truncated.
Besides, it may confuse e.g. concurrent ext4_writepage() with the page
still be dirty (will be cleared by truncate_pagecache() in
ext4_setattr()) but buffers has been freed; and then trigger a bug
show as below:
[ 26.057508] ------------[ cut here ]------------
[ 26.058531] kernel BUG at fs/ext4/inode.c:2134!
...
[ 26.088130] Call trace:
[ 26.088695] ext4_writepage+0x914/0xb28
[ 26.089541] writeout.isra.4+0x1b4/0x2b8
[ 26.090409] move_to_new_page+0x3b0/0x568
[ 26.091338] __unmap_and_move+0x648/0x988
[ 26.092241] unmap_and_move+0x48c/0xbb8
[ 26.093096] migrate_pages+0x220/0xb28
[ 26.093945] kernel_mbind+0x828/0xa18
[ 26.094791] __arm64_sys_mbind+0xc8/0x138
[ 26.095716] el0_svc_common+0x190/0x490
[ 26.096571] el0_svc_handler+0x60/0xd0
[ 26.097423] el0_svc+0x8/0xc
Run the procedure (generate by syzkaller) parallel with ext3.
void main()
{
int fd, fd1, ret;
void *addr;
size_t length = 4096;
int flags;
off_t offset = 0;
char *str = "12345";
fd = open("a", O_RDWR | O_CREAT);
assert(fd >= 0);
/* Truncate to 4k */
ret = ftruncate(fd, length);
assert(ret == 0);
/* Journal data mode */
flags = 0xc00f;
ret = ioctl(fd, _IOW('f', 2, long), &flags);
assert(ret == 0);
/* Truncate to 0 */
fd1 = open("a", O_TRUNC | O_NOATIME);
assert(fd1 >= 0);
addr = mmap(NULL, length, PROT_WRITE | PROT_READ,
MAP_SHARED, fd, offset);
assert(addr != (void *)-1);
memcpy(addr, str, 5);
mbind(addr, length, 0, 0, 0, MPOL_MF_MOVE);
}
And the bug will be triggered once we seen the below order.
reproduce1 reproduce2
... | ...
truncate to 4k |
change to journal data mode |
| memcpy(set page dirty)
truncate to 0: |
ext4_setattr: |
... |
ext4_wait_for_tail_page_commit |
| mbind(trigger bug)
truncate_pagecache(clean dirty)| ...
... |
mbind will call ext4_writepage() since the page still be dirty, and then
report the bug since the buffers has been free. Fix it by return
directly once offset equals to 0 which means the page has been fully
truncated.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: yangerkun <yangerkun@huawei.com>
Link: https://lore.kernel.org/r/20190919063508.1045-1-yangerkun@huawei.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-09-19 14:35:08 +08:00
|
|
|
* buffers). Also we don't need to wait for any commit if all buffers in
|
2022-02-10 04:21:43 +08:00
|
|
|
* the folio remain valid. This is most beneficial for the common case of
|
ext4: fix a bug in ext4_wait_for_tail_page_commit
No need to wait for any commit once the page is fully truncated.
Besides, it may confuse e.g. concurrent ext4_writepage() with the page
still be dirty (will be cleared by truncate_pagecache() in
ext4_setattr()) but buffers has been freed; and then trigger a bug
show as below:
[ 26.057508] ------------[ cut here ]------------
[ 26.058531] kernel BUG at fs/ext4/inode.c:2134!
...
[ 26.088130] Call trace:
[ 26.088695] ext4_writepage+0x914/0xb28
[ 26.089541] writeout.isra.4+0x1b4/0x2b8
[ 26.090409] move_to_new_page+0x3b0/0x568
[ 26.091338] __unmap_and_move+0x648/0x988
[ 26.092241] unmap_and_move+0x48c/0xbb8
[ 26.093096] migrate_pages+0x220/0xb28
[ 26.093945] kernel_mbind+0x828/0xa18
[ 26.094791] __arm64_sys_mbind+0xc8/0x138
[ 26.095716] el0_svc_common+0x190/0x490
[ 26.096571] el0_svc_handler+0x60/0xd0
[ 26.097423] el0_svc+0x8/0xc
Run the procedure (generate by syzkaller) parallel with ext3.
void main()
{
int fd, fd1, ret;
void *addr;
size_t length = 4096;
int flags;
off_t offset = 0;
char *str = "12345";
fd = open("a", O_RDWR | O_CREAT);
assert(fd >= 0);
/* Truncate to 4k */
ret = ftruncate(fd, length);
assert(ret == 0);
/* Journal data mode */
flags = 0xc00f;
ret = ioctl(fd, _IOW('f', 2, long), &flags);
assert(ret == 0);
/* Truncate to 0 */
fd1 = open("a", O_TRUNC | O_NOATIME);
assert(fd1 >= 0);
addr = mmap(NULL, length, PROT_WRITE | PROT_READ,
MAP_SHARED, fd, offset);
assert(addr != (void *)-1);
memcpy(addr, str, 5);
mbind(addr, length, 0, 0, 0, MPOL_MF_MOVE);
}
And the bug will be triggered once we seen the below order.
reproduce1 reproduce2
... | ...
truncate to 4k |
change to journal data mode |
| memcpy(set page dirty)
truncate to 0: |
ext4_setattr: |
... |
ext4_wait_for_tail_page_commit |
| mbind(trigger bug)
truncate_pagecache(clean dirty)| ...
... |
mbind will call ext4_writepage() since the page still be dirty, and then
report the bug since the buffers has been free. Fix it by return
directly once offset equals to 0 which means the page has been fully
truncated.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: yangerkun <yangerkun@huawei.com>
Link: https://lore.kernel.org/r/20190919063508.1045-1-yangerkun@huawei.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-09-19 14:35:08 +08:00
|
|
|
* blocksize == PAGESIZE.
|
2012-12-26 02:29:52 +08:00
|
|
|
*/
|
ext4: fix a bug in ext4_wait_for_tail_page_commit
No need to wait for any commit once the page is fully truncated.
Besides, it may confuse e.g. concurrent ext4_writepage() with the page
still be dirty (will be cleared by truncate_pagecache() in
ext4_setattr()) but buffers has been freed; and then trigger a bug
show as below:
[ 26.057508] ------------[ cut here ]------------
[ 26.058531] kernel BUG at fs/ext4/inode.c:2134!
...
[ 26.088130] Call trace:
[ 26.088695] ext4_writepage+0x914/0xb28
[ 26.089541] writeout.isra.4+0x1b4/0x2b8
[ 26.090409] move_to_new_page+0x3b0/0x568
[ 26.091338] __unmap_and_move+0x648/0x988
[ 26.092241] unmap_and_move+0x48c/0xbb8
[ 26.093096] migrate_pages+0x220/0xb28
[ 26.093945] kernel_mbind+0x828/0xa18
[ 26.094791] __arm64_sys_mbind+0xc8/0x138
[ 26.095716] el0_svc_common+0x190/0x490
[ 26.096571] el0_svc_handler+0x60/0xd0
[ 26.097423] el0_svc+0x8/0xc
Run the procedure (generate by syzkaller) parallel with ext3.
void main()
{
int fd, fd1, ret;
void *addr;
size_t length = 4096;
int flags;
off_t offset = 0;
char *str = "12345";
fd = open("a", O_RDWR | O_CREAT);
assert(fd >= 0);
/* Truncate to 4k */
ret = ftruncate(fd, length);
assert(ret == 0);
/* Journal data mode */
flags = 0xc00f;
ret = ioctl(fd, _IOW('f', 2, long), &flags);
assert(ret == 0);
/* Truncate to 0 */
fd1 = open("a", O_TRUNC | O_NOATIME);
assert(fd1 >= 0);
addr = mmap(NULL, length, PROT_WRITE | PROT_READ,
MAP_SHARED, fd, offset);
assert(addr != (void *)-1);
memcpy(addr, str, 5);
mbind(addr, length, 0, 0, 0, MPOL_MF_MOVE);
}
And the bug will be triggered once we seen the below order.
reproduce1 reproduce2
... | ...
truncate to 4k |
change to journal data mode |
| memcpy(set page dirty)
truncate to 0: |
ext4_setattr: |
... |
ext4_wait_for_tail_page_commit |
| mbind(trigger bug)
truncate_pagecache(clean dirty)| ...
... |
mbind will call ext4_writepage() since the page still be dirty, and then
report the bug since the buffers has been free. Fix it by return
directly once offset equals to 0 which means the page has been fully
truncated.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: yangerkun <yangerkun@huawei.com>
Link: https://lore.kernel.org/r/20190919063508.1045-1-yangerkun@huawei.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2019-09-19 14:35:08 +08:00
|
|
|
if (!offset || offset > (PAGE_SIZE - i_blocksize(inode)))
|
2012-12-26 02:29:52 +08:00
|
|
|
return;
|
|
|
|
while (1) {
|
2022-02-10 04:21:43 +08:00
|
|
|
struct folio *folio = filemap_lock_folio(inode->i_mapping,
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
inode->i_size >> PAGE_SHIFT);
|
2022-02-10 04:21:43 +08:00
|
|
|
if (!folio)
|
2012-12-26 02:29:52 +08:00
|
|
|
return;
|
2022-02-10 04:21:43 +08:00
|
|
|
ret = __ext4_journalled_invalidate_folio(folio, offset,
|
|
|
|
folio_size(folio) - offset);
|
|
|
|
folio_unlock(folio);
|
|
|
|
folio_put(folio);
|
2012-12-26 02:29:52 +08:00
|
|
|
if (ret != -EBUSY)
|
|
|
|
return;
|
|
|
|
commit_tid = 0;
|
|
|
|
read_lock(&journal->j_state_lock);
|
|
|
|
if (journal->j_committing_transaction)
|
|
|
|
commit_tid = journal->j_committing_transaction->t_tid;
|
|
|
|
read_unlock(&journal->j_state_lock);
|
|
|
|
if (commit_tid)
|
|
|
|
jbd2_log_wait_commit(journal, commit_tid);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_setattr()
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* Called from notify_change.
|
|
|
|
*
|
|
|
|
* We want to trap VFS attempts to truncate the file as soon as
|
|
|
|
* possible. In particular, we want to make sure that when the VFS
|
|
|
|
* shrinks i_size, we put the inode on the orphan list and modify
|
|
|
|
* i_disksize immediately, so that during the subsequent flushing of
|
|
|
|
* dirty pages and freeing of disk blocks, we can guarantee that any
|
|
|
|
* commit will leave the blocks being flushed in an unused state on
|
|
|
|
* disk. (On recovery, the inode will get truncated and the blocks will
|
|
|
|
* be freed, so we have a strong guarantee that no future commit will
|
|
|
|
* leave these blocks visible to the user.)
|
|
|
|
*
|
2008-07-12 07:27:31 +08:00
|
|
|
* Another thing we have to assure is that if we are in ordered mode
|
|
|
|
* and inode is still attached to the committing transaction, we must
|
|
|
|
* we start writeout of all the dirty pages which are being truncated.
|
|
|
|
* This way we are sure that all the data written in the previous
|
|
|
|
* transaction are already on disk (truncate waits for pages under
|
|
|
|
* writeback).
|
|
|
|
*
|
2022-01-21 15:06:11 +08:00
|
|
|
* Called with inode->i_rwsem down.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2021-01-21 21:19:43 +08:00
|
|
|
int ext4_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
|
|
|
|
struct iattr *attr)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2015-03-18 06:25:59 +08:00
|
|
|
struct inode *inode = d_inode(dentry);
|
2006-10-11 16:20:50 +08:00
|
|
|
int error, rc = 0;
|
2010-10-28 10:08:46 +08:00
|
|
|
int orphan = 0;
|
2006-10-11 16:20:50 +08:00
|
|
|
const unsigned int ia_valid = attr->ia_valid;
|
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2019-06-10 10:04:33 +08:00
|
|
|
if (unlikely(IS_IMMUTABLE(inode)))
|
|
|
|
return -EPERM;
|
|
|
|
|
|
|
|
if (unlikely(IS_APPEND(inode) &&
|
|
|
|
(ia_valid & (ATTR_MODE | ATTR_UID |
|
|
|
|
ATTR_GID | ATTR_TIMES_SET))))
|
|
|
|
return -EPERM;
|
|
|
|
|
2021-01-21 21:19:57 +08:00
|
|
|
error = setattr_prepare(mnt_userns, dentry, attr);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2017-10-19 08:21:58 +08:00
|
|
|
error = fscrypt_prepare_setattr(dentry, attr);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2019-07-23 00:26:24 +08:00
|
|
|
error = fsverity_prepare_setattr(dentry, attr);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2022-06-21 22:14:54 +08:00
|
|
|
if (is_quota_modification(mnt_userns, inode, attr)) {
|
2015-06-29 22:22:54 +08:00
|
|
|
error = dquot_initialize(inode);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
}
|
2021-12-24 04:21:37 +08:00
|
|
|
|
2022-06-21 22:14:54 +08:00
|
|
|
if (i_uid_needs_update(mnt_userns, attr, inode) ||
|
|
|
|
i_gid_needs_update(mnt_userns, attr, inode)) {
|
2006-10-11 16:20:50 +08:00
|
|
|
handle_t *handle;
|
|
|
|
|
|
|
|
/* (user+group)*(old+new) structure, inode write (sb,
|
|
|
|
* inode block, ? - but truncate inode update has it) */
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_QUOTA,
|
|
|
|
(EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) +
|
|
|
|
EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb)) + 3);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
error = PTR_ERR(handle);
|
|
|
|
goto err_out;
|
|
|
|
}
|
2017-06-22 23:46:48 +08:00
|
|
|
|
|
|
|
/* dquot_transfer() calls back ext4_get_inode_usage() which
|
|
|
|
* counts xattr inode references.
|
|
|
|
*/
|
|
|
|
down_read(&EXT4_I(inode)->xattr_sem);
|
2022-06-21 22:14:54 +08:00
|
|
|
error = dquot_transfer(mnt_userns, inode, attr);
|
2017-06-22 23:46:48 +08:00
|
|
|
up_read(&EXT4_I(inode)->xattr_sem);
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
if (error) {
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_journal_stop(handle);
|
2006-10-11 16:20:50 +08:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
/* Update corresponding info in inode so that everything is in
|
|
|
|
* one transaction */
|
2022-06-21 22:14:54 +08:00
|
|
|
i_uid_update(mnt_userns, attr, inode);
|
|
|
|
i_gid_update(mnt_userns, attr, inode);
|
2006-10-11 16:20:53 +08:00
|
|
|
error = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
ext4_journal_stop(handle);
|
2021-01-17 16:57:32 +08:00
|
|
|
if (unlikely(error)) {
|
2020-04-27 09:34:37 +08:00
|
|
|
return error;
|
2021-01-17 16:57:32 +08:00
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2015-06-22 12:31:26 +08:00
|
|
|
if (attr->ia_valid & ATTR_SIZE) {
|
2013-08-17 22:07:17 +08:00
|
|
|
handle_t *handle;
|
2015-06-22 12:31:26 +08:00
|
|
|
loff_t oldsize = inode->i_size;
|
2022-03-26 14:53:51 +08:00
|
|
|
loff_t old_disksize;
|
2019-05-30 23:56:23 +08:00
|
|
|
int shrink = (attr->ia_size < inode->i_size);
|
2011-06-25 02:29:45 +08:00
|
|
|
|
2010-05-17 10:00:00 +08:00
|
|
|
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
|
2008-01-29 12:58:27 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
|
2020-10-16 04:37:57 +08:00
|
|
|
if (attr->ia_size > sbi->s_bitmap_maxbytes) {
|
2010-07-27 23:56:06 +08:00
|
|
|
return -EFBIG;
|
2020-10-16 04:37:57 +08:00
|
|
|
}
|
2008-01-29 12:58:27 +08:00
|
|
|
}
|
2020-10-16 04:37:57 +08:00
|
|
|
if (!S_ISREG(inode->i_mode)) {
|
2015-06-22 12:31:26 +08:00
|
|
|
return -EINVAL;
|
2020-10-16 04:37:57 +08:00
|
|
|
}
|
2013-11-19 23:17:07 +08:00
|
|
|
|
|
|
|
if (IS_I_VERSION(inode) && attr->ia_size != inode->i_size)
|
|
|
|
inode_inc_iversion(inode);
|
|
|
|
|
2019-05-30 23:56:23 +08:00
|
|
|
if (shrink) {
|
|
|
|
if (ext4_should_order_data(inode)) {
|
|
|
|
error = ext4_begin_ordered_truncate(inode,
|
2008-07-12 07:27:31 +08:00
|
|
|
attr->ia_size);
|
2019-05-30 23:56:23 +08:00
|
|
|
if (error)
|
|
|
|
goto err_out;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Blocks are going to be removed from the inode. Wait
|
|
|
|
* for dio in flight.
|
|
|
|
*/
|
|
|
|
inode_dio_wait(inode);
|
|
|
|
}
|
|
|
|
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_lock(inode->i_mapping);
|
2019-05-30 23:56:23 +08:00
|
|
|
|
|
|
|
rc = ext4_break_layouts(inode);
|
|
|
|
if (rc) {
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(inode->i_mapping);
|
2020-10-16 04:37:57 +08:00
|
|
|
goto err_out;
|
2015-06-22 12:31:26 +08:00
|
|
|
}
|
2019-05-30 23:56:23 +08:00
|
|
|
|
2015-06-22 12:31:26 +08:00
|
|
|
if (attr->ia_size != inode->i_size) {
|
2013-08-17 22:07:17 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_INODE, 3);
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
error = PTR_ERR(handle);
|
2019-05-30 23:56:23 +08:00
|
|
|
goto out_mmap_sem;
|
2013-08-17 22:07:17 +08:00
|
|
|
}
|
2015-06-22 12:31:26 +08:00
|
|
|
if (ext4_handle_valid(handle) && shrink) {
|
2013-08-17 22:07:17 +08:00
|
|
|
error = ext4_orphan_add(handle, inode);
|
|
|
|
orphan = 1;
|
|
|
|
}
|
2015-07-29 03:08:41 +08:00
|
|
|
/*
|
|
|
|
* Update c/mtime on truncate up, ext4_truncate() will
|
|
|
|
* update c/mtime in shrink case below
|
|
|
|
*/
|
|
|
|
if (!shrink) {
|
2016-11-15 10:40:10 +08:00
|
|
|
inode->i_mtime = current_time(inode);
|
2015-07-29 03:08:41 +08:00
|
|
|
inode->i_ctime = inode->i_mtime;
|
|
|
|
}
|
2020-10-16 04:37:57 +08:00
|
|
|
|
|
|
|
if (shrink)
|
2020-11-06 11:58:53 +08:00
|
|
|
ext4_fc_track_range(handle, inode,
|
2020-10-16 04:37:57 +08:00
|
|
|
(attr->ia_size > 0 ? attr->ia_size - 1 : 0) >>
|
|
|
|
inode->i_sb->s_blocksize_bits,
|
2021-12-23 11:23:37 +08:00
|
|
|
EXT_MAX_BLOCKS - 1);
|
2020-10-16 04:37:57 +08:00
|
|
|
else
|
|
|
|
ext4_fc_track_range(
|
2020-11-06 11:58:53 +08:00
|
|
|
handle, inode,
|
2020-10-16 04:37:57 +08:00
|
|
|
(oldsize > 0 ? oldsize - 1 : oldsize) >>
|
|
|
|
inode->i_sb->s_blocksize_bits,
|
|
|
|
(attr->ia_size > 0 ? attr->ia_size - 1 : 0) >>
|
|
|
|
inode->i_sb->s_blocksize_bits);
|
|
|
|
|
2013-08-17 22:09:31 +08:00
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
2022-03-26 14:53:51 +08:00
|
|
|
old_disksize = EXT4_I(inode)->i_disksize;
|
2013-08-17 22:07:17 +08:00
|
|
|
EXT4_I(inode)->i_disksize = attr->ia_size;
|
|
|
|
rc = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
if (!error)
|
|
|
|
error = rc;
|
2013-08-17 22:09:31 +08:00
|
|
|
/*
|
|
|
|
* We have to update i_size under i_data_sem together
|
|
|
|
* with i_disksize to avoid races with writeback code
|
|
|
|
* running ext4_wb_update_i_disksize().
|
|
|
|
*/
|
|
|
|
if (!error)
|
|
|
|
i_size_write(inode, attr->ia_size);
|
2022-03-26 14:53:51 +08:00
|
|
|
else
|
|
|
|
EXT4_I(inode)->i_disksize = old_disksize;
|
2013-08-17 22:09:31 +08:00
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
2013-08-17 22:07:17 +08:00
|
|
|
ext4_journal_stop(handle);
|
2019-05-30 23:56:23 +08:00
|
|
|
if (error)
|
|
|
|
goto out_mmap_sem;
|
|
|
|
if (!shrink) {
|
|
|
|
pagecache_isize_extended(inode, oldsize,
|
|
|
|
inode->i_size);
|
|
|
|
} else if (ext4_should_journal_data(inode)) {
|
|
|
|
ext4_wait_for_tail_page_commit(inode);
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
2014-10-02 09:49:46 +08:00
|
|
|
}
|
2018-07-30 05:00:22 +08:00
|
|
|
|
2013-08-17 22:07:17 +08:00
|
|
|
/*
|
|
|
|
* Truncate pagecache after we've waited for commit
|
|
|
|
* in data=journal mode to make pages freeable.
|
|
|
|
*/
|
2015-02-17 07:59:38 +08:00
|
|
|
truncate_pagecache(inode, inode->i_size);
|
2019-05-30 23:56:23 +08:00
|
|
|
/*
|
|
|
|
* Call ext4_truncate() even if i_size didn't change to
|
|
|
|
* truncate possible preallocated blocks.
|
|
|
|
*/
|
|
|
|
if (attr->ia_size <= oldsize) {
|
2016-11-14 11:02:26 +08:00
|
|
|
rc = ext4_truncate(inode);
|
|
|
|
if (rc)
|
|
|
|
error = rc;
|
|
|
|
}
|
2019-05-30 23:56:23 +08:00
|
|
|
out_mmap_sem:
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(inode->i_mapping);
|
2011-05-24 03:13:02 +08:00
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2016-11-14 11:02:26 +08:00
|
|
|
if (!error) {
|
2021-01-21 21:19:57 +08:00
|
|
|
setattr_copy(mnt_userns, inode, attr);
|
2010-06-04 17:30:02 +08:00
|
|
|
mark_inode_dirty(inode);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the call to ext4_truncate failed to get a transaction handle at
|
|
|
|
* all, we need to clean up the in-core orphan list manually.
|
|
|
|
*/
|
2010-10-28 10:08:46 +08:00
|
|
|
if (orphan && inode->i_nlink)
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_orphan_del(NULL, inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2016-11-14 11:02:26 +08:00
|
|
|
if (!error && (ia_valid & ATTR_MODE))
|
2021-01-21 21:19:57 +08:00
|
|
|
rc = posix_acl_chmod(mnt_userns, inode, inode->i_mode);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
err_out:
|
2020-10-16 04:37:57 +08:00
|
|
|
if (error)
|
|
|
|
ext4_std_error(inode->i_sb, error);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (!error)
|
|
|
|
error = rc;
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2021-01-21 21:19:43 +08:00
|
|
|
int ext4_getattr(struct user_namespace *mnt_userns, const struct path *path,
|
|
|
|
struct kstat *stat, u32 request_mask, unsigned int query_flags)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2017-04-01 01:31:56 +08:00
|
|
|
struct inode *inode = d_inode(path->dentry);
|
|
|
|
struct ext4_inode *raw_inode;
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
unsigned int flags;
|
|
|
|
|
2019-11-29 11:26:51 +08:00
|
|
|
if ((request_mask & STATX_BTIME) &&
|
|
|
|
EXT4_FITS_IN_INODE(raw_inode, ei, i_crtime)) {
|
2017-04-01 01:31:56 +08:00
|
|
|
stat->result_mask |= STATX_BTIME;
|
|
|
|
stat->btime.tv_sec = ei->i_crtime.tv_sec;
|
|
|
|
stat->btime.tv_nsec = ei->i_crtime.tv_nsec;
|
|
|
|
}
|
|
|
|
|
|
|
|
flags = ei->i_flags & EXT4_FL_USER_VISIBLE;
|
|
|
|
if (flags & EXT4_APPEND_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_APPEND;
|
|
|
|
if (flags & EXT4_COMPR_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_COMPRESSED;
|
|
|
|
if (flags & EXT4_ENCRYPT_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_ENCRYPTED;
|
|
|
|
if (flags & EXT4_IMMUTABLE_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_IMMUTABLE;
|
|
|
|
if (flags & EXT4_NODUMP_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_NODUMP;
|
2019-10-30 04:41:39 +08:00
|
|
|
if (flags & EXT4_VERITY_FL)
|
|
|
|
stat->attributes |= STATX_ATTR_VERITY;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2017-04-01 01:32:17 +08:00
|
|
|
stat->attributes_mask |= (STATX_ATTR_APPEND |
|
|
|
|
STATX_ATTR_COMPRESSED |
|
|
|
|
STATX_ATTR_ENCRYPTED |
|
|
|
|
STATX_ATTR_IMMUTABLE |
|
2019-10-30 04:41:39 +08:00
|
|
|
STATX_ATTR_NODUMP |
|
|
|
|
STATX_ATTR_VERITY);
|
2017-04-01 01:32:17 +08:00
|
|
|
|
2021-01-21 21:19:57 +08:00
|
|
|
generic_fillattr(mnt_userns, inode, stat);
|
2017-04-01 01:31:56 +08:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2021-01-21 21:19:43 +08:00
|
|
|
int ext4_file_getattr(struct user_namespace *mnt_userns,
|
|
|
|
const struct path *path, struct kstat *stat,
|
2017-04-01 01:31:56 +08:00
|
|
|
u32 request_mask, unsigned int query_flags)
|
|
|
|
{
|
|
|
|
struct inode *inode = d_inode(path->dentry);
|
|
|
|
u64 delalloc_blocks;
|
|
|
|
|
2021-01-21 21:19:57 +08:00
|
|
|
ext4_getattr(mnt_userns, path, stat, request_mask, query_flags);
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2013-11-12 11:38:12 +08:00
|
|
|
/*
|
|
|
|
* If there is inline data in the inode, the inode will normally not
|
|
|
|
* have data blocks allocated (it may have an external xattr block).
|
|
|
|
* Report at least one sector for such files, so tools like tar, rsync,
|
2017-03-26 05:33:31 +08:00
|
|
|
* others don't incorrectly think the file is completely sparse.
|
2013-11-12 11:38:12 +08:00
|
|
|
*/
|
|
|
|
if (unlikely(ext4_has_inline_data(inode)))
|
|
|
|
stat->blocks += (stat->size + 511) >> 9;
|
|
|
|
|
2008-07-12 07:27:31 +08:00
|
|
|
/*
|
|
|
|
* We can't update i_blocks if the block allocation is delayed
|
|
|
|
* otherwise in the case of system crash before the real block
|
|
|
|
* allocation is done, we will have i_blocks inconsistent with
|
|
|
|
* on-disk file blocks.
|
|
|
|
* We always keep i_blocks updated together with real
|
|
|
|
* allocation. But to not confuse with user, stat
|
|
|
|
* will return the blocks that include the delayed allocation
|
|
|
|
* blocks for this file.
|
|
|
|
*/
|
2012-06-01 10:54:16 +08:00
|
|
|
delalloc_blocks = EXT4_C2B(EXT4_SB(inode->i_sb),
|
2013-11-12 11:38:12 +08:00
|
|
|
EXT4_I(inode)->i_reserved_data_blocks);
|
|
|
|
stat->blocks += delalloc_blocks << (inode->i_sb->s_blocksize_bits - 9);
|
2008-07-12 07:27:31 +08:00
|
|
|
return 0;
|
|
|
|
}
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2013-06-05 01:01:11 +08:00
|
|
|
static int ext4_index_trans_blocks(struct inode *inode, int lblocks,
|
|
|
|
int pextents)
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
{
|
2010-05-17 10:00:00 +08:00
|
|
|
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
|
2013-06-05 01:01:11 +08:00
|
|
|
return ext4_ind_trans_blocks(inode, lblocks);
|
|
|
|
return ext4_ext_index_trans_blocks(inode, pextents);
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
}
|
2008-11-07 05:49:36 +08:00
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
* Account for index blocks, block groups bitmaps and block group
|
|
|
|
* descriptor blocks if modify datablocks and index blocks
|
|
|
|
* worse case, the indexs blocks spread over different block groups
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
* If datablocks are discontiguous, they are possible to spread over
|
2012-09-02 01:31:09 +08:00
|
|
|
* different block groups too. If they are contiguous, with flexbg,
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
* they could still across block group boundary.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
* Also account for superblock, inode, quota and xattr blocks
|
|
|
|
*/
|
2017-06-22 23:44:55 +08:00
|
|
|
static int ext4_meta_trans_blocks(struct inode *inode, int lblocks,
|
2013-06-05 01:01:11 +08:00
|
|
|
int pextents)
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
{
|
2009-05-01 20:50:38 +08:00
|
|
|
ext4_group_t groups, ngroups = ext4_get_groups_count(inode->i_sb);
|
|
|
|
int gdpblocks;
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
int idxblocks;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/*
|
2013-06-05 01:01:11 +08:00
|
|
|
* How many index blocks need to touch to map @lblocks logical blocks
|
|
|
|
* to @pextents physical extents?
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
*/
|
2013-06-05 01:01:11 +08:00
|
|
|
idxblocks = ext4_index_trans_blocks(inode, lblocks, pextents);
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
|
|
|
|
ret = idxblocks;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Now let's see how many group bitmaps and group descriptors need
|
|
|
|
* to account
|
|
|
|
*/
|
2013-06-05 01:01:11 +08:00
|
|
|
groups = idxblocks + pextents;
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
gdpblocks = groups;
|
2009-05-01 20:50:38 +08:00
|
|
|
if (groups > ngroups)
|
|
|
|
groups = ngroups;
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
if (groups > EXT4_SB(inode->i_sb)->s_gdb_count)
|
|
|
|
gdpblocks = EXT4_SB(inode->i_sb)->s_gdb_count;
|
|
|
|
|
|
|
|
/* bitmaps and block group descriptor blocks */
|
|
|
|
ret += groups + gdpblocks;
|
|
|
|
|
|
|
|
/* Blocks for super block, inode, quota and xattr blocks */
|
|
|
|
ret += EXT4_META_TRANS_BLOCKS(inode->i_sb);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2011-03-31 09:57:33 +08:00
|
|
|
* Calculate the total number of credits to reserve to fit
|
2008-08-20 10:16:03 +08:00
|
|
|
* the modification of a single pages into a single transaction,
|
|
|
|
* which may include multiple chunks of block allocations.
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2008-08-20 10:15:58 +08:00
|
|
|
* This could be called via ext4_write_begin()
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
2008-08-20 10:15:58 +08:00
|
|
|
* We need to consider the worse case, when
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
* one new block per extent.
|
2006-10-11 16:20:50 +08:00
|
|
|
*/
|
2006-10-11 16:21:03 +08:00
|
|
|
int ext4_writepage_trans_blocks(struct inode *inode)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2006-10-11 16:20:53 +08:00
|
|
|
int bpp = ext4_journal_blocks_per_page(inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
int ret;
|
|
|
|
|
2013-06-05 01:01:11 +08:00
|
|
|
ret = ext4_meta_trans_blocks(inode, bpp, bpp);
|
2006-10-11 16:21:03 +08:00
|
|
|
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
/* Account for data blocks for journalled mode */
|
2006-10-11 16:20:53 +08:00
|
|
|
if (ext4_should_journal_data(inode))
|
ext4: journal credits calulation cleanup and fix for non-extent writepage
When considering how many journal credits are needed for modifying a
chunk of data, we need to account for the super block, inode block,
quota blocks and xattr block, indirect/index blocks, also, group bitmap
and group descriptor blocks for new allocation (including data and
indirect/index blocks). There are many places in ext4 do the calculation
on their own and often missed one or two meta blocks, and often they
assume single block allocation, and did not considering the multile
chunk of allocation case.
This patch is trying to cleanup current journal credit code, provides
some common helper funtion to calculate the journal credits, to be used
for writepage, writepages, DIO, fallocate, migration, defrag, and for
both nonextent and extent files.
This patch modified the writepage/write_begin credit caculation for
nonextent files, to use the new helper function. It also fixed the
problem that writepage on nonextent files did not consider the case
blocksize <pagesize, thus could possibelly need multiple block
allocation in a single transaction.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2008-08-20 10:16:07 +08:00
|
|
|
ret += bpp;
|
2006-10-11 16:20:50 +08:00
|
|
|
return ret;
|
|
|
|
}
|
2008-08-20 10:16:03 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Calculate the journal credits for a chunk of data modification.
|
|
|
|
*
|
|
|
|
* This is called from DIO, fallocate or whoever calling
|
2010-07-27 23:56:07 +08:00
|
|
|
* ext4_map_blocks() to map/allocate a chunk of contiguous disk blocks.
|
2008-08-20 10:16:03 +08:00
|
|
|
*
|
|
|
|
* journal buffers for data blocks are not included here, as DIO
|
|
|
|
* and fallocate do no need to journal data buffers.
|
|
|
|
*/
|
|
|
|
int ext4_chunk_trans_blocks(struct inode *inode, int nrblocks)
|
|
|
|
{
|
|
|
|
return ext4_meta_trans_blocks(inode, nrblocks, 1);
|
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* The caller must have previously called ext4_reserve_inode_write().
|
2006-10-11 16:20:50 +08:00
|
|
|
* Give this, we know that the caller already has write access to iloc->bh.
|
|
|
|
*/
|
2006-10-11 16:20:53 +08:00
|
|
|
int ext4_mark_iloc_dirty(handle_t *handle,
|
2009-06-15 05:45:34 +08:00
|
|
|
struct inode *inode, struct ext4_iloc *iloc)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
|
2018-11-07 05:49:50 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) {
|
|
|
|
put_bh(iloc->bh);
|
2017-02-05 14:28:48 +08:00
|
|
|
return -EIO;
|
2018-11-07 05:49:50 +08:00
|
|
|
}
|
2020-11-06 11:58:53 +08:00
|
|
|
ext4_fc_track_inode(handle, inode);
|
2020-10-16 04:37:57 +08:00
|
|
|
|
2012-03-03 01:23:11 +08:00
|
|
|
if (IS_I_VERSION(inode))
|
2008-01-29 12:58:27 +08:00
|
|
|
inode_inc_iversion(inode);
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/* the do_update_inode consumes one bh->b_count */
|
|
|
|
get_bh(iloc->bh);
|
|
|
|
|
2006-10-11 16:21:01 +08:00
|
|
|
/* ext4_do_update_inode() does jbd2_journal_dirty_metadata */
|
2009-09-29 22:07:47 +08:00
|
|
|
err = ext4_do_update_inode(handle, inode, iloc);
|
2006-10-11 16:20:50 +08:00
|
|
|
put_bh(iloc->bh);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* On success, We end up with an outstanding reference count against
|
|
|
|
* iloc->bh. This _must_ be cleaned up later.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_reserve_inode_write(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_iloc *iloc)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2009-01-07 13:06:22 +08:00
|
|
|
int err;
|
|
|
|
|
2017-02-05 14:28:48 +08:00
|
|
|
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
|
|
|
|
return -EIO;
|
|
|
|
|
2009-01-07 13:06:22 +08:00
|
|
|
err = ext4_get_inode_loc(inode, iloc);
|
|
|
|
if (!err) {
|
|
|
|
BUFFER_TRACE(iloc->bh, "get_write_access");
|
2021-08-16 17:57:04 +08:00
|
|
|
err = ext4_journal_get_write_access(handle, inode->i_sb,
|
|
|
|
iloc->bh, EXT4_JTR_NONE);
|
2009-01-07 13:06:22 +08:00
|
|
|
if (err) {
|
|
|
|
brelse(iloc->bh);
|
|
|
|
iloc->bh = NULL;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
}
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_std_error(inode->i_sb, err);
|
2006-10-11 16:20:50 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
static int __ext4_expand_extra_isize(struct inode *inode,
|
|
|
|
unsigned int new_extra_isize,
|
|
|
|
struct ext4_iloc *iloc,
|
|
|
|
handle_t *handle, int *no_expand)
|
|
|
|
{
|
|
|
|
struct ext4_inode *raw_inode;
|
|
|
|
struct ext4_xattr_ibody_header *header;
|
2019-11-08 10:43:41 +08:00
|
|
|
unsigned int inode_size = EXT4_INODE_SIZE(inode->i_sb);
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
2017-08-06 13:00:49 +08:00
|
|
|
int error;
|
|
|
|
|
2019-11-08 10:43:41 +08:00
|
|
|
/* this was checked at iget time, but double check for good measure */
|
|
|
|
if ((EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize > inode_size) ||
|
|
|
|
(ei->i_extra_isize & 3)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "bad extra_isize %u (inode size %u)",
|
|
|
|
ei->i_extra_isize,
|
|
|
|
EXT4_INODE_SIZE(inode->i_sb));
|
|
|
|
return -EFSCORRUPTED;
|
|
|
|
}
|
|
|
|
if ((new_extra_isize < ei->i_extra_isize) ||
|
|
|
|
(new_extra_isize < 4) ||
|
|
|
|
(new_extra_isize > inode_size - EXT4_GOOD_OLD_INODE_SIZE))
|
|
|
|
return -EINVAL; /* Should never happen */
|
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
raw_inode = ext4_raw_inode(iloc);
|
|
|
|
|
|
|
|
header = IHDR(inode, raw_inode);
|
|
|
|
|
|
|
|
/* No extended attributes present */
|
|
|
|
if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR) ||
|
|
|
|
header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) {
|
|
|
|
memset((void *)raw_inode + EXT4_GOOD_OLD_INODE_SIZE +
|
|
|
|
EXT4_I(inode)->i_extra_isize, 0,
|
|
|
|
new_extra_isize - EXT4_I(inode)->i_extra_isize);
|
|
|
|
EXT4_I(inode)->i_extra_isize = new_extra_isize;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* try to expand with EAs present */
|
|
|
|
error = ext4_expand_extra_isize_ea(inode, new_extra_isize,
|
|
|
|
raw_inode, handle);
|
|
|
|
if (error) {
|
|
|
|
/*
|
|
|
|
* Inode size expansion failed; don't try again
|
|
|
|
*/
|
|
|
|
*no_expand = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2007-07-18 21:19:57 +08:00
|
|
|
/*
|
|
|
|
* Expand an inode by new_extra_isize bytes.
|
|
|
|
* Returns 0 on success or negative error number on failure.
|
|
|
|
*/
|
2017-08-06 12:40:01 +08:00
|
|
|
static int ext4_try_to_expand_extra_isize(struct inode *inode,
|
|
|
|
unsigned int new_extra_isize,
|
|
|
|
struct ext4_iloc iloc,
|
|
|
|
handle_t *handle)
|
2007-07-18 21:19:57 +08:00
|
|
|
{
|
2017-08-06 12:27:38 +08:00
|
|
|
int no_expand;
|
|
|
|
int error;
|
2007-07-18 21:19:57 +08:00
|
|
|
|
2017-08-06 12:40:01 +08:00
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND))
|
|
|
|
return -EOVERFLOW;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* In nojournal mode, we can immediately attempt to expand
|
|
|
|
* the inode. When journaled, we first need to obtain extra
|
|
|
|
* buffer credits since we may write into the EA block
|
|
|
|
* with this same handle. If journal_extend fails, then it will
|
|
|
|
* only result in a minor loss of functionality for that inode.
|
|
|
|
* If this is felt to be critical, then e2fsck should be run to
|
|
|
|
* force a large enough s_min_extra_isize.
|
|
|
|
*/
|
2019-11-06 00:44:14 +08:00
|
|
|
if (ext4_journal_extend(handle,
|
2019-11-06 00:44:29 +08:00
|
|
|
EXT4_DATA_TRANS_BLOCKS(inode->i_sb), 0) != 0)
|
2017-08-06 12:40:01 +08:00
|
|
|
return -ENOSPC;
|
2007-07-18 21:19:57 +08:00
|
|
|
|
2017-08-06 12:27:38 +08:00
|
|
|
if (ext4_write_trylock_xattr(inode, &no_expand) == 0)
|
2017-08-06 12:40:01 +08:00
|
|
|
return -EBUSY;
|
2017-08-06 12:27:38 +08:00
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
error = __ext4_expand_extra_isize(inode, new_extra_isize, &iloc,
|
|
|
|
handle, &no_expand);
|
|
|
|
ext4_write_unlock_xattr(inode, &no_expand);
|
2007-07-18 21:19:57 +08:00
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
return error;
|
|
|
|
}
|
2007-07-18 21:19:57 +08:00
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
int ext4_expand_extra_isize(struct inode *inode,
|
|
|
|
unsigned int new_extra_isize,
|
|
|
|
struct ext4_iloc *iloc)
|
|
|
|
{
|
|
|
|
handle_t *handle;
|
|
|
|
int no_expand;
|
|
|
|
int error, rc;
|
|
|
|
|
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND)) {
|
|
|
|
brelse(iloc->bh);
|
|
|
|
return -EOVERFLOW;
|
2007-07-18 21:19:57 +08:00
|
|
|
}
|
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_INODE,
|
|
|
|
EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
error = PTR_ERR(handle);
|
|
|
|
brelse(iloc->bh);
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
ext4_write_lock_xattr(inode, &no_expand);
|
|
|
|
|
2019-02-22 00:29:10 +08:00
|
|
|
BUFFER_TRACE(iloc->bh, "get_write_access");
|
2021-08-16 17:57:04 +08:00
|
|
|
error = ext4_journal_get_write_access(handle, inode->i_sb, iloc->bh,
|
|
|
|
EXT4_JTR_NONE);
|
2017-08-06 12:27:38 +08:00
|
|
|
if (error) {
|
2017-08-06 13:00:49 +08:00
|
|
|
brelse(iloc->bh);
|
2019-12-14 02:50:11 +08:00
|
|
|
goto out_unlock;
|
2017-08-06 12:27:38 +08:00
|
|
|
}
|
2017-08-06 12:40:01 +08:00
|
|
|
|
2017-08-06 13:00:49 +08:00
|
|
|
error = __ext4_expand_extra_isize(inode, new_extra_isize, iloc,
|
|
|
|
handle, &no_expand);
|
|
|
|
|
|
|
|
rc = ext4_mark_iloc_dirty(handle, inode, iloc);
|
|
|
|
if (!error)
|
|
|
|
error = rc;
|
|
|
|
|
2019-12-14 02:50:11 +08:00
|
|
|
out_unlock:
|
2017-08-06 13:00:49 +08:00
|
|
|
ext4_write_unlock_xattr(inode, &no_expand);
|
|
|
|
ext4_journal_stop(handle);
|
2017-08-06 12:27:38 +08:00
|
|
|
return error;
|
2007-07-18 21:19:57 +08:00
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:50 +08:00
|
|
|
/*
|
|
|
|
* What we do here is to mark the in-core inode as clean with respect to inode
|
|
|
|
* dirtiness (it may still be data-dirty).
|
|
|
|
* This means that the in-core inode may be reaped by prune_icache
|
|
|
|
* without having to perform any I/O. This is a very good thing,
|
|
|
|
* because *any* task may call prune_icache - even ones which
|
|
|
|
* have a transaction open against a different journal.
|
|
|
|
*
|
|
|
|
* Is this cheating? Not really. Sure, we haven't written the
|
|
|
|
* inode out, but prune_icache isn't a user-visible syncing function.
|
|
|
|
* Whenever the user wants stuff synced (sys_sync, sys_msync, sys_fsync)
|
|
|
|
* we start and wait on commits.
|
|
|
|
*/
|
2020-04-27 09:34:37 +08:00
|
|
|
int __ext4_mark_inode_dirty(handle_t *handle, struct inode *inode,
|
|
|
|
const char *func, unsigned int line)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
2006-10-11 16:20:53 +08:00
|
|
|
struct ext4_iloc iloc;
|
2007-07-18 21:19:57 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2017-08-06 12:40:01 +08:00
|
|
|
int err;
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
might_sleep();
|
2010-11-09 02:51:33 +08:00
|
|
|
trace_ext4_mark_inode_dirty(inode, _RET_IP_);
|
2006-10-11 16:20:53 +08:00
|
|
|
err = ext4_reserve_inode_write(handle, inode, &iloc);
|
2016-03-13 10:40:32 +08:00
|
|
|
if (err)
|
2020-04-27 09:34:37 +08:00
|
|
|
goto out;
|
2017-08-06 12:40:01 +08:00
|
|
|
|
|
|
|
if (EXT4_I(inode)->i_extra_isize < sbi->s_want_extra_isize)
|
|
|
|
ext4_try_to_expand_extra_isize(inode, sbi->s_want_extra_isize,
|
|
|
|
iloc, handle);
|
|
|
|
|
2020-04-27 09:34:37 +08:00
|
|
|
err = ext4_mark_iloc_dirty(handle, inode, &iloc);
|
|
|
|
out:
|
|
|
|
if (unlikely(err))
|
|
|
|
ext4_error_inode_err(inode, func, line, 0, err,
|
|
|
|
"mark_inode_dirty error");
|
|
|
|
return err;
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 16:20:53 +08:00
|
|
|
* ext4_dirty_inode() is called from __mark_inode_dirty()
|
2006-10-11 16:20:50 +08:00
|
|
|
*
|
|
|
|
* We're really interested in the case where a file is being extended.
|
|
|
|
* i_size has been changed by generic_commit_write() and we thus need
|
|
|
|
* to include the updated inode in the current transaction.
|
|
|
|
*
|
2010-03-03 22:05:00 +08:00
|
|
|
* Also, dquot_alloc_block() will always dirty the inode when blocks
|
2006-10-11 16:20:50 +08:00
|
|
|
* are allocated to the file.
|
|
|
|
*
|
|
|
|
* If the inode is marked synchronous, we don't honour that here - doing
|
|
|
|
* so would cause a commit on atime updates, which we don't bother doing.
|
|
|
|
* We handle synchronous inodes at the highest possible level.
|
|
|
|
*/
|
2011-05-27 18:53:02 +08:00
|
|
|
void ext4_dirty_inode(struct inode *inode, int flags)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
handle_t *handle;
|
|
|
|
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_INODE, 2);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (IS_ERR(handle))
|
2021-01-13 03:02:47 +08:00
|
|
|
return;
|
2009-09-30 04:06:01 +08:00
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_journal_stop(handle);
|
2006-10-11 16:20:50 +08:00
|
|
|
}
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
int ext4_change_inode_journal_flag(struct inode *inode, int val)
|
2006-10-11 16:20:50 +08:00
|
|
|
{
|
|
|
|
journal_t *journal;
|
|
|
|
handle_t *handle;
|
|
|
|
int err;
|
2016-04-26 11:22:35 +08:00
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* We have to be very careful here: changing a data block's
|
|
|
|
* journaling status dynamically is dangerous. If we write a
|
|
|
|
* data block to the journal, change the status and then delete
|
|
|
|
* that block, we risk forgetting to revoke the old log record
|
|
|
|
* from the journal and so a subsequent replay can corrupt data.
|
|
|
|
* So, first we make sure that the journal is empty and that
|
|
|
|
* nobody is changing anything.
|
|
|
|
*/
|
|
|
|
|
2006-10-11 16:20:53 +08:00
|
|
|
journal = EXT4_JOURNAL(inode);
|
2009-01-07 13:06:22 +08:00
|
|
|
if (!journal)
|
|
|
|
return 0;
|
2007-07-18 20:33:51 +08:00
|
|
|
if (is_journal_aborted(journal))
|
2006-10-11 16:20:50 +08:00
|
|
|
return -EROFS;
|
|
|
|
|
2012-09-29 12:41:21 +08:00
|
|
|
/* Wait for all existing dio workers */
|
|
|
|
inode_dio_wait(inode);
|
|
|
|
|
ext4: handle unwritten or delalloc buffers before enabling data journaling
We already allocate delalloc blocks before changing the inode mode into
"per-file data journal" mode to prevent delalloc blocks from remaining
not allocated, but another issue concerned with "BH_Unwritten" status
still exists. For example, by fallocate(), several buffers' status
change into "BH_Unwritten", but these buffers cannot be processed by
ext4_alloc_da_blocks(). So, they still remain in unwritten status after
per-file data journaling is enabled and they cannot be changed into
written status any more and, if they are journaled and eventually
checkpointed, these unwritten buffer will cause a kernel panic by the
below BUG_ON() function of submit_bh_wbc() when they are submitted
during checkpointing.
static int submit_bh_wbc(int rw, struct buffer_head *bh,...
{
...
BUG_ON(buffer_unwritten(bh));
Moreover, when "dioread_nolock" option is enabled, the status of a
buffer is changed into "BH_Unwritten" after write_begin() completes and
the "BH_Unwritten" status will be cleared after I/O is done. Therefore,
if a buffer's status is changed into unwrutten but the buffer's I/O is
not submitted and completed, it can cause the same problem after
enabling per-file data journaling. You can easily generate this bug by
executing the following command.
./kvm-xfstests -C 10000 -m nodelalloc,dioread_nolock generic/269
To resolve these problems and define a boundary between the previous
mode and per-file data journaling mode, we need to flush and wait all
the I/O of buffers of a file before enabling per-file data journaling
of the file.
Signed-off-by: Daeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
2016-04-26 11:21:00 +08:00
|
|
|
/*
|
|
|
|
* Before flushing the journal and switching inode's aops, we have
|
|
|
|
* to flush all dirty data the inode has. There can be outstanding
|
|
|
|
* delayed allocations, there can be unwritten extents created by
|
|
|
|
* fallocate or buffered writes in dioread_nolock mode covered by
|
|
|
|
* dirty data which can be converted only after flushing the dirty
|
|
|
|
* data (and journalled aops don't know how to handle these cases).
|
|
|
|
*/
|
|
|
|
if (val) {
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_lock(inode->i_mapping);
|
ext4: handle unwritten or delalloc buffers before enabling data journaling
We already allocate delalloc blocks before changing the inode mode into
"per-file data journal" mode to prevent delalloc blocks from remaining
not allocated, but another issue concerned with "BH_Unwritten" status
still exists. For example, by fallocate(), several buffers' status
change into "BH_Unwritten", but these buffers cannot be processed by
ext4_alloc_da_blocks(). So, they still remain in unwritten status after
per-file data journaling is enabled and they cannot be changed into
written status any more and, if they are journaled and eventually
checkpointed, these unwritten buffer will cause a kernel panic by the
below BUG_ON() function of submit_bh_wbc() when they are submitted
during checkpointing.
static int submit_bh_wbc(int rw, struct buffer_head *bh,...
{
...
BUG_ON(buffer_unwritten(bh));
Moreover, when "dioread_nolock" option is enabled, the status of a
buffer is changed into "BH_Unwritten" after write_begin() completes and
the "BH_Unwritten" status will be cleared after I/O is done. Therefore,
if a buffer's status is changed into unwrutten but the buffer's I/O is
not submitted and completed, it can cause the same problem after
enabling per-file data journaling. You can easily generate this bug by
executing the following command.
./kvm-xfstests -C 10000 -m nodelalloc,dioread_nolock generic/269
To resolve these problems and define a boundary between the previous
mode and per-file data journaling mode, we need to flush and wait all
the I/O of buffers of a file before enabling per-file data journaling
of the file.
Signed-off-by: Daeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
2016-04-26 11:21:00 +08:00
|
|
|
err = filemap_write_and_wait(inode->i_mapping);
|
|
|
|
if (err < 0) {
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(inode->i_mapping);
|
ext4: handle unwritten or delalloc buffers before enabling data journaling
We already allocate delalloc blocks before changing the inode mode into
"per-file data journal" mode to prevent delalloc blocks from remaining
not allocated, but another issue concerned with "BH_Unwritten" status
still exists. For example, by fallocate(), several buffers' status
change into "BH_Unwritten", but these buffers cannot be processed by
ext4_alloc_da_blocks(). So, they still remain in unwritten status after
per-file data journaling is enabled and they cannot be changed into
written status any more and, if they are journaled and eventually
checkpointed, these unwritten buffer will cause a kernel panic by the
below BUG_ON() function of submit_bh_wbc() when they are submitted
during checkpointing.
static int submit_bh_wbc(int rw, struct buffer_head *bh,...
{
...
BUG_ON(buffer_unwritten(bh));
Moreover, when "dioread_nolock" option is enabled, the status of a
buffer is changed into "BH_Unwritten" after write_begin() completes and
the "BH_Unwritten" status will be cleared after I/O is done. Therefore,
if a buffer's status is changed into unwrutten but the buffer's I/O is
not submitted and completed, it can cause the same problem after
enabling per-file data journaling. You can easily generate this bug by
executing the following command.
./kvm-xfstests -C 10000 -m nodelalloc,dioread_nolock generic/269
To resolve these problems and define a boundary between the previous
mode and per-file data journaling mode, we need to flush and wait all
the I/O of buffers of a file before enabling per-file data journaling
of the file.
Signed-off-by: Daeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
2016-04-26 11:21:00 +08:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_down_write(&sbi->s_writepages_rwsem);
|
2006-10-11 16:21:01 +08:00
|
|
|
jbd2_journal_lock_updates(journal);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* OK, there are no updates running now, and all cached data is
|
|
|
|
* synced to disk. We are now in a completely consistent state
|
|
|
|
* which doesn't have anything in the journal, and we know that
|
|
|
|
* no filesystem updates are running, so it is safe to modify
|
|
|
|
* the inode's in-core data-journaling state flag now.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (val)
|
2010-05-17 10:00:00 +08:00
|
|
|
ext4_set_inode_flag(inode, EXT4_INODE_JOURNAL_DATA);
|
2011-12-29 02:55:51 +08:00
|
|
|
else {
|
2021-05-18 23:13:25 +08:00
|
|
|
err = jbd2_journal_flush(journal, 0);
|
2014-10-30 22:53:17 +08:00
|
|
|
if (err < 0) {
|
|
|
|
jbd2_journal_unlock_updates(journal);
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_up_write(&sbi->s_writepages_rwsem);
|
2014-10-30 22:53:17 +08:00
|
|
|
return err;
|
|
|
|
}
|
2010-05-17 10:00:00 +08:00
|
|
|
ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA);
|
2011-12-29 02:55:51 +08:00
|
|
|
}
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_set_aops(inode);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
2006-10-11 16:21:01 +08:00
|
|
|
jbd2_journal_unlock_updates(journal);
|
2020-02-20 02:30:46 +08:00
|
|
|
percpu_up_write(&sbi->s_writepages_rwsem);
|
2016-04-26 11:22:35 +08:00
|
|
|
|
ext4: handle unwritten or delalloc buffers before enabling data journaling
We already allocate delalloc blocks before changing the inode mode into
"per-file data journal" mode to prevent delalloc blocks from remaining
not allocated, but another issue concerned with "BH_Unwritten" status
still exists. For example, by fallocate(), several buffers' status
change into "BH_Unwritten", but these buffers cannot be processed by
ext4_alloc_da_blocks(). So, they still remain in unwritten status after
per-file data journaling is enabled and they cannot be changed into
written status any more and, if they are journaled and eventually
checkpointed, these unwritten buffer will cause a kernel panic by the
below BUG_ON() function of submit_bh_wbc() when they are submitted
during checkpointing.
static int submit_bh_wbc(int rw, struct buffer_head *bh,...
{
...
BUG_ON(buffer_unwritten(bh));
Moreover, when "dioread_nolock" option is enabled, the status of a
buffer is changed into "BH_Unwritten" after write_begin() completes and
the "BH_Unwritten" status will be cleared after I/O is done. Therefore,
if a buffer's status is changed into unwrutten but the buffer's I/O is
not submitted and completed, it can cause the same problem after
enabling per-file data journaling. You can easily generate this bug by
executing the following command.
./kvm-xfstests -C 10000 -m nodelalloc,dioread_nolock generic/269
To resolve these problems and define a boundary between the previous
mode and per-file data journaling mode, we need to flush and wait all
the I/O of buffers of a file before enabling per-file data journaling
of the file.
Signed-off-by: Daeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
2016-04-26 11:21:00 +08:00
|
|
|
if (val)
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock(inode->i_mapping);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
/* Finally we can mark the inode as dirty. */
|
|
|
|
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
|
2006-10-11 16:20:50 +08:00
|
|
|
if (IS_ERR(handle))
|
|
|
|
return PTR_ERR(handle);
|
|
|
|
|
2020-10-16 04:37:57 +08:00
|
|
|
ext4_fc_mark_ineligible(inode->i_sb,
|
2022-01-17 17:36:54 +08:00
|
|
|
EXT4_FC_REASON_JOURNAL_FLAG_CHANGE, handle);
|
2006-10-11 16:20:53 +08:00
|
|
|
err = ext4_mark_inode_dirty(handle, inode);
|
2009-01-07 13:06:22 +08:00
|
|
|
ext4_handle_sync(handle);
|
2006-10-11 16:20:53 +08:00
|
|
|
ext4_journal_stop(handle);
|
|
|
|
ext4_std_error(inode->i_sb, err);
|
2006-10-11 16:20:50 +08:00
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2021-08-16 17:57:04 +08:00
|
|
|
static int ext4_bh_unmapped(handle_t *handle, struct inode *inode,
|
|
|
|
struct buffer_head *bh)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
|
|
|
return !buffer_mapped(bh);
|
|
|
|
}
|
|
|
|
|
2018-10-03 10:20:50 +08:00
|
|
|
vm_fault_t ext4_page_mkwrite(struct vm_fault *vmf)
|
2008-07-12 07:27:31 +08:00
|
|
|
{
|
2017-02-25 06:56:41 +08:00
|
|
|
struct vm_area_struct *vma = vmf->vma;
|
2009-04-01 06:23:21 +08:00
|
|
|
struct page *page = vmf->page;
|
2008-07-12 07:27:31 +08:00
|
|
|
loff_t size;
|
|
|
|
unsigned long len;
|
2018-10-03 10:20:50 +08:00
|
|
|
int err;
|
|
|
|
vm_fault_t ret;
|
2008-07-12 07:27:31 +08:00
|
|
|
struct file *file = vma->vm_file;
|
2013-01-24 06:07:38 +08:00
|
|
|
struct inode *inode = file_inode(file);
|
2008-07-12 07:27:31 +08:00
|
|
|
struct address_space *mapping = inode->i_mapping;
|
2011-06-25 02:29:41 +08:00
|
|
|
handle_t *handle;
|
|
|
|
get_block_t *get_block;
|
|
|
|
int retries = 0;
|
2008-07-12 07:27:31 +08:00
|
|
|
|
2019-06-10 10:04:33 +08:00
|
|
|
if (unlikely(IS_IMMUTABLE(inode)))
|
|
|
|
return VM_FAULT_SIGBUS;
|
|
|
|
|
2012-06-12 22:20:38 +08:00
|
|
|
sb_start_pagefault(inode->i_sb);
|
2012-10-01 11:04:56 +08:00
|
|
|
file_update_time(vma->vm_file);
|
2015-12-08 03:28:03 +08:00
|
|
|
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_lock_shared(mapping);
|
2017-04-30 12:10:50 +08:00
|
|
|
|
2018-10-03 10:20:50 +08:00
|
|
|
err = ext4_convert_inline_data(inode);
|
|
|
|
if (err)
|
2017-04-30 12:10:50 +08:00
|
|
|
goto out_ret;
|
|
|
|
|
ext4: data=journal: fixes for ext4_page_mkwrite()
These are two fixes for data journalling required by
the next patch, discovered while testing it.
First, the optimization to return early if all buffers
are mapped is not appropriate for the next patch:
The inode _must_ be added to the transaction's list in
data=journal mode (so to write-protect pages on commit)
thus we cannot return early there.
Second, once that optimization to reduce transactions
was disabled for data=journal mode, more transactions
happened, and occasionally hit this warning message:
'JBD2: Spotted dirty metadata buffer'.
Reason is, block_page_mkwrite() will set_buffer_dirty()
before do_journal_get_write_access() that is there to
prevent it. This issue was masked by the optimization.
So, on data=journal use __block_write_begin() instead.
This also requires page locking and len recalculation.
(see block_page_mkwrite() for implementation details.)
Finally, as Jan noted there is little sharing between
data=journal and other modes in ext4_page_mkwrite().
However, a prototype of ext4_journalled_page_mkwrite()
showed there still would be lots of duplicated lines
(tens of) that didn't seem worth it.
Thus this patch ends up with an ugly goto to skip all
non-data journalling code (to avoid long indentations,
but that can be changed..) in the beginning, and just
a conditional in the transaction section.
Well, we skip a common part to data journalling which
is the page truncated check, but we do it again after
ext4_journal_start() when we re-acquire the page lock
(so not to acquire the page lock twice needlessly for
data journalling.)
Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
Suggested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20201006004841.600488-4-mfo@canonical.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2020-10-06 08:48:40 +08:00
|
|
|
/*
|
|
|
|
* On data journalling we skip straight to the transaction handle:
|
|
|
|
* there's no delalloc; page truncated will be checked later; the
|
|
|
|
* early return w/ all buffers mapped (calculates size/len) can't
|
|
|
|
* be used; and there's no dioread_nolock, so only ext4_get_block.
|
|
|
|
*/
|
|
|
|
if (ext4_should_journal_data(inode))
|
|
|
|
goto retry_alloc;
|
|
|
|
|
2011-06-25 02:29:41 +08:00
|
|
|
/* Delalloc case is easy... */
|
|
|
|
if (test_opt(inode->i_sb, DELALLOC) &&
|
|
|
|
!ext4_nonda_switch(inode->i_sb)) {
|
|
|
|
do {
|
2018-10-03 10:20:50 +08:00
|
|
|
err = block_page_mkwrite(vma, vmf,
|
2011-06-25 02:29:41 +08:00
|
|
|
ext4_da_get_block_prep);
|
2018-10-03 10:20:50 +08:00
|
|
|
} while (err == -ENOSPC &&
|
2011-06-25 02:29:41 +08:00
|
|
|
ext4_should_retry_alloc(inode->i_sb, &retries));
|
|
|
|
goto out_ret;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
2011-05-19 01:55:20 +08:00
|
|
|
|
|
|
|
lock_page(page);
|
2011-06-25 02:29:41 +08:00
|
|
|
size = i_size_read(inode);
|
|
|
|
/* Page got truncated from under us? */
|
|
|
|
if (page->mapping != mapping || page_offset(page) > size) {
|
|
|
|
unlock_page(page);
|
|
|
|
ret = VM_FAULT_NOPAGE;
|
|
|
|
goto out;
|
2011-05-19 01:55:20 +08:00
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
if (page->index == size >> PAGE_SHIFT)
|
|
|
|
len = size & ~PAGE_MASK;
|
2008-07-12 07:27:31 +08:00
|
|
|
else
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 20:29:47 +08:00
|
|
|
len = PAGE_SIZE;
|
2009-09-10 10:36:03 +08:00
|
|
|
/*
|
2011-06-25 02:29:41 +08:00
|
|
|
* Return if we have all the buffers mapped. This avoids the need to do
|
|
|
|
* journal_start/journal_stop which can block and take a long time
|
ext4: data=journal: fixes for ext4_page_mkwrite()
These are two fixes for data journalling required by
the next patch, discovered while testing it.
First, the optimization to return early if all buffers
are mapped is not appropriate for the next patch:
The inode _must_ be added to the transaction's list in
data=journal mode (so to write-protect pages on commit)
thus we cannot return early there.
Second, once that optimization to reduce transactions
was disabled for data=journal mode, more transactions
happened, and occasionally hit this warning message:
'JBD2: Spotted dirty metadata buffer'.
Reason is, block_page_mkwrite() will set_buffer_dirty()
before do_journal_get_write_access() that is there to
prevent it. This issue was masked by the optimization.
So, on data=journal use __block_write_begin() instead.
This also requires page locking and len recalculation.
(see block_page_mkwrite() for implementation details.)
Finally, as Jan noted there is little sharing between
data=journal and other modes in ext4_page_mkwrite().
However, a prototype of ext4_journalled_page_mkwrite()
showed there still would be lots of duplicated lines
(tens of) that didn't seem worth it.
Thus this patch ends up with an ugly goto to skip all
non-data journalling code (to avoid long indentations,
but that can be changed..) in the beginning, and just
a conditional in the transaction section.
Well, we skip a common part to data journalling which
is the page truncated check, but we do it again after
ext4_journal_start() when we re-acquire the page lock
(so not to acquire the page lock twice needlessly for
data journalling.)
Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
Suggested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20201006004841.600488-4-mfo@canonical.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2020-10-06 08:48:40 +08:00
|
|
|
*
|
|
|
|
* This cannot be done for data journalling, as we have to add the
|
|
|
|
* inode to the transaction's list to writeprotect pages on commit.
|
2009-09-10 10:36:03 +08:00
|
|
|
*/
|
2008-07-12 07:27:31 +08:00
|
|
|
if (page_has_buffers(page)) {
|
2021-08-16 17:57:04 +08:00
|
|
|
if (!ext4_walk_page_buffers(NULL, inode, page_buffers(page),
|
2012-12-11 03:05:51 +08:00
|
|
|
0, len, NULL,
|
|
|
|
ext4_bh_unmapped)) {
|
2011-06-25 02:29:41 +08:00
|
|
|
/* Wait so that we don't change page under IO */
|
mm: only enforce stable page writes if the backing device requires it
Create a helper function to check if a backing device requires stable
page writes and, if so, performs the necessary wait. Then, make it so
that all points in the memory manager that handle making pages writable
use the helper function. This should provide stable page write support
to most filesystems, while eliminating unnecessary waiting for devices
that don't require the feature.
Before this patchset, all filesystems would block, regardless of whether
or not it was necessary. ext3 would wait, but still generate occasional
checksum errors. The network filesystems were left to do their own
thing, so they'd wait too.
After this patchset, all the disk filesystems except ext3 and btrfs will
wait only if the hardware requires it. ext3 (if necessary) snapshots
pages instead of blocking, and btrfs provides its own bdi so the mm will
never wait. Network filesystems haven't been touched, so either they
provide their own stable page guarantees or they don't block at all.
The blocking behavior is back to what it was before 3.0 if you don't
have a disk requiring stable page writes.
Here's the result of using dbench to test latency on ext2:
3.8.0-rc3:
Operation Count AvgLat MaxLat
----------------------------------------
WriteX 109347 0.028 59.817
ReadX 347180 0.004 3.391
Flush 15514 29.828 287.283
Throughput 57.429 MB/sec 4 clients 4 procs max_latency=287.290 ms
3.8.0-rc3 + patches:
WriteX 105556 0.029 4.273
ReadX 335004 0.005 4.112
Flush 14982 30.540 298.634
Throughput 55.4496 MB/sec 4 clients 4 procs max_latency=298.650 ms
As you can see, the maximum write latency drops considerably with this
patch enabled. The other filesystems (ext3/ext4/xfs/btrfs) behave
similarly, but see the cover letter for those results.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Artem Bityutskiy <dedekind1@gmail.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Eric Van Hensbergen <ericvh@gmail.com>
Cc: Ron Minnich <rminnich@sandia.gov>
Cc: Latchesar Ionkov <lucho@ionkov.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-22 08:42:51 +08:00
|
|
|
wait_for_stable_page(page);
|
2011-06-25 02:29:41 +08:00
|
|
|
ret = VM_FAULT_LOCKED;
|
|
|
|
goto out;
|
2009-09-10 10:36:03 +08:00
|
|
|
}
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|
2009-09-10 10:36:03 +08:00
|
|
|
unlock_page(page);
|
2011-06-25 02:29:41 +08:00
|
|
|
/* OK, we need to fill the hole... */
|
|
|
|
if (ext4_should_dioread_nolock(inode))
|
2016-03-09 12:08:10 +08:00
|
|
|
get_block = ext4_get_block_unwritten;
|
2011-06-25 02:29:41 +08:00
|
|
|
else
|
|
|
|
get_block = ext4_get_block;
|
|
|
|
retry_alloc:
|
2013-02-09 10:59:22 +08:00
|
|
|
handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE,
|
|
|
|
ext4_writepage_trans_blocks(inode));
|
2011-06-25 02:29:41 +08:00
|
|
|
if (IS_ERR(handle)) {
|
2009-04-01 06:23:21 +08:00
|
|
|
ret = VM_FAULT_SIGBUS;
|
2011-06-25 02:29:41 +08:00
|
|
|
goto out;
|
|
|
|
}
|
ext4: data=journal: fixes for ext4_page_mkwrite()
These are two fixes for data journalling required by
the next patch, discovered while testing it.
First, the optimization to return early if all buffers
are mapped is not appropriate for the next patch:
The inode _must_ be added to the transaction's list in
data=journal mode (so to write-protect pages on commit)
thus we cannot return early there.
Second, once that optimization to reduce transactions
was disabled for data=journal mode, more transactions
happened, and occasionally hit this warning message:
'JBD2: Spotted dirty metadata buffer'.
Reason is, block_page_mkwrite() will set_buffer_dirty()
before do_journal_get_write_access() that is there to
prevent it. This issue was masked by the optimization.
So, on data=journal use __block_write_begin() instead.
This also requires page locking and len recalculation.
(see block_page_mkwrite() for implementation details.)
Finally, as Jan noted there is little sharing between
data=journal and other modes in ext4_page_mkwrite().
However, a prototype of ext4_journalled_page_mkwrite()
showed there still would be lots of duplicated lines
(tens of) that didn't seem worth it.
Thus this patch ends up with an ugly goto to skip all
non-data journalling code (to avoid long indentations,
but that can be changed..) in the beginning, and just
a conditional in the transaction section.
Well, we skip a common part to data journalling which
is the page truncated check, but we do it again after
ext4_journal_start() when we re-acquire the page lock
(so not to acquire the page lock twice needlessly for
data journalling.)
Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
Suggested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20201006004841.600488-4-mfo@canonical.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2020-10-06 08:48:40 +08:00
|
|
|
/*
|
|
|
|
* Data journalling can't use block_page_mkwrite() because it
|
|
|
|
* will set_buffer_dirty() before do_journal_get_write_access()
|
|
|
|
* thus might hit warning messages for dirty metadata buffers.
|
|
|
|
*/
|
|
|
|
if (!ext4_should_journal_data(inode)) {
|
|
|
|
err = block_page_mkwrite(vma, vmf, get_block);
|
|
|
|
} else {
|
|
|
|
lock_page(page);
|
|
|
|
size = i_size_read(inode);
|
|
|
|
/* Page got truncated from under us? */
|
|
|
|
if (page->mapping != mapping || page_offset(page) > size) {
|
|
|
|
ret = VM_FAULT_NOPAGE;
|
2020-10-06 08:48:41 +08:00
|
|
|
goto out_error;
|
2011-06-25 02:29:41 +08:00
|
|
|
}
|
ext4: data=journal: fixes for ext4_page_mkwrite()
These are two fixes for data journalling required by
the next patch, discovered while testing it.
First, the optimization to return early if all buffers
are mapped is not appropriate for the next patch:
The inode _must_ be added to the transaction's list in
data=journal mode (so to write-protect pages on commit)
thus we cannot return early there.
Second, once that optimization to reduce transactions
was disabled for data=journal mode, more transactions
happened, and occasionally hit this warning message:
'JBD2: Spotted dirty metadata buffer'.
Reason is, block_page_mkwrite() will set_buffer_dirty()
before do_journal_get_write_access() that is there to
prevent it. This issue was masked by the optimization.
So, on data=journal use __block_write_begin() instead.
This also requires page locking and len recalculation.
(see block_page_mkwrite() for implementation details.)
Finally, as Jan noted there is little sharing between
data=journal and other modes in ext4_page_mkwrite().
However, a prototype of ext4_journalled_page_mkwrite()
showed there still would be lots of duplicated lines
(tens of) that didn't seem worth it.
Thus this patch ends up with an ugly goto to skip all
non-data journalling code (to avoid long indentations,
but that can be changed..) in the beginning, and just
a conditional in the transaction section.
Well, we skip a common part to data journalling which
is the page truncated check, but we do it again after
ext4_journal_start() when we re-acquire the page lock
(so not to acquire the page lock twice needlessly for
data journalling.)
Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
Suggested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20201006004841.600488-4-mfo@canonical.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2020-10-06 08:48:40 +08:00
|
|
|
|
|
|
|
if (page->index == size >> PAGE_SHIFT)
|
|
|
|
len = size & ~PAGE_MASK;
|
|
|
|
else
|
|
|
|
len = PAGE_SIZE;
|
|
|
|
|
|
|
|
err = __block_write_begin(page, 0, len, ext4_get_block);
|
|
|
|
if (!err) {
|
2020-10-06 08:48:41 +08:00
|
|
|
ret = VM_FAULT_SIGBUS;
|
2021-08-16 17:57:04 +08:00
|
|
|
if (ext4_walk_page_buffers(handle, inode,
|
|
|
|
page_buffers(page), 0, len, NULL,
|
|
|
|
do_journal_get_write_access))
|
2020-10-06 08:48:41 +08:00
|
|
|
goto out_error;
|
2021-08-16 17:57:04 +08:00
|
|
|
if (ext4_walk_page_buffers(handle, inode,
|
|
|
|
page_buffers(page), 0, len, NULL,
|
|
|
|
write_end_fn))
|
2020-10-06 08:48:41 +08:00
|
|
|
goto out_error;
|
2020-10-27 21:27:51 +08:00
|
|
|
if (ext4_jbd2_inode_add_write(handle, inode,
|
|
|
|
page_offset(page), len))
|
2020-10-06 08:48:41 +08:00
|
|
|
goto out_error;
|
ext4: data=journal: fixes for ext4_page_mkwrite()
These are two fixes for data journalling required by
the next patch, discovered while testing it.
First, the optimization to return early if all buffers
are mapped is not appropriate for the next patch:
The inode _must_ be added to the transaction's list in
data=journal mode (so to write-protect pages on commit)
thus we cannot return early there.
Second, once that optimization to reduce transactions
was disabled for data=journal mode, more transactions
happened, and occasionally hit this warning message:
'JBD2: Spotted dirty metadata buffer'.
Reason is, block_page_mkwrite() will set_buffer_dirty()
before do_journal_get_write_access() that is there to
prevent it. This issue was masked by the optimization.
So, on data=journal use __block_write_begin() instead.
This also requires page locking and len recalculation.
(see block_page_mkwrite() for implementation details.)
Finally, as Jan noted there is little sharing between
data=journal and other modes in ext4_page_mkwrite().
However, a prototype of ext4_journalled_page_mkwrite()
showed there still would be lots of duplicated lines
(tens of) that didn't seem worth it.
Thus this patch ends up with an ugly goto to skip all
non-data journalling code (to avoid long indentations,
but that can be changed..) in the beginning, and just
a conditional in the transaction section.
Well, we skip a common part to data journalling which
is the page truncated check, but we do it again after
ext4_journal_start() when we re-acquire the page lock
(so not to acquire the page lock twice needlessly for
data journalling.)
Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
Suggested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20201006004841.600488-4-mfo@canonical.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2020-10-06 08:48:40 +08:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_JDATA);
|
|
|
|
} else {
|
|
|
|
unlock_page(page);
|
|
|
|
}
|
2011-06-25 02:29:41 +08:00
|
|
|
}
|
|
|
|
ext4_journal_stop(handle);
|
2018-10-03 10:20:50 +08:00
|
|
|
if (err == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
|
2011-06-25 02:29:41 +08:00
|
|
|
goto retry_alloc;
|
|
|
|
out_ret:
|
2018-10-03 10:20:50 +08:00
|
|
|
ret = block_page_mkwrite_return(err);
|
2011-06-25 02:29:41 +08:00
|
|
|
out:
|
2021-02-05 01:05:42 +08:00
|
|
|
filemap_invalidate_unlock_shared(mapping);
|
2012-06-12 22:20:38 +08:00
|
|
|
sb_end_pagefault(inode->i_sb);
|
2008-07-12 07:27:31 +08:00
|
|
|
return ret;
|
2020-10-06 08:48:41 +08:00
|
|
|
out_error:
|
|
|
|
unlock_page(page);
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
goto out;
|
2008-07-12 07:27:31 +08:00
|
|
|
}
|