OpenCloudOS-Kernel/fs/xfs/libxfs/xfs_trans_resv.c

950 lines
31 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
* Copyright (C) 2010 Red Hat, Inc.
* All Rights Reserved.
*/
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_da_format.h"
#include "xfs_da_btree.h"
#include "xfs_inode.h"
#include "xfs_bmap_btree.h"
#include "xfs_quota.h"
#include "xfs_trans.h"
#include "xfs_qm.h"
#include "xfs_trans_space.h"
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
#define _ALLOC true
#define _FREE false
/*
* A buffer has a format structure overhead in the log in addition
* to the data, so we need to take this into account when reserving
* space in a transaction for a buffer. Round the space required up
* to a multiple of 128 bytes so that we don't change the historical
* reservation that has been used for this overhead.
*/
STATIC uint
xfs_buf_log_overhead(void)
{
return round_up(sizeof(struct xlog_op_header) +
sizeof(struct xfs_buf_log_format), 128);
}
/*
* Calculate out transaction log reservation per item in bytes.
*
* The nbufs argument is used to indicate the number of items that
* will be changed in a transaction. size is used to tell how many
* bytes should be reserved per item.
*/
STATIC uint
xfs_calc_buf_res(
uint nbufs,
uint size)
{
return nbufs * (size + xfs_buf_log_overhead());
}
/*
* Per-extent log reservation for the btree changes involved in freeing or
* allocating an extent. In classic XFS there were two trees that will be
* modified (bnobt + cntbt). With rmap enabled, there are three trees
* (rmapbt). With reflink, there are four trees (refcountbt). The number of
* blocks reserved is based on the formula:
*
* num trees * ((2 blocks/level * max depth) - 1)
*
* Keep in mind that max depth is calculated separately for each type of tree.
*/
uint
xfs_allocfree_log_count(
struct xfs_mount *mp,
uint num_ops)
{
uint blocks;
blocks = num_ops * 2 * (2 * mp->m_ag_maxlevels - 1);
if (xfs_sb_version_hasrmapbt(&mp->m_sb))
blocks += num_ops * (2 * mp->m_rmap_maxlevels - 1);
if (xfs_sb_version_hasreflink(&mp->m_sb))
blocks += num_ops * (2 * mp->m_refc_maxlevels - 1);
return blocks;
}
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
/*
* Logging inodes is really tricksy. They are logged in memory format,
* which means that what we write into the log doesn't directly translate into
* the amount of space they use on disk.
*
* Case in point - btree format forks in memory format use more space than the
* on-disk format. In memory, the buffer contains a normal btree block header so
* the btree code can treat it as though it is just another generic buffer.
* However, when we write it to the inode fork, we don't write all of this
* header as it isn't needed. e.g. the root is only ever in the inode, so
* there's no need for sibling pointers which would waste 16 bytes of space.
*
* Hence when we have an inode with a maximally sized btree format fork, then
* amount of information we actually log is greater than the size of the inode
* on disk. Hence we need an inode reservation function that calculates all this
* correctly. So, we log:
*
* - 4 log op headers for object
* - for the ilf, the inode core and 2 forks
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
* - inode log format object
* - the inode core
* - two inode forks containing bmap btree root blocks.
* - the btree data contained by both forks will fit into the inode size,
* hence when combined with the inode core above, we have a total of the
* actual inode size.
* - the BMBT headers need to be accounted separately, as they are
* additional to the records and pointers that fit inside the inode
* forks.
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
*/
STATIC uint
xfs_calc_inode_res(
struct xfs_mount *mp,
uint ninodes)
{
return ninodes *
(4 * sizeof(struct xlog_op_header) +
sizeof(struct xfs_inode_log_format) +
mp->m_sb.sb_inodesize +
2 * XFS_BMBT_BLOCK_LEN(mp));
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
}
/*
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* Inode btree record insertion/removal modifies the inode btree and free space
* btrees (since the inobt does not use the agfl). This requires the following
* reservation:
*
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* the inode btree: max depth * blocksize
* the allocation btrees: 2 trees * (max depth - 1) * block size
*
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* The caller must account for SB and AG header modifications, etc.
*/
STATIC uint
xfs_calc_inobt_res(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(M_IGEO(mp)->inobt_maxlevels,
XFS_FSB_TO_B(mp, 1)) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1));
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
}
/*
* The free inode btree is a conditional feature. The behavior differs slightly
* from that of the traditional inode btree in that the finobt tracks records
* for inode chunks with at least one free inode. A record can be removed from
* the tree during individual inode allocation. Therefore the finobt
* reservation is unconditional for both the inode chunk allocation and
* individual inode allocation (modify) cases.
*
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* Behavior aside, the reservation for finobt modification is equivalent to the
* traditional inobt: cover a full finobt shape change plus block allocation.
*/
STATIC uint
xfs_calc_finobt_res(
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
struct xfs_mount *mp)
{
if (!xfs_sb_version_hasfinobt(&mp->m_sb))
return 0;
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
return xfs_calc_inobt_res(mp);
}
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
/*
* Calculate the reservation required to allocate or free an inode chunk. This
* includes:
*
* the allocation btrees: 2 trees * (max depth - 1) * block size
* the inode chunk: m_ino_geo.ialloc_blks * N
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
*
* The size N of the inode chunk reservation depends on whether it is for
* allocation or free and which type of create transaction is in use. An inode
* chunk free always invalidates the buffers and only requires reservation for
* headers (N == 0). An inode chunk allocation requires a chunk sized
* reservation on v4 and older superblocks to initialize the chunk. No chunk
* reservation is required for allocation on v5 supers, which use ordered
* buffers to initialize.
*/
STATIC uint
xfs_calc_inode_chunk_res(
struct xfs_mount *mp,
bool alloc)
{
uint res, size = 0;
res = xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1));
if (alloc) {
/* icreate tx uses ordered buffers */
if (xfs_sb_version_has_v3inode(&mp->m_sb))
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
return res;
size = XFS_FSB_TO_B(mp, 1);
}
res += xfs_calc_buf_res(M_IGEO(mp)->ialloc_blks, size);
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
return res;
}
/*
* Per-extent log reservation for the btree changes involved in freeing or
* allocating a realtime extent. We have to be able to log as many rtbitmap
* blocks as needed to mark inuse MAXEXTLEN blocks' worth of realtime extents,
* as well as the realtime summary block.
*/
static unsigned int
xfs_rtalloc_log_count(
struct xfs_mount *mp,
unsigned int num_ops)
{
unsigned int blksz = XFS_FSB_TO_B(mp, 1);
unsigned int rtbmp_bytes;
rtbmp_bytes = (MAXEXTLEN / mp->m_sb.sb_rextsize) / NBBY;
return (howmany(rtbmp_bytes, blksz) + 1) * num_ops;
}
/*
* Various log reservation values.
*
* These are based on the size of the file system block because that is what
* most transactions manipulate. Each adds in an additional 128 bytes per
* item logged to try to account for the overhead of the transaction mechanism.
*
* Note: Most of the reservations underestimate the number of allocation
* groups into which they could free extents in the xfs_defer_finish() call.
* This is because the number in the worst case is quite high and quite
* unusual. In order to fix this we need to change xfs_defer_finish() to free
* extents in only a single AG at a time. This will require changes to the
* EFI code as well, however, so that the EFI for the extents not freed is
* logged again in each transaction. See SGI PV #261917.
*
* Reservation functions here avoid a huge stack in xfs_trans_init due to
* register overflow from temporaries in the calculations.
*/
/*
* In a write transaction we can allocate a maximum of 2
* extents. This gives (t1):
* the inode getting the new extents: inode size
* the inode's bmap btree: max depth * block size
* the agfs of the ags from which the extents are allocated: 2 * sector
* the superblock free block counter: sector size
* the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
* Or, if we're writing to a realtime file (t2):
* the inode getting the new extents: inode size
* the inode's bmap btree: max depth * block size
* the agfs of the ags from which the extents are allocated: 2 * sector
* the superblock free block counter: sector size
* the realtime bitmap: ((MAXEXTLEN / rtextsize) / NBBY) bytes
* the realtime summary: 1 block
* the allocation btrees: 2 trees * (2 * max depth - 1) * block size
* And the bmap_finish transaction can free bmap blocks in a join (t3):
* the agfs of the ags containing the blocks: 2 * sector size
* the agfls of the ags containing the blocks: 2 * sector size
* the super block free block counter: sector size
* the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_write_reservation(
struct xfs_mount *mp)
{
unsigned int t1, t2, t3;
unsigned int blksz = XFS_FSB_TO_B(mp, 1);
t1 = xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), blksz) +
xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz);
if (xfs_sb_version_hasrealtime(&mp->m_sb)) {
t2 = xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK),
blksz) +
xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_rtalloc_log_count(mp, 1), blksz) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1), blksz);
} else {
t2 = 0;
}
t3 = xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz);
return XFS_DQUOT_LOGRES(mp) + max3(t1, t2, t3);
}
/*
* In truncating a file we free up to two extents at once. We can modify (t1):
* the inode being truncated: inode size
* the inode's bmap btree: (max depth + 1) * block size
* And the bmap_finish transaction can free the blocks and bmap blocks (t2):
* the agf for each of the ags: 4 * sector size
* the agfl for each of the ags: 4 * sector size
* the super block to reflect the freed blocks: sector size
* worst case split in allocation btrees per extent assuming 4 extents:
* 4 exts * 2 trees * (2 * max depth - 1) * block size
* Or, if it's a realtime file (t3):
* the agf for each of the ags: 2 * sector size
* the agfl for each of the ags: 2 * sector size
* the super block to reflect the freed blocks: sector size
* the realtime bitmap: 2 exts * ((MAXEXTLEN / rtextsize) / NBBY) bytes
* the realtime summary: 2 exts * 1 block
* worst case split in allocation btrees per extent assuming 2 extents:
* 2 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_itruncate_reservation(
struct xfs_mount *mp)
{
unsigned int t1, t2, t3;
unsigned int blksz = XFS_FSB_TO_B(mp, 1);
t1 = xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1, blksz);
t2 = xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 4), blksz);
if (xfs_sb_version_hasrealtime(&mp->m_sb)) {
t3 = xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_rtalloc_log_count(mp, 2), blksz) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz);
} else {
t3 = 0;
}
return XFS_DQUOT_LOGRES(mp) + max3(t1, t2, t3);
}
/*
* In renaming a files we can modify:
* the four inodes involved: 4 * inode size
* the two directory btrees: 2 * (max depth + v2) * dir block size
* the two directory bmap btrees: 2 * max depth * block size
* And the bmap_finish transaction can free dir and bmap blocks (two sets
* of bmap blocks) giving:
* the agf for the ags in which the blocks live: 3 * sector size
* the agfl for the ags in which the blocks live: 3 * sector size
* the superblock for the free block count: sector size
* the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_rename_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
max((xfs_calc_inode_res(mp, 4) +
xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp),
XFS_FSB_TO_B(mp, 1))),
(xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 3),
XFS_FSB_TO_B(mp, 1))));
}
/*
* For removing an inode from unlinked list at first, we can modify:
* the agi hash list and counters: sector size
* the on disk inode before ours in the agi hash list: inode cluster size
* the on disk inode in the agi hash list: inode cluster size
*/
STATIC uint
xfs_calc_iunlink_remove_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
2 * M_IGEO(mp)->inode_cluster_size;
}
/*
* For creating a link to an inode:
* the parent directory inode: inode size
* the linked inode: inode size
* the directory btree could split: (max depth + v2) * dir block size
* the directory bmap btree could join or split: (max depth + v2) * blocksize
* And the bmap_finish transaction can free some bmap blocks giving:
* the agf for the ag in which the blocks live: sector size
* the agfl for the ag in which the blocks live: sector size
* the superblock for the free block count: sector size
* the allocation btrees: 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_link_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs_calc_iunlink_remove_reservation(mp) +
max((xfs_calc_inode_res(mp, 2) +
xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
XFS_FSB_TO_B(mp, 1))),
(xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1))));
}
/*
* For adding an inode to unlinked list we can modify:
* the agi hash list: sector size
* the on disk inode: inode cluster size
*/
STATIC uint
xfs_calc_iunlink_add_reservation(xfs_mount_t *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
M_IGEO(mp)->inode_cluster_size;
}
/*
* For removing a directory entry we can modify:
* the parent directory inode: inode size
* the removed inode: inode size
* the directory btree could join: (max depth + v2) * dir block size
* the directory bmap btree could join or split: (max depth + v2) * blocksize
* And the bmap_finish transaction can free the dir and bmap blocks giving:
* the agf for the ag in which the blocks live: 2 * sector size
* the agfl for the ag in which the blocks live: 2 * sector size
* the superblock for the free block count: sector size
* the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_remove_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs_calc_iunlink_add_reservation(mp) +
max((xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
XFS_FSB_TO_B(mp, 1))),
(xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2),
XFS_FSB_TO_B(mp, 1))));
}
/*
* For create, break it in to the two cases that the transaction
* covers. We start with the modify case - allocation done by modification
* of the state of existing inodes - and the allocation case.
*/
/*
* For create we can modify:
* the parent directory inode: inode size
* the new inode: inode size
* the inode btree entry: block size
* the superblock for the nlink flag: sector size
* the directory btree: (max depth + v2) * dir block size
* the directory inode's bmap btree: (max depth + v2) * block size
* the finobt (record modification and allocation btrees)
*/
STATIC uint
xfs_calc_create_resv_modify(
struct xfs_mount *mp)
{
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
return xfs_calc_inode_res(mp, 2) +
xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
(uint)XFS_FSB_TO_B(mp, 1) +
xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), XFS_FSB_TO_B(mp, 1)) +
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
xfs_calc_finobt_res(mp);
}
/*
* For icreate we can allocate some inodes giving:
* the agi and agf of the ag getting the new inodes: 2 * sectorsize
* the superblock for the nlink flag: sector size
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
* the inode chunk (allocation, optional init)
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* the inobt (record insertion)
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
* the finobt (optional, record insertion)
*/
STATIC uint
xfs_calc_icreate_resv_alloc(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) +
mp->m_sb.sb_sectsize +
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
xfs_calc_inode_chunk_res(mp, _ALLOC) +
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
xfs_calc_inobt_res(mp) +
xfs_calc_finobt_res(mp);
}
STATIC uint
xfs_calc_icreate_reservation(xfs_mount_t *mp)
{
return XFS_DQUOT_LOGRES(mp) +
max(xfs_calc_icreate_resv_alloc(mp),
xfs_calc_create_resv_modify(mp));
}
STATIC uint
xfs_calc_create_tmpfile_reservation(
struct xfs_mount *mp)
{
uint res = XFS_DQUOT_LOGRES(mp);
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
res += xfs_calc_icreate_resv_alloc(mp);
return res + xfs_calc_iunlink_add_reservation(mp);
}
/*
* Making a new directory is the same as creating a new file.
*/
STATIC uint
xfs_calc_mkdir_reservation(
struct xfs_mount *mp)
{
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
return xfs_calc_icreate_reservation(mp);
}
/*
* Making a new symplink is the same as creating a new file, but
* with the added blocks for remote symlink data which can be up to 1kB in
* length (XFS_SYMLINK_MAXLEN).
*/
STATIC uint
xfs_calc_symlink_reservation(
struct xfs_mount *mp)
{
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
return xfs_calc_icreate_reservation(mp) +
xfs_calc_buf_res(1, XFS_SYMLINK_MAXLEN);
}
/*
* In freeing an inode we can modify:
* the inode being freed: inode size
xfs: include inobt buffers in ifree tx log reservation The tr_ifree transaction handles inode unlinks and inode chunk frees. The current transaction calculation does not accurately reflect worst case changes to the inode btree, however. The inobt portion of the current transaction reservation only covers modification of a single inobt buffer (for the particular inode record). This is a historical artifact from the days before XFS supported full inode chunk removal. When support for inode chunk removal was added in commit 254f6311ed1b ("Implement deletion of inode clusters in XFS."), the additional log reservation required for chunk removal was not added correctly. The new reservation only considered the header overhead of associated buffers rather than the full contents of the btrees and AGF and AGFL buffers affected by the transaction. The reservation for the free space btrees was subsequently fixed up in commit 5fe6abb82f76 ("Add space for inode and allocation btrees to ITRUNCATE log reservation"), but the res. for full inobt joins has never been added. Further review of the ifree reservation uncovered a couple more problems: - The undocumented +2 blocks are intended for the AGF and AGFL, but are also not sized correctly and should be logged as full sectors (not FSBs). - The additional single block header is undocumented and serves no apparent purpose. Update xfs_calc_ifree_reservation() to include a full inobt join in the reservation calculation. Refactor the undocumented blocks appropriately and fix up the comments to reflect the current calculation. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:36 +08:00
* the super block free inode counter, AGF and AGFL: sector size
* the on disk inode (agi unlinked list removal)
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
* the inode chunk (invalidated, headers only)
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
* the inode btree
* the finobt (record insertion, removal or modification)
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
*
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
* Note that the inode chunk res. includes an allocfree res. for freeing of the
* inode chunk. This is technically extraneous because the inode chunk free is
* deferred (it occurs after a transaction roll). Include the extra reservation
* anyways since we've had reports of ifree transaction overruns due to too many
* agfl fixups during inode chunk frees.
*/
STATIC uint
xfs_calc_ifree_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 1) +
xfs: include inobt buffers in ifree tx log reservation The tr_ifree transaction handles inode unlinks and inode chunk frees. The current transaction calculation does not accurately reflect worst case changes to the inode btree, however. The inobt portion of the current transaction reservation only covers modification of a single inobt buffer (for the particular inode record). This is a historical artifact from the days before XFS supported full inode chunk removal. When support for inode chunk removal was added in commit 254f6311ed1b ("Implement deletion of inode clusters in XFS."), the additional log reservation required for chunk removal was not added correctly. The new reservation only considered the header overhead of associated buffers rather than the full contents of the btrees and AGF and AGFL buffers affected by the transaction. The reservation for the free space btrees was subsequently fixed up in commit 5fe6abb82f76 ("Add space for inode and allocation btrees to ITRUNCATE log reservation"), but the res. for full inobt joins has never been added. Further review of the ifree reservation uncovered a couple more problems: - The undocumented +2 blocks are intended for the AGF and AGFL, but are also not sized correctly and should be logged as full sectors (not FSBs). - The additional single block header is undocumented and serves no apparent purpose. Update xfs_calc_ifree_reservation() to include a full inobt join in the reservation calculation. Refactor the undocumented blocks appropriately and fix up the comments to reflect the current calculation. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:36 +08:00
xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
xfs_calc_iunlink_remove_reservation(mp) +
xfs: refactor inode chunk alloc/free tx reservation The reservation for the various forms of inode allocation is scattered across several different functions. This includes two variants of chunk allocation (v5 icreate transactions vs. older create transactions) and the inode free transaction. To clean up some of this code and clarify the purpose of specific allocfree reservations, continue the pattern of defining helper functions for smaller operational units of broader transactions. Refactor the reservation into an inode chunk alloc/free helper that considers the various conditions based on filesystem format. An inode chunk free involves an extent free and buffer invalidations. The latter requires reservation for log headers only. An inode chunk allocation modifies the free space btrees and logs the chunk on v4 supers. v5 supers initialize the inode chunk using ordered buffers and so do not log the chunk. As a side effect of this refactoring, add one more allocfree res to the ifree transaction. Technically this does not serve a specific purpose because inode chunks are freed via deferred operations and thus occur after a transaction roll. tr_ifree has a bit of a history of tx overruns caused by too many agfl fixups during sustained file deletion workloads, so add this extra reservation as a form of padding nonetheless. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
xfs_calc_inode_chunk_res(mp, _FREE) +
xfs: include an allocfree res for inobt modifications Analysis of recent reports of log reservation overruns and code inspection has uncovered that the reservations associated with inode operations may not cover the worst case scenarios. In particular, many cases only include one allocfree res. for a particular operation even though said operations may also entail AGFL fixups and inode btree block allocations in addition to the actual inode chunk allocation. This can easily turn into two or three block allocations (or frees) per operation. In theory, the only way to define the worst case reservation is to include an allocfree res for each individual allocation in a transaction. Since that is impractical (we can perform multiple agfl fixups per tx and not every allocation results in a full tree operation), we need to find a reasonable compromise that addresses the deficiency in practice without blowing out the size of the transactions. Since the inode btrees are not filled by the AGFL, record insertion and removal can directly result in block allocations and frees depending on the shape of the tree. These allocations and frees occur in the same transaction context as the inobt update itself, but are separate from the allocation/free that might be required for an inode chunk. Therefore, it makes sense to assume that an [f]inobt insert/remove can directly result in one or more block allocations on behalf of the tree. Refactor the inode transaction reservations to include one allocfree res. per inode btree modification to cover allocations required by the tree itself. This separates the reservation required to allocate the inode chunk from the reservation required for inobt record insertion/removal. Apply the same logic to the finobt. This results in killing off the finobt modify condition because we no longer assume that the broader transaction reservation will cover finobt block allocations and finobt shape changes can occur in either of the inobt allocation or modify situations. Suggested-by: Dave Chinner <david@fromorbit.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:37 +08:00
xfs_calc_inobt_res(mp) +
xfs_calc_finobt_res(mp);
}
/*
* When only changing the inode we log the inode and possibly the superblock
* We also add a bit of slop for the transaction stuff.
*/
STATIC uint
xfs_calc_ichange_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
/*
* Growing the data section of the filesystem.
* superblock
* agi and agf
* allocation btrees
*/
STATIC uint
xfs_calc_growdata_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1));
}
/*
* Growing the rt section of the filesystem.
* In the first set of transactions (ALLOC) we allocate space to the
* bitmap or summary files.
* superblock: sector size
* agf of the ag from which the extent is allocated: sector size
* bmap btree for bitmap/summary inode: max depth * blocksize
* bitmap/summary inode: inode size
* allocation btrees for 1 block alloc: 2 * (2 * maxdepth - 1) * blocksize
*/
STATIC uint
xfs_calc_growrtalloc_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK),
XFS_FSB_TO_B(mp, 1)) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1));
}
/*
* Growing the rt section of the filesystem.
* In the second set of transactions (ZERO) we zero the new metadata blocks.
* one bitmap/summary block: blocksize
*/
STATIC uint
xfs_calc_growrtzero_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_blocksize);
}
/*
* Growing the rt section of the filesystem.
* In the third set of transactions (FREE) we update metadata without
* allocating any new blocks.
* superblock: sector size
* bitmap inode: inode size
* summary inode: inode size
* one bitmap block: blocksize
* summary blocks: new summary size
*/
STATIC uint
xfs_calc_growrtfree_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 2) +
xfs_calc_buf_res(1, mp->m_sb.sb_blocksize) +
xfs_calc_buf_res(1, mp->m_rsumsize);
}
/*
* Logging the inode modification timestamp on a synchronous write.
* inode
*/
STATIC uint
xfs_calc_swrite_reservation(
struct xfs_mount *mp)
{
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
return xfs_calc_inode_res(mp, 1);
}
/*
* Logging the inode mode bits when writing a setuid/setgid file
* inode
*/
STATIC uint
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_writeid_reservation(
struct xfs_mount *mp)
{
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
return xfs_calc_inode_res(mp, 1);
}
/*
* Converting the inode from non-attributed to attributed.
* the inode being converted: inode size
* agf block and superblock (for block allocation)
* the new block (directory sized)
* bmap blocks for the new directory block
* allocation btrees
*/
STATIC uint
xfs_calc_addafork_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(1, mp->m_dir_geo->blksize) +
xfs_calc_buf_res(XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1,
XFS_FSB_TO_B(mp, 1)) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1),
XFS_FSB_TO_B(mp, 1));
}
/*
* Removing the attribute fork of a file
* the inode being truncated: inode size
* the inode's bmap btree: max depth * block size
* And the bmap_finish transaction can free the blocks and bmap blocks:
* the agf for each of the ags: 4 * sector size
* the agfl for each of the ags: 4 * sector size
* the super block to reflect the freed blocks: sector size
* worst case split in allocation btrees per extent assuming 4 extents:
* 4 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_attrinval_reservation(
struct xfs_mount *mp)
{
return max((xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK),
XFS_FSB_TO_B(mp, 1))),
(xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 4),
XFS_FSB_TO_B(mp, 1))));
}
/*
* Setting an attribute at mount time.
* the inode getting the attribute
* the superblock for allocations
* the agfs extents are allocated from
* the attribute btree * max depth
* the inode allocation btree
* Since attribute transaction space is dependent on the size of the attribute,
* the calculation is done partially at mount time and partially at runtime(see
* below).
*/
STATIC uint
xfs_calc_attrsetm_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
xfs: inode log reservations are too small We've been seeing occasional problems with log space leaks and transaction underruns such as this for some time: XFS (dm-0): xlog_write: reservation summary: trans type = FSYNC_TS (36) unit res = 2740 bytes current res = -4 bytes total reg = 0 bytes (o/flow = 0 bytes) ophdrs = 0 (ophdr space = 0 bytes) ophdr + reg = 0 bytes num regions = 0 Turns out that xfstests generic/311 is reliably reproducing this problem with the test it runs at sequence 16 of it execution. It is a 100% reliable reproducer with the mkfs configuration of "-b size=1024 -m crc=1" on a 10GB scratch device. The problem? Inode forks in btree format are logged in memory format, not disk format (i.e. bmbt format, not bmdr format). That means there is a btree block header being logged, when such a structure is never written to the inode fork in bmdr format. The bmdr header in the inode is only 4 bytes, while the bmbt header is 24 bytes for v4 filesystems and 72 bytes for v5 filesystems. We currently reserve the inode size plus the rounded up overhead of a logging a buffer, which is 128 bytes. That means the reservation for a 512 byte inode is 640 bytes. What we can actually log is: inode core, data and attr fork = 512 bytes inode log format + log op header = 56 + 12 = 68 bytes data fork bmbt hdr = 24/72 bytes attr fork bmbt hdr = 24/72 bytes So, for a v2 inodes we can log at least 628 bytes, but if we split that inode over the end of the log across log buffers, we need to also another log op header, which takes us to 640 bytes. If there's another reservation taken out of this that I haven't taken into account (perhaps multiple iclog splits?) or I haven't corectly calculated the bmbt format space used (entirely possible), then we will overun it. For v3 inodes the maximum is actually 724 bytes, and even a single maximally sized btree format fork can blow it (652 bytes). And that's exactly what is happening with the FSYNC_TS transaction in the above output - it's consumed 644 bytes of space after the CIL context took the space reserved for it (2100 bytes). This problem has always been present in the XFS code - the btree format inode forks have always been logged in this manner. Hence there has always been the possibility of an overrun with such a transaction. The CRC code has just exposed it frequently enough to be able to debug and understand the root cause.... So, let's fix all the inode log space reservations. [ I'm so glad we spent the effort to clean up the transaction reservation code. This is an easy fix now. ] Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-28 14:10:35 +08:00
xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(XFS_DA_NODE_MAXDEPTH, XFS_FSB_TO_B(mp, 1));
}
/*
* Setting an attribute at runtime, transaction space unit per block.
* the superblock for allocations: sector size
* the inode bmap btree could join or split: max depth * block size
* Since the runtime attribute transaction space is dependent on the total
* blocks needed for the 1st bmap, here we calculate out the space unit for
* one block so that the caller could figure out the total space according
* to the attibute extent length in blocks by:
* ext * M_RES(mp)->tr_attrsetrt.tr_logres
*/
STATIC uint
xfs_calc_attrsetrt_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK),
XFS_FSB_TO_B(mp, 1));
}
/*
* Removing an attribute.
* the inode: inode size
* the attribute btree could join: max depth * block size
* the inode bmap btree could join or split: max depth * block size
* And the bmap_finish transaction can free the attr blocks freed giving:
* the agf for the ag in which the blocks live: 2 * sector size
* the agfl for the ag in which the blocks live: 2 * sector size
* the superblock for the free block count: sector size
* the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
*/
STATIC uint
xfs_calc_attrrm_reservation(
struct xfs_mount *mp)
{
return XFS_DQUOT_LOGRES(mp) +
max((xfs_calc_inode_res(mp, 1) +
xfs_calc_buf_res(XFS_DA_NODE_MAXDEPTH,
XFS_FSB_TO_B(mp, 1)) +
(uint)XFS_FSB_TO_B(mp,
XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) +
xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), 0)),
(xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) +
xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2),
XFS_FSB_TO_B(mp, 1))));
}
/*
* Clearing a bad agino number in an agi hash bucket.
*/
STATIC uint
xfs_calc_clear_agi_bucket_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
/*
* Adjusting quota limits.
* the disk quota buffer: sizeof(struct xfs_disk_dquot)
*/
STATIC uint
xfs_calc_qm_setqlim_reservation(void)
{
return xfs_calc_buf_res(1, sizeof(struct xfs_disk_dquot));
}
/*
* Allocating quota on disk if needed.
* the write transaction log space for quota file extent allocation
* the unit of quota allocation: one system block size
*/
STATIC uint
xfs_calc_qm_dqalloc_reservation(
struct xfs_mount *mp)
{
return xfs_calc_write_reservation(mp) +
xfs_calc_buf_res(1,
XFS_FSB_TO_B(mp, XFS_DQUOT_CLUSTER_SIZE_FSB) - 1);
}
/*
* Turning off quotas.
* the quota off logitems: sizeof(struct xfs_qoff_logitem) * 2
* the superblock for the quota flags: sector size
*/
STATIC uint
xfs_calc_qm_quotaoff_reservation(
struct xfs_mount *mp)
{
return sizeof(struct xfs_qoff_logitem) * 2 +
xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
/*
* End of turning off quotas.
* the quota off logitems: sizeof(struct xfs_qoff_logitem) * 2
*/
STATIC uint
xfs_calc_qm_quotaoff_end_reservation(void)
{
return sizeof(struct xfs_qoff_logitem) * 2;
}
/*
* Syncing the incore super block changes to disk.
* the super block to reflect the changes: sector size
*/
STATIC uint
xfs_calc_sb_reservation(
struct xfs_mount *mp)
{
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
void
xfs_trans_resv_calc(
struct xfs_mount *mp,
struct xfs_trans_resv *resp)
{
/*
* The following transactions are logged in physical format and
* require a permanent reservation on space.
*/
resp->tr_write.tr_logres = xfs_calc_write_reservation(mp);
if (xfs_sb_version_hasreflink(&mp->m_sb))
resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT_REFLINK;
else
resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT;
resp->tr_write.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_itruncate.tr_logres = xfs_calc_itruncate_reservation(mp);
if (xfs_sb_version_hasreflink(&mp->m_sb))
resp->tr_itruncate.tr_logcount =
XFS_ITRUNCATE_LOG_COUNT_REFLINK;
else
resp->tr_itruncate.tr_logcount = XFS_ITRUNCATE_LOG_COUNT;
resp->tr_itruncate.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
resp->tr_rename.tr_logcount = XFS_RENAME_LOG_COUNT;
resp->tr_rename.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_link.tr_logres = xfs_calc_link_reservation(mp);
resp->tr_link.tr_logcount = XFS_LINK_LOG_COUNT;
resp->tr_link.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_remove.tr_logres = xfs_calc_remove_reservation(mp);
resp->tr_remove.tr_logcount = XFS_REMOVE_LOG_COUNT;
resp->tr_remove.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_symlink.tr_logres = xfs_calc_symlink_reservation(mp);
resp->tr_symlink.tr_logcount = XFS_SYMLINK_LOG_COUNT;
resp->tr_symlink.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
xfs: eliminate duplicate icreate tx reservation functions The create transaction reservation calculation has two different branches of code depending on whether the filesystem is a v5 format fs or older. Each branch considers the max reservation between the allocation case (new chunk allocation + record insert) and the modify case (chunk exists, record modification) of inode allocation. The modify case is the same for both superblock versions with the exception of the finobt. The finobt helper checks the feature bit, however, and so the modify case already shares the same code. Now that inode chunk allocation has been refactored into a helper that checks the superblock version to calculate the appropriate reservation for the create transaction, the only remaining difference between the create and icreate branches is the call to the finobt helper. As noted above, the finobt helper is a no-op when the feature is not enabled. Therefore, these branches are effectively duplicate and can be condensed. Remove the xfs_calc_create_*() branch of functions and update the various callers to use the xfs_calc_icreate_*() variant. The latter creates the same reservation size for v4 create transactions as the removed branch. As such, this patch does not result in transaction reservation changes. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-01-09 02:41:38 +08:00
resp->tr_create.tr_logres = xfs_calc_icreate_reservation(mp);
resp->tr_create.tr_logcount = XFS_CREATE_LOG_COUNT;
resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_create_tmpfile.tr_logres =
xfs_calc_create_tmpfile_reservation(mp);
resp->tr_create_tmpfile.tr_logcount = XFS_CREATE_TMPFILE_LOG_COUNT;
resp->tr_create_tmpfile.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_ifree.tr_logres = xfs_calc_ifree_reservation(mp);
resp->tr_ifree.tr_logcount = XFS_INACTIVE_LOG_COUNT;
resp->tr_ifree.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_addafork.tr_logres = xfs_calc_addafork_reservation(mp);
resp->tr_addafork.tr_logcount = XFS_ADDAFORK_LOG_COUNT;
resp->tr_addafork.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_attrinval.tr_logres = xfs_calc_attrinval_reservation(mp);
resp->tr_attrinval.tr_logcount = XFS_ATTRINVAL_LOG_COUNT;
resp->tr_attrinval.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_attrsetm.tr_logres = xfs_calc_attrsetm_reservation(mp);
resp->tr_attrsetm.tr_logcount = XFS_ATTRSET_LOG_COUNT;
resp->tr_attrsetm.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_attrrm.tr_logres = xfs_calc_attrrm_reservation(mp);
resp->tr_attrrm.tr_logcount = XFS_ATTRRM_LOG_COUNT;
resp->tr_attrrm.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_growrtalloc.tr_logres = xfs_calc_growrtalloc_reservation(mp);
resp->tr_growrtalloc.tr_logcount = XFS_DEFAULT_PERM_LOG_COUNT;
resp->tr_growrtalloc.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
resp->tr_qm_dqalloc.tr_logres = xfs_calc_qm_dqalloc_reservation(mp);
if (xfs_sb_version_hasreflink(&mp->m_sb))
resp->tr_qm_dqalloc.tr_logcount = XFS_WRITE_LOG_COUNT_REFLINK;
else
resp->tr_qm_dqalloc.tr_logcount = XFS_WRITE_LOG_COUNT;
resp->tr_qm_dqalloc.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
/*
* The following transactions are logged in logical format with
* a default log count.
*/
resp->tr_qm_setqlim.tr_logres = xfs_calc_qm_setqlim_reservation();
resp->tr_qm_setqlim.tr_logcount = XFS_DEFAULT_LOG_COUNT;
resp->tr_qm_quotaoff.tr_logres = xfs_calc_qm_quotaoff_reservation(mp);
resp->tr_qm_quotaoff.tr_logcount = XFS_DEFAULT_LOG_COUNT;
resp->tr_qm_equotaoff.tr_logres =
xfs_calc_qm_quotaoff_end_reservation();
resp->tr_qm_equotaoff.tr_logcount = XFS_DEFAULT_LOG_COUNT;
resp->tr_sb.tr_logres = xfs_calc_sb_reservation(mp);
resp->tr_sb.tr_logcount = XFS_DEFAULT_LOG_COUNT;
/* growdata requires permanent res; it can free space to the last AG */
resp->tr_growdata.tr_logres = xfs_calc_growdata_reservation(mp);
resp->tr_growdata.tr_logcount = XFS_DEFAULT_PERM_LOG_COUNT;
resp->tr_growdata.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
/* The following transaction are logged in logical format */
resp->tr_ichange.tr_logres = xfs_calc_ichange_reservation(mp);
resp->tr_fsyncts.tr_logres = xfs_calc_swrite_reservation(mp);
resp->tr_writeid.tr_logres = xfs_calc_writeid_reservation(mp);
resp->tr_attrsetrt.tr_logres = xfs_calc_attrsetrt_reservation(mp);
resp->tr_clearagi.tr_logres = xfs_calc_clear_agi_bucket_reservation(mp);
resp->tr_growrtzero.tr_logres = xfs_calc_growrtzero_reservation(mp);
resp->tr_growrtfree.tr_logres = xfs_calc_growrtfree_reservation(mp);
}