linux/fs/xfs/libxfs/xfs_defer.h

89 lines
3.2 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
/*
* Copyright (C) 2016 Oracle. All Rights Reserved.
* Author: Darrick J. Wong <darrick.wong@oracle.com>
*/
#ifndef __XFS_DEFER_H__
#define __XFS_DEFER_H__
struct xfs_defer_op_type;
/*
* Save a log intent item and a list of extents, so that we can replay
* whatever action had to happen to the extent list and file the log done
* item.
*/
struct xfs_defer_pending {
const struct xfs_defer_op_type *dfp_type; /* function pointers */
struct list_head dfp_list; /* pending items */
void *dfp_intent; /* log intent item */
void *dfp_done; /* log done item */
struct list_head dfp_work; /* work items */
unsigned int dfp_count; /* # extent items */
};
/*
* Header for deferred operation list.
*
* dop_low is used by the allocator to activate the lowspace algorithm -
* when free space is running low the extent allocator may choose to
* allocate an extent from an AG without leaving sufficient space for
* a btree split when inserting the new extent. In this case the allocator
* will enable the lowspace algorithm which is supposed to allow further
* allocations (such as btree splits and newroots) to allocate from
* sequential AGs. In order to avoid locking AGs out of order the lowspace
* algorithm will start searching for free space from AG 0. If the correct
* transaction reservations have been made then this algorithm will eventually
* find all the space it needs.
*/
enum xfs_defer_ops_type {
XFS_DEFER_OPS_TYPE_BMAP,
XFS_DEFER_OPS_TYPE_REFCOUNT,
XFS_DEFER_OPS_TYPE_RMAP,
XFS_DEFER_OPS_TYPE_FREE,
xfs: defer agfl block frees when dfops is available The AGFL fixup code executes before every block allocation/free and rectifies the AGFL based on the current, dynamic allocation requirements of the fs. The AGFL must hold a minimum number of blocks to satisfy a worst case split of the free space btrees caused by the impending allocation operation. The AGFL is also updated to maintain the implicit requirement for a minimum number of free slots to satisfy a worst case join of the free space btrees. Since the AGFL caches individual blocks, AGFL reduction typically involves multiple, single block frees. We've had reports of transaction overrun problems during certain workloads that boil down to AGFL reduction freeing multiple blocks and consuming more space in the log than was reserved for the transaction. Since the objective of freeing AGFL blocks is to ensure free AGFL free slots are available for the upcoming allocation, one way to address this problem is to release surplus blocks from the AGFL immediately but defer the free of those blocks (similar to how file-mapped blocks are unmapped from the file in one transaction and freed via a deferred operation) until the transaction is rolled. This turns AGFL reduction into an operation with predictable log reservation consumption. Add the capability to defer AGFL block frees when a deferred ops list is available to the AGFL fixup code. Add a dfops pointer to the transaction to carry dfops through various contexts to the allocator context. Deferring AGFL frees is conditional behavior based on whether the transaction pointer is populated. The long term objective is to reuse the transaction pointer to clean up all unrelated callchains that pass dfops on the stack along with a transaction and in doing so, consistently defer AGFL blocks from the allocator. A bit of customization is required to handle deferred completion processing because AGFL blocks are accounted against a per-ag reservation pool and AGFL blocks are not inserted into the extent busy list when freed (they are inserted when used and released back to the AGFL). Reuse the majority of the existing deferred extent free infrastructure and customize it appropriately to handle AGFL blocks. Note that this patch only adds infrastructure. It does not change behavior because no callers have been updated to pass ->t_agfl_dfops into the allocation code. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-05-08 08:38:47 +08:00
XFS_DEFER_OPS_TYPE_AGFL_FREE,
XFS_DEFER_OPS_TYPE_MAX,
};
#define XFS_DEFER_OPS_NR_INODES 2 /* join up to two inodes */
#define XFS_DEFER_OPS_NR_BUFS 2 /* join up to two buffers */
struct xfs_defer_ops {
struct list_head dop_intake; /* unlogged pending work */
struct list_head dop_pending; /* logged pending work */
/* relog these with each roll */
struct xfs_inode *dop_inodes[XFS_DEFER_OPS_NR_INODES];
struct xfs_buf *dop_bufs[XFS_DEFER_OPS_NR_BUFS];
bool dop_low; /* alloc in low mode */
};
void xfs_defer_add(struct xfs_defer_ops *dop, enum xfs_defer_ops_type type,
struct list_head *h);
int xfs_defer_finish(struct xfs_trans **tp, struct xfs_defer_ops *dop);
void xfs_defer_cancel(struct xfs_defer_ops *dop);
void xfs_defer_init(struct xfs_trans *tp, struct xfs_defer_ops *dop);
bool xfs_defer_has_unfinished_work(struct xfs_defer_ops *dop);
int xfs_defer_ijoin(struct xfs_defer_ops *dop, struct xfs_inode *ip);
int xfs_defer_bjoin(struct xfs_defer_ops *dop, struct xfs_buf *bp);
/* Description of a deferred type. */
struct xfs_defer_op_type {
enum xfs_defer_ops_type type;
unsigned int max_items;
void (*abort_intent)(void *);
void *(*create_done)(struct xfs_trans *, void *, unsigned int);
int (*finish_item)(struct xfs_trans *, struct xfs_defer_ops *,
struct list_head *, void *, void **);
void (*finish_cleanup)(struct xfs_trans *, void *, int);
void (*cancel_item)(struct list_head *);
int (*diff_items)(void *, struct list_head *, struct list_head *);
void *(*create_intent)(struct xfs_trans *, uint);
void (*log_item)(struct xfs_trans *, void *, struct list_head *);
};
void xfs_defer_init_op_type(const struct xfs_defer_op_type *type);
#endif /* __XFS_DEFER_H__ */