mirror_zfs/include/sys/abd.h
Brian Atkinson a10e552b99
Adding Direct IO Support
Adding O_DIRECT support to ZFS to bypass the ARC for writes/reads.

O_DIRECT support in ZFS will always ensure there is coherency between
buffered and O_DIRECT IO requests. This ensures that all IO requests,
whether buffered or direct, will see the same file contents at all
times. Just as in other FS's , O_DIRECT does not imply O_SYNC. While
data is written directly to VDEV disks, metadata will not be synced
until the associated  TXG is synced.
For both O_DIRECT read and write request the offset and request sizes,
at a minimum, must be PAGE_SIZE aligned. In the event they are not,
then EINVAL is returned unless the direct property is set to always (see
below).

For O_DIRECT writes:
The request also must be block aligned (recordsize) or the write
request will take the normal (buffered) write path. In the event that
request is block aligned and a cached copy of the buffer in the ARC,
then it will be discarded from the ARC forcing all further reads to
retrieve the data from disk.

For O_DIRECT reads:
The only alignment restrictions are PAGE_SIZE alignment. In the event
that the requested data is in buffered (in the ARC) it will just be
copied from the ARC into the user buffer.

For both O_DIRECT writes and reads the O_DIRECT flag will be ignored in
the event that file contents are mmap'ed. In this case, all requests
that are at least PAGE_SIZE aligned will just fall back to the buffered
paths. If the request however is not PAGE_SIZE aligned, EINVAL will
be returned as always regardless if the file's contents are mmap'ed.

Since O_DIRECT writes go through the normal ZIO pipeline, the
following operations are supported just as with normal buffered writes:
Checksum
Compression
Encryption
Erasure Coding
There is one caveat for the data integrity of O_DIRECT writes that is
distinct for each of the OS's supported by ZFS.
FreeBSD - FreeBSD is able to place user pages under write protection so
          any data in the user buffers and written directly down to the
	  VDEV disks is guaranteed to not change. There is no concern
	  with data integrity and O_DIRECT writes.
Linux - Linux is not able to place anonymous user pages under write
        protection. Because of this, if the user decides to manipulate
	the page contents while the write operation is occurring, data
	integrity can not be guaranteed. However, there is a module
	parameter `zfs_vdev_direct_write_verify` that controls the
	if a O_DIRECT writes that can occur to a top-level VDEV before
	a checksum verify is run before the contents of the I/O buffer
        are committed to disk. In the event of a checksum verification
	failure the write will return EIO. The number of O_DIRECT write
	checksum verification errors can be observed by doing
	`zpool status -d`, which will list all verification errors that
	have occurred on a top-level VDEV. Along with `zpool status`, a
	ZED event will be issues as `dio_verify` when a checksum
	verification error occurs.

ZVOLs and dedup is not currently supported with Direct I/O.

A new dataset property `direct` has been added with the following 3
allowable values:
disabled - Accepts O_DIRECT flag, but silently ignores it and treats
	   the request as a buffered IO request.
standard - Follows the alignment restrictions  outlined above for
	   write/read IO requests when the O_DIRECT flag is used.
always   - Treats every write/read IO request as though it passed
           O_DIRECT and will do O_DIRECT if the alignment restrictions
	   are met otherwise will redirect through the ARC. This
	   property will not allow a request to fail.

There is also a module parameter zfs_dio_enabled that can be used to
force all reads and writes through the ARC. By setting this module
parameter to 0, it mimics as if the  direct dataset property is set to
disabled.

Reviewed-by: Brian Behlendorf <behlendorf@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Atkinson <batkinson@lanl.gov>
Co-authored-by: Mark Maybee <mark.maybee@delphix.com>
Co-authored-by: Matt Macy <mmacy@FreeBSD.org>
Co-authored-by: Brian Behlendorf <behlendorf@llnl.gov>
Closes #10018
2024-09-14 13:47:59 -07:00

223 lines
5.9 KiB
C

/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or https://opensource.org/licenses/CDDL-1.0.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2014 by Chunwei Chen. All rights reserved.
* Copyright (c) 2016, 2019 by Delphix. All rights reserved.
*/
#ifndef _ABD_H
#define _ABD_H
#include <sys/isa_defs.h>
#include <sys/debug.h>
#include <sys/zfs_refcount.h>
#include <sys/uio.h>
#include <sys/abd_os.h>
#ifdef __cplusplus
extern "C" {
#endif
typedef enum abd_flags {
ABD_FLAG_LINEAR = 1 << 0, /* is buffer linear (or scattered)? */
ABD_FLAG_OWNER = 1 << 1, /* does it own its data buffers? */
ABD_FLAG_META = 1 << 2, /* does this represent FS metadata? */
ABD_FLAG_MULTI_ZONE = 1 << 3, /* pages split over memory zones */
ABD_FLAG_MULTI_CHUNK = 1 << 4, /* pages split over multiple chunks */
ABD_FLAG_LINEAR_PAGE = 1 << 5, /* linear but allocd from page */
ABD_FLAG_GANG = 1 << 6, /* mult ABDs chained together */
ABD_FLAG_GANG_FREE = 1 << 7, /* gang ABD is responsible for mem */
ABD_FLAG_ALLOCD = 1 << 8, /* we allocated the abd_t */
ABD_FLAG_FROM_PAGES = 1 << 9, /* does not own pages */
} abd_flags_t;
typedef struct abd {
abd_flags_t abd_flags;
uint_t abd_size; /* excludes scattered abd_offset */
list_node_t abd_gang_link;
#ifdef ZFS_DEBUG
struct abd *abd_parent;
zfs_refcount_t abd_children;
#endif
kmutex_t abd_mtx;
union {
struct abd_scatter abd_scatter;
struct abd_linear abd_linear;
struct abd_gang {
list_t abd_gang_chain;
} abd_gang;
} abd_u;
} abd_t;
typedef int abd_iter_func_t(void *buf, size_t len, void *priv);
typedef int abd_iter_func2_t(void *bufa, void *bufb, size_t len, void *priv);
extern int zfs_abd_scatter_enabled;
/*
* Allocations and deallocations
*/
__attribute__((malloc))
abd_t *abd_alloc(size_t, boolean_t);
__attribute__((malloc))
abd_t *abd_alloc_linear(size_t, boolean_t);
__attribute__((malloc))
abd_t *abd_alloc_gang(void);
__attribute__((malloc))
abd_t *abd_alloc_for_io(size_t, boolean_t);
__attribute__((malloc))
abd_t *abd_alloc_sametype(abd_t *, size_t);
boolean_t abd_size_alloc_linear(size_t);
void abd_gang_add(abd_t *, abd_t *, boolean_t);
void abd_free(abd_t *);
abd_t *abd_get_offset(abd_t *, size_t);
abd_t *abd_get_offset_size(abd_t *, size_t, size_t);
abd_t *abd_get_offset_struct(abd_t *, abd_t *, size_t, size_t);
abd_t *abd_get_zeros(size_t);
abd_t *abd_get_from_buf(void *, size_t);
abd_t *abd_get_from_buf_struct(abd_t *, void *, size_t);
void abd_cache_reap_now(void);
/*
* Conversion to and from a normal buffer
*/
void *abd_to_buf(abd_t *);
void *abd_borrow_buf(abd_t *, size_t);
void *abd_borrow_buf_copy(abd_t *, size_t);
void abd_return_buf(abd_t *, void *, size_t);
void abd_return_buf_copy(abd_t *, void *, size_t);
void abd_take_ownership_of_buf(abd_t *, boolean_t);
void abd_release_ownership_of_buf(abd_t *);
/*
* ABD operations
*/
int abd_iterate_func(abd_t *, size_t, size_t, abd_iter_func_t *, void *);
int abd_iterate_func2(abd_t *, abd_t *, size_t, size_t, size_t,
abd_iter_func2_t *, void *);
void abd_copy_off(abd_t *, abd_t *, size_t, size_t, size_t);
void abd_copy_from_buf_off(abd_t *, const void *, size_t, size_t);
void abd_copy_to_buf_off(void *, abd_t *, size_t, size_t);
int abd_cmp(abd_t *, abd_t *);
int abd_cmp_buf_off(abd_t *, const void *, size_t, size_t);
int abd_cmp_zero_off(abd_t *, size_t, size_t);
void abd_zero_off(abd_t *, size_t, size_t);
void abd_verify(abd_t *);
void abd_raidz_gen_iterate(abd_t **cabds, abd_t *dabd, size_t off,
size_t csize, size_t dsize, const unsigned parity,
void (*func_raidz_gen)(void **, const void *, size_t, size_t));
void abd_raidz_rec_iterate(abd_t **cabds, abd_t **tabds,
size_t tsize, const unsigned parity,
void (*func_raidz_rec)(void **t, const size_t tsize, void **c,
const unsigned *mul),
const unsigned *mul);
/*
* Wrappers for calls with offsets of 0
*/
static inline void
abd_copy(abd_t *dabd, abd_t *sabd, size_t size)
{
abd_copy_off(dabd, sabd, 0, 0, size);
}
static inline void
abd_copy_from_buf(abd_t *abd, const void *buf, size_t size)
{
abd_copy_from_buf_off(abd, buf, 0, size);
}
static inline void
abd_copy_to_buf(void* buf, abd_t *abd, size_t size)
{
abd_copy_to_buf_off(buf, abd, 0, size);
}
static inline int
abd_cmp_buf(abd_t *abd, const void *buf, size_t size)
{
return (abd_cmp_buf_off(abd, buf, 0, size));
}
static inline void
abd_zero(abd_t *abd, size_t size)
{
abd_zero_off(abd, 0, size);
}
static inline int
abd_cmp_zero(abd_t *abd, size_t size)
{
return (abd_cmp_zero_off(abd, 0, size));
}
/*
* ABD type check functions
*/
static inline boolean_t
abd_is_linear(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_LINEAR) ? B_TRUE : B_FALSE);
}
static inline boolean_t
abd_is_linear_page(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_LINEAR_PAGE) ? B_TRUE : B_FALSE);
}
static inline boolean_t
abd_is_gang(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_GANG) ? B_TRUE : B_FALSE);
}
static inline uint_t
abd_get_size(abd_t *abd)
{
return (abd->abd_size);
}
static inline boolean_t
abd_is_from_pages(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_FROM_PAGES) ? B_TRUE : B_FALSE);
}
/*
* Module lifecycle
* Defined in each specific OS's abd_os.c
*/
void abd_init(void);
void abd_fini(void);
#ifdef __cplusplus
}
#endif
#endif /* _ABD_H */