Archive-Team
/
zfs
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 103 Wiki Activity
zfs/include/sys/abd.h

236 lines
6.3 KiB
C
Raw Permalink Normal View History

DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
/*
* 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
Replace dead opensolaris.org license link The commit replaces all findings of the link: http://www.opensolaris.org/os/licensing with this one: https://opensource.org/licenses/CDDL-1.0 Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Tino Reichardt <milky-zfs@mcmilk.de> Closes #13619
2022-07-11 21:16:13 +00:00
* or https://opensource.org/licenses/CDDL-1.0.
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
* 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.
single-chunk scatter ABDs can be treated as linear Scatter ABD's are allocated from a number of pages. In contrast to linear ABD's, these pages are disjoint in the kernel's virtual address space, so they can't be accessed as a contiguous buffer. Therefore routines that need a linear buffer (e.g. abd_borrow_buf() and friends) must allocate a separate linear buffer (with zio_buf_alloc()), and copy the contents of the pages to/from the linear buffer. This can have a measurable performance overhead on some workloads. https://github.com/zfsonlinux/zfs/commit/87c25d567fb7969b44c7d8af63990e ("abd_alloc should use scatter for >1K allocations") increased the use of scatter ABD's, specifically switching 1.5K through 4K (inclusive) buffers from linear to scatter. For workloads that access blocks whose compressed sizes are in this range, that commit introduced an additional copy into the read code path. For example, the sequential_reads_arc_cached tests in the test suite were reduced by around 5% (this is doing reads of 8K-logical blocks, compressed to 3K, which are cached in the ARC). This commit treats single-chunk scattered buffers as linear buffers, because they are contiguous in the kernel's virtual address space. All single-page (4K) ABD's can be represented this way. Some multi-page ABD's can also be represented this way, if we were able to allocate a single "chunk" (higher-order "page" which represents a power-of-2 series of physically-contiguous pages). This is often the case for 2-page (8K) ABD's. Representing a single-entry scatter ABD as a linear ABD has the performance advantage of avoiding the copy (and allocation) in abd_borrow_buf_copy / abd_return_buf_copy. A performance increase of around 5% has been observed for ARC-cached reads (of small blocks which can take advantage of this), fixing the regression introduced by 87c25d567. Note that this optimization is only possible because all physical memory is always mapped into the kernel's address space. This is not the case for HIGHMEM pages, so the optimization can not be made on 32-bit systems. Reviewed-by: Chunwei Chen <tuxoko@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #8580
2019-06-11 16:02:31 +00:00
* Copyright (c) 2016, 2019 by Delphix. All rights reserved.
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
*/
#ifndef _ABD_H
#define _ABD_H
#include <sys/isa_defs.h>
#include <sys/debug.h>
Rename refcount.h to zfs_refcount.h Renamed to avoid conflicting with refcount.h when a different implementation is already provided by the platform. Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matt Macy <mmacy@FreeBSD.org> Closes #10620
2020-07-29 23:35:33 +00:00
#include <sys/zfs_refcount.h>
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
#include <sys/uio.h>
#ifdef __cplusplus
extern "C" {
#endif
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
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_ZEROS = 1 << 8, /* ABD for zero-filled buffer */
ABD_FLAG_ALLOCD = 1 << 9, /* we allocated the abd_t */
} 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;
The abd child/parent relationship does not need to be tracked ABD's currently track their parent/child relationship. This applies to `abd_get_offset()` and `abd_borrow_buf()`. However, nothing depends on knowing this relationship, it's only used for consistency checks to verify that we are not destroying an ABD that's still in use. When we are creating/destroying ABD's frequently, the performance impact of maintaining these data structures (in particular the atomic increment/decrement operations) can be measurable. This commit removes this verification code on production builds, but keeps it when ZFS_DEBUG is set. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #11535
2021-01-30 18:04:42 +00:00
#ifdef ZFS_DEBUG
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
struct abd *abd_parent;
zfs_refcount_t abd_children;
The abd child/parent relationship does not need to be tracked ABD's currently track their parent/child relationship. This applies to `abd_get_offset()` and `abd_borrow_buf()`. However, nothing depends on knowing this relationship, it's only used for consistency checks to verify that we are not destroying an ABD that's still in use. When we are creating/destroying ABD's frequently, the performance impact of maintaining these data structures (in particular the atomic increment/decrement operations) can be measurable. This commit removes this verification code on production builds, but keeps it when ZFS_DEBUG is set. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #11535
2021-01-30 18:04:42 +00:00
#endif
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
kmutex_t abd_mtx;
union {
struct abd_scatter {
uint_t abd_offset;
#if defined(__FreeBSD__) && defined(_KERNEL)
void *abd_chunks[1]; /* actually variable-length */
#else
uint_t abd_nents;
struct scatterlist *abd_sgl;
#endif
} abd_scatter;
struct abd_linear {
void *abd_buf;
struct scatterlist *abd_sgl; /* for LINEAR_PAGE */
} abd_linear;
struct abd_gang {
list_t abd_gang_chain;
} abd_gang;
} abd_u;
} abd_t;
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
Improve compatibility with C++ consumers C++ is a little picky about not using keywords for names, or string constness. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Jorgen Lundman <lundman@lundman.net> Signed-off-by: Ryan Moeller <ryan@iXsystems.com> Closes #10409
2020-06-06 19:54:04 +00:00
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);
abd: add page iterator The regular ABD iterators yield data buffers, so they have to map and unmap pages into kernel memory. If the caller only wants to count chunks, or can use page pointers directly, then the map/unmap is just unnecessary overhead. This adds adb_iterate_page_func, which yields unmapped struct page instead. Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <rob.norris@klarasystems.com> Sponsored-by: Klara, Inc. Sponsored-by: Wasabi Technology, Inc. Closes #15533 Closes #15588 (cherry picked from commit 390b448726c580999dd337be7a40b0e95cf1d50b)
2023-12-11 05:05:54 +00:00
#if defined(__linux__) && defined(_KERNEL)
typedef int abd_iter_page_func_t(struct page *, size_t, size_t, void *);
#endif
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
extern int zfs_abd_scatter_enabled;
/*
* Allocations and deallocations
*/
Use __attribute__((malloc)) on memory allocation functions This informs the C compiler that pointers returned from these functions do not alias other functions, which allows it to do better code optimization and should make the compiled code smaller. References: https://stackoverflow.com/a/53654773 https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute https://clang.llvm.org/docs/AttributeReference.html#malloc Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #14827
2023-05-26 22:47:52 +00:00
__attribute__((malloc))
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
abd_t *abd_alloc(size_t, boolean_t);
Use __attribute__((malloc)) on memory allocation functions This informs the C compiler that pointers returned from these functions do not alias other functions, which allows it to do better code optimization and should make the compiled code smaller. References: https://stackoverflow.com/a/53654773 https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute https://clang.llvm.org/docs/AttributeReference.html#malloc Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #14827
2023-05-26 22:47:52 +00:00
__attribute__((malloc))
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
abd_t *abd_alloc_linear(size_t, boolean_t);
Use __attribute__((malloc)) on memory allocation functions This informs the C compiler that pointers returned from these functions do not alias other functions, which allows it to do better code optimization and should make the compiled code smaller. References: https://stackoverflow.com/a/53654773 https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute https://clang.llvm.org/docs/AttributeReference.html#malloc Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #14827
2023-05-26 22:47:52 +00:00
__attribute__((malloc))
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
abd_t *abd_alloc_gang(void);
Use __attribute__((malloc)) on memory allocation functions This informs the C compiler that pointers returned from these functions do not alias other functions, which allows it to do better code optimization and should make the compiled code smaller. References: https://stackoverflow.com/a/53654773 https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute https://clang.llvm.org/docs/AttributeReference.html#malloc Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #14827
2023-05-26 22:47:52 +00:00
__attribute__((malloc))
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
abd_t *abd_alloc_for_io(size_t, boolean_t);
Use __attribute__((malloc)) on memory allocation functions This informs the C compiler that pointers returned from these functions do not alias other functions, which allows it to do better code optimization and should make the compiled code smaller. References: https://stackoverflow.com/a/53654773 https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute https://clang.llvm.org/docs/AttributeReference.html#malloc Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu> Closes #14827
2023-05-26 22:47:52 +00:00
__attribute__((malloc))
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
abd_t *abd_alloc_sametype(abd_t *, size_t);
Avoid small buffer copying on write It is wrong for arc_write_ready() to use zfs_abd_scatter_enabled to decide whether to reallocate/copy the buffer, because the answer is OS-specific and depends on the buffer size. Instead of that use abd_size_alloc_linear(), moved into public header. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Signed-off-by: Alexander Motin <mav@FreeBSD.org> Closes #12425
2021-07-27 23:05:47 +00:00
boolean_t abd_size_alloc_linear(size_t);
Gang ABD Type Adding the gang ABD type, which allows for linear and scatter ABDs to be chained together into a single ABD. This can be used to avoid doing memory copies to/from ABDs. An example of this can be found in vdev_queue.c in the vdev_queue_aggregate() function. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Co-authored-by: Brian <bwa@clemson.edu> Co-authored-by: Mark Maybee <mmaybee@cray.com> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10069
2020-05-21 01:06:09 +00:00
void abd_gang_add(abd_t *, abd_t *, boolean_t);
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
void abd_free(abd_t *);
abd_t *abd_get_offset(abd_t *, size_t);
ABD changes for vectorized RAIDZ * userspace: aligned buffers. Minimum of 32B alignment is needed for AVX2. Kernel buffers are aligned 512B or more. * add abd_get_offset_size() interface * abd_iter_map(): fix calculation of iter_mapsize * add abd_raidz_gen_iterate() and abd_raidz_rec_iterate() Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 13:42:51 +00:00
abd_t *abd_get_offset_size(abd_t *, size_t, size_t);
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
abd_t *abd_get_offset_struct(abd_t *, abd_t *, size_t, size_t);
Gang ABD Type Adding the gang ABD type, which allows for linear and scatter ABDs to be chained together into a single ABD. This can be used to avoid doing memory copies to/from ABDs. An example of this can be found in vdev_queue.c in the vdev_queue_aggregate() function. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Co-authored-by: Brian <bwa@clemson.edu> Co-authored-by: Mark Maybee <mmaybee@cray.com> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10069
2020-05-21 01:06:09 +00:00
abd_t *abd_get_zeros(size_t);
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
abd_t *abd_get_from_buf(void *, size_t);
Add abd_cache_reap_now for abd_chunk_cache users Apparently missed in the initial port integration was the need to reap the abd_chunk_cache on FreeBSD. This change addresses that oversight. Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Alexander Motin <mav@FreeBSD.org> Signed-off-by: Matt Macy <mmacy@FreeBSD.org> Closes #10474
2020-06-18 04:44:13 +00:00
void abd_cache_reap_now(void);
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
/*
* 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 *);
abd: add page iterator The regular ABD iterators yield data buffers, so they have to map and unmap pages into kernel memory. If the caller only wants to count chunks, or can use page pointers directly, then the map/unmap is just unnecessary overhead. This adds adb_iterate_page_func, which yields unmapped struct page instead. Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <rob.norris@klarasystems.com> Sponsored-by: Klara, Inc. Sponsored-by: Wasabi Technology, Inc. Closes #15533 Closes #15588 (cherry picked from commit 390b448726c580999dd337be7a40b0e95cf1d50b)
2023-12-11 05:05:54 +00:00
#if defined(__linux__) && defined(_KERNEL)
int abd_iterate_page_func(abd_t *, size_t, size_t, abd_iter_page_func_t *,
void *);
#endif
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
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);
void abd_zero_off(abd_t *, size_t, size_t);
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
void abd_verify(abd_t *);
ABD page support to vdev_disk.c Signed-off-by: Isaac Huang <he.huang@intel.com>
2016-08-31 06:26:43 +00:00
ABD changes for vectorized RAIDZ * userspace: aligned buffers. Minimum of 32B alignment is needed for AVX2. Kernel buffers are aligned 512B or more. * add abd_get_offset_size() interface * abd_iter_map(): fix calculation of iter_mapsize * add abd_raidz_gen_iterate() and abd_raidz_rec_iterate() Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 13:42:51 +00:00
void abd_raidz_gen_iterate(abd_t **cabds, abd_t *dabd,
ssize_t csize, ssize_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,
ssize_t tsize, const unsigned parity,
void (*func_raidz_rec)(void **t, const size_t tsize, void **c,
const unsigned *mul),
const unsigned *mul);
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
/*
* 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
OpenZFS 8426 - mark immutable buffer arguments as such in abd.h Authored by: Andriy Gapon <avg@FreeBSD.org> Reviewed by: Matt Ahrens <mahrens@delphix.com> Approved by: Robert Mustacchi <rm@joyent.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: George Melikov <mail@gmelikov.ru> Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov> OpenZFS-issue: https://www.illumos.org/issues/8426 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/37359a6 Closes #6287
2017-06-26 10:46:45 +00:00
abd_copy_from_buf(abd_t *abd, const void *buf, size_t size)
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
{
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
OpenZFS 8426 - mark immutable buffer arguments as such in abd.h Authored by: Andriy Gapon <avg@FreeBSD.org> Reviewed by: Matt Ahrens <mahrens@delphix.com> Approved by: Robert Mustacchi <rm@joyent.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: George Melikov <mail@gmelikov.ru> Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov> OpenZFS-issue: https://www.illumos.org/issues/8426 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/37359a6 Closes #6287
2017-06-26 10:46:45 +00:00
abd_cmp_buf(abd_t *abd, const void *buf, size_t size)
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
{
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);
}
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
/*
* ABD type check functions
*/
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
static inline boolean_t
abd_is_linear(abd_t *abd)
{
Make inline ABD predicates compatible with C++ FreeBSD's zfsd fails to build after e2af2acce3 due to strict type checking errors from the implicit conversion between bool and boolean_t in the inline predicate definitions in abd.h. Use conditionals to return the correct value type from these functions. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Igor Kozhukhov <igor@dilos.org> Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org> Closes #11592
2021-02-15 18:15:50 +00:00
return ((abd->abd_flags & ABD_FLAG_LINEAR) ? B_TRUE : B_FALSE);
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
}
static inline boolean_t
abd_is_linear_page(abd_t *abd)
{
Make inline ABD predicates compatible with C++ FreeBSD's zfsd fails to build after e2af2acce3 due to strict type checking errors from the implicit conversion between bool and boolean_t in the inline predicate definitions in abd.h. Use conditionals to return the correct value type from these functions. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Igor Kozhukhov <igor@dilos.org> Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org> Closes #11592
2021-02-15 18:15:50 +00:00
return ((abd->abd_flags & ABD_FLAG_LINEAR_PAGE) ? B_TRUE : B_FALSE);
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
}
static inline boolean_t
abd_is_gang(abd_t *abd)
{
Make inline ABD predicates compatible with C++ FreeBSD's zfsd fails to build after e2af2acce3 due to strict type checking errors from the implicit conversion between bool and boolean_t in the inline predicate definitions in abd.h. Use conditionals to return the correct value type from these functions. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Igor Kozhukhov <igor@dilos.org> Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org> Closes #11592
2021-02-15 18:15:50 +00:00
return ((abd->abd_flags & ABD_FLAG_GANG) ? B_TRUE : B_FALSE);
allow callers to allocate and provide the abd_t struct The `abd_get_offset_*()` routines create an abd_t that references another abd_t, and doesn't allocate any pages/buffers of its own. In some workloads, these routines may be called frequently, to create many abd_t's representing small pieces of a single large abd_t. In particular, the upcoming RAIDZ Expansion project makes heavy use of these routines. This commit adds the ability for the caller to allocate and provide the abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the cost of allocating the abd_t and performing the accounting associated with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for the `rc_abd`, which references the zio's abd. The upcoming RAIDZ Expansion project will leverage this infrastructure to increase performance of reads post-expansion by around 50%. Additionally, some of the interfaces around creating and destroying abd_t's are cleaned up. Most significantly, the distinction between `abd_put()` and `abd_free()` is eliminated; all types of abd_t's are now disposed of with `abd_free()`. Reviewed-by: Brian Atkinson <batkinson@lanl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Issue #8853 Closes #11439
2021-01-20 19:24:37 +00:00
}
static inline uint_t
abd_get_size(abd_t *abd)
{
return (abd->abd_size);
}
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
/*
* Module lifecycle
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
* Defined in each specific OS's abd_os.c
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
*/
void abd_init(void);
void abd_fini(void);
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
/*
* Linux ABD bio functions
vdev_disk: rename existing functions to vdev_classic_* This is just renaming the existing functions we're about to replace and grouping them together to make the next commits easier to follow. Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Rob Norris <rob.norris@klarasystems.com> Sponsored-by: Klara, Inc. Sponsored-by: Wasabi Technology, Inc. Closes #15533 Closes #15588 (cherry picked from commit f3b85d706bae82957d2e3e0ef1d53a1cfab60eb4)
2024-01-09 01:12:56 +00:00
* Note: these are only needed to support vdev_classic. See comment in
* vdev_disk.c.
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
*/
#if defined(__linux__) && defined(_KERNEL)
Gang ABD Type Adding the gang ABD type, which allows for linear and scatter ABDs to be chained together into a single ABD. This can be used to avoid doing memory copies to/from ABDs. An example of this can be found in vdev_queue.c in the vdev_queue_aggregate() function. Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Co-authored-by: Brian <bwa@clemson.edu> Co-authored-by: Mark Maybee <mmaybee@cray.com> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10069
2020-05-21 01:06:09 +00:00
unsigned int abd_bio_map_off(struct bio *, abd_t *, unsigned int, size_t);
Combine OS-independent ABD Code into Common Source File Reorganizing ABD code base so OS-independent ABD code has been placed into a common abd.c file. OS-dependent ABD code has been left in each OS's ABD source files, and these source files have been renamed to abd_os. The OS-independent ABD code is now under: module/zfs/abd.c With the OS-dependent code in: module/os/linux/zfs/abd_os.c module/os/freebsd/zfs/abd_os.c Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Atkinson <batkinson@lanl.gov> Closes #10293
2020-05-10 19:23:52 +00:00
unsigned long abd_nr_pages_off(abd_t *, unsigned int, size_t);
#endif
DLPX-44812 integrate EP-220 large memory scalability
2016-07-22 15:52:49 +00:00
#ifdef __cplusplus
}
#endif
#endif /* _ABD_H */