Commit 54a611b6 authored by Liam R. Howlett's avatar Liam R. Howlett Committed by Andrew Morton

Maple Tree: add new data structure

Patch series "Introducing the Maple Tree"

The maple tree is an RCU-safe range based B-tree designed to use modern
processor cache efficiently.  There are a number of places in the kernel
that a non-overlapping range-based tree would be beneficial, especially
one with a simple interface.  If you use an rbtree with other data
structures to improve performance or an interval tree to track
non-overlapping ranges, then this is for you.

The tree has a branching factor of 10 for non-leaf nodes and 16 for leaf
nodes.  With the increased branching factor, it is significantly shorter
than the rbtree so it has fewer cache misses.  The removal of the linked
list between subsequent entries also reduces the cache misses and the need
to pull in the previous and next VMA during many tree alterations.

The first user that is covered in this patch set is the vm_area_struct,
where three data structures are replaced by the maple tree: the augmented
rbtree, the vma cache, and the linked list of VMAs in the mm_struct.  The
long term goal is to reduce or remove the mmap_lock contention.

The plan is to get to the point where we use the maple tree in RCU mode.
Readers will not block for writers.  A single write operation will be
allowed at a time.  A reader re-walks if stale data is encountered.  VMAs
would be RCU enabled and this mode would be entered once multiple tasks
are using the mm_struct.

Davidlor said

: Yes I like the maple tree, and at this stage I don't think we can ask for
: more from this series wrt the MM - albeit there seems to still be some
: folks reporting breakage.  Fundamentally I see Liam's work to (re)move
: complexity out of the MM (not to say that the actual maple tree is not
: complex) by consolidating the three complimentary data structures very
: much worth it considering performance does not take a hit.  This was very
: much a turn off with the range locking approach, which worst case scenario
: incurred in prohibitive overhead.  Also as Liam and Matthew have
: mentioned, RCU opens up a lot of nice performance opportunities, and in
: addition academia[1] has shown outstanding scalability of address spaces
: with the foundation of replacing the locked rbtree with RCU aware trees.

A similar work has been discovered in the academic press

	https://pdos.csail.mit.edu/papers/rcuvm:asplos12.pdf

Sheer coincidence.  We designed our tree with the intention of solving the
hardest problem first.  Upon settling on a b-tree variant and a rough
outline, we researched ranged based b-trees and RCU b-trees and did find
that article.  So it was nice to find reassurances that we were on the
right path, but our design choice of using ranges made that paper unusable
for us.

This patch (of 70):

The maple tree is an RCU-safe range based B-tree designed to use modern
processor cache efficiently.  There are a number of places in the kernel
that a non-overlapping range-based tree would be beneficial, especially
one with a simple interface.  If you use an rbtree with other data
structures to improve performance or an interval tree to track
non-overlapping ranges, then this is for you.

The tree has a branching factor of 10 for non-leaf nodes and 16 for leaf
nodes.  With the increased branching factor, it is significantly shorter
than the rbtree so it has fewer cache misses.  The removal of the linked
list between subsequent entries also reduces the cache misses and the need
to pull in the previous and next VMA during many tree alterations.

The first user that is covered in this patch set is the vm_area_struct,
where three data structures are replaced by the maple tree: the augmented
rbtree, the vma cache, and the linked list of VMAs in the mm_struct.  The
long term goal is to reduce or remove the mmap_lock contention.

The plan is to get to the point where we use the maple tree in RCU mode.
Readers will not block for writers.  A single write operation will be
allowed at a time.  A reader re-walks if stale data is encountered.  VMAs
would be RCU enabled and this mode would be entered once multiple tasks
are using the mm_struct.

There is additional BUG_ON() calls added within the tree, most of which
are in debug code.  These will be replaced with a WARN_ON() call in the
future.  There is also additional BUG_ON() calls within the code which
will also be reduced in number at a later date.  These exist to catch
things such as out-of-range accesses which would crash anyways.

Link: https://lkml.kernel.org/r/20220906194824.2110408-1-Liam.Howlett@oracle.com
Link: https://lkml.kernel.org/r/20220906194824.2110408-2-Liam.Howlett@oracle.comSigned-off-by: default avatarLiam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: default avatarMatthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: default avatarDavid Howells <dhowells@redhat.com>
Tested-by: default avatarSven Schnelle <svens@linux.ibm.com>
Tested-by: default avatarYu Zhao <yuzhao@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: David Hildenbrand <david@redhat.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: SeongJae Park <sj@kernel.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
parent 9832fb87
...@@ -36,6 +36,7 @@ Library functionality that is used throughout the kernel. ...@@ -36,6 +36,7 @@ Library functionality that is used throughout the kernel.
kref kref
assoc_array assoc_array
xarray xarray
maple_tree
idr idr
circular-buffers circular-buffers
rbtree rbtree
......
.. SPDX-License-Identifier: GPL-2.0+
==========
Maple Tree
==========
:Author: Liam R. Howlett
Overview
========
The Maple Tree is a B-Tree data type which is optimized for storing
non-overlapping ranges, including ranges of size 1. The tree was designed to
be simple to use and does not require a user written search method. It
supports iterating over a range of entries and going to the previous or next
entry in a cache-efficient manner. The tree can also be put into an RCU-safe
mode of operation which allows reading and writing concurrently. Writers must
synchronize on a lock, which can be the default spinlock, or the user can set
the lock to an external lock of a different type.
The Maple Tree maintains a small memory footprint and was designed to use
modern processor cache efficiently. The majority of the users will be able to
use the normal API. An :ref:`maple-tree-advanced-api` exists for more complex
scenarios. The most important usage of the Maple Tree is the tracking of the
virtual memory areas.
The Maple Tree can store values between ``0`` and ``ULONG_MAX``. The Maple
Tree reserves values with the bottom two bits set to '10' which are below 4096
(ie 2, 6, 10 .. 4094) for internal use. If the entries may use reserved
entries then the users can convert the entries using xa_mk_value() and convert
them back by calling xa_to_value(). If the user needs to use a reserved
value, then the user can convert the value when using the
:ref:`maple-tree-advanced-api`, but are blocked by the normal API.
The Maple Tree can also be configured to support searching for a gap of a given
size (or larger).
Pre-allocating of nodes is also supported using the
:ref:`maple-tree-advanced-api`. This is useful for users who must guarantee a
successful store operation within a given
code segment when allocating cannot be done. Allocations of nodes are
relatively small at around 256 bytes.
.. _maple-tree-normal-api:
Normal API
==========
Start by initialising a maple tree, either with DEFINE_MTREE() for statically
allocated maple trees or mt_init() for dynamically allocated ones. A
freshly-initialised maple tree contains a ``NULL`` pointer for the range ``0``
- ``ULONG_MAX``. There are currently two types of maple trees supported: the
allocation tree and the regular tree. The regular tree has a higher branching
factor for internal nodes. The allocation tree has a lower branching factor
but allows the user to search for a gap of a given size or larger from either
``0`` upwards or ``ULONG_MAX`` down. An allocation tree can be used by
passing in the ``MT_FLAGS_ALLOC_RANGE`` flag when initialising the tree.
You can then set entries using mtree_store() or mtree_store_range().
mtree_store() will overwrite any entry with the new entry and return 0 on
success or an error code otherwise. mtree_store_range() works in the same way
but takes a range. mtree_load() is used to retrieve the entry stored at a
given index. You can use mtree_erase() to erase an entire range by only
knowing one value within that range, or mtree_store() call with an entry of
NULL may be used to partially erase a range or many ranges at once.
If you want to only store a new entry to a range (or index) if that range is
currently ``NULL``, you can use mtree_insert_range() or mtree_insert() which
return -EEXIST if the range is not empty.
You can search for an entry from an index upwards by using mt_find().
You can walk each entry within a range by calling mt_for_each(). You must
provide a temporary variable to store a cursor. If you want to walk each
element of the tree then ``0`` and ``ULONG_MAX`` may be used as the range. If
the caller is going to hold the lock for the duration of the walk then it is
worth looking at the mas_for_each() API in the :ref:`maple-tree-advanced-api`
section.
Sometimes it is necessary to ensure the next call to store to a maple tree does
not allocate memory, please see :ref:`maple-tree-advanced-api` for this use case.
Finally, you can remove all entries from a maple tree by calling
mtree_destroy(). If the maple tree entries are pointers, you may wish to free
the entries first.
Allocating Nodes
----------------
The allocations are handled by the internal tree code. See
:ref:`maple-tree-advanced-alloc` for other options.
Locking
-------
You do not have to worry about locking. See :ref:`maple-tree-advanced-locks`
for other options.
The Maple Tree uses RCU and an internal spinlock to synchronise access:
Takes RCU read lock:
* mtree_load()
* mt_find()
* mt_for_each()
* mt_next()
* mt_prev()
Takes ma_lock internally:
* mtree_store()
* mtree_store_range()
* mtree_insert()
* mtree_insert_range()
* mtree_erase()
* mtree_destroy()
* mt_set_in_rcu()
* mt_clear_in_rcu()
If you want to take advantage of the internal lock to protect the data
structures that you are storing in the Maple Tree, you can call mtree_lock()
before calling mtree_load(), then take a reference count on the object you
have found before calling mtree_unlock(). This will prevent stores from
removing the object from the tree between looking up the object and
incrementing the refcount. You can also use RCU to avoid dereferencing
freed memory, but an explanation of that is beyond the scope of this
document.
.. _maple-tree-advanced-api:
Advanced API
============
The advanced API offers more flexibility and better performance at the
cost of an interface which can be harder to use and has fewer safeguards.
You must take care of your own locking while using the advanced API.
You can use the ma_lock, RCU or an external lock for protection.
You can mix advanced and normal operations on the same array, as long
as the locking is compatible. The :ref:`maple-tree-normal-api` is implemented
in terms of the advanced API.
The advanced API is based around the ma_state, this is where the 'mas'
prefix originates. The ma_state struct keeps track of tree operations to make
life easier for both internal and external tree users.
Initialising the maple tree is the same as in the :ref:`maple-tree-normal-api`.
Please see above.
The maple state keeps track of the range start and end in mas->index and
mas->last, respectively.
mas_walk() will walk the tree to the location of mas->index and set the
mas->index and mas->last according to the range for the entry.
You can set entries using mas_store(). mas_store() will overwrite any entry
with the new entry and return the first existing entry that is overwritten.
The range is passed in as members of the maple state: index and last.
You can use mas_erase() to erase an entire range by setting index and
last of the maple state to the desired range to erase. This will erase
the first range that is found in that range, set the maple state index
and last as the range that was erased and return the entry that existed
at that location.
You can walk each entry within a range by using mas_for_each(). If you want
to walk each element of the tree then ``0`` and ``ULONG_MAX`` may be used as
the range. If the lock needs to be periodically dropped, see the locking
section mas_pause().
Using a maple state allows mas_next() and mas_prev() to function as if the
tree was a linked list. With such a high branching factor the amortized
performance penalty is outweighed by cache optimization. mas_next() will
return the next entry which occurs after the entry at index. mas_prev()
will return the previous entry which occurs before the entry at index.
mas_find() will find the first entry which exists at or above index on
the first call, and the next entry from every subsequent calls.
mas_find_rev() will find the fist entry which exists at or below the last on
the first call, and the previous entry from every subsequent calls.
If the user needs to yield the lock during an operation, then the maple state
must be paused using mas_pause().
There are a few extra interfaces provided when using an allocation tree.
If you wish to search for a gap within a range, then mas_empty_area()
or mas_empty_area_rev() can be used. mas_empty_area() searches for a gap
starting at the lowest index given up to the maximum of the range.
mas_empty_area_rev() searches for a gap starting at the highest index given
and continues downward to the lower bound of the range.
.. _maple-tree-advanced-alloc:
Advanced Allocating Nodes
-------------------------
Allocations are usually handled internally to the tree, however if allocations
need to occur before a write occurs then calling mas_expected_entries() will
allocate the worst-case number of needed nodes to insert the provided number of
ranges. This also causes the tree to enter mass insertion mode. Once
insertions are complete calling mas_destroy() on the maple state will free the
unused allocations.
.. _maple-tree-advanced-locks:
Advanced Locking
----------------
The maple tree uses a spinlock by default, but external locks can be used for
tree updates as well. To use an external lock, the tree must be initialized
with the ``MT_FLAGS_LOCK_EXTERN flag``, this is usually done with the
MTREE_INIT_EXT() #define, which takes an external lock as an argument.
Functions and structures
========================
.. kernel-doc:: include/linux/maple_tree.h
.. kernel-doc:: lib/maple_tree.c
...@@ -12092,6 +12092,18 @@ L: linux-man@vger.kernel.org ...@@ -12092,6 +12092,18 @@ L: linux-man@vger.kernel.org
S: Maintained S: Maintained
W: http://www.kernel.org/doc/man-pages W: http://www.kernel.org/doc/man-pages
MAPLE TREE
M: Liam R. Howlett <Liam.Howlett@oracle.com>
L: linux-mm@kvack.org
S: Supported
F: Documentation/core-api/maple_tree.rst
F: include/linux/maple_tree.h
F: include/trace/events/maple_tree.h
F: lib/maple_tree.c
F: lib/test_maple_tree.c
F: tools/testing/radix-tree/linux/maple_tree.h
F: tools/testing/radix-tree/maple.c
MARDUK (CREATOR CI40) DEVICE TREE SUPPORT MARDUK (CREATOR CI40) DEVICE TREE SUPPORT
M: Rahul Bedarkar <rahulbedarkar89@gmail.com> M: Rahul Bedarkar <rahulbedarkar89@gmail.com>
L: linux-mips@vger.kernel.org L: linux-mips@vger.kernel.org
......
/* SPDX-License-Identifier: GPL-2.0+ */
#ifndef _LINUX_MAPLE_TREE_H
#define _LINUX_MAPLE_TREE_H
/*
* Maple Tree - An RCU-safe adaptive tree for storing ranges
* Copyright (c) 2018-2022 Oracle
* Authors: Liam R. Howlett <Liam.Howlett@Oracle.com>
* Matthew Wilcox <willy@infradead.org>
*/
#include <linux/kernel.h>
#include <linux/rcupdate.h>
#include <linux/spinlock.h>
/* #define CONFIG_MAPLE_RCU_DISABLED */
/* #define CONFIG_DEBUG_MAPLE_TREE_VERBOSE */
/*
* Allocated nodes are mutable until they have been inserted into the tree,
* at which time they cannot change their type until they have been removed
* from the tree and an RCU grace period has passed.
*
* Removed nodes have their ->parent set to point to themselves. RCU readers
* check ->parent before relying on the value that they loaded from the
* slots array. This lets us reuse the slots array for the RCU head.
*
* Nodes in the tree point to their parent unless bit 0 is set.
*/
#if defined(CONFIG_64BIT) || defined(BUILD_VDSO32_64)
/* 64bit sizes */
#define MAPLE_NODE_SLOTS 31 /* 256 bytes including ->parent */
#define MAPLE_RANGE64_SLOTS 16 /* 256 bytes */
#define MAPLE_ARANGE64_SLOTS 10 /* 240 bytes */
#define MAPLE_ARANGE64_META_MAX 15 /* Out of range for metadata */
#define MAPLE_ALLOC_SLOTS (MAPLE_NODE_SLOTS - 1)
#else
/* 32bit sizes */
#define MAPLE_NODE_SLOTS 63 /* 256 bytes including ->parent */
#define MAPLE_RANGE64_SLOTS 32 /* 256 bytes */
#define MAPLE_ARANGE64_SLOTS 21 /* 240 bytes */
#define MAPLE_ARANGE64_META_MAX 31 /* Out of range for metadata */
#define MAPLE_ALLOC_SLOTS (MAPLE_NODE_SLOTS - 2)
#endif /* defined(CONFIG_64BIT) || defined(BUILD_VDSO32_64) */
#define MAPLE_NODE_MASK 255UL
/*
* The node->parent of the root node has bit 0 set and the rest of the pointer
* is a pointer to the tree itself. No more bits are available in this pointer
* (on m68k, the data structure may only be 2-byte aligned).
*
* Internal non-root nodes can only have maple_range_* nodes as parents. The
* parent pointer is 256B aligned like all other tree nodes. When storing a 32
* or 64 bit values, the offset can fit into 4 bits. The 16 bit values need an
* extra bit to store the offset. This extra bit comes from a reuse of the last
* bit in the node type. This is possible by using bit 1 to indicate if bit 2
* is part of the type or the slot.
*
* Once the type is decided, the decision of an allocation range type or a range
* type is done by examining the immutable tree flag for the MAPLE_ALLOC_RANGE
* flag.
*
* Node types:
* 0x??1 = Root
* 0x?00 = 16 bit nodes
* 0x010 = 32 bit nodes
* 0x110 = 64 bit nodes
*
* Slot size and location in the parent pointer:
* type : slot location
* 0x??1 : Root
* 0x?00 : 16 bit values, type in 0-1, slot in 2-6
* 0x010 : 32 bit values, type in 0-2, slot in 3-6
* 0x110 : 64 bit values, type in 0-2, slot in 3-6
*/
/*
* This metadata is used to optimize the gap updating code and in reverse
* searching for gaps or any other code that needs to find the end of the data.
*/
struct maple_metadata {
unsigned char end;
unsigned char gap;
};
/*
* Leaf nodes do not store pointers to nodes, they store user data. Users may
* store almost any bit pattern. As noted above, the optimisation of storing an
* entry at 0 in the root pointer cannot be done for data which have the bottom
* two bits set to '10'. We also reserve values with the bottom two bits set to
* '10' which are below 4096 (ie 2, 6, 10 .. 4094) for internal use. Some APIs
* return errnos as a negative errno shifted right by two bits and the bottom
* two bits set to '10', and while choosing to store these values in the array
* is not an error, it may lead to confusion if you're testing for an error with
* mas_is_err().
*
* Non-leaf nodes store the type of the node pointed to (enum maple_type in bits
* 3-6), bit 2 is reserved. That leaves bits 0-1 unused for now.
*
* In regular B-Tree terms, pivots are called keys. The term pivot is used to
* indicate that the tree is specifying ranges, Pivots may appear in the
* subtree with an entry attached to the value whereas keys are unique to a
* specific position of a B-tree. Pivot values are inclusive of the slot with
* the same index.
*/
struct maple_range_64 {
struct maple_pnode *parent;
unsigned long pivot[MAPLE_RANGE64_SLOTS - 1];
union {
void __rcu *slot[MAPLE_RANGE64_SLOTS];
struct {
void __rcu *pad[MAPLE_RANGE64_SLOTS - 1];
struct maple_metadata meta;
};
};
};
/*
* At tree creation time, the user can specify that they're willing to trade off
* storing fewer entries in a tree in return for storing more information in
* each node.
*
* The maple tree supports recording the largest range of NULL entries available
* in this node, also called gaps. This optimises the tree for allocating a
* range.
*/
struct maple_arange_64 {
struct maple_pnode *parent;
unsigned long pivot[MAPLE_ARANGE64_SLOTS - 1];
void __rcu *slot[MAPLE_ARANGE64_SLOTS];
unsigned long gap[MAPLE_ARANGE64_SLOTS];
struct maple_metadata meta;
};
struct maple_alloc {
unsigned long total;
unsigned char node_count;
unsigned int request_count;
struct maple_alloc *slot[MAPLE_ALLOC_SLOTS];
};
struct maple_topiary {
struct maple_pnode *parent;
struct maple_enode *next; /* Overlaps the pivot */
};
enum maple_type {
maple_dense,
maple_leaf_64,
maple_range_64,
maple_arange_64,
};
/**
* DOC: Maple tree flags
*
* * MT_FLAGS_ALLOC_RANGE - Track gaps in this tree
* * MT_FLAGS_USE_RCU - Operate in RCU mode
* * MT_FLAGS_HEIGHT_OFFSET - The position of the tree height in the flags
* * MT_FLAGS_HEIGHT_MASK - The mask for the maple tree height value
* * MT_FLAGS_LOCK_MASK - How the mt_lock is used
* * MT_FLAGS_LOCK_IRQ - Acquired irq-safe
* * MT_FLAGS_LOCK_BH - Acquired bh-safe
* * MT_FLAGS_LOCK_EXTERN - mt_lock is not used
*
* MAPLE_HEIGHT_MAX The largest height that can be stored
*/
#define MT_FLAGS_ALLOC_RANGE 0x01
#define MT_FLAGS_USE_RCU 0x02
#define MT_FLAGS_HEIGHT_OFFSET 0x02
#define MT_FLAGS_HEIGHT_MASK 0x7C
#define MT_FLAGS_LOCK_MASK 0x300
#define MT_FLAGS_LOCK_IRQ 0x100
#define MT_FLAGS_LOCK_BH 0x200
#define MT_FLAGS_LOCK_EXTERN 0x300
#define MAPLE_HEIGHT_MAX 31
#define MAPLE_NODE_TYPE_MASK 0x0F
#define MAPLE_NODE_TYPE_SHIFT 0x03
#define MAPLE_RESERVED_RANGE 4096
#ifdef CONFIG_LOCKDEP
typedef struct lockdep_map *lockdep_map_p;
#define mt_lock_is_held(mt) lock_is_held(mt->ma_external_lock)
#define mt_set_external_lock(mt, lock) \
(mt)->ma_external_lock = &(lock)->dep_map
#else
typedef struct { /* nothing */ } lockdep_map_p;
#define mt_lock_is_held(mt) 1
#define mt_set_external_lock(mt, lock) do { } while (0)
#endif
/*
* If the tree contains a single entry at index 0, it is usually stored in
* tree->ma_root. To optimise for the page cache, an entry which ends in '00',
* '01' or '11' is stored in the root, but an entry which ends in '10' will be
* stored in a node. Bits 3-6 are used to store enum maple_type.
*
* The flags are used both to store some immutable information about this tree
* (set at tree creation time) and dynamic information set under the spinlock.
*
* Another use of flags are to indicate global states of the tree. This is the
* case with the MAPLE_USE_RCU flag, which indicates the tree is currently in
* RCU mode. This mode was added to allow the tree to reuse nodes instead of
* re-allocating and RCU freeing nodes when there is a single user.
*/
struct maple_tree {
union {
spinlock_t ma_lock;
lockdep_map_p ma_external_lock;
};
void __rcu *ma_root;
unsigned int ma_flags;
};
/**
* MTREE_INIT() - Initialize a maple tree
* @name: The maple tree name
* @__flags: The maple tree flags
*
*/
#define MTREE_INIT(name, __flags) { \
.ma_lock = __SPIN_LOCK_UNLOCKED((name).ma_lock), \
.ma_flags = __flags, \
.ma_root = NULL, \
}
/**
* MTREE_INIT_EXT() - Initialize a maple tree with an external lock.
* @name: The tree name
* @__flags: The maple tree flags
* @__lock: The external lock
*/
#ifdef CONFIG_LOCKDEP
#define MTREE_INIT_EXT(name, __flags, __lock) { \
.ma_external_lock = &(__lock).dep_map, \
.ma_flags = (__flags), \
.ma_root = NULL, \
}
#else
#define MTREE_INIT_EXT(name, __flags, __lock) MTREE_INIT(name, __flags)
#endif
#define DEFINE_MTREE(name) \
struct maple_tree name = MTREE_INIT(name, 0)
#define mtree_lock(mt) spin_lock((&(mt)->ma_lock))
#define mtree_unlock(mt) spin_unlock((&(mt)->ma_lock))
/*
* The Maple Tree squeezes various bits in at various points which aren't
* necessarily obvious. Usually, this is done by observing that pointers are
* N-byte aligned and thus the bottom log_2(N) bits are available for use. We
* don't use the high bits of pointers to store additional information because
* we don't know what bits are unused on any given architecture.
*
* Nodes are 256 bytes in size and are also aligned to 256 bytes, giving us 8
* low bits for our own purposes. Nodes are currently of 4 types:
* 1. Single pointer (Range is 0-0)
* 2. Non-leaf Allocation Range nodes
* 3. Non-leaf Range nodes
* 4. Leaf Range nodes All nodes consist of a number of node slots,
* pivots, and a parent pointer.
*/
struct maple_node {
union {
struct {
struct maple_pnode *parent;
void __rcu *slot[MAPLE_NODE_SLOTS];
};
struct {
void *pad;
struct rcu_head rcu;
struct maple_enode *piv_parent;
unsigned char parent_slot;
enum maple_type type;
unsigned char slot_len;
unsigned int ma_flags;
};
struct maple_range_64 mr64;
struct maple_arange_64 ma64;
struct maple_alloc alloc;
};
};
/*
* More complicated stores can cause two nodes to become one or three and
* potentially alter the height of the tree. Either half of the tree may need
* to be rebalanced against the other. The ma_topiary struct is used to track
* which nodes have been 'cut' from the tree so that the change can be done
* safely at a later date. This is done to support RCU.
*/
struct ma_topiary {
struct maple_enode *head;
struct maple_enode *tail;
struct maple_tree *mtree;
};
void *mtree_load(struct maple_tree *mt, unsigned long index);
int mtree_insert(struct maple_tree *mt, unsigned long index,
void *entry, gfp_t gfp);
int mtree_insert_range(struct maple_tree *mt, unsigned long first,
unsigned long last, void *entry, gfp_t gfp);
int mtree_alloc_range(struct maple_tree *mt, unsigned long *startp,
void *entry, unsigned long size, unsigned long min,
unsigned long max, gfp_t gfp);
int mtree_alloc_rrange(struct maple_tree *mt, unsigned long *startp,
void *entry, unsigned long size, unsigned long min,
unsigned long max, gfp_t gfp);
int mtree_store_range(struct maple_tree *mt, unsigned long first,
unsigned long last, void *entry, gfp_t gfp);
int mtree_store(struct maple_tree *mt, unsigned long index,
void *entry, gfp_t gfp);
void *mtree_erase(struct maple_tree *mt, unsigned long index);
void mtree_destroy(struct maple_tree *mt);
void __mt_destroy(struct maple_tree *mt);
/**
* mtree_empty() - Determine if a tree has any present entries.
* @mt: Maple Tree.
*
* Context: Any context.
* Return: %true if the tree contains only NULL pointers.
*/
static inline bool mtree_empty(const struct maple_tree *mt)
{
return mt->ma_root == NULL;
}
/* Advanced API */
/*
* The maple state is defined in the struct ma_state and is used to keep track
* of information during operations, and even between operations when using the
* advanced API.
*
* If state->node has bit 0 set then it references a tree location which is not
* a node (eg the root). If bit 1 is set, the rest of the bits are a negative
* errno. Bit 2 (the 'unallocated slots' bit) is clear. Bits 3-6 indicate the
* node type.
*
* state->alloc either has a request number of nodes or an allocated node. If
* stat->alloc has a requested number of nodes, the first bit will be set (0x1)
* and the remaining bits are the value. If state->alloc is a node, then the
* node will be of type maple_alloc. maple_alloc has MAPLE_NODE_SLOTS - 1 for
* storing more allocated nodes, a total number of nodes allocated, and the
* node_count in this node. node_count is the number of allocated nodes in this
* node. The scaling beyond MAPLE_NODE_SLOTS - 1 is handled by storing further
* nodes into state->alloc->slot[0]'s node. Nodes are taken from state->alloc
* by removing a node from the state->alloc node until state->alloc->node_count
* is 1, when state->alloc is returned and the state->alloc->slot[0] is promoted
* to state->alloc. Nodes are pushed onto state->alloc by putting the current
* state->alloc into the pushed node's slot[0].
*
* The state also contains the implied min/max of the state->node, the depth of
* this search, and the offset. The implied min/max are either from the parent
* node or are 0-oo for the root node. The depth is incremented or decremented
* every time a node is walked down or up. The offset is the slot/pivot of
* interest in the node - either for reading or writing.
*
* When returning a value the maple state index and last respectively contain
* the start and end of the range for the entry. Ranges are inclusive in the
* Maple Tree.
*/
struct ma_state {
struct maple_tree *tree; /* The tree we're operating in */
unsigned long index; /* The index we're operating on - range start */
unsigned long last; /* The last index we're operating on - range end */
struct maple_enode *node; /* The node containing this entry */
unsigned long min; /* The minimum index of this node - implied pivot min */
unsigned long max; /* The maximum index of this node - implied pivot max */
struct maple_alloc *alloc; /* Allocated nodes for this operation */
unsigned char depth; /* depth of tree descent during write */
unsigned char offset;
unsigned char mas_flags;
};
struct ma_wr_state {
struct ma_state *mas;
struct maple_node *node; /* Decoded mas->node */
unsigned long r_min; /* range min */
unsigned long r_max; /* range max */
enum maple_type type; /* mas->node type */
unsigned char offset_end; /* The offset where the write ends */
unsigned char node_end; /* mas->node end */
unsigned long *pivots; /* mas->node->pivots pointer */
unsigned long end_piv; /* The pivot at the offset end */
void __rcu **slots; /* mas->node->slots pointer */
void *entry; /* The entry to write */
void *content; /* The existing entry that is being overwritten */
};
#define mas_lock(mas) spin_lock(&((mas)->tree->ma_lock))
#define mas_unlock(mas) spin_unlock(&((mas)->tree->ma_lock))
/*
* Special values for ma_state.node.
* MAS_START means we have not searched the tree.
* MAS_ROOT means we have searched the tree and the entry we found lives in
* the root of the tree (ie it has index 0, length 1 and is the only entry in
* the tree).
* MAS_NONE means we have searched the tree and there is no node in the
* tree for this entry. For example, we searched for index 1 in an empty
* tree. Or we have a tree which points to a full leaf node and we
* searched for an entry which is larger than can be contained in that
* leaf node.
* MA_ERROR represents an errno. After dropping the lock and attempting
* to resolve the error, the walk would have to be restarted from the
* top of the tree as the tree may have been modified.
*/
#define MAS_START ((struct maple_enode *)1UL)
#define MAS_ROOT ((struct maple_enode *)5UL)
#define MAS_NONE ((struct maple_enode *)9UL)
#define MAS_PAUSE ((struct maple_enode *)17UL)
#define MA_ERROR(err) \
((struct maple_enode *)(((unsigned long)err << 2) | 2UL))
#define MA_STATE(name, mt, first, end) \
struct ma_state name = { \
.tree = mt, \
.index = first, \
.last = end, \
.node = MAS_START, \
.min = 0, \
.max = ULONG_MAX, \
.alloc = NULL, \
}
#define MA_WR_STATE(name, ma_state, wr_entry) \
struct ma_wr_state name = { \
.mas = ma_state, \
.content = NULL, \
.entry = wr_entry, \
}
#define MA_TOPIARY(name, tree) \
struct ma_topiary name = { \
.head = NULL, \
.tail = NULL, \
.mtree = tree, \
}
void *mas_walk(struct ma_state *mas);
void *mas_store(struct ma_state *mas, void *entry);
void *mas_erase(struct ma_state *mas);
int mas_store_gfp(struct ma_state *mas, void *entry, gfp_t gfp);
void mas_store_prealloc(struct ma_state *mas, void *entry);
void *mas_find(struct ma_state *mas, unsigned long max);
void *mas_find_rev(struct ma_state *mas, unsigned long min);
int mas_preallocate(struct ma_state *mas, void *entry, gfp_t gfp);
bool mas_is_err(struct ma_state *mas);
bool mas_nomem(struct ma_state *mas, gfp_t gfp);
void mas_pause(struct ma_state *mas);
void maple_tree_init(void);
void mas_destroy(struct ma_state *mas);
int mas_expected_entries(struct ma_state *mas, unsigned long nr_entries);
void *mas_prev(struct ma_state *mas, unsigned long min);
void *mas_next(struct ma_state *mas, unsigned long max);
int mas_empty_area(struct ma_state *mas, unsigned long min, unsigned long max,
unsigned long size);
/* Checks if a mas has not found anything */
static inline bool mas_is_none(struct ma_state *mas)
{
return mas->node == MAS_NONE;
}
/* Checks if a mas has been paused */
static inline bool mas_is_paused(struct ma_state *mas)
{
return mas->node == MAS_PAUSE;
}
void mas_dup_tree(struct ma_state *oldmas, struct ma_state *mas);
void mas_dup_store(struct ma_state *mas, void *entry);
/*
* This finds an empty area from the highest address to the lowest.
* AKA "Topdown" version,
*/
int mas_empty_area_rev(struct ma_state *mas, unsigned long min,
unsigned long max, unsigned long size);
/**
* mas_reset() - Reset a Maple Tree operation state.
* @mas: Maple Tree operation state.
*
* Resets the error or walk state of the @mas so future walks of the
* array will start from the root. Use this if you have dropped the
* lock and want to reuse the ma_state.
*
* Context: Any context.
*/
static inline void mas_reset(struct ma_state *mas)
{
mas->node = MAS_START;
}
/**
* mas_for_each() - Iterate over a range of the maple tree.
* @__mas: Maple Tree operation state (maple_state)
* @__entry: Entry retrieved from the tree
* @__max: maximum index to retrieve from the tree
*
* When returned, mas->index and mas->last will hold the entire range for the
* entry.
*
* Note: may return the zero entry.
*
*/
#define mas_for_each(__mas, __entry, __max) \
while (((__entry) = mas_find((__mas), (__max))) != NULL)
/**
* mas_set_range() - Set up Maple Tree operation state for a different index.
* @mas: Maple Tree operation state.
* @start: New start of range in the Maple Tree.
* @last: New end of range in the Maple Tree.
*
* Move the operation state to refer to a different range. This will
* have the effect of starting a walk from the top; see mas_next()
* to move to an adjacent index.
*/
static inline
void mas_set_range(struct ma_state *mas, unsigned long start, unsigned long last)
{
mas->index = start;
mas->last = last;
mas->node = MAS_START;
}
/**
* mas_set() - Set up Maple Tree operation state for a different index.
* @mas: Maple Tree operation state.
* @index: New index into the Maple Tree.
*
* Move the operation state to refer to a different index. This will
* have the effect of starting a walk from the top; see mas_next()
* to move to an adjacent index.
*/
static inline void mas_set(struct ma_state *mas, unsigned long index)
{
mas_set_range(mas, index, index);
}
static inline bool mt_external_lock(const struct maple_tree *mt)
{
return (mt->ma_flags & MT_FLAGS_LOCK_MASK) == MT_FLAGS_LOCK_EXTERN;
}
/**
* mt_init_flags() - Initialise an empty maple tree with flags.
* @mt: Maple Tree
* @flags: maple tree flags.
*
* If you need to initialise a Maple Tree with special flags (eg, an
* allocation tree), use this function.
*
* Context: Any context.
*/
static inline void mt_init_flags(struct maple_tree *mt, unsigned int flags)
{
mt->ma_flags = flags;
if (!mt_external_lock(mt))
spin_lock_init(&mt->ma_lock);
rcu_assign_pointer(mt->ma_root, NULL);
}
/**
* mt_init() - Initialise an empty maple tree.
* @mt: Maple Tree
*
* An empty Maple Tree.
*
* Context: Any context.
*/
static inline void mt_init(struct maple_tree *mt)
{
mt_init_flags(mt, 0);
}
static inline bool mt_in_rcu(struct maple_tree *mt)
{
#ifdef CONFIG_MAPLE_RCU_DISABLED
return false;
#endif
return mt->ma_flags & MT_FLAGS_USE_RCU;
}
/**
* mt_clear_in_rcu() - Switch the tree to non-RCU mode.
* @mt: The Maple Tree
*/
static inline void mt_clear_in_rcu(struct maple_tree *mt)
{
if (!mt_in_rcu(mt))
return;
if (mt_external_lock(mt)) {
BUG_ON(!mt_lock_is_held(mt));
mt->ma_flags &= ~MT_FLAGS_USE_RCU;
} else {
mtree_lock(mt);
mt->ma_flags &= ~MT_FLAGS_USE_RCU;
mtree_unlock(mt);
}
}
/**
* mt_set_in_rcu() - Switch the tree to RCU safe mode.
* @mt: The Maple Tree
*/
static inline void mt_set_in_rcu(struct maple_tree *mt)
{
if (mt_in_rcu(mt))
return;
if (mt_external_lock(mt)) {
BUG_ON(!mt_lock_is_held(mt));
mt->ma_flags |= MT_FLAGS_USE_RCU;
} else {
mtree_lock(mt);
mt->ma_flags |= MT_FLAGS_USE_RCU;
mtree_unlock(mt);
}
}
void *mt_find(struct maple_tree *mt, unsigned long *index, unsigned long max);
void *mt_find_after(struct maple_tree *mt, unsigned long *index,
unsigned long max);
void *mt_prev(struct maple_tree *mt, unsigned long index, unsigned long min);
void *mt_next(struct maple_tree *mt, unsigned long index, unsigned long max);
/**
* mt_for_each - Iterate over each entry starting at index until max.
* @__tree: The Maple Tree
* @__entry: The current entry
* @__index: The index to update to track the location in the tree
* @__max: The maximum limit for @index
*
* Note: Will not return the zero entry.
*/
#define mt_for_each(__tree, __entry, __index, __max) \
for (__entry = mt_find(__tree, &(__index), __max); \
__entry; __entry = mt_find_after(__tree, &(__index), __max))
#ifdef CONFIG_DEBUG_MAPLE_TREE
extern atomic_t maple_tree_tests_run;
extern atomic_t maple_tree_tests_passed;
void mt_dump(const struct maple_tree *mt);
void mt_validate(struct maple_tree *mt);
#define MT_BUG_ON(__tree, __x) do { \
atomic_inc(&maple_tree_tests_run); \
if (__x) { \
pr_info("BUG at %s:%d (%u)\n", \
__func__, __LINE__, __x); \
mt_dump(__tree); \
pr_info("Pass: %u Run:%u\n", \
atomic_read(&maple_tree_tests_passed), \
atomic_read(&maple_tree_tests_run)); \
dump_stack(); \
} else { \
atomic_inc(&maple_tree_tests_passed); \
} \
} while (0)
#else
#define MT_BUG_ON(__tree, __x) BUG_ON(__x)
#endif /* CONFIG_DEBUG_MAPLE_TREE */
#endif /*_LINUX_MAPLE_TREE_H */
/* SPDX-License-Identifier: GPL-2.0 */
#undef TRACE_SYSTEM
#define TRACE_SYSTEM maple_tree
#if !defined(_TRACE_MM_H) || defined(TRACE_HEADER_MULTI_READ)
#define _TRACE_MM_H
#include <linux/tracepoint.h>
struct ma_state;
TRACE_EVENT(ma_op,
TP_PROTO(const char *fn, struct ma_state *mas),
TP_ARGS(fn, mas),
TP_STRUCT__entry(
__field(const char *, fn)
__field(unsigned long, min)
__field(unsigned long, max)
__field(unsigned long, index)
__field(unsigned long, last)
__field(void *, node)
),
TP_fast_assign(
__entry->fn = fn;
__entry->min = mas->min;
__entry->max = mas->max;
__entry->index = mas->index;
__entry->last = mas->last;
__entry->node = mas->node;
),
TP_printk("%s\tNode: %p (%lu %lu) range: %lu-%lu",
__entry->fn,
(void *) __entry->node,
(unsigned long) __entry->min,
(unsigned long) __entry->max,
(unsigned long) __entry->index,
(unsigned long) __entry->last
)
)
TRACE_EVENT(ma_read,
TP_PROTO(const char *fn, struct ma_state *mas),
TP_ARGS(fn, mas),
TP_STRUCT__entry(
__field(const char *, fn)
__field(unsigned long, min)
__field(unsigned long, max)
__field(unsigned long, index)
__field(unsigned long, last)
__field(void *, node)
),
TP_fast_assign(
__entry->fn = fn;
__entry->min = mas->min;
__entry->max = mas->max;
__entry->index = mas->index;
__entry->last = mas->last;
__entry->node = mas->node;
),
TP_printk("%s\tNode: %p (%lu %lu) range: %lu-%lu",
__entry->fn,
(void *) __entry->node,
(unsigned long) __entry->min,
(unsigned long) __entry->max,
(unsigned long) __entry->index,
(unsigned long) __entry->last
)
)
TRACE_EVENT(ma_write,
TP_PROTO(const char *fn, struct ma_state *mas, unsigned long piv,
void *val),
TP_ARGS(fn, mas, piv, val),
TP_STRUCT__entry(
__field(const char *, fn)
__field(unsigned long, min)
__field(unsigned long, max)
__field(unsigned long, index)
__field(unsigned long, last)
__field(unsigned long, piv)
__field(void *, val)
__field(void *, node)
),
TP_fast_assign(
__entry->fn = fn;
__entry->min = mas->min;
__entry->max = mas->max;
__entry->index = mas->index;
__entry->last = mas->last;
__entry->piv = piv;
__entry->val = val;
__entry->node = mas->node;
),
TP_printk("%s\tNode %p (%lu %lu) range:%lu-%lu piv (%lu) val %p",
__entry->fn,
(void *) __entry->node,
(unsigned long) __entry->min,
(unsigned long) __entry->max,
(unsigned long) __entry->index,
(unsigned long) __entry->last,
(unsigned long) __entry->piv,
(void *) __entry->val
)
)
#endif /* _TRACE_MM_H */
/* This part must be outside protection */
#include <trace/define_trace.h>
...@@ -117,6 +117,7 @@ static int kernel_init(void *); ...@@ -117,6 +117,7 @@ static int kernel_init(void *);
extern void init_IRQ(void); extern void init_IRQ(void);
extern void radix_tree_init(void); extern void radix_tree_init(void);
extern void maple_tree_init(void);
/* /*
* Debug helper: via this flag we know that we are in 'early bootup code' * Debug helper: via this flag we know that we are in 'early bootup code'
...@@ -1005,6 +1006,7 @@ asmlinkage __visible void __init __no_sanitize_address start_kernel(void) ...@@ -1005,6 +1006,7 @@ asmlinkage __visible void __init __no_sanitize_address start_kernel(void)
"Interrupts were enabled *very* early, fixing it\n")) "Interrupts were enabled *very* early, fixing it\n"))
local_irq_disable(); local_irq_disable();
radix_tree_init(); radix_tree_init();
maple_tree_init();
/* /*
* Set up housekeeping before setting up workqueues to allow the unbound * Set up housekeeping before setting up workqueues to allow the unbound
......
...@@ -820,6 +820,13 @@ config DEBUG_VM_VMACACHE ...@@ -820,6 +820,13 @@ config DEBUG_VM_VMACACHE
can cause significant overhead, so only enable it in non-production can cause significant overhead, so only enable it in non-production
environments. environments.
config DEBUG_VM_MAPLE_TREE
bool "Debug VM maple trees"
depends on DEBUG_VM
select DEBUG_MAPLE_TREE
help
Enable VM maple tree debugging information and extra validations.
If unsure, say N. If unsure, say N.
config DEBUG_VM_RB config DEBUG_VM_RB
...@@ -1635,6 +1642,14 @@ config BUG_ON_DATA_CORRUPTION ...@@ -1635,6 +1642,14 @@ config BUG_ON_DATA_CORRUPTION
If unsure, say N. If unsure, say N.
config DEBUG_MAPLE_TREE
bool "Debug maple trees"
depends on DEBUG_KERNEL
help
Enable maple tree debugging information and extra validations.
If unsure, say N.
endmenu endmenu
config DEBUG_CREDENTIALS config DEBUG_CREDENTIALS
......
...@@ -29,7 +29,7 @@ endif ...@@ -29,7 +29,7 @@ endif
lib-y := ctype.o string.o vsprintf.o cmdline.o \ lib-y := ctype.o string.o vsprintf.o cmdline.o \
rbtree.o radix-tree.o timerqueue.o xarray.o \ rbtree.o radix-tree.o timerqueue.o xarray.o \
idr.o extable.o irq_regs.o argv_split.o \ maple_tree.o idr.o extable.o irq_regs.o argv_split.o \
flex_proportions.o ratelimit.o show_mem.o \ flex_proportions.o ratelimit.o show_mem.o \
is_single_threaded.o plist.o decompress.o kobject_uevent.o \ is_single_threaded.o plist.o decompress.o kobject_uevent.o \
earlycpio.o seq_buf.o siphash.o dec_and_lock.o \ earlycpio.o seq_buf.o siphash.o dec_and_lock.o \
......
This source diff could not be displayed because it is too large. You can view the blob instead.
...@@ -6,3 +6,5 @@ main ...@@ -6,3 +6,5 @@ main
multiorder multiorder
radix-tree.c radix-tree.c
xarray xarray
maple
ma_xa_benchmark
#define CONFIG_XARRAY_MULTI 1 #define CONFIG_XARRAY_MULTI 1
#define CONFIG_64BIT 1
/* SPDX-License-Identifier: GPL-2.0+ */
#define atomic_t int32_t
#include "../../../../include/linux/maple_tree.h"
#define atomic_inc(x) uatomic_inc(x)
#define atomic_read(x) uatomic_read(x)
#define atomic_set(x, y) do {} while (0)
#define U8_MAX UCHAR_MAX
// SPDX-License-Identifier: GPL-2.0+
/*
* maple_tree.c: Userspace shim for maple tree test-suite
* Copyright (c) 2018 Liam R. Howlett <Liam.Howlett@Oracle.com>
*/
#define CONFIG_DEBUG_MAPLE_TREE
#define CONFIG_MAPLE_SEARCH
#include "test.h"
#define module_init(x)
#define module_exit(x)
#define MODULE_AUTHOR(x)
#define MODULE_LICENSE(x)
#define dump_stack() assert(0)
#include "../../../lib/maple_tree.c"
#undef CONFIG_DEBUG_MAPLE_TREE
#include "../../../lib/test_maple_tree.c"
void farmer_tests(void)
{
struct maple_node *node;
DEFINE_MTREE(tree);
mt_dump(&tree);
tree.ma_root = xa_mk_value(0);
mt_dump(&tree);
node = mt_alloc_one(GFP_KERNEL);
node->parent = (void *)((unsigned long)(&tree) | 1);
node->slot[0] = xa_mk_value(0);
node->slot[1] = xa_mk_value(1);
node->mr64.pivot[0] = 0;
node->mr64.pivot[1] = 1;
node->mr64.pivot[2] = 0;
tree.ma_root = mt_mk_node(node, maple_leaf_64);
mt_dump(&tree);
ma_free_rcu(node);
}
void maple_tree_tests(void)
{
farmer_tests();
maple_tree_seed();
maple_tree_harvest();
}
int __weak main(void)
{
maple_tree_init();
maple_tree_tests();
rcu_barrier();
if (nr_allocated)
printf("nr_allocated = %d\n", nr_allocated);
return 0;
}
/* SPDX-License-Identifier: GPL-2.0+ */
#define trace_ma_op(a, b) do {} while (0)
#define trace_ma_read(a, b) do {} while (0)
#define trace_ma_write(a, b, c, d) do {} while (0)
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment