Blame view
lib/list_sort.c
8.37 KB
b24413180 License cleanup: ... |
1 |
// SPDX-License-Identifier: GPL-2.0 |
2c761270d lib: Introduce ge... |
2 |
#include <linux/kernel.h> |
7259fa042 lib/list_sort.c: ... |
3 4 5 6 |
#include <linux/bug.h> #include <linux/compiler.h> #include <linux/export.h> #include <linux/string.h> |
2c761270d lib: Introduce ge... |
7 |
#include <linux/list_sort.h> |
2c761270d lib: Introduce ge... |
8 |
#include <linux/list.h> |
043b3f7b6 lib/list_sort: si... |
9 10 |
typedef int __attribute__((nonnull(2,3))) (*cmp_func)(void *, struct list_head const *, struct list_head const *); |
835cc0c84 lib: more scalabl... |
11 12 13 14 15 16 |
/* * Returns a list organized in an intermediate format suited * to chaining of merge() calls: null-terminated, no reserved or * sentinel head node, "prev" links not maintained. */ |
043b3f7b6 lib/list_sort: si... |
17 18 |
__attribute__((nonnull(2,3,4))) static struct list_head *merge(void *priv, cmp_func cmp, |
835cc0c84 lib: more scalabl... |
19 20 |
struct list_head *a, struct list_head *b) { |
043b3f7b6 lib/list_sort: si... |
21 |
struct list_head *head, **tail = &head; |
835cc0c84 lib: more scalabl... |
22 |
|
043b3f7b6 lib/list_sort: si... |
23 |
for (;;) { |
835cc0c84 lib: more scalabl... |
24 |
/* if equal, take 'a' -- important for sort stability */ |
043b3f7b6 lib/list_sort: si... |
25 26 27 |
if (cmp(priv, a, b) <= 0) { *tail = a; tail = &a->next; |
835cc0c84 lib: more scalabl... |
28 |
a = a->next; |
043b3f7b6 lib/list_sort: si... |
29 30 31 32 |
if (!a) { *tail = b; break; } |
835cc0c84 lib: more scalabl... |
33 |
} else { |
043b3f7b6 lib/list_sort: si... |
34 35 |
*tail = b; tail = &b->next; |
835cc0c84 lib: more scalabl... |
36 |
b = b->next; |
043b3f7b6 lib/list_sort: si... |
37 38 39 40 |
if (!b) { *tail = a; break; } |
835cc0c84 lib: more scalabl... |
41 |
} |
835cc0c84 lib: more scalabl... |
42 |
} |
043b3f7b6 lib/list_sort: si... |
43 |
return head; |
835cc0c84 lib: more scalabl... |
44 45 46 47 48 49 50 51 52 |
} /* * Combine final list merge with restoration of standard doubly-linked * list structure. This approach duplicates code from merge(), but * runs faster than the tidier alternatives of either a separate final * prev-link restoration pass, or maintaining the prev links * throughout. */ |
043b3f7b6 lib/list_sort: si... |
53 54 55 |
__attribute__((nonnull(2,3,4,5))) static void merge_final(void *priv, cmp_func cmp, struct list_head *head, struct list_head *a, struct list_head *b) |
835cc0c84 lib: more scalabl... |
56 57 |
{ struct list_head *tail = head; |
61b3d6c48 lib: list_sort.c:... |
58 |
u8 count = 0; |
835cc0c84 lib: more scalabl... |
59 |
|
043b3f7b6 lib/list_sort: si... |
60 |
for (;;) { |
835cc0c84 lib: more scalabl... |
61 |
/* if equal, take 'a' -- important for sort stability */ |
043b3f7b6 lib/list_sort: si... |
62 |
if (cmp(priv, a, b) <= 0) { |
835cc0c84 lib: more scalabl... |
63 64 |
tail->next = a; a->prev = tail; |
043b3f7b6 lib/list_sort: si... |
65 |
tail = a; |
835cc0c84 lib: more scalabl... |
66 |
a = a->next; |
043b3f7b6 lib/list_sort: si... |
67 68 |
if (!a) break; |
835cc0c84 lib: more scalabl... |
69 70 71 |
} else { tail->next = b; b->prev = tail; |
043b3f7b6 lib/list_sort: si... |
72 |
tail = b; |
835cc0c84 lib: more scalabl... |
73 |
b = b->next; |
043b3f7b6 lib/list_sort: si... |
74 75 76 77 |
if (!b) { b = a; break; } |
835cc0c84 lib: more scalabl... |
78 |
} |
835cc0c84 lib: more scalabl... |
79 |
} |
835cc0c84 lib: more scalabl... |
80 |
|
043b3f7b6 lib/list_sort: si... |
81 82 |
/* Finish linking remainder of list b on to tail */ tail->next = b; |
835cc0c84 lib: more scalabl... |
83 84 |
do { /* |
043b3f7b6 lib/list_sort: si... |
85 86 |
* If the merge is highly unbalanced (e.g. the input is * already sorted), this loop may run many iterations. |
835cc0c84 lib: more scalabl... |
87 88 89 90 |
* Continue callbacks to the client even though no * element comparison is needed, so the client's cmp() * routine can invoke cond_resched() periodically. */ |
043b3f7b6 lib/list_sort: si... |
91 92 93 94 95 96 97 98 |
if (unlikely(!++count)) cmp(priv, b, b); b->prev = tail; tail = b; b = b->next; } while (b); /* And the final links to make a circular doubly-linked list */ |
835cc0c84 lib: more scalabl... |
99 100 101 |
tail->next = head; head->prev = tail; } |
2c761270d lib: Introduce ge... |
102 |
/** |
02b12b7a2 lib: revise list_... |
103 104 |
* list_sort - sort a list * @priv: private data, opaque to list_sort(), passed to @cmp |
2c761270d lib: Introduce ge... |
105 106 107 |
* @head: the list to sort * @cmp: the elements comparison function * |
043b3f7b6 lib/list_sort: si... |
108 109 110 111 112 113 |
* The comparison funtion @cmp must return > 0 if @a should sort after * @b ("@a > @b" if you want an ascending sort), and <= 0 if @a should * sort before @b *or* their original order should be preserved. It is * always called with the element that came first in the input in @a, * and list_sort is a stable sort, so it is not necessary to distinguish * the @a < @b and @a == @b cases. |
2c761270d lib: Introduce ge... |
114 |
* |
043b3f7b6 lib/list_sort: si... |
115 116 117 118 119 120 |
* This is compatible with two styles of @cmp function: * - The traditional style which returns <0 / =0 / >0, or * - Returning a boolean 0/1. * The latter offers a chance to save a few cycles in the comparison * (which is used by e.g. plug_ctx_cmp() in block/blk-mq.c). * |
f35a1abd9 lib/list_sort: fi... |
121 122 |
* A good way to write a multi-word comparison is:: * |
043b3f7b6 lib/list_sort: si... |
123 124 125 126 127 |
* if (a->high != b->high) * return a->high > b->high; * if (a->middle != b->middle) * return a->middle > b->middle; * return a->low > b->low; |
b5c56e0cd lib/list_sort: op... |
128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
* * * This mergesort is as eager as possible while always performing at least * 2:1 balanced merges. Given two pending sublists of size 2^k, they are * merged to a size-2^(k+1) list as soon as we have 2^k following elements. * * Thus, it will avoid cache thrashing as long as 3*2^k elements can * fit into the cache. Not quite as good as a fully-eager bottom-up * mergesort, but it does use 0.2*n fewer comparisons, so is faster in * the common case that everything fits into L1. * * * The merging is controlled by "count", the number of elements in the * pending lists. This is beautiully simple code, but rather subtle. * * Each time we increment "count", we set one bit (bit k) and clear * bits k-1 .. 0. Each time this happens (except the very first time * for each bit, when count increments to 2^k), we merge two lists of * size 2^k into one list of size 2^(k+1). * * This merge happens exactly when the count reaches an odd multiple of * 2^k, which is when we have 2^k elements pending in smaller lists, * so it's safe to merge away two lists of size 2^k. * * After this happens twice, we have created two lists of size 2^(k+1), * which will be merged into a list of size 2^(k+2) before we create * a third list of size 2^(k+1), so there are never more than two pending. * * The number of pending lists of size 2^k is determined by the * state of bit k of "count" plus two extra pieces of information: |
4ae5b8f21 lib: list_sort.c:... |
158 |
* |
b5c56e0cd lib/list_sort: op... |
159 160 161 |
* - The state of bit k-1 (when k == 0, consider bit -1 always set), and * - Whether the higher-order bits are zero or non-zero (i.e. * is count >= 2^(k+1)). |
4ae5b8f21 lib: list_sort.c:... |
162 |
* |
b5c56e0cd lib/list_sort: op... |
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 |
* There are six states we distinguish. "x" represents some arbitrary * bits, and "y" represents some arbitrary non-zero bits: * 0: 00x: 0 pending of size 2^k; x pending of sizes < 2^k * 1: 01x: 0 pending of size 2^k; 2^(k-1) + x pending of sizes < 2^k * 2: x10x: 0 pending of size 2^k; 2^k + x pending of sizes < 2^k * 3: x11x: 1 pending of size 2^k; 2^(k-1) + x pending of sizes < 2^k * 4: y00x: 1 pending of size 2^k; 2^k + x pending of sizes < 2^k * 5: y01x: 2 pending of size 2^k; 2^(k-1) + x pending of sizes < 2^k * (merge and loop back to state 2) * * We gain lists of size 2^k in the 2->3 and 4->5 transitions (because * bit k-1 is set while the more significant bits are non-zero) and * merge them away in the 5->2 transition. Note in particular that just * before the 5->2 transition, all lower-order bits are 11 (state 3), * so there is one list of each smaller size. * * When we reach the end of the input, we merge all the pending * lists, from smallest to largest. If you work through cases 2 to * 5 above, you can see that the number of elements we merge with a list * of size 2^k varies from 2^(k-1) (cases 3 and 5 when x == 0) to * 2^(k+1) - 1 (second merge of case 5 when x == 2^(k-1) - 1). |
2c761270d lib: Introduce ge... |
184 |
*/ |
043b3f7b6 lib/list_sort: si... |
185 |
__attribute__((nonnull(2,3))) |
2c761270d lib: Introduce ge... |
186 |
void list_sort(void *priv, struct list_head *head, |
835cc0c84 lib: more scalabl... |
187 188 |
int (*cmp)(void *priv, struct list_head *a, struct list_head *b)) |
2c761270d lib: Introduce ge... |
189 |
{ |
043b3f7b6 lib/list_sort: si... |
190 191 |
struct list_head *list = head->next, *pending = NULL; size_t count = 0; /* Count of pending */ |
2c761270d lib: Introduce ge... |
192 |
|
043b3f7b6 lib/list_sort: si... |
193 |
if (list == head->prev) /* Zero or one elements */ |
2c761270d lib: Introduce ge... |
194 |
return; |
043b3f7b6 lib/list_sort: si... |
195 |
/* Convert to a null-terminated singly-linked list. */ |
835cc0c84 lib: more scalabl... |
196 |
head->prev->next = NULL; |
2c761270d lib: Introduce ge... |
197 |
|
043b3f7b6 lib/list_sort: si... |
198 199 200 201 202 203 |
/* * Data structure invariants: * - All lists are singly linked and null-terminated; prev * pointers are not maintained. * - pending is a prev-linked "list of lists" of sorted * sublists awaiting further merging. |
b5c56e0cd lib/list_sort: op... |
204 |
* - Each of the sorted sublists is power-of-two in size. |
043b3f7b6 lib/list_sort: si... |
205 |
* - Sublists are sorted by size and age, smallest & newest at front. |
b5c56e0cd lib/list_sort: op... |
206 207 208 209 210 211 212 213 214 |
* - There are zero to two sublists of each size. * - A pair of pending sublists are merged as soon as the number * of following pending elements equals their size (i.e. * each time count reaches an odd multiple of that size). * That ensures each later final merge will be at worst 2:1. * - Each round consists of: * - Merging the two sublists selected by the highest bit * which flips when count is incremented, and * - Adding an element from the input as a size-1 sublist. |
043b3f7b6 lib/list_sort: si... |
215 216 217 |
*/ do { size_t bits; |
b5c56e0cd lib/list_sort: op... |
218 |
struct list_head **tail = &pending; |
043b3f7b6 lib/list_sort: si... |
219 |
|
b5c56e0cd lib/list_sort: op... |
220 221 222 223 224 225 |
/* Find the least-significant clear bit in count */ for (bits = count; bits & 1; bits >>= 1) tail = &(*tail)->prev; /* Do the indicated merge */ if (likely(bits)) { struct list_head *a = *tail, *b = a->prev; |
835cc0c84 lib: more scalabl... |
226 |
|
b5c56e0cd lib/list_sort: op... |
227 228 229 230 |
a = merge(priv, (cmp_func)cmp, b, a); /* Install the merged result in place of the inputs */ a->prev = b->prev; *tail = a; |
835cc0c84 lib: more scalabl... |
231 |
} |
b5c56e0cd lib/list_sort: op... |
232 233 234 235 236 237 |
/* Move one element from input list to pending */ list->prev = pending; pending = list; list = list->next; pending->next = NULL; |
043b3f7b6 lib/list_sort: si... |
238 |
count++; |
b5c56e0cd lib/list_sort: op... |
239 240 241 242 243 244 245 |
} while (list); /* End of input; merge together all the pending lists. */ list = pending; pending = pending->prev; for (;;) { struct list_head *next = pending->prev; |
043b3f7b6 lib/list_sort: si... |
246 |
|
b5c56e0cd lib/list_sort: op... |
247 248 |
if (!next) break; |
043b3f7b6 lib/list_sort: si... |
249 |
list = merge(priv, (cmp_func)cmp, pending, list); |
b5c56e0cd lib/list_sort: op... |
250 |
pending = next; |
835cc0c84 lib: more scalabl... |
251 |
} |
043b3f7b6 lib/list_sort: si... |
252 253 |
/* The final merge, rebuilding prev links */ merge_final(priv, (cmp_func)cmp, head, pending, list); |
835cc0c84 lib: more scalabl... |
254 255 |
} EXPORT_SYMBOL(list_sort); |