1 /*- 2 * Copyright (c) 2002-2005, 2009, 2013 Jeffrey Roberson <jeff@FreeBSD.org> 3 * Copyright (c) 2004, 2005 Bosko Milekic <bmilekic@FreeBSD.org> 4 * All rights reserved. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice unmodified, this list of conditions, and the following 11 * disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 17 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 18 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 19 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, 20 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 21 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 22 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 23 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 24 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 25 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 26 * 27 * $FreeBSD$ 28 * 29 */ 30 31 #include <sys/_task.h> 32 33 /* 34 * This file includes definitions, structures, prototypes, and inlines that 35 * should not be used outside of the actual implementation of UMA. 36 */ 37 38 /* 39 * Here's a quick description of the relationship between the objects: 40 * 41 * Kegs contain lists of slabs which are stored in either the full bin, empty 42 * bin, or partially allocated bin, to reduce fragmentation. They also contain 43 * the user supplied value for size, which is adjusted for alignment purposes 44 * and rsize is the result of that. The Keg also stores information for 45 * managing a hash of page addresses that maps pages to uma_slab_t structures 46 * for pages that don't have embedded uma_slab_t's. 47 * 48 * The uma_slab_t may be embedded in a UMA_SLAB_SIZE chunk of memory or it may 49 * be allocated off the page from a special slab zone. The free list within a 50 * slab is managed with a bitmask. For item sizes that would yield more than 51 * 10% memory waste we potentially allocate a separate uma_slab_t if this will 52 * improve the number of items per slab that will fit. 53 * 54 * The only really gross cases, with regards to memory waste, are for those 55 * items that are just over half the page size. You can get nearly 50% waste, 56 * so you fall back to the memory footprint of the power of two allocator. I 57 * have looked at memory allocation sizes on many of the machines available to 58 * me, and there does not seem to be an abundance of allocations at this range 59 * so at this time it may not make sense to optimize for it. This can, of 60 * course, be solved with dynamic slab sizes. 61 * 62 * Kegs may serve multiple Zones but by far most of the time they only serve 63 * one. When a Zone is created, a Keg is allocated and setup for it. While 64 * the backing Keg stores slabs, the Zone caches Buckets of items allocated 65 * from the slabs. Each Zone is equipped with an init/fini and ctor/dtor 66 * pair, as well as with its own set of small per-CPU caches, layered above 67 * the Zone's general Bucket cache. 68 * 69 * The PCPU caches are protected by critical sections, and may be accessed 70 * safely only from their associated CPU, while the Zones backed by the same 71 * Keg all share a common Keg lock (to coalesce contention on the backing 72 * slabs). The backing Keg typically only serves one Zone but in the case of 73 * multiple Zones, one of the Zones is considered the Master Zone and all 74 * Zone-related stats from the Keg are done in the Master Zone. For an 75 * example of a Multi-Zone setup, refer to the Mbuf allocation code. 76 */ 77 78 /* 79 * This is the representation for normal (Non OFFPAGE slab) 80 * 81 * i == item 82 * s == slab pointer 83 * 84 * <---------------- Page (UMA_SLAB_SIZE) ------------------> 85 * ___________________________________________________________ 86 * | _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___________ | 87 * ||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i| |slab header|| 88 * ||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_| |___________|| 89 * |___________________________________________________________| 90 * 91 * 92 * This is an OFFPAGE slab. These can be larger than UMA_SLAB_SIZE. 93 * 94 * ___________________________________________________________ 95 * | _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ | 96 * ||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i||i| | 97 * ||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_||_| | 98 * |___________________________________________________________| 99 * ___________ ^ 100 * |slab header| | 101 * |___________|---* 102 * 103 */ 104 105 #ifndef VM_UMA_INT_H 106 #define VM_UMA_INT_H 107 108 #define UMA_SLAB_SIZE PAGE_SIZE /* How big are our slabs? */ 109 #define UMA_SLAB_MASK (PAGE_SIZE - 1) /* Mask to get back to the page */ 110 #define UMA_SLAB_SHIFT PAGE_SHIFT /* Number of bits PAGE_MASK */ 111 112 #define UMA_BOOT_PAGES 64 /* Pages allocated for startup */ 113 #define UMA_BOOT_PAGES_ZONES 32 /* Multiplier for pages to reserve */ 114 /* if uma_zone > PAGE_SIZE */ 115 116 /* Max waste percentage before going to off page slab management */ 117 #define UMA_MAX_WASTE 10 118 119 /* 120 * I doubt there will be many cases where this is exceeded. This is the initial 121 * size of the hash table for uma_slabs that are managed off page. This hash 122 * does expand by powers of two. Currently it doesn't get smaller. 123 */ 124 #define UMA_HASH_SIZE_INIT 32 125 126 /* 127 * I should investigate other hashing algorithms. This should yield a low 128 * number of collisions if the pages are relatively contiguous. 129 */ 130 131 #define UMA_HASH(h, s) ((((uintptr_t)s) >> UMA_SLAB_SHIFT) & (h)->uh_hashmask) 132 133 #define UMA_HASH_INSERT(h, s, mem) \ 134 SLIST_INSERT_HEAD(&(h)->uh_slab_hash[UMA_HASH((h), \ 135 (mem))], (s), us_hlink) 136 #define UMA_HASH_REMOVE(h, s, mem) \ 137 SLIST_REMOVE(&(h)->uh_slab_hash[UMA_HASH((h), \ 138 (mem))], (s), uma_slab, us_hlink) 139 140 /* Hash table for freed address -> slab translation */ 141 142 SLIST_HEAD(slabhead, uma_slab); 143 144 struct uma_hash { 145 struct slabhead *uh_slab_hash; /* Hash table for slabs */ 146 int uh_hashsize; /* Current size of the hash table */ 147 int uh_hashmask; /* Mask used during hashing */ 148 }; 149 150 /* 151 * align field or structure to cache line 152 */ 153 #if defined(__amd64__) 154 #define UMA_ALIGN __aligned(CACHE_LINE_SIZE) 155 #else 156 #define UMA_ALIGN 157 #endif 158 159 /* 160 * Structures for per cpu queues. 161 */ 162 163 struct uma_bucket { 164 LIST_ENTRY(uma_bucket) ub_link; /* Link into the zone */ 165 int16_t ub_cnt; /* Count of free items. */ 166 int16_t ub_entries; /* Max items. */ 167 void *ub_bucket[]; /* actual allocation storage */ 168 }; 169 170 typedef struct uma_bucket * uma_bucket_t; 171 172 struct uma_cache { 173 uma_bucket_t uc_freebucket; /* Bucket we're freeing to */ 174 uma_bucket_t uc_allocbucket; /* Bucket to allocate from */ 175 uint64_t uc_allocs; /* Count of allocations */ 176 uint64_t uc_frees; /* Count of frees */ 177 } UMA_ALIGN; 178 179 typedef struct uma_cache * uma_cache_t; 180 181 /* 182 * Keg management structure 183 * 184 * TODO: Optimize for cache line size 185 * 186 */ 187 struct uma_keg { 188 struct mtx_padalign uk_lock; /* Lock for the keg */ 189 struct uma_hash uk_hash; 190 191 LIST_HEAD(,uma_zone) uk_zones; /* Keg's zones */ 192 LIST_HEAD(,uma_slab) uk_part_slab; /* partially allocated slabs */ 193 LIST_HEAD(,uma_slab) uk_free_slab; /* empty slab list */ 194 LIST_HEAD(,uma_slab) uk_full_slab; /* full slabs */ 195 196 uint32_t uk_align; /* Alignment mask */ 197 uint32_t uk_pages; /* Total page count */ 198 uint32_t uk_free; /* Count of items free in slabs */ 199 uint32_t uk_reserve; /* Number of reserved items. */ 200 uint32_t uk_size; /* Requested size of each item */ 201 uint32_t uk_rsize; /* Real size of each item */ 202 uint32_t uk_maxpages; /* Maximum number of pages to alloc */ 203 204 uma_init uk_init; /* Keg's init routine */ 205 uma_fini uk_fini; /* Keg's fini routine */ 206 uma_alloc uk_allocf; /* Allocation function */ 207 uma_free uk_freef; /* Free routine */ 208 209 u_long uk_offset; /* Next free offset from base KVA */ 210 vm_offset_t uk_kva; /* Zone base KVA */ 211 uma_zone_t uk_slabzone; /* Slab zone backing us, if OFFPAGE */ 212 213 uint16_t uk_pgoff; /* Offset to uma_slab struct */ 214 uint16_t uk_ppera; /* pages per allocation from backend */ 215 uint16_t uk_ipers; /* Items per slab */ 216 uint32_t uk_flags; /* Internal flags */ 217 218 /* Least used fields go to the last cache line. */ 219 const char *uk_name; /* Name of creating zone. */ 220 LIST_ENTRY(uma_keg) uk_link; /* List of all kegs */ 221 }; 222 typedef struct uma_keg * uma_keg_t; 223 224 /* 225 * Free bits per-slab. 226 */ 227 #define SLAB_SETSIZE (PAGE_SIZE / UMA_SMALLEST_UNIT) 228 BITSET_DEFINE(slabbits, SLAB_SETSIZE); 229 230 /* 231 * The slab structure manages a single contiguous allocation from backing 232 * store and subdivides it into individually allocatable items. 233 */ 234 struct uma_slab { 235 uma_keg_t us_keg; /* Keg we live in */ 236 union { 237 LIST_ENTRY(uma_slab) _us_link; /* slabs in zone */ 238 unsigned long _us_size; /* Size of allocation */ 239 } us_type; 240 SLIST_ENTRY(uma_slab) us_hlink; /* Link for hash table */ 241 uint8_t *us_data; /* First item */ 242 struct slabbits us_free; /* Free bitmask. */ 243 #ifdef INVARIANTS 244 struct slabbits us_debugfree; /* Debug bitmask. */ 245 #endif 246 uint16_t us_freecount; /* How many are free? */ 247 uint8_t us_flags; /* Page flags see uma.h */ 248 uint8_t us_pad; /* Pad to 32bits, unused. */ 249 }; 250 251 #define us_link us_type._us_link 252 #define us_size us_type._us_size 253 254 typedef struct uma_slab * uma_slab_t; 255 typedef uma_slab_t (*uma_slaballoc)(uma_zone_t, uma_keg_t, int); 256 257 struct uma_klink { 258 LIST_ENTRY(uma_klink) kl_link; 259 uma_keg_t kl_keg; 260 }; 261 typedef struct uma_klink *uma_klink_t; 262 263 /* 264 * Zone management structure 265 * 266 * TODO: Optimize for cache line size 267 * 268 */ 269 struct uma_zone { 270 struct mtx_padalign uz_lock; /* Lock for the zone */ 271 struct mtx_padalign *uz_lockptr; 272 const char *uz_name; /* Text name of the zone */ 273 274 LIST_ENTRY(uma_zone) uz_link; /* List of all zones in keg */ 275 LIST_HEAD(,uma_bucket) uz_buckets; /* full buckets */ 276 277 LIST_HEAD(,uma_klink) uz_kegs; /* List of kegs. */ 278 struct uma_klink uz_klink; /* klink for first keg. */ 279 280 uma_slaballoc uz_slab; /* Allocate a slab from the backend. */ 281 uma_ctor uz_ctor; /* Constructor for each allocation */ 282 uma_dtor uz_dtor; /* Destructor */ 283 uma_init uz_init; /* Initializer for each item */ 284 uma_fini uz_fini; /* Finalizer for each item. */ 285 uma_import uz_import; /* Import new memory to cache. */ 286 uma_release uz_release; /* Release memory from cache. */ 287 void *uz_arg; /* Import/release argument. */ 288 289 uint32_t uz_flags; /* Flags inherited from kegs */ 290 uint32_t uz_size; /* Size inherited from kegs */ 291 292 volatile u_long uz_allocs UMA_ALIGN; /* Total number of allocations */ 293 volatile u_long uz_fails; /* Total number of alloc failures */ 294 volatile u_long uz_frees; /* Total number of frees */ 295 uint64_t uz_sleeps; /* Total number of alloc sleeps */ 296 uint16_t uz_count; /* Amount of items in full bucket */ 297 uint16_t uz_count_min; /* Minimal amount of items there */ 298 299 /* The next two fields are used to print a rate-limited warnings. */ 300 const char *uz_warning; /* Warning to print on failure */ 301 struct timeval uz_ratecheck; /* Warnings rate-limiting */ 302 303 struct task uz_maxaction; /* Task to run when at limit */ 304 305 /* 306 * This HAS to be the last item because we adjust the zone size 307 * based on NCPU and then allocate the space for the zones. 308 */ 309 struct uma_cache uz_cpu[1]; /* Per cpu caches */ 310 }; 311 312 /* 313 * These flags must not overlap with the UMA_ZONE flags specified in uma.h. 314 */ 315 #define UMA_ZFLAG_MULTI 0x04000000 /* Multiple kegs in the zone. */ 316 #define UMA_ZFLAG_DRAINING 0x08000000 /* Running zone_drain. */ 317 #define UMA_ZFLAG_BUCKET 0x10000000 /* Bucket zone. */ 318 #define UMA_ZFLAG_INTERNAL 0x20000000 /* No offpage no PCPU. */ 319 #define UMA_ZFLAG_FULL 0x40000000 /* Reached uz_maxpages */ 320 #define UMA_ZFLAG_CACHEONLY 0x80000000 /* Don't ask VM for buckets. */ 321 322 #define UMA_ZFLAG_INHERIT \ 323 (UMA_ZFLAG_INTERNAL | UMA_ZFLAG_CACHEONLY | UMA_ZFLAG_BUCKET) 324 325 static inline uma_keg_t 326 zone_first_keg(uma_zone_t zone) 327 { 328 uma_klink_t klink; 329 330 klink = LIST_FIRST(&zone->uz_kegs); 331 return (klink != NULL) ? klink->kl_keg : NULL; 332 } 333 334 #undef UMA_ALIGN 335 336 #ifdef _KERNEL 337 /* Internal prototypes */ 338 static __inline uma_slab_t hash_sfind(struct uma_hash *hash, uint8_t *data); 339 void *uma_large_malloc(vm_size_t size, int wait); 340 void uma_large_free(uma_slab_t slab); 341 342 /* Lock Macros */ 343 344 #define KEG_LOCK_INIT(k, lc) \ 345 do { \ 346 if ((lc)) \ 347 mtx_init(&(k)->uk_lock, (k)->uk_name, \ 348 (k)->uk_name, MTX_DEF | MTX_DUPOK); \ 349 else \ 350 mtx_init(&(k)->uk_lock, (k)->uk_name, \ 351 "UMA zone", MTX_DEF | MTX_DUPOK); \ 352 } while (0) 353 354 #define KEG_LOCK_FINI(k) mtx_destroy(&(k)->uk_lock) 355 #define KEG_LOCK(k) mtx_lock(&(k)->uk_lock) 356 #define KEG_UNLOCK(k) mtx_unlock(&(k)->uk_lock) 357 358 #define ZONE_LOCK_INIT(z, lc) \ 359 do { \ 360 if ((lc)) \ 361 mtx_init(&(z)->uz_lock, (z)->uz_name, \ 362 (z)->uz_name, MTX_DEF | MTX_DUPOK); \ 363 else \ 364 mtx_init(&(z)->uz_lock, (z)->uz_name, \ 365 "UMA zone", MTX_DEF | MTX_DUPOK); \ 366 } while (0) 367 368 #define ZONE_LOCK(z) mtx_lock((z)->uz_lockptr) 369 #define ZONE_TRYLOCK(z) mtx_trylock((z)->uz_lockptr) 370 #define ZONE_UNLOCK(z) mtx_unlock((z)->uz_lockptr) 371 #define ZONE_LOCK_FINI(z) mtx_destroy(&(z)->uz_lock) 372 373 /* 374 * Find a slab within a hash table. This is used for OFFPAGE zones to lookup 375 * the slab structure. 376 * 377 * Arguments: 378 * hash The hash table to search. 379 * data The base page of the item. 380 * 381 * Returns: 382 * A pointer to a slab if successful, else NULL. 383 */ 384 static __inline uma_slab_t 385 hash_sfind(struct uma_hash *hash, uint8_t *data) 386 { 387 uma_slab_t slab; 388 int hval; 389 390 hval = UMA_HASH(hash, data); 391 392 SLIST_FOREACH(slab, &hash->uh_slab_hash[hval], us_hlink) { 393 if ((uint8_t *)slab->us_data == data) 394 return (slab); 395 } 396 return (NULL); 397 } 398 399 static __inline uma_slab_t 400 vtoslab(vm_offset_t va) 401 { 402 vm_page_t p; 403 404 p = PHYS_TO_VM_PAGE(pmap_kextract(va)); 405 return ((uma_slab_t)p->plinks.s.pv); 406 } 407 408 static __inline void 409 vsetslab(vm_offset_t va, uma_slab_t slab) 410 { 411 vm_page_t p; 412 413 p = PHYS_TO_VM_PAGE(pmap_kextract(va)); 414 p->plinks.s.pv = slab; 415 } 416 417 /* 418 * The following two functions may be defined by architecture specific code 419 * if they can provide more efficient allocation functions. This is useful 420 * for using direct mapped addresses. 421 */ 422 void *uma_small_alloc(uma_zone_t zone, vm_size_t bytes, uint8_t *pflag, 423 int wait); 424 void uma_small_free(void *mem, vm_size_t size, uint8_t flags); 425 #endif /* _KERNEL */ 426 427 #endif /* VM_UMA_INT_H */ 428