1 /* -----------------------------------------------------------------------------
2 * $Id: Storage.c,v 1.23 2000/02/14 10:58:05 sewardj Exp $
4 * (c) The GHC Team, 1998-1999
6 * Storage manager front end
8 * ---------------------------------------------------------------------------*/
15 #include "BlockAlloc.h"
23 #include "StoragePriv.h"
26 nat nursery_blocks; /* number of blocks in the nursery */
29 StgClosure *caf_list = NULL;
31 bdescr *small_alloc_list; /* allocate()d small objects */
32 bdescr *large_alloc_list; /* allocate()d large objects */
33 nat alloc_blocks; /* number of allocate()d blocks since GC */
34 nat alloc_blocks_lim; /* approximate limit on alloc_blocks */
36 StgPtr alloc_Hp = NULL; /* next free byte in small_alloc_list */
37 StgPtr alloc_HpLim = NULL; /* end of block at small_alloc_list */
39 generation *generations; /* all the generations */
40 generation *g0; /* generation 0, for convenience */
41 generation *oldest_gen; /* oldest generation, for convenience */
42 step *g0s0; /* generation 0, step 0, for convenience */
45 * Storage manager mutex: protects all the above state from
46 * simultaneous access by two STG threads.
49 pthread_mutex_t sm_mutex = PTHREAD_MUTEX_INITIALIZER;
55 static void *stgAllocForGMP (size_t size_in_bytes);
56 static void *stgReallocForGMP (void *ptr, size_t old_size, size_t new_size);
57 static void stgDeallocForGMP (void *ptr, size_t size);
66 /* If we're doing heap profiling, we want a two-space heap with a
67 * fixed-size allocation area so that we get roughly even-spaced
70 #if defined(PROFILING) || defined(DEBUG)
71 if (RtsFlags.ProfFlags.doHeapProfile) {
72 RtsFlags.GcFlags.generations = 1;
73 RtsFlags.GcFlags.steps = 1;
74 RtsFlags.GcFlags.oldGenFactor = 0;
75 RtsFlags.GcFlags.heapSizeSuggestion = 0;
79 if (RtsFlags.GcFlags.heapSizeSuggestion >
80 RtsFlags.GcFlags.maxHeapSize) {
81 RtsFlags.GcFlags.maxHeapSize = RtsFlags.GcFlags.heapSizeSuggestion;
86 /* allocate generation info array */
87 generations = (generation *)stgMallocBytes(RtsFlags.GcFlags.generations
88 * sizeof(struct _generation),
91 /* Initialise all generations */
92 for(g = 0; g < RtsFlags.GcFlags.generations; g++) {
93 gen = &generations[g];
95 gen->mut_list = END_MUT_LIST;
96 gen->mut_once_list = END_MUT_LIST;
98 gen->failed_promotions = 0;
102 /* A couple of convenience pointers */
103 g0 = &generations[0];
104 oldest_gen = &generations[RtsFlags.GcFlags.generations-1];
106 /* Allocate step structures in each generation */
107 if (RtsFlags.GcFlags.generations > 1) {
108 /* Only for multiple-generations */
110 /* Oldest generation: one step */
111 oldest_gen->n_steps = 1;
113 stgMallocBytes(1 * sizeof(struct _step), "initStorage: last step");
115 /* set up all except the oldest generation with 2 steps */
116 for(g = 0; g < RtsFlags.GcFlags.generations-1; g++) {
117 generations[g].n_steps = RtsFlags.GcFlags.steps;
118 generations[g].steps =
119 stgMallocBytes (RtsFlags.GcFlags.steps * sizeof(struct _step),
120 "initStorage: steps");
124 /* single generation, i.e. a two-space collector */
126 g0->steps = stgMallocBytes (sizeof(struct _step), "initStorage: steps");
129 /* Initialise all steps */
130 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
131 for (s = 0; s < generations[g].n_steps; s++) {
132 step = &generations[g].steps[s];
136 step->gen = &generations[g];
141 step->scan_bd = NULL;
142 step->large_objects = NULL;
143 step->new_large_objects = NULL;
144 step->scavenged_large_objects = NULL;
148 /* Set up the destination pointers in each younger gen. step */
149 for (g = 0; g < RtsFlags.GcFlags.generations-1; g++) {
150 for (s = 0; s < generations[g].n_steps-1; s++) {
151 generations[g].steps[s].to = &generations[g].steps[s+1];
153 generations[g].steps[s].to = &generations[g+1].steps[0];
156 /* The oldest generation has one step and its destination is the
158 oldest_gen->steps[0].to = &oldest_gen->steps[0];
160 /* generation 0 is special: that's the nursery */
161 generations[0].max_blocks = 0;
163 /* G0S0: the allocation area. Policy: keep the allocation area
164 * small to begin with, even if we have a large suggested heap
165 * size. Reason: we're going to do a major collection first, and we
166 * don't want it to be a big one. This vague idea is borne out by
167 * rigorous experimental evidence.
169 g0s0 = &generations[0].steps[0];
173 weak_ptr_list = NULL;
176 /* initialise the allocate() interface */
177 small_alloc_list = NULL;
178 large_alloc_list = NULL;
180 alloc_blocks_lim = RtsFlags.GcFlags.minAllocAreaSize;
182 /* Tell GNU multi-precision pkg about our custom alloc functions */
183 mp_set_memory_functions(stgAllocForGMP, stgReallocForGMP, stgDeallocForGMP);
186 pthread_mutex_init(&sm_mutex, NULL);
189 IF_DEBUG(gc, stat_describe_gens());
195 stat_exit(calcAllocated());
199 newCAF(StgClosure* caf)
201 /* Put this CAF on the mutable list for the old generation.
202 * This is a HACK - the IND_STATIC closure doesn't really have
203 * a mut_link field, but we pretend it has - in fact we re-use
204 * the STATIC_LINK field for the time being, because when we
205 * come to do a major GC we won't need the mut_link field
206 * any more and can use it as a STATIC_LINK.
208 ACQUIRE_LOCK(&sm_mutex);
209 ((StgMutClosure *)caf)->mut_link = oldest_gen->mut_once_list;
210 oldest_gen->mut_once_list = (StgMutClosure *)caf;
214 const StgInfoTable *info;
216 info = get_itbl(caf);
217 ASSERT(info->type == IND_STATIC);
219 STATIC_LINK2(info,caf) = caf_list;
224 RELEASE_LOCK(&sm_mutex);
227 /* -----------------------------------------------------------------------------
229 -------------------------------------------------------------------------- */
232 allocNurseries( void )
241 for (cap = free_capabilities; cap != NULL; cap = cap->link) {
242 cap->rNursery = allocNursery(NULL, RtsFlags.GcFlags.minAllocAreaSize);
243 cap->rCurrentNursery = cap->rNursery;
244 for (bd = cap->rNursery; bd != NULL; bd = bd->link) {
245 bd->back = (bdescr *)cap;
248 /* Set the back links to be equal to the Capability,
249 * so we can do slightly better informed locking.
253 nursery_blocks = RtsFlags.GcFlags.minAllocAreaSize;
254 g0s0->blocks = allocNursery(NULL, nursery_blocks);
255 g0s0->n_blocks = nursery_blocks;
256 g0s0->to_space = NULL;
257 MainRegTable.rNursery = g0s0->blocks;
258 MainRegTable.rCurrentNursery = g0s0->blocks;
259 /* hp, hpLim, hp_bd, to_space etc. aren't used in G0S0 */
264 resetNurseries( void )
270 /* All tasks must be stopped */
271 ASSERT(n_free_capabilities == RtsFlags.ParFlags.nNodes);
273 for (cap = free_capabilities; cap != NULL; cap = cap->link) {
274 for (bd = cap->rNursery; bd; bd = bd->link) {
275 bd->free = bd->start;
276 ASSERT(bd->gen == g0);
277 ASSERT(bd->step == g0s0);
278 IF_DEBUG(sanity,memset(bd->start, 0xaa, BLOCK_SIZE));
280 cap->rCurrentNursery = cap->rNursery;
283 for (bd = g0s0->blocks; bd; bd = bd->link) {
284 bd->free = bd->start;
285 ASSERT(bd->gen == g0);
286 ASSERT(bd->step == g0s0);
287 IF_DEBUG(sanity,memset(bd->start, 0xaa, BLOCK_SIZE));
289 MainRegTable.rNursery = g0s0->blocks;
290 MainRegTable.rCurrentNursery = g0s0->blocks;
295 allocNursery (bdescr *last_bd, nat blocks)
300 /* Allocate a nursery */
301 for (i=0; i < blocks; i++) {
307 bd->free = bd->start;
314 resizeNursery ( nat blocks )
319 barf("resizeNursery: can't resize in SMP mode");
322 if (nursery_blocks == blocks) {
323 ASSERT(g0s0->n_blocks == blocks);
327 else if (nursery_blocks < blocks) {
328 IF_DEBUG(gc, fprintf(stderr, "Increasing size of nursery to %d blocks\n",
330 g0s0->blocks = allocNursery(g0s0->blocks, blocks-nursery_blocks);
336 IF_DEBUG(gc, fprintf(stderr, "Decreasing size of nursery to %d blocks\n",
338 for (bd = g0s0->blocks; nursery_blocks > blocks; nursery_blocks--) {
346 g0s0->n_blocks = nursery_blocks = blocks;
349 /* -----------------------------------------------------------------------------
350 The allocate() interface
352 allocate(n) always succeeds, and returns a chunk of memory n words
353 long. n can be larger than the size of a block if necessary, in
354 which case a contiguous block group will be allocated.
355 -------------------------------------------------------------------------- */
363 ACQUIRE_LOCK(&sm_mutex);
365 TICK_ALLOC_HEAP_NOCTR(n);
368 /* big allocation (>LARGE_OBJECT_THRESHOLD) */
369 /* ToDo: allocate directly into generation 1 */
370 if (n >= LARGE_OBJECT_THRESHOLD/sizeof(W_)) {
371 nat req_blocks = (lnat)BLOCK_ROUND_UP(n*sizeof(W_)) / BLOCK_SIZE;
372 bd = allocGroup(req_blocks);
373 dbl_link_onto(bd, &g0s0->large_objects);
377 bd->free = bd->start;
378 /* don't add these blocks to alloc_blocks, since we're assuming
379 * that large objects are likely to remain live for quite a while
380 * (eg. running threads), so garbage collecting early won't make
383 RELEASE_LOCK(&sm_mutex);
386 /* small allocation (<LARGE_OBJECT_THRESHOLD) */
387 } else if (small_alloc_list == NULL || alloc_Hp + n > alloc_HpLim) {
388 if (small_alloc_list) {
389 small_alloc_list->free = alloc_Hp;
392 bd->link = small_alloc_list;
393 small_alloc_list = bd;
397 alloc_Hp = bd->start;
398 alloc_HpLim = bd->start + BLOCK_SIZE_W;
404 RELEASE_LOCK(&sm_mutex);
408 lnat allocated_bytes(void)
410 return (alloc_blocks * BLOCK_SIZE_W - (alloc_HpLim - alloc_Hp));
413 /* -----------------------------------------------------------------------------
414 Allocation functions for GMP.
416 These all use the allocate() interface - we can't have any garbage
417 collection going on during a gmp operation, so we use allocate()
418 which always succeeds. The gmp operations which might need to
419 allocate will ask the storage manager (via doYouWantToGC()) whether
420 a garbage collection is required, in case we get into a loop doing
421 only allocate() style allocation.
422 -------------------------------------------------------------------------- */
425 stgAllocForGMP (size_t size_in_bytes)
428 nat data_size_in_words, total_size_in_words;
430 /* should be a multiple of sizeof(StgWord) (whole no. of limbs) */
431 ASSERT(size_in_bytes % sizeof(W_) == 0);
433 data_size_in_words = size_in_bytes / sizeof(W_);
434 total_size_in_words = sizeofW(StgArrWords) + data_size_in_words;
436 /* allocate and fill it in. */
437 arr = (StgArrWords *)allocate(total_size_in_words);
438 SET_ARR_HDR(arr, &ARR_WORDS_info, CCCS, data_size_in_words);
440 /* and return a ptr to the goods inside the array */
441 return(BYTE_ARR_CTS(arr));
445 stgReallocForGMP (void *ptr, size_t old_size, size_t new_size)
447 void *new_stuff_ptr = stgAllocForGMP(new_size);
449 char *p = (char *) ptr;
450 char *q = (char *) new_stuff_ptr;
452 for (; i < old_size; i++, p++, q++) {
456 return(new_stuff_ptr);
460 stgDeallocForGMP (void *ptr STG_UNUSED,
461 size_t size STG_UNUSED)
463 /* easy for us: the garbage collector does the dealloc'n */
466 /* -----------------------------------------------------------------------------
468 * -------------------------------------------------------------------------- */
470 /* -----------------------------------------------------------------------------
473 * Approximate how much we've allocated: number of blocks in the
474 * nursery + blocks allocated via allocate() - unused nusery blocks.
475 * This leaves a little slop at the end of each block, and doesn't
476 * take into account large objects (ToDo).
477 * -------------------------------------------------------------------------- */
480 calcAllocated( void )
488 /* All tasks must be stopped. Can't assert that all the
489 capabilities are owned by the scheduler, though: one or more
490 tasks might have been stopped while they were running (non-main)
492 /* ASSERT(n_free_capabilities == RtsFlags.ParFlags.nNodes); */
495 n_free_capabilities * RtsFlags.GcFlags.minAllocAreaSize * BLOCK_SIZE_W
498 for (cap = free_capabilities; cap != NULL; cap = cap->link) {
499 for ( bd = cap->rCurrentNursery->link; bd != NULL; bd = bd->link ) {
500 allocated -= BLOCK_SIZE_W;
502 if (cap->rCurrentNursery->free < cap->rCurrentNursery->start
504 allocated -= (cap->rCurrentNursery->start + BLOCK_SIZE_W)
505 - cap->rCurrentNursery->free;
510 bdescr *current_nursery = MainRegTable.rCurrentNursery;
512 allocated = (nursery_blocks * BLOCK_SIZE_W) + allocated_bytes();
513 for ( bd = current_nursery->link; bd != NULL; bd = bd->link ) {
514 allocated -= BLOCK_SIZE_W;
516 if (current_nursery->free < current_nursery->start + BLOCK_SIZE_W) {
517 allocated -= (current_nursery->start + BLOCK_SIZE_W)
518 - current_nursery->free;
525 /* Approximate the amount of live data in the heap. To be called just
526 * after garbage collection (see GarbageCollect()).
535 if (RtsFlags.GcFlags.generations == 1) {
536 live = (g0s0->to_blocks - 1) * BLOCK_SIZE_W +
537 ((lnat)g0s0->hp_bd->free - (lnat)g0s0->hp_bd->start) / sizeof(W_);
541 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
542 for (s = 0; s < generations[g].n_steps; s++) {
543 /* approximate amount of live data (doesn't take into account slop
544 * at end of each block).
546 if (g == 0 && s == 0) {
549 step = &generations[g].steps[s];
550 live += (step->n_blocks - 1) * BLOCK_SIZE_W +
551 ((lnat)step->hp_bd->free - (lnat)step->hp_bd->start) / sizeof(W_);
557 /* Approximate the number of blocks that will be needed at the next
558 * garbage collection.
560 * Assume: all data currently live will remain live. Steps that will
561 * be collected next time will therefore need twice as many blocks
562 * since all the data will be copied.
571 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
572 for (s = 0; s < generations[g].n_steps; s++) {
573 if (g == 0 && s == 0) { continue; }
574 step = &generations[g].steps[s];
575 if (generations[g].steps[0].n_blocks > generations[g].max_blocks) {
576 needed += 2 * step->n_blocks;
578 needed += step->n_blocks;
585 /* -----------------------------------------------------------------------------
588 memInventory() checks for memory leaks by counting up all the
589 blocks we know about and comparing that to the number of blocks
590 allegedly floating around in the system.
591 -------------------------------------------------------------------------- */
601 lnat total_blocks = 0, free_blocks = 0;
603 /* count the blocks we current have */
605 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
606 for (s = 0; s < generations[g].n_steps; s++) {
607 step = &generations[g].steps[s];
608 total_blocks += step->n_blocks;
609 if (RtsFlags.GcFlags.generations == 1) {
610 /* two-space collector has a to-space too :-) */
611 total_blocks += g0s0->to_blocks;
613 for (bd = step->large_objects; bd; bd = bd->link) {
614 total_blocks += bd->blocks;
615 /* hack for megablock groups: they have an extra block or two in
616 the second and subsequent megablocks where the block
617 descriptors would normally go.
619 if (bd->blocks > BLOCKS_PER_MBLOCK) {
620 total_blocks -= (MBLOCK_SIZE / BLOCK_SIZE - BLOCKS_PER_MBLOCK)
621 * (bd->blocks/(MBLOCK_SIZE/BLOCK_SIZE));
627 /* any blocks held by allocate() */
628 for (bd = small_alloc_list; bd; bd = bd->link) {
629 total_blocks += bd->blocks;
631 for (bd = large_alloc_list; bd; bd = bd->link) {
632 total_blocks += bd->blocks;
635 /* count the blocks on the free list */
636 free_blocks = countFreeList();
638 ASSERT(total_blocks + free_blocks == mblocks_allocated * BLOCKS_PER_MBLOCK);
641 if (total_blocks + free_blocks != mblocks_allocated *
643 fprintf(stderr, "Blocks: %ld live + %ld free = %ld total (%ld around)\n",
644 total_blocks, free_blocks, total_blocks + free_blocks,
645 mblocks_allocated * BLOCKS_PER_MBLOCK);
650 /* Full heap sanity check. */
657 if (RtsFlags.GcFlags.generations == 1) {
658 checkHeap(g0s0->to_space, NULL);
659 checkChain(g0s0->large_objects);
662 for (g = 0; g <= N; g++) {
663 for (s = 0; s < generations[g].n_steps; s++) {
664 if (g == 0 && s == 0) { continue; }
665 checkHeap(generations[g].steps[s].blocks, NULL);
668 for (g = N+1; g < RtsFlags.GcFlags.generations; g++) {
669 for (s = 0; s < generations[g].n_steps; s++) {
670 checkHeap(generations[g].steps[s].blocks,
671 generations[g].steps[s].blocks->start);
672 checkChain(generations[g].steps[s].large_objects);
675 checkFreeListSanity();