1 /* -----------------------------------------------------------------------------
3 * (c) The GHC Team, 1998-2004
5 * Storage manager front end
7 * ---------------------------------------------------------------------------*/
9 #include "PosixSource.h"
15 #include "BlockAlloc.h"
20 #include "OSThreads.h"
21 #include "Capability.h"
24 #include "RetainerProfile.h" // for counting memory blocks (memInventory)
30 * All these globals require sm_mutex to access in THREADED_RTS mode.
32 StgClosure *caf_list = NULL;
33 StgClosure *revertible_caf_list = NULL;
36 bdescr *small_alloc_list; /* allocate()d small objects */
37 bdescr *pinned_object_block; /* allocate pinned objects into this block */
38 nat alloc_blocks; /* number of allocate()d blocks since GC */
39 nat alloc_blocks_lim; /* approximate limit on alloc_blocks */
41 StgPtr alloc_Hp = NULL; /* next free byte in small_alloc_list */
42 StgPtr alloc_HpLim = NULL; /* end of block at small_alloc_list */
44 generation *generations = NULL; /* all the generations */
45 generation *g0 = NULL; /* generation 0, for convenience */
46 generation *oldest_gen = NULL; /* oldest generation, for convenience */
47 step *g0s0 = NULL; /* generation 0, step 0, for convenience */
49 ullong total_allocated = 0; /* total memory allocated during run */
51 nat n_nurseries = 0; /* == RtsFlags.ParFlags.nNodes, convenience */
52 step *nurseries = NULL; /* array of nurseries, >1 only if THREADED_RTS */
56 * Storage manager mutex: protects all the above state from
57 * simultaneous access by two STG threads.
61 * This mutex is used by atomicModifyMutVar# only
63 Mutex atomic_modify_mutvar_mutex;
70 static void *stgAllocForGMP (size_t size_in_bytes);
71 static void *stgReallocForGMP (void *ptr, size_t old_size, size_t new_size);
72 static void stgDeallocForGMP (void *ptr, size_t size);
75 initStep (step *stp, int g, int s)
80 stp->old_blocks = NULL;
81 stp->n_old_blocks = 0;
82 stp->gen = &generations[g];
88 stp->scavd_hpLim = NULL;
91 stp->large_objects = NULL;
92 stp->n_large_blocks = 0;
93 stp->new_large_objects = NULL;
94 stp->scavenged_large_objects = NULL;
95 stp->n_scavenged_large_blocks = 0;
96 stp->is_compacted = 0;
106 if (generations != NULL) {
107 // multi-init protection
111 /* Sanity check to make sure the LOOKS_LIKE_ macros appear to be
112 * doing something reasonable.
114 ASSERT(LOOKS_LIKE_INFO_PTR(&stg_BLACKHOLE_info));
115 ASSERT(LOOKS_LIKE_CLOSURE_PTR(&stg_dummy_ret_closure));
116 ASSERT(!HEAP_ALLOCED(&stg_dummy_ret_closure));
118 if (RtsFlags.GcFlags.maxHeapSize != 0 &&
119 RtsFlags.GcFlags.heapSizeSuggestion >
120 RtsFlags.GcFlags.maxHeapSize) {
121 RtsFlags.GcFlags.maxHeapSize = RtsFlags.GcFlags.heapSizeSuggestion;
124 if (RtsFlags.GcFlags.maxHeapSize != 0 &&
125 RtsFlags.GcFlags.minAllocAreaSize >
126 RtsFlags.GcFlags.maxHeapSize) {
127 errorBelch("maximum heap size (-M) is smaller than minimum alloc area size (-A)");
131 initBlockAllocator();
133 #if defined(THREADED_RTS)
134 initMutex(&sm_mutex);
135 initMutex(&atomic_modify_mutvar_mutex);
140 /* allocate generation info array */
141 generations = (generation *)stgMallocBytes(RtsFlags.GcFlags.generations
142 * sizeof(struct generation_),
143 "initStorage: gens");
145 /* Initialise all generations */
146 for(g = 0; g < RtsFlags.GcFlags.generations; g++) {
147 gen = &generations[g];
149 gen->mut_list = allocBlock();
150 gen->collections = 0;
151 gen->failed_promotions = 0;
155 /* A couple of convenience pointers */
156 g0 = &generations[0];
157 oldest_gen = &generations[RtsFlags.GcFlags.generations-1];
159 /* Allocate step structures in each generation */
160 if (RtsFlags.GcFlags.generations > 1) {
161 /* Only for multiple-generations */
163 /* Oldest generation: one step */
164 oldest_gen->n_steps = 1;
166 stgMallocBytes(1 * sizeof(struct step_), "initStorage: last step");
168 /* set up all except the oldest generation with 2 steps */
169 for(g = 0; g < RtsFlags.GcFlags.generations-1; g++) {
170 generations[g].n_steps = RtsFlags.GcFlags.steps;
171 generations[g].steps =
172 stgMallocBytes (RtsFlags.GcFlags.steps * sizeof(struct step_),
173 "initStorage: steps");
177 /* single generation, i.e. a two-space collector */
179 g0->steps = stgMallocBytes (sizeof(struct step_), "initStorage: steps");
183 n_nurseries = n_capabilities;
184 nurseries = stgMallocBytes (n_nurseries * sizeof(struct step_),
185 "initStorage: nurseries");
188 nurseries = g0->steps; // just share nurseries[0] with g0s0
191 /* Initialise all steps */
192 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
193 for (s = 0; s < generations[g].n_steps; s++) {
194 initStep(&generations[g].steps[s], g, s);
199 for (s = 0; s < n_nurseries; s++) {
200 initStep(&nurseries[s], 0, s);
204 /* Set up the destination pointers in each younger gen. step */
205 for (g = 0; g < RtsFlags.GcFlags.generations-1; g++) {
206 for (s = 0; s < generations[g].n_steps-1; s++) {
207 generations[g].steps[s].to = &generations[g].steps[s+1];
209 generations[g].steps[s].to = &generations[g+1].steps[0];
211 oldest_gen->steps[0].to = &oldest_gen->steps[0];
214 for (s = 0; s < n_nurseries; s++) {
215 nurseries[s].to = generations[0].steps[0].to;
219 /* The oldest generation has one step. */
220 if (RtsFlags.GcFlags.compact) {
221 if (RtsFlags.GcFlags.generations == 1) {
222 errorBelch("WARNING: compaction is incompatible with -G1; disabled");
224 oldest_gen->steps[0].is_compacted = 1;
229 if (RtsFlags.GcFlags.generations == 1) {
230 errorBelch("-G1 is incompatible with -threaded");
231 stg_exit(EXIT_FAILURE);
235 /* generation 0 is special: that's the nursery */
236 generations[0].max_blocks = 0;
238 /* G0S0: the allocation area. Policy: keep the allocation area
239 * small to begin with, even if we have a large suggested heap
240 * size. Reason: we're going to do a major collection first, and we
241 * don't want it to be a big one. This vague idea is borne out by
242 * rigorous experimental evidence.
244 g0s0 = &generations[0].steps[0];
248 weak_ptr_list = NULL;
250 revertible_caf_list = NULL;
252 /* initialise the allocate() interface */
253 small_alloc_list = NULL;
255 alloc_blocks_lim = RtsFlags.GcFlags.minAllocAreaSize;
257 /* Tell GNU multi-precision pkg about our custom alloc functions */
258 mp_set_memory_functions(stgAllocForGMP, stgReallocForGMP, stgDeallocForGMP);
260 IF_DEBUG(gc, statDescribeGens());
268 stat_exit(calcAllocated());
272 /* -----------------------------------------------------------------------------
275 The entry code for every CAF does the following:
277 - builds a CAF_BLACKHOLE in the heap
278 - pushes an update frame pointing to the CAF_BLACKHOLE
279 - invokes UPD_CAF(), which:
280 - calls newCaf, below
281 - updates the CAF with a static indirection to the CAF_BLACKHOLE
283 Why do we build a BLACKHOLE in the heap rather than just updating
284 the thunk directly? It's so that we only need one kind of update
285 frame - otherwise we'd need a static version of the update frame too.
287 newCaf() does the following:
289 - it puts the CAF on the oldest generation's mut-once list.
290 This is so that we can treat the CAF as a root when collecting
293 For GHCI, we have additional requirements when dealing with CAFs:
295 - we must *retain* all dynamically-loaded CAFs ever entered,
296 just in case we need them again.
297 - we must be able to *revert* CAFs that have been evaluated, to
298 their pre-evaluated form.
300 To do this, we use an additional CAF list. When newCaf() is
301 called on a dynamically-loaded CAF, we add it to the CAF list
302 instead of the old-generation mutable list, and save away its
303 old info pointer (in caf->saved_info) for later reversion.
305 To revert all the CAFs, we traverse the CAF list and reset the
306 info pointer to caf->saved_info, then throw away the CAF list.
307 (see GC.c:revertCAFs()).
311 -------------------------------------------------------------------------- */
314 newCAF(StgClosure* caf)
321 // If we are in GHCi _and_ we are using dynamic libraries,
322 // then we can't redirect newCAF calls to newDynCAF (see below),
323 // so we make newCAF behave almost like newDynCAF.
324 // The dynamic libraries might be used by both the interpreted
325 // program and GHCi itself, so they must not be reverted.
326 // This also means that in GHCi with dynamic libraries, CAFs are not
327 // garbage collected. If this turns out to be a problem, we could
328 // do another hack here and do an address range test on caf to figure
329 // out whether it is from a dynamic library.
330 ((StgIndStatic *)caf)->saved_info = (StgInfoTable *)caf->header.info;
331 ((StgIndStatic *)caf)->static_link = caf_list;
336 /* Put this CAF on the mutable list for the old generation.
337 * This is a HACK - the IND_STATIC closure doesn't really have
338 * a mut_link field, but we pretend it has - in fact we re-use
339 * the STATIC_LINK field for the time being, because when we
340 * come to do a major GC we won't need the mut_link field
341 * any more and can use it as a STATIC_LINK.
343 ((StgIndStatic *)caf)->saved_info = NULL;
344 recordMutableGen(caf, oldest_gen);
350 /* If we are PAR or DIST then we never forget a CAF */
352 //debugBelch("<##> Globalising CAF %08x %s",caf,info_type(caf));
353 newGA=makeGlobal(caf,rtsTrue); /*given full weight*/
359 // An alternate version of newCaf which is used for dynamically loaded
360 // object code in GHCi. In this case we want to retain *all* CAFs in
361 // the object code, because they might be demanded at any time from an
362 // expression evaluated on the command line.
363 // Also, GHCi might want to revert CAFs, so we add these to the
364 // revertible_caf_list.
366 // The linker hackily arranges that references to newCaf from dynamic
367 // code end up pointing to newDynCAF.
369 newDynCAF(StgClosure *caf)
373 ((StgIndStatic *)caf)->saved_info = (StgInfoTable *)caf->header.info;
374 ((StgIndStatic *)caf)->static_link = revertible_caf_list;
375 revertible_caf_list = caf;
380 /* -----------------------------------------------------------------------------
382 -------------------------------------------------------------------------- */
385 allocNursery (step *stp, bdescr *tail, nat blocks)
390 // Allocate a nursery: we allocate fresh blocks one at a time and
391 // cons them on to the front of the list, not forgetting to update
392 // the back pointer on the tail of the list to point to the new block.
393 for (i=0; i < blocks; i++) {
396 processNursery() in LdvProfile.c assumes that every block group in
397 the nursery contains only a single block. So, if a block group is
398 given multiple blocks, change processNursery() accordingly.
402 // double-link the nursery: we might need to insert blocks
409 bd->free = bd->start;
417 assignNurseriesToCapabilities (void)
422 for (i = 0; i < n_nurseries; i++) {
423 capabilities[i].r.rNursery = &nurseries[i];
424 capabilities[i].r.rCurrentNursery = nurseries[i].blocks;
425 capabilities[i].r.rCurrentAlloc = NULL;
427 #else /* THREADED_RTS */
428 MainCapability.r.rNursery = &nurseries[0];
429 MainCapability.r.rCurrentNursery = nurseries[0].blocks;
430 MainCapability.r.rCurrentAlloc = NULL;
435 allocNurseries( void )
439 for (i = 0; i < n_nurseries; i++) {
440 nurseries[i].blocks =
441 allocNursery(&nurseries[i], NULL,
442 RtsFlags.GcFlags.minAllocAreaSize);
443 nurseries[i].n_blocks = RtsFlags.GcFlags.minAllocAreaSize;
444 nurseries[i].old_blocks = NULL;
445 nurseries[i].n_old_blocks = 0;
446 /* hp, hpLim, hp_bd, to_space etc. aren't used in the nursery */
448 assignNurseriesToCapabilities();
452 resetNurseries( void )
458 for (i = 0; i < n_nurseries; i++) {
460 for (bd = stp->blocks; bd; bd = bd->link) {
461 bd->free = bd->start;
462 ASSERT(bd->gen_no == 0);
463 ASSERT(bd->step == stp);
464 IF_DEBUG(sanity,memset(bd->start, 0xaa, BLOCK_SIZE));
467 assignNurseriesToCapabilities();
471 countNurseryBlocks (void)
476 for (i = 0; i < n_nurseries; i++) {
477 blocks += nurseries[i].n_blocks;
483 resizeNursery ( step *stp, nat blocks )
488 nursery_blocks = stp->n_blocks;
489 if (nursery_blocks == blocks) return;
491 if (nursery_blocks < blocks) {
492 IF_DEBUG(gc, debugBelch("Increasing size of nursery to %d blocks\n",
494 stp->blocks = allocNursery(stp, stp->blocks, blocks-nursery_blocks);
499 IF_DEBUG(gc, debugBelch("Decreasing size of nursery to %d blocks\n",
503 while (nursery_blocks > blocks) {
505 next_bd->u.back = NULL;
506 nursery_blocks -= bd->blocks; // might be a large block
511 // might have gone just under, by freeing a large block, so make
512 // up the difference.
513 if (nursery_blocks < blocks) {
514 stp->blocks = allocNursery(stp, stp->blocks, blocks-nursery_blocks);
518 stp->n_blocks = blocks;
519 ASSERT(countBlocks(stp->blocks) == stp->n_blocks);
523 // Resize each of the nurseries to the specified size.
526 resizeNurseriesFixed (nat blocks)
529 for (i = 0; i < n_nurseries; i++) {
530 resizeNursery(&nurseries[i], blocks);
535 // Resize the nurseries to the total specified size.
538 resizeNurseries (nat blocks)
540 // If there are multiple nurseries, then we just divide the number
541 // of available blocks between them.
542 resizeNurseriesFixed(blocks / n_nurseries);
545 /* -----------------------------------------------------------------------------
546 The allocate() interface
548 allocate(n) always succeeds, and returns a chunk of memory n words
549 long. n can be larger than the size of a block if necessary, in
550 which case a contiguous block group will be allocated.
551 -------------------------------------------------------------------------- */
561 TICK_ALLOC_HEAP_NOCTR(n);
564 /* big allocation (>LARGE_OBJECT_THRESHOLD) */
565 /* ToDo: allocate directly into generation 1 */
566 if (n >= LARGE_OBJECT_THRESHOLD/sizeof(W_)) {
567 nat req_blocks = (lnat)BLOCK_ROUND_UP(n*sizeof(W_)) / BLOCK_SIZE;
568 bd = allocGroup(req_blocks);
569 dbl_link_onto(bd, &g0s0->large_objects);
570 g0s0->n_large_blocks += req_blocks;
573 bd->flags = BF_LARGE;
574 bd->free = bd->start + n;
575 alloc_blocks += req_blocks;
579 /* small allocation (<LARGE_OBJECT_THRESHOLD) */
580 } else if (small_alloc_list == NULL || alloc_Hp + n > alloc_HpLim) {
581 if (small_alloc_list) {
582 small_alloc_list->free = alloc_Hp;
585 bd->link = small_alloc_list;
586 small_alloc_list = bd;
590 alloc_Hp = bd->start;
591 alloc_HpLim = bd->start + BLOCK_SIZE_W;
602 allocated_bytes( void )
606 allocated = alloc_blocks * BLOCK_SIZE_W - (alloc_HpLim - alloc_Hp);
607 if (pinned_object_block != NULL) {
608 allocated -= (pinned_object_block->start + BLOCK_SIZE_W) -
609 pinned_object_block->free;
616 tidyAllocateLists (void)
618 if (small_alloc_list != NULL) {
619 ASSERT(alloc_Hp >= small_alloc_list->start &&
620 alloc_Hp <= small_alloc_list->start + BLOCK_SIZE);
621 small_alloc_list->free = alloc_Hp;
625 /* -----------------------------------------------------------------------------
628 This allocates memory in the current thread - it is intended for
629 use primarily from STG-land where we have a Capability. It is
630 better than allocate() because it doesn't require taking the
631 sm_mutex lock in the common case.
633 Memory is allocated directly from the nursery if possible (but not
634 from the current nursery block, so as not to interfere with
636 -------------------------------------------------------------------------- */
639 allocateLocal (Capability *cap, nat n)
644 TICK_ALLOC_HEAP_NOCTR(n);
647 /* big allocation (>LARGE_OBJECT_THRESHOLD) */
648 /* ToDo: allocate directly into generation 1 */
649 if (n >= LARGE_OBJECT_THRESHOLD/sizeof(W_)) {
650 nat req_blocks = (lnat)BLOCK_ROUND_UP(n*sizeof(W_)) / BLOCK_SIZE;
652 bd = allocGroup(req_blocks);
653 dbl_link_onto(bd, &g0s0->large_objects);
654 g0s0->n_large_blocks += req_blocks;
657 bd->flags = BF_LARGE;
658 bd->free = bd->start + n;
659 alloc_blocks += req_blocks;
663 /* small allocation (<LARGE_OBJECT_THRESHOLD) */
666 bd = cap->r.rCurrentAlloc;
667 if (bd == NULL || bd->free + n > bd->start + BLOCK_SIZE_W) {
669 // The CurrentAlloc block is full, we need to find another
670 // one. First, we try taking the next block from the
672 bd = cap->r.rCurrentNursery->link;
674 if (bd == NULL || bd->free + n > bd->start + BLOCK_SIZE_W) {
675 // The nursery is empty, or the next block is already
676 // full: allocate a fresh block (we can't fail here).
679 cap->r.rNursery->n_blocks++;
682 bd->step = cap->r.rNursery;
685 // we have a block in the nursery: take it and put
686 // it at the *front* of the nursery list, and use it
687 // to allocate() from.
688 cap->r.rCurrentNursery->link = bd->link;
689 if (bd->link != NULL) {
690 bd->link->u.back = cap->r.rCurrentNursery;
693 dbl_link_onto(bd, &cap->r.rNursery->blocks);
694 cap->r.rCurrentAlloc = bd;
695 IF_DEBUG(sanity, checkNurserySanity(cap->r.rNursery));
703 /* ---------------------------------------------------------------------------
704 Allocate a fixed/pinned object.
706 We allocate small pinned objects into a single block, allocating a
707 new block when the current one overflows. The block is chained
708 onto the large_object_list of generation 0 step 0.
710 NOTE: The GC can't in general handle pinned objects. This
711 interface is only safe to use for ByteArrays, which have no
712 pointers and don't require scavenging. It works because the
713 block's descriptor has the BF_LARGE flag set, so the block is
714 treated as a large object and chained onto various lists, rather
715 than the individual objects being copied. However, when it comes
716 to scavenge the block, the GC will only scavenge the first object.
717 The reason is that the GC can't linearly scan a block of pinned
718 objects at the moment (doing so would require using the
719 mostly-copying techniques). But since we're restricting ourselves
720 to pinned ByteArrays, not scavenging is ok.
722 This function is called by newPinnedByteArray# which immediately
723 fills the allocated memory with a MutableByteArray#.
724 ------------------------------------------------------------------------- */
727 allocatePinned( nat n )
730 bdescr *bd = pinned_object_block;
732 // If the request is for a large object, then allocate()
733 // will give us a pinned object anyway.
734 if (n >= LARGE_OBJECT_THRESHOLD/sizeof(W_)) {
740 TICK_ALLOC_HEAP_NOCTR(n);
743 // we always return 8-byte aligned memory. bd->free must be
744 // 8-byte aligned to begin with, so we just round up n to
745 // the nearest multiple of 8 bytes.
746 if (sizeof(StgWord) == 4) {
750 // If we don't have a block of pinned objects yet, or the current
751 // one isn't large enough to hold the new object, allocate a new one.
752 if (bd == NULL || (bd->free + n) > (bd->start + BLOCK_SIZE_W)) {
753 pinned_object_block = bd = allocBlock();
754 dbl_link_onto(bd, &g0s0->large_objects);
757 bd->flags = BF_PINNED | BF_LARGE;
758 bd->free = bd->start;
768 /* -----------------------------------------------------------------------------
769 This is the write barrier for MUT_VARs, a.k.a. IORefs. A
770 MUT_VAR_CLEAN object is not on the mutable list; a MUT_VAR_DIRTY
771 is. When written to, a MUT_VAR_CLEAN turns into a MUT_VAR_DIRTY
772 and is put on the mutable list.
773 -------------------------------------------------------------------------- */
776 dirty_MUT_VAR(StgRegTable *reg, StgClosure *p)
778 Capability *cap = regTableToCapability(reg);
780 if (p->header.info == &stg_MUT_VAR_CLEAN_info) {
781 p->header.info = &stg_MUT_VAR_DIRTY_info;
782 bd = Bdescr((StgPtr)p);
783 if (bd->gen_no > 0) recordMutableCap(p,cap,bd->gen_no);
787 /* -----------------------------------------------------------------------------
788 Allocation functions for GMP.
790 These all use the allocate() interface - we can't have any garbage
791 collection going on during a gmp operation, so we use allocate()
792 which always succeeds. The gmp operations which might need to
793 allocate will ask the storage manager (via doYouWantToGC()) whether
794 a garbage collection is required, in case we get into a loop doing
795 only allocate() style allocation.
796 -------------------------------------------------------------------------- */
799 stgAllocForGMP (size_t size_in_bytes)
802 nat data_size_in_words, total_size_in_words;
804 /* round up to a whole number of words */
805 data_size_in_words = (size_in_bytes + sizeof(W_) + 1) / sizeof(W_);
806 total_size_in_words = sizeofW(StgArrWords) + data_size_in_words;
808 /* allocate and fill it in. */
809 #if defined(THREADED_RTS)
810 arr = (StgArrWords *)allocateLocal(myTask()->cap, total_size_in_words);
812 arr = (StgArrWords *)allocateLocal(&MainCapability, total_size_in_words);
814 SET_ARR_HDR(arr, &stg_ARR_WORDS_info, CCCS, data_size_in_words);
816 /* and return a ptr to the goods inside the array */
821 stgReallocForGMP (void *ptr, size_t old_size, size_t new_size)
823 void *new_stuff_ptr = stgAllocForGMP(new_size);
825 char *p = (char *) ptr;
826 char *q = (char *) new_stuff_ptr;
828 for (; i < old_size; i++, p++, q++) {
832 return(new_stuff_ptr);
836 stgDeallocForGMP (void *ptr STG_UNUSED,
837 size_t size STG_UNUSED)
839 /* easy for us: the garbage collector does the dealloc'n */
842 /* -----------------------------------------------------------------------------
844 * -------------------------------------------------------------------------- */
846 /* -----------------------------------------------------------------------------
849 * Approximate how much we've allocated: number of blocks in the
850 * nursery + blocks allocated via allocate() - unused nusery blocks.
851 * This leaves a little slop at the end of each block, and doesn't
852 * take into account large objects (ToDo).
853 * -------------------------------------------------------------------------- */
856 calcAllocated( void )
861 allocated = allocated_bytes();
862 allocated += countNurseryBlocks() * BLOCK_SIZE_W;
867 for (i = 0; i < n_nurseries; i++) {
869 for ( bd = capabilities[i].r.rCurrentNursery->link;
870 bd != NULL; bd = bd->link ) {
871 allocated -= BLOCK_SIZE_W;
873 cap = &capabilities[i];
874 if (cap->r.rCurrentNursery->free <
875 cap->r.rCurrentNursery->start + BLOCK_SIZE_W) {
876 allocated -= (cap->r.rCurrentNursery->start + BLOCK_SIZE_W)
877 - cap->r.rCurrentNursery->free;
881 bdescr *current_nursery = MainCapability.r.rCurrentNursery;
883 for ( bd = current_nursery->link; bd != NULL; bd = bd->link ) {
884 allocated -= BLOCK_SIZE_W;
886 if (current_nursery->free < current_nursery->start + BLOCK_SIZE_W) {
887 allocated -= (current_nursery->start + BLOCK_SIZE_W)
888 - current_nursery->free;
893 total_allocated += allocated;
897 /* Approximate the amount of live data in the heap. To be called just
898 * after garbage collection (see GarbageCollect()).
907 if (RtsFlags.GcFlags.generations == 1) {
908 live = (g0s0->n_blocks - 1) * BLOCK_SIZE_W +
909 ((lnat)g0s0->hp_bd->free - (lnat)g0s0->hp_bd->start) / sizeof(W_);
913 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
914 for (s = 0; s < generations[g].n_steps; s++) {
915 /* approximate amount of live data (doesn't take into account slop
916 * at end of each block).
918 if (g == 0 && s == 0) {
921 stp = &generations[g].steps[s];
922 live += (stp->n_large_blocks + stp->n_blocks - 1) * BLOCK_SIZE_W;
923 if (stp->hp_bd != NULL) {
924 live += ((lnat)stp->hp_bd->free - (lnat)stp->hp_bd->start)
927 if (stp->scavd_hp != NULL) {
928 live -= (P_)(BLOCK_ROUND_UP(stp->scavd_hp)) - stp->scavd_hp;
935 /* Approximate the number of blocks that will be needed at the next
936 * garbage collection.
938 * Assume: all data currently live will remain live. Steps that will
939 * be collected next time will therefore need twice as many blocks
940 * since all the data will be copied.
949 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
950 for (s = 0; s < generations[g].n_steps; s++) {
951 if (g == 0 && s == 0) { continue; }
952 stp = &generations[g].steps[s];
953 if (generations[g].steps[0].n_blocks +
954 generations[g].steps[0].n_large_blocks
955 > generations[g].max_blocks
956 && stp->is_compacted == 0) {
957 needed += 2 * stp->n_blocks;
959 needed += stp->n_blocks;
966 /* -----------------------------------------------------------------------------
969 memInventory() checks for memory leaks by counting up all the
970 blocks we know about and comparing that to the number of blocks
971 allegedly floating around in the system.
972 -------------------------------------------------------------------------- */
977 stepBlocks (step *stp)
982 total_blocks = stp->n_blocks;
983 total_blocks += stp->n_old_blocks;
984 for (bd = stp->large_objects; bd; bd = bd->link) {
985 total_blocks += bd->blocks;
986 /* hack for megablock groups: they have an extra block or two in
987 the second and subsequent megablocks where the block
988 descriptors would normally go.
990 if (bd->blocks > BLOCKS_PER_MBLOCK) {
991 total_blocks -= (MBLOCK_SIZE / BLOCK_SIZE - BLOCKS_PER_MBLOCK)
992 * (bd->blocks/(MBLOCK_SIZE/BLOCK_SIZE));
1004 lnat total_blocks = 0, free_blocks = 0;
1006 /* count the blocks we current have */
1008 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
1009 for (i = 0; i < n_capabilities; i++) {
1010 for (bd = capabilities[i].mut_lists[g]; bd != NULL; bd = bd->link) {
1011 total_blocks += bd->blocks;
1014 for (bd = generations[g].mut_list; bd != NULL; bd = bd->link) {
1015 total_blocks += bd->blocks;
1017 for (s = 0; s < generations[g].n_steps; s++) {
1018 if (g==0 && s==0) continue;
1019 stp = &generations[g].steps[s];
1020 total_blocks += stepBlocks(stp);
1024 for (i = 0; i < n_nurseries; i++) {
1025 total_blocks += stepBlocks(&nurseries[i]);
1028 // We put pinned object blocks in g0s0, so better count blocks there too.
1029 total_blocks += stepBlocks(g0s0);
1032 /* any blocks held by allocate() */
1033 for (bd = small_alloc_list; bd; bd = bd->link) {
1034 total_blocks += bd->blocks;
1038 if (RtsFlags.ProfFlags.doHeapProfile == HEAP_BY_RETAINER) {
1039 total_blocks += retainerStackBlocks();
1043 // count the blocks allocated by the arena allocator
1044 total_blocks += arenaBlocks();
1046 /* count the blocks on the free list */
1047 free_blocks = countFreeList();
1049 if (total_blocks + free_blocks != mblocks_allocated *
1050 BLOCKS_PER_MBLOCK) {
1051 debugBelch("Blocks: %ld live + %ld free = %ld total (%ld around)\n",
1052 total_blocks, free_blocks, total_blocks + free_blocks,
1053 mblocks_allocated * BLOCKS_PER_MBLOCK);
1056 ASSERT(total_blocks + free_blocks == mblocks_allocated * BLOCKS_PER_MBLOCK);
1061 countBlocks(bdescr *bd)
1064 for (n=0; bd != NULL; bd=bd->link) {
1070 /* Full heap sanity check. */
1076 if (RtsFlags.GcFlags.generations == 1) {
1077 checkHeap(g0s0->blocks);
1078 checkChain(g0s0->large_objects);
1081 for (g = 0; g < RtsFlags.GcFlags.generations; g++) {
1082 for (s = 0; s < generations[g].n_steps; s++) {
1083 if (g == 0 && s == 0) { continue; }
1084 ASSERT(countBlocks(generations[g].steps[s].blocks)
1085 == generations[g].steps[s].n_blocks);
1086 ASSERT(countBlocks(generations[g].steps[s].large_objects)
1087 == generations[g].steps[s].n_large_blocks);
1088 checkHeap(generations[g].steps[s].blocks);
1089 checkChain(generations[g].steps[s].large_objects);
1091 checkMutableList(generations[g].mut_list, g);
1096 for (s = 0; s < n_nurseries; s++) {
1097 ASSERT(countBlocks(nurseries[s].blocks)
1098 == nurseries[s].n_blocks);
1099 ASSERT(countBlocks(nurseries[s].large_objects)
1100 == nurseries[s].n_large_blocks);
1103 checkFreeListSanity();
1107 /* Nursery sanity check */
1109 checkNurserySanity( step *stp )
1115 for (bd = stp->blocks; bd != NULL; bd = bd->link) {
1116 ASSERT(bd->u.back == prev);
1118 blocks += bd->blocks;
1120 ASSERT(blocks == stp->n_blocks);
1123 // handy function for use in gdb, because Bdescr() is inlined.
1124 extern bdescr *_bdescr( StgPtr p );