/* * Copyright (c) 1991 Regents of the University of California. * All rights reserved. * Copyright (c) 1994 John S. Dyson * All rights reserved. * Copyright (c) 1994 David Greenman * All rights reserved. * * This code is derived from software contributed to Berkeley by * The Mach Operating System project at Carnegie-Mellon University. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * from: @(#)vm_pageout.c 7.4 (Berkeley) 5/7/91 * * * Copyright (c) 1987, 1990 Carnegie-Mellon University. * All rights reserved. * * Authors: Avadis Tevanian, Jr., Michael Wayne Young * * Permission to use, copy, modify and distribute this software and * its documentation is hereby granted, provided that both the copyright * notice and this permission notice appear in all copies of the * software, derivative works or modified versions, and any portions * thereof, and that both notices appear in supporting documentation. * * CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS "AS IS" * CONDITION. CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND * FOR ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE. * * Carnegie Mellon requests users of this software to return to * * Software Distribution Coordinator or Software.Distribution@CS.CMU.EDU * School of Computer Science * Carnegie Mellon University * Pittsburgh PA 15213-3890 * * any improvements or extensions that they make and grant Carnegie the * rights to redistribute these changes. * * $Id: vm_pageout.c,v 1.19 1994/10/18 14:59:20 davidg Exp $ */ /* * The proverbial page-out daemon. */ #include #include #include #include #include #include #include #include #include extern vm_map_t kmem_map; int vm_pages_needed; /* Event on which pageout daemon sleeps */ int vm_pagescanner; /* Event on which pagescanner sleeps */ int vm_pageout_free_min = 0; /* Stop pageout to wait for pagers at this free level */ int vm_pageout_pages_needed = 0; /* flag saying that the pageout daemon needs pages */ int vm_page_pagesfreed; int vm_desired_cache_size; extern int npendingio; extern int hz; int vm_pageout_proc_limit; extern int nswiodone; extern int swap_pager_full; extern int vm_swap_size; extern int swap_pager_ready(); #define MAXREF 32767 #define MAXSCAN 512 /* maximum number of pages to scan in active queue */ /* set the "clock" hands to be (MAXSCAN * 4096) Bytes */ #define ACT_DECLINE 1 #define ACT_ADVANCE 3 #define ACT_MAX 100 #define LOWATER ((2048*1024)/NBPG) #define VM_PAGEOUT_PAGE_COUNT 8 int vm_pageout_page_count = VM_PAGEOUT_PAGE_COUNT; int vm_pageout_req_do_stats; int vm_page_max_wired = 0; /* XXX max # of wired pages system-wide */ /* * vm_pageout_clean: * cleans a vm_page */ int vm_pageout_clean(m, sync) register vm_page_t m; int sync; { /* * Clean the page and remove it from the * laundry. * * We set the busy bit to cause * potential page faults on this page to * block. * * And we set pageout-in-progress to keep * the object from disappearing during * pageout. This guarantees that the * page won't move from the inactive * queue. (However, any other page on * the inactive queue may move!) */ register vm_object_t object; register vm_pager_t pager; int pageout_status[VM_PAGEOUT_PAGE_COUNT]; vm_page_t ms[VM_PAGEOUT_PAGE_COUNT]; int pageout_count; int anyok=0; int i; vm_offset_t offset = m->offset; object = m->object; if (!object) { printf("pager: object missing\n"); return 0; } /* * Try to collapse the object before * making a pager for it. We must * unlock the page queues first. * We try to defer the creation of a pager * until all shadows are not paging. This * allows vm_object_collapse to work better and * helps control swap space size. * (J. Dyson 11 Nov 93) */ if (!object->pager && cnt.v_free_count < vm_pageout_free_min) return 0; if (!object->pager && object->shadow && object->shadow->paging_in_progress) return 0; if( !sync) { if (object->shadow) { vm_object_collapse(object); if (!vm_page_lookup(object, offset)) return 0; } if ((m->busy != 0) || (m->flags & PG_BUSY) || (m->hold_count != 0)) { return 0; } } pageout_count = 1; ms[0] = m; pager = object->pager; if (pager) { for (i = 1; i < vm_pageout_page_count; i++) { ms[i] = vm_page_lookup(object, offset+i*NBPG); if (ms[i]) { if (( ((ms[i]->flags & (PG_CLEAN|PG_INACTIVE|PG_BUSY)) == PG_INACTIVE) || ( (ms[i]->flags & (PG_CLEAN|PG_BUSY)) == 0 && sync == VM_PAGEOUT_FORCE)) && (ms[i]->wire_count == 0) && (ms[i]->busy == 0) && (ms[i]->hold_count == 0)) pageout_count++; else break; } else break; } for(i=0;iflags |= PG_BUSY; pmap_page_protect(VM_PAGE_TO_PHYS(ms[i]), VM_PROT_READ); } object->paging_in_progress += pageout_count; } else { m->flags |= PG_BUSY; pmap_page_protect(VM_PAGE_TO_PHYS(m), VM_PROT_READ); object->paging_in_progress++; pager = vm_pager_allocate(PG_DFLT, (caddr_t)0, object->size, VM_PROT_ALL, 0); if (pager != NULL) { vm_object_setpager(object, pager, 0, FALSE); } } /* * If there is no pager for the page, * use the default pager. If there's * no place to put the page at the * moment, leave it in the laundry and * hope that there will be paging space * later. */ if ((pager && pager->pg_type == PG_SWAP) || cnt.v_free_count >= vm_pageout_free_min) { if( pageout_count == 1) { pageout_status[0] = pager ? vm_pager_put(pager, m, ((sync || (object == kernel_object)) ? TRUE: FALSE)) : VM_PAGER_FAIL; } else { if( !pager) { for(i=0;iflags &= ~PG_LAUNDRY; ++anyok; break; case VM_PAGER_PEND: ms[i]->flags &= ~PG_LAUNDRY; ++anyok; break; case VM_PAGER_BAD: /* * Page outside of range of object. * Right now we essentially lose the * changes by pretending it worked. */ ms[i]->flags &= ~PG_LAUNDRY; ms[i]->flags |= PG_CLEAN; pmap_clear_modify(VM_PAGE_TO_PHYS(ms[i])); break; case VM_PAGER_ERROR: case VM_PAGER_FAIL: /* * If page couldn't be paged out, then * reactivate the page so it doesn't * clog the inactive list. (We will * try paging out it again later). */ if (ms[i]->flags & PG_INACTIVE) vm_page_activate(ms[i]); break; case VM_PAGER_AGAIN: break; } /* * If the operation is still going, leave * the page busy to block all other accesses. * Also, leave the paging in progress * indicator set so that we don't attempt an * object collapse. */ if (pageout_status[i] != VM_PAGER_PEND) { PAGE_WAKEUP(ms[i]); if (--object->paging_in_progress == 0) wakeup((caddr_t) object); if ((ms[i]->flags & PG_REFERENCED) || pmap_is_referenced(VM_PAGE_TO_PHYS(ms[i]))) { pmap_clear_reference(VM_PAGE_TO_PHYS(ms[i])); ms[i]->flags &= ~PG_REFERENCED; if( ms[i]->flags & PG_INACTIVE) vm_page_activate(ms[i]); } } } return anyok; } /* * vm_pageout_object_deactivate_pages * * deactivate enough pages to satisfy the inactive target * requirements or if vm_page_proc_limit is set, then * deactivate all of the pages in the object and its * shadows. * * The object and map must be locked. */ int vm_pageout_object_deactivate_pages(map, object, count) vm_map_t map; vm_object_t object; int count; { register vm_page_t p, next; int rcount; int dcount; dcount = 0; if (count == 0) count = 1; if (object->shadow) { int scount = count; if( object->shadow->ref_count > 1) scount /= object->shadow->ref_count; if( scount) dcount += vm_pageout_object_deactivate_pages(map, object->shadow, scount); } if (object->paging_in_progress) return dcount; /* * scan the objects entire memory queue */ rcount = object->resident_page_count; p = object->memq.tqh_first; while (p && (rcount-- > 0)) { next = p->listq.tqe_next; cnt.v_pdpages++; vm_page_lock_queues(); /* * if a page is active, not wired and is in the processes pmap, * then deactivate the page. */ if ((p->flags & (PG_ACTIVE|PG_BUSY)) == PG_ACTIVE && p->wire_count == 0 && p->hold_count == 0 && p->busy == 0 && pmap_page_exists(vm_map_pmap(map), VM_PAGE_TO_PHYS(p))) { if (!pmap_is_referenced(VM_PAGE_TO_PHYS(p)) && (p->flags & PG_REFERENCED) == 0) { p->act_count -= min(p->act_count, ACT_DECLINE); /* * if the page act_count is zero -- then we deactivate */ if (!p->act_count) { vm_page_deactivate(p); pmap_page_protect(VM_PAGE_TO_PHYS(p), VM_PROT_NONE); /* * else if on the next go-around we will deactivate the page * we need to place the page on the end of the queue to age * the other pages in memory. */ } else { TAILQ_REMOVE(&vm_page_queue_active, p, pageq); TAILQ_INSERT_TAIL(&vm_page_queue_active, p, pageq); TAILQ_REMOVE(&object->memq, p, listq); TAILQ_INSERT_TAIL(&object->memq, p, listq); } /* * see if we are done yet */ if (p->flags & PG_INACTIVE) { --count; ++dcount; if (count <= 0 && cnt.v_inactive_count > cnt.v_inactive_target) { vm_page_unlock_queues(); return dcount; } } } else { /* * Move the page to the bottom of the queue. */ pmap_clear_reference(VM_PAGE_TO_PHYS(p)); p->flags &= ~PG_REFERENCED; if (p->act_count < ACT_MAX) p->act_count += ACT_ADVANCE; TAILQ_REMOVE(&vm_page_queue_active, p, pageq); TAILQ_INSERT_TAIL(&vm_page_queue_active, p, pageq); TAILQ_REMOVE(&object->memq, p, listq); TAILQ_INSERT_TAIL(&object->memq, p, listq); } } vm_page_unlock_queues(); p = next; } return dcount; } /* * deactivate some number of pages in a map, try to do it fairly, but * that is really hard to do. */ void vm_pageout_map_deactivate_pages(map, entry, count, freeer) vm_map_t map; vm_map_entry_t entry; int *count; int (*freeer)(vm_map_t, vm_object_t, int); { vm_map_t tmpm; vm_map_entry_t tmpe; vm_object_t obj; if (*count <= 0) return; vm_map_reference(map); if (!lock_try_read(&map->lock)) { vm_map_deallocate(map); return; } if (entry == 0) { tmpe = map->header.next; while (tmpe != &map->header && *count > 0) { vm_pageout_map_deactivate_pages(map, tmpe, count, freeer); tmpe = tmpe->next; }; } else if (entry->is_sub_map || entry->is_a_map) { tmpm = entry->object.share_map; tmpe = tmpm->header.next; while (tmpe != &tmpm->header && *count > 0) { vm_pageout_map_deactivate_pages(tmpm, tmpe, count, freeer); tmpe = tmpe->next; }; } else if ((obj = entry->object.vm_object) != 0) { *count -= (*freeer)(map, obj, *count); } lock_read_done(&map->lock); vm_map_deallocate(map); return; } /* * vm_pageout_scan does the dirty work for the pageout daemon. */ int vm_pageout_scan() { vm_page_t m; int page_shortage, maxscan, maxlaunder; int pages_freed; int desired_free; vm_page_t next; struct proc *p, *bigproc; vm_offset_t size, bigsize; vm_object_t object; int force_wakeup = 0; int cache_size, orig_cache_size; /* * We manage the cached memory by attempting to keep it * at about the desired level. * We deactivate the pages for the oldest cached objects * first. This keeps pages that are "cached" from hogging * physical memory. */ orig_cache_size = 0; object = vm_object_cached_list.tqh_first; /* calculate the total cached size */ while( object) { orig_cache_size += object->resident_page_count; object = object->cached_list.tqe_next; } redeact: cache_size = orig_cache_size; object = vm_object_cached_list.tqh_first; vm_object_cache_lock(); while ( object && (cnt.v_inactive_count < cnt.v_inactive_target) && (cache_size >= (vm_swap_size?vm_desired_cache_size:0))) { vm_object_cache_unlock(); /* * if there are no resident pages -- get rid of the object */ if( object->resident_page_count == 0) { if (object != vm_object_lookup(object->pager)) panic("vm_object_deactivate: I'm sooo confused."); pager_cache(object, FALSE); goto redeact; } else { /* * if there are resident pages -- deactivate them */ vm_object_deactivate_pages(object); cache_size -= object->resident_page_count; object = object->cached_list.tqe_next; } vm_object_cache_lock(); } vm_object_cache_unlock(); morefree: /* * now swap processes out if we are in low memory conditions */ if ((cnt.v_free_count <= cnt.v_free_min) && !swap_pager_full && vm_swap_size) { /* * swap out inactive processes */ swapout_threads(); } /* * scan the processes for exceeding their rlimits or if process * is swapped out -- deactivate pages */ for (p = (struct proc *)allproc; p != NULL; p = p->p_next) { int overage; quad_t limit; /* * if this is a system process or if we have already * looked at this process, skip it. */ if (p->p_flag & (P_SYSTEM|P_WEXIT)) { continue; } /* * if the process is in a non-running type state, * don't touch it. */ if (p->p_stat != SRUN && p->p_stat != SSLEEP) { continue; } /* * get a limit */ limit = qmin(p->p_rlimit[RLIMIT_RSS].rlim_cur, p->p_rlimit[RLIMIT_RSS].rlim_max); /* * let processes that are swapped out really be swapped out * set the limit to nothing (will force a swap-out.) */ if ((p->p_flag & P_INMEM) == 0) limit = 0; size = p->p_vmspace->vm_pmap.pm_stats.resident_count * NBPG; if (limit >= 0 && size >= limit) { overage = (size - limit) / NBPG; vm_pageout_map_deactivate_pages(&p->p_vmspace->vm_map, (vm_map_entry_t) 0, &overage, vm_pageout_object_deactivate_pages); } } if (((cnt.v_free_count + cnt.v_inactive_count) >= (cnt.v_inactive_target + cnt.v_free_target)) && (cnt.v_free_count >= cnt.v_free_target)) return force_wakeup; pages_freed = 0; desired_free = cnt.v_free_target; /* * Start scanning the inactive queue for pages we can free. * We keep scanning until we have enough free pages or * we have scanned through the entire queue. If we * encounter dirty pages, we start cleaning them. */ maxlaunder = (cnt.v_free_target - cnt.v_free_count); maxscan = cnt.v_inactive_count; rescan1: m = vm_page_queue_inactive.tqh_first; while (m && (maxscan-- > 0) && (cnt.v_free_count < desired_free) ) { vm_page_t next; cnt.v_pdpages++; next = m->pageq.tqe_next; if( (m->flags & PG_INACTIVE) == 0) { printf("vm_pageout_scan: page not inactive?"); continue; } /* * activate held pages */ if (m->hold_count != 0) { vm_page_activate(m); m = next; continue; } /* * dont mess with busy pages */ if (m->busy || (m->flags & PG_BUSY)) { m = next; continue; } /* * NOTE: PG_CLEAN doesn't guarantee that the page is clean. */ if (m->flags & PG_CLEAN) { /* * If we're not low on memory and the page has been reference, * or if the page has been modified, then reactivate the page. */ if (((cnt.v_free_count > vm_pageout_free_min) && (pmap_is_referenced(VM_PAGE_TO_PHYS(m)) || ((m->flags & PG_REFERENCED) != 0))) || pmap_is_modified(VM_PAGE_TO_PHYS(m))) { m->flags &= ~PG_REFERENCED; vm_page_activate(m); } else if (!m->act_count) { pmap_page_protect(VM_PAGE_TO_PHYS(m), VM_PROT_NONE); vm_page_free(m); ++cnt.v_dfree; ++pages_freed; } else { m->act_count -= min(m->act_count, ACT_DECLINE); TAILQ_REMOVE(&vm_page_queue_inactive, m, pageq); TAILQ_INSERT_TAIL(&vm_page_queue_inactive, m, pageq); } } else if ((m->flags & PG_LAUNDRY) && maxlaunder > 0) { int written; if (pmap_is_referenced(VM_PAGE_TO_PHYS(m)) || ((m->flags & PG_REFERENCED) != 0)) { pmap_clear_reference(VM_PAGE_TO_PHYS(m)); vm_page_activate(m); m->flags &= ~PG_REFERENCED; m = next; continue; } /* * If a page is dirty, then it is either * being washed (but not yet cleaned) * or it is still in the laundry. If it is * still in the laundry, then we start the * cleaning operation. */ written = vm_pageout_clean(m,0); if (written) maxlaunder -= written; if (!next) break; /* * if the next page has been re-activated, start scanning again */ if ((next->flags & PG_INACTIVE) == 0) goto rescan1; } else if ((m->flags & PG_REFERENCED) || pmap_is_referenced(VM_PAGE_TO_PHYS(m))) { pmap_clear_reference(VM_PAGE_TO_PHYS(m)); m->flags &= ~PG_REFERENCED; vm_page_activate(m); } m = next; } /* * Compute the page shortage. If we are still very low on memory * be sure that we will move a minimal amount of pages from active * to inactive. */ page_shortage = cnt.v_inactive_target - (cnt.v_free_count + cnt.v_inactive_count); if (page_shortage <= 0) { if (pages_freed == 0) { if( cnt.v_free_count < cnt.v_free_min) { page_shortage = cnt.v_free_min - cnt.v_free_count + 1; } else if(((cnt.v_free_count + cnt.v_inactive_count) < (cnt.v_free_min + cnt.v_inactive_target))) { page_shortage = 1; } else { page_shortage = 0; } } } maxscan = cnt.v_active_count; m = vm_page_queue_active.tqh_first; while (m && maxscan-- && (page_shortage > 0)) { cnt.v_pdpages++; next = m->pageq.tqe_next; /* * Don't deactivate pages that are busy. */ if ((m->busy != 0) || (m->flags & PG_BUSY) || (m->hold_count != 0)) { m = next; continue; } if ((m->flags & PG_REFERENCED) || pmap_is_referenced(VM_PAGE_TO_PHYS(m))) { pmap_clear_reference(VM_PAGE_TO_PHYS(m)); m->flags &= ~PG_REFERENCED; if (m->act_count < ACT_MAX) m->act_count += ACT_ADVANCE; TAILQ_REMOVE(&vm_page_queue_active, m, pageq); TAILQ_INSERT_TAIL(&vm_page_queue_active, m, pageq); TAILQ_REMOVE(&m->object->memq, m, listq); TAILQ_INSERT_TAIL(&m->object->memq, m, listq); } else { m->act_count -= min(m->act_count, ACT_DECLINE); /* * if the page act_count is zero -- then we deactivate */ if (!m->act_count) { vm_page_deactivate(m); --page_shortage; /* * else if on the next go-around we will deactivate the page * we need to place the page on the end of the queue to age * the other pages in memory. */ } else { TAILQ_REMOVE(&vm_page_queue_active, m, pageq); TAILQ_INSERT_TAIL(&vm_page_queue_active, m, pageq); TAILQ_REMOVE(&m->object->memq, m, listq); TAILQ_INSERT_TAIL(&m->object->memq, m, listq); } } m = next; } /* * if we have not freed any pages and we are desparate for memory * then we keep trying until we get some (any) memory. */ if (!force_wakeup && (swap_pager_full || !force_wakeup || (pages_freed == 0 && (cnt.v_free_count < cnt.v_free_min)))){ vm_pager_sync(); force_wakeup = 1; goto morefree; } /* * make sure that we have swap space -- if we are low on * memory and swap -- then kill the biggest process. */ if ((vm_swap_size == 0 || swap_pager_full) && (cnt.v_free_count < cnt.v_free_min)) { bigproc = NULL; bigsize = 0; for (p = (struct proc *)allproc; p != NULL; p = p->p_next) { /* * if this is a system process, skip it */ if ((p->p_flag & P_SYSTEM) || ((p->p_pid < 48) && (vm_swap_size != 0))) { continue; } /* * if the process is in a non-running type state, * don't touch it. */ if (p->p_stat != SRUN && p->p_stat != SSLEEP) { continue; } /* * get the process size */ size = p->p_vmspace->vm_pmap.pm_stats.resident_count; /* * if the this process is bigger than the biggest one * remember it. */ if (size > bigsize) { bigproc = p; bigsize = size; } } if (bigproc != NULL) { printf("Process %lu killed by vm_pageout -- out of swap\n", (u_long)bigproc->p_pid); psignal(bigproc, SIGKILL); bigproc->p_estcpu = 0; bigproc->p_nice = PRIO_MIN; resetpriority(bigproc); wakeup( (caddr_t) &cnt.v_free_count); } } vm_page_pagesfreed += pages_freed; return force_wakeup; } /* * vm_pageout is the high level pageout daemon. */ void vm_pageout() { (void) spl0(); /* * Initialize some paging parameters. */ cnt.v_free_min = 12; /* * free_reserved needs to include enough for the largest * swap pager structures plus enough for any pv_entry * structs when paging. */ cnt.v_free_reserved = 4 + cnt.v_page_count / 1024; if (cnt.v_free_min < 8) cnt.v_free_min = 8; if (cnt.v_free_min > 32) cnt.v_free_min = 32; vm_pageout_free_min = cnt.v_free_reserved; cnt.v_free_target = 2*cnt.v_free_min + cnt.v_free_reserved; cnt.v_inactive_target = cnt.v_free_count / 12; cnt.v_free_min += cnt.v_free_reserved; vm_desired_cache_size = cnt.v_page_count / 3; /* XXX does not really belong here */ if (vm_page_max_wired == 0) vm_page_max_wired = cnt.v_free_count / 3; (void) swap_pager_alloc(0, 0, 0, 0); /* * The pageout daemon is never done, so loop * forever. */ while (TRUE) { int force_wakeup; /* cnt.v_free_min = 12 + averunnable.ldavg[0] / 1024; cnt.v_free_target = 2*cnt.v_free_min + cnt.v_free_reserved; cnt.v_inactive_target = cnt.v_free_target*2; */ tsleep((caddr_t) &vm_pages_needed, PVM, "psleep", 0); cnt.v_pdwakeups++; vm_pager_sync(); /* * The force wakeup hack added to eliminate delays and potiential * deadlock. It was possible for the page daemon to indefintely * postpone waking up a process that it might be waiting for memory * on. The putmulti stuff seems to have aggravated the situation. */ force_wakeup = vm_pageout_scan(); vm_pager_sync(); if( force_wakeup) wakeup( (caddr_t) &cnt.v_free_count); wakeup((caddr_t) kmem_map); } }