^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * linux/mm/page_isolation.c
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #include <linux/page-isolation.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) #include <linux/pageblock-flags.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include <linux/memory.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) #include <linux/hugetlb.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include <linux/page_owner.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <linux/migrate.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include "internal.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #define CREATE_TRACE_POINTS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include <trace/events/page_isolation.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) struct zone *zone = page_zone(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) struct page *unmovable;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) spin_lock_irqsave(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) * We assume the caller intended to SET migrate type to isolate.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) * If it is already set, then someone else must have raced and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * set it before us.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) if (is_migrate_isolate_page(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) spin_unlock_irqrestore(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * FIXME: Now, memory hotplug doesn't call shrink_slab() by itself.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) * We just check MOVABLE pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) unmovable = has_unmovable_pages(zone, page, migratetype, isol_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) if (!unmovable) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) unsigned long nr_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) int mt = get_pageblock_migratetype(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) set_pageblock_migratetype(page, MIGRATE_ISOLATE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) zone->nr_isolate_pageblock++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) nr_pages = move_freepages_block(zone, page, MIGRATE_ISOLATE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) __mod_zone_freepage_state(zone, -nr_pages, mt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) spin_unlock_irqrestore(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) spin_unlock_irqrestore(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) if (isol_flags & REPORT_FAILURE) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) * printk() with zone->lock held will likely trigger a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) * lockdep splat, so defer it here.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) dump_page(unmovable, "unmovable page");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) struct zone *zone;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) unsigned long flags, nr_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) bool isolated_page = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) unsigned int order;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) unsigned long pfn, buddy_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) struct page *buddy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) zone = page_zone(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) spin_lock_irqsave(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) if (!is_migrate_isolate_page(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) * Because freepage with more than pageblock_order on isolated
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) * pageblock is restricted to merge due to freepage counting problem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) * it is possible that there is free buddy page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) * move_freepages_block() doesn't care of merge so we need other
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) * approach in order to merge them. Isolation and free will make
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) * these pages to be merged.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) if (PageBuddy(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) order = buddy_order(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) if (order >= pageblock_order) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) pfn = page_to_pfn(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) buddy_pfn = __find_buddy_pfn(pfn, order);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) buddy = page + (buddy_pfn - pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) if (pfn_valid_within(buddy_pfn) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) !is_migrate_isolate_page(buddy)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) __isolate_free_page(page, order);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) isolated_page = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * If we isolate freepage with more than pageblock_order, there
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) * should be no freepage in the range, so we could avoid costly
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) * pageblock scanning for freepage moving.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) * We didn't actually touch any of the isolated pages, so place them
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) * to the tail of the freelist. This is an optimization for memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) * onlining - just onlined memory won't immediately be considered for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) * allocation.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) if (!isolated_page) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) nr_pages = move_freepages_block(zone, page, migratetype, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) __mod_zone_freepage_state(zone, nr_pages, migratetype);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) set_pageblock_migratetype(page, migratetype);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) if (isolated_page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) __putback_isolated_page(page, order, migratetype);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) zone->nr_isolate_pageblock--;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) spin_unlock_irqrestore(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) static inline struct page *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) __first_valid_page(unsigned long pfn, unsigned long nr_pages)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) for (i = 0; i < nr_pages; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) page = pfn_to_online_page(pfn + i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) return page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) * start_isolate_page_range() - make page-allocation-type of range of pages to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) * be MIGRATE_ISOLATE.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) * @start_pfn: The lower PFN of the range to be isolated.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) * @end_pfn: The upper PFN of the range to be isolated.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) * start_pfn/end_pfn must be aligned to pageblock_order.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) * @migratetype: Migrate type to set in error recovery.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) * @flags: The following flags are allowed (they can be combined in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) * a bit mask)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) * MEMORY_OFFLINE - isolate to offline (!allocate) memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) * e.g., skip over PageHWPoison() pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) * and PageOffline() pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) * REPORT_FAILURE - report details about the failure to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) * isolate the range
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) * Making page-allocation-type to be MIGRATE_ISOLATE means free pages in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) * the range will never be allocated. Any free pages and pages freed in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) * future will not be allocated again. If specified range includes migrate types
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) * other than MOVABLE or CMA, this will fail with -EBUSY. For isolating all
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) * pages in the range finally, the caller have to free all pages in the range.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) * test_page_isolated() can be used for test it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) * There is no high level synchronization mechanism that prevents two threads
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) * from trying to isolate overlapping ranges. If this happens, one thread
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) * will notice pageblocks in the overlapping range already set to isolate.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) * This happens in set_migratetype_isolate, and set_migratetype_isolate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) * returns an error. We then clean up by restoring the migration type on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) * pageblocks we may have modified and return -EBUSY to caller. This
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) * prevents two threads from simultaneously working on overlapping ranges.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) * Please note that there is no strong synchronization with the page allocator
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) * either. Pages might be freed while their page blocks are marked ISOLATED.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) * A call to drain_all_pages() after isolation can flush most of them. However
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) * in some cases pages might still end up on pcp lists and that would allow
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) * for their allocation even when they are in fact isolated already. Depending
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) * on how strong of a guarantee the caller needs, further drain_all_pages()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) * might be needed (e.g. __offline_pages will need to call it after check for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) * isolated range for a next retry).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) * Return: 0 on success and -EBUSY if any part of range cannot be isolated.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) unsigned migratetype, int flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) unsigned long *failed_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) unsigned long pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) unsigned long undo_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) BUG_ON(!IS_ALIGNED(start_pfn, pageblock_nr_pages));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) BUG_ON(!IS_ALIGNED(end_pfn, pageblock_nr_pages));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) for (pfn = start_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) pfn < end_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) pfn += pageblock_nr_pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) page = __first_valid_page(pfn, pageblock_nr_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) if (page) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) if (set_migratetype_isolate(page, migratetype, flags)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) undo_pfn = pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) if (failed_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) *failed_pfn = page_to_pfn(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) goto undo;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) undo:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) for (pfn = start_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) pfn < undo_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) pfn += pageblock_nr_pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) struct page *page = pfn_to_online_page(pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) unset_migratetype_isolate(page, migratetype);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) * Make isolated pages available again.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) void undo_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) unsigned migratetype)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) unsigned long pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) BUG_ON(!IS_ALIGNED(start_pfn, pageblock_nr_pages));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) BUG_ON(!IS_ALIGNED(end_pfn, pageblock_nr_pages));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) for (pfn = start_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) pfn < end_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) pfn += pageblock_nr_pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) page = __first_valid_page(pfn, pageblock_nr_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) if (!page || !is_migrate_isolate_page(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) unset_migratetype_isolate(page, migratetype);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) * Test all pages in the range is free(means isolated) or not.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) * all pages in [start_pfn...end_pfn) must be in the same zone.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) * zone->lock must be held before call this.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) * Returns the last tested pfn.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) static unsigned long
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) __test_page_isolated_in_pageblock(unsigned long pfn, unsigned long end_pfn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) int flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) while (pfn < end_pfn) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) if (!pfn_valid_within(pfn)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) pfn++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) page = pfn_to_page(pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) if (PageBuddy(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) * If the page is on a free list, it has to be on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) * the correct MIGRATE_ISOLATE freelist. There is no
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) * simple way to verify that as VM_BUG_ON(), though.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) pfn += 1 << buddy_order(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) else if ((flags & MEMORY_OFFLINE) && PageHWPoison(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) /* A HWPoisoned page cannot be also PageBuddy */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) pfn++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) else if ((flags & MEMORY_OFFLINE) && PageOffline(page) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) !page_count(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) * The responsible driver agreed to skip PageOffline()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) * pages when offlining memory by dropping its
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) * reference in MEM_GOING_OFFLINE.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) pfn++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) #ifdef CONFIG_NO_GKI
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) if (pfn < end_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) dump_page_owner(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) return pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) /* Caller should ensure that requested range is in a single zone */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) int test_pages_isolated(unsigned long start_pfn, unsigned long end_pfn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) int isol_flags, unsigned long *failed_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) unsigned long pfn, flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) struct zone *zone;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) * Note: pageblock_nr_pages != MAX_ORDER. Then, chunks of free pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) * are not aligned to pageblock_nr_pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) * Then we just check migratetype first.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) for (pfn = start_pfn; pfn < end_pfn; pfn += pageblock_nr_pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) page = __first_valid_page(pfn, pageblock_nr_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) if (page && !is_migrate_isolate_page(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) page = __first_valid_page(start_pfn, end_pfn - start_pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) if ((pfn < end_pfn) || !page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) /* Check all pages are free or marked as ISOLATED */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) zone = page_zone(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) spin_lock_irqsave(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) pfn = __test_page_isolated_in_pageblock(start_pfn, end_pfn, isol_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) spin_unlock_irqrestore(&zone->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) trace_test_pages_isolated(start_pfn, end_pfn, pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) if (pfn < end_pfn) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) page_pinner_failure_detect(pfn_to_page(pfn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) if (failed_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) *failed_pfn = pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) }