linux-5.15/isolate_lru_pages() の履歴(No.2)

• 履歴一覧
• 差分 を表示
• 現在との差分 を表示
• ソース を表示
• linux-5.15/isolate_lru_pages() へ行く。
  • 1 (2025-09-11 (木) 20:09:00)
  • 2 (2025-09-11 (木) 20:14:40)
  • 3 (2025-10-24 (金) 17:06:57)

参照元†

• linux-5.15/shrink_active_list()
• linux-5.15/shrink_inactive_list()

説明†

• パス: linux-5.15/mm/vmscan.c

• FIXME: これは何？
  • 説明

引数†

• unsigned long nr_to_scan
• struct lruvec *lruvec
  • linux-5.15/lruvec
• struct list_head *dst
  • linux-5.15/list_head
• unsigned long *nr_scanned
• struct scan_control *sc
  • linux-5.15/scan_control
• enum lru_list lru
  • linux-5.15/lru_list

返り値†

• unsigned long

参考†

実装†

/*
 * Isolating page from the lruvec to fill in @dst list by nr_to_scan times.
 *
 * lruvec->lru_lock is heavily contended.  Some of the functions that
 * shrink the lists perform better by taking out a batch of pages
 * and working on them outside the LRU lock.
 *
 * For pagecache intensive workloads, this function is the hottest
 * spot in the kernel (apart from copy_*_user functions).
 *
 * Lru_lock must be held before calling this function.
 *
 * @nr_to_scan:	The number of eligible pages to look through on the list.
 * @lruvec:	The LRU vector to pull pages from.
 * @dst:	The temp list to put pages on to.
 * @nr_scanned:	The number of pages that were scanned.
 * @sc:		The scan_control struct for this reclaim session
 * @lru:	LRU list id for isolating
 *
 * returns how many pages were moved onto *@dst.
 */
static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
		struct lruvec *lruvec, struct list_head *dst,
		unsigned long *nr_scanned, struct scan_control *sc,
		enum lru_list lru)
{
	struct list_head *src = &lruvec->lists[lru];
	unsigned long nr_taken = 0;
	unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
	unsigned long skipped = 0;
	unsigned long scan, total_scan, nr_pages;
	LIST_HEAD(pages_skipped);
	isolate_mode_t mode = (sc->may_unmap ? 0 : ISOLATE_UNMAPPED);

• • linux-5.15/list_head
  • linux-5.15/LIST_HEAD()

	total_scan = 0;
	scan = 0;
	while (scan < nr_to_scan && !list_empty(src)) {
		struct page *page;

		page = lru_to_page(src);
		prefetchw_prev_lru_page(page, src, flags);

		nr_pages = compound_nr(page);
		total_scan += nr_pages;

		if (page_zonenum(page) > sc->reclaim_idx) {
			list_move(&page->lru, &pages_skipped);
			nr_skipped[page_zonenum(page)] += nr_pages;
			continue;
		}

• • linux-5.15/list_empty()
  • linux-5.15/lru_to_page()
  • linux-5.15/prefetchw_prev_lru_page()
  • linux-5.15/page_zonenum()
  • linux-5.15/list_move()

		/*
		 * Do not count skipped pages because that makes the function
		 * return with no isolated pages if the LRU mostly contains
		 * ineligible pages.  This causes the VM to not reclaim any
		 * pages, triggering a premature OOM.
		 *
		 * Account all tail pages of THP.  This would not cause
		 * premature OOM since __isolate_lru_page() returns -EBUSY
		 * only when the page is being freed somewhere else.
		 */
		scan += nr_pages;
		if (!__isolate_lru_page_prepare(page, mode)) {
			/* It is being freed elsewhere */
			list_move(&page->lru, src);
			continue;
		}

• • linux-5.15/isolate_lru_page_prepare()
  • linux-5.15/list_move()

		/*
		 * Be careful not to clear PageLRU until after we're
		 * sure the page is not being freed elsewhere -- the
		 * page release code relies on it.
		 */
		if (unlikely(!get_page_unless_zero(page))) {
			list_move(&page->lru, src);
			continue;
		}

• • linux-5.15/unlikely()
  • linux-5.15/get_page_unless_zero()
  • linux-5.15/list_move()

		if (!TestClearPageLRU(page)) {
			/* Another thread is already isolating this page */
			put_page(page);
			list_move(&page->lru, src);
			continue;
		}

• • linux-5.15/TestClearPageLRU()
  • linux-5.15/put_page()
  • linux-5.15/list_move()

		nr_taken += nr_pages;
		nr_zone_taken[page_zonenum(page)] += nr_pages;
		list_move(&page->lru, dst);
	}

• • linux-5.15/list_move()

	/*
	 * Splice any skipped pages to the start of the LRU list. Note that
	 * this disrupts the LRU order when reclaiming for lower zones but
	 * we cannot splice to the tail. If we did then the SWAP_CLUSTER_MAX
	 * scanning would soon rescan the same pages to skip and put the
	 * system at risk of premature OOM.
	 */
	if (!list_empty(&pages_skipped)) {
		int zid;

		list_splice(&pages_skipped, src);
		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
			if (!nr_skipped[zid])
				continue;

			__count_zid_vm_events(PGSCAN_SKIP, zid, nr_skipped[zid]);
			skipped += nr_skipped[zid];
		}
	}
	*nr_scanned = total_scan;
	trace_mm_vmscan_lru_isolate(sc->reclaim_idx, sc->order, nr_to_scan,
				    total_scan, skipped, nr_taken, mode, lru);
	update_lru_sizes(lruvec, lru, nr_zone_taken);
	return nr_taken;
}

• • linux-5.15/list_empty()
  • linux-5.15/list_splice()
  • linux-5.15/__count_zid_vm_events()
  • linux-5.15/trace_mm_vmscan_lru_isolate()
  • linux-5.15/update_lru_sizes()

コメント†