Re: [PATCH 09/10] mm: page_isolation: prepare for hygienic freelists

[Johannes Weiner <hannes@cmpxchg.org>]

On Thu, Mar 21, 2024 at 09:13:57PM +0800, kernel test robot wrote:
> Hi Johannes,
> 
> kernel test robot noticed the following build warnings:
> 
> [auto build test WARNING on akpm-mm/mm-everything]
> 
> url:    https://github.com/intel-lab-lkp/linux/commits/Johannes-Weiner/mm-page_alloc-remove-pcppage-migratetype-caching/20240321-020814
> base:   https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-everything
> patch link:    https://lore.kernel.org/r/20240320180429.678181-10-hannes%40cmpxchg.org
> patch subject: [PATCH 09/10] mm: page_isolation: prepare for hygienic freelists
> config: i386-randconfig-003-20240321 (https://download.01.org/0day-ci/archive/20240321/202403212118.ye7lcKjD-lkp@intel.com/config)
> compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240321/202403212118.ye7lcKjD-lkp@intel.com/reproduce)
> 
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@intel.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202403212118.ye7lcKjD-lkp@intel.com/
> 
> All warnings (new ones prefixed by >>):
> 
>    mm/page_alloc.c: In function 'move_freepages_block_isolate':
> >> mm/page_alloc.c:688:17: warning: array subscript 11 is above array bounds of 'struct free_area[11]' [-Warray-bounds]
>      688 |  zone->free_area[order].nr_free--;
>          |  ~~~~~~~~~~~~~~~^~~~~~~
> >> mm/page_alloc.c:688:17: warning: array subscript 11 is above array bounds of 'struct free_area[11]' [-Warray-bounds]

I think this is a bug in the old gcc.

We have this in move_freepages_block_isolate():

	/* We're the starting block of a larger buddy */
	if (PageBuddy(page) && buddy_order(page) > pageblock_order) {
		int mt = get_pfnblock_migratetype(page, pfn);
		int order = buddy_order(page);

		if (!is_migrate_isolate(mt))
			__mod_zone_freepage_state(zone, -(1UL << order), mt);
		del_page_from_free_list(page, zone, order);

And this config doesn't have hugetlb enabled, so:

/* If huge pages are not used, group by MAX_ORDER_NR_PAGES */
#define pageblock_order         MAX_PAGE_ORDER

If buddies were indeed >MAX_PAGE_ORDER, this would be an out-of-bounds
access when delete updates the freelist count. Of course, buddies per
definition cannot be larger than MAX_PAGE_ORDER. But the older gcc
doesn't seem to realize this branch in this configuration is dead.

Maybe we can help it out and make the impossible scenario a bit more
explicit? Does this fixlet silence the warning?

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index efb2581ac142..4cdc356e73f6 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -1698,6 +1698,10 @@ bool move_freepages_block_isolate(struct zone *zone, struct page *page,
 				       NULL, NULL))
 		return false;
 
+	/* No splits needed if buddies can't span multiple blocks */
+	if (pageblock_order == MAX_PAGE_ORDER)
+		goto move;
+
 	/* We're a tail block in a larger buddy */
 	pfn = find_large_buddy(start_pfn);
 	if (pfn != start_pfn) {
@@ -1725,7 +1729,7 @@ bool move_freepages_block_isolate(struct zone *zone, struct page *page,
 		split_large_buddy(zone, page, pfn, order);
 		return true;
 	}
-
+move:
 	mt = get_pfnblock_migratetype(page, start_pfn);
 	nr_moved = move_freepages(zone, start_pfn, end_pfn, migratetype);
 	if (!is_migrate_isolate(mt))

Zi Yan, does this look sane to you as well?