]> www.infradead.org Git - users/jedix/linux-maple.git/commitdiff
mm/memory_hotplug: fix hwpoisoned large folio handling in do_migrate_range()
authorJinjiang Tu <tujinjiang@huawei.com>
Fri, 27 Jun 2025 12:57:47 +0000 (20:57 +0800)
committerAndrew Morton <akpm@linux-foundation.org>
Thu, 4 Sep 2025 00:10:35 +0000 (17:10 -0700)
In do_migrate_range(), the hwpoisoned folio may be large folio, which
can't be handled by unmap_poisoned_folio().

I can reproduce this issue in qemu after adding delay in memory_failure()

BUG: kernel NULL pointer dereference, address: 0000000000000000
Workqueue: kacpi_hotplug acpi_hotplug_work_fn
RIP: 0010:try_to_unmap_one+0x16a/0xfc0
  <TASK>
  rmap_walk_anon+0xda/0x1f0
  try_to_unmap+0x78/0x80
  ? __pfx_try_to_unmap_one+0x10/0x10
  ? __pfx_folio_not_mapped+0x10/0x10
  ? __pfx_folio_lock_anon_vma_read+0x10/0x10
  unmap_poisoned_folio+0x60/0x140
  do_migrate_range+0x4d1/0x600
  ? slab_memory_callback+0x6a/0x190
  ? notifier_call_chain+0x56/0xb0
  offline_pages+0x3e6/0x460
  memory_subsys_offline+0x130/0x1f0
  device_offline+0xba/0x110
  acpi_bus_offline+0xb7/0x130
  acpi_scan_hot_remove+0x77/0x290
  acpi_device_hotplug+0x1e0/0x240
  acpi_hotplug_work_fn+0x1a/0x30
  process_one_work+0x186/0x340

Besides, do_migrate_range() may be called between memory_failure set
hwpoison flag and isolate the folio from lru, so remove WARN_ON(). In other
places, unmap_poisoned_folio() is called when the folio is isolated, obey
it in do_migrate_range() too.

[david@redhat.com: don't abort offlining, fixed typo, add comment]
Link: https://lkml.kernel.org/r/3c214dff-9649-4015-840f-10de0e03ebe4@redhat.com
Fixes: b15c87263a69 ("hwpoison, memory_hotplug: allow hwpoisoned pages to be offlined")
Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Luis Chamberalin <mcgrof@kernel.org>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Pankaj Raghav <kernel@pankajraghav.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
mm/memory_hotplug.c

index 1f15af712bc3465df3c732299716ab0d6289b9a1..74318c7877156734c6e78d6d1dfb80d96aea64bc 100644 (file)
@@ -1815,8 +1815,14 @@ static void do_migrate_range(unsigned long start_pfn, unsigned long end_pfn)
                        pfn = folio_pfn(folio) + folio_nr_pages(folio) - 1;
 
                if (folio_contain_hwpoisoned_page(folio)) {
-                       if (WARN_ON(folio_test_lru(folio)))
-                               folio_isolate_lru(folio);
+                       /*
+                        * unmap_poisoned_folio() cannot handle large folios
+                        * in all cases yet.
+                        */
+                       if (folio_test_large(folio) && !folio_test_hugetlb(folio))
+                               goto put_folio;
+                       if (folio_test_lru(folio) && !folio_isolate_lru(folio))
+                               goto put_folio;
                        if (folio_mapped(folio)) {
                                folio_lock(folio);
                                unmap_poisoned_folio(folio, pfn, false);