Hi Andrew, The following diff applied to the second patch[1] fixes this error. The diff is functionally a no-op; the get_pmd_pfn() immediately following the code here would return -1 in the !pmd_present() case. I can send a brand new patch if you'd prefer. [1]: https://lore.kernel.org/linux-mm/20241019012940.3656292-3-jthoughton@xxxxxxxxxx/ diff --git a/mm/vmscan.c b/mm/vmscan.c index 29c098790b01..70604eef0dd3 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -3486,7 +3486,10 @@ static void walk_pmd_range_locked(pud_t *pud, unsigned long addr, struct vm_area /* don't round down the first address */ addr = i ? (*first & PMD_MASK) + i * PMD_SIZE : *first; - if (pmd_present(pmd[i]) && !pmd_trans_huge(pmd[i])) { + if (!pmd_present(pmd[i])) + goto next; + + if (!pmd_trans_huge(pmd[i])) { if (!walk->force_scan && should_clear_pmd_young() && !mm_has_notifiers(args->mm)) pmdp_test_and_clear_young(vma, addr, pmd + i);
