diff mbox series

mm/swapfile: skip HugeTLB pages for unuse_vma

Message ID 20241015014521.570237-1-liushixin2@huawei.com (mailing list archive)
State New
Headers show
Series mm/swapfile: skip HugeTLB pages for unuse_vma | expand

Commit Message

Liu Shixin Oct. 15, 2024, 1:45 a.m. UTC 
I got a bad pud error and lost a 1GB HugeTLB when calling swapoff.
The problem can be reproduced by the following steps:

 1. Allocate an anonymous 1GB HugeTLB and some other anonymous memory.
 2. Swapout the above anonymous memory.
 3. run swapoff and we will get a bad pud error in kernel message:

  mm/pgtable-generic.c:42: bad pud 00000000743d215d(84000001400000e7)

We can tell that pud_clear_bad is called by pud_none_or_clear_bad
in unuse_pud_range() by ftrace. And therefore the HugeTLB pages will
never be freed because we lost it from page table. We can skip
HugeTLB pages for unuse_vma to fix it.

Fixes: 0fe6e20b9c4c ("hugetlb, rmap: add reverse mapping for hugepage")
Signed-off-by: Liu Shixin <liushixin2@huawei.com>
---
 mm/swapfile.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Muchun Song Oct. 15, 2024, 2:33 a.m. UTC | #1 
> On Oct 15, 2024, at 09:45, Liu Shixin <liushixin2@huawei.com> wrote:
> 
> I got a bad pud error and lost a 1GB HugeTLB when calling swapoff.
> The problem can be reproduced by the following steps:
> 
> 1. Allocate an anonymous 1GB HugeTLB and some other anonymous memory.
> 2. Swapout the above anonymous memory.
> 3. run swapoff and we will get a bad pud error in kernel message:
> 
>  mm/pgtable-generic.c:42: bad pud 00000000743d215d(84000001400000e7)
> 
> We can tell that pud_clear_bad is called by pud_none_or_clear_bad
> in unuse_pud_range() by ftrace. And therefore the HugeTLB pages will
> never be freed because we lost it from page table. We can skip
> HugeTLB pages for unuse_vma to fix it.
> 
> Fixes: 0fe6e20b9c4c ("hugetlb, rmap: add reverse mapping for hugepage")
> Signed-off-by: Liu Shixin <liushixin2@huawei.com>

Acked-by: Muchun Song <muchun.song@linux.dev>

Thanks.
diff mbox series

Patch

diff --git a/mm/swapfile.c b/mm/swapfile.c
index 0cded32414a1..f4ef91513fc9 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -2312,7 +2312,7 @@  static int unuse_mm(struct mm_struct *mm, unsigned int type)
 
 	mmap_read_lock(mm);
 	for_each_vma(vmi, vma) {
-		if (vma->anon_vma) {
+		if (vma->anon_vma && !is_vm_hugetlb_page(vma)) {
 			ret = unuse_vma(vma, type);
 			if (ret)
 				break;