Skip to contentSkip to conversation

mm/MADV_COLLAPSE: don't expand collapse when vm_end is past requested end

PATCH Superseded Zach O'Keefe 3 participants 6 messages Dec 23, 2022

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mm/MADV_COLLAPSE: don't expand collapse when vm_end is past requested end

MADV_COLLAPSE acts on one hugepage-aligned/sized region at a time, until it has collapsed all eligible memory contained within the bounds supplied by the user.

At the top of each hugepage iteration we (re)lock mmap_lock and (re)validate the VMA for eligibility and update variables that might have changed while mmap_lock was dropped. One thing that might occur, is that the VMA could be resized, and as such, we refetch vma->vm_end to make sure we don't collapse past the end of the VMA.

However, it's possible that during this refetch that we expand the region acted on by MADV_COLLAPSE if vma->vm_end is greater than the end of the user-supplied range.

Don't expand the acted-on region when refetching vma->vm_end.

Fixes 4d24de9425f7 ("mm: MADV_COLLAPSE: refetch vm_end after reacquiring mmap_lock")
Cc Yang Shi <[email protected]>
---
 mm/khugepaged.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
mm/khugepaged.c +1 -1
index 5cb401aa2b9d..b4d2ec0a94ed 100644
@@ -2649,7 +2649,7 @@ int madvise_collapse(struct vm_area_struct *vma, struct vm_area_struct **prev,
 				goto out_nolock;
 			}
 
-			hend = vma->vm_end & HPAGE_PMD_MASK;
+			hend = min(hend, vma->vm_end & HPAGE_PMD_MASK);
 		}
 		mmap_assert_locked(mm);
 		memset(cc->node_load, 0, sizeof(cc->node_load));
Signature · 1 line
2.39.0.314.g84b9a713c41-goog
maintainer
#1 lore raw
Quoted 10 lines · On Thu, 22 Dec 2022 16:39:53 -0800 "Zach O'Keefe" <[email protected]> wrote:

MADV_COLLAPSE acts on one hugepage-aligned/sized region at a time, until it has collapsed all eligible memory contained within the bounds supplied by the user.

At the top of each hugepage iteration we (re)lock mmap_lock and (re)validate the VMA for eligibility and update variables that might have changed while mmap_lock was dropped. One thing that might occur, is that the VMA could be resized, and as such, we refetch vma->vm_end to make sure we don't collapse past the end of the VMA.

Quoted 5 lines

However, it's possible that during this refetch that we expand the region acted on by MADV_COLLAPSE if vma->vm_end is greater than the end of the user-supplied range.

Don't expand the acted-on region when refetching vma->vm_end.

What are the user-visible effects of this?

Quoted 1 line
Fixes 4d24de9425f7 ("mm: MADV_COLLAPSE: refetch vm_end after reacquiring mmap_lock")

Should we backport "mm/shmem: restore SHMEM_HUGE_DENY precedence over MADV_COLLAPSE" and/or this patch into 6.1.x?

#2 lore raw
Quoted 12 lines · Andrew Morton
Quoted 10 lines · On Thu, 22 Dec 2022 16:39:53 -0800 "Zach O'Keefe" <[email protected]> wrote:

MADV_COLLAPSE acts on one hugepage-aligned/sized region at a time, until it has collapsed all eligible memory contained within the bounds supplied by the user.

At the top of each hugepage iteration we (re)lock mmap_lock and (re)validate the VMA for eligibility and update variables that might have changed while mmap_lock was dropped. One thing that might occur, is that the VMA could be resized, and as such, we refetch vma->vm_end to make sure we don't collapse past the end of the VMA.

Quoted 7 lines
Quoted 5 lines

However, it's possible that during this refetch that we expand the region acted on by MADV_COLLAPSE if vma->vm_end is greater than the end of the user-supplied range.

Don't expand the acted-on region when refetching vma->vm_end.

What are the user-visible effects of this?

Not any kernel crash, I think; but in my case (I was trying to check something else about MADV_COLLAPSE, and so was first verifying that it worked in the simple case) I kept getting EINVAL back from it, even when I'd fixed all my own userspace mistakes.

It turned out to be that my mmap was bigger than the file itself, and I was only trying to collapse the file length; but because of the mis-adjustment to vm_end, it ran off the end of file and got into EINVAL territory (in a different context, would be EFAULT or SIGBUS).

So in my case, unexpected failure. But I guess another case would be too much success: I suppose that if you try to collapse the first 2M of a 2T file, the mis-adjustment would cause it to spend a very long time doing much more work than you asked for.

Quoted 5 lines
Quoted 2 lines
Fixes 4d24de9425f7 ("mm: MADV_COLLAPSE: refetch vm_end after reacquiring mmap_lock")

Should we backport "mm/shmem: restore SHMEM_HUGE_DENY precedence over MADV_COLLAPSE" and/or this patch into 6.1.x?

Yes, please do Cc stable for them both in 6.1.x: I only just now realized the nasty "too much success" possibility, which does seem well worth stable; and I'd particularly like the precedence of SHMEM_HUGE_DENY asserted in 6.1.x, because doing it later it would become a UAPI change - I'm sorry I didn't catch it sooner, Zach did ask me to check but I was head down on other things.

Thanks,
Hugh

#3 lore raw
Quoted 32 lines · Hugh Dickins
Quoted 12 lines · Andrew Morton
Quoted 10 lines · On Thu, 22 Dec 2022 16:39:53 -0800 "Zach O'Keefe" <[email protected]> wrote:

MADV_COLLAPSE acts on one hugepage-aligned/sized region at a time, until it has collapsed all eligible memory contained within the bounds supplied by the user.

At the top of each hugepage iteration we (re)lock mmap_lock and (re)validate the VMA for eligibility and update variables that might have changed while mmap_lock was dropped. One thing that might occur, is that the VMA could be resized, and as such, we refetch vma->vm_end to make sure we don't collapse past the end of the VMA.

Quoted 7 lines
Quoted 5 lines

However, it's possible that during this refetch that we expand the region acted on by MADV_COLLAPSE if vma->vm_end is greater than the end of the user-supplied range.

Don't expand the acted-on region when refetching vma->vm_end.

What are the user-visible effects of this?

Not any kernel crash, I think; but in my case (I was trying to check something else about MADV_COLLAPSE, and so was first verifying that it worked in the simple case) I kept getting EINVAL back from it, even when I'd fixed all my own userspace mistakes.

It turned out to be that my mmap was bigger than the file itself, and I was only trying to collapse the file length; but because of the mis-adjustment to vm_end, it ran off the end of file and got into EINVAL territory (in a different context, would be EFAULT or SIGBUS).

Quoted 4 lines

So in my case, unexpected failure. But I guess another case would be too much success: I suppose that if you try to collapse the first 2M of a 2T file, the mis-adjustment would cause it to spend a very long time doing much more work than you asked for.

Thanks Hugh,

Andrew -- I should have clarified this question in the description -- apologies there. As Hugh mentions, I don't believe there is a kernel stability concern here as we always (re)validate the VMA / region accordingly. Also as Hugh mentions, the user-visible effects are: we try to collapse more memory than requested by the user, and/or failing an operation that should have otherwise succeeded. An example is trying to collapse a 4MiB file contained within a 12MiB VMA.

Quoted 10 lines
Quoted 1 line
Fixes 4d24de9425f7 ("mm: MADV_COLLAPSE: refetch vm_end after reacquiring mmap_lock")

Should we backport "mm/shmem: restore SHMEM_HUGE_DENY precedence over MADV_COLLAPSE" and/or this patch into 6.1.x?

Quoted 5 lines
Quoted 2 lines
Fixes 4d24de9425f7 ("mm: MADV_COLLAPSE: refetch vm_end after reacquiring mmap_lock")

Should we backport "mm/shmem: restore SHMEM_HUGE_DENY precedence over MADV_COLLAPSE" and/or this patch into 6.1.x?

Quoted 6 lines

Yes, please do Cc stable for them both in 6.1.x: I only just now realized the nasty "too much success" possibility, which does seem well worth stable; and I'd particularly like the precedence of SHMEM_HUGE_DENY asserted in 6.1.x, because doing it later it would become a UAPI change - I'm sorry I didn't catch it sooner, Zach did ask me to check but I was head down on other things.

Thanks Hugh. Yes, I'm planning to backport these both to 6.1 stable after they were deemed acceptable.

Quoted 2 lines

Thanks,
Hugh

maintainer
#4 lore raw
Quoted 2 lines · On Thu, 22 Dec 2022 17:33:00 -0800 "Zach O'Keefe" <[email protected]> wrote:

Andrew -- I should have clarified this question in the description -- apologies there.

OK. Could you please send along revised changelogs which make clearer our reasons for backporting?

#5 lore raw
Quoted 8 lines · Andrew Morton
Quoted 2 lines · On Thu, 22 Dec 2022 17:33:00 -0800 "Zach O'Keefe" <[email protected]> wrote:

Andrew -- I should have clarified this question in the description -- apologies there.

OK. Could you please send along revised changelogs which make clearer our reasons for backporting?

Yep! Sounds good — will do this today. Thanks Andrew