Skip to contentSkip to conversation

mm, hwpoison: enable 1GB hugepage support (v2)

PATCH In review Naoya Horiguchi 5 participants 45 messages Jun 23, 2022 – Jun 30, 2022

Subsystems (8)
Memory management x86 MM core x86 MM +4 more subsystems
lore
Share & tools
lore

Naoya Horiguchi

author reviewer
#0 lore raw

v2 cover mm, hwpoison: enable 1GB hugepage support (v2)

Here is v2 of "enabling memory error handling on 1GB hugepage" patchset.

Major updates:

- (patch 3/9) I made pud_huge() and follow_huge_pud() aware of non-present pud entry (based on Miaohe's comment).

- (patch 4/9 and patch 5/9) I extended the mechanism to save raw error info to support multiple error subpages in a hugepage. Additionally, I added a "unreliable" flag which prevents freeing hwpoison hugetlb if any raw error info is lost.

- (patch 1/9 and 2/9) During testing some common cases for 1GB hugepage, I found a few issues in existing code, so this series starts with fixing them.

The remaining patches should have only minor updates since v1.

Patch dependency:

- "mm/memory-failure: disable unpoison once hw error happens" (actually the conflict is not logical one, but adding MF_SIMULATED to mf_flags conflicts with patch 6/9.)

v1: https://lore.kernel.org/linux-mm/[email protected]/T/#u

Thanks,
Naoya Horiguchi
---
Summary:

Naoya Horiguchi (9):
mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages() mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range() mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry mm, hwpoison, hugetlb: support saving mechanism of raw error pages mm, hwpoison: make unpoison aware of raw error info in hwpoisoned hugepage mm, hwpoison: set PG_hwpoison for busy hugetlb pages
mm, hwpoison: make __page_handle_poison returns int
mm, hwpoison: skip raw hwpoison page in freeing 1GB hugepage
mm, hwpoison: enable memory error handling on 1GB hugepage

 arch/x86/mm/hugetlbpage.c |   3 +-
 include/linux/hugetlb.h   |  13 ++++
 include/linux/mm.h        |   2 +-
 include/linux/swapops.h   |   9 +++
 include/ras/ras_event.h   |   1 -
 mm/hugetlb.c              |  78 ++++++++++++++--------
 mm/memory-failure.c       | 163 +++++++++++++++++++++++++++++++++++++---------
 7 files changed, 209 insertions(+), 60 deletions(-)
author reviewer
#1 lore raw

PATCH v2 1/9 mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

---
 mm/hugetlb.c | 4 ----
 1 file changed, 4 deletions(-)
mm/hugetlb.c +0 -4
index a57e1be41401..c538278170a2 100644
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-
 	/*
 	 * Part (or even all) of the reservation could have been backed
 	 * by pre-allocated pages. Only free surplus pages.
Signature · 1 line
2.25.1
maintainer
#2 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3
Quoted 8 lines

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

s/unnecessary/necessary/

Quoted 11 lines
---
 mm/hugetlb.c | 4 ----
 1 file changed, 4 deletions(-)
mm/hugetlb.c +0 -0
index a57e1be41401..c538278170a2 100644

mm/hugetlb.c static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 7 lines
mm/hugetlb.c +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;

But I might be miss something.

Thanks.

Quoted 4 lines
mm/hugetlb.c +0 -0
 	/*
 	 * Part (or even all) of the reservation could have been backed
 	 * by pre-allocated pages. Only free surplus pages.
#3 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 53 lines · Miaohe Lin
Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3
Quoted 8 lines

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

s/unnecessary/necessary/

Quoted 11 lines
---
 mm/hugetlb.c | 4 ----
 1 file changed, 4 deletions(-)
mm/hugetlb.c +0 -0
index a57e1be41401..c538278170a2 100644

mm/hugetlb.c static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 7 lines
mm/hugetlb.c +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

Quoted 7 lines
	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Thanks.

Quoted 10 lines

But I might be miss something.

Thanks.

Quoted 4 lines
mm/hugetlb.c +0 -0
 	/*
 	 * Part (or even all) of the reservation could have been backed
 	 * by pre-allocated pages. Only free surplus pages.
maintainer
#4 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 62 lines · Muchun Song
Quoted 53 lines · Miaohe Lin
Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3
Quoted 8 lines

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

s/unnecessary/necessary/

Quoted 11 lines
---
 mm/hugetlb.c | 4 ----
 1 file changed, 4 deletions(-)
mm/hugetlb.c +0 -0
index a57e1be41401..c538278170a2 100644

mm/hugetlb.c static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 7 lines
mm/hugetlb.c +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

Quoted 7 lines
	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Quoted 2 lines

Thanks.

Thanks for reviewing.

Quoted 13 lines
Quoted 11 lines

But I might be miss something.

Thanks.

Quoted 4 lines
mm/hugetlb.c +0 -0
 	/*
 	 * Part (or even all) of the reservation could have been backed
 	 * by pre-allocated pages. Only free surplus pages.

.

#5 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 28 lines · Miaohe Lin
Quoted 27 lines · Muchun Song
Quoted 26 lines · Miaohe Lin
Quoted 25 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

s/unnecessary/necessary/

Thanks, I fixed it.

Quoted 29 lines
Quoted 29 lines
Quoted 28 lines
Quoted 12 lines
---
 mm/hugetlb.c | 4 ----
 1 file changed, 4 deletions(-)
mm/hugetlb.c +0 -0
index a57e1be41401..c538278170a2 100644

mm/hugetlb.c static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 7 lines
mm/hugetlb.c +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

Thanks,
Naoya Horiguchi

#6 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 25 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 24 lines · Miaohe Lin
Quoted 23 lines · Muchun Song
Quoted 22 lines · Miaohe Lin
Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

Thank you for finding this!!!

Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 18 lines

Review static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 9 lines

+++ b/mm/hugetlb.c

(quoted diff) +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported(). To avoid this situation, perhaps we should change set_max_huge_pages as follows (not tested)?

Signature · 1 line
Mike Kravetz
mm/hugetlb.c +15 -1
index 5eabb8009d8a..c78d6c20e6b0 100644
@@ -3416,7 +3416,8 @@ static int set_max_huge_pages(struct hstate *h, unsigned long count, int nid,
 	 * the user tries to allocate gigantic pages but let the user free the
 	 * boottime allocated gigantic pages.
 	 */
-	if (hstate_is_gigantic(h) && !IS_ENABLED(CONFIG_CONTIG_ALLOC)) {
+	if (hstate_is_gigantic(h) && (!IS_ENABLED(CONFIG_CONTIG_ALLOC) ||
+					!gigantic_page_runtime_supported())) {
 		if (count > persistent_huge_pages(h)) {
 			spin_unlock_irq(&hugetlb_lock);
 			mutex_unlock(&h->resize_lock);
@@ -3464,6 +3465,19 @@ static int set_max_huge_pages(struct hstate *h, unsigned long count, int nid,
 			goto out;
 	}
 
+	/*
+	 * We can not decrease gigantic pool size if runtime modification
+	 * is not supported.
+	 */
+	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported()) {
+		if (count < persistent_huge_pages(h)) {
+			spin_unlock_irq(&hugetlb_lock);
+			mutex_unlock(&h->resize_lock);
+			NODEMASK_FREE(node_alloc_noretry);
+			return -EINVAL;
+		}
+	}
+
 	/*
 	 * Decrease the pool size
 	 * First return free pages to the buddy allocator (being careful
#7 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 66 lines · Mike Kravetz
Quoted 25 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 24 lines · Miaohe Lin
Quoted 23 lines · Muchun Song
Quoted 22 lines · Miaohe Lin
Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

I found a weird state of 1GB hugepage pool, caused by the following procedure:

- run a process reserving all free 1GB hugepages,
- shrink free 1GB hugepage pool to zero (i.e. writing 0 to

    /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages), then

- kill the reserving process.

, then all the hugepages are free *and* surplus at the same time.

  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/free_hugepages
  3
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/resv_hugepages
  0
  $ cat /sys/kernel/mm/hugepages/hugepages-1048576kB/surplus_hugepages
  3
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

This state is resolved by reserving and allocating the pages then freeing them again, so this seems not to result in serious problem. But it's a little surprizing (shrinking pool suddenly fails).

This behavior is caused by hstate_is_gigantic() check in return_unused_surplus_pages(). This was introduced so long ago in 2008 by commit aa888a74977a ("hugetlb: support larger than MAX_ORDER"), and it seems to me that this check is no longer unnecessary. Let's remove it.

Thank you for finding this!!!

Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 18 lines

Review static void return_unused_surplus_pages(struct hstate *h, · line 2438

Quoted 9 lines

+++ b/mm/hugetlb.c

(quoted diff) +0 -4
@@ -2432,10 +2432,6 @@ static void return_unused_surplus_pages(struct hstate *h,
 	/* Uncommit the reservation */
 	h->resv_huge_pages -= unused_resv_pages;
 
-	/* Cannot return gigantic pages currently */
-	if (hstate_is_gigantic(h))
-		goto out;
-

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

Quoted 7 lines

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported().

I have the same understanding as the above.

Quoted 2 lines

To avoid this situation,
perhaps we should change set_max_huge_pages as follows (not tested)?

The suggested diff looks clearer about what it does, so I'd like to take it in the next version. Then, what do we do on the "if (hstate_if_gigantic())" check in return_unused_surplus_pages in the original suggestion? Should it be kept as is, or removed, or checked with !gigantic_page_runtime_supported()? I guess that the new checks prevent calling return_unused_surplus_pages() during pool shrinking, so the check seems not necessary any more.

Thanks,
Naoya Horiguchi

Quoted 39 lines
Signature · 2 lines
-- 
Mike Kravetz
mm/hugetlb.c +15 -1
index 5eabb8009d8a..c78d6c20e6b0 100644
@@ -3416,7 +3416,8 @@ static int set_max_huge_pages(struct hstate *h, unsigned long count, int nid,
 	 * the user tries to allocate gigantic pages but let the user free the
 	 * boottime allocated gigantic pages.
 	 */
-	if (hstate_is_gigantic(h) && !IS_ENABLED(CONFIG_CONTIG_ALLOC)) {
+	if (hstate_is_gigantic(h) && (!IS_ENABLED(CONFIG_CONTIG_ALLOC) ||
+					!gigantic_page_runtime_supported())) {
 		if (count > persistent_huge_pages(h)) {
 			spin_unlock_irq(&hugetlb_lock);
 			mutex_unlock(&h->resize_lock);
@@ -3464,6 +3465,19 @@ static int set_max_huge_pages(struct hstate *h, unsigned long count, int nid,
 			goto out;
 	}
 
+	/*
+	 * We can not decrease gigantic pool size if runtime modification
+	 * is not supported.
+	 */
+	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported()) {
+		if (count < persistent_huge_pages(h)) {
+			spin_unlock_irq(&hugetlb_lock);
+			mutex_unlock(&h->resize_lock);
+			NODEMASK_FREE(node_alloc_noretry);
+			return -EINVAL;
+		}
+	}
+
 	/*
 	 * Decrease the pool size
 	 * First return free pages to the buddy allocator (being careful
#8 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 40 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 26 lines · Mike Kravetz
Quoted 15 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 14 lines · Miaohe Lin
Quoted 13 lines · Muchun Song
Quoted 11 lines · Miaohe Lin
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

Quoted 7 lines

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported().

I have the same understanding as the above.

Quoted 2 lines

To avoid this situation,
perhaps we should change set_max_huge_pages as follows (not tested)?

Quoted 6 lines

The suggested diff looks clearer about what it does, so I'd like to take it in the next version. Then, what do we do on the "if (hstate_if_gigantic())" check in return_unused_surplus_pages in the original suggestion? Should it be kept as is, or removed, or checked with !gigantic_page_runtime_supported()? I guess that the new checks prevent calling return_unused_surplus_pages() during pool shrinking, so the check seems not necessary any more.

My first thought was to keep the check in return_unused_surplus_pages() as it is called in other code paths. However, it SHOULD only try to free surplus hugetlb pages. With the modifications to set_max_huge_pages we will not have any surplus gigantic pages if !gigantic_page_runtime_supported, so the check can be removed.

Also note that we never try to dynamically allocate surplus gigantic pages. This also is left over from the time when we could not easily allocate a gigantic page at runtime. It would not surprise me if someone found a use case to ease this restriction in the future. Especially so if 1G THP support is ever added. If this happens, the check would be necessary and I would guess that it would not be added.

Sorry for thinking our loud!!! Although not necessary, it 'might' be a good idea to leave the check because it would be overlooked if dynamic allocation of gigantic surplus pages is ever added. I do not have a strong opinion.

P.S. This also reminds me that a similar check should be added to the demote hugetlb code path. It would be bad if !gigantic_page_runtime_supported and we demoted a gigantic page into numerous non-gigantic pages. I will send a patch.

Signature · 1 line
Mike Kravetz
maintainer
#9 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 61 lines · Mike Kravetz
Quoted 40 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 26 lines · Mike Kravetz
Quoted 15 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 14 lines · Miaohe Lin
Quoted 13 lines · Muchun Song
Quoted 11 lines · Miaohe Lin
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

Quoted 7 lines

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported().

I have the same understanding as the above.

Quoted 2 lines

To avoid this situation,
perhaps we should change set_max_huge_pages as follows (not tested)?

Quoted 6 lines

The suggested diff looks clearer about what it does, so I'd like to take it in the next version. Then, what do we do on the "if (hstate_if_gigantic())" check in return_unused_surplus_pages in the original suggestion? Should it be kept as is, or removed, or checked with !gigantic_page_runtime_supported()? I guess that the new checks prevent calling return_unused_surplus_pages() during pool shrinking, so the check seems not necessary any more.

My first thought was to keep the check in return_unused_surplus_pages() as it is called in other code paths. However, it SHOULD only try to free surplus hugetlb pages. With the modifications to set_max_huge_pages we will not have any surplus gigantic pages if !gigantic_page_runtime_supported, so the check can be removed.

Also note that we never try to dynamically allocate surplus gigantic pages. This also is left over from the time when we could not easily allocate a gigantic page at runtime. It would not surprise me if someone found a use case to ease this restriction in the future. Especially so if 1G THP support is ever added. If this happens, the check would be necessary and I would guess that it would not be added.

Quoted 8 lines

Sorry for thinking our loud!!! Although not necessary, it 'might' be a good idea to leave the check because it would be overlooked if dynamic allocation of gigantic surplus pages is ever added. I do not have a strong opinion.

P.S. This also reminds me that a similar check should be added to the demote hugetlb code path. It would be bad if !gigantic_page_runtime_supported and we demoted a gigantic page into numerous non-gigantic pages. I will send a patch.

Out-of-topic
There're some places check "if (hstate_is_gigantic(h))" while others check "if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())". Should we make it explicit in some manner when gigantic_page_runtime_supported is needed to make code easier to follow?

Just a trivial suggestion. Thanks!

Quoted 1 line
#10 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

In reply to Mike Kravetz
Quoted 61 lines · Mike Kravetz
Quoted 40 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 26 lines · Mike Kravetz
Quoted 15 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 14 lines · Miaohe Lin
Quoted 13 lines · Muchun Song
Quoted 11 lines · Miaohe Lin
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

Quoted 7 lines

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported().

I have the same understanding as the above.

Quoted 2 lines

To avoid this situation,
perhaps we should change set_max_huge_pages as follows (not tested)?

Quoted 6 lines

The suggested diff looks clearer about what it does, so I'd like to take it in the next version. Then, what do we do on the "if (hstate_if_gigantic())" check in return_unused_surplus_pages in the original suggestion? Should it be kept as is, or removed, or checked with !gigantic_page_runtime_supported()? I guess that the new checks prevent calling return_unused_surplus_pages() during pool shrinking, so the check seems not necessary any more.

My first thought was to keep the check in return_unused_surplus_pages() as it is called in other code paths. However, it SHOULD only try to free surplus hugetlb pages. With the modifications to set_max_huge_pages we will not have any surplus gigantic pages if !gigantic_page_runtime_supported, so the check can be removed.

Also note that we never try to dynamically allocate surplus gigantic pages. This also is left over from the time when we could not easily allocate a gigantic page at runtime. It would not surprise me if someone found a use case to ease this restriction in the future. Especially so if 1G THP support is ever added. If this happens, the check would be necessary and I would guess that it would not be added.

Quoted 3 lines

Sorry for thinking our loud!!! Although not necessary, it 'might' be a good idea to leave the check because it would be overlooked if dynamic allocation of gigantic surplus pages is ever added. I do not have a strong opinion.

OK, so let's keep the check.

Quoted 5 lines

P.S. This also reminds me that a similar check should be added to the demote hugetlb code path. It would be bad if !gigantic_page_runtime_supported and we demoted a gigantic page into numerous non-gigantic pages. I will send a patch.

Sounds nice.

Thanks,
Naoya Horiguchi

#11 lore raw

Re: [PATCH v2 1/9] mm/hugetlb: remove checking hstate_is_gigantic() in return_unused_surplus_pages()

Quoted 62 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 61 lines · Mike Kravetz
Quoted 40 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 26 lines · Mike Kravetz
Quoted 15 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 14 lines · Miaohe Lin
Quoted 13 lines · Muchun Song
Quoted 11 lines · Miaohe Lin
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

IIUC it might be better to do the below check:

	/*
	 * Cannot return gigantic pages currently if runtime gigantic page
	 * allocation is not supported.
	 */
	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
		goto out;
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

The change looks good to me. However, the comments above is unnecessary since gigantic_page_runtime_supported() is straightforward.

Agree. The comments can be removed.

Thank you, both. Adding !gigantic_page_runtime_supported without comment makes sense to me.

The change above makes sense to me. However, ...

Quoted 7 lines

If we make the change above, will we have the same strange situation described in the commit message when !gigantic_page_runtime_supported() is true?

IIUC, !gigantic_page_runtime_supported implies that gigantic hugetlb pages can not be allocated or freed at run time. They can only be allocated at boot time. So, there should NEVER be surplus gigantic pages if !gigantic_page_runtime_supported().

I have the same understanding as the above.

Quoted 2 lines

To avoid this situation,
perhaps we should change set_max_huge_pages as follows (not tested)?

Quoted 6 lines

The suggested diff looks clearer about what it does, so I'd like to take it in the next version. Then, what do we do on the "if (hstate_if_gigantic())" check in return_unused_surplus_pages in the original suggestion? Should it be kept as is, or removed, or checked with !gigantic_page_runtime_supported()? I guess that the new checks prevent calling return_unused_surplus_pages() during pool shrinking, so the check seems not necessary any more.

My first thought was to keep the check in return_unused_surplus_pages() as it is called in other code paths. However, it SHOULD only try to free surplus hugetlb pages. With the modifications to set_max_huge_pages we will not have any surplus gigantic pages if !gigantic_page_runtime_supported, so the check can be removed.

Also note that we never try to dynamically allocate surplus gigantic pages. This also is left over from the time when we could not easily allocate a gigantic page at runtime. It would not surprise me if someone found a use case to ease this restriction in the future. Especially so if 1G THP support is ever added. If this happens, the check would be necessary and I would guess that it would not be added.

Quoted 5 lines
Quoted 3 lines

Sorry for thinking our loud!!! Although not necessary, it 'might' be a good idea to leave the check because it would be overlooked if dynamic allocation of gigantic surplus pages is ever added. I do not have a strong opinion.

OK, so let's keep the check.

Sorry, I found that keeping the check doesn't fix the problem. I'll update the check with !gigantic_page_runtime_supported().

- Naoya Horiguchi

author reviewer
#12 lore raw

PATCH v2 2/9 mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

Separate the if branch into one for hwpoison entries and one for migration entries.

Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Cc <[email protected]> # 5.18
---
 mm/hugetlb.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
mm/hugetlb.c +7 -2
index c538278170a2..f59f43c06601 100644
@@ -4784,8 +4784,13 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			 * sharing with another vma.
 			 */
 			;
-		} else if (unlikely(is_hugetlb_entry_migration(entry) ||
-				    is_hugetlb_entry_hwpoisoned(entry))) {
+		} else if (unlikely(is_hugetlb_entry_hwpoisoned(entry))) {
+			bool uffd_wp = huge_pte_uffd_wp(entry);
+
+			if (!userfaultfd_wp(dst_vma) && uffd_wp)
+				entry = huge_pte_clear_uffd_wp(entry);
+			set_huge_swap_pte_at(dst, addr, dst_pte, entry, sz);
+		} else if (unlikely(is_hugetlb_entry_migration(entry))) {
 			swp_entry_t swp_entry = pte_to_swp_entry(entry);
 			bool uffd_wp = huge_pte_uffd_wp(entry);
 
Signature · 1 line
2.25.1
maintainer Reviewed-by on v2 2/9
#13 lore raw

Re: [PATCH v2 2/9] mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

Quoted 3 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone

s/hwpoisone/hwpoisoned/

Quoted 6 lines

entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

s/foking/forking/

Quoted 7 lines

Separate the if branch into one for hwpoison entries and one for migration entries.

Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Cc <[email protected]> # 5.18

This makes sense to me. Thanks for fixing this.

Quoted 25 lines
---
 mm/hugetlb.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
mm/hugetlb.c +7 -2
index c538278170a2..f59f43c06601 100644
@@ -4784,8 +4784,13 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			 * sharing with another vma.
 			 */
 			;
-		} else if (unlikely(is_hugetlb_entry_migration(entry) ||
-				    is_hugetlb_entry_hwpoisoned(entry))) {
+		} else if (unlikely(is_hugetlb_entry_hwpoisoned(entry))) {
+			bool uffd_wp = huge_pte_uffd_wp(entry);
+
+			if (!userfaultfd_wp(dst_vma) && uffd_wp)
+				entry = huge_pte_clear_uffd_wp(entry);
+			set_huge_swap_pte_at(dst, addr, dst_pte, entry, sz);
+		} else if (unlikely(is_hugetlb_entry_migration(entry))) {
 			swp_entry_t swp_entry = pte_to_swp_entry(entry);
 			bool uffd_wp = huge_pte_uffd_wp(entry);
 
#14 lore raw

Re: [PATCH v2 2/9] mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

Quoted 7 lines · Miaohe Lin
Quoted 3 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone

s/hwpoisone/hwpoisoned/

Quoted 8 lines
Quoted 6 lines

entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

s/foking/forking/

Sorry for many typos. I somehow forgot to run spell checker.

Quoted 5 lines
Quoted 5 lines

Separate the if branch into one for hwpoison entries and one for migration entries.

Quoted 7 lines
Quoted 3 lines
Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Cc <[email protected]> # 5.18

This makes sense to me. Thanks for fixing this.

Thank you for reviewing!

- Naoya Horiguchi

Quoted 26 lines
Quoted 26 lines
---
 mm/hugetlb.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
mm/hugetlb.c +7 -2
index c538278170a2..f59f43c06601 100644
@@ -4784,8 +4784,13 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			 * sharing with another vma.
 			 */
 			;
-		} else if (unlikely(is_hugetlb_entry_migration(entry) ||
-				    is_hugetlb_entry_hwpoisoned(entry))) {
+		} else if (unlikely(is_hugetlb_entry_hwpoisoned(entry))) {
+			bool uffd_wp = huge_pte_uffd_wp(entry);
+
+			if (!userfaultfd_wp(dst_vma) && uffd_wp)
+				entry = huge_pte_clear_uffd_wp(entry);
+			set_huge_swap_pte_at(dst, addr, dst_pte, entry, sz);
+		} else if (unlikely(is_hugetlb_entry_migration(entry))) {
 			swp_entry_t swp_entry = pte_to_swp_entry(entry);
 			bool uffd_wp = huge_pte_uffd_wp(entry);
 
Reviewed-by on v2 2/9
#15 lore raw

Re: [PATCH v2 2/9] mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

In reply to Naoya Horiguchi
Quoted 13 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

Separate the if branch into one for hwpoison entries and one for migration entries.

Quoted 6 lines
Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Cc <[email protected]> # 5.18
---
 mm/hugetlb.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Thank you!

(with typos pointed out by Miaohe Lin)

Just curious, are there any hwpoisoned tests in any test suite? I run libhugetlbfs tests and ltp on a regular basis which sometimes catch regressions. If there are no tests in any suite today, this might be something we want to consider for future work.

Signature · 1 line
Mike Kravetz
#16 lore raw

Re: [PATCH v2 2/9] mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

Quoted 23 lines · Mike Kravetz
Quoted 13 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

Separate the if branch into one for hwpoison entries and one for migration entries.

Quoted 6 lines
Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Cc <[email protected]> # 5.18
---
 mm/hugetlb.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Thank you!

Quoted 7 lines

(with typos pointed out by Miaohe Lin)

Just curious, are there any hwpoisoned tests in any test suite? I run libhugetlbfs tests and ltp on a regular basis which sometimes catch regressions. If there are no tests in any suite today, this might be something we want to consider for future work.

mce-tests (https://git.kernel.org/pub/scm/utils/cpu/mce/mce-test.git) has some test cases about hwpoison (under cases/function/hwpoison/), but this tool mostly focuses on MCE and does not have enough hwpoison testcases. So I'm maintaining my own test suite
(https://github.com/nhoriguchi/mm_regression) for long to help my own hwpoison development/maintenance. I'd like to make this tool more handy so that other developers or testsuites can easily run the hwpoison testing, although I need more work for it.

Thanks,
Naoya Horiguchi

Reviewed-by on v2 2/9
#17 lore raw

Re: [PATCH v2 2/9] mm/hugetlb: separate path for hwpoison entry in copy_hugetlb_page_range()

In reply to Naoya Horiguchi
Quoted 10 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Originally copy_hugetlb_page_range() handles migration entries and hwpoisone entries in similar manner. But recently the related code path has more code for migration entries, and when is_writable_migration_entry() was converted to !is_readable_migration_entry(), hwpoison entries on source processes got to be unexpectedly updated (which is legitimate for migration entries, but not for hwpoison entries). This results in unexpected serious issues like kernel panic when foking processes with hwpoison entries in pmd.

Quoted 5 lines

Separate the if branch into one for hwpoison entries and one for migration entries.

Fixes 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")

Thanks for fixing it.

author reviewer
#18 lore raw

PATCH v2 3/9 mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

---
 arch/x86/mm/hugetlbpage.c |  3 ++-
 mm/hugetlb.c              | 26 +++++++++++++++++++++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)
arch/x86/mm/hugetlbpage.c +2 -1
index a0d023cb4292..5fb86fb49ba8 100644
@@ -70,7 +70,8 @@ int pmd_huge(pmd_t pmd)
 
 int pud_huge(pud_t pud)
 {
-	return !!(pud_val(pud) & _PAGE_PSE);
+	return !pud_none(pud) &&
+		(pud_val(pud) & (_PAGE_PRESENT|_PAGE_PSE)) != _PAGE_PRESENT;
 }
 #endif
 
mm/hugetlb.c +25 -1
index f59f43c06601..b7ae5f73f3b2 100644
@@ -6946,10 +6946,34 @@ struct page * __weak
 follow_huge_pud(struct mm_struct *mm, unsigned long address,
 		pud_t *pud, int flags)
 {
+	struct page *page = NULL;
+	spinlock_t *ptl;
+	pte_t pte;
+
 	if (flags & (FOLL_GET | FOLL_PIN))
 		return NULL;
 
-	return pte_page(*(pte_t *)pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+retry:
+	ptl = huge_pte_lock(hstate_sizelog(PUD_SHIFT), mm, (pte_t *)pud);
+	if (!pud_huge(*pud))
+		goto out;
+	pte = huge_ptep_get((pte_t *)pud);
+	if (pte_present(pte)) {
+		page = pud_page(*pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+		if (WARN_ON_ONCE(!try_grab_page(page, flags))) {
+			page = NULL;
+			goto out;
+		}
+	} else {
+		if (is_hugetlb_entry_migration(pte)) {
+			spin_unlock(ptl);
+			__migration_entry_wait(mm, (pte_t *)pud, ptl);
+			goto retry;
+		}
+	}
+out:
+	spin_unlock(ptl);
+	return page;
 }
 
 struct page * __weak
Signature · 1 line
2.25.1
#19 lore raw

Re: [PATCH v2 3/9] mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

Sorry, I found that $SUBJECT mentions wrong function name (I meant follow_huge_pud(), not huge_pud()), this will be fixed in the later version.

Quoted 9 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The

here the same typo, too.

- Naoya Horiguchi

Quoted 6 lines

changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

---

#20 lore raw

Re: [PATCH v2 3/9] mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

In reply to Naoya Horiguchi
Quoted 13 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

Quoted 5 lines
---
 arch/x86/mm/hugetlbpage.c |  3 ++-
 mm/hugetlb.c              | 26 +++++++++++++++++++++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)

Thanks. Overall looks good with typos corrected that you already noticed. One question below.

Quoted 29 lines
arch/x86/mm/hugetlbpage.c +2 -1
index a0d023cb4292..5fb86fb49ba8 100644
@@ -70,7 +70,8 @@ int pmd_huge(pmd_t pmd)
 
 int pud_huge(pud_t pud)
 {
-	return !!(pud_val(pud) & _PAGE_PSE);
+	return !pud_none(pud) &&
+		(pud_val(pud) & (_PAGE_PRESENT|_PAGE_PSE)) != _PAGE_PRESENT;
 }
 #endif
 
mm/hugetlb.c +4 -0
index f59f43c06601..b7ae5f73f3b2 100644
@@ -6946,10 +6946,34 @@ struct page * __weak
 follow_huge_pud(struct mm_struct *mm, unsigned long address,
 		pud_t *pud, int flags)
 {
+	struct page *page = NULL;
+	spinlock_t *ptl;
+	pte_t pte;
+
 	if (flags & (FOLL_GET | FOLL_PIN))
 		return NULL;

mm/hugetlb.c struct page * __weak · line 6963

Quoted 10 lines
mm/hugetlb.c +8 -1
 
-	return pte_page(*(pte_t *)pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+retry:
+	ptl = huge_pte_lock(hstate_sizelog(PUD_SHIFT), mm, (pte_t *)pud);
+	if (!pud_huge(*pud))
+		goto out;
+	pte = huge_ptep_get((pte_t *)pud);
+	if (pte_present(pte)) {
+		page = pud_page(*pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+		if (WARN_ON_ONCE(!try_grab_page(page, flags))) {

The call to try_grab_page() seems a bit strange since flags will not include FOLL_GET | FOLL_PIN as tested above. Is try_grab_page() always going be a noop here?

Signature · 1 line
Mike Kravetz
Quoted 19 lines
mm/hugetlb.c +13 -0
+			page = NULL;
+			goto out;
+		}
+	} else {
+		if (is_hugetlb_entry_migration(pte)) {
+			spin_unlock(ptl);
+			__migration_entry_wait(mm, (pte_t *)pud, ptl);
+			goto retry;
+		}
+	}
+out:
+	spin_unlock(ptl);
+	return page;
 }
 
 struct page * __weak
Signature · 1 line
2.25.1
#21 lore raw

Re: [PATCH v2 3/9] mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

Quoted 52 lines · Mike Kravetz
Quoted 13 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

Quoted 5 lines
---
 arch/x86/mm/hugetlbpage.c |  3 ++-
 mm/hugetlb.c              | 26 +++++++++++++++++++++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)

Thanks. Overall looks good with typos corrected that you already noticed. One question below.

Quoted 30 lines
arch/x86/mm/hugetlbpage.c +2 -1
index a0d023cb4292..5fb86fb49ba8 100644
@@ -70,7 +70,8 @@ int pmd_huge(pmd_t pmd)
 
 int pud_huge(pud_t pud)
 {
-	return !!(pud_val(pud) & _PAGE_PSE);
+	return !pud_none(pud) &&
+		(pud_val(pud) & (_PAGE_PRESENT|_PAGE_PSE)) != _PAGE_PRESENT;
 }
 #endif
 
mm/hugetlb.c +4 -0
index f59f43c06601..b7ae5f73f3b2 100644
@@ -6946,10 +6946,34 @@ struct page * __weak
 follow_huge_pud(struct mm_struct *mm, unsigned long address,
 		pud_t *pud, int flags)
 {
+	struct page *page = NULL;
+	spinlock_t *ptl;
+	pte_t pte;
+
 	if (flags & (FOLL_GET | FOLL_PIN))
 		return NULL;
 
Quoted 13 lines

mm/hugetlb.c struct page * __weak · line 6963

Quoted 9 lines
mm/hugetlb.c +8 -1
-	return pte_page(*(pte_t *)pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+retry:
+	ptl = huge_pte_lock(hstate_sizelog(PUD_SHIFT), mm, (pte_t *)pud);
+	if (!pud_huge(*pud))
+		goto out;
+	pte = huge_ptep_get((pte_t *)pud);
+	if (pte_present(pte)) {
+		page = pud_page(*pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+		if (WARN_ON_ONCE(!try_grab_page(page, flags))) {

The call to try_grab_page() seems a bit strange since flags will not include FOLL_GET | FOLL_PIN as tested above. Is try_grab_page() always going be a noop here?

Thanks, you're right. "flags & (FOLL_GET | FOLL_PIN)" check should be updated. Probably it's to be aligned to what follow_huge_pmd() checks first.

Thanks,
Naoya Horiguchi

Quoted 23 lines
Signature · 2 lines
-- 
Mike Kravetz
Quoted 19 lines
mm/hugetlb.c +13 -0
+			page = NULL;
+			goto out;
+		}
+	} else {
+		if (is_hugetlb_entry_migration(pte)) {
+			spin_unlock(ptl);
+			__migration_entry_wait(mm, (pte_t *)pud, ptl);
+			goto retry;
+		}
+	}
+out:
+	spin_unlock(ptl);
+	return page;
 }
 
 struct page * __weak
Signature · 1 line
2.25.1
maintainer
#22 lore raw

Re: [PATCH v2 3/9] mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

In reply to Naoya Horiguchi
Quoted 24 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

---
 arch/x86/mm/hugetlbpage.c |  3 ++-
 mm/hugetlb.c              | 26 +++++++++++++++++++++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)
arch/x86/mm/hugetlbpage.c +0 -0
index a0d023cb4292..5fb86fb49ba8 100644

arch/x86/mm/hugetlbpage.c int pmd_huge(pmd_t pmd) · line 70

Quoted 1 line
arch/x86/mm/hugetlbpage.c +0 -0
@@ -70,7 +70,8 @@ int pmd_huge(pmd_t pmd)
 

No strong opinion but a comment similar to pmd_huge might be better?

/*
* pmd_huge() returns 1 if @pmd is hugetlb related entry, that is normal
* hugetlb entry or non-present (migration or hwpoisoned) hugetlb entry.
* Otherwise, returns 0.
*/

Quoted 13 lines
arch/x86/mm/hugetlbpage.c +2 -1
 int pud_huge(pud_t pud)
 {
-	return !!(pud_val(pud) & _PAGE_PSE);
+	return !pud_none(pud) &&
+		(pud_val(pud) & (_PAGE_PRESENT|_PAGE_PSE)) != _PAGE_PRESENT;
 }
 #endif
 
mm/hugetlb.c +0 -0
index f59f43c06601..b7ae5f73f3b2 100644

mm/hugetlb.c struct page * __weak · line 6954

Quoted 9 lines
mm/hugetlb.c +4 -0
@@ -6946,10 +6946,34 @@ struct page * __weak
 follow_huge_pud(struct mm_struct *mm, unsigned long address,
 		pud_t *pud, int flags)
 {
+	struct page *page = NULL;
+	spinlock_t *ptl;
+	pte_t pte;
+
 	if (flags & (FOLL_GET | FOLL_PIN))
 		return NULL;

Should the above check be modified? It seems the below try_grab_page might not grab the page as expected (as Mike pointed out). Or the extra page refcnt is unneeded?

Quoted 9 lines
mm/hugetlb.c +7 -1
-	return pte_page(*(pte_t *)pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+retry:
+	ptl = huge_pte_lock(hstate_sizelog(PUD_SHIFT), mm, (pte_t *)pud);
+	if (!pud_huge(*pud))
+		goto out;
+	pte = huge_ptep_get((pte_t *)pud);
+	if (pte_present(pte)) {
+		page = pud_page(*pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);

mm/hugetlb.c struct page * __weak · line 6972

Quoted 10 lines
mm/hugetlb.c +10 -0
+		if (WARN_ON_ONCE(!try_grab_page(page, flags))) {
+			page = NULL;
+			goto out;
+		}
+	} else {
+		if (is_hugetlb_entry_migration(pte)) {
+			spin_unlock(ptl);
+			__migration_entry_wait(mm, (pte_t *)pud, ptl);
+			goto retry;
+		}

Again. No strong opinion but a comment similar to follow_huge_pmd might be better?

/*
* hwpoisoned entry is treated as no_page_table in
* follow_page_mask().
*/

Thanks!

Quoted 8 lines
mm/hugetlb.c +4 -0
+	}
+out:
+	spin_unlock(ptl);
+	return page;
 }
 
 struct page * __weak
 
#23 lore raw

Re: [PATCH v2 3/9] mm/hugetlb: make pud_huge() and huge_pud() aware of non-present pud entry

Quoted 29 lines · Miaohe Lin
Quoted 24 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

follow_pud_mask() does not support non-present pud entry now. As long as I tested on x86_64 server, follow_pud_mask() still simply returns no_page_table() for non-present_pud_entry() due to pud_bad(), so no severe user-visible effect should happen. But generally we should call follow_huge_pud() for non-present pud entry for 1GB hugetlb page.

Update pud_huge() and huge_pud() to handle non-present pud entries. The changes are similar to previous works for pud entries commit e66f17ff7177 ("mm/hugetlb: take page table lock in follow_huge_pmd()") and commit cbef8478bee5 ("mm/hugetlb: pmd_huge() returns true for non-present hugepage").

---
 arch/x86/mm/hugetlbpage.c |  3 ++-
 mm/hugetlb.c              | 26 +++++++++++++++++++++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)
arch/x86/mm/hugetlbpage.c +0 -0
index a0d023cb4292..5fb86fb49ba8 100644

arch/x86/mm/hugetlbpage.c int pmd_huge(pmd_t pmd) · line 70

Quoted 1 line
arch/x86/mm/hugetlbpage.c +0 -0
@@ -70,7 +70,8 @@ int pmd_huge(pmd_t pmd)
 

No strong opinion but a comment similar to pmd_huge might be better?

Quoted 5 lines

/*
* pmd_huge() returns 1 if @pmd is hugetlb related entry, that is normal
* hugetlb entry or non-present (migration or hwpoisoned) hugetlb entry.
* Otherwise, returns 0.
*/

OK, I'll add some.

Quoted 18 lines
Quoted 18 lines
arch/x86/mm/hugetlbpage.c +2 -1
 int pud_huge(pud_t pud)
 {
-	return !!(pud_val(pud) & _PAGE_PSE);
+	return !pud_none(pud) &&
+		(pud_val(pud) & (_PAGE_PRESENT|_PAGE_PSE)) != _PAGE_PRESENT;
 }
 #endif
 
mm/hugetlb.c +1 -0
index f59f43c06601..b7ae5f73f3b2 100644
@@ -6946,10 +6946,34 @@ struct page * __weak
 follow_huge_pud(struct mm_struct *mm, unsigned long address,
 		pud_t *pud, int flags)
 {
+	struct page *page = NULL;
Quoted 8 lines

mm/hugetlb.c struct page * __weak · line 6954

Quoted 5 lines
mm/hugetlb.c +3 -0
+	spinlock_t *ptl;
+	pte_t pte;
+
 	if (flags & (FOLL_GET | FOLL_PIN))
 		return NULL;

Should the above check be modified? It seems the below try_grab_page might not grab the page as expected (as Mike pointed out). Or the extra page refcnt is unneeded?

Yes, this check should be updated.

Quoted 23 lines
Quoted 10 lines
mm/hugetlb.c +7 -1
-	return pte_page(*(pte_t *)pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);
+retry:
+	ptl = huge_pte_lock(hstate_sizelog(PUD_SHIFT), mm, (pte_t *)pud);
+	if (!pud_huge(*pud))
+		goto out;
+	pte = huge_ptep_get((pte_t *)pud);
+	if (pte_present(pte)) {
+		page = pud_page(*pud) + ((address & ~PUD_MASK) >> PAGE_SHIFT);

mm/hugetlb.c struct page * __weak · line 6972

Quoted 10 lines
mm/hugetlb.c +10 -0
+		if (WARN_ON_ONCE(!try_grab_page(page, flags))) {
+			page = NULL;
+			goto out;
+		}
+	} else {
+		if (is_hugetlb_entry_migration(pte)) {
+			spin_unlock(ptl);
+			__migration_entry_wait(mm, (pte_t *)pud, ptl);
+			goto retry;
+		}

Again. No strong opinion but a comment similar to follow_huge_pmd might be better?

Quoted 4 lines

/*
* hwpoisoned entry is treated as no_page_table in
* follow_page_mask().
*/

Will add comment on this too. Thank you.

- Naoya Horiguchi

author reviewer
#24 lore raw

PATCH v2 4/9 mm, hwpoison, hugetlb: support saving mechanism of raw error pages

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

---
v1 -> v2:
- support hwpoison hugepage with multiple errors,
- moved the new interface functions to mm/memory-failure.c,
- define additional subpage index SUBPAGE_INDEX_HWPOISON_UNRELIABLE,
- stop freeing/dissolving hwpoison hugepages with unreliable raw error info,
- drop hugetlb_clear_page_hwpoison() in dissolve_free_huge_page() because that's done in update_and_free_page(),
- move setting/clearing PG_hwpoison flag to the new interfaces,
- checking already hwpoisoned or not on a subpage basis.

ChangeLog since previous post on 4/27:
- fixed typo in patch description (by Miaohe)
- fixed config value in #ifdef statement (by Miaohe)
- added sentences about "multiple hwpoison pages" scenario in patch description

---
 include/linux/hugetlb.h | 13 ++++++
 mm/hugetlb.c            | 39 +++++++++--------
 mm/memory-failure.c     | 95 ++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 125 insertions(+), 22 deletions(-)
include/linux/hugetlb.h +13 -0
index e4cff27d1198..ac13c2022517 100644
@@ -42,6 +42,10 @@ enum {
 	SUBPAGE_INDEX_CGROUP,		/* reuse page->private */
 	SUBPAGE_INDEX_CGROUP_RSVD,	/* reuse page->private */
 	__MAX_CGROUP_SUBPAGE_INDEX = SUBPAGE_INDEX_CGROUP_RSVD,
+#endif
+#ifdef CONFIG_MEMORY_FAILURE
+	SUBPAGE_INDEX_HWPOISON,
+	SUBPAGE_INDEX_HWPOISON_UNRELIABLE,
 #endif
 	__NR_USED_SUBPAGE,
 };
@@ -798,6 +802,15 @@ extern int dissolve_free_huge_page(struct page *page);
 extern int dissolve_free_huge_pages(unsigned long start_pfn,
 				    unsigned long end_pfn);
 
+#ifdef CONFIG_MEMORY_FAILURE
+extern int hugetlb_clear_page_hwpoison(struct page *hpage);
+#else
+static inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
+	return 0;
+}
+#endif
+
 #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION
 #ifndef arch_hugetlb_migration_supported
 static inline bool arch_hugetlb_migration_supported(struct hstate *h)
mm/hugetlb.c +19 -20
index b7ae5f73f3b2..19ef427aa1e8 100644
@@ -1541,17 +1541,15 @@ static void __update_and_free_page(struct hstate *h, struct page *page)
 	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
 		return;
 
-	if (hugetlb_vmemmap_alloc(h, page)) {
-		spin_lock_irq(&hugetlb_lock);
-		/*
-		 * If we cannot allocate vmemmap pages, just refuse to free the
-		 * page and put the page back on the hugetlb free list and treat
-		 * as a surplus page.
-		 */
-		add_hugetlb_page(h, page, true);
-		spin_unlock_irq(&hugetlb_lock);
-		return;
-	}
+	if (hugetlb_vmemmap_alloc(h, page))
+		goto fail;
+
+	/*
+	 * Move PageHWPoison flag from head page to the raw error pages,
+	 * which makes any healthy subpages reusable.
+	 */
+	if (unlikely(PageHWPoison(page) && hugetlb_clear_page_hwpoison(page)))
+		goto fail;
 
 	for (i = 0; i < pages_per_huge_page(h);
 	     i++, subpage = mem_map_next(subpage, page, i)) {
@@ -1572,6 +1570,16 @@ static void __update_and_free_page(struct hstate *h, struct page *page)
 	} else {
 		__free_pages(page, huge_page_order(h));
 	}
+	return;
+fail:
+	spin_lock_irq(&hugetlb_lock);
+	/*
+	 * If we cannot allocate vmemmap pages or cannot identify raw hwpoison
+	 * subpages reliably, just refuse to free the page and put the page
+	 * back on the hugetlb free list and treat as a surplus page.
+	 */
+	add_hugetlb_page(h, page, true);
+	spin_unlock_irq(&hugetlb_lock);
 }
 
 /*
@@ -2115,15 +2123,6 @@ int dissolve_free_huge_page(struct page *page)
 		 */
 		rc = hugetlb_vmemmap_alloc(h, head);
 		if (!rc) {
-			/*
-			 * Move PageHWPoison flag from head page to the raw
-			 * error page, which makes any subpages rather than
-			 * the error page reusable.
-			 */
-			if (PageHWPoison(head) && page != head) {
-				SetPageHWPoison(page);
-				ClearPageHWPoison(head);
-			}
 			update_and_free_page(h, head, false);
 		} else {
 			spin_lock_irq(&hugetlb_lock);
mm/memory-failure.c +93 -2

113 lines not on this page · raw message

Signature · 1 line
2.25.1
maintainer
#25 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Many thanks for your patch. This patch looks good to me. Some nits below.

Quoted 1 line

<snip>

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1540

Quoted 12 lines
mm/memory-failure.c +11 -0
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,

It should be s/mean/meaning/ ?

Quoted 19 lines
mm/memory-failure.c +19 -0
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);
+	if (raw_hwp) {

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1569

Quoted 10 lines
mm/memory-failure.c +10 -0
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;

This "return ret" can be combined into the below one?

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 11 lines
mm/memory-failure.c +11 -0
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

Can we try freeing the memory of raw_hwp_list to save possible memory? It seems raw_hwp_list becomes unneeded when raw_hwp_unreliable.

Thanks.

Quoted 34 lines
mm/memory-failure.c +14 -2
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		SetPageHWPoison(p->page);
+		kfree(p);
+	}
+	llist_del_all(head);
+	return 0;
+}
+
 /*
  * Called from hugetlb code with hugetlb_lock held.
  *
@@ -1533,7 +1624,7 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 		goto out;
 	}
 
-	if (TestSetPageHWPoison(head)) {
+	if (hugetlb_set_page_hwpoison(head, page)) {
 		ret = -EHWPOISON;
 		goto out;
 	}
@@ -1585,7 +1676,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	lock_page(head);
 
 	if (hwpoison_filter(p)) {
-		ClearPageHWPoison(head);
+		hugetlb_clear_page_hwpoison(head);
 		res = -EOPNOTSUPP;
 		goto out;
 	}
#26 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 30 lines · Miaohe Lin
Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Many thanks for your patch. This patch looks good to me. Some nits below.

Quoted 1 line

<snip>

Quoted 1 line
Quoted 13 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1540

Quoted 11 lines
mm/memory-failure.c +11 -0
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,

It should be s/mean/meaning/ ?

Right, fixed.

Quoted 24 lines
Quoted 24 lines
mm/memory-failure.c +23 -0
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);
+	if (raw_hwp) {
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);
+	} else {
+		/*
Quoted 9 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1569

Quoted 6 lines
mm/memory-failure.c +6 -0
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;

This "return ret" can be combined into the below one?

Right, fixed.

Quoted 6 lines
Quoted 6 lines
mm/memory-failure.c +6 -0
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

Can we try freeing the memory of raw_hwp_list to save possible memory? It seems raw_hwp_list becomes unneeded when raw_hwp_unreliable.

Yes, we can. I try to change like that.
Thanks for the suggestion.

- Naoya Horiguchi

#27 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

In reply to Naoya Horiguchi
Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Thanks for your work on this. I have several quesions below.

Quoted 141 lines

---
v1 -> v2:
- support hwpoison hugepage with multiple errors,
- moved the new interface functions to mm/memory-failure.c,
- define additional subpage index SUBPAGE_INDEX_HWPOISON_UNRELIABLE,
- stop freeing/dissolving hwpoison hugepages with unreliable raw error info,
- drop hugetlb_clear_page_hwpoison() in dissolve_free_huge_page() because that's done in update_and_free_page(),
- move setting/clearing PG_hwpoison flag to the new interfaces,
- checking already hwpoisoned or not on a subpage basis.

ChangeLog since previous post on 4/27:
- fixed typo in patch description (by Miaohe)
- fixed config value in #ifdef statement (by Miaohe)
- added sentences about "multiple hwpoison pages" scenario in patch description

---
 include/linux/hugetlb.h | 13 ++++++
 mm/hugetlb.c            | 39 +++++++++--------
 mm/memory-failure.c     | 95 ++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 125 insertions(+), 22 deletions(-)
include/linux/hugetlb.h +13 -0
index e4cff27d1198..ac13c2022517 100644
@@ -42,6 +42,10 @@ enum {
 	SUBPAGE_INDEX_CGROUP,		/* reuse page->private */
 	SUBPAGE_INDEX_CGROUP_RSVD,	/* reuse page->private */
 	__MAX_CGROUP_SUBPAGE_INDEX = SUBPAGE_INDEX_CGROUP_RSVD,
+#endif
+#ifdef CONFIG_MEMORY_FAILURE
+	SUBPAGE_INDEX_HWPOISON,
+	SUBPAGE_INDEX_HWPOISON_UNRELIABLE,
 #endif
 	__NR_USED_SUBPAGE,
 };
@@ -798,6 +802,15 @@ extern int dissolve_free_huge_page(struct page *page);
 extern int dissolve_free_huge_pages(unsigned long start_pfn,
 				    unsigned long end_pfn);
 
+#ifdef CONFIG_MEMORY_FAILURE
+extern int hugetlb_clear_page_hwpoison(struct page *hpage);
+#else
+static inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
+	return 0;
+}
+#endif
+
 #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION
 #ifndef arch_hugetlb_migration_supported
 static inline bool arch_hugetlb_migration_supported(struct hstate *h)
mm/hugetlb.c +19 -20
index b7ae5f73f3b2..19ef427aa1e8 100644
@@ -1541,17 +1541,15 @@ static void __update_and_free_page(struct hstate *h, struct page *page)
 	if (hstate_is_gigantic(h) && !gigantic_page_runtime_supported())
 		return;
 
-	if (hugetlb_vmemmap_alloc(h, page)) {
-		spin_lock_irq(&hugetlb_lock);
-		/*
-		 * If we cannot allocate vmemmap pages, just refuse to free the
-		 * page and put the page back on the hugetlb free list and treat
-		 * as a surplus page.
-		 */
-		add_hugetlb_page(h, page, true);
-		spin_unlock_irq(&hugetlb_lock);
-		return;
-	}
+	if (hugetlb_vmemmap_alloc(h, page))
+		goto fail;
+
+	/*
+	 * Move PageHWPoison flag from head page to the raw error pages,
+	 * which makes any healthy subpages reusable.
+	 */
+	if (unlikely(PageHWPoison(page) && hugetlb_clear_page_hwpoison(page)))
+		goto fail;
 
 	for (i = 0; i < pages_per_huge_page(h);
 	     i++, subpage = mem_map_next(subpage, page, i)) {
@@ -1572,6 +1570,16 @@ static void __update_and_free_page(struct hstate *h, struct page *page)
 	} else {
 		__free_pages(page, huge_page_order(h));
 	}
+	return;
+fail:
+	spin_lock_irq(&hugetlb_lock);
+	/*
+	 * If we cannot allocate vmemmap pages or cannot identify raw hwpoison
+	 * subpages reliably, just refuse to free the page and put the page
+	 * back on the hugetlb free list and treat as a surplus page.
+	 */
+	add_hugetlb_page(h, page, true);
+	spin_unlock_irq(&hugetlb_lock);
 }
 
 /*
@@ -2115,15 +2123,6 @@ int dissolve_free_huge_page(struct page *page)
 		 */
 		rc = hugetlb_vmemmap_alloc(h, head);
 		if (!rc) {
-			/*
-			 * Move PageHWPoison flag from head page to the raw
-			 * error page, which makes any subpages rather than
-			 * the error page reusable.
-			 */
-			if (PageHWPoison(head) && page != head) {
-				SetPageHWPoison(page);
-				ClearPageHWPoison(head);
-			}
 			update_and_free_page(h, head, false);
 		} else {
 			spin_lock_irq(&hugetlb_lock);
mm/memory-failure.c +14 -0
index fb3feb1f363e..af571fa6f2af 100644
@@ -1499,6 +1499,97 @@ static int try_to_split_thp_page(struct page *page, const char *msg)
 }
 
 #ifdef CONFIG_HUGETLB_PAGE
+/*
+ * Struct raw_hwp_page represents information about "raw error page",
+ * constructing singly linked list originated from ->private field of
+ * SUBPAGE_INDEX_HWPOISON-th tail page.
+ */
+struct raw_hwp_page {
+	struct llist_node node;
+	struct page *page;
+};
+
+static inline struct llist_head *raw_hwp_list_head(struct page *hpage)
+{
+	return (struct llist_head *)&page_private(hpage + SUBPAGE_INDEX_HWPOISON);
+}

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1525

Quoted 10 lines
mm/memory-failure.c +10 -0
+
+static inline int raw_hwp_unreliable(struct page *hpage)
+{
+	return page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE);
+}
+
+static inline void set_raw_hwp_unreliable(struct page *hpage)
+{
+	set_page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE, 1);
+}

Why not use HPAGEFLAG(HwpoisonUnreliable, hwpoison_unreliable) directly?

Quoted 23 lines
mm/memory-failure.c +23 -0
+
+/*
+ * about race consideration
+ */
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1558

Quoted 10 lines
mm/memory-failure.c +10 -0
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);

This function can be called in atomic context, GFP_ATOMIC should be used here.

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1561

Quoted 3 lines
mm/memory-failure.c +3 -0
+	if (raw_hwp) {
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);

The maximum amount of items in the list is one, right?

Quoted 9 lines
mm/memory-failure.c +9 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 10 lines
mm/memory-failure.c +10 -0
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 3 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1589

Quoted 6 lines
mm/memory-failure.c +6 -0
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		SetPageHWPoison(p->page);
+		kfree(p);
+	}
+	llist_del_all(head);

If the above issue exists, moving ClearPageHWPoison(hpage) to here could fix it. We should clear PageHWPoison carefully since the head page is also can be poisoned.

Thanks.

Quoted 28 lines
mm/memory-failure.c +5 -2
+	return 0;
+}
+
 /*
  * Called from hugetlb code with hugetlb_lock held.
  *
@@ -1533,7 +1624,7 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 		goto out;
 	}
 
-	if (TestSetPageHWPoison(head)) {
+	if (hugetlb_set_page_hwpoison(head, page)) {
 		ret = -EHWPOISON;
 		goto out;
 	}
@@ -1585,7 +1676,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	lock_page(head);
 
 	if (hwpoison_filter(p)) {
-		ClearPageHWPoison(head);
+		hugetlb_clear_page_hwpoison(head);
 		res = -EOPNOTSUPP;
 		goto out;
 	}
Signature · 1 line
2.25.1
#28 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 27 lines · Muchun Song
Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Thanks for your work on this. I have several quesions below.

...

Quoted 22 lines
Quoted 22 lines
mm/memory-failure.c +18 -0
@@ -1499,6 +1499,97 @@ static int try_to_split_thp_page(struct page *page, const char *msg)
 }
 
 #ifdef CONFIG_HUGETLB_PAGE
+/*
+ * Struct raw_hwp_page represents information about "raw error page",
+ * constructing singly linked list originated from ->private field of
+ * SUBPAGE_INDEX_HWPOISON-th tail page.
+ */
+struct raw_hwp_page {
+	struct llist_node node;
+	struct page *page;
+};
+
+static inline struct llist_head *raw_hwp_list_head(struct page *hpage)
+{
+	return (struct llist_head *)&page_private(hpage + SUBPAGE_INDEX_HWPOISON);
+}
+
+static inline int raw_hwp_unreliable(struct page *hpage)
+{
+	return page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE);
Quoted 9 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1525

Quoted 6 lines
mm/memory-failure.c +6 -0
+}
+
+static inline void set_raw_hwp_unreliable(struct page *hpage)
+{
+	set_page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE, 1);
+}

Why not use HPAGEFLAG(HwpoisonUnreliable, hwpoison_unreliable) directly?

OK, that sounds better, I'll do it.

Quoted 28 lines
Quoted 28 lines
mm/memory-failure.c +28 -0
+
+/*
+ * about race consideration
+ */
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1558

Quoted 5 lines
mm/memory-failure.c +5 -0
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);

This function can be called in atomic context, GFP_ATOMIC should be used here.

OK, I'll use GFP_ATOMIC.

Quoted 6 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1561

Quoted 4 lines
mm/memory-failure.c +3 -0
+	if (raw_hwp) {
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);

The maximum amount of items in the list is one, right?

The maximum is the number of subpages in the hugepage (512 for 2MB hugepage, 262144 for 1GB hugepage).

Quoted 15 lines
Quoted 15 lines
mm/memory-failure.c +14 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

Sorry if I might miss your point, but raw_hwp_unreliable is set when allocating raw_hwp_page failed. hugetlb_set_page_hwpoison() can be called multiple times on a hugepage and if one of the calls fails, the hwpoisoned hugepage becomes unreliable.

BTW, as you pointed out above, if we switch to passing GFP_ATOMIC to kmalloc(), the kmalloc() never fails, so we no longer have to implement this unreliable flag, so things get simpler.

Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 4 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

Maybe you are mentioning the race like below. Yes, that's possible.

CPU 0 CPU 1

free_huge_page
lock hugetlb_lock ClearHPageMigratable unlock hugetlb_lock
get_huge_page_for_hwpoison

    lock hugetlb_lock
    __get_huge_page_for_hwpoison
      hugetlb_set_page_hwpoison
        allocate raw_hwp_page
        TestSetPageHWPoison
                                     update_and_free_page
                                       __update_and_free_page
                                         if (PageHWPoison)
                                           hugetlb_clear_page_hwpoison
                                             TestClearPageHWPoison
                                             // remove all list items
        llist_add
    unlock hugetlb_lock

The end result seems not critical (leaking raced raw_hwp_page?), but we need fix.

Quoted 11 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1589

Quoted 7 lines
mm/memory-failure.c +6 -0
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		SetPageHWPoison(p->page);
+		kfree(p);
+	}
+	llist_del_all(head);

If the above issue exists, moving ClearPageHWPoison(hpage) to here could fix it. We should clear PageHWPoison carefully since the head page is also can be poisoned.

The reason why I put ClearPageHWPoison(hpage) before llist_for_each_safe() was that raw error page can be the head page. But this can be solved with some additional code to remember whether raw_hwp_page list has an item associated with the head page.

Or another approach in my mind now is to call hugetlb_clear_page_hwpoison() with taking mf_mutex.

Quoted 2 lines

Thanks.

Thank you for valuable feedbacks.

- Naoya Horiguchi

Quoted 29 lines
Quoted 29 lines
mm/memory-failure.c +5 -2
+	return 0;
+}
+
 /*
  * Called from hugetlb code with hugetlb_lock held.
  *
@@ -1533,7 +1624,7 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 		goto out;
 	}
 
-	if (TestSetPageHWPoison(head)) {
+	if (hugetlb_set_page_hwpoison(head, page)) {
 		ret = -EHWPOISON;
 		goto out;
 	}
@@ -1585,7 +1676,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	lock_page(head);
 
 	if (hwpoison_filter(p)) {
-		ClearPageHWPoison(head);
+		hugetlb_clear_page_hwpoison(head);
 		res = -EOPNOTSUPP;
 		goto out;
 	}
Signature · 1 line
2.25.1
#29 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 104 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 27 lines · Muchun Song
Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Thanks for your work on this. I have several quesions below.

...

Quoted 22 lines
Quoted 22 lines
mm/memory-failure.c +18 -0
@@ -1499,6 +1499,97 @@ static int try_to_split_thp_page(struct page *page, const char *msg)
 }
 
 #ifdef CONFIG_HUGETLB_PAGE
+/*
+ * Struct raw_hwp_page represents information about "raw error page",
+ * constructing singly linked list originated from ->private field of
+ * SUBPAGE_INDEX_HWPOISON-th tail page.
+ */
+struct raw_hwp_page {
+	struct llist_node node;
+	struct page *page;
+};
+
+static inline struct llist_head *raw_hwp_list_head(struct page *hpage)
+{
+	return (struct llist_head *)&page_private(hpage + SUBPAGE_INDEX_HWPOISON);
+}
+
+static inline int raw_hwp_unreliable(struct page *hpage)
+{
+	return page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE);
Quoted 9 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1525

Quoted 6 lines
mm/memory-failure.c +6 -0
+}
+
+static inline void set_raw_hwp_unreliable(struct page *hpage)
+{
+	set_page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE, 1);
+}

Why not use HPAGEFLAG(HwpoisonUnreliable, hwpoison_unreliable) directly?

OK, that sounds better, I'll do it.

Quoted 28 lines
Quoted 28 lines
mm/memory-failure.c +28 -0
+
+/*
+ * about race consideration
+ */
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1558

Quoted 5 lines
mm/memory-failure.c +5 -0
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);

This function can be called in atomic context, GFP_ATOMIC should be used here.

OK, I'll use GFP_ATOMIC.

Quoted 1 line
Quoted 9 lines
Quoted 5 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1561

Quoted 3 lines
mm/memory-failure.c +3 -0
+	if (raw_hwp) {
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);

The maximum amount of items in the list is one, right?

The maximum is the number of subpages in the hugepage (512 for 2MB hugepage, 262144 for 1GB hugepage).

You are right. I have missed something yesterday.

Quoted 24 lines
Quoted 15 lines
Quoted 15 lines
mm/memory-failure.c +14 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

Quoted 2 lines

Sorry if I might miss your point, but raw_hwp_unreliable is set when allocating raw_hwp_page failed. hugetlb_set_page_hwpoison() can be called

Sorry. I have missed this. Thanks for your clarification.

Quoted 5 lines

multiple times on a hugepage and if one of the calls fails, the hwpoisoned hugepage becomes unreliable.

BTW, as you pointed out above, if we switch to passing GFP_ATOMIC to kmalloc(), the kmalloc() never fails, so we no longer have to implement this unreliable

No. kmalloc() with GFP_ATOMIC can fail unless I miss something important.

Quoted 11 lines

flag, so things get simpler.

Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 4 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

Quoted 2 lines

Maybe you are mentioning the race like below. Yes, that's possible.

Sorry, ignore my previous comments, I'm thinking something wrong.

Quoted 5 lines

CPU 0 CPU 1

free_huge_page
lock hugetlb_lock ClearHPageMigratable

				       remove_hugetlb_page()
				       // the page is non-HugeTLB now
Quoted 4 lines

unlock hugetlb_lock
get_huge_page_for_hwpoison
lock hugetlb_lock __get_huge_page_for_hwpoison

	// cannot reach here since it is not a HugeTLB page now.
	// So this race is impossible. Then we fallback to normal
	// page handling. Seems there is a new issue here.
	//
	// memory_failure()
	//	try_memory_failure_hugetlb()
	//	if (hugetlb)
	//		goto unlock_mutex;
	//	if (TestSetPageHWPoison(p)) {
	//	// This non-HugeTLB page's vmemmap is still optimized.

Setting COMPOUND_PAGE_DTOR after hugetlb_vmemmap_restore() might fix this issue, but we will encounter this race as you mentioned below.

Thanks.

Quoted 72 lines
      hugetlb_set_page_hwpoison
        allocate raw_hwp_page
        TestSetPageHWPoison
                                     update_and_free_page
                                       __update_and_free_page
                                         if (PageHWPoison)
                                           hugetlb_clear_page_hwpoison
                                             TestClearPageHWPoison
                                             // remove all list items
        llist_add
    unlock hugetlb_lock

The end result seems not critical (leaking raced raw_hwp_page?), but we need fix.

Quoted 11 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1589

Quoted 7 lines
mm/memory-failure.c +6 -0
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		SetPageHWPoison(p->page);
+		kfree(p);
+	}
+	llist_del_all(head);

If the above issue exists, moving ClearPageHWPoison(hpage) to here could fix it. We should clear PageHWPoison carefully since the head page is also can be poisoned.

The reason why I put ClearPageHWPoison(hpage) before llist_for_each_safe() was that raw error page can be the head page. But this can be solved with some additional code to remember whether raw_hwp_page list has an item associated with the head page.

Or another approach in my mind now is to call hugetlb_clear_page_hwpoison() with taking mf_mutex.

Quoted 2 lines

Thanks.

Thank you for valuable feedbacks.

- Naoya Horiguchi

Quoted 29 lines
Quoted 29 lines
mm/memory-failure.c +5 -2
+	return 0;
+}
+
 /*
  * Called from hugetlb code with hugetlb_lock held.
  *
@@ -1533,7 +1624,7 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 		goto out;
 	}
 
-	if (TestSetPageHWPoison(head)) {
+	if (hugetlb_set_page_hwpoison(head, page)) {
 		ret = -EHWPOISON;
 		goto out;
 	}
@@ -1585,7 +1676,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	lock_page(head);
 
 	if (hwpoison_filter(p)) {
-		ClearPageHWPoison(head);
+		hugetlb_clear_page_hwpoison(head);
 		res = -EOPNOTSUPP;
 		goto out;
 	}
Signature · 1 line
2.25.1
#30 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 193 lines · Muchun Song
Quoted 104 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 27 lines · Muchun Song
Quoted 18 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

When handling memory error on a hugetlb page, the error handler tries to dissolve and turn it into 4kB pages. If it's successfully dissolved, PageHWPoison flag is moved to the raw error page, so that's all right. However, dissolve sometimes fails, then the error page is left as hwpoisoned hugepage. It's useful if we can retry to dissolve it to save healthy pages, but that's not possible now because the information about where the raw error pages is lost.

Use the private field of a few tail pages to keep that information. The code path of shrinking hugepage pool uses this info to try delayed dissolve. In order to remember multiple errors in a hugepage, a singly-linked list originated from SUBPAGE_INDEX_HWPOISON-th tail page is constructed. Only simple operations (adding an entry or clearing all) are required and the list is assumed not to be very long, so this simple data structure should be enough.

Quoted 5 lines

If we failed to save raw error info, the hwpoison hugepage has errors on unknown subpage, then this new saving mechanism does not work any more, so disable saving new raw error info and freeing hwpoison hugepages.

Thanks for your work on this. I have several quesions below.

...

Quoted 22 lines
Quoted 22 lines
mm/memory-failure.c +18 -0
@@ -1499,6 +1499,97 @@ static int try_to_split_thp_page(struct page *page, const char *msg)
 }
 
 #ifdef CONFIG_HUGETLB_PAGE
+/*
+ * Struct raw_hwp_page represents information about "raw error page",
+ * constructing singly linked list originated from ->private field of
+ * SUBPAGE_INDEX_HWPOISON-th tail page.
+ */
+struct raw_hwp_page {
+	struct llist_node node;
+	struct page *page;
+};
+
+static inline struct llist_head *raw_hwp_list_head(struct page *hpage)
+{
+	return (struct llist_head *)&page_private(hpage + SUBPAGE_INDEX_HWPOISON);
+}
+
+static inline int raw_hwp_unreliable(struct page *hpage)
+{
+	return page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE);
Quoted 9 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1525

Quoted 6 lines
mm/memory-failure.c +6 -0
+}
+
+static inline void set_raw_hwp_unreliable(struct page *hpage)
+{
+	set_page_private(hpage + SUBPAGE_INDEX_HWPOISON_UNRELIABLE, 1);
+}

Why not use HPAGEFLAG(HwpoisonUnreliable, hwpoison_unreliable) directly?

OK, that sounds better, I'll do it.

Quoted 28 lines
Quoted 28 lines
mm/memory-failure.c +28 -0
+
+/*
+ * about race consideration
+ */
+static inline int hugetlb_set_page_hwpoison(struct page *hpage,
+					struct page *page)
+{
+	struct llist_head *head;
+	struct raw_hwp_page *raw_hwp;
+	struct llist_node *t, *tnode;
+	int ret;
+
+	/*
+	 * Once the hwpoison hugepage has lost reliable raw error info,
+	 * there is little mean to keep additional error info precisely,
+	 * so skip to add additional raw error info.
+	 */
+	if (raw_hwp_unreliable(hpage))
+		return -EHWPOISON;
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		if (p->page == page)
+			return -EHWPOISON;
+	}
+
+	ret = TestSetPageHWPoison(hpage) ? -EHWPOISON : 0;
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1558

Quoted 5 lines
mm/memory-failure.c +5 -0
+	/* the first error event will be counted in action_result(). */
+	if (ret)
+		num_poisoned_pages_inc();
+
+	raw_hwp = kmalloc(sizeof(struct raw_hwp_page), GFP_KERNEL);

This function can be called in atomic context, GFP_ATOMIC should be used here.

OK, I'll use GFP_ATOMIC.

Quoted 1 line
Quoted 9 lines
Quoted 5 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1561

Quoted 3 lines
mm/memory-failure.c +3 -0
+	if (raw_hwp) {
+		raw_hwp->page = page;
+		llist_add(&raw_hwp->node, head);

The maximum amount of items in the list is one, right?

The maximum is the number of subpages in the hugepage (512 for 2MB hugepage, 262144 for 1GB hugepage).

You are right. I have missed something yesterday.

Quoted 24 lines
Quoted 15 lines
Quoted 15 lines
mm/memory-failure.c +14 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

Quoted 2 lines

Sorry if I might miss your point, but raw_hwp_unreliable is set when allocating raw_hwp_page failed. hugetlb_set_page_hwpoison() can be called

Sorry. I have missed this. Thanks for your clarification.

Quoted 5 lines

multiple times on a hugepage and if one of the calls fails, the hwpoisoned hugepage becomes unreliable.

BTW, as you pointed out above, if we switch to passing GFP_ATOMIC to kmalloc(), the kmalloc() never fails, so we no longer have to implement this unreliable

No. kmalloc() with GFP_ATOMIC can fail unless I miss something important.

Quoted 11 lines

flag, so things get simpler.

Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 4 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

Quoted 2 lines

Maybe you are mentioning the race like below. Yes, that's possible.

Sorry, ignore my previous comments, I'm thinking something wrong.

Quoted 5 lines

CPU 0 CPU 1

free_huge_page
lock hugetlb_lock ClearHPageMigratable

				       remove_hugetlb_page()
				       // the page is non-HugeTLB now
Quoted 4 lines

unlock hugetlb_lock
get_huge_page_for_hwpoison
lock hugetlb_lock __get_huge_page_for_hwpoison

	// cannot reach here since it is not a HugeTLB page now.
	// So this race is impossible. Then we fallback to normal
	// page handling. Seems there is a new issue here.
	//
	// memory_failure()
	//	try_memory_failure_hugetlb()
	//	if (hugetlb)
	//		goto unlock_mutex;
	//	if (TestSetPageHWPoison(p)) {
	//	// This non-HugeTLB page's vmemmap is still optimized.
Quoted 1 line

Setting COMPOUND_PAGE_DTOR after hugetlb_vmemmap_restore() might fix this

^^^
I mean hugetlb_vmemmap_alloc().

Quoted 77 lines

issue, but we will encounter this race as you mentioned below.

Thanks.

Quoted 72 lines
      hugetlb_set_page_hwpoison
        allocate raw_hwp_page
        TestSetPageHWPoison
                                     update_and_free_page
                                       __update_and_free_page
                                         if (PageHWPoison)
                                           hugetlb_clear_page_hwpoison
                                             TestClearPageHWPoison
                                             // remove all list items
        llist_add
    unlock hugetlb_lock

The end result seems not critical (leaking raced raw_hwp_page?), but we need fix.

Quoted 11 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1589

Quoted 7 lines
mm/memory-failure.c +6 -0
+		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
+
+		SetPageHWPoison(p->page);
+		kfree(p);
+	}
+	llist_del_all(head);

If the above issue exists, moving ClearPageHWPoison(hpage) to here could fix it. We should clear PageHWPoison carefully since the head page is also can be poisoned.

The reason why I put ClearPageHWPoison(hpage) before llist_for_each_safe() was that raw error page can be the head page. But this can be solved with some additional code to remember whether raw_hwp_page list has an item associated with the head page.

Or another approach in my mind now is to call hugetlb_clear_page_hwpoison() with taking mf_mutex.

Quoted 2 lines

Thanks.

Thank you for valuable feedbacks.

- Naoya Horiguchi

Quoted 29 lines
Quoted 29 lines
mm/memory-failure.c +5 -2
+	return 0;
+}
+
 /*
  * Called from hugetlb code with hugetlb_lock held.
  *
@@ -1533,7 +1624,7 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 		goto out;
 	}
 
-	if (TestSetPageHWPoison(head)) {
+	if (hugetlb_set_page_hwpoison(head, page)) {
 		ret = -EHWPOISON;
 		goto out;
 	}
@@ -1585,7 +1676,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	lock_page(head);
 
 	if (hwpoison_filter(p)) {
-		ClearPageHWPoison(head);
+		hugetlb_clear_page_hwpoison(head);
 		res = -EOPNOTSUPP;
 		goto out;
 	}
Signature · 1 line
2.25.1
#31 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

In reply to Muchun Song
Quoted 4 lines · Muchun Song
Quoted 3 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 2 lines · Muchun Song
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

...

Quoted 28 lines
Quoted 23 lines
Quoted 14 lines
Quoted 14 lines
mm/memory-failure.c +14 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

Quoted 2 lines

Sorry if I might miss your point, but raw_hwp_unreliable is set when allocating raw_hwp_page failed. hugetlb_set_page_hwpoison() can be called

Sorry. I have missed this. Thanks for your clarification.

Quoted 7 lines
Quoted 5 lines

multiple times on a hugepage and if one of the calls fails, the hwpoisoned hugepage becomes unreliable.

BTW, as you pointed out above, if we switch to passing GFP_ATOMIC to kmalloc(), the kmalloc() never fails, so we no longer have to implement this unreliable

No. kmalloc() with GFP_ATOMIC can fail unless I miss something important.

OK, I've interpretted the comment about GFP_ATOMIC wrongly.

* %GFP_ATOMIC users can not sleep and need the allocation to succeed. A lower
* watermark is applied to allow access to "atomic reserves".

Quoted 16 lines
Quoted 11 lines

flag, so things get simpler.

Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 4 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

Quoted 2 lines

Maybe you are mentioning the race like below. Yes, that's possible.

Sorry, ignore my previous comments, I'm thinking something wrong.

Quoted 7 lines
Quoted 5 lines

CPU 0 CPU 1

free_huge_page
lock hugetlb_lock ClearHPageMigratable

				       remove_hugetlb_page()
				       // the page is non-HugeTLB now

Oh, I missed that.

Quoted 16 lines
Quoted 4 lines

unlock hugetlb_lock
get_huge_page_for_hwpoison
lock hugetlb_lock __get_huge_page_for_hwpoison

	// cannot reach here since it is not a HugeTLB page now.
	// So this race is impossible. Then we fallback to normal
	// page handling. Seems there is a new issue here.
	//
	// memory_failure()
	//	try_memory_failure_hugetlb()
	//	if (hugetlb)
	//		goto unlock_mutex;
	//	if (TestSetPageHWPoison(p)) {
	//	// This non-HugeTLB page's vmemmap is still optimized.
Quoted 2 lines

Setting COMPOUND_PAGE_DTOR after hugetlb_vmemmap_restore() might fix this issue, but we will encounter this race as you mentioned below.

I don't have clear ideas about this now (I don't test vmemmap-optimized case yet), so I will think more about this case. Maybe memory_failure() need detect it because memory_failure() heaviliy depends on the status of struct page.

Thanks,
Naoya Horiguchi

Quoted 18 lines

Thanks.

Quoted 15 lines
      hugetlb_set_page_hwpoison
        allocate raw_hwp_page
        TestSetPageHWPoison
                                     update_and_free_page
                                       __update_and_free_page
                                         if (PageHWPoison)
                                           hugetlb_clear_page_hwpoison
                                             TestClearPageHWPoison
                                             // remove all list items
        llist_add
    unlock hugetlb_lock

The end result seems not critical (leaking raced raw_hwp_page?), but we need fix.

#32 lore raw

Re: [PATCH v2 4/9] mm, hwpoison, hugetlb: support saving mechanism of raw error pages

Quoted 90 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 4 lines · Muchun Song
Quoted 3 lines · HORIGUCHI NAOYA(堀口 直也)
Quoted 2 lines · Muchun Song
Quoted 1 line · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

...

Quoted 28 lines
Quoted 23 lines
Quoted 14 lines
Quoted 14 lines
mm/memory-failure.c +14 -0
+	} else {
+		/*
+		 * Failed to save raw error info.  We no longer trace all
+		 * hwpoisoned subpages, and we need refuse to free/dissolve
+		 * this hwpoisoned hugepage.
+		 */
+		set_raw_hwp_unreliable(hpage);
+		return ret;
+	}
+	return ret;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1580

Quoted 5 lines
mm/memory-failure.c +5 -0
+	struct llist_head *head;
+	struct llist_node *t, *tnode;
+
+	if (raw_hwp_unreliable(hpage))
+		return -EBUSY;

IIUC, we use head page's PageHWPoison to synchronize hugetlb_clear_page_hwpoison() and hugetlb_set_page_hwpoison(), right? If so, who can set hwp_unreliable here?

Quoted 2 lines

Sorry if I might miss your point, but raw_hwp_unreliable is set when allocating raw_hwp_page failed. hugetlb_set_page_hwpoison() can be called

Sorry. I have missed this. Thanks for your clarification.

Quoted 7 lines
Quoted 5 lines

multiple times on a hugepage and if one of the calls fails, the hwpoisoned hugepage becomes unreliable.

BTW, as you pointed out above, if we switch to passing GFP_ATOMIC to kmalloc(), the kmalloc() never fails, so we no longer have to implement this unreliable

No. kmalloc() with GFP_ATOMIC can fail unless I miss something important.

OK, I've interpretted the comment about GFP_ATOMIC wrongly.

* %GFP_ATOMIC users can not sleep and need the allocation to succeed. A lower
* watermark is applied to allow access to "atomic reserves".

Quoted 16 lines
Quoted 11 lines

flag, so things get simpler.

Quoted 8 lines

mm/memory-failure.c static int try_to_split_thp_page(struct page *page, const char *msg) · line 1583

Quoted 4 lines
mm/memory-failure.c +3 -0
+	ClearPageHWPoison(hpage);
+	head = raw_hwp_list_head(hpage);
+	llist_for_each_safe(tnode, t, head->first) {

Is it possible that a new item is added hugetlb_set_page_hwpoison() and we do not traverse it (we have cleared page's PageHWPoison)? Then we ignored a real hwpoison page, right?

Quoted 2 lines

Maybe you are mentioning the race like below. Yes, that's possible.

Sorry, ignore my previous comments, I'm thinking something wrong.

Quoted 7 lines
Quoted 5 lines

CPU 0 CPU 1

free_huge_page
lock hugetlb_lock ClearHPageMigratable

				       remove_hugetlb_page()
				       // the page is non-HugeTLB now

Oh, I missed that.

Quoted 16 lines
Quoted 4 lines

unlock hugetlb_lock
get_huge_page_for_hwpoison
lock hugetlb_lock __get_huge_page_for_hwpoison

	// cannot reach here since it is not a HugeTLB page now.
	// So this race is impossible. Then we fallback to normal
	// page handling. Seems there is a new issue here.
	//
	// memory_failure()
	//	try_memory_failure_hugetlb()
	//	if (hugetlb)
	//		goto unlock_mutex;
	//	if (TestSetPageHWPoison(p)) {
	//	// This non-HugeTLB page's vmemmap is still optimized.
Quoted 8 lines
Quoted 2 lines

Setting COMPOUND_PAGE_DTOR after hugetlb_vmemmap_restore() might fix this issue, but we will encounter this race as you mentioned below.

I don't have clear ideas about this now (I don't test vmemmap-optimized case yet), so I will think more about this case. Maybe memory_failure() need detect it because memory_failure() heaviliy depends on the status of struct page.

Because HVO (HugeTLB Vmemmap Optimization) will map all tail vmemmap pages with read-only, we cannot write any data to some tail struct pages. It is a new issue unrelated to this patch.

Thanks.

Quoted 21 lines

Thanks,
Naoya Horiguchi

Quoted 18 lines

Thanks.

Quoted 15 lines
      hugetlb_set_page_hwpoison
        allocate raw_hwp_page
        TestSetPageHWPoison
                                     update_and_free_page
                                       __update_and_free_page
                                         if (PageHWPoison)
                                           hugetlb_clear_page_hwpoison
                                             TestClearPageHWPoison
                                             // remove all list items
        llist_add
    unlock hugetlb_lock

The end result seems not critical (leaking raced raw_hwp_page?), but we need fix.

author reviewer
#33 lore raw

PATCH v2 5/9 mm, hwpoison: make unpoison aware of raw error info in hwpoisoned hugepage

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Raw error info list needs to be removed when hwpoisoned hugetlb is unpoisioned. And unpoison handler needs to know how many errors there are in the target hugepage. So add them.

---
 include/linux/swapops.h |  9 +++++++++
 mm/memory-failure.c     | 34 +++++++++++++++++++++++++++-------
 2 files changed, 36 insertions(+), 7 deletions(-)
include/linux/swapops.h +9 -0
index 9584c2e1c54a..ad62776ee99c 100644
@@ -494,6 +494,11 @@ static inline void num_poisoned_pages_dec(void)
 	atomic_long_dec(&num_poisoned_pages);
 }
 
+static inline void num_poisoned_pages_sub(long i)
+{
+	atomic_long_sub(i, &num_poisoned_pages);
+}
+
 #else
 
 static inline swp_entry_t make_hwpoison_entry(struct page *page)
@@ -514,6 +519,10 @@ static inline struct page *hwpoison_entry_to_page(swp_entry_t entry)
 static inline void num_poisoned_pages_inc(void)
 {
 }
+
+static inline void num_poisoned_pages_sub(long i)
+{
+}
 #endif
 
 static inline int non_swap_entry(swp_entry_t entry)
mm/memory-failure.c +27 -7
index af571fa6f2af..6005e953d011 100644
@@ -1571,23 +1571,34 @@ static inline int hugetlb_set_page_hwpoison(struct page *hpage,
 	return ret;
 }
 
-inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+static inline long free_raw_hwp_pages(struct page *hpage, bool move_flag)
 {
 	struct llist_head *head;
 	struct llist_node *t, *tnode;
+	long count = 0;
 
-	if (raw_hwp_unreliable(hpage))
-		return -EBUSY;
-	ClearPageHWPoison(hpage);
 	head = raw_hwp_list_head(hpage);
 	llist_for_each_safe(tnode, t, head->first) {
 		struct raw_hwp_page *p = container_of(tnode, struct raw_hwp_page, node);
 
-		SetPageHWPoison(p->page);
+		if (move_flag)
+			SetPageHWPoison(p->page);
 		kfree(p);
+		count++;
 	}
 	llist_del_all(head);
-	return 0;
+	return count;
+}
+
+inline int hugetlb_clear_page_hwpoison(struct page *hpage)
+{
+	int ret = -EBUSY;
+
+	if (raw_hwp_unreliable(hpage))
+		return ret;
+	ret = !TestClearPageHWPoison(hpage);
+	free_raw_hwp_pages(hpage, true);
+	return ret;
 }
 
 /*
@@ -1729,6 +1740,10 @@ static inline int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *
 {
 	return 0;
 }
+
+static inline void free_raw_hwp_pages(struct page *hpage, bool move_flag)
+{
+}
 #endif
 
 static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
@@ -2188,6 +2203,7 @@ int unpoison_memory(unsigned long pfn)
 	struct page *p;
 	int ret = -EBUSY;
 	int freeit = 0;
+	long count = 1;
 	static DEFINE_RATELIMIT_STATE(unpoison_rs, DEFAULT_RATELIMIT_INTERVAL,
 					DEFAULT_RATELIMIT_BURST);
 
@@ -2235,6 +2251,8 @@ int unpoison_memory(unsigned long pfn)
 
 	ret = get_hwpoison_page(p, MF_UNPOISON);
 	if (!ret) {
+		if (PageHuge(p))
+			count = free_raw_hwp_pages(page, false);
 		ret = TestClearPageHWPoison(page) ? 0 : -EBUSY;
 	} else if (ret < 0) {
 		if (ret == -EHWPOISON) {
@@ -2243,6 +2261,8 @@ int unpoison_memory(unsigned long pfn)
 			unpoison_pr_info("Unpoison: failed to grab page %#lx\n",
 					 pfn, &unpoison_rs);
 	} else {
+		if (PageHuge(p))
+			count = free_raw_hwp_pages(page, false);
 		freeit = !!TestClearPageHWPoison(p);
 
 		put_page(page);
@@ -2255,7 +2275,7 @@ int unpoison_memory(unsigned long pfn)
 unlock_mutex:
 	mutex_unlock(&mf_mutex);
 	if (!ret || freeit) {
-		num_poisoned_pages_dec();
+		num_poisoned_pages_sub(count);
 		unpoison_pr_info("Unpoison: Software-unpoisoned page %#lx\n",
 				 page_to_pfn(p), &unpoison_rs);
 	}
Signature · 1 line
2.25.1
maintainer
#34 lore raw

Re: [PATCH v2 5/9] mm, hwpoison: make unpoison aware of raw error info in hwpoisoned hugepage

Quoted 7 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Raw error info list needs to be removed when hwpoisoned hugetlb is unpoisioned. And unpoison handler needs to know how many errors there are in the target hugepage. So add them.

snip

mm/memory-failure.c int unpoison_memory(unsigned long pfn) · line 2278

Quoted 6 lines
mm/memory-failure.c +1 -1
@@ -2255,7 +2275,7 @@ int unpoison_memory(unsigned long pfn)
 unlock_mutex:
 	mutex_unlock(&mf_mutex);
 	if (!ret || freeit) {
-		num_poisoned_pages_dec();
+		num_poisoned_pages_sub(count);

IIUC, num_poisoned_pages will only be incremented once for hugetlb page. If many subpages are hwpoisoned, they will reach the "else if (res == -EHWPOISON)" path in try_memory_failure_hugetlb and thus num_poisoned_pages_inc is ignored. Maybe that should be changed so subpages can contribute to the num_poisoned_pages or should we just do num_poisoned_pages_dec here? Or am I miss something?

Thanks!

Quoted 4 lines
mm/memory-failure.c +0 -0
 		unpoison_pr_info("Unpoison: Software-unpoisoned page %#lx\n",
 				 page_to_pfn(p), &unpoison_rs);
 	}
maintainer
#35 lore raw

Re: [PATCH v2 5/9] mm, hwpoison: make unpoison aware of raw error info in hwpoisoned hugepage

Quoted 11 lines · Miaohe Lin
Quoted 7 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Raw error info list needs to be removed when hwpoisoned hugetlb is unpoisioned. And unpoison handler needs to know how many errors there are in the target hugepage. So add them.

snip

Quoted 12 lines

mm/memory-failure.c int unpoison_memory(unsigned long pfn) · line 2278

Quoted 6 lines
mm/memory-failure.c +1 -1
@@ -2255,7 +2275,7 @@ int unpoison_memory(unsigned long pfn)
 unlock_mutex:
 	mutex_unlock(&mf_mutex);
 	if (!ret || freeit) {
-		num_poisoned_pages_dec();
+		num_poisoned_pages_sub(count);

IIUC, num_poisoned_pages will only be incremented once for hugetlb page. If many subpages are hwpoisoned, they will reach the "else if (res == -EHWPOISON)" path in try_memory_failure_hugetlb and thus num_poisoned_pages_inc is ignored. Maybe that should be changed so subpages can contribute to the num_poisoned_pages or should we just do num_poisoned_pages_dec here? Or am I miss something?

Sorry. I re-read patch 4/9 and I found hugetlb_set_page_hwpoison will increment the num_poisoned_pages for each subpages. Please ignore this question.

Quoted 8 lines

Thanks!

Quoted 4 lines
mm/memory-failure.c +0 -0
 		unpoison_pr_info("Unpoison: Software-unpoisoned page %#lx\n",
 				 page_to_pfn(p), &unpoison_rs);
 	}
author reviewer
#36 lore raw

PATCH v2 6/9 mm, hwpoison: set PG_hwpoison for busy hugetlb pages

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

If memory_failure() fails to grab page refcount on a hugetlb page because it's busy, it returns without setting PG_hwpoison on it. This not only loses a chance of error containment, but breaks the rule that action_result() should be called only when memory_failure() do any of handling work (even if that's just setting PG_hwpoison). This inconsistency could harm code maintainability.

So set PG_hwpoison and call hugetlb_set_page_hwpoison() for such a case.

Fixes 405ce051236c ("mm/hwpoison: fix race between hugetlb free/demotion and memory_failure_hugetlb()")
---
 include/linux/mm.h  | 1 +
 mm/memory-failure.c | 8 ++++----
 2 files changed, 5 insertions(+), 4 deletions(-)
include/linux/mm.h +1 -0
index 4346e51484ba..044dc5a2e361 100644
@@ -3233,6 +3233,7 @@ enum mf_flags {
 	MF_SOFT_OFFLINE = 1 << 3,
 	MF_UNPOISON = 1 << 4,
 	MF_SW_SIMULATED = 1 << 5,
+	MF_NO_RETRY = 1 << 6,
 };
 extern int memory_failure(unsigned long pfn, int flags);
 extern void memory_failure_queue(unsigned long pfn, int flags);
mm/memory-failure.c +4 -4
index 6005e953d011..ce045d0d6115 100644
@@ -1632,7 +1632,8 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 			count_increased = true;
 	} else {
 		ret = -EBUSY;
-		goto out;
+		if (!(flags & MF_NO_RETRY))
+			goto out;
 	}
 
 	if (hugetlb_set_page_hwpoison(head, page)) {
@@ -1659,7 +1660,6 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	struct page *p = pfn_to_page(pfn);
 	struct page *head;
 	unsigned long page_flags;
-	bool retry = true;
 
 	*hugetlb = 1;
 retry:
@@ -1675,8 +1675,8 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 		}
 		return res;
 	} else if (res == -EBUSY) {
-		if (retry) {
-			retry = false;
+		if (!(flags & MF_NO_RETRY)) {
+			flags |= MF_NO_RETRY;
 			goto retry;
 		}
 		action_result(pfn, MF_MSG_UNKNOWN, MF_IGNORED);
Signature · 1 line
2.25.1
maintainer Reviewed-by on v2 6/9
#37 lore raw

Re: [PATCH v2 6/9] mm, hwpoison: set PG_hwpoison for busy hugetlb pages

Quoted 9 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

If memory_failure() fails to grab page refcount on a hugetlb page because it's busy, it returns without setting PG_hwpoison on it. This not only loses a chance of error containment, but breaks the rule that action_result() should be called only when memory_failure() do any of handling work (even if that's just setting PG_hwpoison). This inconsistency could harm code maintainability.

Quoted 4 lines

So set PG_hwpoison and call hugetlb_set_page_hwpoison() for such a case.

Fixes 405ce051236c ("mm/hwpoison: fix race between hugetlb free/demotion and memory_failure_hugetlb()")

Thanks!

Quoted 51 lines
---
 include/linux/mm.h  | 1 +
 mm/memory-failure.c | 8 ++++----
 2 files changed, 5 insertions(+), 4 deletions(-)
include/linux/mm.h +1 -0
index 4346e51484ba..044dc5a2e361 100644
@@ -3233,6 +3233,7 @@ enum mf_flags {
 	MF_SOFT_OFFLINE = 1 << 3,
 	MF_UNPOISON = 1 << 4,
 	MF_SW_SIMULATED = 1 << 5,
+	MF_NO_RETRY = 1 << 6,
 };
 extern int memory_failure(unsigned long pfn, int flags);
 extern void memory_failure_queue(unsigned long pfn, int flags);
mm/memory-failure.c +4 -4
index 6005e953d011..ce045d0d6115 100644
@@ -1632,7 +1632,8 @@ int __get_huge_page_for_hwpoison(unsigned long pfn, int flags)
 			count_increased = true;
 	} else {
 		ret = -EBUSY;
-		goto out;
+		if (!(flags & MF_NO_RETRY))
+			goto out;
 	}
 
 	if (hugetlb_set_page_hwpoison(head, page)) {
@@ -1659,7 +1660,6 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	struct page *p = pfn_to_page(pfn);
 	struct page *head;
 	unsigned long page_flags;
-	bool retry = true;
 
 	*hugetlb = 1;
 retry:
@@ -1675,8 +1675,8 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 		}
 		return res;
 	} else if (res == -EBUSY) {
-		if (retry) {
-			retry = false;
+		if (!(flags & MF_NO_RETRY)) {
+			flags |= MF_NO_RETRY;
 			goto retry;
 		}
 		action_result(pfn, MF_MSG_UNKNOWN, MF_IGNORED);
author reviewer
#38 lore raw

PATCH v2 7/9 mm, hwpoison: make __page_handle_poison returns int

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

__page_handle_poison() returns bool that shows whether take_page_off_buddy() has passed or not now. But we will want to distinguish another case of "dissolve has passed but taking off failed" by its return value. So change the type of the return value. No functional change.

---
 mm/memory-failure.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
mm/memory-failure.c +11 -6
index ce045d0d6115..db85f644a1e3 100644
@@ -71,7 +71,13 @@ atomic_long_t num_poisoned_pages __read_mostly = ATOMIC_LONG_INIT(0);
 
 static bool hw_memory_failure __read_mostly = false;
 
-static bool __page_handle_poison(struct page *page)
+/*
+ * Return values:
+ *   1:   the page is dissolved (if needed) and taken off from buddy,
+ *   0:   the page is dissolved (if needed) and not taken off from buddy,
+ *   < 0: failed to dissolve.
+ */
+static int __page_handle_poison(struct page *page)
 {
 	int ret;
 
@@ -81,7 +87,7 @@ static bool __page_handle_poison(struct page *page)
 		ret = take_page_off_buddy(page);
 	zone_pcp_enable(page_zone(page));
 
-	return ret > 0;
+	return ret;
 }
 
 static bool page_handle_poison(struct page *page, bool hugepage_or_freepage, bool release)
@@ -91,7 +97,7 @@ static bool page_handle_poison(struct page *page, bool hugepage_or_freepage, boo
 		 * Doing this check for free pages is also fine since dissolve_free_huge_page
 		 * returns 0 for non-hugetlb pages as well.
 		 */
-		if (!__page_handle_poison(page))
+		if (__page_handle_poison(page) <= 0)
 			/*
 			 * We could fail to take off the target page from buddy
 			 * for example due to racy page allocation, but that's
@@ -1048,7 +1054,7 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		 * subpages.
 		 */
 		put_page(hpage);
-		if (__page_handle_poison(p)) {
+		if (__page_handle_poison(p) > 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
 		}
@@ -1698,8 +1704,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	 */
 	if (res == 0) {
 		unlock_page(head);
-		res = MF_FAILED;
-		if (__page_handle_poison(p)) {
+		if (__page_handle_poison(p) > 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
 		}
Signature · 1 line
2.25.1
maintainer Reviewed-by on v2 7/9
#39 lore raw

Re: [PATCH v2 7/9] mm, hwpoison: make __page_handle_poison returns int

Quoted 60 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

__page_handle_poison() returns bool that shows whether take_page_off_buddy() has passed or not now. But we will want to distinguish another case of "dissolve has passed but taking off failed" by its return value. So change the type of the return value. No functional change.

---
 mm/memory-failure.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
mm/memory-failure.c +10 -4
index ce045d0d6115..db85f644a1e3 100644
@@ -71,7 +71,13 @@ atomic_long_t num_poisoned_pages __read_mostly = ATOMIC_LONG_INIT(0);
 
 static bool hw_memory_failure __read_mostly = false;
 
-static bool __page_handle_poison(struct page *page)
+/*
+ * Return values:
+ *   1:   the page is dissolved (if needed) and taken off from buddy,
+ *   0:   the page is dissolved (if needed) and not taken off from buddy,
+ *   < 0: failed to dissolve.
+ */
+static int __page_handle_poison(struct page *page)
 {
 	int ret;
 
@@ -81,7 +87,7 @@ static bool __page_handle_poison(struct page *page)
 		ret = take_page_off_buddy(page);
 	zone_pcp_enable(page_zone(page));
 
-	return ret > 0;
+	return ret;
 }
 
 static bool page_handle_poison(struct page *page, bool hugepage_or_freepage, bool release)
@@ -91,7 +97,7 @@ static bool page_handle_poison(struct page *page, bool hugepage_or_freepage, boo
 		 * Doing this check for free pages is also fine since dissolve_free_huge_page
 		 * returns 0 for non-hugetlb pages as well.
 		 */
-		if (!__page_handle_poison(page))
+		if (__page_handle_poison(page) <= 0)
 			/*
 			 * We could fail to take off the target page from buddy
 			 * for example due to racy page allocation, but that's
@@ -1048,7 +1054,7 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		 * subpages.
 		 */
 		put_page(hpage);
-		if (__page_handle_poison(p)) {
+		if (__page_handle_poison(p) > 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
 		}

mm/memory-failure.c static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb · line 1701

Quoted 4 lines
mm/memory-failure.c +0 -1
@@ -1698,8 +1704,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	 */
 	if (res == 0) {
 		unlock_page(head);
-		res = MF_FAILED;

It seems the previous discussion in [1] is missed. But that doesn't matter as pointed out by [1]. :)

Thanks.

[1]: https://lkml.org/lkml/2022/6/8/10

Quoted 6 lines
mm/memory-failure.c +1 -1
-		if (__page_handle_poison(p)) {
+		if (__page_handle_poison(p) > 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
 		}
#40 lore raw

Re: [PATCH v2 7/9] mm, hwpoison: make __page_handle_poison returns int

Quoted 11 lines · Miaohe Lin
Quoted 10 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

__page_handle_poison() returns bool that shows whether take_page_off_buddy() has passed or not now. But we will want to distinguish another case of "dissolve has passed but taking off failed" by its return value. So change the type of the return value. No functional change.

---

...

Quoted 7 lines

mm/memory-failure.c static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb · line 1701

Quoted 5 lines
mm/memory-failure.c +0 -1
@@ -1698,8 +1704,7 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	 */
 	if (res == 0) {
 		unlock_page(head);
-		res = MF_FAILED;

It seems the previous discussion in [1] is missed. But that doesn't matter as pointed out by [1]. :)

Sorry for missing this, I updated this line to be removed in patch 8/9.

Quoted 2 lines

Thank you.

- Naoya Horiguchi

Quoted 4 lines

Thanks.

[1]: https://lkml.org/lkml/2022/6/8/10

author reviewer
#41 lore raw

PATCH v2 8/9 mm, hwpoison: skip raw hwpoison page in freeing 1GB hugepage

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Currently if memory_failure() (modified to remove blocking code with subsequent patch) is called on a page in some 1GB hugepage, memory error handling fails and the raw error page gets into leaked state. The impact is small in production systems (just leaked single 4kB page), but this limits the testability because unpoison doesn't work for it. We can no longer create 1GB hugepage on the 1GB physical address range with such leaked pages, that's not useful when testing on small systems.

When a hwpoison page in a 1GB hugepage is handled, it's caught by the PageHWPoison check in free_pages_prepare() because the 1GB hugepage is broken down into raw error pages before coming to this point:

        if (unlikely(PageHWPoison(page)) && !order) {
                ...
                return false;
        }

Then, the page is not sent to buddy and the page refcount is left 0.

Originally this check is supposed to work when the error page is freed from page_handle_poison() (that is called from soft-offline), but now we are opening another path to call it, so the callers of __page_handle_poison() need to handle the case by considering the return value 0 as success. Then page refcount for hwpoison is properly incremented so unpoison works.

---
 mm/memory-failure.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
mm/memory-failure.c +6 -3
index db85f644a1e3..fc7b83cb6468 100644
@@ -1046,7 +1046,6 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		res = truncate_error_page(hpage, page_to_pfn(p), mapping);
 		unlock_page(hpage);
 	} else {
-		res = MF_FAILED;
 		unlock_page(hpage);
 		/*
 		 * migration entry prevents later access on error hugepage,
@@ -1054,9 +1053,11 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		 * subpages.
 		 */
 		put_page(hpage);
-		if (__page_handle_poison(p) > 0) {
+		if (__page_handle_poison(p) >= 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
+		} else {
+			res = MF_FAILED;
 		}
 	}
 
@@ -1704,9 +1705,11 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	 */
 	if (res == 0) {
 		unlock_page(head);
-		if (__page_handle_poison(p) > 0) {
+		if (__page_handle_poison(p) >= 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
+		} else {
+			res = MF_FAILED;
 		}
 		action_result(pfn, MF_MSG_FREE_HUGE, res);
 		return res == MF_RECOVERED ? 0 : -EBUSY;
Signature · 1 line
2.25.1
maintainer Reviewed-by on v2 8/9
#42 lore raw

Re: [PATCH v2 8/9] mm, hwpoison: skip raw hwpoison page in freeing 1GB hugepage

Quoted 21 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Currently if memory_failure() (modified to remove blocking code with subsequent patch) is called on a page in some 1GB hugepage, memory error handling fails and the raw error page gets into leaked state. The impact is small in production systems (just leaked single 4kB page), but this limits the testability because unpoison doesn't work for it. We can no longer create 1GB hugepage on the 1GB physical address range with such leaked pages, that's not useful when testing on small systems.

When a hwpoison page in a 1GB hugepage is handled, it's caught by the PageHWPoison check in free_pages_prepare() because the 1GB hugepage is broken down into raw error pages before coming to this point:

        if (unlikely(PageHWPoison(page)) && !order) {
                ...
                return false;
        }

Then, the page is not sent to buddy and the page refcount is left 0.

Quoted 7 lines

Originally this check is supposed to work when the error page is freed from page_handle_poison() (that is called from soft-offline), but now we are opening another path to call it, so the callers of __page_handle_poison() need to handle the case by considering the return value 0 as success. Then page refcount for hwpoison is properly incremented so unpoison works.

It seems I misunderstand the commit log in [1]. But I hope I get the point this time. :)

Thanks!

[1]https://lore.kernel.org/linux-mm/[email protected]/

Quoted 43 lines
---
 mm/memory-failure.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
mm/memory-failure.c +6 -3
index db85f644a1e3..fc7b83cb6468 100644
@@ -1046,7 +1046,6 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		res = truncate_error_page(hpage, page_to_pfn(p), mapping);
 		unlock_page(hpage);
 	} else {
-		res = MF_FAILED;
 		unlock_page(hpage);
 		/*
 		 * migration entry prevents later access on error hugepage,
@@ -1054,9 +1053,11 @@ static int me_huge_page(struct page_state *ps, struct page *p)
 		 * subpages.
 		 */
 		put_page(hpage);
-		if (__page_handle_poison(p) > 0) {
+		if (__page_handle_poison(p) >= 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
+		} else {
+			res = MF_FAILED;
 		}
 	}
 
@@ -1704,9 +1705,11 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 	 */
 	if (res == 0) {
 		unlock_page(head);
-		if (__page_handle_poison(p) > 0) {
+		if (__page_handle_poison(p) >= 0) {
 			page_ref_inc(p);
 			res = MF_RECOVERED;
+		} else {
+			res = MF_FAILED;
 		}
 		action_result(pfn, MF_MSG_FREE_HUGE, res);
 		return res == MF_RECOVERED ? 0 : -EBUSY;
author reviewer
#43 lore raw

PATCH v2 9/9 mm, hwpoison: enable memory error handling on 1GB hugepage

In reply to Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Now error handling code is prepared, so remove the blocking code and enable memory error handling on 1GB hugepage.

---
 include/linux/mm.h      |  1 -
 include/ras/ras_event.h |  1 -
 mm/memory-failure.c     | 16 ----------------
 3 files changed, 18 deletions(-)
include/linux/mm.h +0 -1
index 044dc5a2e361..9d7e9b5a4d1d 100644
@@ -3284,7 +3284,6 @@ enum mf_action_page_type {
 	MF_MSG_DIFFERENT_COMPOUND,
 	MF_MSG_HUGE,
 	MF_MSG_FREE_HUGE,
-	MF_MSG_NON_PMD_HUGE,
 	MF_MSG_UNMAP_FAILED,
 	MF_MSG_DIRTY_SWAPCACHE,
 	MF_MSG_CLEAN_SWAPCACHE,
include/ras/ras_event.h +0 -1
index d0337a41141c..cbd3ddd7c33d 100644
@@ -360,7 +360,6 @@ TRACE_EVENT(aer_event,
 	EM ( MF_MSG_DIFFERENT_COMPOUND, "different compound page after locking" ) \
 	EM ( MF_MSG_HUGE, "huge page" )					\
 	EM ( MF_MSG_FREE_HUGE, "free huge page" )			\
-	EM ( MF_MSG_NON_PMD_HUGE, "non-pmd-sized huge page" )		\
 	EM ( MF_MSG_UNMAP_FAILED, "unmapping failed page" )		\
 	EM ( MF_MSG_DIRTY_SWAPCACHE, "dirty swapcache page" )		\
 	EM ( MF_MSG_CLEAN_SWAPCACHE, "clean swapcache page" )		\
mm/memory-failure.c +0 -16
index fc7b83cb6468..33521e059f7f 100644
@@ -728,7 +728,6 @@ static const char * const action_page_types[] = {
 	[MF_MSG_DIFFERENT_COMPOUND]	= "different compound page after locking",
 	[MF_MSG_HUGE]			= "huge page",
 	[MF_MSG_FREE_HUGE]		= "free huge page",
-	[MF_MSG_NON_PMD_HUGE]		= "non-pmd-sized huge page",
 	[MF_MSG_UNMAP_FAILED]		= "unmapping failed page",
 	[MF_MSG_DIRTY_SWAPCACHE]	= "dirty swapcache page",
 	[MF_MSG_CLEAN_SWAPCACHE]	= "clean swapcache page",
@@ -1717,21 +1716,6 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 
 	page_flags = head->flags;
 
-	/*
-	 * TODO: hwpoison for pud-sized hugetlb doesn't work right now, so
-	 * simply disable it. In order to make it work properly, we need
-	 * make sure that:
-	 *  - conversion of a pud that maps an error hugetlb into hwpoison
-	 *    entry properly works, and
-	 *  - other mm code walking over page table is aware of pud-aligned
-	 *    hwpoison entries.
-	 */
-	if (huge_page_size(page_hstate(head)) > PMD_SIZE) {
-		action_result(pfn, MF_MSG_NON_PMD_HUGE, MF_IGNORED);
-		res = -EBUSY;
-		goto out;
-	}
-
 	if (!hwpoison_user_mappings(p, pfn, flags, head)) {
 		action_result(pfn, MF_MSG_UNMAP_FAILED, MF_IGNORED);
 		res = -EBUSY;
Signature · 1 line
2.25.1
maintainer Reviewed-by on v2 9/9
#44 lore raw

Re: [PATCH v2 9/9] mm, hwpoison: enable memory error handling on 1GB hugepage

Quoted 6 lines · Naoya Horiguchi

From: Naoya Horiguchi <[email protected]>

Now error handling code is prepared, so remove the blocking code and enable memory error handling on 1GB hugepage.

The patch itself looks good to me.

Thanks!

BTW: There might still be some places need to be changed. Maybe we can process them altogether after this series is done?

Quoted 65 lines
---
 include/linux/mm.h      |  1 -
 include/ras/ras_event.h |  1 -
 mm/memory-failure.c     | 16 ----------------
 3 files changed, 18 deletions(-)
include/linux/mm.h +0 -1
index 044dc5a2e361..9d7e9b5a4d1d 100644
@@ -3284,7 +3284,6 @@ enum mf_action_page_type {
 	MF_MSG_DIFFERENT_COMPOUND,
 	MF_MSG_HUGE,
 	MF_MSG_FREE_HUGE,
-	MF_MSG_NON_PMD_HUGE,
 	MF_MSG_UNMAP_FAILED,
 	MF_MSG_DIRTY_SWAPCACHE,
 	MF_MSG_CLEAN_SWAPCACHE,
include/ras/ras_event.h +0 -1
index d0337a41141c..cbd3ddd7c33d 100644
@@ -360,7 +360,6 @@ TRACE_EVENT(aer_event,
 	EM ( MF_MSG_DIFFERENT_COMPOUND, "different compound page after locking" ) \
 	EM ( MF_MSG_HUGE, "huge page" )					\
 	EM ( MF_MSG_FREE_HUGE, "free huge page" )			\
-	EM ( MF_MSG_NON_PMD_HUGE, "non-pmd-sized huge page" )		\
 	EM ( MF_MSG_UNMAP_FAILED, "unmapping failed page" )		\
 	EM ( MF_MSG_DIRTY_SWAPCACHE, "dirty swapcache page" )		\
 	EM ( MF_MSG_CLEAN_SWAPCACHE, "clean swapcache page" )		\
mm/memory-failure.c +0 -16
index fc7b83cb6468..33521e059f7f 100644
@@ -728,7 +728,6 @@ static const char * const action_page_types[] = {
 	[MF_MSG_DIFFERENT_COMPOUND]	= "different compound page after locking",
 	[MF_MSG_HUGE]			= "huge page",
 	[MF_MSG_FREE_HUGE]		= "free huge page",
-	[MF_MSG_NON_PMD_HUGE]		= "non-pmd-sized huge page",
 	[MF_MSG_UNMAP_FAILED]		= "unmapping failed page",
 	[MF_MSG_DIRTY_SWAPCACHE]	= "dirty swapcache page",
 	[MF_MSG_CLEAN_SWAPCACHE]	= "clean swapcache page",
@@ -1717,21 +1716,6 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb
 
 	page_flags = head->flags;
 
-	/*
-	 * TODO: hwpoison for pud-sized hugetlb doesn't work right now, so
-	 * simply disable it. In order to make it work properly, we need
-	 * make sure that:
-	 *  - conversion of a pud that maps an error hugetlb into hwpoison
-	 *    entry properly works, and
-	 *  - other mm code walking over page table is aware of pud-aligned
-	 *    hwpoison entries.
-	 */
-	if (huge_page_size(page_hstate(head)) > PMD_SIZE) {
-		action_result(pfn, MF_MSG_NON_PMD_HUGE, MF_IGNORED);
-		res = -EBUSY;
-		goto out;
-	}
-
 	if (!hwpoison_user_mappings(p, pfn, flags, head)) {
 		action_result(pfn, MF_MSG_UNMAP_FAILED, MF_IGNORED);
 		res = -EBUSY;