Skip to contentSkip to conversation

mm/migrate: fix shmem xarray update during migration

PATCH Superseded Zi Yan 4 participants 10 messages Feb 28, 2025 – Mar 5, 2025

Subsystems (3)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Zi Yan

author reviewer
#0 lore raw

PATCH v2 1/1 mm/migrate: fix shmem xarray update during migration

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +5 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2
Reviewed-by on v2 1/1
#1 lore raw
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

LGTM. Thanks.

Quoted 21 lines
---
  mm/migrate.c | 6 +++++-
  1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +5 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
  			folio_set_swapcache(newfolio);
  			newfolio->private = folio_get_private(folio);
  		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
+			entries = nr;
  	} else {
  		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
  		entries = 1;
author reviewer
#2 lore raw
In reply to Zi Yan
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

+Cc:stable

Quoted 23 lines
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +5 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2

Best Regards,
Yan, Zi

maintainer
#3 lore raw
Quoted 9 lines · Zi Yan
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 8 lines
Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

+Cc:stable

<formletter>

This is not the correct way to submit patches for inclusion in the stable kernel tree. Please read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

</formletter>

author reviewer
#4 lore raw
Quoted 19 lines · Greg KH
Quoted 9 lines · Zi Yan
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 8 lines
Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

+Cc:stable

Quoted 8 lines

<formletter>

This is not the correct way to submit patches for inclusion in the stable kernel tree. Please read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

</formletter>

Sure. And this is not the right fix. I will resend a new one. Sorry for the noise.

Best Regards,
Yan, Zi

#5 lore raw
In reply to Zi Yan
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

It's a great find (I think), and your commit message is okay:
but unless I'm much mistaken, NAK to the patch itself.

First, I say "(I think)" there, because I don't actually know what the loop writing the same folio nr times to the multi-index entry does to the xarray: I can imagine it as being completely harmless, just nr times more work than was needed.

But I guess it does something bad, since Matthew was horrified, and we have all found that your patch appears to improve behaviour (or at least improve behaviour in the context of your folio_split()
series: none of us noticed a problem before that, but it may be that your new series is widening our exposure to existing bugs).

Maybe your orginal patch, with the shmem_mapping(mapping) check there, was good, and it's only wrong when changed to !folio_test_anon(folio); but TBH I find it too confusing, with the conditionals the way they are. See my preferred alternative below.

The vital point is that multi-index entries are not used in swap cache:
whether the folio in question orginates from anon or from shmem. And it's easier to understand once you remember that a shmem folio is never in both page cache and swap cache at the same time (well, there may be an instant of transition from one to other while that folio is held locked) - once it's in swap cache, folio->mapping is NULL and it's no longer recognizable as from a shmem mapping.

The way I read your patch originally, I thought it meant that shmem folios go into the swap cache as multi-index, but anon folios do not; which seemed a worrying mixture to me. But crashes on the VM_BUG_ON_PAGE(entry != folio, entry) in __delete_from_swap_cache() yesterday (with your patch in) led me to see how add_to_swap_cache() inserts multiple non-multi-index entries, whether for anon or for shmem.

If this patch really is needed in old releases, then I suspect that mm/huge_memory.c needs correction there too; but let me explain in a response to your folio_split() series.

Quoted 23 lines
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +5 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2

NAK to that patch above, here's how I think it should be:

---
 mm/migrate.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
mm/migrate.c +5 -5
index fb19a18892c8..822776819ca6 100644
@@ -518,12 +518,12 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 	if (folio_test_anon(folio) && folio_test_large(folio))
 		mod_mthp_stat(folio_order(folio), MTHP_STAT_NR_ANON, 1);
 	folio_ref_add(newfolio, nr); /* add cache reference */
-	if (folio_test_swapbacked(folio)) {
+	if (folio_test_swapbacked(folio))
 		__folio_set_swapbacked(newfolio);
-		if (folio_test_swapcache(folio)) {
-			folio_set_swapcache(newfolio);
-			newfolio->private = folio_get_private(folio);
-		}
+
+	if (folio_test_swapcache(folio)) {
+		folio_set_swapcache(newfolio);
+		newfolio->private = folio_get_private(folio);
 		entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
Signature · 1 line
2.43.0
author reviewer
#6 lore raw
Quoted 9 lines · Hugh Dickins
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 8 lines
Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

It's a great find (I think), and your commit message is okay:
but unless I'm much mistaken, NAK to the patch itself.

Got it. Thank you for the review.

Quoted 17 lines

First, I say "(I think)" there, because I don't actually know what the loop writing the same folio nr times to the multi-index entry does to the xarray: I can imagine it as being completely harmless, just nr times more work than was needed.

But I guess it does something bad, since Matthew was horrified, and we have all found that your patch appears to improve behaviour (or at least improve behaviour in the context of your folio_split()
series: none of us noticed a problem before that, but it may be that your new series is widening our exposure to existing bugs).

Maybe your orginal patch, with the shmem_mapping(mapping) check there, was good, and it's only wrong when changed to !folio_test_anon(folio); but TBH I find it too confusing, with the conditionals the way they are. See my preferred alternative below.

Quoted 7 lines

The vital point is that multi-index entries are not used in swap cache:
whether the folio in question orginates from anon or from shmem. And it's easier to understand once you remember that a shmem folio is never in both page cache and swap cache at the same time (well, there may be an instant of transition from one to other while that folio is held locked) - once it's in swap cache, folio->mapping is NULL and it's no longer recognizable as from a shmem mapping.

Got it. Now it all makes sense to me. Thank you for the explanation.

Quoted 7 lines

The way I read your patch originally, I thought it meant that shmem folios go into the swap cache as multi-index, but anon folios do not; which seemed a worrying mixture to me. But crashes on the VM_BUG_ON_PAGE(entry != folio, entry) in __delete_from_swap_cache() yesterday (with your patch in) led me to see how add_to_swap_cache() inserts multiple non-multi-index entries, whether for anon or for shmem.

Thanks for the pointer.

Quoted 22 lines

If this patch really is needed in old releases, then I suspect that mm/huge_memory.c needs correction there too; but let me explain in a response to your folio_split() series.

Quoted 17 lines
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +4 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
Quoted 8 lines
Quoted 6 lines
mm/migrate.c +1 -0
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2

NAK to that patch above, here's how I think it should be:

OK. I will resend your fix with __split_huge_page() fixes against Linus’s tree. My folio_split() will conflict with the fix, but the merge fix should be simple, since the related patch just deletes __split_huge_page() entirely.

Quoted 30 lines
---
 mm/migrate.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
mm/migrate.c +5 -5
index fb19a18892c8..822776819ca6 100644
@@ -518,12 +518,12 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 	if (folio_test_anon(folio) && folio_test_large(folio))
 		mod_mthp_stat(folio_order(folio), MTHP_STAT_NR_ANON, 1);
 	folio_ref_add(newfolio, nr); /* add cache reference */
-	if (folio_test_swapbacked(folio)) {
+	if (folio_test_swapbacked(folio))
 		__folio_set_swapbacked(newfolio);
-		if (folio_test_swapcache(folio)) {
-			folio_set_swapcache(newfolio);
-			newfolio->private = folio_get_private(folio);
-		}
+
+	if (folio_test_swapcache(folio)) {
+		folio_set_swapcache(newfolio);
+		newfolio->private = folio_get_private(folio);
 		entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
Signature · 1 line
2.43.0

Best Regards,
Yan, Zi

author reviewer
#7 lore raw
Quoted 17 lines · Zi Yan
Quoted 15 lines · Hugh Dickins
Quoted 12 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
Quoted 10 lines
Quoted 2 lines

It's a great find (I think), and your commit message is okay:
but unless I'm much mistaken, NAK to the patch itself.

Got it. Thank you for the review.

Quoted 5 lines

First, I say "(I think)" there, because I don't actually know what the loop writing the same folio nr times to the multi-index entry does to the xarray: I can imagine it as being completely harmless, just nr times more work than was needed.

It seems that you are right on this one. I am trying to reproduce the issue on mainline but could not and I did see shmem hits the entries = nr. So it is likely there is no bug in mainline just inefficiency.

This fix might just mask the bugs introduced in my folio_split() patchset, since I reverted my xas_try_split() in shmem_large_split_entry() patch and still hit the issue. Let me do more debugging and get back.

Quoted 63 lines
Quoted 12 lines

But I guess it does something bad, since Matthew was horrified, and we have all found that your patch appears to improve behaviour (or at least improve behaviour in the context of your folio_split()
series: none of us noticed a problem before that, but it may be that your new series is widening our exposure to existing bugs).

Maybe your orginal patch, with the shmem_mapping(mapping) check there, was good, and it's only wrong when changed to !folio_test_anon(folio); but TBH I find it too confusing, with the conditionals the way they are. See my preferred alternative below.

Quoted 7 lines

The vital point is that multi-index entries are not used in swap cache:
whether the folio in question orginates from anon or from shmem. And it's easier to understand once you remember that a shmem folio is never in both page cache and swap cache at the same time (well, there may be an instant of transition from one to other while that folio is held locked) - once it's in swap cache, folio->mapping is NULL and it's no longer recognizable as from a shmem mapping.

Got it. Now it all makes sense to me. Thank you for the explanation.

Quoted 7 lines

The way I read your patch originally, I thought it meant that shmem folios go into the swap cache as multi-index, but anon folios do not; which seemed a worrying mixture to me. But crashes on the VM_BUG_ON_PAGE(entry != folio, entry) in __delete_from_swap_cache() yesterday (with your patch in) led me to see how add_to_swap_cache() inserts multiple non-multi-index entries, whether for anon or for shmem.

Thanks for the pointer.

Quoted 22 lines

If this patch really is needed in old releases, then I suspect that mm/huge_memory.c needs correction there too; but let me explain in a response to your folio_split() series.

Quoted 17 lines
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +4 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
Quoted 8 lines
Quoted 6 lines
mm/migrate.c +1 -0
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2

NAK to that patch above, here's how I think it should be:

Quoted 3 lines

OK. I will resend your fix with __split_huge_page() fixes against Linus’s tree. My folio_split() will conflict with the fix, but the merge fix should be simple, since the related patch just deletes __split_huge_page() entirely.

Best Regards,
Yan, Zi

author reviewer
#8 lore raw
Quoted 30 lines · Zi Yan
Quoted 17 lines · Zi Yan
Quoted 15 lines · Hugh Dickins
Quoted 12 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
Quoted 10 lines
Quoted 2 lines

It's a great find (I think), and your commit message is okay:
but unless I'm much mistaken, NAK to the patch itself.

Got it. Thank you for the review.

Quoted 5 lines

First, I say "(I think)" there, because I don't actually know what the loop writing the same folio nr times to the multi-index entry does to the xarray: I can imagine it as being completely harmless, just nr times more work than was needed.

Quoted 7 lines

It seems that you are right on this one. I am trying to reproduce the issue on mainline but could not and I did see shmem hits the entries = nr. So it is likely there is no bug in mainline just inefficiency.

This fix might just mask the bugs introduced in my folio_split() patchset, since I reverted my xas_try_split() in shmem_large_split_entry() patch and still hit the issue. Let me do more debugging and get back.

I need to take this back. It turns out I did not turn on large folio on shmem when I was testing 6.14-rc5. After turning on 64KB only large folio on shmem, shmem swapin got stuck using the repro from Liu Shixin (running compact_memory all the time then doing linear shmem swapin). But if I turn on 2MB large folio on shmem, there is no issue.

I get no issue with v6.13 either. So this issue seems from 6.14-rc. I am going to rebase my folio_split() patchset on v6.13 to test the uniform split part (the non-uniform part would need Baolin’s patchset).

Quoted 64 lines
Quoted 64 lines
Quoted 13 lines

But I guess it does something bad, since Matthew was horrified, and we have all found that your patch appears to improve behaviour (or at least improve behaviour in the context of your folio_split()
series: none of us noticed a problem before that, but it may be that your new series is widening our exposure to existing bugs).

Maybe your orginal patch, with the shmem_mapping(mapping) check there, was good, and it's only wrong when changed to !folio_test_anon(folio); but TBH I find it too confusing, with the conditionals the way they are. See my preferred alternative below.

Quoted 7 lines

The vital point is that multi-index entries are not used in swap cache:
whether the folio in question orginates from anon or from shmem. And it's easier to understand once you remember that a shmem folio is never in both page cache and swap cache at the same time (well, there may be an instant of transition from one to other while that folio is held locked) - once it's in swap cache, folio->mapping is NULL and it's no longer recognizable as from a shmem mapping.

Got it. Now it all makes sense to me. Thank you for the explanation.

Quoted 7 lines

The way I read your patch originally, I thought it meant that shmem folios go into the swap cache as multi-index, but anon folios do not; which seemed a worrying mixture to me. But crashes on the VM_BUG_ON_PAGE(entry != folio, entry) in __delete_from_swap_cache() yesterday (with your patch in) led me to see how add_to_swap_cache() inserts multiple non-multi-index entries, whether for anon or for shmem.

Thanks for the pointer.

Quoted 30 lines

If this patch really is needed in old releases, then I suspect that mm/huge_memory.c needs correction there too; but let me explain in a response to your folio_split() series.

Quoted 23 lines
---
 mm/migrate.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
mm/migrate.c +5 -1
index 365c6daa8d1b..2c9669135a38 100644
@@ -524,7 +524,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (!folio_test_anon(folio))
+			entries = 1;
+		else
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
Signature · 1 line
2.47.2

NAK to that patch above, here's how I think it should be:

Quoted 6 lines
Quoted 3 lines

OK. I will resend your fix with __split_huge_page() fixes against Linus’s tree. My folio_split() will conflict with the fix, but the merge fix should be simple, since the related patch just deletes __split_huge_page() entirely.

Best Regards,
Yan, Zi

Best Regards,
Yan, Zi

author reviewer
#9 lore raw
In reply to Hugh Dickins
Quoted 49 lines · Hugh Dickins
Quoted 7 lines · Zi Yan

Pagecache uses multi-index entries for large folio, so does shmem. Only swap cache still stores multiple entries for a single large folio. Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly") fixed swap cache but got shmem wrong by storing multiple entries for a large shmem folio. Fix it by storing a single entry for a shmem folio.

Quoted 5 lines
Fixes fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")

It's a great find (I think), and your commit message is okay:
but unless I'm much mistaken, NAK to the patch itself.

First, I say "(I think)" there, because I don't actually know what the loop writing the same folio nr times to the multi-index entry does to the xarray: I can imagine it as being completely harmless, just nr times more work than was needed.

But I guess it does something bad, since Matthew was horrified, and we have all found that your patch appears to improve behaviour (or at least improve behaviour in the context of your folio_split()
series: none of us noticed a problem before that, but it may be that your new series is widening our exposure to existing bugs).

Maybe your orginal patch, with the shmem_mapping(mapping) check there, was good, and it's only wrong when changed to !folio_test_anon(folio); but TBH I find it too confusing, with the conditionals the way they are. See my preferred alternative below.

The vital point is that multi-index entries are not used in swap cache:
whether the folio in question orginates from anon or from shmem. And it's easier to understand once you remember that a shmem folio is never in both page cache and swap cache at the same time (well, there may be an instant of transition from one to other while that folio is held locked) - once it's in swap cache, folio->mapping is NULL and it's no longer recognizable as from a shmem mapping.

The way I read your patch originally, I thought it meant that shmem folios go into the swap cache as multi-index, but anon folios do not; which seemed a worrying mixture to me. But crashes on the VM_BUG_ON_PAGE(entry != folio, entry) in __delete_from_swap_cache() yesterday (with your patch in) led me to see how add_to_swap_cache() inserts multiple non-multi-index entries, whether for anon or for shmem.

Quoted 3 lines

If this patch really is needed in old releases, then I suspect that mm/huge_memory.c needs correction there too; but let me explain in a response to your folio_split() series.

Actually, mm/huge_memory.c does not need the change, because split_huge_page*() bails with EBUSY when folio->mapping is NULL. So __split_huge_page() will not encounter large shmem folio in swap cache case.

Best Regards,
Yan, Zi