s390/mm: Allow hugepage migration for s390
Gerald Schaefer
maintainerPATCH v2 1/1 s390/mm: Allow hugepage migration for s390
After hugetlbfs PTE_MARKER support for s390 introduced region-third and segment table swap entries, it is possible to also allow hugepage migration for s390.
This fixes the problem that hugetlbfs pages could not be allocated from ZONE_MOVABLE memory, which is often used in combination with memory hotplug. Therefore, without support for hugepage migration, a large portion of system memory could be unavailable for hugetlbfs page allocations.
Besides selecting ARCH_ENABLE_HUGEPAGE_MIGRATION, this only requires definition of arch_hugetlb_migration_supported(), which can make use of the existing arch_hugetlb_valid_size() function, similar to other architectures.
--- arch/s390/Kconfig | 1 + arch/s390/include/asm/hugetlb.h | 5 +++++ arch/s390/mm/hugetlbpage.c | 14 +++++++++++++- 3 files changed, 19 insertions(+), 1 deletion(-)
---
Changes in v2:
- Use "#if IS_ENABLED()" instead of "#ifdef" in
arch/s390/mm/hugetlbpage.c
arch/s390/Kconfig +1 -0
@@ -83,6 +83,7 @@ config S390 select ALTERNATE_USER_ADDRESS_SPACE select ARCH_32BIT_USTAT_F_TINODE select ARCH_CORRECT_STACKTRACE_ON_KRETPROBE + select ARCH_ENABLE_HUGEPAGE_MIGRATION if HUGETLB_PAGE && MIGRATION select ARCH_ENABLE_MEMORY_HOTPLUG if SPARSEMEM select ARCH_ENABLE_SPLIT_PMD_PTLOCK if PGTABLE_LEVELS > 2 select ARCH_HAS_PMD_SOFTLEAVES if TRANSPARENT_HUGEPAGE
arch/s390/include/asm/hugetlb.h +5 -0
@@ -17,6 +17,11 @@#define hugepages_supported () cpu_has_edat1() +#ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION +bool arch_hugetlb_migration_supported (struct hstate *h); +#define arch_hugetlb_migration_supported arch_hugetlb_migration_supported +#endif +#define __HAVE_ARCH_HUGE_SET_HUGE_PTE_ATvoid set_huge_pte_at (struct mm_struct *mm,unsigned long addr,pte_t *ptep,pte_t pte,unsigned long sz);
arch/s390/mm/hugetlbpage.c +13 -1
@@ -224,7 +224,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm,return (pte_t *) pmdp; } -bool __init arch_hugetlb_valid_size (unsigned long size) +static bool __hugetlb_valid_size (unsigned long size) {if (cpu_has_edat1 () && size ==PMD_SIZE )return true;
@@ -241,3 +241,15 @@ unsigned int __init arch_hugetlb_cma_order(void)return 0 ; } + +bool __init arch_hugetlb_valid_size (unsigned long size) +{ +return __hugetlb_valid_size (size); +} + +#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{ +return __hugetlb_valid_size (huge_page_size (h)); +} +#endif
Signature · 1 line
2.53.0
[email protected]
AI reviewThank you for your contribution! Sashiko AI review found 2 potential issues to consider: - [High] s390/mm: guest storage key loss on 1M/2G hugetlbfs page migration - [Medium] s390/mm: loss of…
Thank you for your contribution! Sashiko AI review found 2 potential issues to consider:
- [High] s390/mm: guest storage key loss on 1M/2G hugetlbfs page migration
- [Medium] s390/mm: loss of soft-dirty bit in __pte_to_rste() and __rste_to_pte() on hugetlb migration
--
commit d03e5331e4e0b88aa75246fc7dc6aef18e7c8cdd
Author: Gerald Schaefer <[email protected]>
s390/mm: Allow hugepage migration for s390
This commit selects ARCH_ENABLE_HUGEPAGE_MIGRATION on s390 and defines arch_hugetlb_migration_supported() for 1M and 2G hugepages so hugetlbfs pages can be migrated and allocated from ZONE_MOVABLE.
Quoted 19 lines
arch/s390/mm/hugetlbpage.c +3
-1@@ -224,7 +224,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm,return (pte_t *) pmdp; } -bool __init arch_hugetlb_valid_size (unsigned long size) +static bool __hugetlb_valid_size (unsigned long size) {if (cpu_has_edat1 () && size ==PMD_SIZE )return true;@@ -241,3 +241,15 @@ unsigned int __init arch_hugetlb_cma_order(void)return 0 ; } + +bool __init arch_hugetlb_valid_size (unsigned long size)
arch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 255
Quoted 10 lines
arch/s390/mm/hugetlbpage.c +10
-0+{ +return __hugetlb_valid_size (size); +} + +#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{ +return __hugetlb_valid_size (huge_page_size (h)); +} +#endif
[Severity: High]
Does migrating a 1M or 2G hugetlbfs folio backing a KVM guest lose the guest's hardware storage keys?
For 1M and 2G mappings (KVM_CAP_S390_HPAGE_1M / KVM_CAP_S390_HPAGE_2G), the KVM gmap uses large leaf CRSTEs without PGSTEs, and guest storage keys live only in the physical page frames. When unmap_and_move_hugetlb_folio() unmaps the folio via mmu_notifier_invalidate_range_start(), _gmap_unmap_crste() clears the large CRSTE to _CRSTE_EMPTY without saving or splitting the hardware storage keys:
arch/s390/kvm/gmap/gmap.c:_gmap_unmap_crste() {
...
if (!old.h.fc)
return 0;
if (old.s.fc1.pr && test_bit(GMAP_FLAG_EXPORT_ON_UNMAP, &priv->gmap->flags))
folio = phys_to_folio(crste_origin_large(old));
ok = gmap_crstep_xchg_atomic(priv->gmap, crstep, old, _CRSTE_EMPTY(old.h.tt), gfn);
...
}folio_copy() -> copy_highpage() -> copy_page() then copies only the 4K page contents via mvcl and does not copy the hardware storage keys to the destination frames, leaving the guest with stale storage keys and causing guest protection exceptions or memory corruption.
[Severity: Medium]
Does migrating a hugetlbfs page also drop its soft-dirty bit in __pte_to_rste() and __rste_to_pte()?
When try_to_migrate_one() installs a hugetlb migration entry, it sets _PAGE_SOFT_DIRTY via pte_swp_mksoft_dirty() and calls set_huge_pte_at(). However, the swap entry branches in __pte_to_rste() and __rste_to_pte() do not transfer _PAGE_SOFT_DIRTY to and from _SEGMENT_ENTRY_SOFT_DIRTY:
arch/s390/mm/hugetlbpage.c:__pte_to_rste() {
...
} else if (!pte_none(pte)) {
/* swap pte */
arch_entry = __pte_to_swp_entry(pte);
rste = mk_swap_rste(__swp_type(arch_entry), __swp_offset(arch_entry));
...
}arch/s390/mm/hugetlbpage.c:__rste_to_pte() {
...
} else if (!none) {
/* swap rste */
arch_entry = __rste_to_swp_entry(rste);
pte = mk_swap_pte(__swp_type_rste(arch_entry), __swp_offset_rste(arch_entry));
pteval = pte_val(pte);
...
}As a result, remove_migration_pte() reads back old_pte via huge_ptep_get() with pte_swp_soft_dirty(old_pte) == 0 and clears _PAGE_SOFT_DIRTY on the restored hugepage PTE, breaking soft-dirty tracking via /proc/pid/pagemap.
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
Alexander Gordeev
maintainerarch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 252
Quoted 3 lines · Gerald Schaefer
arch/s390/mm/hugetlbpage.c +3
-0+#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{
Oh, would you agree to such one? ;)
return IS_ENABLED(CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION) && __hugetlb_valid_size(huge_page_size(h));
Quoted 3 lines
arch/s390/mm/hugetlbpage.c +3
-0+ return __hugetlb_valid_size(huge_page_size(h)); +} +#endif
Gerald Schaefer
maintainerQuoted 4 lines · On Wed, 11 Nov 2026 19:14:22 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 2 potential issues to consider:
- [High] s390/mm: guest storage key loss on 1M/2G hugetlbfs page migration
@Claudio, could you please have a look, if this is a valid issue?
Quoted 2 lines
- [Medium] s390/mm: loss of soft-dirty bit in
__pte_to_rste()and__rste_to_pte()on hugetlb migration
No problem, soft-dirty tracking not supported for hugetlbfs, see below.
Quoted 96 lines
--
commit
d03e5331e4e0b88aa75246fc7dc6aef18e7c8cdd
Author: Gerald Schaefer <[email protected]>s390/mm: Allow hugepage migration for s390
This commit selects
ARCH_ENABLE_HUGEPAGE_MIGRATIONon s390 and definesarch_hugetlb_migration_supported()for 1M and 2G hugepages so hugetlbfs pages can be migrated and allocated fromZONE_MOVABLE.Quoted 19 lines
arch/s390/mm/hugetlbpage.c +3
-1@@ -224,7 +224,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm,return (pte_t *) pmdp; } -bool __init arch_hugetlb_valid_size (unsigned long size) +static bool __hugetlb_valid_size (unsigned long size) {if (cpu_has_edat1 () && size ==PMD_SIZE )return true;@@ -241,3 +241,15 @@ unsigned int __init arch_hugetlb_cma_order(void)return 0 ; } + +bool __init arch_hugetlb_valid_size (unsigned long size)arch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 255
Quoted 10 lines
arch/s390/mm/hugetlbpage.c +10
-0+{ +return __hugetlb_valid_size (size); +} + +#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{ +return __hugetlb_valid_size (huge_page_size (h)); +} +#endif [Severity: High]
Does migrating a 1M or 2G hugetlbfs folio backing a KVM guest lose the guest's hardware storage keys?For 1M and 2G mappings (
KVM_CAP_S390_HPAGE_1M/KVM_CAP_S390_HPAGE_2G), the KVM gmap uses large leaf CRSTEs without PGSTEs, and guest storage keys live only in the physical page frames. Whenunmap_and_move_hugetlb_folio()unmaps the folio viammu_notifier_invalidate_range_start(),_gmap_unmap_crste()clears the large CRSTE to_CRSTE_EMPTYwithout saving or splitting the hardware storage keys:arch/s390/kvm/gmap/gmap.c:_gmap_unmap_crste() { ... if (!old.h.fc) return 0; if (old.s.fc1.pr && test_bit(GMAP_FLAG_EXPORT_ON_UNMAP, &priv->gmap->flags)) folio = phys_to_folio(crste_origin_large(old)); ok = gmap_crstep_xchg_atomic(priv->gmap, crstep, old, _CRSTE_EMPTY(old.h.tt), gfn); ... }
folio_copy()->copy_highpage()->copy_page()then copies only the 4K page contents via mvcl and does not copy the hardware storage keys to the destination frames, leaving the guest with stale storage keys and causing guest protection exceptions or memory corruption.[Severity: Medium]
Does migrating a hugetlbfs page also drop its soft-dirty bit in__pte_to_rste()and__rste_to_pte()?When
try_to_migrate_one()installs a hugetlb migration entry, it sets_PAGE_SOFT_DIRTYviapte_swp_mksoft_dirty()and callsset_huge_pte_at(). However, the swap entry branches in__pte_to_rste()and__rste_to_pte()do not transfer_PAGE_SOFT_DIRTYto and from_SEGMENT_ENTRY_SOFT_DIRTY:arch/s390/mm/hugetlbpage.c:__pte_to_rste() { ... } else if (!pte_none(pte)) { /* swap pte */ arch_entry = __pte_to_swp_entry(pte); rste = mk_swap_rste(__swp_type(arch_entry), __swp_offset(arch_entry)); ... }arch/s390/mm/hugetlbpage.c:__rste_to_pte() { ... } else if (!none) { /* swap rste */ arch_entry = __rste_to_swp_entry(rste); pte = mk_swap_pte(__swp_type_rste(arch_entry), __swp_offset_rste(arch_entry)); pteval = pte_val(pte); ... }
Quoted 3 lines
As a result,
remove_migration_pte()reads backold_pteviahuge_ptep_get()withpte_swp_soft_dirty(old_pte)== 0 and clears_PAGE_SOFT_DIRTYon the restored hugepage PTE, breaking soft-dirty tracking via/proc/pid/pagemap.
Hugetlbfs does not support soft-dirty page tracking in the kernel:vma_wants_writenotify() in mm/vma.c explicitly excludes hugetlb VMAs from soft-dirty tracking, clear_refs_walk_ops in fs/proc/task_mmu.c does not implement a hugetlb_entry callback, and mm/hugetlb.c never sets or preserves soft-dirty bits on hugetlb PTEs or migration entries.
The RSTE swap entry soft-dirty bit (bit 62) exists for THP PMD swap/migration entries rather than hugetlbfs. Generic hugetlb migration handling in mm/hugetlb.c (copy_hugetlb_page_range() and hugetlb_change_protection()) also does not preserve pte_swp_soft_dirty() when rebuilding hugetlb migration swap entries.
However, I was not aware that try_to_migrate_one() actually does set _PAGE_SOFT_DIRTY via pte_swp_mksoft_dirty() also for hugetlbfs pages, even though it is not being used anywhere yet. This bit is indeed lost after __pte_to_rste() conversion from a swap PTE. But it does not break soft-dirty tracking for hugetlbfs, because that does not exist. It also does not conflict with or corrupt the RSTE swap type or offset calculation.
Of course, it probably would be good practice to already preserve _PAGE_SOFT_DIRTY now, even if it is not needed yet, and further changes to generic code would be required if it should ever be supported. Will look into this in v3.
Heiko Carstens
maintainerQuoted 8 lines · Alexander Gordeev
arch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 252
Quoted 3 lines · Gerald Schaefer
arch/s390/mm/hugetlbpage.c +3
-0+#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{Oh, would you agree to such one? ;)
Quoted 6 lines
return IS_ENABLED(CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION) && __hugetlb_valid_size(huge_page_size(h));arch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 255
Quoted 3 lines
arch/s390/mm/hugetlbpage.c +3
-0+ return __hugetlb_valid_size(huge_page_size(h)); +} +#endif
Out of curiosity: why?
All other architectures do ifdef, instead if IS_ENABLED to make sure the function exists only if the config option is enabled (IS_ENABLED would also make sense if "m" would be a possible option, but that is not the case here).
Then if this is only inline within the function, like suggested above, the prototype in the header does not exist (rightfully), leading to compile time warnings. The arch_hugetlb_migration_supported define also indicates the availability of the feature.
So I'm wondering what the point of this exercise is and what the deviation compared to other architectures would buy us.
Quoted 9 lines · On Thu, 12 Nov 2026 12:13:37 +0100 Gerald Schaefer…
Quoted 4 lines · On Wed, 11 Nov 2026 19:14:22 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 2 potential issues to consider:
- [High] s390/mm: guest storage key loss on 1M/2G hugetlbfs page migration@Claudio, could you please have a look, if this is a valid issue?
yes, it's a valid issue, I was already aware of it.
it's in my pipeline, will be fixed as soon as possible
Quoted 129 lines
Quoted 3 lines
- [Medium] s390/mm: loss of soft-dirty bit in
__pte_to_rste()and__rste_to_pte()on hugetlb migrationNo problem, soft-dirty tracking not supported for hugetlbfs, see below.
Quoted 96 lines
--
commit
d03e5331e4e0b88aa75246fc7dc6aef18e7c8cdd
Author: Gerald Schaefer <[email protected]>s390/mm: Allow hugepage migration for s390
This commit selects
ARCH_ENABLE_HUGEPAGE_MIGRATIONon s390 and definesarch_hugetlb_migration_supported()for 1M and 2G hugepages so hugetlbfs pages can be migrated and allocated fromZONE_MOVABLE.Quoted 19 lines
arch/s390/mm/hugetlbpage.c +3
-1@@ -224,7 +224,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm,return (pte_t *) pmdp; } -bool __init arch_hugetlb_valid_size (unsigned long size) +static bool __hugetlb_valid_size (unsigned long size) {if (cpu_has_edat1 () && size ==PMD_SIZE )return true;@@ -241,3 +241,15 @@ unsigned int __init arch_hugetlb_cma_order(void)return 0 ; } + +bool __init arch_hugetlb_valid_size (unsigned long size)arch/s390/mm/hugetlbpage.c unsigned int __init arch_hugetlb_cma_order(void) · line 255
Quoted 10 lines
arch/s390/mm/hugetlbpage.c +10
-0+{ +return __hugetlb_valid_size (size); +} + +#if IS_ENABLED (CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION ) +bool arch_hugetlb_migration_supported (struct hstate *h) +{ +return __hugetlb_valid_size (huge_page_size (h)); +} +#endif [Severity: High]
Does migrating a 1M or 2G hugetlbfs folio backing a KVM guest lose the guest's hardware storage keys?For 1M and 2G mappings (
KVM_CAP_S390_HPAGE_1M/KVM_CAP_S390_HPAGE_2G), the KVM gmap uses large leaf CRSTEs without PGSTEs, and guest storage keys live only in the physical page frames. Whenunmap_and_move_hugetlb_folio()unmaps the folio viammu_notifier_invalidate_range_start(),_gmap_unmap_crste()clears the large CRSTE to_CRSTE_EMPTYwithout saving or splitting the hardware storage keys:arch/s390/kvm/gmap/gmap.c:_gmap_unmap_crste() { ... if (!old.h.fc) return 0; if (old.s.fc1.pr && test_bit(GMAP_FLAG_EXPORT_ON_UNMAP, &priv->gmap->flags)) folio = phys_to_folio(crste_origin_large(old)); ok = gmap_crstep_xchg_atomic(priv->gmap, crstep, old, _CRSTE_EMPTY(old.h.tt), gfn); ... }
folio_copy()->copy_highpage()->copy_page()then copies only the 4K page contents via mvcl and does not copy the hardware storage keys to the destination frames, leaving the guest with stale storage keys and causing guest protection exceptions or memory corruption.[Severity: Medium]
Does migrating a hugetlbfs page also drop its soft-dirty bit in__pte_to_rste()and__rste_to_pte()?When
try_to_migrate_one()installs a hugetlb migration entry, it sets_PAGE_SOFT_DIRTYviapte_swp_mksoft_dirty()and callsset_huge_pte_at(). However, the swap entry branches in__pte_to_rste()and__rste_to_pte()do not transfer_PAGE_SOFT_DIRTYto and from_SEGMENT_ENTRY_SOFT_DIRTY:arch/s390/mm/hugetlbpage.c:__pte_to_rste() { ... } else if (!pte_none(pte)) { /* swap pte */ arch_entry = __pte_to_swp_entry(pte); rste = mk_swap_rste(__swp_type(arch_entry), __swp_offset(arch_entry)); ... }arch/s390/mm/hugetlbpage.c:__rste_to_pte() { ... } else if (!none) { /* swap rste */ arch_entry = __rste_to_swp_entry(rste); pte = mk_swap_pte(__swp_type_rste(arch_entry), __swp_offset_rste(arch_entry)); pteval = pte_val(pte); ... }Quoted 3 lines
As a result,
remove_migration_pte()reads backold_pteviahuge_ptep_get()withpte_swp_soft_dirty(old_pte)== 0 and clears_PAGE_SOFT_DIRTYon the restored hugepage PTE, breaking soft-dirty tracking via/proc/pid/pagemap.Hugetlbfs does not support soft-dirty page tracking in the kernel:
vma_wants_writenotify()inmm/vma.cexplicitly excludes hugetlb VMAs from soft-dirty tracking,clear_refs_walk_opsinfs/proc/task_mmu.cdoes not implement ahugetlb_entrycallback, andmm/hugetlb.cnever sets or preserves soft-dirty bits on hugetlb PTEs or migration entries.The RSTE swap entry soft-dirty bit (bit 62) exists for THP PMD swap/migration entries rather than hugetlbfs. Generic hugetlb migration handling in
mm/hugetlb.c(copy_hugetlb_page_range()andhugetlb_change_protection()) also does not preservepte_swp_soft_dirty()when rebuilding hugetlb migration swap entries.However, I was not aware that
try_to_migrate_one()actually does set_PAGE_SOFT_DIRTYviapte_swp_mksoft_dirty()also for hugetlbfs pages, even though it is not being used anywhere yet. This bit is indeed lost after__pte_to_rste()conversion from a swap PTE. But it does not break soft-dirty tracking for hugetlbfs, because that does not exist. It also does not conflict with or corrupt the RSTE swap type or offset calculation.Of course, it probably would be good practice to already preserve
_PAGE_SOFT_DIRTYnow, even if it is not needed yet, and further changes to generic code would be required if it should ever be supported. Will look into this in v3.