mm: don't check VMA write permissions if the PTE/PMD indicates write permissions
Subsystems (3)
PATCH 1/1 mm: don't check VMA write permissions if the PTE/PMD indicates write permissions
Staring at the comment "Recheck VMA as permissions can change since migration started" in remove_migration_pte() can result in confusion, because if the source PTE/PMD indicates write permissions, then there should be no need to check VMA write permissions when restoring migration entries or PTE-mapping a PMD.
Commit d3cb8bf6081b ("mm: migrate: Close race between migration completion and mprotect") introduced the maybe_mkwrite() handling in remove_migration_pte() in 2014, stating that a race between mprotect() and migration finishing would be possible, and that we could end up with a writable PTE that should be readable.
However, mprotect() code first updates vma->vm_flags / vma->vm_page_prot and then walks the page tables to (a) set all present writable PTEs to read-only and (b) convert all writable migration entries to readable migration entries. While walking the page tables and modifying the entries, migration code has to grab the PT locks to synchronize against concurrent page table modifications.
Assuming migration would find a writable migration entry (while holding the PT lock) and replace it with a writable present PTE, surely mprotect() code didn't stumble over the writable migration entry yet (converting it into a readable migration entry) and would instead wait for the PT lock to convert the now present writable PTE into a read-only PTE. As mprotect() didn't finish yet, the behavior is just like migration didn't happen: a writable PTE will be converted to a read-only PTE.
So it's fine to rely on the writability information in the source PTE/PMD and not recheck against the VMA as long as we're holding the PT lock to synchronize with anyone who concurrently wants to downgrade write permissions (like mprotect()) by first adjusting vma->vm_flags / vma->vm_page_prot to then walk over the page tables to adjust the page table entries.
Running test cases that should reveal such races -- mprotect(PROT_READ) racing with page migration or THP splitting -- for multiple hours did not reveal an issue with this cleanup.
---
This is a follow-up cleanup to [1]:
[PATCH v1 RESEND 0/6] mm: (pte|pmd)_mkdirty() should not unconditionally allow for write access
I wanted to be a bit careful and write some test cases to convince myself that I am not missing something important. Of course, there is still the possibility that my test cases are buggy ;)
Test cases I'm running:
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_migration.c
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_thp_split.c
[1] https://lkml.kernel.org/r/[email protected]
--- mm/huge_memory.c | 4 ++-- mm/migrate.c | 5 +---- 2 files changed, 3 insertions(+), 6 deletions(-)
mm/huge_memory.c +2 -2
@@ -2234,7 +2234,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, }else { entry =mk_pte (page + i,READ_ONCE (vma->vm_page_prot));if (write) - entry =maybe_mkwrite (entry, vma); + entry =pte_mkwrite (entry);if (anon_exclusive)SetPageAnonExclusive (page + i);if (!young)
@@ -3271,7 +3271,7 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new)if (pmd_swp_soft_dirty (*pvmw->pmd)) pmde =pmd_mksoft_dirty (pmde);if (is_writable_migration_entry (entry)) - pmde =maybe_pmd_mkwrite (pmde, vma); + pmde =pmd_mkwrite (pmde);if (pmd_swp_uffd_wp (*pvmw->pmd)) pmde =pmd_mkuffd_wp (pmde);if (!is_migration_entry_young (entry))
mm/migrate.c +1 -4
@@ -213,16 +213,13 @@ static bool remove_migration_pte(struct folio *folio,if (pte_swp_soft_dirty (*pvmw.pte)) pte =pte_mksoft_dirty (pte); -/* -* Recheck VMA as permissions can change since migration started -*/ entry =pte_to_swp_entry (*pvmw.pte);if (!is_migration_entry_young (entry)) pte =pte_mkold (pte);if (folio_test_dirty (folio) &&is_migration_entry_dirty (entry)) pte =pte_mkdirty (pte);if (is_writable_migration_entry (entry)) - pte =maybe_mkwrite (pte, vma); + pte =pte_mkwrite (pte);else if (pte_swp_uffd_wp (*pvmw.pte)) pte =pte_mkuffd_wp (pte);
Signature · 1 line
2.39.2
Quoted 12 lines · David Hildenbrand
Staring at the comment "Recheck VMA as permissions can change since migration started" in
remove_migration_pte()can result in confusion, because if the source PTE/PMD indicates write permissions, then there should be no need to check VMA write permissions when restoring migration entries or PTE-mapping a PMD.Commit
d3cb8bf6081b("mm: migrate: Close race between migration completion and mprotect") introduced themaybe_mkwrite()handling inremove_migration_pte()in 2014, stating that a race betweenmprotect()and migration finishing would be possible, and that we could end up with a writable PTE that should be readable.
Quoted 6 lines
However,
mprotect()code first updatesvma->vm_flags/vma->vm_page_protand then walks the page tables to (a) set all present writable PTEs to read-only and (b) convert all writable migration entries to readable migration entries. While walking the page tables and modifying the entries, migration code has to grab the PT locks to synchronize against concurrent page table modifications.
Makes sense to me.
Quoted 51 lines
Assuming migration would find a writable migration entry (while holding the PT lock) and replace it with a writable present PTE, surely
mprotect()code didn't stumble over the writable migration entry yet (converting it into a readable migration entry) and would instead wait for the PT lock to convert the now present writable PTE into a read-only PTE. Asmprotect()didn't finish yet, the behavior is just like migration didn't happen: a writable PTE will be converted to a read-only PTE.So it's fine to rely on the writability information in the source PTE/PMD and not recheck against the VMA as long as we're holding the PT lock to synchronize with anyone who concurrently wants to downgrade write permissions (like
mprotect()) by first adjustingvma->vm_flags/vma->vm_page_protto then walk over the page tables to adjust the page table entries.Running test cases that should reveal such races --
mprotect(PROT_READ)racing with page migration or THP splitting -- for multiple hours did not reveal an issue with this cleanup.Cc Andrew Morton <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Xu <[email protected]>Signed-off-by David Hildenbrand <[email protected]>---
This is a follow-up cleanup to [1]:
[PATCH v1 RESEND 0/6] mm: (pte|pmd)_mkdirty() should not unconditionally allow for write accessI wanted to be a bit careful and write some test cases to convince myself that I am not missing something important. Of course, there is still the possibility that my test cases are buggy ;)
Test cases I'm running:
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_migration.c
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_thp_split.c[1] https://lkml.kernel.org/r/[email protected]
--- mm/huge_memory.c | 4 ++-- mm/migrate.c | 5 +---- 2 files changed, 3 insertions(+), 6 deletions(-)mm/huge_memory.c +0
-0
mm/huge_memory.c static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, · line 2237
Quoted 5 lines
mm/huge_memory.c +1
-1@@ -2234,7 +2234,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, }else { entry =mk_pte (page + i,READ_ONCE (vma->vm_page_prot));if (write) - entry =maybe_mkwrite (entry, vma); + entry =pte_mkwrite (entry);
This is another change besides page migration. I also don't know why it's needed, but it's there since day 1 of thp split in eef1b3ba053, so maybe worthwhile to copy Kirill too (which I did).
Quoted 37 lines
mm/huge_memory.c +1
-1if (anon_exclusive)SetPageAnonExclusive (page + i);if (!young)@@ -3271,7 +3271,7 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new)if (pmd_swp_soft_dirty (*pvmw->pmd)) pmde =pmd_mksoft_dirty (pmde);if (is_writable_migration_entry (entry)) - pmde =maybe_pmd_mkwrite (pmde, vma); + pmde =pmd_mkwrite (pmde);if (pmd_swp_uffd_wp (*pvmw->pmd)) pmde =pmd_mkuffd_wp (pmde);if (!is_migration_entry_young (entry))mm/migrate.c +1
-4@@ -213,16 +213,13 @@ static bool remove_migration_pte(struct folio *folio,if (pte_swp_soft_dirty (*pvmw.pte)) pte =pte_mksoft_dirty (pte); -/* -* Recheck VMA as permissions can change since migration started -*/ entry =pte_to_swp_entry (*pvmw.pte);if (!is_migration_entry_young (entry)) pte =pte_mkold (pte);if (folio_test_dirty (folio) &&is_migration_entry_dirty (entry)) pte =pte_mkdirty (pte);if (is_writable_migration_entry (entry)) - pte =maybe_mkwrite (pte, vma); + pte =pte_mkwrite (pte);else if (pte_swp_uffd_wp (*pvmw.pte)) pte =pte_mkuffd_wp (pte);Signature · 1 line
2.39.2
Signature · 1 line
Peter Xu
Quoted 68 lines · Peter Xu
Quoted 12 lines · David Hildenbrand
Staring at the comment "Recheck VMA as permissions can change since migration started" in
remove_migration_pte()can result in confusion, because if the source PTE/PMD indicates write permissions, then there should be no need to check VMA write permissions when restoring migration entries or PTE-mapping a PMD.Commit
d3cb8bf6081b("mm: migrate: Close race between migration completion and mprotect") introduced themaybe_mkwrite()handling inremove_migration_pte()in 2014, stating that a race betweenmprotect()and migration finishing would be possible, and that we could end up with a writable PTE that should be readable.Quoted 6 lines
However,
mprotect()code first updatesvma->vm_flags/vma->vm_page_protand then walks the page tables to (a) set all present writable PTEs to read-only and (b) convert all writable migration entries to readable migration entries. While walking the page tables and modifying the entries, migration code has to grab the PT locks to synchronize against concurrent page table modifications.Makes sense to me.
Quoted 46 lines
Assuming migration would find a writable migration entry (while holding the PT lock) and replace it with a writable present PTE, surely
mprotect()code didn't stumble over the writable migration entry yet (converting it into a readable migration entry) and would instead wait for the PT lock to convert the now present writable PTE into a read-only PTE. Asmprotect()didn't finish yet, the behavior is just like migration didn't happen: a writable PTE will be converted to a read-only PTE.So it's fine to rely on the writability information in the source PTE/PMD and not recheck against the VMA as long as we're holding the PT lock to synchronize with anyone who concurrently wants to downgrade write permissions (like
mprotect()) by first adjustingvma->vm_flags/vma->vm_page_protto then walk over the page tables to adjust the page table entries.Running test cases that should reveal such races --
mprotect(PROT_READ)racing with page migration or THP splitting -- for multiple hours did not reveal an issue with this cleanup.Cc Andrew Morton <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Xu <[email protected]>Signed-off-by David Hildenbrand <[email protected]>---
This is a follow-up cleanup to [1]:
[PATCH v1 RESEND 0/6] mm: (pte|pmd)_mkdirty() should not unconditionally allow for write accessI wanted to be a bit careful and write some test cases to convince myself that I am not missing something important. Of course, there is still the possibility that my test cases are buggy ;)
Test cases I'm running:
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_migration.c
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_thp_split.c[1] https://lkml.kernel.org/r/[email protected]
--- mm/huge_memory.c | 4 ++-- mm/migrate.c | 5 +---- 2 files changed, 3 insertions(+), 6 deletions(-)
Quoted 14 lines
mm/huge_memory.c static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, · line 2237
Quoted 10 lines
mm/huge_memory.c +1
-1@@ -2234,7 +2234,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, }else { entry =mk_pte (page + i,READ_ONCE (vma->vm_page_prot));if (write) - entry =maybe_mkwrite (entry, vma); + entry =pte_mkwrite (entry);This is another change besides page migration. I also don't know why it's needed, but it's there since day 1 of thp split in
eef1b3ba053, so maybe worthwhile to copy Kirill too (which I did).
Indeed (I wanted but forgot ...), thanks Peter!
Signature · 3 lines
Thanks, David / dhildenb
Kirill A. Shutemov
Acked-byQuoted 68 lines · Peter Xu
Quoted 12 lines · David Hildenbrand
Staring at the comment "Recheck VMA as permissions can change since migration started" in
remove_migration_pte()can result in confusion, because if the source PTE/PMD indicates write permissions, then there should be no need to check VMA write permissions when restoring migration entries or PTE-mapping a PMD.Commit
d3cb8bf6081b("mm: migrate: Close race between migration completion and mprotect") introduced themaybe_mkwrite()handling inremove_migration_pte()in 2014, stating that a race betweenmprotect()and migration finishing would be possible, and that we could end up with a writable PTE that should be readable.Quoted 6 lines
However,
mprotect()code first updatesvma->vm_flags/vma->vm_page_protand then walks the page tables to (a) set all present writable PTEs to read-only and (b) convert all writable migration entries to readable migration entries. While walking the page tables and modifying the entries, migration code has to grab the PT locks to synchronize against concurrent page table modifications.Makes sense to me.
Quoted 46 lines
Assuming migration would find a writable migration entry (while holding the PT lock) and replace it with a writable present PTE, surely
mprotect()code didn't stumble over the writable migration entry yet (converting it into a readable migration entry) and would instead wait for the PT lock to convert the now present writable PTE into a read-only PTE. Asmprotect()didn't finish yet, the behavior is just like migration didn't happen: a writable PTE will be converted to a read-only PTE.So it's fine to rely on the writability information in the source PTE/PMD and not recheck against the VMA as long as we're holding the PT lock to synchronize with anyone who concurrently wants to downgrade write permissions (like
mprotect()) by first adjustingvma->vm_flags/vma->vm_page_protto then walk over the page tables to adjust the page table entries.Running test cases that should reveal such races --
mprotect(PROT_READ)racing with page migration or THP splitting -- for multiple hours did not reveal an issue with this cleanup.Cc Andrew Morton <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Xu <[email protected]>Signed-off-by David Hildenbrand <[email protected]>---
This is a follow-up cleanup to [1]:
[PATCH v1 RESEND 0/6] mm: (pte|pmd)_mkdirty() should not unconditionally allow for write accessI wanted to be a bit careful and write some test cases to convince myself that I am not missing something important. Of course, there is still the possibility that my test cases are buggy ;)
Test cases I'm running:
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_migration.c
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_thp_split.c[1] https://lkml.kernel.org/r/[email protected]
--- mm/huge_memory.c | 4 ++-- mm/migrate.c | 5 +---- 2 files changed, 3 insertions(+), 6 deletions(-)
Quoted 14 lines
mm/huge_memory.c static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, · line 2237
Quoted 10 lines
mm/huge_memory.c +1
-1@@ -2234,7 +2234,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, }else { entry =mk_pte (page + i,READ_ONCE (vma->vm_page_prot));if (write) - entry =maybe_mkwrite (entry, vma); + entry =pte_mkwrite (entry);This is another change besides page migration. I also don't know why it's needed, but it's there since day 1 of thp split in
eef1b3ba053, so maybe worthwhile to copy Kirill too (which I did).
I was concentrated on the correctness at the point and this small inefficency didn't catch my eyes.
I was curious how we serialize here against mprotect().
Looks safe to me:
CPU0 CPU1
__split_huge_pmd()
pmd_lock()
__split_huge_pmd_locked()
pmdp_invalidate()
// PMD is non-present, but huge at this point
change_protection()
change_pmd_range()
pmd_none_or_clear_bad_unless_trans_huge() == 0 // not skipped
change_huge_pmd()
__pmd_trans_huge_lock()
pmd_lock() // serialized against __split_huge_pmd()Signature · 1 line
Kiryl Shutsemau / Kirill A. Shutemov
Quoted 5 lines · David Hildenbrand
Staring at the comment "Recheck VMA as permissions can change since migration started" in
remove_migration_pte()can result in confusion, because if the source PTE/PMD indicates write permissions, then there should be no need to check VMA write permissions when restoring migration entries or PTE-mapping a PMD.
Thanks David, I have oft wondered about that but not stared at it to the point of confusion. The change looks correct to me so feel free to add:
For the mm/migrate.c parts. Also presumably if mprotect(PROT_READ) was a problem then mprotect(PROT_NONE) would also need some kind of special handling which I don't see.
Quoted 101 lines
Commit
d3cb8bf6081b("mm: migrate: Close race between migration completion and mprotect") introduced themaybe_mkwrite()handling inremove_migration_pte()in 2014, stating that a race betweenmprotect()and migration finishing would be possible, and that we could end up with a writable PTE that should be readable.However,
mprotect()code first updatesvma->vm_flags/vma->vm_page_protand then walks the page tables to (a) set all present writable PTEs to read-only and (b) convert all writable migration entries to readable migration entries. While walking the page tables and modifying the entries, migration code has to grab the PT locks to synchronize against concurrent page table modifications.Assuming migration would find a writable migration entry (while holding the PT lock) and replace it with a writable present PTE, surely
mprotect()code didn't stumble over the writable migration entry yet (converting it into a readable migration entry) and would instead wait for the PT lock to convert the now present writable PTE into a read-only PTE. Asmprotect()didn't finish yet, the behavior is just like migration didn't happen: a writable PTE will be converted to a read-only PTE.So it's fine to rely on the writability information in the source PTE/PMD and not recheck against the VMA as long as we're holding the PT lock to synchronize with anyone who concurrently wants to downgrade write permissions (like
mprotect()) by first adjustingvma->vm_flags/vma->vm_page_protto then walk over the page tables to adjust the page table entries.Running test cases that should reveal such races --
mprotect(PROT_READ)racing with page migration or THP splitting -- for multiple hours did not reveal an issue with this cleanup.Cc Andrew Morton <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Xu <[email protected]>Signed-off-by David Hildenbrand <[email protected]>---
This is a follow-up cleanup to [1]:
[PATCH v1 RESEND 0/6] mm: (pte|pmd)_mkdirty() should not unconditionally allow for write accessI wanted to be a bit careful and write some test cases to convince myself that I am not missing something important. Of course, there is still the possibility that my test cases are buggy ;)
Test cases I'm running:
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_migration.c
https://gitlab.com/davidhildenbrand/scratchspace/-/raw/main/test_mprotect_thp_split.c[1] https://lkml.kernel.org/r/[email protected]
--- mm/huge_memory.c | 4 ++-- mm/migrate.c | 5 +---- 2 files changed, 3 insertions(+), 6 deletions(-)mm/huge_memory.c +2
-2@@ -2234,7 +2234,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, }else { entry =mk_pte (page + i,READ_ONCE (vma->vm_page_prot));if (write) - entry =maybe_mkwrite (entry, vma); + entry =pte_mkwrite (entry);if (anon_exclusive)SetPageAnonExclusive (page + i);if (!young)@@ -3271,7 +3271,7 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new)if (pmd_swp_soft_dirty (*pvmw->pmd)) pmde =pmd_mksoft_dirty (pmde);if (is_writable_migration_entry (entry)) - pmde =maybe_pmd_mkwrite (pmde, vma); + pmde =pmd_mkwrite (pmde);if (pmd_swp_uffd_wp (*pvmw->pmd)) pmde =pmd_mkuffd_wp (pmde);if (!is_migration_entry_young (entry))mm/migrate.c +1
-4@@ -213,16 +213,13 @@ static bool remove_migration_pte(struct folio *folio,if (pte_swp_soft_dirty (*pvmw.pte)) pte =pte_mksoft_dirty (pte); -/* -* Recheck VMA as permissions can change since migration started -*/ entry =pte_to_swp_entry (*pvmw.pte);if (!is_migration_entry_young (entry)) pte =pte_mkold (pte);if (folio_test_dirty (folio) &&is_migration_entry_dirty (entry)) pte =pte_mkdirty (pte);if (is_writable_migration_entry (entry)) - pte =maybe_mkwrite (pte, vma); + pte =pte_mkwrite (pte);else if (pte_swp_uffd_wp (*pvmw.pte)) pte =pte_mkuffd_wp (pte);