[PATCH v3 00/13] Some cleanups for memory-failure

[PATCH v3 00/13] Some cleanups for memory-failure

[Miaohe Lin]

Hi everyone,
This series contains a few cleanup patches to avoid exporting unused
function, add helper macro, fix some obsolete comments and so on. More
details can be found in the respective changelogs.
Thanks!
---
v3:
 Fix build error per kernel test robot. Thanks.
 s/extern// for function declarations per Matthew. Thanks.
v2:
 Collect Reviewed-by tag and use folio_mapped() per David. Thanks.
 Fix build error per kernel test robot. Thanks.
---
Miaohe Lin (13):
  mm/memory-failure: simplify put_ref_page()
  mm/memory-failure: remove MF_MSG_SLAB
  mm/memory-failure: add macro GET_PAGE_MAX_RETRY_NUM
  mm/memory-failure: save some page_folio() calls
  mm/memory-failure: remove unneeded empty string
  mm/memory-failure: remove confusing initialization to count
  mm/memory-failure: don't export hwpoison_filter() when
    !CONFIG_HWPOISON_INJECT
  mm/memory-failure: use helper macro task_pid_nr()
  mm/memory-failure: remove obsolete comment in unpoison_memory()
  mm/memory-failure: move some function declarations into internal.h
  mm/memory-failure: fix comment of get_hwpoison_page()
  mm/memory-failure: remove obsolete comment in kill_proc()
  mm/memory-failure: correct comment in me_swapcache_dirty

 include/linux/mm.h         |  8 -------
 include/linux/page-flags.h |  5 -----
 include/linux/rmap.h       |  2 --
 include/ras/ras_event.h    |  1 -
 mm/internal.h              | 10 +++++++++
 mm/memory-failure.c        | 46 +++++++++++++++-----------------------
 6 files changed, 28 insertions(+), 44 deletions(-)


base-commit: 8a58962d482f5ab03d698f56fcbd5449757accfa
-- 
2.33.0

[PATCH v3 01/13] mm/memory-failure: simplify put_ref_page()

[Miaohe Lin]

Remove unneeded page != NULL check. pfn_to_page() won't return NULL.
No functional change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index f679b579d45d..2e6038c73119 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -2120,14 +2120,10 @@ static inline unsigned long folio_free_raw_hwp(struct folio *folio, bool flag)
 /* Drop the extra refcount in case we come from madvise() */
 static void put_ref_page(unsigned long pfn, int flags)
 {
-	struct page *page;
-
 	if (!(flags & MF_COUNT_INCREASED))
 		return;
 
-	page = pfn_to_page(pfn);
-	if (page)
-		put_page(page);
+	put_page(pfn_to_page(pfn));
 }
 
 static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
-- 
2.33.0

[PATCH v3 02/13] mm/memory-failure: remove MF_MSG_SLAB

[Miaohe Lin]

Since commit 46df8e73a4a3 ("mm: free up PG_slab"), MF_MSG_SLAB becomes
unused. Remove it. No functional change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 include/linux/mm.h      | 1 -
 include/ras/ras_event.h | 1 -
 mm/memory-failure.c     | 1 -
 3 files changed, 3 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 25b541974134..0bf0f22b32c8 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -4048,7 +4048,6 @@ enum mf_result {
 enum mf_action_page_type {
 	MF_MSG_KERNEL,
 	MF_MSG_KERNEL_HIGH_ORDER,
-	MF_MSG_SLAB,
 	MF_MSG_DIFFERENT_COMPOUND,
 	MF_MSG_HUGE,
 	MF_MSG_FREE_HUGE,
diff --git a/include/ras/ras_event.h b/include/ras/ras_event.h
index cf7f19b7ce64..9bc707fe8819 100644
--- a/include/ras/ras_event.h
+++ b/include/ras/ras_event.h
@@ -356,7 +356,6 @@ TRACE_EVENT(aer_event,
 #define MF_PAGE_TYPE		\
 	EM ( MF_MSG_KERNEL, "reserved kernel page" )			\
 	EM ( MF_MSG_KERNEL_HIGH_ORDER, "high-order kernel page" )	\
-	EM ( MF_MSG_SLAB, "kernel slab page" )				\
 	EM ( MF_MSG_DIFFERENT_COMPOUND, "different compound page after locking" ) \
 	EM ( MF_MSG_HUGE, "huge page" )					\
 	EM ( MF_MSG_FREE_HUGE, "free huge page" )			\
diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 2e6038c73119..4b9a9298d478 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -911,7 +911,6 @@ static const char *action_name[] = {
 static const char * const action_page_types[] = {
 	[MF_MSG_KERNEL]			= "reserved kernel page",
 	[MF_MSG_KERNEL_HIGH_ORDER]	= "high-order kernel page",
-	[MF_MSG_SLAB]			= "kernel slab page",
 	[MF_MSG_DIFFERENT_COMPOUND]	= "different compound page after locking",
 	[MF_MSG_HUGE]			= "huge page",
 	[MF_MSG_FREE_HUGE]		= "free huge page",
-- 
2.33.0

[PATCH v3 03/13] mm/memory-failure: add macro GET_PAGE_MAX_RETRY_NUM

[Miaohe Lin]

Add helper macro GET_PAGE_MAX_RETRY_NUM to replace magic number 3.
No functional change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 4b9a9298d478..958b17a4b0f5 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1417,6 +1417,8 @@ static int __get_hwpoison_page(struct page *page, unsigned long flags)
 	return 0;
 }
 
+#define GET_PAGE_MAX_RETRY_NUM 3
+
 static int get_any_page(struct page *p, unsigned long flags)
 {
 	int ret = 0, pass = 0;
@@ -1431,12 +1433,12 @@ static int get_any_page(struct page *p, unsigned long flags)
 		if (!ret) {
 			if (page_count(p)) {
 				/* We raced with an allocation, retry. */
-				if (pass++ < 3)
+				if (pass++ < GET_PAGE_MAX_RETRY_NUM)
 					goto try_again;
 				ret = -EBUSY;
 			} else if (!PageHuge(p) && !is_free_buddy_page(p)) {
 				/* We raced with put_page, retry. */
-				if (pass++ < 3)
+				if (pass++ < GET_PAGE_MAX_RETRY_NUM)
 					goto try_again;
 				ret = -EIO;
 			}
@@ -1462,7 +1464,7 @@ static int get_any_page(struct page *p, unsigned long flags)
 		 * A page we cannot handle. Check whether we can turn
 		 * it into something we can handle.
 		 */
-		if (pass++ < 3) {
+		if (pass++ < GET_PAGE_MAX_RETRY_NUM) {
 			put_page(p);
 			shake_page(p);
 			count_increased = false;
-- 
2.33.0

[PATCH v3 04/13] mm/memory-failure: save some page_folio() calls

[Miaohe Lin]

Use local variable folio directly to save a page_folio() call. Also
use folio_mapped() to save more page_folio() calls. No functional
change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
---
 mm/memory-failure.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 958b17a4b0f5..470c570d779c 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1575,7 +1575,7 @@ static bool hwpoison_user_mappings(struct folio *folio, struct page *p,
 	 * This check implies we don't kill processes if their pages
 	 * are in the swap cache early. Those are always late kills.
 	 */
-	if (!page_mapped(p))
+	if (!folio_mapped(folio))
 		return true;
 
 	if (folio_test_swapcache(folio)) {
@@ -1626,10 +1626,10 @@ static bool hwpoison_user_mappings(struct folio *folio, struct page *p,
 		try_to_unmap(folio, ttu);
 	}
 
-	unmap_success = !page_mapped(p);
+	unmap_success = !folio_mapped(folio);
 	if (!unmap_success)
 		pr_err("%#lx: failed to unmap page (folio mapcount=%d)\n",
-		       pfn, folio_mapcount(page_folio(p)));
+		       pfn, folio_mapcount(folio));
 
 	/*
 	 * try_to_unmap() might put mlocked page in lru cache, so call
-- 
2.33.0

[PATCH v3 05/13] mm/memory-failure: remove unneeded empty string

[Miaohe Lin]

Remove unneeded empty string in definition of macro pr_fmt. No functional
change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 470c570d779c..ec482524158e 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -2518,7 +2518,7 @@ static int __init memory_failure_init(void)
 core_initcall(memory_failure_init);
 
 #undef pr_fmt
-#define pr_fmt(fmt)	"" fmt
+#define pr_fmt(fmt)	fmt
 #define unpoison_pr_info(fmt, pfn, rs)			\
 ({							\
 	if (__ratelimit(rs))				\
-- 
2.33.0

[PATCH v3 06/13] mm/memory-failure: remove confusing initialization to count

[Miaohe Lin]

It's meaningless and confusing to init local variable count to 1.
Remove it. No functional change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index ec482524158e..339752d768d8 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -2542,7 +2542,7 @@ int unpoison_memory(unsigned long pfn)
 	struct folio *folio;
 	struct page *p;
 	int ret = -EBUSY, ghp;
-	unsigned long count = 1;
+	unsigned long count;
 	bool huge = false;
 	static DEFINE_RATELIMIT_STATE(unpoison_rs, DEFAULT_RATELIMIT_INTERVAL,
 					DEFAULT_RATELIMIT_BURST);
-- 
2.33.0

[PATCH v3 07/13] mm/memory-failure: don't export hwpoison_filter() when !CONFIG_HWPOISON_INJECT

[Miaohe Lin]

When CONFIG_HWPOISON_INJECT is not enabled, there is no user of the
hwpoison_filter() outside memory-failure. So there is no need to export
it in that case.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202406070136.hGQwVbsv-lkp@intel.com/
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 339752d768d8..ee6ec63534b7 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -294,6 +294,7 @@ int hwpoison_filter(struct page *p)
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(hwpoison_filter);
 #else
 int hwpoison_filter(struct page *p)
 {
@@ -301,8 +302,6 @@ int hwpoison_filter(struct page *p)
 }
 #endif
 
-EXPORT_SYMBOL_GPL(hwpoison_filter);
-
 /*
  * Kill all processes that have a poisoned page mapped and then isolate
  * the page.
-- 
2.33.0

[PATCH v3 08/13] mm/memory-failure: use helper macro task_pid_nr()

[Miaohe Lin]

Use helper macro task_pid_nr() to get the pid of a task. No functional
change intended.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index ee6ec63534b7..bd3ab367ba2f 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -343,7 +343,7 @@ static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags)
 	int ret = 0;
 
 	pr_err("%#lx: Sending SIGBUS to %s:%d due to hardware memory corruption\n",
-			pfn, t->comm, t->pid);
+			pfn, t->comm, task_pid_nr(t));
 
 	if ((flags & MF_ACTION_REQUIRED) && (t == current))
 		ret = force_sig_mceerr(BUS_MCEERR_AR,
@@ -361,7 +361,7 @@ static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags)
 				      addr_lsb, t);
 	if (ret < 0)
 		pr_info("Error sending signal to %s:%d: %d\n",
-			t->comm, t->pid, ret);
+			t->comm, task_pid_nr(t), ret);
 	return ret;
 }
 
@@ -523,7 +523,7 @@ static void kill_procs(struct list_head *to_kill, int forcekill,
 		if (forcekill) {
 			if (tk->addr == -EFAULT) {
 				pr_err("%#lx: forcibly killing %s:%d because of failure to unmap corrupted page\n",
-				       pfn, tk->tsk->comm, tk->tsk->pid);
+				       pfn, tk->tsk->comm, task_pid_nr(tk->tsk));
 				do_send_sig_info(SIGKILL, SEND_SIG_PRIV,
 						 tk->tsk, PIDTYPE_PID);
 			}
@@ -536,7 +536,7 @@ static void kill_procs(struct list_head *to_kill, int forcekill,
 			 */
 			else if (kill_proc(tk, pfn, flags) < 0)
 				pr_err("%#lx: Cannot send advisory machine check signal to %s:%d\n",
-				       pfn, tk->tsk->comm, tk->tsk->pid);
+				       pfn, tk->tsk->comm, task_pid_nr(tk->tsk));
 		}
 		list_del(&tk->nd);
 		put_task_struct(tk->tsk);
-- 
2.33.0

[PATCH v3 09/13] mm/memory-failure: remove obsolete comment in unpoison_memory()

[Miaohe Lin]

Since commit 130d4df57390 ("mm/sl[au]b: rearrange struct slab fields to
allow larger rcu_head"), folio->_mapcount is not overloaded with SLAB.
Update corresponding comment.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
---
 mm/memory-failure.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index bd3ab367ba2f..bb8e79381c1f 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -2584,10 +2584,6 @@ int unpoison_memory(unsigned long pfn)
 	    folio_test_reserved(folio) || folio_test_offline(folio))
 		goto unlock_mutex;
 
-	/*
-	 * Note that folio->_mapcount is overloaded in SLAB, so the simple test
-	 * in folio_mapped() has to be done after folio_test_slab() is checked.
-	 */
 	if (folio_mapped(folio)) {
 		unpoison_pr_info("Unpoison: Someone maps the hwpoison page %#lx\n",
 				 pfn, &unpoison_rs);
-- 
2.33.0

[PATCH v3 10/13] mm/memory-failure: move some function declarations into internal.h

[Miaohe Lin]

There are some functions only used inside mm. Move them into internal.h.
No functional change intended.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202405251049.hxjwX7zO-lkp@intel.com/
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 include/linux/mm.h         |  7 -------
 include/linux/page-flags.h |  5 -----
 include/linux/rmap.h       |  2 --
 mm/internal.h              | 10 ++++++++++
 4 files changed, 10 insertions(+), 14 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 0bf0f22b32c8..587d34879865 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -3981,7 +3981,6 @@ extern int __get_huge_page_for_hwpoison(unsigned long pfn, int flags,
 					bool *migratable_cleared);
 void num_poisoned_pages_inc(unsigned long pfn);
 void num_poisoned_pages_sub(unsigned long pfn, long i);
-struct task_struct *task_early_kill(struct task_struct *tsk, int force_early);
 #else
 static inline void memory_failure_queue(unsigned long pfn, int flags)
 {
@@ -4002,12 +4001,6 @@ static inline void num_poisoned_pages_sub(unsigned long pfn, long i)
 }
 #endif
 
-#if defined(CONFIG_MEMORY_FAILURE) && defined(CONFIG_KSM)
-void add_to_kill_ksm(struct task_struct *tsk, struct page *p,
-		     struct vm_area_struct *vma, struct list_head *to_kill,
-		     unsigned long ksm_addr);
-#endif
-
 #if defined(CONFIG_MEMORY_FAILURE) && defined(CONFIG_MEMORY_HOTPLUG)
 extern void memblk_nr_poison_inc(unsigned long pfn);
 extern void memblk_nr_poison_sub(unsigned long pfn, long i);
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 0c9fd5250687..9a98b657ba07 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -611,11 +611,6 @@ PAGEFLAG_FALSE(Uncached, uncached)
 PAGEFLAG(HWPoison, hwpoison, PF_ANY)
 TESTSCFLAG(HWPoison, hwpoison, PF_ANY)
 #define __PG_HWPOISON (1UL << PG_hwpoison)
-#define MAGIC_HWPOISON	0x48575053U	/* HWPS */
-extern void SetPageHWPoisonTakenOff(struct page *page);
-extern void ClearPageHWPoisonTakenOff(struct page *page);
-extern bool take_page_off_buddy(struct page *page);
-extern bool put_page_back_buddy(struct page *page);
 #else
 PAGEFLAG_FALSE(HWPoison, hwpoison)
 #define __PG_HWPOISON 0
diff --git a/include/linux/rmap.h b/include/linux/rmap.h
index cae38a2a643d..b1bbe237ea4c 100644
--- a/include/linux/rmap.h
+++ b/include/linux/rmap.h
@@ -745,8 +745,6 @@ int pfn_mkclean_range(unsigned long pfn, unsigned long nr_pages, pgoff_t pgoff,
 
 void remove_migration_ptes(struct folio *src, struct folio *dst, bool locked);
 
-unsigned long page_mapped_in_vma(struct page *page, struct vm_area_struct *vma);
-
 /*
  * rmap_walk_control: To control rmap traversing for specific needs
  *
diff --git a/mm/internal.h b/mm/internal.h
index 14bab8a41baf..c9dc056a47a3 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -1078,6 +1078,16 @@ extern u64 hwpoison_filter_flags_mask;
 extern u64 hwpoison_filter_flags_value;
 extern u64 hwpoison_filter_memcg;
 extern u32 hwpoison_filter_enable;
+#define MAGIC_HWPOISON	0x48575053U	/* HWPS */
+void SetPageHWPoisonTakenOff(struct page *page);
+void ClearPageHWPoisonTakenOff(struct page *page);
+bool take_page_off_buddy(struct page *page);
+bool put_page_back_buddy(struct page *page);
+struct task_struct *task_early_kill(struct task_struct *tsk, int force_early);
+void add_to_kill_ksm(struct task_struct *tsk, struct page *p,
+		     struct vm_area_struct *vma, struct list_head *to_kill,
+		     unsigned long ksm_addr);
+unsigned long page_mapped_in_vma(struct page *page, struct vm_area_struct *vma);
 
 extern unsigned long  __must_check vm_mmap_pgoff(struct file *, unsigned long,
         unsigned long, unsigned long,
-- 
2.33.0

[PATCH v3 11/13] mm/memory-failure: fix comment of get_hwpoison_page()

[Miaohe Lin]

When return value is 0, it could also means the page is free hugetlb
page or free buddy page. Fix the corresponding comment.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index bb8e79381c1f..475524f79550 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1525,7 +1525,7 @@ static int __get_unpoison_page(struct page *page)
  * the given page has PG_hwpoison. So it's never reused for other page
  * allocations, and __get_unpoison_page() never races with them.
  *
- * Return: 0 on failure,
+ * Return: 0 on failure or free buddy (hugetlb) page,
  *         1 on success for in-use pages in a well-defined state,
  *         -EIO for pages on which we can not handle memory errors,
  *         -EBUSY when get_hwpoison_page() has raced with page lifecycle
-- 
2.33.0

[PATCH v3 12/13] mm/memory-failure: remove obsolete comment in kill_proc()

[Miaohe Lin]

When user sets SIGBUS to SIG_IGN, it won't cause loop now. For action
required mce error, SIGBUS cannot be blocked. Also when a hwpoisoned
page is re-accessed, kill_accessing_process() will be called to kill
the process.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 475524f79550..12b516895ee7 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -354,8 +354,6 @@ static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags)
 		 * PF_MCE_EARLY set.
 		 * Don't use force here, it's convenient if the signal
 		 * can be temporarily blocked.
-		 * This could cause a loop when the user sets SIGBUS
-		 * to SIG_IGN, but hopefully no one will do that?
 		 */
 		ret = send_sig_mceerr(BUS_MCEERR_AO, (void __user *)tk->addr,
 				      addr_lsb, t);
-- 
2.33.0

[PATCH v3 13/13] mm/memory-failure: correct comment in me_swapcache_dirty

[Miaohe Lin]

Dirty swap cache page could live both in page table (not page cache) and
swap cache when freshly swapped in. Correct comment.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/memory-failure.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 12b516895ee7..2cf7acc286de 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1126,7 +1126,7 @@ static int me_pagecache_dirty(struct page_state *ps, struct page *p)
  * Clean and dirty swap cache.
  *
  * Dirty swap cache page is tricky to handle. The page could live both in page
- * cache and swap cache(ie. page is freshly swapped in). So it could be
+ * table and swap cache(ie. page is freshly swapped in). So it could be
  * referenced concurrently by 2 types of PTEs:
  * normal PTEs and swap PTEs. We try to handle them consistently by calling
  * try_to_unmap(!TTU_HWPOISON) to convert the normal PTEs to swap PTEs,
-- 
2.33.0