The khugepaged_enter_vma() function requires handling in two specific
scenarios:
1. New VMA creation
When a new VMA is created, if vma->vm_mm is not present in
khugepaged_mm_slot, it must be added. In this case,
khugepaged_enter_vma() is called after vma->vm_flags have been set,
allowing direct use of the VMA's flags.
2. VMA flag modification
When vma->vm_flags are modified (particularly when VM_HUGEPAGE is set),
the system must recheck whether to add vma->vm_mm to khugepaged_mm_slot.
Currently, khugepaged_enter_vma() is called before the flag update, so
the call must be relocated to occur after vma->vm_flags have been set.
Additionally, khugepaged_enter_vma() is invoked in other contexts, such as
during VMA merging. However, these calls are unnecessary because the
existing VMA already ensures that vma->vm_mm is registered in
khugepaged_mm_slot. While removing these redundant calls represents a
potential optimization, that change should be addressed separately.
Because VMA merging only occurs when the vm_flags of both VMAs are
identical (excluding special flags like VM_SOFTDIRTY), we can safely use
target->vm_flags instead.
After this change, we can further remove vm_flags parameter from
thp_vma_allowable_order(). That will be handled in a followup patch.
Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Cc: Yang Shi <shy828301@gmail.com>
---
include/linux/khugepaged.h | 6 ++----
mm/huge_memory.c | 2 +-
mm/khugepaged.c | 11 ++---------
mm/madvise.c | 7 +++++++
mm/vma.c | 6 +++---
5 files changed, 15 insertions(+), 17 deletions(-)
diff --git a/include/linux/khugepaged.h b/include/linux/khugepaged.h
index f14680cd9854..b30814d3d665 100644
--- a/include/linux/khugepaged.h
+++ b/include/linux/khugepaged.h
@@ -13,8 +13,7 @@ extern void khugepaged_destroy(void);
extern int start_stop_khugepaged(void);
extern void __khugepaged_enter(struct mm_struct *mm);
extern void __khugepaged_exit(struct mm_struct *mm);
-extern void khugepaged_enter_vma(struct vm_area_struct *vma,
- vm_flags_t vm_flags);
+extern void khugepaged_enter_vma(struct vm_area_struct *vma);
extern void khugepaged_enter_mm(struct mm_struct *mm);
extern void khugepaged_min_free_kbytes_update(void);
extern bool current_is_khugepaged(void);
@@ -39,8 +38,7 @@ static inline void khugepaged_fork(struct mm_struct *mm, struct mm_struct *oldmm
static inline void khugepaged_exit(struct mm_struct *mm)
{
}
-static inline void khugepaged_enter_vma(struct vm_area_struct *vma,
- vm_flags_t vm_flags)
+static inline void khugepaged_enter_vma(struct vm_area_struct *vma)
{
}
static inline void khugepaged_enter_mm(struct mm_struct *mm)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 1b81680b4225..ac6601f30e65 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1346,7 +1346,7 @@ vm_fault_t do_huge_pmd_anonymous_page(struct vm_fault *vmf)
ret = vmf_anon_prepare(vmf);
if (ret)
return ret;
- khugepaged_enter_vma(vma, vma->vm_flags);
+ khugepaged_enter_vma(vma);
if (!(vmf->flags & FAULT_FLAG_WRITE) &&
!mm_forbids_zeropage(vma->vm_mm) &&
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index f47ac8c19447..04121ae7d18d 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -353,12 +353,6 @@ int hugepage_madvise(struct vm_area_struct *vma,
#endif
*vm_flags &= ~VM_NOHUGEPAGE;
*vm_flags |= VM_HUGEPAGE;
- /*
- * If the vma become good for khugepaged to scan,
- * register it here without waiting a page fault that
- * may not happen any time soon.
- */
- khugepaged_enter_vma(vma, *vm_flags);
break;
case MADV_NOHUGEPAGE:
*vm_flags &= ~VM_HUGEPAGE;
@@ -467,10 +461,9 @@ void khugepaged_enter_mm(struct mm_struct *mm)
__khugepaged_enter(mm);
}
-void khugepaged_enter_vma(struct vm_area_struct *vma,
- vm_flags_t vm_flags)
+void khugepaged_enter_vma(struct vm_area_struct *vma)
{
- if (!thp_vma_allowable_order(vma, vm_flags, TVA_KHUGEPAGED, PMD_ORDER))
+ if (!thp_vma_allowable_order(vma, vma->vm_flags, TVA_KHUGEPAGED, PMD_ORDER))
return;
khugepaged_enter_mm(vma->vm_mm);
diff --git a/mm/madvise.c b/mm/madvise.c
index 35ed4ab0d7c5..ab8b5d47badb 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -1425,6 +1425,13 @@ static int madvise_vma_behavior(struct madvise_behavior *madv_behavior)
VM_WARN_ON_ONCE(madv_behavior->lock_mode != MADVISE_MMAP_WRITE_LOCK);
error = madvise_update_vma(new_flags, madv_behavior);
+ /*
+ * If the vma become good for khugepaged to scan,
+ * register it here without waiting a page fault that
+ * may not happen any time soon.
+ */
+ if (!error && new_flags & VM_HUGEPAGE)
+ khugepaged_enter_mm(vma->vm_mm);
out:
/*
* madvise() returns EAGAIN if kernel resources, such as
diff --git a/mm/vma.c b/mm/vma.c
index a1ec405bda25..6a548b0d64cd 100644
--- a/mm/vma.c
+++ b/mm/vma.c
@@ -973,7 +973,7 @@ static __must_check struct vm_area_struct *vma_merge_existing_range(
if (err || commit_merge(vmg))
goto abort;
- khugepaged_enter_vma(vmg->target, vmg->vm_flags);
+ khugepaged_enter_vma(vmg->target);
vmg->state = VMA_MERGE_SUCCESS;
return vmg->target;
@@ -1093,7 +1093,7 @@ struct vm_area_struct *vma_merge_new_range(struct vma_merge_struct *vmg)
* following VMA if we have VMAs on both sides.
*/
if (vmg->target && !vma_expand(vmg)) {
- khugepaged_enter_vma(vmg->target, vmg->vm_flags);
+ khugepaged_enter_vma(vmg->target);
vmg->state = VMA_MERGE_SUCCESS;
return vmg->target;
}
@@ -2520,7 +2520,7 @@ static int __mmap_new_vma(struct mmap_state *map, struct vm_area_struct **vmap)
* call covers the non-merge case.
*/
if (!vma_is_anonymous(vma))
- khugepaged_enter_vma(vma, map->vm_flags);
+ khugepaged_enter_vma(vma);
*vmap = vma;
return 0;
--
2.47.3
On 26/09/2025 10:33, Yafang Shao wrote:
> The khugepaged_enter_vma() function requires handling in two specific
> scenarios:
> 1. New VMA creation
> When a new VMA is created, if vma->vm_mm is not present in
> khugepaged_mm_slot, it must be added. In this case,
> khugepaged_enter_vma() is called after vma->vm_flags have been set,
> allowing direct use of the VMA's flags.
> 2. VMA flag modification
> When vma->vm_flags are modified (particularly when VM_HUGEPAGE is set),
> the system must recheck whether to add vma->vm_mm to khugepaged_mm_slot.
> Currently, khugepaged_enter_vma() is called before the flag update, so
> the call must be relocated to occur after vma->vm_flags have been set.
>
> Additionally, khugepaged_enter_vma() is invoked in other contexts, such as
> during VMA merging. However, these calls are unnecessary because the
> existing VMA already ensures that vma->vm_mm is registered in
> khugepaged_mm_slot. While removing these redundant calls represents a
> potential optimization, that change should be addressed separately.
> Because VMA merging only occurs when the vm_flags of both VMAs are
> identical (excluding special flags like VM_SOFTDIRTY), we can safely use
> target->vm_flags instead.
>
The patch looks good to me, but if we are sure that khugepaged_enter_vma
is not needed in VMA merging case, we should remove it in this patch itself.
If the reason we are removing what flags are being considered when calling
khugepaged_enter_vma in VMA merging case is because the calls are unnecessary,
then we should just remove the calls and not modify them
(if its safe and functionally correct :))
> After this change, we can further remove vm_flags parameter from
> thp_vma_allowable_order(). That will be handled in a followup patch.
>
> Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> Cc: Yang Shi <shy828301@gmail.com>
> ---
> include/linux/khugepaged.h | 6 ++----
> mm/huge_memory.c | 2 +-
> mm/khugepaged.c | 11 ++---------
> mm/madvise.c | 7 +++++++
> mm/vma.c | 6 +++---
> 5 files changed, 15 insertions(+), 17 deletions(-)
>
> diff --git a/include/linux/khugepaged.h b/include/linux/khugepaged.h
> index f14680cd9854..b30814d3d665 100644
> --- a/include/linux/khugepaged.h
> +++ b/include/linux/khugepaged.h
> @@ -13,8 +13,7 @@ extern void khugepaged_destroy(void);
> extern int start_stop_khugepaged(void);
> extern void __khugepaged_enter(struct mm_struct *mm);
> extern void __khugepaged_exit(struct mm_struct *mm);
> -extern void khugepaged_enter_vma(struct vm_area_struct *vma,
> - vm_flags_t vm_flags);
> +extern void khugepaged_enter_vma(struct vm_area_struct *vma);
> extern void khugepaged_enter_mm(struct mm_struct *mm);
> extern void khugepaged_min_free_kbytes_update(void);
> extern bool current_is_khugepaged(void);
> @@ -39,8 +38,7 @@ static inline void khugepaged_fork(struct mm_struct *mm, struct mm_struct *oldmm
> static inline void khugepaged_exit(struct mm_struct *mm)
> {
> }
> -static inline void khugepaged_enter_vma(struct vm_area_struct *vma,
> - vm_flags_t vm_flags)
> +static inline void khugepaged_enter_vma(struct vm_area_struct *vma)
> {
> }
> static inline void khugepaged_enter_mm(struct mm_struct *mm)
> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> index 1b81680b4225..ac6601f30e65 100644
> --- a/mm/huge_memory.c
> +++ b/mm/huge_memory.c
> @@ -1346,7 +1346,7 @@ vm_fault_t do_huge_pmd_anonymous_page(struct vm_fault *vmf)
> ret = vmf_anon_prepare(vmf);
> if (ret)
> return ret;
> - khugepaged_enter_vma(vma, vma->vm_flags);
> + khugepaged_enter_vma(vma);
>
> if (!(vmf->flags & FAULT_FLAG_WRITE) &&
> !mm_forbids_zeropage(vma->vm_mm) &&
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index f47ac8c19447..04121ae7d18d 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -353,12 +353,6 @@ int hugepage_madvise(struct vm_area_struct *vma,
> #endif
> *vm_flags &= ~VM_NOHUGEPAGE;
> *vm_flags |= VM_HUGEPAGE;
> - /*
> - * If the vma become good for khugepaged to scan,
> - * register it here without waiting a page fault that
> - * may not happen any time soon.
> - */
> - khugepaged_enter_vma(vma, *vm_flags);
> break;
> case MADV_NOHUGEPAGE:
> *vm_flags &= ~VM_HUGEPAGE;
> @@ -467,10 +461,9 @@ void khugepaged_enter_mm(struct mm_struct *mm)
> __khugepaged_enter(mm);
> }
>
> -void khugepaged_enter_vma(struct vm_area_struct *vma,
> - vm_flags_t vm_flags)
> +void khugepaged_enter_vma(struct vm_area_struct *vma)
> {
> - if (!thp_vma_allowable_order(vma, vm_flags, TVA_KHUGEPAGED, PMD_ORDER))
> + if (!thp_vma_allowable_order(vma, vma->vm_flags, TVA_KHUGEPAGED, PMD_ORDER))
> return;
>
> khugepaged_enter_mm(vma->vm_mm);
> diff --git a/mm/madvise.c b/mm/madvise.c
> index 35ed4ab0d7c5..ab8b5d47badb 100644
> --- a/mm/madvise.c
> +++ b/mm/madvise.c
> @@ -1425,6 +1425,13 @@ static int madvise_vma_behavior(struct madvise_behavior *madv_behavior)
> VM_WARN_ON_ONCE(madv_behavior->lock_mode != MADVISE_MMAP_WRITE_LOCK);
>
> error = madvise_update_vma(new_flags, madv_behavior);
> + /*
> + * If the vma become good for khugepaged to scan,
> + * register it here without waiting a page fault that
> + * may not happen any time soon.
> + */
> + if (!error && new_flags & VM_HUGEPAGE)
> + khugepaged_enter_mm(vma->vm_mm);
> out:
> /*
> * madvise() returns EAGAIN if kernel resources, such as
> diff --git a/mm/vma.c b/mm/vma.c
> index a1ec405bda25..6a548b0d64cd 100644
> --- a/mm/vma.c
> +++ b/mm/vma.c
> @@ -973,7 +973,7 @@ static __must_check struct vm_area_struct *vma_merge_existing_range(
> if (err || commit_merge(vmg))
> goto abort;
>
> - khugepaged_enter_vma(vmg->target, vmg->vm_flags);
> + khugepaged_enter_vma(vmg->target);
> vmg->state = VMA_MERGE_SUCCESS;
> return vmg->target;
>
> @@ -1093,7 +1093,7 @@ struct vm_area_struct *vma_merge_new_range(struct vma_merge_struct *vmg)
> * following VMA if we have VMAs on both sides.
> */
> if (vmg->target && !vma_expand(vmg)) {
> - khugepaged_enter_vma(vmg->target, vmg->vm_flags);
> + khugepaged_enter_vma(vmg->target);
> vmg->state = VMA_MERGE_SUCCESS;
> return vmg->target;
> }
> @@ -2520,7 +2520,7 @@ static int __mmap_new_vma(struct mmap_state *map, struct vm_area_struct **vmap)
> * call covers the non-merge case.
> */
> if (!vma_is_anonymous(vma))
> - khugepaged_enter_vma(vma, map->vm_flags);
> + khugepaged_enter_vma(vma);
> *vmap = vma;
> return 0;
>
On Fri, Sep 26, 2025 at 10:50 PM Usama Arif <usamaarif642@gmail.com> wrote: > > > > On 26/09/2025 10:33, Yafang Shao wrote: > > The khugepaged_enter_vma() function requires handling in two specific > > scenarios: > > 1. New VMA creation > > When a new VMA is created, if vma->vm_mm is not present in > > khugepaged_mm_slot, it must be added. In this case, > > khugepaged_enter_vma() is called after vma->vm_flags have been set, > > allowing direct use of the VMA's flags. > > 2. VMA flag modification > > When vma->vm_flags are modified (particularly when VM_HUGEPAGE is set), > > the system must recheck whether to add vma->vm_mm to khugepaged_mm_slot. > > Currently, khugepaged_enter_vma() is called before the flag update, so > > the call must be relocated to occur after vma->vm_flags have been set. > > > > Additionally, khugepaged_enter_vma() is invoked in other contexts, such as > > during VMA merging. However, these calls are unnecessary because the > > existing VMA already ensures that vma->vm_mm is registered in > > khugepaged_mm_slot. While removing these redundant calls represents a > > potential optimization, that change should be addressed separately. > > Because VMA merging only occurs when the vm_flags of both VMAs are > > identical (excluding special flags like VM_SOFTDIRTY), we can safely use > > target->vm_flags instead. > > > > The patch looks good to me, but if we are sure that khugepaged_enter_vma > is not needed in VMA merging case, Calling khugepaged_enter_vma() is unnecessary during VMA merging because it's already handled: for non-anonymous VMAs, it's called upon creation, and for anonymous VMAs, it's handled at page fault. > we should remove it in this patch itself. I'd prefer to handle this cleanup separately. The goal is to keep the THP changes minimal, even though I've already made significant modifications ;-) > If the reason we are removing what flags are being considered when calling > khugepaged_enter_vma in VMA merging case is because the calls are unnecessary, Actually, the rationale is that the flags can be removed because: Because VMA merging only occurs when the vm_flags of both VMAs are identical (excluding special flags like VM_SOFTDIRTY), we can safely use target->vm_flags instead. I will update the commit log to clarify this point. > then we should just remove the calls and not modify them > (if its safe and functionally correct :)) -- Regards Yafang
© 2016 - 2025 Red Hat, Inc.