| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <146b95bd-e0f0-4e6b-a9fa-5a8f11355268@gmail.com>
Date: Fri, 26 Sep 2025 15:49:59 +0100
From: Usama Arif <usamaarif642@...il.com>
To: Yafang Shao <laoar.shao@...il.com>, akpm@...ux-foundation.org,
david@...hat.com, ziy@...dia.com, baolin.wang@...ux.alibaba.com,
lorenzo.stoakes@...cle.com, Liam.Howlett@...cle.com, npache@...hat.com,
ryan.roberts@....com, dev.jain@....com, hannes@...xchg.org,
gutierrez.asier@...wei-partners.com, willy@...radead.org, ast@...nel.org,
daniel@...earbox.net, andrii@...nel.org, ameryhung@...il.com,
rientjes@...gle.com, corbet@....net, 21cnbao@...il.com,
shakeel.butt@...ux.dev, tj@...nel.org, lance.yang@...ux.dev
Cc: bpf@...r.kernel.org, linux-mm@...ck.org, linux-doc@...r.kernel.org,
linux-kernel@...r.kernel.org, Yang Shi <shy828301@...il.com>
Subject: Re: [PATCH v8 mm-new 02/12] mm: thp: remove vm_flags parameter from
khugepaged_enter_vma()
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@...il.com>
> Cc: Yang Shi <shy828301@...il.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;
>
Powered by blists - more mailing lists