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
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <2157a865-5481-43fd-8139-4b9662239eb4@lucifer.local>
Date: Thu, 11 Sep 2025 11:06:29 +0100
From: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
To: "Liam R. Howlett" <Liam.Howlett@...cle.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>, maple-tree@...ts.infradead.org,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        David Hildenbrand <david@...hat.com>, Vlastimil Babka <vbabka@...e.cz>,
        Suren Baghdasaryan <surenb@...gle.com>, Michal Hocko <mhocko@...e.com>,
        Jann Horn <jannh@...gle.com>, Pedro Falcato <pfalcato@...e.de>,
        Charan Teja Kalla <quic_charante@...cinc.com>,
        shikemeng@...weicloud.com, kasong@...cent.com, nphamcs@...il.com,
        bhe@...hat.com, baohua@...nel.org, chrisl@...nel.org,
        Matthew Wilcox <willy@...radead.org>
Subject: Re: [PATCH v1 9/9] mm: Use unmap_desc struct for freeing page tables.

On Tue, Sep 09, 2025 at 03:09:45PM -0400, Liam R. Howlett wrote:
> Pass through the unmap_desc to free_pgtables() because it almost has
> everything necessary and is already on the stack.
>
> Updates testing code as necessary.
>
> No functional changes intended.
>
> Signed-off-by: Liam R. Howlett <Liam.Howlett@...cle.com>

To be clear, I _love_ the use of this helper struct. All commentary is more
so details. what you're doing here is great... :)

Other than nits below this looks fine so on basis of addressing them:

Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>

> ---
>  mm/internal.h                    |  5 +----
>  mm/memory.c                      | 21 ++++++++++-----------
>  mm/mmap.c                        |  6 +++---
>  mm/vma.c                         |  7 ++-----
>  tools/testing/vma/vma_internal.h | 11 ++---------
>  5 files changed, 18 insertions(+), 32 deletions(-)
>
> diff --git a/mm/internal.h b/mm/internal.h
> index 1239944f2830a..f22329967e908 100644
> --- a/mm/internal.h
> +++ b/mm/internal.h
> @@ -445,10 +445,7 @@ bool __folio_end_writeback(struct folio *folio);
>  void deactivate_file_folio(struct folio *folio);
>  void folio_activate(struct folio *folio);
>
> -void free_pgtables(struct mmu_gather *tlb, struct ma_state *mas,
> -		   struct vm_area_struct *start_vma, unsigned long floor,
> -		   unsigned long ceiling, unsigned long tree_max,
> -		   bool mm_wr_locked);
> +void free_pgtables(struct mmu_gather *tlb, struct unmap_desc *desc);
>
>  void pmd_install(struct mm_struct *mm, pmd_t *pmd, pgtable_t *pte);
>
> diff --git a/mm/memory.c b/mm/memory.c
> index 8d4d976311037..98c5ffd28a109 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -387,15 +387,14 @@ void free_pgd_range(struct mmu_gather *tlb,
>   * The tree_max differs from the ceiling when a dup_mmap() failed and the tree
>   * has unrelated data to the mm_struct being torn down.
>   */
> -void free_pgtables(struct mmu_gather *tlb, struct ma_state *mas,
> -		   struct vm_area_struct *vma, unsigned long floor,
> -		   unsigned long ceiling, unsigned long tree_max,
> -		   bool mm_wr_locked)
> +void free_pgtables(struct mmu_gather *tlb, struct unmap_desc *desc)

God this is nice.

>  {
>  	struct unlink_vma_file_batch vb;
> +	struct ma_state *mas = desc->mas;
> +	struct vm_area_struct *vma = desc->first;
>
>  	/* underflow can happen and is fine */
> -	WARN_ON_ONCE(tree_max - 1 > ceiling - 1);
> +	WARN_ON_ONCE(desc->tree_max - 1 > desc->last_pgaddr - 1);
>
>  	tlb_free_vmas(tlb);
>
> @@ -407,13 +406,13 @@ void free_pgtables(struct mmu_gather *tlb, struct ma_state *mas,
>  		 * Note: USER_PGTABLES_CEILING may be passed as ceiling and may
>  		 * be 0.  This will underflow and is okay.
>  		 */
> -		next = mas_find(mas, tree_max - 1);
> +		next = mas_find(mas, desc->tree_max - 1);
>
>  		/*
>  		 * Hide vma from rmap and truncate_pagecache before freeing
>  		 * pgtables
>  		 */
> -		if (mm_wr_locked)
> +		if (desc->mm_wr_locked)
>  			vma_start_write(vma);
>  		unlink_anon_vmas(vma);
>
> @@ -425,16 +424,16 @@ void free_pgtables(struct mmu_gather *tlb, struct ma_state *mas,
>  		 */
>  		while (next && next->vm_start <= vma->vm_end + PMD_SIZE) {
>  			vma = next;
> -			next = mas_find(mas, tree_max - 1);
> -			if (mm_wr_locked)
> +			next = mas_find(mas, desc->tree_max - 1);
> +			if (desc->mm_wr_locked)
>  				vma_start_write(vma);
>  			unlink_anon_vmas(vma);
>  			unlink_file_vma_batch_add(&vb, vma);
>  		}
>  		unlink_file_vma_batch_final(&vb);
>
> -		free_pgd_range(tlb, addr, vma->vm_end,
> -			floor, next ? next->vm_start : ceiling);
> +		free_pgd_range(tlb, addr, vma->vm_end, desc->first_pgaddr,
> +			       next ? next->vm_start : desc->last_pgaddr);

Hm, actually, isn't ceiling exclusive? So 'last_pgaddr' seems a bit
misleading?

>  		vma = next;
>  	} while (vma);
>  }
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 6011f62b0a294..9908481452780 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -1311,10 +1311,10 @@ void exit_mmap(struct mm_struct *mm)
>  	 */
>  	mm_flags_set(MMF_OOM_SKIP, mm);
>  	mmap_write_lock(mm);
> +	unmap.mm_wr_locked = true;
>  	mt_clear_in_rcu(&mm->mm_mt);
> -	vma_iter_set(&vmi, vma->vm_end);
> -	free_pgtables(&tlb, &vmi.mas, vma, FIRST_USER_ADDRESS,
> -		      USER_PGTABLES_CEILING, USER_PGTABLES_CEILING, true);
> +	vma_iter_set(&vmi, unmap.tree_reset);
> +	free_pgtables(&tlb, &unmap);

Lovely!

>  	tlb_finish_mmu(&tlb);
>
>  	/*
> diff --git a/mm/vma.c b/mm/vma.c
> index ad64cd9795ef3..ba155a539d160 100644
> --- a/mm/vma.c
> +++ b/mm/vma.c
> @@ -476,16 +476,13 @@ void remove_vma(struct vm_area_struct *vma)
>  void unmap_region(struct unmap_desc *desc)
>  {
>  	struct mm_struct *mm = desc->first->vm_mm;
> -	struct ma_state *mas = desc->mas;
>  	struct mmu_gather tlb;
>
>  	tlb_gather_mmu(&tlb, mm);
>  	update_hiwater_rss(mm);
>  	unmap_vmas(&tlb, desc);
> -	mas_set(mas, desc->tree_reset);
> -	free_pgtables(&tlb, mas, desc->first, desc->first_pgaddr,
> -		      desc->last_pgaddr, desc->tree_max,
> -		      desc->mm_wr_locked);
> +	mas_set(desc->mas, desc->tree_reset);
> +	free_pgtables(&tlb, desc);
>  	tlb_finish_mmu(&tlb);
>  }
>
> diff --git a/tools/testing/vma/vma_internal.h b/tools/testing/vma/vma_internal.h
> index d73ad4747d40a..435c5a24480bc 100644
> --- a/tools/testing/vma/vma_internal.h
> +++ b/tools/testing/vma/vma_internal.h
> @@ -892,17 +892,10 @@ static inline void unmap_vmas(struct mmu_gather *tlb, struct unmap_desc *unmap)
>  	(void)unmap;
>  }
>
> -static inline void free_pgtables(struct mmu_gather *tlb, struct ma_state *mas,
> -		   struct vm_area_struct *vma, unsigned long floor,
> -		   unsigned long ceiling, unsigned long tree_max,
> -		   bool mm_wr_locked)
> +static inline void free_pgtables(struct mmu_gather *tlb, struct unmap_desc *desc)
>  {
>  	(void)tlb;
> -	(void)mas;
> -	(void)vma;
> -	(void)floor;
> -	(void)ceiling;
> -	(void)mm_wr_locked;
> +	(void)desc;
>  }

Again, let's drop the (void)'s here. Will let Vlasta fix up the conflict I
guess ;)

>
>  static inline void mapping_unmap_writable(struct address_space *)
> --
> 2.47.2
>

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ