diff mbox series

[v7,12/21] mm/vma: Track start and end for munmap in vma_munmap_struct

Message ID 20240822192543.3359552-13-Liam.Howlett@oracle.com (mailing list archive)
State New
Headers show
Series Avoid MAP_FIXED gap exposure | expand

Commit Message

Liam R. Howlett Aug. 22, 2024, 7:25 p.m. UTC 
From: "Liam R. Howlett" <Liam.Howlett@Oracle.com>

Set the start and end address for munmap when the prev and next are
gathered.  This is needed to avoid incorrect addresses being used during
the vms_complete_munmap_vmas() function if the prev/next vma are
expanded.

Add a new helper vms_complete_pte_clear(), which is needed later and
will avoid growing the argument list to unmap_region() beyond the 9 it
already has.

Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
---
 mm/vma.c | 32 +++++++++++++++++++++++++-------
 mm/vma.h |  4 ++++
 2 files changed, 29 insertions(+), 7 deletions(-)

Comments

Geert Uytterhoeven Aug. 26, 2024, 2:01 p.m. UTC | #1 
Hi Liam,

On Thu, Aug 22, 2024 at 9:27 PM Liam R. Howlett <Liam.Howlett@oracle.com> wrote:
> From: "Liam R. Howlett" <Liam.Howlett@Oracle.com>
>
> Set the start and end address for munmap when the prev and next are
> gathered.  This is needed to avoid incorrect addresses being used during
> the vms_complete_munmap_vmas() function if the prev/next vma are
> expanded.
>
> Add a new helper vms_complete_pte_clear(), which is needed later and
> will avoid growing the argument list to unmap_region() beyond the 9 it
> already has.
>
> Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Thanks for your patch, which is now commit ca39aca8db2d78ff ("mm/vma:
track start and end for munmap in vma_munmap_struct") in next-20240826.

> --- a/mm/vma.h
> +++ b/mm/vma.h
> @@ -38,6 +38,8 @@ struct vma_munmap_struct {
>         struct list_head *uf;           /* Userfaultfd list_head */
>         unsigned long start;            /* Aligned start addr (inclusive) */
>         unsigned long end;              /* Aligned end addr (exclusive) */
> +       unsigned long unmap_start;      /* Unmap PTE start */
> +       unsigned long unmap_end;        /* Unmap PTE end */
>         int vma_count;                  /* Number of vmas that will be removed */
>         unsigned long nr_pages;         /* Number of pages being removed */
>         unsigned long locked_vm;        /* Number of locked pages */
> @@ -108,6 +110,8 @@ static inline void init_vma_munmap(struct vma_munmap_struct *vms,
>         vms->vma_count = 0;
>         vms->nr_pages = vms->locked_vm = vms->nr_accounted = 0;
>         vms->exec_vm = vms->stack_vm = vms->data_vm = 0;
> +       vms->unmap_start = FIRST_USER_ADDRESS;
> +       vms->unmap_end = USER_PGTABLES_CEILING;

noreply@ellerman.id.au reported build failures for m5272c3_defconfig
http://kisskb.ellerman.id.au/kisskb/buildresult/15224802/

$ make ARCH=m68k m5272c3_defconfig mm/filemap.o
In file included from mm/internal.h:22,
                 from mm/filemap.c:52:
mm/vma.h: In function ‘init_vma_munmap’:
mm/vma.h:113:21: error: ‘FIRST_USER_ADDRESS’ undeclared (first use in
this function)
  113 |  vms->unmap_start = FIRST_USER_ADDRESS;
      |                     ^~~~~~~~~~~~~~~~~~
mm/vma.h:113:21: note: each undeclared identifier is reported only
once for each function it appears in
mm/vma.h:114:19: error: ‘USER_PGTABLES_CEILING’ undeclared (first use
in this function)
  114 |  vms->unmap_end = USER_PGTABLES_CEILING;
      |                   ^~~~~~~~~~~~~~~~~~~~~

Both are defined in include/linux/pgtable.h inside #ifdef CONFIG_MMU,
so they are not available on nommu.


>  }
>
>  int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,

Gr{oetje,eeting}s,

                        Geert


--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
Lorenzo Stoakes Aug. 26, 2024, 2:12 p.m. UTC | #2 
On Mon, Aug 26, 2024 at 04:01:10PM GMT, Geert Uytterhoeven wrote:
> Hi Liam,
>
> On Thu, Aug 22, 2024 at 9:27 PM Liam R. Howlett <Liam.Howlett@oracle.com> wrote:
> > From: "Liam R. Howlett" <Liam.Howlett@Oracle.com>
> >
> > Set the start and end address for munmap when the prev and next are
> > gathered.  This is needed to avoid incorrect addresses being used during
> > the vms_complete_munmap_vmas() function if the prev/next vma are
> > expanded.
> >
> > Add a new helper vms_complete_pte_clear(), which is needed later and
> > will avoid growing the argument list to unmap_region() beyond the 9 it
> > already has.
> >
> > Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
> > Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
>
> Thanks for your patch, which is now commit ca39aca8db2d78ff ("mm/vma:
> track start and end for munmap in vma_munmap_struct") in next-20240826.
>
> > --- a/mm/vma.h
> > +++ b/mm/vma.h
> > @@ -38,6 +38,8 @@ struct vma_munmap_struct {
> >         struct list_head *uf;           /* Userfaultfd list_head */
> >         unsigned long start;            /* Aligned start addr (inclusive) */
> >         unsigned long end;              /* Aligned end addr (exclusive) */
> > +       unsigned long unmap_start;      /* Unmap PTE start */
> > +       unsigned long unmap_end;        /* Unmap PTE end */
> >         int vma_count;                  /* Number of vmas that will be removed */
> >         unsigned long nr_pages;         /* Number of pages being removed */
> >         unsigned long locked_vm;        /* Number of locked pages */
> > @@ -108,6 +110,8 @@ static inline void init_vma_munmap(struct vma_munmap_struct *vms,
> >         vms->vma_count = 0;
> >         vms->nr_pages = vms->locked_vm = vms->nr_accounted = 0;
> >         vms->exec_vm = vms->stack_vm = vms->data_vm = 0;
> > +       vms->unmap_start = FIRST_USER_ADDRESS;
> > +       vms->unmap_end = USER_PGTABLES_CEILING;
>
> noreply@ellerman.id.au reported build failures for m5272c3_defconfig
> http://kisskb.ellerman.id.au/kisskb/buildresult/15224802/
>
> $ make ARCH=m68k m5272c3_defconfig mm/filemap.o
> In file included from mm/internal.h:22,
>                  from mm/filemap.c:52:
> mm/vma.h: In function ‘init_vma_munmap’:
> mm/vma.h:113:21: error: ‘FIRST_USER_ADDRESS’ undeclared (first use in
> this function)
>   113 |  vms->unmap_start = FIRST_USER_ADDRESS;
>       |                     ^~~~~~~~~~~~~~~~~~
> mm/vma.h:113:21: note: each undeclared identifier is reported only
> once for each function it appears in
> mm/vma.h:114:19: error: ‘USER_PGTABLES_CEILING’ undeclared (first use
> in this function)
>   114 |  vms->unmap_end = USER_PGTABLES_CEILING;
>       |                   ^~~~~~~~~~~~~~~~~~~~~
>
> Both are defined in include/linux/pgtable.h inside #ifdef CONFIG_MMU,
> so they are not available on nommu.

Thanks for the report, this was already resolved (or should be :) via fix-patch
in [0].

[0}:https://lore.kernel.org/all/7d0ea994-f750-49c5-b392-ae7117369cf3@lucifer.local/


>
>
> >  }
> >
> >  int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
diff mbox series

Patch

diff --git a/mm/vma.c b/mm/vma.c
index 9de41e1bf3b2..dda0dae069e2 100644
--- a/mm/vma.c
+++ b/mm/vma.c
@@ -646,6 +646,26 @@  void vma_complete(struct vma_prepare *vp,
 		uprobe_mmap(vp->insert);
 }
 
+static void vms_complete_pte_clear(struct vma_munmap_struct *vms,
+		struct ma_state *mas_detach, bool mm_wr_locked)
+{
+	struct mmu_gather tlb;
+
+	/*
+	 * We can free page tables without write-locking mmap_lock because VMAs
+	 * were isolated before we downgraded mmap_lock.
+	 */
+	mas_set(mas_detach, 1);
+	lru_add_drain();
+	tlb_gather_mmu(&tlb, vms->mm);
+	update_hiwater_rss(vms->mm);
+	unmap_vmas(&tlb, mas_detach, vms->vma, vms->start, vms->end, vms->vma_count, mm_wr_locked);
+	mas_set(mas_detach, 1);
+	/* start and end may be different if there is no prev or next vma. */
+	free_pgtables(&tlb, mas_detach, vms->vma, vms->unmap_start, vms->unmap_end, mm_wr_locked);
+	tlb_finish_mmu(&tlb);
+}
+
 /*
  * vms_complete_munmap_vmas() - Finish the munmap() operation
  * @vms: The vma munmap struct
@@ -667,13 +687,7 @@  void vms_complete_munmap_vmas(struct vma_munmap_struct *vms,
 	if (vms->unlock)
 		mmap_write_downgrade(mm);
 
-	/*
-	 * We can free page tables without write-locking mmap_lock because VMAs
-	 * were isolated before we downgraded mmap_lock.
-	 */
-	mas_set(mas_detach, 1);
-	unmap_region(mm, mas_detach, vms->vma, vms->prev, vms->next,
-		     vms->start, vms->end, vms->vma_count, !vms->unlock);
+	vms_complete_pte_clear(vms, mas_detach, !vms->unlock);
 	/* Update high watermark before we lower total_vm */
 	update_hiwater_vm(mm);
 	/* Stat accounting */
@@ -746,6 +760,8 @@  int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,
 			goto start_split_failed;
 	}
 	vms->prev = vma_prev(vms->vmi);
+	if (vms->prev)
+		vms->unmap_start = vms->prev->vm_end;
 
 	/*
 	 * Detach a range of VMAs from the mm. Using next as a temp variable as
@@ -810,6 +826,8 @@  int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,
 	}
 
 	vms->next = vma_next(vms->vmi);
+	if (vms->next)
+		vms->unmap_end = vms->next->vm_start;
 
 #if defined(CONFIG_DEBUG_VM_MAPLE_TREE)
 	/* Make sure no VMAs are about to be lost. */
diff --git a/mm/vma.h b/mm/vma.h
index c85fc7c888a8..7bc0f9e7751b 100644
--- a/mm/vma.h
+++ b/mm/vma.h
@@ -38,6 +38,8 @@  struct vma_munmap_struct {
 	struct list_head *uf;           /* Userfaultfd list_head */
 	unsigned long start;            /* Aligned start addr (inclusive) */
 	unsigned long end;              /* Aligned end addr (exclusive) */
+	unsigned long unmap_start;      /* Unmap PTE start */
+	unsigned long unmap_end;        /* Unmap PTE end */
 	int vma_count;                  /* Number of vmas that will be removed */
 	unsigned long nr_pages;         /* Number of pages being removed */
 	unsigned long locked_vm;        /* Number of locked pages */
@@ -108,6 +110,8 @@  static inline void init_vma_munmap(struct vma_munmap_struct *vms,
 	vms->vma_count = 0;
 	vms->nr_pages = vms->locked_vm = vms->nr_accounted = 0;
 	vms->exec_vm = vms->stack_vm = vms->data_vm = 0;
+	vms->unmap_start = FIRST_USER_ADDRESS;
+	vms->unmap_end = USER_PGTABLES_CEILING;
 }
 
 int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,