[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <b3308387-464a-52ae-114b-34ab94e3b5c6@nvidia.com>
Date: Wed, 7 Dec 2022 15:19:55 -0800
From: John Hubbard <jhubbard@...dia.com>
To: Peter Xu <peterx@...hat.com>, <linux-mm@...ck.org>,
<linux-kernel@...r.kernel.org>
CC: Muchun Song <songmuchun@...edance.com>,
Andrea Arcangeli <aarcange@...hat.com>,
James Houghton <jthoughton@...gle.com>,
Jann Horn <jannh@...gle.com>, Rik van Riel <riel@...riel.com>,
Miaohe Lin <linmiaohe@...wei.com>,
Andrew Morton <akpm@...ux-foundation.org>,
"Mike Kravetz" <mike.kravetz@...cle.com>,
David Hildenbrand <david@...hat.com>,
Nadav Amit <nadav.amit@...il.com>
Subject: Re: [PATCH v2 05/10] mm/hugetlb: Make userfaultfd_huge_must_wait()
safe to pmd unshare
On 12/7/22 12:30, Peter Xu wrote:
> We can take the hugetlb walker lock, here taking vma lock directly.
>
> Reviewed-by: David Hildenbrand <david@...hat.com>
> Reviewed-by: Mike Kravetz <mike.kravetz@...cle.com>
> Signed-off-by: Peter Xu <peterx@...hat.com>
> ---
> fs/userfaultfd.c | 18 ++++++++++++++----
> 1 file changed, 14 insertions(+), 4 deletions(-)
>
> diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
> index 07c81ab3fd4d..a602f008dde5 100644
> --- a/fs/userfaultfd.c
> +++ b/fs/userfaultfd.c
> @@ -376,7 +376,8 @@ static inline unsigned int userfaultfd_get_blocking_state(unsigned int flags)
> */
> vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)
> {
> - struct mm_struct *mm = vmf->vma->vm_mm;
> + struct vm_area_struct *vma = vmf->vma;
> + struct mm_struct *mm = vma->vm_mm;
> struct userfaultfd_ctx *ctx;
> struct userfaultfd_wait_queue uwq;
> vm_fault_t ret = VM_FAULT_SIGBUS;
> @@ -403,7 +404,7 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)
> */
> mmap_assert_locked(mm);
>
> - ctx = vmf->vma->vm_userfaultfd_ctx.ctx;
> + ctx = vma->vm_userfaultfd_ctx.ctx;
> if (!ctx)
> goto out;
>
> @@ -493,6 +494,13 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)
>
> blocking_state = userfaultfd_get_blocking_state(vmf->flags);
>
> + /*
> + * This stablizes pgtable for hugetlb on e.g. pmd unsharing. Need
> + * to be before setting current state.
> + */
Looking at this code, I am not able to come up with a reason for why the
vma lock/unlock placement is exactly where it is. It looks quite arbitrary.
Why not, for example, take and drop the vma lock within
userfaultfd_huge_must_wait()? That makes more sense to me, but I'm not familiar
with userfaultfd so of course I'm missing something.
But the comment above certainly doesn't supply that something.
thanks,
--
John Hubbard
NVIDIA
> + if (is_vm_hugetlb_page(vma))
> + hugetlb_vma_lock_read(vma);
> +
> spin_lock_irq(&ctx->fault_pending_wqh.lock);
> /*
> * After the __add_wait_queue the uwq is visible to userland
> @@ -507,13 +515,15 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason)
> set_current_state(blocking_state);
> spin_unlock_irq(&ctx->fault_pending_wqh.lock);
>
> - if (!is_vm_hugetlb_page(vmf->vma))
> + if (!is_vm_hugetlb_page(vma))
> must_wait = userfaultfd_must_wait(ctx, vmf->address, vmf->flags,
> reason);
> else
> - must_wait = userfaultfd_huge_must_wait(ctx, vmf->vma,
> + must_wait = userfaultfd_huge_must_wait(ctx, vma,
> vmf->address,
> vmf->flags, reason);
> + if (is_vm_hugetlb_page(vma))
> + hugetlb_vma_unlock_read(vma);
> mmap_read_unlock(mm);
>
> if (likely(must_wait && !READ_ONCE(ctx->released))) {
Powered by blists - more mailing lists