[PATCH 0/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[PATCH 0/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[Andrea Arcangeli]

Hello,

The juxtaposition with the other bugchecks didn't make it apparent
that the this WARN_ON was one line too soon.

An app trying to unregister a not-yet-registered range will trigger an
_harmless_ false positive WARN_ON. No real app would do that so it
went unnoticed during testing.

This should be applied on top of 29ec90660d68 ("userfaultfd:
shmem/hugetlbfs: only allow to register VM_MAYWRITE vmas") to shut off
the false positive warning.

Thanks,
Andrea

Andrea Arcangeli (1):
  userfaultfd: check VM_MAYWRITE was set after verifying the uffd is
    registered

 fs/userfaultfd.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

[PATCH 1/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[Andrea Arcangeli]

Calling UFFDIO_UNREGISTER on virtual ranges not yet registered in uffd
could trigger an harmless false positive WARN_ON. Check the vma is
already registered before checking VM_MAYWRITE to shut off the
false positive warning.

Cc: <stable@vger.kernel.org>
Fixes: 29ec90660d68 ("userfaultfd: shmem/hugetlbfs: only allow to register VM_MAYWRITE vmas")
Reported-by: syzbot+06c7092e7d71218a2c16@syzkaller.appspotmail.com
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
---
 fs/userfaultfd.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
index cd58939dc977..7a85e609fc27 100644
--- a/fs/userfaultfd.c
+++ b/fs/userfaultfd.c
@@ -1566,7 +1566,6 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
 		cond_resched();
 
 		BUG_ON(!vma_can_userfault(vma));
-		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
 
 		/*
 		 * Nothing to do: this vma is already registered into this
@@ -1575,6 +1574,8 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
 		if (!vma->vm_userfaultfd_ctx.ctx)
 			goto skip;
 
+		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
+
 		if (vma->vm_start > start)
 			start = vma->vm_start;
 		vma_end = min(end, vma->vm_end);

Re: [PATCH 1/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[Mike Rapoport]

On Thu, Dec 06, 2018 at 04:20:28PM -0500, Andrea Arcangeli wrote:
> Calling UFFDIO_UNREGISTER on virtual ranges not yet registered in uffd
> could trigger an harmless false positive WARN_ON. Check the vma is
> already registered before checking VM_MAYWRITE to shut off the
> false positive warning.
> 
> Cc: <stable@vger.kernel.org>
> Fixes: 29ec90660d68 ("userfaultfd: shmem/hugetlbfs: only allow to register VM_MAYWRITE vmas")
> Reported-by: syzbot+06c7092e7d71218a2c16@syzkaller.appspotmail.com
> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>

Acked-by: Mike Rapoport <rppt@linux.ibm.com>

> ---
>  fs/userfaultfd.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
> index cd58939dc977..7a85e609fc27 100644
> --- a/fs/userfaultfd.c
> +++ b/fs/userfaultfd.c
> @@ -1566,7 +1566,6 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		cond_resched();
> 
>  		BUG_ON(!vma_can_userfault(vma));
> -		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
> 
>  		/*
>  		 * Nothing to do: this vma is already registered into this
> @@ -1575,6 +1574,8 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		if (!vma->vm_userfaultfd_ctx.ctx)
>  			goto skip;
> 
> +		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
> +
>  		if (vma->vm_start > start)
>  			start = vma->vm_start;
>  		vma_end = min(end, vma->vm_end);
> 

-- 
Sincerely yours,
Mike.

Re: [PATCH 1/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[Hugh Dickins]

On Thu, 6 Dec 2018, Andrea Arcangeli wrote:

> Calling UFFDIO_UNREGISTER on virtual ranges not yet registered in uffd
> could trigger an harmless false positive WARN_ON. Check the vma is
> already registered before checking VM_MAYWRITE to shut off the
> false positive warning.
> 
> Cc: <stable@vger.kernel.org>
> Fixes: 29ec90660d68 ("userfaultfd: shmem/hugetlbfs: only allow to register VM_MAYWRITE vmas")
> Reported-by: syzbot+06c7092e7d71218a2c16@syzkaller.appspotmail.com
> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>

Acked-by: Hugh Dickins <hughd@google.com>

> ---
>  fs/userfaultfd.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
> index cd58939dc977..7a85e609fc27 100644
> --- a/fs/userfaultfd.c
> +++ b/fs/userfaultfd.c
> @@ -1566,7 +1566,6 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		cond_resched();
>  
>  		BUG_ON(!vma_can_userfault(vma));
> -		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
>  
>  		/*
>  		 * Nothing to do: this vma is already registered into this
> @@ -1575,6 +1574,8 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		if (!vma->vm_userfaultfd_ctx.ctx)
>  			goto skip;
>  
> +		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
> +
>  		if (vma->vm_start > start)
>  			start = vma->vm_start;
>  		vma_end = min(end, vma->vm_end);
> 

Re: [PATCH 1/1] userfaultfd: check VM_MAYWRITE was set after verifying the uffd is registered

[Peter Xu]

On Thu, Dec 06, 2018 at 04:20:28PM -0500, Andrea Arcangeli wrote:
> Calling UFFDIO_UNREGISTER on virtual ranges not yet registered in uffd
> could trigger an harmless false positive WARN_ON. Check the vma is
> already registered before checking VM_MAYWRITE to shut off the
> false positive warning.
> 
> Cc: <stable@vger.kernel.org>
> Fixes: 29ec90660d68 ("userfaultfd: shmem/hugetlbfs: only allow to register VM_MAYWRITE vmas")
> Reported-by: syzbot+06c7092e7d71218a2c16@syzkaller.appspotmail.com
> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
> ---
>  fs/userfaultfd.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
> index cd58939dc977..7a85e609fc27 100644
> --- a/fs/userfaultfd.c
> +++ b/fs/userfaultfd.c
> @@ -1566,7 +1566,6 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		cond_resched();
>  
>  		BUG_ON(!vma_can_userfault(vma));
> -		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
>  
>  		/*
>  		 * Nothing to do: this vma is already registered into this

Maybe we want to fix this comment too some day:

		/*
		 * Nothing to do: this vma is already registered into this
		 * userfaultfd and with the right tracking mode too.
		 */

But I don't think it's anything urgent since it's clear it means the
other way round and it can potentially be touched up in any further
cleanup/fixes of uffd.

Acked-by: Peter Xu <peterx@redhat.com>

> @@ -1575,6 +1574,8 @@ static int userfaultfd_unregister(struct userfaultfd_ctx *ctx,
>  		if (!vma->vm_userfaultfd_ctx.ctx)
>  			goto skip;
>  
> +		WARN_ON(!(vma->vm_flags & VM_MAYWRITE));
> +
>  		if (vma->vm_start > start)
>  			start = vma->vm_start;
>  		vma_end = min(end, vma->vm_end);

Thanks,

-- 
Peter Xu