[PATCH] fuse: Fix a potential double free in virtio_fs_get_tree

[PATCH] fuse: Fix a potential double free in virtio_fs_get_tree

[Lv Yunlong]

In virtio_fs_get_tree, fm is allocated by kzalloc() and
assigned to fsc->s_fs_info by fsc->s_fs_info=fm statement.
If the kzalloc() failed, it will goto err directly, so that
fsc->s_fs_info must be non-NULL and fm will be freed.

But later fm is freed again when virtio_fs_fill_super() fialed.
I think the statement if (fsc->s_fs_info) {kfree(fm);} is
misplaced.

My patch puts this statement in the correct palce to avoid
double free.

Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
---
 fs/fuse/virtio_fs.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
index 8868ac31a3c0..727cf436828f 100644
--- a/fs/fuse/virtio_fs.c
+++ b/fs/fuse/virtio_fs.c
@@ -1437,10 +1437,7 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
 
 	fsc->s_fs_info = fm;
 	sb = sget_fc(fsc, virtio_fs_test_super, set_anon_super_fc);
-	if (fsc->s_fs_info) {
-		fuse_conn_put(fc);
-		kfree(fm);
-	}
+
 	if (IS_ERR(sb))
 		return PTR_ERR(sb);
 
@@ -1457,6 +1454,11 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
 		sb->s_flags |= SB_ACTIVE;
 	}
 
+	if (fsc->s_fs_info) {
+		fuse_conn_put(fc);
+		kfree(fm);
+	}
+
 	WARN_ON(fsc->root);
 	fsc->root = dget(sb->s_root);
 	return 0;
-- 
2.25.1

Re: [PATCH] fuse: Fix a potential double free in virtio_fs_get_tree

[Connor Kuehl]

On 3/23/21 12:18 AM, Lv Yunlong wrote:
> In virtio_fs_get_tree, fm is allocated by kzalloc() and
> assigned to fsc->s_fs_info by fsc->s_fs_info=fm statement.
> If the kzalloc() failed, it will goto err directly, so that

Right, I follow this so far.

> fsc->s_fs_info must be non-NULL and fm will be freed.

But this I don't follow in the context of the stuff that happens in out_err.

> But later fm is freed again when virtio_fs_fill_super() fialed.
> I think the statement if (fsc->s_fs_info) {kfree(fm);} is
> misplaced.

I'm not sure this can double free, because:

* If fm = kzalloc[..] fails, the function bails early.

* If sget_fc() fails, the function cleans up fm and fc and bails early.

* If sget_fc() succeeds and allocated a new superblock, fc->s_fs_info 
pointer is moved to sb->s_fs_info and fc->s_fs_info is set to NULL, so 
the first free hasn't happened yet.

* If sget_fc() succeeds and somehow returns an existing superblock 
(which I think is tested by checking if fc->s_fs_info is not NULL, since 
otherwise it'd have been moved to the superblock and set to NULL in 
sget_fc), I think sb->s_root would not be NULL, therefore the flow of 
control wouldn't enter the if-block where virtio_fs_fill_super could 
fail which means the code won't reach the double free.

That's just my reading of it though, and I'm wondering if that makes 
sense to others :-)

One last comment inline:

> My patch puts this statement in the correct palce to avoid
> double free.
> 
> Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
> ---
>   fs/fuse/virtio_fs.c | 10 ++++++----
>   1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
> index 8868ac31a3c0..727cf436828f 100644
> --- a/fs/fuse/virtio_fs.c
> +++ b/fs/fuse/virtio_fs.c
> @@ -1437,10 +1437,7 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
>   
>   	fsc->s_fs_info = fm;
>   	sb = sget_fc(fsc, virtio_fs_test_super, set_anon_super_fc);
> -	if (fsc->s_fs_info) {
> -		fuse_conn_put(fc);
> -		kfree(fm);
> -	}
> +
>   	if (IS_ERR(sb))
>   		return PTR_ERR(sb);

By removing the check from here, it now looks like if sget_fc() fails, 
then this early return will leak fm's memory and fc's reference.

Connor

>   
> @@ -1457,6 +1454,11 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
>   		sb->s_flags |= SB_ACTIVE;
>   	}
>   
> +	if (fsc->s_fs_info) {
> +		fuse_conn_put(fc);
> +		kfree(fm);
> +	}
> +
>   	WARN_ON(fsc->root);
>   	fsc->root = dget(sb->s_root);
>   	return 0;
> 

Re: [PATCH] fuse: Fix a potential double free in virtio_fs_get_tree

[Vivek Goyal]

On Mon, Mar 22, 2021 at 10:18:31PM -0700, Lv Yunlong wrote:
> In virtio_fs_get_tree, fm is allocated by kzalloc() and
> assigned to fsc->s_fs_info by fsc->s_fs_info=fm statement.
> If the kzalloc() failed, it will goto err directly, so that
> fsc->s_fs_info must be non-NULL and fm will be freed.

sget_fc() will either consume fsc->s_fs_info in case a new super
block is allocated and set fsc->s_fs_info. In that case we don't
free fc or fm.

Or, sget_fc() will return with fsc->s_fs_info set in case we already
found a super block. In that case we need to free fc and fm.

In case of error from sget_fc(), fc/fm need to be freed first and
then error needs to be returned to caller.

        if (IS_ERR(sb))
                return PTR_ERR(sb);


If we allocated a new super block in sget_fc(), then next step is
to initialize it.

        if (!sb->s_root) {
                err = virtio_fs_fill_super(sb, fsc);
	}

If we run into errors here, then fc/fm need to be freed.

So current code looks fine to me.

Vivek

> 
> But later fm is freed again when virtio_fs_fill_super() fialed.
> I think the statement if (fsc->s_fs_info) {kfree(fm);} is
> misplaced.
> 
> My patch puts this statement in the correct palce to avoid
> double free.
> 
> Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
> ---
>  fs/fuse/virtio_fs.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
> index 8868ac31a3c0..727cf436828f 100644
> --- a/fs/fuse/virtio_fs.c
> +++ b/fs/fuse/virtio_fs.c
> @@ -1437,10 +1437,7 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
>  
>  	fsc->s_fs_info = fm;
>  	sb = sget_fc(fsc, virtio_fs_test_super, set_anon_super_fc);
> -	if (fsc->s_fs_info) {
> -		fuse_conn_put(fc);
> -		kfree(fm);
> -	}
> +
>  	if (IS_ERR(sb))
>  		return PTR_ERR(sb);
>  
> @@ -1457,6 +1454,11 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
>  		sb->s_flags |= SB_ACTIVE;
>  	}
>  
> +	if (fsc->s_fs_info) {
> +		fuse_conn_put(fc);
> +		kfree(fm);
> +	}
> +
>  	WARN_ON(fsc->root);
>  	fsc->root = dget(sb->s_root);
>  	return 0;
> -- 
> 2.25.1
> 
> 

Re: Re: [PATCH] fuse: Fix a potential double free in virtio_fs_get_tree

[lyl2019]

> -----原始邮件-----
> 发件人: "Vivek Goyal" <vgoyal@redhat.com>
> 发送时间: 2021-03-24 01:10:03 (星期三)
> 收件人: "Lv Yunlong" <lyl2019@mail.ustc.edu.cn>
> 抄送: stefanha@redhat.com, miklos@szeredi.hu, virtualization@lists.linux-foundation.org, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
> 主题: Re: [PATCH] fuse: Fix a potential double free in virtio_fs_get_tree
> 
> On Mon, Mar 22, 2021 at 10:18:31PM -0700, Lv Yunlong wrote:
> > In virtio_fs_get_tree, fm is allocated by kzalloc() and
> > assigned to fsc->s_fs_info by fsc->s_fs_info=fm statement.
> > If the kzalloc() failed, it will goto err directly, so that
> > fsc->s_fs_info must be non-NULL and fm will be freed.
> 
> sget_fc() will either consume fsc->s_fs_info in case a new super
> block is allocated and set fsc->s_fs_info. In that case we don't
> free fc or fm.
> 
> Or, sget_fc() will return with fsc->s_fs_info set in case we already
> found a super block. In that case we need to free fc and fm.
> 
> In case of error from sget_fc(), fc/fm need to be freed first and
> then error needs to be returned to caller.
> 
>         if (IS_ERR(sb))
>                 return PTR_ERR(sb);
> 
> 
> If we allocated a new super block in sget_fc(), then next step is
> to initialize it.
> 
>         if (!sb->s_root) {
>                 err = virtio_fs_fill_super(sb, fsc);
> 	}
> 
> If we run into errors here, then fc/fm need to be freed.
> 
> So current code looks fine to me.
> 
> Vivek
> 
> > 
> > But later fm is freed again when virtio_fs_fill_super() fialed.
> > I think the statement if (fsc->s_fs_info) {kfree(fm);} is
> > misplaced.
> > 
> > My patch puts this statement in the correct palce to avoid
> > double free.
> > 
> > Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
> > ---
> >  fs/fuse/virtio_fs.c | 10 ++++++----
> >  1 file changed, 6 insertions(+), 4 deletions(-)
> > 
> > diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
> > index 8868ac31a3c0..727cf436828f 100644
> > --- a/fs/fuse/virtio_fs.c
> > +++ b/fs/fuse/virtio_fs.c
> > @@ -1437,10 +1437,7 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
> >  
> >  	fsc->s_fs_info = fm;
> >  	sb = sget_fc(fsc, virtio_fs_test_super, set_anon_super_fc);
> > -	if (fsc->s_fs_info) {
> > -		fuse_conn_put(fc);
> > -		kfree(fm);
> > -	}
> > +
> >  	if (IS_ERR(sb))
> >  		return PTR_ERR(sb);
> >  
> > @@ -1457,6 +1454,11 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
> >  		sb->s_flags |= SB_ACTIVE;
> >  	}
> >  
> > +	if (fsc->s_fs_info) {
> > +		fuse_conn_put(fc);
> > +		kfree(fm);
> > +	}
> > +
> >  	WARN_ON(fsc->root);
> >  	fsc->root = dget(sb->s_root);
> >  	return 0;
> > -- 
> > 2.25.1
> > 
> > 
> 


Ok, thanks.
It should be a false positive.