2018-07-07 17:53:49

by Bernd Edlinger

Subject: [PATCHv3] Fix range checks in kernfs_get_target_path

The terminating NUL byte is only there because the buffer is
allocated with kzalloc(PAGE_SIZE, GFP_KERNEL), but since the
range-check is off-by-one, and PAGE_SIZE==PATH_MAX, the
returned string may not be zero-terminated if it is exactly
PATH_MAX characters long. Furthermore also the initial loop
may theoretically exceed PATH_MAX and cause a fault.

Signed-off-by: Bernd Edlinger <[email protected]>
---
fs/kernfs/symlink.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/fs/kernfs/symlink.c b/fs/kernfs/symlink.c
index 08ccabd..774bada 100644
--- a/fs/kernfs/symlink.c
+++ b/fs/kernfs/symlink.c
@@ -63,6 +63,9 @@ static int kernfs_get_target_path(struct kernfs_node
if (base == kn)
break;

+ if ((s - path) + 3 >= PATH_MAX)
+ return -ENAMETOOLONG;
+
strcpy(s, "../");
s += 3;
base = base->parent;
@@ -79,7 +82,7 @@ static int kernfs_get_target_path(struct kernfs_node
if (len < 2)
return -EINVAL;
len--;
- if ((s - path) + len > PATH_MAX)
+ if ((s - path) + len >= PATH_MAX)
return -ENAMETOOLONG;

/* reverse fillup of target string from target to base */
--
1.9.1

2018-08-26 10:57:53

by Bernd Edlinger

Subject: Re: [PATCHv3] Fix range checks in kernfs_get_target_path

Ping...
Sorry, I had actually completely forgotten about this one.

On 07/07/18 19:52, Bernd Edlinger wrote:
> The terminating NUL byte is only there because the buffer is
> allocated with kzalloc(PAGE_SIZE, GFP_KERNEL), but since the
> range-check is off-by-one, and PAGE_SIZE==PATH_MAX, the
> returned string may not be zero-terminated if it is exactly
> PATH_MAX characters long.  Furthermore also the initial loop
> may theoretically exceed PATH_MAX and cause a fault.
>
> Signed-off-by: Bernd Edlinger <[email protected]>
> ---
>  fs/kernfs/symlink.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/fs/kernfs/symlink.c b/fs/kernfs/symlink.c
> index 08ccabd..774bada 100644
> --- a/fs/kernfs/symlink.c
> +++ b/fs/kernfs/symlink.c
> @@ -63,6 +63,9 @@ static int kernfs_get_target_path(struct kernfs_node
>          if (base == kn)
>              break;
>
> +        if ((s - path) + 3 >= PATH_MAX)
> +            return -ENAMETOOLONG;
> +
>          strcpy(s, "../");
>          s += 3;
>          base = base->parent;
> @@ -79,7 +82,7 @@ static int kernfs_get_target_path(struct kernfs_node
>      if (len < 2)
>          return -EINVAL;
>      len--;
> -    if ((s - path) + len > PATH_MAX)
> +    if ((s - path) + len >= PATH_MAX)
>          return -ENAMETOOLONG;
>
>      /* reverse fillup of target string from target to base */

2018-08-27 14:05:26

by Tejun Heo

Subject: Re: [PATCHv3] Fix range checks in kernfs_get_target_path

On Sun, Aug 26, 2018 at 10:55:24AM +0000, Bernd Edlinger wrote:
> Ping...
> Sorry, I had actually completely forgotten about this one.
>
> On 07/07/18 19:52, Bernd Edlinger wrote:
> > The terminating NUL byte is only there because the buffer is
> > allocated with kzalloc(PAGE_SIZE, GFP_KERNEL), but since the
> > range-check is off-by-one, and PAGE_SIZE==PATH_MAX, the
> > returned string may not be zero-terminated if it is exactly
> > PATH_MAX characters long.? Furthermore also the initial loop
> > may theoretically exceed PATH_MAX and cause a fault.
> >
> > Signed-off-by: Bernd Edlinger <[email protected]>

Acked-by: Tejun Heo <[email protected]>

Thanks.

--
tejun