[PATCH 1/1] swapfile: swap_info_get: Check for swap_info[type] == NULL

[PATCH 1/1] swapfile: swap_info_get: Check for swap_info[type] == NULL

[Kautuk Consul]

From: Kautuk Consul <consul.kautuk@gmail.com>

If the swapfile type encoded within entry.val is corrupted in
such a way that the swap_info[type] == NULL, then the code in
swap_info_get will cause a NULL pointer exception.

Assuming that the code in swap_info_get attempts to validate the
swapfile type by checking its range, another bad_nofile check would
be to check for check whether the swap_info[type] pointer is NULL.

Adding a NULL check for swap_info[type] to be reagrded as a "bad_nofile"
error scenario.

Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
---
 mm/swapfile.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/mm/swapfile.c b/mm/swapfile.c
index b1cd120..7bdbe91 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -483,6 +483,8 @@ static struct swap_info_struct *swap_info_get(swp_entry_t entry)
 	if (type >= nr_swapfiles)
 		goto bad_nofile;
 	p = swap_info[type];
+	if (!p)
+		goto bad_nofile;
 	if (!(p->flags & SWP_USED))
 		goto bad_device;
 	offset = swp_offset(entry);
-- 
1.7.6

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 1/1] swapfile: swap_info_get: Check for swap_info[type] == NULL

[Kautuk Consul]

Hi,

Sorry, please ignore the patch file below.

I made a mistake in understanding the code logic.
I now understand that the first "bad_nofile" validation check
is good enough for checking the validity of the value in the "type" variable.


On Mon, Dec 26, 2011 at 6:56 AM, Kautuk Consul <consul.kautuk@gmail.com> wrote:
> From: Kautuk Consul <consul.kautuk@gmail.com>
>
> If the swapfile type encoded within entry.val is corrupted in
> such a way that the swap_info[type] == NULL, then the code in
> swap_info_get will cause a NULL pointer exception.
>
> Assuming that the code in swap_info_get attempts to validate the
> swapfile type by checking its range, another bad_nofile check would
> be to check for check whether the swap_info[type] pointer is NULL.
>
> Adding a NULL check for swap_info[type] to be reagrded as a "bad_nofile"
> error scenario.
>
> Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
> ---
>  mm/swapfile.c |    2 ++
>  1 files changed, 2 insertions(+), 0 deletions(-)
>
> diff --git a/mm/swapfile.c b/mm/swapfile.c
> index b1cd120..7bdbe91 100644
> --- a/mm/swapfile.c
> +++ b/mm/swapfile.c
> @@ -483,6 +483,8 @@ static struct swap_info_struct *swap_info_get(swp_entry_t entry)
>        if (type >= nr_swapfiles)
>                goto bad_nofile;
>        p = swap_info[type];
> +       if (!p)
> +               goto bad_nofile;
>        if (!(p->flags & SWP_USED))
>                goto bad_device;
>        offset = swp_offset(entry);
> --
> 1.7.6
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>