[PATCH v18 07/15] fs/namespace: untag user pointers in copy_mount_options
catalin.marinas at arm.com
Mon Jun 24 17:50:09 UTC 2019
On Mon, Jun 24, 2019 at 04:32:52PM +0200, Andrey Konovalov wrote:
> This patch is a part of a series that extends kernel ABI to allow to pass
> tagged user pointers (with the top byte set to something else other than
> 0x00) as syscall arguments.
> In copy_mount_options a user address is being subtracted from TASK_SIZE.
> If the address is lower than TASK_SIZE, the size is calculated to not
> allow the exact_copy_from_user() call to cross TASK_SIZE boundary.
> However if the address is tagged, then the size will be calculated
> Untag the address before subtracting.
> Reviewed-by: Khalid Aziz <khalid.aziz at oracle.com>
> Reviewed-by: Vincenzo Frascino <vincenzo.frascino at arm.com>
> Reviewed-by: Kees Cook <keescook at chromium.org>
> Reviewed-by: Catalin Marinas <catalin.marinas at arm.com>
> Signed-off-by: Andrey Konovalov <andreyknvl at google.com>
> fs/namespace.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
> diff --git a/fs/namespace.c b/fs/namespace.c
> index 7660c2749c96..ec78f7223917 100644
> --- a/fs/namespace.c
> +++ b/fs/namespace.c
> @@ -2994,7 +2994,7 @@ void *copy_mount_options(const void __user * data)
> * the remainder of the page.
> /* copy_from_user cannot cross TASK_SIZE ! */
> - size = TASK_SIZE - (unsigned long)data;
> + size = TASK_SIZE - (unsigned long)untagged_addr(data);
> if (size > PAGE_SIZE)
> size = PAGE_SIZE;
I think this patch needs an ack from Al Viro (cc'ed).
More information about the amd-gfx