Re: [patch] Staging: tidspbridge: make mmap root-only so it is not a security problem

[Dan Carpenter <dan.carpenter@oracle.com>]

On Sun, Dec 01, 2013 at 01:10:06PM +0100, Pavel Machek wrote:
> diff --git a/drivers/staging/tidspbridge/rmgr/drv_interface.c b/drivers/staging/tidspbridge/rmgr/drv_interface.c
> index 1aa4a3f..a8e86cf 100644
> --- a/drivers/staging/tidspbridge/rmgr/drv_interface.c
> +++ b/drivers/staging/tidspbridge/rmgr/drv_interface.c
> @@ -258,7 +258,17 @@ err:
>  /* This function maps kernel space memory to user space memory. */
>  static int bridge_mmap(struct file *filp, struct vm_area_struct *vma)
>  {
> -	u32 status;
> +	int status;
> +	struct omap_dsp_platform_data *pdata =
> +	    omap_dspbridge_dev->dev.platform_data;
> +	unsigned long start = vma->vm_pgoff << PAGE_SHIFT;
> +		
> +	if (start < pdata->phys_mempool_base)
> +		return -EINVAL;
> +
> +	if (vma->vm_end - vma->vm_start + (start - pdata->phys_mempool_base) 
> +	    > pdata->phys_mempool_size)

This test is vulnerable to integer overflows if you pick a very high
value for start.  Consider using the vm_iomap_memory() helper function
instead of calling remap_pfn_range() directly.  Commit 7314e613d5ff
('Fix a few incorrectly checked [io_]remap_pfn_range() calls') has an
example of how the conversion works.

regards,
dan carpenter