Re: [PATCH][V2] usb: storage: return on error to avoid a null pointer dereference

From: Alan Stern
Date: Thu Jul 06 2017 - 11:48:47 EST


On Thu, 6 Jul 2017, Colin King wrote:

> From: Colin Ian King <colin.king@xxxxxxxxxxxxx>
>
> When us->extra is null the driver is not initialized, however, a
> later call to osd200_scsi_to_ata is made that dereferences
> us->extra, causing a null pointer dereference. The code
> currently detects and reports that the driver is not initialized;
> add a return to avoid the subsequent dereference issue in this
> check.
>
> Thanks to Alan Stern for pointing out that srb->result needs setting
> to DID_ERROR << 16
>
> Detected by CoverityScan, CID#100308 ("Dereference after null check")
>
> Signed-off-by: Colin Ian King <colin.king@xxxxxxxxxxxxx>

Acked-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>

> ---
> drivers/usb/storage/isd200.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/usb/storage/isd200.c b/drivers/usb/storage/isd200.c
> index fba4005dd737..6a7720e66595 100644
> --- a/drivers/usb/storage/isd200.c
> +++ b/drivers/usb/storage/isd200.c
> @@ -1529,8 +1529,11 @@ static void isd200_ata_command(struct scsi_cmnd *srb, struct us_data *us)
>
> /* Make sure driver was initialized */
>
> - if (us->extra == NULL)
> + if (us->extra == NULL) {
> usb_stor_dbg(us, "ERROR Driver not initialized\n");
> + srb->result = DID_ERROR << 16;
> + return;
> + }
>
> scsi_set_resid(srb, 0);
> /* scsi_bufflen might change in protocol translation to ata */
>