Re: [dpdk-dev] [PATCH] examples/vhost_scsi: fix buffer not terminated

[Bruce Richardson <bruce.richardson@intel.com>]

On Thu, Oct 05, 2017 at 02:42:24PM +0200, Maxime Coquelin wrote:
> 
> 
> On 10/05/2017 02:35 PM, Piasecki, JacekX wrote:
> > -----Original Message-----
> > From: Maxime Coquelin [mailto:maxime.coquelin@redhat.com]
> > Sent: Monday, 2 October, 2017 17:07
> > To: Jastrzebski, MichalX K <michalx.k.jastrzebski@intel.com>; yliu@fridaylinux.org
> > Cc: dev@dpdk.org; Jain, Deepak K <deepak.k.jain@intel.com>; Piasecki, JacekX <jacekx.piasecki@intel.com>; Liu, Changpeng <changpeng.liu@intel.com>; stable@dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH] examples/vhost_scsi: fix buffer not terminated
> > 
> > 
> > 
> > On 10/02/2017 03:50 PM, Jastrzebski, MichalX K wrote:
> > > > -----Original Message-----
> > > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Michal
> > > > Jastrzebski
> > > > Sent: Friday, September 22, 2017 3:08 PM
> > > > To: yliu@fridaylinux.org; maxime.coquelin@redhat.com
> > > > Cc: dev@dpdk.org; Jain, Deepak K <deepak.k.jain@intel.com>; Piasecki,
> > > > JacekX <jacekx.piasecki@intel.com>; Liu, Changpeng
> > > > <changpeng.liu@intel.com>; stable@dpdk.org
> > > > Subject: [dpdk-dev] [PATCH] examples/vhost_scsi: fix buffer not
> > > > terminated
> > > > 
> > > > From: Jacek Piasecki <jacekx.piasecki@intel.com>
> > > > 
> > > > Fix size of buffer in strcpy. There was possible to get not
> > > > terminated string after copy operation.
> > > > 
> > > > Coverity issue: 158631
> > > > Fixes: db75c7af19bb ("examples/vhost_scsi: introduce a new sample
> > > > app")
> > > > Cc: changpeng.liu@intel.com
> > > > Cc: stable@dpdk.org
> > > > 
> > > > Signed-off-by: Jacek Piasecki <jacekx.piasecki@intel.com>
> > > > ---
> > > >    examples/vhost_scsi/scsi.c | 3 ++-
> > > >    1 file changed, 2 insertions(+), 1 deletion(-)
> > > > 
> > > > diff --git a/examples/vhost_scsi/scsi.c b/examples/vhost_scsi/scsi.c
> > > > index 54d3104..de9639a 100644
> > > > --- a/examples/vhost_scsi/scsi.c
> > > > +++ b/examples/vhost_scsi/scsi.c
> > > > @@ -307,7 +307,8 @@
> > > >    		strncpy((char *)inqdata->t10_vendor_id, "INTEL", 8);
> > > > 
> > > >    		/* PRODUCT IDENTIFICATION */
> > > > -		strncpy((char *)inqdata->product_id, bdev->product_name,
> > > > 16);
> > > > +		strncpy((char *)inqdata->product_id, bdev->product_name,
> > > > +				ARRAY_SIZE(inqdata->product_id) - 1);
> > 
> > Does it assume that product_id is memzero'ed before?
> > IIUC strncpy manpage, it wouldn't protect against non-null terminated strings if it is not the case:
> > 
> > Yes, I assumed that this area is zeroed before strncpy. Earlier there is a casting from *task to *inqdata (*task is zmalloced).
> Ok. I think the assumption is dangerous.
> 
> > To be sure that the destination area is clear I propose to add:
> > memset(inqdata->product_id, 0, ARRAY_SIZE(inqdata->product_id));
> > before strncpy. Would that be OK?
> 
> Or call strncpy with ARRAY_SIZE(inqdata->product_id) as max length,
> and then do
> inqdata->product_id[ARRAY_SIZE(inqdata->product_id) - 1] = '\0';
> 
> Regards,
> Maxime
> > 
Or use snprintf instead of strncpy, as we do in many places in DPDK.
Still only one line of code, with nice null-termination, and truncation
semantics.

/Bruce