[PATCH 1/1] bfa: Fixes for 0-terminated strncpy and possible null pointer dereference

[PATCH 1/1] bfa: Fixes for 0-terminated strncpy and possible null pointer dereference

[Jakob Normark]

This patch fixes two cppcheck errors in drivers/scsi/bfa/bfad_im.c
Kernel version: v3.10-rc1

Signed-off-by: Jakob Normark <jakobnormark@gmail.com>
---
 drivers/scsi/bfa/bfad_im.c |    9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c
index 5864f98..9489c56 100644
--- a/drivers/scsi/bfa/bfad_im.c
+++ b/drivers/scsi/bfa/bfad_im.c
@@ -944,13 +944,15 @@ static int
 bfad_im_slave_alloc(struct scsi_device *sdev)
 {
 	struct fc_rport *rport = starget_to_rport(scsi_target(sdev));
-	struct bfad_itnim_data_s *itnim_data =
-				(struct bfad_itnim_data_s *) rport->dd_data;
-	struct bfa_s *bfa = itnim_data->itnim->bfa_itnim->bfa;
+	struct bfad_itnim_data_s *itnim_data;
+	struct bfa_s *bfa;
 
 	if (!rport || fc_remote_port_chkready(rport))
 		return -ENXIO;
 
+	itnim_data = (struct bfad_itnim_data_s *) rport->dd_data;
+	bfa = itnim_data->itnim->bfa_itnim->bfa;
+
 	if (bfa_get_lun_mask_status(bfa) == BFA_LUNMASK_ENABLED) {
 		/*
 		 * We should not mask LUN 0 - since this will translate
@@ -1037,6 +1039,7 @@ bfad_fc_host_init(struct bfad_im_port_s *im_port)
 
 	strncpy(symname, bfad->bfa_fcs.fabric.bport.port_cfg.sym_name.symname,
 		BFA_SYMNAME_MAXLEN);
+	symname[BFA_SYMNAME_MAXLEN - 1] = '\0';
 	sprintf(fc_host_symbolic_name(host), "%s", symname);
 
 	fc_host_supported_speeds(host) = bfad_im_supported_speeds(&bfad->bfa);
-- 
1.7.9.5

RE: [PATCH 1/1] bfa: Fixes for 0-terminated strncpy and possible null pointer dereference

[Vijay Mohan Guvva]

> -----Original Message-----
> From: linux-scsi-owner@vger.kernel.org [mailto:linux-scsi-
> owner@vger.kernel.org] On Behalf Of Jakob Normark
> Sent: Thursday, May 16, 2013 1:12 AM
> To: Anil Gurumurthy; Vijay Mohan Guvva; James E.J. Bottomley
> Cc: linux-scsi@vger.kernel.org; linux-kernel@vger.kernel.org; Jakob Normark
> Subject: [PATCH 1/1] bfa: Fixes for 0-terminated strncpy and possible null
> pointer dereference
> 
> This patch fixes two cppcheck errors in drivers/scsi/bfa/bfad_im.c Kernel
> version: v3.10-rc1
> 
> Signed-off-by: Jakob Normark <jakobnormark@gmail.com>
> ---
>  drivers/scsi/bfa/bfad_im.c |    9 ++++++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c index
> 5864f98..9489c56 100644
> --- a/drivers/scsi/bfa/bfad_im.c
> +++ b/drivers/scsi/bfa/bfad_im.c
> @@ -944,13 +944,15 @@ static int
>  bfad_im_slave_alloc(struct scsi_device *sdev)  {
>  	struct fc_rport *rport = starget_to_rport(scsi_target(sdev));
> -	struct bfad_itnim_data_s *itnim_data =
> -				(struct bfad_itnim_data_s *) rport->dd_data;
> -	struct bfa_s *bfa = itnim_data->itnim->bfa_itnim->bfa;
> +	struct bfad_itnim_data_s *itnim_data;
> +	struct bfa_s *bfa;
> 
>  	if (!rport || fc_remote_port_chkready(rport))
>  		return -ENXIO;
> 
> +	itnim_data = (struct bfad_itnim_data_s *) rport->dd_data;
> +	bfa = itnim_data->itnim->bfa_itnim->bfa;
> +
>  	if (bfa_get_lun_mask_status(bfa) == BFA_LUNMASK_ENABLED) {
>  		/*
>  		 * We should not mask LUN 0 - since this will translate @@ -
> 1037,6 +1039,7 @@ bfad_fc_host_init(struct bfad_im_port_s *im_port)
> 
>  	strncpy(symname, bfad-
> >bfa_fcs.fabric.bport.port_cfg.sym_name.symname,
>  		BFA_SYMNAME_MAXLEN);
> +	symname[BFA_SYMNAME_MAXLEN - 1] = '\0';
>  	sprintf(fc_host_symbolic_name(host), "%s", symname);
> 
>  	fc_host_supported_speeds(host) =
> bfad_im_supported_speeds(&bfad->bfa);
> --
> 1.7.9.5

Changes looks good. Thanks for the patch.
Acked-by: Vijay Mohan Guvva <vmohan@brocade.com>

Re: [PATCH 1/1] bfa: Fixes for 0-terminated strncpy and possible null pointer dereference

[James Bottomley]

On Wed, 2013-05-15 at 21:41 +0200, Jakob Normark wrote:
> This patch fixes two cppcheck errors in drivers/scsi/bfa/bfad_im.c
> 
> Signed-off-by: Jakob Normark <jakobnormark@gmail.com>
> ---
>  drivers/scsi/bfa/bfad_im.c |    9 ++++++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c
> index 5864f98..9489c56 100644
> --- a/drivers/scsi/bfa/bfad_im.c
> +++ b/drivers/scsi/bfa/bfad_im.c
> @@ -944,13 +944,15 @@ static int
>  bfad_im_slave_alloc(struct scsi_device *sdev)
>  {
>  	struct fc_rport *rport = starget_to_rport(scsi_target(sdev));
> -	struct bfad_itnim_data_s *itnim_data =
> -				(struct bfad_itnim_data_s *) rport->dd_data;
> -	struct bfa_s *bfa = itnim_data->itnim->bfa_itnim->bfa;
> +	struct bfad_itnim_data_s *itnim_data;
> +	struct bfa_s *bfa;
>  
>  	if (!rport || fc_remote_port_chkready(rport))
>  		return -ENXIO;
>  
> +	itnim_data = (struct bfad_itnim_data_s *) rport->dd_data;
> +	bfa = itnim_data->itnim->bfa_itnim->bfa;
> +
>  	if (bfa_get_lun_mask_status(bfa) == BFA_LUNMASK_ENABLED) {
>  		/*
>  		 * We should not mask LUN 0 - since this will translate
> @@ -1037,6 +1039,7 @@ bfad_fc_host_init(struct bfad_im_port_s *im_port)
>  
>  	strncpy(symname, bfad->bfa_fcs.fabric.bport.port_cfg.sym_name.symname,
>  		BFA_SYMNAME_MAXLEN);
> +	symname[BFA_SYMNAME_MAXLEN - 1] = '\0';

This isn't correct, it's effectively a handrolled version of strlcpy.
The correct fix is to replace strncpy with strlcpy.  I fixed it up in
the patch version.

James