Re: [PATCH 1/1] net: thunderx: avoid null pointer dereference

On 05/17/2016 10:21 PM, David Daney wrote:

On 05/17/2016 12:40 PM, Heinrich Schuchardt wrote:

quoted

In function bgx_lmac_handler only use a member of
lmac after checking it is not null.

Were you seeing and actual OOPs, or is this only a theoretical problem
flagged by some sort of static code checker?

Cppcheck identified the code inconsistency.

If the struct netdev * parameter cannot be NULL, you could eliminate the
check.

Other drivers do no have check against a NULL parameter in the
adjust_link handler function (e.g. bgmac_adjust_link,
pxa168_eth_adjust_link).

Best regards

Heinrich

quoted

Signed-off-by: Heinrich Schuchardt <redacted>

In any event, it seems sane to make this change, so:

Acked-by: David Daney <redacted>

quoted

---
  drivers/net/ethernet/cavium/thunder/thunder_bgx.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c

b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
index d20539a..3ed2198 100644

--- a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c

@@ -274,12 +274,14 @@ static void bgx_sgmii_change_link_state(struct

lmac *lmac)
  static void bgx_lmac_handler(struct net_device *netdev)
  {
      struct lmac *lmac = container_of(netdev, struct lmac, netdev);
-    struct phy_device *phydev = lmac->phydev;
+    struct phy_device *phydev;
      int link_changed = 0;

      if (!lmac)
          return;

+    phydev = lmac->phydev;
+
      if (!phydev->link && lmac->last_link)
          link_changed = -1;