> -----Original Message----- > From: Brian King <brk...@linux.vnet.ibm.com> > Sent: Tuesday, July 16, 2019 3:12 AM > To: GR-everest-linux-l2 <gr-everest-linux...@marvell.com> > Cc: Sudarsana Reddy Kalluru <skall...@marvell.com>; Ariel Elior > <ael...@marvell.com>; netdev@vger.kernel.org; Brian King > <brk...@linux.vnet.ibm.com> > Subject: [EXT] [PATCH] bnx2x: Prevent load reordering in tx completion > processing > > External Email > > ---------------------------------------------------------------------- > This patch fixes an issue seen on Power systems with bnx2x which results in > the skb is NULL WARN_ON in bnx2x_free_tx_pkt firing due to the skb pointer > getting loaded in bnx2x_free_tx_pkt prior to the hw_cons load in > bnx2x_tx_int. Adding a read memory barrier resolves the issue. > > Signed-off-by: Brian King <brk...@linux.vnet.ibm.com> > --- > drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c > b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c > index 656ed80..e2be5a6 100644 > --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c > +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c > @@ -285,6 +285,9 @@ int bnx2x_tx_int(struct bnx2x *bp, struct > bnx2x_fp_txdata *txdata) > hw_cons = le16_to_cpu(*txdata->tx_cons_sb); > sw_cons = txdata->tx_pkt_cons; > > + /* Ensure subsequent loads occur after hw_cons */ > + smp_rmb(); > + > while (sw_cons != hw_cons) { > u16 pkt_cons; > > -- > 1.8.3.1
Could you please explain a bit in detail what could have caused skb to NULL exactly ? Curious that if skb would have been NULL for some reason it did not cause NULL pointer dereference in bnx2x_free_tx_pkt() on below call - prefetch(&skb->end); Which is prior to the said WARN_ON(!skb) in bnx2x_free_tx_pkt().