[PATCH] net: wan: fix error return code of uhdlc_init()

[Jia-Ju Bai]

When priv->rx_skbuff or priv->tx_skbuff is NULL, no error return code of
uhdlc_init() is assigned.
To fix this bug, ret is assigned with -ENOMEM in these cases.

Reported-by: TOTE Robot 
Signed-off-by: Jia-Ju Bai 
---
 drivers/net/wan/fsl_ucc_hdlc.c | 8 ++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/net/wan/fsl_ucc_hdlc.c b/drivers/net/wan/fsl_ucc_hdlc.c
index dca97cd7c4e7..7eac6a3e1cde 100644
--- a/drivers/net/wan/fsl_ucc_hdlc.c
+++ b/drivers/net/wan/fsl_ucc_hdlc.c
@@ -204,14 +204,18 @@ static int uhdlc_init(struct ucc_hdlc_private *priv)
priv->rx_skbuff = kcalloc(priv->rx_ring_size,
  sizeof(*priv->rx_skbuff),
  GFP_KERNEL);
-   if (!priv->rx_skbuff)
+   if (!priv->rx_skbuff) {
+   ret = -ENOMEM;
goto free_ucc_pram;
+   }
 
priv->tx_skbuff = kcalloc(priv->tx_ring_size,
  sizeof(*priv->tx_skbuff),
  GFP_KERNEL);
-   if (!priv->tx_skbuff)
+   if (!priv->tx_skbuff) {
+   ret = -ENOMEM;
goto free_rx_skbuff;
+   }
 
priv->skb_curtx = 0;
priv->skb_dirtytx = 0;
-- 
2.17.1

[BUG] drivers/tty/hvc: a possible sleep-in-atomic bug in xen_pv_console_init

[Jia-Ju Bai]

According to drivers/tty/hvc/hvc_xen.c, the kernel module may sleep 
under a spinlock.

The function call path is:
xen_pv_console_init (acquire the spinlock)
  xencons_info_pv_init
gfn_to_virt (gfn_to_pfn) --> may sleep

I do not find a good way to fix it, so I only report.
This possible bug is found by my static analysis tool (DSAC) and my code 
review.



Thanks,
Jia-Ju Bai