From: Wong Vee Khee vee.khee.wong@linux.intel.com
[ Upstream commit 82a44ae113b7b35850f4542f0443fcab221e376a ]
In the case of taprio offload is not enabled, the error handling path causes a kernel crash due to kernel NULL pointer deference.
Fix this by adding check for NULL before attempt to access 'plat->est' on the mutex_lock() call.
The following kernel panic is observed without this patch:
RIP: 0010:mutex_lock+0x10/0x20 Call Trace: tc_setup_taprio+0x482/0x560 [stmmac] kmem_cache_alloc_trace+0x13f/0x490 taprio_disable_offload.isra.0+0x9d/0x180 [sch_taprio] taprio_destroy+0x6c/0x100 [sch_taprio] qdisc_create+0x2e5/0x4f0 tc_modify_qdisc+0x126/0x740 rtnetlink_rcv_msg+0x12b/0x380 _raw_spin_lock_irqsave+0x19/0x40 _raw_spin_unlock_irqrestore+0x18/0x30 create_object+0x212/0x340 rtnl_calcit.isra.0+0x110/0x110 netlink_rcv_skb+0x50/0x100 netlink_unicast+0x191/0x230 netlink_sendmsg+0x243/0x470 sock_sendmsg+0x5e/0x60 ____sys_sendmsg+0x20b/0x280 copy_msghdr_from_user+0x5c/0x90 __mod_memcg_state+0x87/0xf0 ___sys_sendmsg+0x7c/0xc0 lru_cache_add+0x7f/0xa0 _raw_spin_unlock+0x16/0x30 wp_page_copy+0x449/0x890 handle_mm_fault+0x921/0xfc0 __sys_sendmsg+0x59/0xa0 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 ---[ end trace b1f19b24368a96aa ]---
Fixes: b60189e0392f ("net: stmmac: Integrate EST with TAPRIO scheduler API") Cc: stable@vger.kernel.org # 5.10.x Signed-off-by: Wong Vee Khee vee.khee.wong@linux.intel.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Sasha Levin sashal@kernel.org --- drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c index 8c2eae2a7efd..22c34474e617 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c @@ -781,11 +781,13 @@ static int tc_setup_taprio(struct stmmac_priv *priv, return 0;
disable: - mutex_lock(&priv->plat->est->lock); - priv->plat->est->enable = false; - stmmac_est_configure(priv, priv->ioaddr, priv->plat->est, - priv->plat->clk_ptp_rate); - mutex_unlock(&priv->plat->est->lock); + if (priv->plat->est) { + mutex_lock(&priv->plat->est->lock); + priv->plat->est->enable = false; + stmmac_est_configure(priv, priv->ioaddr, priv->plat->est, + priv->plat->clk_ptp_rate); + mutex_unlock(&priv->plat->est->lock); + }
return ret; }