[PATCH AUTOSEL 5.5 006/106] net/mlx5e: Enforce setting of a single FEC mode

[PATCH AUTOSEL 5.5 006/106] net/mlx5e: Enforce setting of a single FEC mode

[Sasha Levin]

From: Aya Levin <ayal@mellanox.com>

[ Upstream commit 4bd9d5070b92da012f2715cf8e4859acb78b8f35 ]

Ethtool command allow setting of several FEC modes in a single set
command. The driver can only set a single FEC mode at a time. With this
patch driver will reply not-supported on setting several FEC modes.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index c6776f308d5e6..ce63f9bd10fa8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -1554,6 +1554,10 @@ static int mlx5e_set_fecparam(struct net_device *netdev,
 	int mode;
 	int err;
 
+	if (bitmap_weight((unsigned long *)&fecparam->fec,
+			  ETHTOOL_FEC_BASER_BIT + 1) > 1)
+		return -EOPNOTSUPP;
+
 	for (mode = 0; mode < ARRAY_SIZE(pplm_fec_2_ethtool); mode++) {
 		if (!(pplm_fec_2_ethtool[mode] & fecparam->fec))
 			continue;
-- 
2.20.1

Re: [PATCH AUTOSEL 5.5 006/106] net/mlx5e: Enforce setting of a single FEC mode

[Or Gerlitz]

On Thu, Apr 16, 2020 at 2:56 AM Sasha Levin <sashal@kernel.org> wrote:
> From: Aya Levin <ayal@mellanox.com>
> [ Upstream commit 4bd9d5070b92da012f2715cf8e4859acb78b8f35 ]
>
> Ethtool command allow setting of several FEC modes in a single set
> command. The driver can only set a single FEC mode at a time. With this
> patch driver will reply not-supported on setting several FEC modes.
>
> Signed-off-by: Aya Levin <ayal@mellanox.com>
> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
>  drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 4 ++++

Haven't we agreed that drivers/net/ethernet/mellanox/mlx5 is not
subject to autosel anymore?!

Re: [PATCH AUTOSEL 5.5 006/106] net/mlx5e: Enforce setting of a single FEC mode

[Sasha Levin]

On Sat, Apr 18, 2020 at 10:52:59PM +0300, Or Gerlitz wrote:
>On Thu, Apr 16, 2020 at 2:56 AM Sasha Levin <sashal@kernel.org> wrote:
>> From: Aya Levin <ayal@mellanox.com>
>> [ Upstream commit 4bd9d5070b92da012f2715cf8e4859acb78b8f35 ]
>>
>> Ethtool command allow setting of several FEC modes in a single set
>> command. The driver can only set a single FEC mode at a time. With this
>> patch driver will reply not-supported on setting several FEC modes.
>>
>> Signed-off-by: Aya Levin <ayal@mellanox.com>
>> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
>> Signed-off-by: Sasha Levin <sashal@kernel.org>
>> ---
>>  drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 4 ++++
>
>Haven't we agreed that drivers/net/ethernet/mellanox/mlx5 is not
>subject to autosel anymore?!

On Thu, Apr 16, 2020 at 09:08:06PM +0000, Saeed Mahameed wrote:
>Please don't opt mlx5 out just yet ;-), i need to do some more research
>and make up my mind..

It would be awesome if the Mellanox folks could agree between
themselves whether they want it or not and let us know which option they
pick.

Again, I really don't care whether mlx5 is opted out or not, I'm not
going to argue with anyone if you want to opt out, I just want to know
whether you're out or not :)

-- 
Thanks,
Sasha

Re: [PATCH AUTOSEL 5.5 006/106] net/mlx5e: Enforce setting of a single FEC mode

[Or Gerlitz]

On Sun, Apr 19, 2020 at 12:20 AM Sasha Levin <sashal@kernel.org> wrote:
> On Sat, Apr 18, 2020 at 10:52:59PM +0300, Or Gerlitz wrote:
> >On Thu, Apr 16, 2020 at 2:56 AM Sasha Levin <sashal@kernel.org> wrote:
> >> From: Aya Levin <ayal@mellanox.com>
> >> [ Upstream commit 4bd9d5070b92da012f2715cf8e4859acb78b8f35 ]
> >>
> >> Ethtool command allow setting of several FEC modes in a single set
> >> command. The driver can only set a single FEC mode at a time. With this
> >> patch driver will reply not-supported on setting several FEC modes.
> >>
> >> Signed-off-by: Aya Levin <ayal@mellanox.com>
> >> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
> >> Signed-off-by: Sasha Levin <sashal@kernel.org>
> >> ---
> >>  drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 4 ++++
> >
> >Haven't we agreed that drivers/net/ethernet/mellanox/mlx5 is not
> >subject to autosel anymore?!
>
> On Thu, Apr 16, 2020 at 09:08:06PM +0000, Saeed Mahameed wrote:
> >Please don't opt mlx5 out just yet ;-), i need to do some more research
> >and make up my mind..
>
> It would be awesome if the Mellanox folks could agree between
> themselves whether they want it or not and let us know which option they
> pick.

My reply did had a question mark. I missed remembering this response
from Saeed over the other thread and remembered your "no problem opting
it out" response.

> Again, I really don't care whether mlx5 is opted out or not, I'm not
> going to argue with anyone if you want to opt out, I just want to know
> whether you're out or not :)

let us continue on the other thread, I think whole netdev (net + drivers)
should be opted out by default and only on those folders you get ack
from the subsystem maintainer and driver maintainer you can opt in.

[PATCH AUTOSEL 5.5 009/106] bpf: Reliably preserve btf_trace_xxx types

[Sasha Levin]

From: Andrii Nakryiko <andriin@fb.com>

[ Upstream commit 441420a1f0b3031f228453697406c86f110e59d4 ]

btf_trace_xxx types, crucial for tp_btf BPF programs (raw tracepoint with
verifier-checked direct memory access), have to be preserved in kernel BTF to
allow verifier do its job and enforce type/memory safety. It was reported
([0]) that for kernels built with Clang current type-casting approach doesn't
preserve these types.

This patch fixes it by declaring an anonymous union for each registered
tracepoint, capturing both struct bpf_raw_event_map information, as well as
recording btf_trace_##call type reliably. Structurally, it's still the same
content as for a plain struct bpf_raw_event_map, so no other changes are
necessary.

  [0] https://github.com/iovisor/bcc/issues/2770#issuecomment-591007692

Fixes: e8c423fb31fa ("bpf: Add typecast to raw_tracepoints to help BTF generation")
Reported-by: Wenbo Zhang <ethercflow@gmail.com>
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200301081045.3491005-2-andriin@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/trace/bpf_probe.h | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/include/trace/bpf_probe.h b/include/trace/bpf_probe.h
index b04c292709730..1ce3be63add1f 100644
--- a/include/trace/bpf_probe.h
+++ b/include/trace/bpf_probe.h
@@ -75,13 +75,17 @@ static inline void bpf_test_probe_##call(void)				\
 	check_trace_callback_type_##call(__bpf_trace_##template);	\
 }									\
 typedef void (*btf_trace_##call)(void *__data, proto);			\
-static struct bpf_raw_event_map	__used					\
-	__attribute__((section("__bpf_raw_tp_map")))			\
-__bpf_trace_tp_map_##call = {						\
-	.tp		= &__tracepoint_##call,				\
-	.bpf_func	= (void *)(btf_trace_##call)__bpf_trace_##template,	\
-	.num_args	= COUNT_ARGS(args),				\
-	.writable_size	= size,						\
+static union {								\
+	struct bpf_raw_event_map event;					\
+	btf_trace_##call handler;					\
+} __bpf_trace_tp_map_##call __used					\
+__attribute__((section("__bpf_raw_tp_map"))) = {			\
+	.event = {							\
+		.tp		= &__tracepoint_##call,			\
+		.bpf_func	= __bpf_trace_##template,		\
+		.num_args	= COUNT_ARGS(args),			\
+		.writable_size	= size,					\
+	},								\
 };
 
 #define FIRST(x, ...) x
-- 
2.20.1

[PATCH AUTOSEL 5.5 055/106] net: phy: at803x: fix clock sink configuration on ATH8030 and ATH8035

[Sasha Levin]

From: Oleksij Rempel <o.rempel@pengutronix.de>

[ Upstream commit b1f4c209d84057b6d40b939b6e4404854271d797 ]

The masks in priv->clk_25m_reg and priv->clk_25m_mask are one-bits-set
for the values that comprise the fields, not zero-bits-set.

This patch fixes the clock frequency configuration for ATH8030 and
ATH8035 Atheros PHYs by removing the erroneous "~".

To reproduce this bug, configure the PHY  with the device tree binding
"qca,clk-out-frequency" and remove the machine specific PHY fixups.

Fixes: 2f664823a47021 ("net: phy: at803x: add device tree binding")
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reported-by: Russell King <rmk+kernel@armlinux.org.uk>
Reviewed-by: Russell King <rmk+kernel@armlinux.org.uk>
Tested-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/phy/at803x.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/phy/at803x.c b/drivers/net/phy/at803x.c
index 481cf48c9b9e4..31f731e6df720 100644
--- a/drivers/net/phy/at803x.c
+++ b/drivers/net/phy/at803x.c
@@ -425,8 +425,8 @@ static int at803x_parse_dt(struct phy_device *phydev)
 		 */
 		if (at803x_match_phy_id(phydev, ATH8030_PHY_ID) ||
 		    at803x_match_phy_id(phydev, ATH8035_PHY_ID)) {
-			priv->clk_25m_reg &= ~AT8035_CLK_OUT_MASK;
-			priv->clk_25m_mask &= ~AT8035_CLK_OUT_MASK;
+			priv->clk_25m_reg &= AT8035_CLK_OUT_MASK;
+			priv->clk_25m_mask &= AT8035_CLK_OUT_MASK;
 		}
 	}
 
-- 
2.20.1

[PATCH AUTOSEL 5.5 056/106] cxgb4: fix MPS index overwrite when setting MAC address

[Sasha Levin]

From: Herat Ramani <herat@chelsio.com>

[ Upstream commit 41aa8561ca3fc5748391f08cc5f3e561923da52c ]

cxgb4_update_mac_filt() earlier requests firmware to add a new MAC
address into MPS TCAM. The MPS TCAM index returned by firmware is
stored in pi->xact_addr_filt. However, the saved MPS TCAM index gets
overwritten again with the return value of cxgb4_update_mac_filt(),
which is wrong.

When trying to update to another MAC address later, the wrong MPS TCAM
index is sent to firmware, which causes firmware to return error,
because it's not the same MPS TCAM index that firmware had sent
earlier to driver.

So, fix by removing the wrong overwrite being done after call to
cxgb4_update_mac_filt().

Fixes: 3f8cfd0d95e6 ("cxgb4/cxgb4vf: Program hash region for {t4/t4vf}_change_mac()")
Signed-off-by: Herat Ramani <herat@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index b11ba4565c20b..9934c8b0f0b07 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -3110,7 +3110,6 @@ static int cxgb_set_mac_addr(struct net_device *dev, void *p)
 		return ret;
 
 	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
-	pi->xact_addr_filt = ret;
 	return 0;
 }
 
-- 
2.20.1

[PATCH AUTOSEL 5.5 057/106] slcan: Don't transmit uninitialized stack data in padding

[Sasha Levin]

From: Richard Palethorpe <rpalethorpe@suse.com>

[ Upstream commit b9258a2cece4ec1f020715fe3554bc2e360f6264 ]

struct can_frame contains some padding which is not explicitly zeroed in
slc_bump. This uninitialized data will then be transmitted if the stack
initialization hardening feature is not enabled (CONFIG_INIT_STACK_ALL).

This commit just zeroes the whole struct including the padding.

Signed-off-by: Richard Palethorpe <rpalethorpe@suse.com>
Fixes: a1044e36e457 ("can: add slcan driver for serial/USB-serial CAN adapters")
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: linux-can@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: security@kernel.org
Cc: wg@grandegger.com
Cc: mkl@pengutronix.de
Cc: davem@davemloft.net
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/can/slcan.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/can/slcan.c b/drivers/net/can/slcan.c
index a3664281a33fc..4dfa459ef5c73 100644
--- a/drivers/net/can/slcan.c
+++ b/drivers/net/can/slcan.c
@@ -148,7 +148,7 @@ static void slc_bump(struct slcan *sl)
 	u32 tmpid;
 	char *cmd = sl->rbuff;
 
-	cf.can_id = 0;
+	memset(&cf, 0, sizeof(cf));
 
 	switch (*cmd) {
 	case 'r':
@@ -187,8 +187,6 @@ static void slc_bump(struct slcan *sl)
 	else
 		return;
 
-	*(u64 *) (&cf.data) = 0; /* clear payload */
-
 	/* RTR frames may have a dlc > 0 but they never have any data bytes */
 	if (!(cf.can_id & CAN_RTR_FLAG)) {
 		for (i = 0; i < cf.can_dlc; i++) {
-- 
2.20.1

[PATCH AUTOSEL 5.5 058/106] net: qualcomm: rmnet: Allow configuration updates to existing devices

[Sasha Levin]

From: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>

[ Upstream commit 2abb5792387eb188b12051337d5dcd2cba615cb0 ]

This allows the changelink operation to succeed if the mux_id was
specified as an argument. Note that the mux_id must match the
existing mux_id of the rmnet device or should be an unused mux_id.

Fixes: 1dc49e9d164c ("net: rmnet: do not allow to change mux id if mux id is duplicated")
Reported-and-tested-by: Alex Elder <elder@linaro.org>
Signed-off-by: Sean Tranchetti <stranche@codeaurora.org>
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 .../ethernet/qualcomm/rmnet/rmnet_config.c    | 31 ++++++++++++-------
 1 file changed, 19 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c b/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c
index fbf4cbcf1a654..02cdbb22d3355 100644
--- a/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c
+++ b/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c
@@ -279,7 +279,6 @@ static int rmnet_changelink(struct net_device *dev, struct nlattr *tb[],
 {
 	struct rmnet_priv *priv = netdev_priv(dev);
 	struct net_device *real_dev;
-	struct rmnet_endpoint *ep;
 	struct rmnet_port *port;
 	u16 mux_id;
 
@@ -294,19 +293,27 @@ static int rmnet_changelink(struct net_device *dev, struct nlattr *tb[],
 
 	if (data[IFLA_RMNET_MUX_ID]) {
 		mux_id = nla_get_u16(data[IFLA_RMNET_MUX_ID]);
-		if (rmnet_get_endpoint(port, mux_id)) {
-			NL_SET_ERR_MSG_MOD(extack, "MUX ID already exists");
-			return -EINVAL;
-		}
-		ep = rmnet_get_endpoint(port, priv->mux_id);
-		if (!ep)
-			return -ENODEV;
 
-		hlist_del_init_rcu(&ep->hlnode);
-		hlist_add_head_rcu(&ep->hlnode, &port->muxed_ep[mux_id]);
+		if (mux_id != priv->mux_id) {
+			struct rmnet_endpoint *ep;
+
+			ep = rmnet_get_endpoint(port, priv->mux_id);
+			if (!ep)
+				return -ENODEV;
 
-		ep->mux_id = mux_id;
-		priv->mux_id = mux_id;
+			if (rmnet_get_endpoint(port, mux_id)) {
+				NL_SET_ERR_MSG_MOD(extack,
+						   "MUX ID already exists");
+				return -EINVAL;
+			}
+
+			hlist_del_init_rcu(&ep->hlnode);
+			hlist_add_head_rcu(&ep->hlnode,
+					   &port->muxed_ep[mux_id]);
+
+			ep->mux_id = mux_id;
+			priv->mux_id = mux_id;
+		}
 	}
 
 	if (data[IFLA_RMNET_FLAGS]) {
-- 
2.20.1

[PATCH AUTOSEL 5.5 060/106] net: stmmac: dwmac1000: fix out-of-bounds mac address reg setting

[Sasha Levin]

From: Jisheng Zhang <Jisheng.Zhang@synaptics.com>

[ Upstream commit 3e1221acf6a8f8595b5ce354bab4327a69d54d18 ]

Commit 9463c4455900 ("net: stmmac: dwmac1000: Clear unused address
entries") cleared the unused mac address entries, but introduced an
out-of bounds mac address register programming bug -- After setting
the secondary unicast mac addresses, the "reg" value has reached
netdev_uc_count() + 1, thus we should only clear address entries
if (addr < perfect_addr_number)

Fixes: 9463c4455900 ("net: stmmac: dwmac1000: Clear unused address entries")
Signed-off-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/stmicro/stmmac/dwmac1000_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_core.c b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_core.c
index 542784300620f..efc6ec1b8027c 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_core.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_core.c
@@ -207,7 +207,7 @@ static void dwmac1000_set_filter(struct mac_device_info *hw,
 			reg++;
 		}
 
-		while (reg <= perfect_addr_number) {
+		while (reg < perfect_addr_number) {
 			writel(0, ioaddr + GMAC_ADDR_HIGH(reg));
 			writel(0, ioaddr + GMAC_ADDR_LOW(reg));
 			reg++;
-- 
2.20.1

[PATCH AUTOSEL 5.5 061/106] net: dsa: mt7530: fix null pointer dereferencing in port5 setup

[Sasha Levin]

From: Chuanhong Guo <gch981213@gmail.com>

[ Upstream commit 0452800f6db4ed0a42ffb15867c0acfd68829f6a ]

The 2nd gmac of mediatek soc ethernet may not be connected to a PHY
and a phy-handle isn't always available.
Unfortunately, mt7530 dsa driver assumes that the 2nd gmac is always
connected to switch port 5 and setup mt7530 according to phy address
of 2nd gmac node, causing null pointer dereferencing when phy-handle
isn't defined in dts.
This commit fix this setup code by checking return value of
of_parse_phandle before using it.

Fixes: 38f790a80560 ("net: dsa: mt7530: Add support for port 5")
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Tested-by: René van Dorst <opensource@vdorst.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/dsa/mt7530.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 60a8a68567baa..936b9b65accab 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -1355,6 +1355,9 @@ mt7530_setup(struct dsa_switch *ds)
 				continue;
 
 			phy_node = of_parse_phandle(mac_np, "phy-handle", 0);
+			if (!phy_node)
+				continue;
+
 			if (phy_node->parent == priv->dev->of_node->parent) {
 				ret = of_get_phy_mode(mac_np, &interface);
 				if (ret && ret != -ENODEV)
-- 
2.20.1

[PATCH AUTOSEL 5.5 062/106] tun: Don't put_page() for all negative return values from XDP program

[Sasha Levin]

From: Will Deacon <will@kernel.org>

[ Upstream commit bee348907d19d654e8524d3a946dcd25b693aa7e ]

When an XDP program is installed, tun_build_skb() grabs a reference to
the current page fragment page if the program returns XDP_REDIRECT or
XDP_TX. However, since tun_xdp_act() passes through negative return
values from the XDP program, it is possible to trigger the error path by
mistake and accidentally drop a reference to the fragments page without
taking one, leading to a spurious free. This is believed to be the cause
of some KASAN use-after-free reports from syzbot [1], although without a
reproducer it is not possible to confirm whether this patch fixes the
problem.

Ensure that we only drop a reference to the fragments page if the XDP
transmit or redirect operations actually fail.

[1] https://syzkaller.appspot.com/bug?id=e76a6af1be4acd727ff6bbca669833f98cbf5d95

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
CC: Eric Dumazet <edumazet@google.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Fixes: 8ae1aff0b331 ("tuntap: split out XDP logic")
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/tun.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 35e884a8242d9..6d3317d868d23 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -1715,8 +1715,12 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
 			alloc_frag->offset += buflen;
 		}
 		err = tun_xdp_act(tun, xdp_prog, &xdp, act);
-		if (err < 0)
-			goto err_xdp;
+		if (err < 0) {
+			if (act == XDP_REDIRECT || act == XDP_TX)
+				put_page(alloc_frag->page);
+			goto out;
+		}
+
 		if (err == XDP_REDIRECT)
 			xdp_do_flush_map();
 		if (err != XDP_PASS)
@@ -1730,8 +1734,6 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
 
 	return __tun_build_skb(tfile, alloc_frag, buf, buflen, len, pad);
 
-err_xdp:
-	put_page(alloc_frag->page);
 out:
 	rcu_read_unlock();
 	local_bh_enable();
-- 
2.20.1

[PATCH AUTOSEL 5.5 063/106] mlxsw: spectrum_flower: Do not stop at FLOW_ACTION_VLAN_MANGLE

[Sasha Levin]

From: Petr Machata <petrm@mellanox.com>

[ Upstream commit ccfc569347f870830e7c7cf854679a06cf9c45b5 ]

The handler for FLOW_ACTION_VLAN_MANGLE ends by returning whatever the
lower-level function that it calls returns. If there are more actions lined
up after this action, those are never offloaded. Fix by only bailing out
when the called function returns an error.

Fixes: a150201a70da ("mlxsw: spectrum: Add support for vlan modify TC action")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c
index b607919c8ad02..498de6ef68705 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c
@@ -123,9 +123,12 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp,
 			u8 prio = act->vlan.prio;
 			u16 vid = act->vlan.vid;
 
-			return mlxsw_sp_acl_rulei_act_vlan(mlxsw_sp, rulei,
-							   act->id, vid,
-							   proto, prio, extack);
+			err = mlxsw_sp_acl_rulei_act_vlan(mlxsw_sp, rulei,
+							  act->id, vid,
+							  proto, prio, extack);
+			if (err)
+				return err;
+			break;
 			}
 		default:
 			NL_SET_ERR_MSG_MOD(extack, "Unsupported action");
-- 
2.20.1

[PATCH AUTOSEL 5.5 064/106] net: dsa: bcm_sf2: Do not register slave MDIO bus with OF

[Sasha Levin]

From: Florian Fainelli <f.fainelli@gmail.com>

[ Upstream commit 536fab5bf5826404534a6c271f622ad2930d9119 ]

We were registering our slave MDIO bus with OF and doing so with
assigning the newly created slave_mii_bus of_node to the master MDIO bus
controller node. This is a bad thing to do for a number of reasons:

- we are completely lying about the slave MII bus is arranged and yet we
  still want to control which MDIO devices it probes. It was attempted
  before to play tricks with the bus_mask to perform that:
  https://www.spinics.net/lists/netdev/msg429420.html but the approach
  was rightfully rejected

- the device_node reference counting is messed up and we are effectively
  doing a double probe on the devices we already probed using the
  master, this messes up all resources reference counts (such as clocks)

The proper fix for this as indicated by David in his reply to the
thread above is to use a platform data style registration so as to
control exactly which devices we probe:
https://www.spinics.net/lists/netdev/msg430083.html

By using mdiobus_register(), our slave_mii_bus->phy_mask value is used
as intended, and all the PHY addresses that must be redirected towards
our slave MDIO bus is happening while other addresses get redirected
towards the master MDIO bus.

Fixes: 461cd1b03e32 ("net: dsa: bcm_sf2: Register our slave MDIO bus")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/dsa/bcm_sf2.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index b0f5280a83cb6..4481afd323a97 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -472,7 +472,7 @@ static int bcm_sf2_mdio_register(struct dsa_switch *ds)
 	priv->slave_mii_bus->parent = ds->dev->parent;
 	priv->slave_mii_bus->phy_mask = ~priv->indir_phy_mask;
 
-	err = of_mdiobus_register(priv->slave_mii_bus, dn);
+	err = mdiobus_register(priv->slave_mii_bus);
 	if (err && dn)
 		of_node_put(dn);
 
-- 
2.20.1

[PATCH AUTOSEL 5.5 074/106] net: dsa: bcm_sf2: Ensure correct sub-node is parsed

[Sasha Levin]

From: Florian Fainelli <f.fainelli@gmail.com>

[ Upstream commit afa3b592953bfaecfb4f2f335ec5f935cff56804 ]

When the bcm_sf2 was converted into a proper platform device driver and
used the new dsa_register_switch() interface, we would still be parsing
the legacy DSA node that contained all the port information since the
platform firmware has intentionally maintained backward and forward
compatibility to client programs. Ensure that we do parse the correct
node, which is "ports" per the revised DSA binding.

Fixes: d9338023fb8e ("net: dsa: bcm_sf2: Make it a real platform device driver")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/dsa/bcm_sf2.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 4481afd323a97..e93c81c4062ed 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -1069,6 +1069,7 @@ static int bcm_sf2_sw_probe(struct platform_device *pdev)
 	const struct bcm_sf2_of_data *data;
 	struct b53_platform_data *pdata;
 	struct dsa_switch_ops *ops;
+	struct device_node *ports;
 	struct bcm_sf2_priv *priv;
 	struct b53_device *dev;
 	struct dsa_switch *ds;
@@ -1136,7 +1137,11 @@ static int bcm_sf2_sw_probe(struct platform_device *pdev)
 	set_bit(0, priv->cfp.used);
 	set_bit(0, priv->cfp.unique);
 
-	bcm_sf2_identify_ports(priv, dn->child);
+	ports = of_find_node_by_name(dn, "ports");
+	if (ports) {
+		bcm_sf2_identify_ports(priv, ports);
+		of_node_put(ports);
+	}
 
 	priv->irq0 = irq_of_parse_and_map(dn, 0);
 	priv->irq1 = irq_of_parse_and_map(dn, 1);
-- 
2.20.1

[PATCH AUTOSEL 5.5 076/106] net: phy: micrel: kszphy_resume(): add delay after genphy_resume() before accessing PHY registers

[Sasha Levin]

From: Oleksij Rempel <o.rempel@pengutronix.de>

[ Upstream commit 6110dff776f7fa65c35850ef65b41d3b39e2fac2 ]

After the power-down bit is cleared, the chip internally triggers a
global reset. According to the KSZ9031 documentation, we have to wait at
least 1ms for the reset to finish.

If the chip is accessed during reset, read will return 0xffff, while
write will be ignored. Depending on the system performance and MDIO bus
speed, we may or may not run in to this issue.

This bug was discovered on an iMX6QP system with KSZ9031 PHY and
attached PHY interrupt line. If IRQ was used, the link status update was
lost. In polling mode, the link status update was always correct.

The investigation showed, that during a read-modify-write access, the
read returned 0xffff (while the chip was still in reset) and
corresponding write hit the chip _after_ reset and triggered (due to the
0xffff) another reset in an undocumented bit (register 0x1f, bit 1),
resulting in the next write being lost due to the new reset cycle.

This patch fixes the issue by adding a 1...2 ms sleep after the
genphy_resume().

Fixes: 836384d2501d ("net: phy: micrel: Add specific suspend")
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/phy/micrel.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c
index 63dedec0433de..51b64f0877172 100644
--- a/drivers/net/phy/micrel.c
+++ b/drivers/net/phy/micrel.c
@@ -25,6 +25,7 @@
 #include <linux/micrel_phy.h>
 #include <linux/of.h>
 #include <linux/clk.h>
+#include <linux/delay.h>
 
 /* Operation Mode Strap Override */
 #define MII_KSZPHY_OMSO				0x16
@@ -902,6 +903,12 @@ static int kszphy_resume(struct phy_device *phydev)
 
 	genphy_resume(phydev);
 
+	/* After switching from power-down to normal mode, an internal global
+	 * reset is automatically generated. Wait a minimum of 1 ms before
+	 * read/write access to the PHY registers.
+	 */
+	usleep_range(1000, 2000);
+
 	ret = kszphy_config_reset(phydev);
 	if (ret)
 		return ret;
-- 
2.20.1

[PATCH AUTOSEL 5.5 078/106] net: stmmac: xgmac: Fix VLAN register handling

[Sasha Levin]

From: Jose Abreu <Jose.Abreu@synopsys.com>

[ Upstream commit 21f64e72e7073199a6f8d7d8efe52cd814d7d665 ]

Commit 907a076881f1, forgot that we need to clear old values of
XGMAC_VLAN_TAG register when we switch from VLAN perfect matching to
HASH matching.

Fix it.

Fixes: 907a076881f1 ("net: stmmac: xgmac: fix incorrect XGMAC_VLAN_TAG register writting")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c
index e9bf54a579df6..f0b9c43f6e31e 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c
@@ -576,8 +576,13 @@ static void dwxgmac2_update_vlan_hash(struct mac_device_info *hw, u32 hash,
 			value |= XGMAC_VLAN_EDVLP;
 			value |= XGMAC_VLAN_ESVL;
 			value |= XGMAC_VLAN_DOVLTC;
+		} else {
+			value &= ~XGMAC_VLAN_EDVLP;
+			value &= ~XGMAC_VLAN_ESVL;
+			value &= ~XGMAC_VLAN_DOVLTC;
 		}
 
+		value &= ~XGMAC_VLAN_VID;
 		writel(value, ioaddr + XGMAC_VLAN_TAG);
 	} else if (perfect_match) {
 		u32 value = readl(ioaddr + XGMAC_PACKET_FILTER);
@@ -588,13 +593,19 @@ static void dwxgmac2_update_vlan_hash(struct mac_device_info *hw, u32 hash,
 
 		value = readl(ioaddr + XGMAC_VLAN_TAG);
 
+		value &= ~XGMAC_VLAN_VTHM;
 		value |= XGMAC_VLAN_ETV;
 		if (is_double) {
 			value |= XGMAC_VLAN_EDVLP;
 			value |= XGMAC_VLAN_ESVL;
 			value |= XGMAC_VLAN_DOVLTC;
+		} else {
+			value &= ~XGMAC_VLAN_EDVLP;
+			value &= ~XGMAC_VLAN_ESVL;
+			value &= ~XGMAC_VLAN_DOVLTC;
 		}
 
+		value &= ~XGMAC_VLAN_VID;
 		writel(value | perfect_match, ioaddr + XGMAC_VLAN_TAG);
 	} else {
 		u32 value = readl(ioaddr + XGMAC_PACKET_FILTER);
-- 
2.20.1

[PATCH AUTOSEL 5.5 080/106] cxgb4: free MQPRIO resources in shutdown path

[Sasha Levin]

From: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>

[ Upstream commit cef8dac96bc108633f5090bb3a9988d734dc1ee0 ]

Perform missing MQPRIO resource cleanup in PCI shutdown path. Also,
fix MQPRIO MSIX bitmap leak in resource cleanup.

Fixes: b1396c2bd675 ("cxgb4: parse and configure TC-MQPRIO offload")
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 .../net/ethernet/chelsio/cxgb4/cxgb4_main.c   |  4 ++++
 .../ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c  | 23 +++++++++++++++++++
 .../ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h  |  1 +
 3 files changed, 28 insertions(+)

diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 9934c8b0f0b07..a2fa4b1fd4435 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -6651,6 +6651,10 @@ static void shutdown_one(struct pci_dev *pdev)
 			if (adapter->port[i]->reg_state == NETREG_REGISTERED)
 				cxgb_close(adapter->port[i]);
 
+		rtnl_lock();
+		cxgb4_mqprio_stop_offload(adapter);
+		rtnl_unlock();
+
 		if (is_uld(adapter)) {
 			detach_ulds(adapter);
 			t4_uld_clean_up(adapter);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c
index ec3eb45ee3b48..e6af4906d6743 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c
@@ -301,6 +301,7 @@ static void cxgb4_mqprio_free_hw_resources(struct net_device *dev)
 			cxgb4_clear_msix_aff(eorxq->msix->vec,
 					     eorxq->msix->aff_mask);
 			free_irq(eorxq->msix->vec, &eorxq->rspq);
+			cxgb4_free_msix_idx_in_bmap(adap, eorxq->msix->idx);
 		}
 
 		free_rspq_fl(adap, &eorxq->rspq, &eorxq->fl);
@@ -611,6 +612,28 @@ int cxgb4_setup_tc_mqprio(struct net_device *dev,
 	return ret;
 }
 
+void cxgb4_mqprio_stop_offload(struct adapter *adap)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct net_device *dev;
+	u8 i;
+
+	if (!adap->tc_mqprio || !adap->tc_mqprio->port_mqprio)
+		return;
+
+	for_each_port(adap, i) {
+		dev = adap->port[i];
+		if (!dev)
+			continue;
+
+		tc_port_mqprio = &adap->tc_mqprio->port_mqprio[i];
+		if (!tc_port_mqprio->mqprio.qopt.num_tc)
+			continue;
+
+		cxgb4_mqprio_disable_offload(dev);
+	}
+}
+
 int cxgb4_init_tc_mqprio(struct adapter *adap)
 {
 	struct cxgb4_tc_port_mqprio *tc_port_mqprio, *port_mqprio;
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h
index c532f1ef84517..ff8794132b22b 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h
@@ -38,6 +38,7 @@ struct cxgb4_tc_mqprio {
 
 int cxgb4_setup_tc_mqprio(struct net_device *dev,
 			  struct tc_mqprio_qopt_offload *mqprio);
+void cxgb4_mqprio_stop_offload(struct adapter *adap);
 int cxgb4_init_tc_mqprio(struct adapter *adap);
 void cxgb4_cleanup_tc_mqprio(struct adapter *adap);
 #endif /* __CXGB4_TC_MQPRIO_H__ */
-- 
2.20.1

[PATCH AUTOSEL 5.5 090/106] SUNRPC: fix krb5p mount to provide large enough buffer in rq_rcvsize

[Sasha Levin]

From: Olga Kornievskaia <olga.kornievskaia@gmail.com>

[ Upstream commit df513a7711712758b9cb1a48d86712e7e1ee03f4 ]

Ever since commit 2c94b8eca1a2 ("SUNRPC: Use au_rslack when computing
reply buffer size"). It changed how "req->rq_rcvsize" is calculated. It
used to use au_cslack value which was nice and large and changed it to
au_rslack value which turns out to be too small.

Since 5.1, v3 mount with sec=krb5p fails against an Ontap server
because client's receive buffer it too small.

For gss krb5p, we need to account for the mic token in the verifier,
and the wrap token in the wrap token.

RFC 4121 defines:
mic token
Octet no   Name        Description
         --------------------------------------------------------------
         0..1     TOK_ID     Identification field.  Tokens emitted by
                             GSS_GetMIC() contain the hex value 04 04
                             expressed in big-endian order in this
                             field.
         2        Flags      Attributes field, as described in section
                             4.2.2.
         3..7     Filler     Contains five octets of hex value FF.
         8..15    SND_SEQ    Sequence number field in clear text,
                             expressed in big-endian order.
         16..last SGN_CKSUM  Checksum of the "to-be-signed" data and
                             octet 0..15, as described in section 4.2.4.

that's 16bytes (GSS_KRB5_TOK_HDR_LEN) + chksum

wrap token
Octet no   Name        Description
         --------------------------------------------------------------
          0..1     TOK_ID    Identification field.  Tokens emitted by
                             GSS_Wrap() contain the hex value 05 04
                             expressed in big-endian order in this
                             field.
          2        Flags     Attributes field, as described in section
                             4.2.2.
          3        Filler    Contains the hex value FF.
          4..5     EC        Contains the "extra count" field, in big-
                             endian order as described in section 4.2.3.
          6..7     RRC       Contains the "right rotation count" in big-
                             endian order, as described in section
                             4.2.5.
          8..15    SND_SEQ   Sequence number field in clear text,
                             expressed in big-endian order.
          16..last Data      Encrypted data for Wrap tokens with
                             confidentiality, or plaintext data followed
                             by the checksum for Wrap tokens without
                             confidentiality, as described in section
                             4.2.4.

Also 16bytes of header (GSS_KRB5_TOK_HDR_LEN), encrypted data, and cksum
(other things like padding)

RFC 3961 defines known cksum sizes:
Checksum type              sumtype        checksum         section or
                                value            size         reference
   ---------------------------------------------------------------------
   CRC32                            1               4           6.1.3
   rsa-md4                          2              16           6.1.2
   rsa-md4-des                      3              24           6.2.5
   des-mac                          4              16           6.2.7
   des-mac-k                        5               8           6.2.8
   rsa-md4-des-k                    6              16           6.2.6
   rsa-md5                          7              16           6.1.1
   rsa-md5-des                      8              24           6.2.4
   rsa-md5-des3                     9              24             ??
   sha1 (unkeyed)                  10              20             ??
   hmac-sha1-des3-kd               12              20            6.3
   hmac-sha1-des3                  13              20             ??
   sha1 (unkeyed)                  14              20             ??
   hmac-sha1-96-aes128             15              20         [KRB5-AES]
   hmac-sha1-96-aes256             16              20         [KRB5-AES]
   [reserved]                  0x8003               ?         [GSS-KRB5]

Linux kernel now mainly supports type 15,16 so max cksum size is 20bytes.
(GSS_KRB5_MAX_CKSUM_LEN)

Re-use already existing define of GSS_KRB5_MAX_SLACK_NEEDED that's used
for encoding the gss_wrap tokens (same tokens are used in reply).

Fixes: 2c94b8eca1a2 ("SUNRPC: Use au_rslack when computing reply buffer size")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/sunrpc/auth_gss/auth_gss.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index d75fddca44c94..f9e1a7e61eda0 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -20,6 +20,7 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/auth.h>
 #include <linux/sunrpc/auth_gss.h>
+#include <linux/sunrpc/gss_krb5.h>
 #include <linux/sunrpc/svcauth_gss.h>
 #include <linux/sunrpc/gss_err.h>
 #include <linux/workqueue.h>
@@ -1050,7 +1051,7 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt)
 		goto err_put_mech;
 	auth = &gss_auth->rpc_auth;
 	auth->au_cslack = GSS_CRED_SLACK >> 2;
-	auth->au_rslack = GSS_VERF_SLACK >> 2;
+	auth->au_rslack = GSS_KRB5_MAX_SLACK_NEEDED >> 2;
 	auth->au_verfsize = GSS_VERF_SLACK >> 2;
 	auth->au_ralign = GSS_VERF_SLACK >> 2;
 	auth->au_flags = 0;
-- 
2.20.1

[PATCH AUTOSEL 5.5 098/106] sunrpc: Fix gss_unwrap_resp_integ() again

[Sasha Levin]

From: Chuck Lever <chuck.lever@oracle.com>

[ Upstream commit 4047aa909c4a40fceebc36fff708d465a4d3c6e2 ]

xdr_buf_read_mic() tries to find unused contiguous space in a
received xdr_buf in order to linearize the checksum for the call
to gss_verify_mic. However, the corner cases in this code are
numerous and we seem to keep missing them. I've just hit yet
another buffer overrun related to it.

This overrun is at the end of xdr_buf_read_mic():

1284         if (buf->tail[0].iov_len != 0)
1285                 mic->data = buf->tail[0].iov_base + buf->tail[0].iov_len;
1286         else
1287                 mic->data = buf->head[0].iov_base + buf->head[0].iov_len;
1288         __read_bytes_from_xdr_buf(&subbuf, mic->data, mic->len);
1289         return 0;

This logic assumes the transport has set the length of the tail
based on the size of the received message. base + len is then
supposed to be off the end of the message but still within the
actual buffer.

In fact, the length of the tail is set by the upper layer when the
Call is encoded so that the end of the tail is actually the end of
the allocated buffer itself. This causes the logic above to set
mic->data to point past the end of the receive buffer.

The "mic->data = head" arm of this if statement is no less fragile.

As near as I can tell, this has been a problem forever. I'm not sure
that minimizing au_rslack recently changed this pathology much.

So instead, let's use a more straightforward approach: kmalloc a
separate buffer to linearize the checksum. This is similar to
how gss_validate() currently works.

Coming back to this code, I had some trouble understanding what
was going on. So I've cleaned up the variable naming and added
a few comments that point back to the XDR definition in RFC 2203
to help guide future spelunkers, including myself.

As an added clean up, the functionality that was in
xdr_buf_read_mic() is folded directly into gss_unwrap_resp_integ(),
as that is its only caller.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/sunrpc/auth_gss/auth_gss.c | 77 +++++++++++++++++++++++++---------
 1 file changed, 58 insertions(+), 19 deletions(-)

diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index f9e1a7e61eda0..ff5fcb3e12084 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -1935,35 +1935,69 @@ gss_unwrap_resp_auth(struct rpc_cred *cred)
 	return 0;
 }
 
+/*
+ * RFC 2203, Section 5.3.2.2
+ *
+ *	struct rpc_gss_integ_data {
+ *		opaque databody_integ<>;
+ *		opaque checksum<>;
+ *	};
+ *
+ *	struct rpc_gss_data_t {
+ *		unsigned int seq_num;
+ *		proc_req_arg_t arg;
+ *	};
+ */
 static int
 gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred,
 		      struct gss_cl_ctx *ctx, struct rpc_rqst *rqstp,
 		      struct xdr_stream *xdr)
 {
-	struct xdr_buf integ_buf, *rcv_buf = &rqstp->rq_rcv_buf;
-	u32 data_offset, mic_offset, integ_len, maj_stat;
+	struct xdr_buf gss_data, *rcv_buf = &rqstp->rq_rcv_buf;
 	struct rpc_auth *auth = cred->cr_auth;
+	u32 len, offset, seqno, maj_stat;
 	struct xdr_netobj mic;
-	__be32 *p;
+	int ret;
 
-	p = xdr_inline_decode(xdr, 2 * sizeof(*p));
-	if (unlikely(!p))
+	ret = -EIO;
+	mic.data = NULL;
+
+	/* opaque databody_integ<>; */
+	if (xdr_stream_decode_u32(xdr, &len))
 		goto unwrap_failed;
-	integ_len = be32_to_cpup(p++);
-	if (integ_len & 3)
+	if (len & 3)
 		goto unwrap_failed;
-	data_offset = (u8 *)(p) - (u8 *)rcv_buf->head[0].iov_base;
-	mic_offset = integ_len + data_offset;
-	if (mic_offset > rcv_buf->len)
+	offset = rcv_buf->len - xdr_stream_remaining(xdr);
+	if (xdr_stream_decode_u32(xdr, &seqno))
 		goto unwrap_failed;
-	if (be32_to_cpup(p) != rqstp->rq_seqno)
+	if (seqno != rqstp->rq_seqno)
 		goto bad_seqno;
+	if (xdr_buf_subsegment(rcv_buf, &gss_data, offset, len))
+		goto unwrap_failed;
 
-	if (xdr_buf_subsegment(rcv_buf, &integ_buf, data_offset, integ_len))
+	/*
+	 * The xdr_stream now points to the beginning of the
+	 * upper layer payload, to be passed below to
+	 * rpcauth_unwrap_resp_decode(). The checksum, which
+	 * follows the upper layer payload in @rcv_buf, is
+	 * located and parsed without updating the xdr_stream.
+	 */
+
+	/* opaque checksum<>; */
+	offset += len;
+	if (xdr_decode_word(rcv_buf, offset, &len))
+		goto unwrap_failed;
+	offset += sizeof(__be32);
+	if (offset + len > rcv_buf->len)
 		goto unwrap_failed;
-	if (xdr_buf_read_mic(rcv_buf, &mic, mic_offset))
+	mic.len = len;
+	mic.data = kmalloc(len, GFP_NOFS);
+	if (!mic.data)
+		goto unwrap_failed;
+	if (read_bytes_from_xdr_buf(rcv_buf, offset, mic.data, mic.len))
 		goto unwrap_failed;
-	maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &integ_buf, &mic);
+
+	maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &gss_data, &mic);
 	if (maj_stat == GSS_S_CONTEXT_EXPIRED)
 		clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags);
 	if (maj_stat != GSS_S_COMPLETE)
@@ -1971,16 +2005,21 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred,
 
 	auth->au_rslack = auth->au_verfsize + 2 + 1 + XDR_QUADLEN(mic.len);
 	auth->au_ralign = auth->au_verfsize + 2;
-	return 0;
+	ret = 0;
+
+out:
+	kfree(mic.data);
+	return ret;
+
 unwrap_failed:
 	trace_rpcgss_unwrap_failed(task);
-	return -EIO;
+	goto out;
 bad_seqno:
-	trace_rpcgss_bad_seqno(task, rqstp->rq_seqno, be32_to_cpup(p));
-	return -EIO;
+	trace_rpcgss_bad_seqno(task, rqstp->rq_seqno, seqno);
+	goto out;
 bad_mic:
 	trace_rpcgss_verify_mic(task, maj_stat);
-	return -EIO;
+	goto out;
 }
 
 static int
-- 
2.20.1