Commits · db849faa9bef993a1379dc510623f750a72fa7ce · Kirill Smelkov / linux

11 Jul, 2019 3 commits

net/mlx5e: Rx, Fix checksum calculation for new hardware · db849faa

Saeed Mahameed authored May 03, 2019

CQE checksum full mode in new HW, provides a full checksum of rx frame.
Covering bytes starting from eth protocol up to last byte in the received
frame (frame_size - ETH_HLEN), as expected by the stack.

Fixing up skb->csum by the driver is not required in such case. This fix
is to avoid wrong checksum calculation in drivers which already support
the new hardware with the new checksum mode.

Fixes: 85327a9c ("net/mlx5: Update the list of the PCI supported devices")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

db849faa

net/mlx5e: Fix port tunnel GRE entropy control · 914adbb1

Eli Britstein authored Jun 02, 2019

GRE entropy calculation is a single bit per card, and not per port.
Force disable GRE entropy calculation upon the first GRE encap rule,
and release the force at the last GRE encap rule removal. This is done
per port.

Fixes: 97417f61 ("net/mlx5e: Fix GRE key by controlling port tunnel entropy calculation")
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

914adbb1

net/mlx5: E-Switch, Fix default encap mode · 9a64144d

Maor Gottlieb authored Jun 17, 2019

Encap mode is related to switchdev mode only. Move the init of
the encap mode to eswitch_offloads. Before this change, we reported
that eswitch supports encap, even tough the device was in non
SRIOV mode.

Fixes: 7768d197 ('net/mlx5: E-Switch, Add control for encapsulation')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

9a64144d

09 Jul, 2019 1 commit

tcp: Reset bytes_acked and bytes_received when disconnecting · e858faf5

Christoph Paasch authored Jul 06, 2019

If an app is playing tricks to reuse a socket via tcp_disconnect(),
bytes_acked/received needs to be reset to 0. Otherwise tcp_info will
report the sum of the current and the old connection..

Cc: Eric Dumazet <edumazet@google.com>
Fixes: 0df48c26 ("tcp: add tcpi_bytes_acked to tcp_info")
Fixes: bdd1f9ed ("tcp: add tcpi_bytes_received to tcp_info")
Signed-off-by: Christoph Paasch <cpaasch@apple.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

e858faf5

08 Jul, 2019 13 commits

selftests: txring_overwrite: fix incorrect test of mmap() return value · cecaa76b

Frank de Brabander authored Jul 05, 2019

If mmap() fails it returns MAP_FAILED, which is defined as ((void *) -1).
The current if-statement incorrectly tests if *ring is NULL.

Fixes: 358be656 ("selftests/net: add txring_overwrite")
Signed-off-by: Frank de Brabander <debrabander@gmail.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

cecaa76b

nfc: fix potential illegal memory access · dd006fc4

Yang Wei authored Jul 08, 2019

The frags_q is not properly initialized, it may result in illegal memory
access when conn_info is NULL.
The "goto free_exit" should be replaced by "goto exit".
Signed-off-by: Yang Wei <albin_yang@163.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

dd006fc4

net: stmmac: Re-work the queue selection for TSO packets · 4993e5b3

Jose Abreu authored Jul 08, 2019

Ben Hutchings says:
	"This is the wrong place to change the queue mapping.
	stmmac_xmit() is called with a specific TX queue locked,
	and accessing a different TX queue results in a data race
	for all of that queue's state.

	I think this commit should be reverted upstream and in all
	stable branches.  Instead, the driver should implement the
	ndo_select_queue operation and override the queue mapping there."

Fixes: c5acdbee ("net: stmmac: Send TSO packets always from Queue 0")
Suggested-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jose Abreu <joabreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

4993e5b3

net: bpfilter: print umh messages to /dev/kmsg · 36c4357c

Gary Lin authored Jul 05, 2019

bpfilter_umh currently printed all messages to /dev/console and this
might interfere the user activity(*).

This commit changes the output device to /dev/kmsg so that the messages
from bpfilter_umh won't show on the console directly.

(*) https://bugzilla.suse.com/show_bug.cgi?id=1140221Signed-off-by: Gary Lin <glin@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

36c4357c

net: nixge: Fix misuse of strlcpy · 6b4ddf99

Joe Perches authored Jul 04, 2019

Probable cut&paste typo - use the correct field size.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

6b4ddf99

net: ethernet: sun4i-emac: Fix misuse of strlcpy · 7391324b

Joe Perches authored Jul 04, 2019

Probable cut&paste typo - use the correct field size.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7391324b

Merge branch 'gtp-fix-several-bugs' · 07996783

David S. Miller authored Jul 07, 2019

Taehee Yoo says:

====================
gtp: fix several bugs

This patch series fixes several bugs in the gtp module.

First patch fixes suspicious RCU usage.
The problem is to use rcu_dereference_sk_user_data() outside of
RCU read critical section.

Second patch fixes use-after-free.
gtp_encap_destroy() is called twice.
gtp_encap_destroy() use both gtp->sk0 and gtp->sk1u.
these pointers can be freed in gtp_encap_destroy().
So, gtp_encap_destroy() should avoid using freed sk pointer.

Third patch removes duplicate code in gtp_dellink().
gtp_dellink() calls gtp_encap_disable() twice.
So, remove one of them.

Fourth patch fixes usage of GFP_KERNEL.
GFP_KERNEL can not be used in RCU read critical section.
This patch make ipv4_pdp_add() to use GFP_ATOMIC instead of GFP_KERNEL.

Fifth patch fixes use-after-free in gtp_newlink().
gtp_newlink() uses gtp_net which would be destroyed by the __exit_net
routine.
So, gtp_newlink should not be called after the __exit_net routine.

Sixth patch adds missing error handling routine in gtp_encap_enable().
gtp_encap_enable() will fail, if invalid role value is sent from
user-space. if so, gtp_encap_enable() should execute error handling
routine.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

07996783