]> git.proxmox.com Git - mirror_ubuntu-kernels.git/log
mirror_ubuntu-kernels.git
3 years agomath: Export mul_u64_u64_div_u64
David S. Miller [Wed, 24 Mar 2021 23:42:54 +0000 (16:42 -0700)]
math: Export mul_u64_u64_div_u64

Fixes: f51d7bf1dbe5 ("ptp_qoriq: fix overflow in ptp_qoriq_adjfine() u64 calcalation")
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoch_ktls: fix enum-conversion warning
Arnd Bergmann [Tue, 23 Mar 2021 21:52:50 +0000 (22:52 +0100)]
ch_ktls: fix enum-conversion warning

gcc points out an incorrect enum assignment:

drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c: In function 'chcr_ktls_cpl_set_tcb_rpl':
drivers/net/ethernet/chelsio/inline_crypto/ch_ktls/chcr_ktls.c:684:22: warning: implicit conversion from 'enum <anonymous>' to 'enum ch_ktls_open_state' [-Wenum-conversion]

This appears harmless, and should apparently use 'CH_KTLS_OPEN_SUCCESS'
instead of 'false', with the same value '0'.

Fixes: efca3878a5fb ("ch_ktls: Issue if connection offload fails")
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Fix memory leak of object buf
Colin Ian King [Tue, 23 Mar 2021 12:32:45 +0000 (12:32 +0000)]
octeontx2-af: Fix memory leak of object buf

Currently the error return path when lfs fails to allocate is not free'ing
the memory allocated to buf. Fix this by adding the missing kfree.

Addresses-Coverity: ("Resource leak")
Fixes: f7884097141b ("octeontx2-af: Formatting debugfs entry rsrc_alloc.")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoptp_qoriq: fix overflow in ptp_qoriq_adjfine() u64 calcalation
Yangbo Lu [Tue, 23 Mar 2021 08:02:29 +0000 (16:02 +0800)]
ptp_qoriq: fix overflow in ptp_qoriq_adjfine() u64 calcalation

Current calculation for diff of TMR_ADD register value may have
64-bit overflow in this code line, when long type scaled_ppm is
large.

adj *= scaled_ppm;

This patch is to resolve it by using mul_u64_u64_div_u64().

Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: bridge: don't notify switchdev for local FDB addresses
Vladimir Oltean [Mon, 22 Mar 2021 18:21:08 +0000 (20:21 +0200)]
net: bridge: don't notify switchdev for local FDB addresses

As explained in this discussion:
https://lore.kernel.org/netdev/20210117193009.io3nungdwuzmo5f7@skbuf/

the switchdev notifiers for FDB entries managed to have a zero-day bug.
The bridge would not say that this entry is local:

ip link add br0 type bridge
ip link set swp0 master br0
bridge fdb add dev swp0 00:01:02:03:04:05 master local

and the switchdev driver would be more than happy to offload it as a
normal static FDB entry. This is despite the fact that 'local' and
non-'local' entries have completely opposite directions: a local entry
is locally terminated and not forwarded, whereas a static entry is
forwarded and not locally terminated. So, for example, DSA would install
this entry on swp0 instead of installing it on the CPU port as it should.

There is an even sadder part, which is that the 'local' flag is implicit
if 'static' is not specified, meaning that this command produces the
same result of adding a 'local' entry:

bridge fdb add dev swp0 00:01:02:03:04:05 master

I've updated the man pages for 'bridge', and after reading it now, it
should be pretty clear to any user that the commands above were broken
and should have never resulted in the 00:01:02:03:04:05 address being
forwarded (this behavior is coherent with non-switchdev interfaces):
https://patchwork.kernel.org/project/netdevbpf/cover/20210211104502.2081443-1-olteanv@gmail.com/
If you're a user reading this and this is what you want, just use:

bridge fdb add dev swp0 00:01:02:03:04:05 master static

Because switchdev should have given drivers the means from day one to
classify FDB entries as local/non-local, but didn't, it means that all
drivers are currently broken. So we can just as well omit the switchdev
notifications for local FDB entries, which is exactly what this patch
does to close the bug in stable trees. For further development work
where drivers might want to trap the local FDB entries to the host, we
can add a 'bool is_local' to br_switchdev_fdb_call_notifiers(), and
selectively make drivers act upon that bit, while all the others ignore
those entries if the 'is_local' bit is set.

Fixes: 6b26b51b1d13 ("net: bridge: Add support for notifying devices about FDB add/del")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/sched: act_ct: clear post_ct if doing ct_clear
Marcelo Ricardo Leitner [Mon, 22 Mar 2021 18:13:22 +0000 (15:13 -0300)]
net/sched: act_ct: clear post_ct if doing ct_clear

Invalid detection works with two distinct moments: act_ct tries to find
a conntrack entry and set post_ct true, indicating that that was
attempted. Then, when flow dissector tries to dissect CT info and no
entry is there, it knows that it was tried and no entry was found, and
synthesizes/sets
                  key->ct_state = TCA_FLOWER_KEY_CT_FLAGS_TRACKED |
                                  TCA_FLOWER_KEY_CT_FLAGS_INVALID;
mimicing what OVS does.

OVS has this a bit more streamlined, as it recomputes the key after
trying to find a conntrack entry for it.

Issue here is, when we have 'tc action ct clear', it didn't clear
post_ct, causing a subsequent match on 'ct_state -trk' to fail, due to
the above. The fix, thus, is to clear it.

Reproducer rules:
tc filter add dev enp130s0f0np0_0 ingress prio 1 chain 0 \
protocol ip flower ip_proto tcp ct_state -trk \
action ct zone 1 pipe \
action goto chain 2
tc filter add dev enp130s0f0np0_0 ingress prio 1 chain 2 \
protocol ip flower \
action ct clear pipe \
action goto chain 4
tc filter add dev enp130s0f0np0_0 ingress prio 1 chain 4 \
protocol ip flower ct_state -trk \
action mirred egress redirect dev enp130s0f1np1_0

With the fix, the 3rd rule matches, like it does with OVS kernel
datapath.

Fixes: 7baf2429a1a9 ("net/sched: cls_flower add CT_FLAGS_INVALID flag support")
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Reviewed-by: wenxu <wenxu@ucloud.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dsa: don't assign an error value to tag_ops
George McCollister [Mon, 22 Mar 2021 20:26:50 +0000 (15:26 -0500)]
net: dsa: don't assign an error value to tag_ops

Use a temporary variable to hold the return value from
dsa_tag_driver_get() instead of assigning it to dst->tag_ops. Leaving
an error value in dst->tag_ops can result in deferencing an invalid
pointer when a deferred switch configuration happens later.

Fixes: 357f203bb3b5 ("net: dsa: keep a copy of the tagging protocol in the DSA switch tree")
Signed-off-by: George McCollister <george.mccollister@gmail.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge tag 'mlx5-fixes-2021-03-22' of git://git.kernel.org/pub/scm/linux/kernel/git...
David S. Miller [Tue, 23 Mar 2021 00:00:48 +0000 (17:00 -0700)]
Merge tag 'mlx5-fixes-2021-03-22' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5 fixes 2021-03-22

This series introduces some fixes to mlx5 driver.
Please pull and let me know if there is any problem.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoisdn: capi: fix mismatched prototypes
Arnd Bergmann [Mon, 22 Mar 2021 16:44:29 +0000 (17:44 +0100)]
isdn: capi: fix mismatched prototypes

gcc-11 complains about a prototype declaration that is different
from the function definition:

drivers/isdn/capi/kcapi.c:724:44: error: argument 2 of type ‘u8 *’ {aka ‘unsigned char *’} declared as a pointer [-Werror=array-parameter=]
  724 | u16 capi20_get_manufacturer(u32 contr, u8 *buf)
      |                                        ~~~~^~~
In file included from drivers/isdn/capi/kcapi.c:13:
drivers/isdn/capi/kcapi.h:62:43: note: previously declared as an array ‘u8[64]’ {aka ‘unsigned char[64]’}
   62 | u16 capi20_get_manufacturer(u32 contr, u8 buf[CAPI_MANUFACTURER_LEN]);
      |                                        ~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/isdn/capi/kcapi.c:790:38: error: argument 2 of type ‘u8 *’ {aka ‘unsigned char *’} declared as a pointer [-Werror=array-parameter=]
  790 | u16 capi20_get_serial(u32 contr, u8 *serial)
      |                                  ~~~~^~~~~~
In file included from drivers/isdn/capi/kcapi.c:13:
drivers/isdn/capi/kcapi.h:64:37: note: previously declared as an array ‘u8[8]’ {aka ‘unsigned char[8]’}
   64 | u16 capi20_get_serial(u32 contr, u8 serial[CAPI_SERIAL_LEN]);
      |                                  ~~~^~~~~~~~~~~~~~~~~~~~~~~

Change the definition to make them match.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/mlx5: SF, do not use ecpu bit for vhca state processing
Parav Pandit [Thu, 11 Mar 2021 18:19:25 +0000 (20:19 +0200)]
net/mlx5: SF, do not use ecpu bit for vhca state processing

Device firmware doesn't handle ecpu bit for vhca state processing
events and commands. Instead device firmware refers to the unique
function id to distinguish SF of different PCI functions.

When ecpu bit is used, firmware returns a syndrome.

mlx5_cmd_check:780:(pid 872): MODIFY_VHCA_STATE(0xb0e) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0x263211)
mlx5_sf_dev_table_create:248:(pid 872): SF DEV table create err = -22

Hence, avoid using ecpu bit.

Fixes: 8f0105418668 ("net/mlx5: SF, Add port add delete functionality")
Fixes: 90d010b8634b ("net/mlx5: SF, Add auxiliary device support")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Vu Pham <vuhuong@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Fix division by 0 in mlx5e_select_queue
Maxim Mikityanskiy [Mon, 1 Mar 2021 18:59:59 +0000 (20:59 +0200)]
net/mlx5e: Fix division by 0 in mlx5e_select_queue

mlx5e_select_queue compares num_tc_x_num_ch to real_num_tx_queues to
determine if HTB and/or PTP offloads are active. If they are, it
calculates netdev_pick_tx() % num_tc_x_num_ch to prevent it from
selecting HTB and PTP queues for regular traffic. However, before the
channels are first activated, num_tc_x_num_ch is zero. If
ndo_select_queue gets called at this point, the HTB/PTP check will pass,
and mlx5e_select_queue will attempt to take a modulo by num_tc_x_num_ch,
which equals to zero.

This commit fixes the bug by assigning num_tc_x_num_ch to a non-zero
value before registering the netdev.

Fixes: 214baf22870c ("net/mlx5e: Support HTB offload")
Reported-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Fix error path for ethtool set-priv-flag
Aya Levin [Thu, 11 Mar 2021 15:46:35 +0000 (17:46 +0200)]
net/mlx5e: Fix error path for ethtool set-priv-flag

Expose error value when failing to comply to command:
$ ethtool --set-priv-flags eth2 rx_cqe_compress [on/off]

Fixes: be7e87f92b58 ("net/mlx5e: Fail safe cqe compressing/moderation mode setting")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Offload tuple rewrite for non-CT flows
Dima Chumak [Thu, 4 Mar 2021 19:28:11 +0000 (21:28 +0200)]
net/mlx5e: Offload tuple rewrite for non-CT flows

Setting connection tracking OVS flows and then setting non-CT flows that
use tuple rewrite action (e.g. mod_tp_dst), causes the latter flows not
being offloaded.

Fix by using a stricter condition in modify_header_match_supported() to
check tuple rewrite support only for flows with CT action. The check is
factored out into standalone modify_tuple_supported() function to aid
readability.

Fixes: 7e36feeb0467 ("net/mlx5e: CT: Don't offload tuple rewrites for established tuples")
Signed-off-by: Dima Chumak <dchumak@nvidia.com>
Reviewed-by: Paul Blakey <paulb@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Allow to match on MPLS parameters only for MPLS over UDP
Alaa Hleihel [Wed, 10 Mar 2021 15:01:46 +0000 (17:01 +0200)]
net/mlx5e: Allow to match on MPLS parameters only for MPLS over UDP

Currently, we support hardware offload only for MPLS over UDP.
However, rules matching on MPLS parameters are now wrongly offloaded
for regular MPLS, without actually taking the parameters into
consideration when doing the offload.
Fix it by rejecting such unsupported rules.

Fixes: 72046a91d134 ("net/mlx5e: Allow to match on mpls parameters")
Signed-off-by: Alaa Hleihel <alaa@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: Add back multicast stats for uplink representor
Huy Nguyen [Fri, 19 Mar 2021 01:33:19 +0000 (20:33 -0500)]
net/mlx5: Add back multicast stats for uplink representor

The multicast counter got removed from uplink representor due to the
cited patch.

Fixes: 47c97e6b10a1 ("net/mlx5e: Fix multicast counter not up-to-date in "ip -s"")
Signed-off-by: Huy Nguyen <huyn@nvidia.com>
Reviewed-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet: ipconfig: ic_dev can be NULL in ic_close_devs
Vladimir Oltean [Mon, 22 Mar 2021 00:26:37 +0000 (02:26 +0200)]
net: ipconfig: ic_dev can be NULL in ic_close_devs

ic_close_dev contains a generalization of the logic to not close a
network interface if it's the host port for a DSA switch. This logic is
disguised behind an iteration through the lowers of ic_dev in
ic_close_dev.

When no interface for ipconfig can be found, ic_dev is NULL, and
ic_close_dev:
- dereferences a NULL pointer when assigning selected_dev
- would attempt to search through the lower interfaces of a NULL
  net_device pointer

So we should protect against that case.

The "lower_dev" iterator variable was shortened to "lower" in order to
keep the 80 character limit.

Fixes: f68cbaed67cb ("net: ipconfig: avoid use-after-free in ic_close_devs")
Fixes: 46acf7bdbc72 ("Revert "net: ipv4: handle DSA enabled master network devices"")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Heiko Thiery <heiko.thiery@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMAINTAINERS: Combine "QLOGIC QLGE 10Gb ETHERNET DRIVER" sections into one
Jonathan Neuschäfer [Sun, 21 Mar 2021 13:00:01 +0000 (14:00 +0100)]
MAINTAINERS: Combine "QLOGIC QLGE 10Gb ETHERNET DRIVER" sections into one

There ended up being two sections with the same title. Combine the two
into one section.

Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Cc: Manish Chopra <manishc@marvell.com>
Cc: Coiby Xu <coiby.xu@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodocs: networking: Fix a typo
Bhaskar Chowdhury [Sat, 20 Mar 2021 21:07:03 +0000 (02:37 +0530)]
docs: networking: Fix a typo

s/subsytem/subsystem/

Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agor8169: fix DMA being used after buffer free if WoL is enabled
Heiner Kallweit [Sat, 20 Mar 2021 20:40:08 +0000 (21:40 +0100)]
r8169: fix DMA being used after buffer free if WoL is enabled

IOMMU errors have been reported if WoL is enabled and interface is
brought down. It turned out that the network chip triggers DMA
transfers after the DMA buffers have been freed. For WoL to work we
need to leave rx enabled, therefore simply stop the chip from being
a DMA busmaster.

Fixes: 567ca57faa62 ("r8169: add rtl8169_up")
Tested-by: Paul Blazejowski <paulb@blazebox.homeip.net>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge tag 'linux-can-fixes-for-5.12-20210320' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Sun, 21 Mar 2021 01:58:56 +0000 (18:58 -0700)]
Merge tag 'linux-can-fixes-for-5.12-20210320' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can

Marc Kleine-Budde says:

====================
pull-request: can 2021-03-20

this is a pull request of 2 patches for net/master.

The first patch is by Oliver Hartkopp. He fixes the TX-path in the
ISO-TP protocol by properly initializing the outgoing CAN frames.

The second patch is by me and reverts a patch from my previous pull
request which added MODULE_SUPPORTED_DEVICE to the peak_usb driver. In
the mean time in Linus's tree the entirely MODULE_SUPPORTED_DEVICE was
removed. So this reverts the adding of the new MODULE_SUPPORTED_DEVICE
to avoid the merge conflict.

If you prefer to resolve the merge conflict by hand, I'll send a new
pull request without that patch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'pa-fox-validation'
David S. Miller [Sun, 21 Mar 2021 01:53:41 +0000 (18:53 -0700)]
Merge branch 'pa-fox-validation'

Alex Elder says:

====================
ipa: fix validation

There is sanity checking code in the IPA driver that's meant to be
enabled only during development.  This allows the driver to make
certain assumptions, but not have to verify those assumptions are
true at (operational) runtime.  This code is built conditional on
IPA_VALIDATION, set (if desired) inside the IPA makefile.

Unfortunately, this validation code has some errors.  First, there
are some mismatched arguments supplied to some dev_err() calls in
ipa_cmd_table_valid() and ipa_cmd_header_valid(), and these are
exposed if validation is enabled.  Second, the tag that enables
this conditional code isn't used consistently (it's IPA_VALIDATE
in some spots and IPA_VALIDATION in others).

This series fixes those two problems with the conditional validation
code.

Version 2 removes the two patches that introduced ipa_assert().  It
also modifies the description in the first patch so that it mentions
the changes made to ipa_cmd_table_valid().
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ipa: fix init header command validation
Alex Elder [Sat, 20 Mar 2021 14:17:28 +0000 (09:17 -0500)]
net: ipa: fix init header command validation

We use ipa_cmd_header_valid() to ensure certain values we will
program into hardware are within range, well in advance of when we
actually program them.  This way we avoid having to check for errors
when we actually program the hardware.

Unfortunately the dev_err() call for a bad offset value does not
supply the arguments to match the format specifiers properly.
Fix this.

There was also supposed to be a check to ensure the size to be
programmed fits in the field that holds it.  Add this missing check.

Rearrange the way we ensure the header table fits in overall IPA
memory range.

Finally, update ipa_cmd_table_valid() so the format of messages
printed for errors matches what's done in ipa_cmd_header_valid().

Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
David S. Miller [Sat, 20 Mar 2021 22:18:06 +0000 (15:18 -0700)]
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

Alexei Starovoitov says:

====================
pull-request: bpf 2021-03-20

The following pull-request contains BPF updates for your *net* tree.

We've added 5 non-merge commits during the last 3 day(s) which contain
a total of 8 files changed, 155 insertions(+), 12 deletions(-).

The main changes are:

1) Use correct nops in fexit trampoline, from Stanislav.

2) Fix BTF dump, from Jean-Philippe.

3) Fix umd memory leak, from Zqiang.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agocan: peak_usb: Revert "can: peak_usb: add forgotten supported devices"
Marc Kleine-Budde [Sat, 20 Mar 2021 19:21:54 +0000 (20:21 +0100)]
can: peak_usb: Revert "can: peak_usb: add forgotten supported devices"

In commit 6417f03132a6 ("module: remove never implemented
MODULE_SUPPORTED_DEVICE") the MODULE_SUPPORTED_DEVICE macro was
removed from the kerne entirely. Shortly before this patch was applied
mainline the commit 59ec7b89ed3e ("can: peak_usb: add forgotten
supported devices") was added to net/master. As this would result in a
merge conflict, let's revert this patch.

Fixes: 59ec7b89ed3e ("can: peak_usb: add forgotten supported devices")
Link: https://lore.kernel.org/r/20210320192649.341832-1-mkl@pengutronix.de
Suggested-by: Leon Romanovsky <leon@kernel.org>
Cc: Stephane Grosjean <s.grosjean@peak-system.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: isotp: tx-path: zero initialize outgoing CAN frames
Oliver Hartkopp [Fri, 19 Mar 2021 10:06:19 +0000 (11:06 +0100)]
can: isotp: tx-path: zero initialize outgoing CAN frames

Commit d4eb538e1f48 ("can: isotp: TX-path: ensure that CAN frame flags are
initialized") ensured the TX flags to be properly set for outgoing CAN
frames.

In fact the root cause of the issue results from a missing initialization
of outgoing CAN frames created by isotp. This is no problem on the CAN bus
as the CAN driver only picks the correctly defined content from the struct
can(fd)_frame. But when the outgoing frames are monitored (e.g. with
candump) we potentially leak some bytes in the unused content of
struct can(fd)_frame.

Fixes: e057dd3fc20f ("can: add ISO 15765-2:2016 transport protocol")
Cc: Marc Kleine-Budde <mkl@pengutronix.de>
Link: https://lore.kernel.org/r/20210319100619.10858-1-socketcan@hartkopp.net
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agobpf: Use NOP_ATOMIC5 instead of emit_nops(&prog, 5) for BPF_TRAMP_F_CALL_ORIG
Stanislav Fomichev [Sat, 20 Mar 2021 00:00:01 +0000 (17:00 -0700)]
bpf: Use NOP_ATOMIC5 instead of emit_nops(&prog, 5) for BPF_TRAMP_F_CALL_ORIG

__bpf_arch_text_poke does rewrite only for atomic nop5, emit_nops(xxx, 5)
emits non-atomic one which breaks fentry/fexit with k8 atomics:

P6_NOP5 == P6_NOP5_ATOMIC (0f1f440000 == 0f1f440000)
K8_NOP5 != K8_NOP5_ATOMIC (6666906690 != 6666666690)

Can be reproduced by doing "ideal_nops = k8_nops" in "arch_init_ideal_nops()
and running fexit_bpf2bpf selftest.

Fixes: e21aa341785c ("bpf: Fix fexit trampoline.")
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210320000001.915366-1-sdf@google.com
3 years agobpf: Fix umd memory leak in copy_process()
Zqiang [Wed, 17 Mar 2021 03:09:15 +0000 (11:09 +0800)]
bpf: Fix umd memory leak in copy_process()

The syzbot reported a memleak as follows:

BUG: memory leak
unreferenced object 0xffff888101b41d00 (size 120):
  comm "kworker/u4:0", pid 8, jiffies 4294944270 (age 12.780s)
  backtrace:
    [<ffffffff8125dc56>] alloc_pid+0x66/0x560
    [<ffffffff81226405>] copy_process+0x1465/0x25e0
    [<ffffffff81227943>] kernel_clone+0xf3/0x670
    [<ffffffff812281a1>] kernel_thread+0x61/0x80
    [<ffffffff81253464>] call_usermodehelper_exec_work
    [<ffffffff81253464>] call_usermodehelper_exec_work+0xc4/0x120
    [<ffffffff812591c9>] process_one_work+0x2c9/0x600
    [<ffffffff81259ab9>] worker_thread+0x59/0x5d0
    [<ffffffff812611c8>] kthread+0x178/0x1b0
    [<ffffffff8100227f>] ret_from_fork+0x1f/0x30

unreferenced object 0xffff888110ef5c00 (size 232):
  comm "kworker/u4:0", pid 8414, jiffies 4294944270 (age 12.780s)
  backtrace:
    [<ffffffff8154a0cf>] kmem_cache_zalloc
    [<ffffffff8154a0cf>] __alloc_file+0x1f/0xf0
    [<ffffffff8154a809>] alloc_empty_file+0x69/0x120
    [<ffffffff8154a8f3>] alloc_file+0x33/0x1b0
    [<ffffffff8154ab22>] alloc_file_pseudo+0xb2/0x140
    [<ffffffff81559218>] create_pipe_files+0x138/0x2e0
    [<ffffffff8126c793>] umd_setup+0x33/0x220
    [<ffffffff81253574>] call_usermodehelper_exec_async+0xb4/0x1b0
    [<ffffffff8100227f>] ret_from_fork+0x1f/0x30

After the UMD process exits, the pipe_to_umh/pipe_from_umh and
tgid need to be released.

Fixes: d71fa5c9763c ("bpf: Add kernel module with user mode driver that populates bpffs.")
Reported-by: syzbot+44908bb56d2bfe56b28e@syzkaller.appspotmail.com
Signed-off-by: Zqiang <qiang.zhang@windriver.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210317030915.2865-1-qiang.zhang@windriver.com
3 years agoMerge branch 'libbpf: Fix BTF dump of pointer-to-array-of-struct'
Andrii Nakryiko [Fri, 19 Mar 2021 21:14:44 +0000 (14:14 -0700)]
Merge branch 'libbpf: Fix BTF dump of pointer-to-array-of-struct'

Jean-Philippe Brucker says:

====================

Fix an issue with the libbpf BTF dump, see patch 1 for details.

Since [v1] I added the selftest in patch 2, though I couldn't figure out
a way to make it independent from the order in which debug info is
issued by the compiler.

[v1]: https://lore.kernel.org/bpf/20210318122700.396574-1-jean-philippe@linaro.org/
====================

Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
3 years agoselftests/bpf: Add selftest for pointer-to-array-of-struct BTF dump
Jean-Philippe Brucker [Fri, 19 Mar 2021 11:25:55 +0000 (12:25 +0100)]
selftests/bpf: Add selftest for pointer-to-array-of-struct BTF dump

Bpftool used to issue forward declarations for a struct used as part of
a pointer to array, which is invalid. Add a test to check that the
struct is fully defined in this case:

@@ -134,9 +134,9 @@
  };
 };

-struct struct_in_array {};
+struct struct_in_array;

-struct struct_in_array_typed {};
+struct struct_in_array_typed;

 typedef struct struct_in_array_typed struct_in_array_t[2];

@@ -189,3 +189,7 @@
  struct struct_with_embedded_stuff _14;
 };

+struct struct_in_array {};
+
+struct struct_in_array_typed {};
+
...
#13/1 btf_dump: syntax:FAIL

Suggested-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210319112554.794552-3-jean-philippe@linaro.org
3 years agolibbpf: Fix BTF dump of pointer-to-array-of-struct
Jean-Philippe Brucker [Fri, 19 Mar 2021 11:25:54 +0000 (12:25 +0100)]
libbpf: Fix BTF dump of pointer-to-array-of-struct

The vmlinux.h generated from BTF is invalid when building
drivers/phy/ti/phy-gmii-sel.c with clang:

vmlinux.h:61702:27: error: array type has incomplete element type ‘struct reg_field’
61702 |  const struct reg_field (*regfields)[3];
      |                           ^~~~~~~~~

bpftool generates a forward declaration for this struct regfield, which
compilers aren't happy about. Here's a simplified reproducer:

struct inner {
int val;
};
struct outer {
struct inner (*ptr_to_array)[2];
} A;

After build with clang -> bpftool btf dump c -> clang/gcc:
./def-clang.h:11:23: error: array has incomplete element type 'struct inner'
        struct inner (*ptr_to_array)[2];

Member ptr_to_array of struct outer is a pointer to an array of struct
inner. In the DWARF generated by clang, struct outer appears before
struct inner, so when converting BTF of struct outer into C, bpftool
issues a forward declaration to struct inner. With GCC the DWARF info is
reversed so struct inner gets fully defined.

That forward declaration is not sufficient when compilers handle an
array of the struct, even when it's only used through a pointer. Note
that we can trigger the same issue with an intermediate typedef:

struct inner {
        int val;
};
typedef struct inner inner2_t[2];
struct outer {
        inner2_t *ptr_to_array;
} A;

Becomes:

struct inner;
typedef struct inner inner2_t[2];

And causes:

./def-clang.h:10:30: error: array has incomplete element type 'struct inner'
typedef struct inner inner2_t[2];

To fix this, clear through_ptr whenever we encounter an intermediate
array, to make the inner struct part of a strong link and force full
declaration.

Fixes: 351131b51c7a ("libbpf: add btf_dump API for BTF-to-C conversion")
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210319112554.794552-2-jean-philippe@linaro.org
3 years agoselftests: forwarding: vxlan_bridge_1d: Fix vxlan ecn decapsulate value
Hangbin Liu [Fri, 19 Mar 2021 14:33:14 +0000 (22:33 +0800)]
selftests: forwarding: vxlan_bridge_1d: Fix vxlan ecn decapsulate value

The ECN bit defines ECT(1) = 1, ECT(0) = 2. So inner 0x02 + outer 0x01
should be inner ECT(0) + outer ECT(1). Based on the description of
__INET_ECN_decapsulate, the final decapsulate value should be
ECT(1). So fix the test expect value to 0x01.

Before the fix:
TEST: VXLAN: ECN decap: 01/02->0x02                                 [FAIL]
        Expected to capture 10 packets, got 0.

After the fix:
TEST: VXLAN: ECN decap: 01/02->0x01                                 [ OK ]

Fixes: a0b61f3d8ebf ("selftests: forwarding: vxlan_bridge_1d: Add an ECN decap test")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomptcp: Change mailing list address
Mat Martineau [Fri, 19 Mar 2021 18:33:02 +0000 (11:33 -0700)]
mptcp: Change mailing list address

The mailing list for MPTCP maintenance has moved to the
kernel.org-supported mptcp@lists.linux.dev address.

Complete, combined archives for both lists are now hosted at
https://lore.kernel.org/mptcp

Cc: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Acked-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
David S. Miller [Fri, 19 Mar 2021 20:49:25 +0000 (13:49 -0700)]
Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue

Tony Nguyen says:

====================
Intel Wired LAN Driver Updates 2021-03-19

This series contains updates to e1000e and igb drivers.

Tom Seewald fixes duplicate guard issues by including the driver name in
the guard for e1000e and igb.

Jesse adds checks that timestamping is on and valid to avoid possible
issues with a misinterpreted time stamp for igb.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoselinux: vsock: Set SID for socket returned by accept()
David Brazdil [Fri, 19 Mar 2021 13:05:41 +0000 (13:05 +0000)]
selinux: vsock: Set SID for socket returned by accept()

For AF_VSOCK, accept() currently returns sockets that are unlabelled.
Other socket families derive the child's SID from the SID of the parent
and the SID of the incoming packet. This is typically done as the
connected socket is placed in the queue that accept() removes from.

Reuse the existing 'security_sk_clone' hook to copy the SID from the
parent (server) socket to the child. There is no packet SID in this
case.

Fixes: d021c344051a ("VSOCK: Introduce VM Sockets")
Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: stmmac: dwmac-sun8i: Provide TX and RX fifo sizes
Corentin Labbe [Fri, 19 Mar 2021 13:44:22 +0000 (13:44 +0000)]
net: stmmac: dwmac-sun8i: Provide TX and RX fifo sizes

MTU cannot be changed on dwmac-sun8i. (ip link set eth0 mtu xxx returning EINVAL)
This is due to tx_fifo_size being 0, since this value is used to compute valid
MTU range.
Like dwmac-sunxi (with commit 806fd188ce2a ("net: stmmac: dwmac-sunxi: Provide TX and RX fifo sizes"))
dwmac-sun8i need to have tx and rx fifo sizes set.
I have used values from datasheets.
After this patch, setting a non-default MTU (like 1000) value works and network is still useable.

Tested-on: sun8i-h3-orangepi-pc
Tested-on: sun8i-r40-bananapi-m2-ultra
Tested-on: sun50i-a64-bananapi-m64
Tested-on: sun50i-h5-nanopi-neo-plus2
Tested-on: sun50i-h6-pine-h64
Fixes: 9f93ac8d408 ("net-next: stmmac: Add dwmac-sun8i")
Reported-by: Belisko Marek <marek.belisko@gmail.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agor8152: limit the RX buffer size of RTL8153A for USB 2.0
Hayes Wang [Fri, 19 Mar 2021 07:37:21 +0000 (15:37 +0800)]
r8152: limit the RX buffer size of RTL8153A for USB 2.0

If the USB host controller is EHCI, the throughput is reduced from
300Mb/s to 60Mb/s, when the rx buffer size is modified from 16K to
32K.

According to the EHCI spec, the maximum size of the qTD is 20K.
Therefore, when the driver uses more than 20K buffer, the latency
time of EHCI would be increased. And, it let the RTL8153A get worse
throughput.

However, the driver uses alloc_pages() for rx buffer, so I limit
the rx buffer to 16K rather than 20K.

BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=205923
Fixes: ec5791c202ac ("r8152: separate the rx buffer size")
Reported-by: Robert Davies <robdavies1977@gmail.com>
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agosch_red: Fix a typo
Bhaskar Chowdhury [Fri, 19 Mar 2021 04:46:23 +0000 (10:16 +0530)]
sch_red: Fix a typo

s/recalcultion/recalculation/

Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agosctp: move sk_route_caps check and set into sctp_outq_flush_transports
Xin Long [Fri, 19 Mar 2021 03:52:41 +0000 (11:52 +0800)]
sctp: move sk_route_caps check and set into sctp_outq_flush_transports

The sk's sk_route_caps is set in sctp_packet_config, and later it
only needs to change when traversing the transport_list in a loop,
as the dst might be changed in the tx path.

So move sk_route_caps check and set into sctp_outq_flush_transports
from sctp_packet_transmit. This also fixes a dst leak reported by
Chen Yi:

  https://bugzilla.kernel.org/show_bug.cgi?id=212227

As calling sk_setup_caps() in sctp_packet_transmit may also set the
sk_route_caps for the ctrl sock in a netns. When the netns is being
deleted, the ctrl sock's releasing is later than dst dev's deleting,
which will cause this dev's deleting to hang and dmesg error occurs:

  unregister_netdevice: waiting for xxx to become free. Usage count = 1

Reported-by: Chen Yi <yiche@redhat.com>
Fixes: bcd623d8e9fa ("sctp: call sk_setup_caps in sctp_packet_transmit instead")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoigb: check timestamp validity
Jesse Brandeburg [Wed, 3 Mar 2021 20:51:03 +0000 (12:51 -0800)]
igb: check timestamp validity

Add a couple of checks to make sure timestamping is on and that the
timestamp value from DMA is valid. This avoids any functional issues
that could come from a misinterpreted time stamp.

One of the functions changed doesn't need a return value added because
there was no value in checking from the calling locations.

While here, fix a couple of reverse christmas tree issues next to
the code being changed.

Fixes: f56e7bba22fa ("igb: Pull timestamp from fragment before adding it to skb")
Fixes: 9cbc948b5a20 ("igb: add XDP support")
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Dave Switzer <david.switzer@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoigb: Fix duplicate include guard
Tom Seewald [Mon, 22 Feb 2021 04:00:05 +0000 (22:00 -0600)]
igb: Fix duplicate include guard

The include guard "_E1000_HW_H_" is used by two separate header files in
two different drivers (e1000/e1000_hw.h and igb/e1000_hw.h). Using the
same include guard macro in more than one header file may cause
unexpected behavior from the compiler. Fix this by renaming the
duplicate guard in the igb driver.

Fixes: 9d5c824399de ("igb: PCI-Express 82575 Gigabit Ethernet driver")
Signed-off-by: Tom Seewald <tseewald@gmail.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agoe1000e: Fix duplicate include guard
Tom Seewald [Mon, 22 Feb 2021 04:00:04 +0000 (22:00 -0600)]
e1000e: Fix duplicate include guard

The include guard "_E1000_HW_H_" is used by header files in three
different drivers (e1000/e1000_hw.h, e1000e/hw.h, and igb/e1000_hw.h).
Using the same include guard macro in more than one header file may
cause unexpected behavior from the compiler. Fix the duplicate include
guard in the e1000e driver by renaming it.

Fixes: bc7f75fa9788 ("[E1000E]: New pci-express e1000 driver (currently for ICH9 devices only)")
Signed-off-by: Tom Seewald <tseewald@gmail.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Acked-by: Sasha Neftin <sasha.neftin@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
3 years agonet: cdc-phonet: fix data-interface release on probe failure
Johan Hovold [Thu, 18 Mar 2021 15:57:49 +0000 (16:57 +0100)]
net: cdc-phonet: fix data-interface release on probe failure

Set the disconnected flag before releasing the data interface in case
netdev registration fails to avoid having the disconnect callback try to
deregister the never registered netdev (and trigger a WARN_ON()).

Fixes: 87cf65601e17 ("USB host CDC Phonet network interface driver")
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf
David S. Miller [Fri, 19 Mar 2021 02:19:06 +0000 (19:19 -0700)]
Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

1) Several patches to testore use of memory barriers instead of RCU to
   ensure consistent access to ruleset, from Mark Tomlinson.

2) Fix dump of expectation via ctnetlink, from Florian Westphal.

3) GRE helper works for IPv6, from Ludovic Senecaux.

4) Set error on unsupported flowtable flags.

5) Use delayed instead of deferrable workqueue in the flowtable,
   from Yinjun Zhang.

6) Fix spurious EEXIST in case of add-after-delete flowtable in
   the same batch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: check all name nodes in __dev_alloc_name
Jiri Bohac [Thu, 18 Mar 2021 03:42:53 +0000 (04:42 +0100)]
net: check all name nodes in __dev_alloc_name

__dev_alloc_name(), when supplied with a name containing '%d',
will search for the first available device number to generate a
unique device name.

Since commit ff92741270bf8b6e78aa885f166b68c7a67ab13a ("net:
introduce name_node struct to be used in hashlist") network
devices may have alternate names.  __dev_alloc_name() does take
these alternate names into account, possibly generating a name
that is already taken and failing with -ENFILE as a result.

This demonstrates the bug:

    # rmmod dummy 2>/dev/null
    # ip link property add dev lo altname dummy0
    # modprobe dummy numdummies=1
    modprobe: ERROR: could not insert 'dummy': Too many open files in system

Instead of creating a device named dummy1, modprobe fails.

Fix this by checking all the names in the d->name_node list, not just d->name.

Signed-off-by: Jiri Bohac <jbohac@suse.cz>
Fixes: ff92741270bf ("net: introduce name_node struct to be used in hashlist")
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: marvell: Remove reference to CONFIG_MV64X60
Christophe Leroy [Thu, 18 Mar 2021 17:25:08 +0000 (17:25 +0000)]
net: marvell: Remove reference to CONFIG_MV64X60

Commit 92c8c16f3457 ("powerpc/embedded6xx: Remove C2K board support")
removed last selector of CONFIG_MV64X60.

As it is not a user selectable config item, all references to it
are stale. Remove them.

Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'octeontx2-fixes'
David S. Miller [Thu, 18 Mar 2021 21:12:42 +0000 (14:12 -0700)]
Merge branch 'octeontx2-fixes'

Hariprasad Kelam says:

====================
octeontx2: miscellaneous fixes

This series of patches fixes various issues related to NPC MCAM entry
management, debugfs, devlink, CGX LMAC mapping, RSS config etc

Change-log:
v2:
Fixed below review comments
- corrected Fixed tag syntax with 12 digits SHA1
          and providing space between SHA1 and subject line
- remove code improvement patch
- make commit description more clear
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Fix uninitialized variable warning
Subbaraya Sundeep [Thu, 18 Mar 2021 14:15:49 +0000 (19:45 +0530)]
octeontx2-af: Fix uninitialized variable warning

Initialize l4_key_offset variable to fix uninitialized
variable compiler warning.

Fixes: b9b7421a01d8 ("octeontx2-af: Support ESP/AH RSS hashing")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: fix infinite loop in unmapping NPC counter
Hariprasad Kelam [Thu, 18 Mar 2021 14:15:48 +0000 (19:45 +0530)]
octeontx2-af: fix infinite loop in unmapping NPC counter

unmapping npc counter works in a way by traversing all mcam
entries to find which mcam rule is associated with counter.
But loop cursor variable 'entry' is not incremented before
checking next mcam entry which resulting in infinite loop.

This in turn hogs the kworker thread forever and no other
mbox message is processed by AF driver after that.
Fix this by updating entry value before checking next
mcam entry.

Fixes: a958dd59f9ce ("octeontx2-af: Map or unmap NPC MCAM entry and counter")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-pf: Clear RSS enable flag on interace down
Geetha sowjanya [Thu, 18 Mar 2021 14:15:47 +0000 (19:45 +0530)]
octeontx2-pf: Clear RSS enable flag on interace down

RSS configuration can not be get/set when interface is in down state
as they required mbox communication. RSS enable flag status
is used for set/get configuration. Current code do not clear the
RSS enable flag on interface down which lead to mbox error while
trying to set/get RSS configuration.

Fixes: 85069e95e531 ("octeontx2-pf: Receive side scaling support")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Fix irq free in rvu teardown
Geetha sowjanya [Thu, 18 Mar 2021 14:15:46 +0000 (19:45 +0530)]
octeontx2-af: Fix irq free in rvu teardown

Current devlink code try to free already freed irqs as the
irq_allocate flag is not cleared after free leading to kernel
crash while removing rvu driver. The patch fixes the irq free
sequence and clears the irq_allocate flag on free.

Fixes: 7304ac4567bc ("octeontx2-af: Add mailbox IRQ and msg handlers")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Return correct CGX RX fifo size
Subbaraya Sundeep [Thu, 18 Mar 2021 14:15:45 +0000 (19:45 +0530)]
octeontx2-af: Return correct CGX RX fifo size

CGX receive buffer size is a constant value and
cannot be read from CGX0 block always since
CGX0 may not enabled everytime. Hence return CGX
receive buffer size from first enabled CGX block
instead of CGX0.

Fixes: 6e54e1c5399a ("octeontx2-af: cn10K: MTU configuration")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Remove TOS field from MKEX TX
Subbaraya Sundeep [Thu, 18 Mar 2021 14:15:44 +0000 (19:45 +0530)]
octeontx2-af: Remove TOS field from MKEX TX

The MKEX profile describes what packet fields need to be extracted from
the input packet and how to place those packet fields in the output key
for MCAM matching.  The MKEX profile can be in a way where higher layer
packet fields can overwrite lower layer packet fields in output MCAM
Key.
Hence MKEX profile is always ensured that there are no overlaps between
any of the layers. But the commit 42006910b5ea
("octeontx2-af: cleanup KPU config data") introduced TX TOS field which
overlaps with DMAC in MCAM key.
This led to AF driver returning error when TX rule is installed with
DMAC as match criteria since DMAC gets overwritten and cannot be
supported. This patch fixes the issue by removing TOS field from MKEX TX
profile.

Fixes: 42006910b5ea ("octeontx2-af: cleanup KPU config data")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Formatting debugfs entry rsrc_alloc.
Rakesh Babu [Thu, 18 Mar 2021 14:15:43 +0000 (19:45 +0530)]
octeontx2-af: Formatting debugfs entry rsrc_alloc.

With the existing rsrc_alloc's format, there is misalignment for the
pcifunc entries whose VF's index is a double digit. This patch fixes
this.

    pcifunc     NPA         NIX0        NIX1        SSO GROUP   SSOWS
    TIM         CPT0        CPT1        REE0        REE1
    PF0:VF0     8           5
    PF0:VF1     9                       3
    PF0:VF10    18          10
    PF0:VF11    19                      8
    PF0:VF12    20          11
    PF0:VF13    21                      9
    PF0:VF14    22          12
    PF0:VF15    23                      10
    PF1         0           0

Fixes: 23205e6d06d4 ("octeontx2-af: Dump current resource provisioning status")
Signed-off-by: Rakesh Babu <rsaladi2@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-pf: Do not modify number of rules
Subbaraya Sundeep [Thu, 18 Mar 2021 14:15:42 +0000 (19:45 +0530)]
octeontx2-pf: Do not modify number of rules

In the ETHTOOL_GRXCLSRLALL ioctl ethtool uses
below structure to read number of rules from the driver.

    struct ethtool_rxnfc {
            __u32                           cmd;
            __u32                           flow_type;
            __u64                           data;
            struct ethtool_rx_flow_spec     fs;
            union {
                    __u32                   rule_cnt;
                    __u32                   rss_context;
            };
            __u32                           rule_locs[0];
    };

Driver must not modify rule_cnt member. But currently driver
modifies it by modifying rss_context. Hence fix it by using a
local variable.

Fixes: 81a4362016e7 ("octeontx2-pf: Add RSS multi group support")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonetsec: restore phy power state after controller reset
Mian Yousaf Kaukab [Thu, 18 Mar 2021 08:50:26 +0000 (09:50 +0100)]
netsec: restore phy power state after controller reset

Since commit 8e850f25b581 ("net: socionext: Stop PHY before resetting
netsec") netsec_netdev_init() power downs phy before resetting the
controller. However, the state is not restored once the reset is
complete. As a result it is not possible to bring up network on a
platform with Broadcom BCM5482 phy.

Fix the issue by restoring phy power state after controller reset is
complete.

Fixes: 8e850f25b581 ("net: socionext: Stop PHY before resetting netsec")
Cc: stable@vger.kernel.org
Signed-off-by: Mian Yousaf Kaukab <ykaukab@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoipv6: weaken the v4mapped source check
Jakub Kicinski [Wed, 17 Mar 2021 16:55:15 +0000 (09:55 -0700)]
ipv6: weaken the v4mapped source check

This reverts commit 6af1799aaf3f1bc8defedddfa00df3192445bbf3.

Commit 6af1799aaf3f ("ipv6: drop incoming packets having a v4mapped
source address") introduced an input check against v4mapped addresses.
Use of such addresses on the wire is indeed questionable and not
allowed on public Internet. As the commit pointed out

  https://tools.ietf.org/html/draft-itojun-v6ops-v4mapped-harmful-02

lists potential issues.

Unfortunately there are applications which use v4mapped addresses,
and breaking them is a clear regression. For example v4mapped
addresses (or any semi-valid addresses, really) may be used
for uni-direction event streams or packet export.

Since the issue which sparked the addition of the check was with
TCP and request_socks in particular push the check down to TCPv6
and DCCP. This restores the ability to receive UDPv6 packets with
v4mapped address as the source.

Keep using the IPSTATS_MIB_INHDRERRORS statistic to minimize the
user-visible changes.

Fixes: 6af1799aaf3f ("ipv6: drop incoming packets having a v4mapped source address")
Reported-by: Sunyi Shao <sunyishao@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoselftest/bpf: Add a test to check trampoline freeing logic.
Alexei Starovoitov [Thu, 18 Mar 2021 00:45:23 +0000 (17:45 -0700)]
selftest/bpf: Add a test to check trampoline freeing logic.

Add a selftest for commit e21aa341785c ("bpf: Fix fexit trampoline.")
to make sure that attaching fexit prog to a sleeping kernel function
will trigger appropriate trampoline and program destruction.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210318004523.55908-1-alexei.starovoitov@gmail.com
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
David S. Miller [Thu, 18 Mar 2021 01:36:34 +0000 (18:36 -0700)]
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2021-03-18

The following pull-request contains BPF updates for your *net* tree.

We've added 10 non-merge commits during the last 4 day(s) which contain
a total of 14 files changed, 336 insertions(+), 94 deletions(-).

The main changes are:

1) Fix fexit/fmod_ret trampoline for sleepable programs, and also fix a ftrace
   splat in modify_ftrace_direct() on address change, from Alexei Starovoitov.

2) Fix two oob speculation possibilities that allows unprivileged to leak mem
   via side-channel, from Piotr Krysiuk and Daniel Borkmann.

3) Fix libbpf's netlink handling wrt SOCK_CLOEXEC, from Kumar Kartikeya Dwivedi.

4) Fix libbpf's error handling on failure in getting section names, from Namhyung Kim.

5) Fix tunnel collect_md BPF selftest wrt Geneve option handling, from Hangbin Liu.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonetfilter: nftables: skip hook overlap logic if flowtable is stale
Pablo Neira Ayuso [Wed, 17 Mar 2021 20:19:57 +0000 (21:19 +0100)]
netfilter: nftables: skip hook overlap logic if flowtable is stale

If the flowtable has been previously removed in this batch, skip the
hook overlap checks. This fixes spurious EEXIST errors when removing and
adding the flowtable in the same batch.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
3 years agolibbpf: Use SOCK_CLOEXEC when opening the netlink socket
Kumar Kartikeya Dwivedi [Wed, 17 Mar 2021 11:58:58 +0000 (17:28 +0530)]
libbpf: Use SOCK_CLOEXEC when opening the netlink socket

Otherwise, there exists a small window between the opening and closing
of the socket fd where it may leak into processes launched by some other
thread.

Fixes: 949abbe88436 ("libbpf: add function to setup XDP")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20210317115857.6536-1-memxor@gmail.com
3 years agolibbpf: Fix error path in bpf_object__elf_init()
Namhyung Kim [Wed, 17 Mar 2021 14:54:14 +0000 (23:54 +0900)]
libbpf: Fix error path in bpf_object__elf_init()

When it failed to get section names, it should call into
bpf_object__elf_finish() like others.

Fixes: 88a82120282b ("libbpf: Factor out common ELF operations and improve logging")
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210317145414.884817-1-namhyung@kernel.org
3 years agonetfilter: flowtable: Make sure GC works periodically in idle system
Yinjun Zhang [Wed, 17 Mar 2021 12:42:24 +0000 (13:42 +0100)]
netfilter: flowtable: Make sure GC works periodically in idle system

Currently flowtable's GC work is initialized as deferrable, which
means GC cannot work on time when system is idle. So the hardware
offloaded flow may be deleted for timeout, since its used time is
not timely updated.

Resolve it by initializing the GC work as delayed work instead of
deferrable.

Fixes: c29f74e0df7a ("netfilter: nf_flow_table: hardware offload support")
Signed-off-by: Yinjun Zhang <yinjun.zhang@corigine.com>
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
3 years agonetfilter: nftables: allow to update flowtable flags
Pablo Neira Ayuso [Wed, 17 Mar 2021 11:54:57 +0000 (12:54 +0100)]
netfilter: nftables: allow to update flowtable flags

Honor flowtable flags from the control update path. Disallow disabling
to toggle hardware offload support though.

Fixes: 8bb69f3b2918 ("netfilter: nf_tables: add flowtable offload control plane")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
3 years agonetfilter: nftables: report EOPNOTSUPP on unsupported flowtable flags
Pablo Neira Ayuso [Wed, 17 Mar 2021 10:31:55 +0000 (11:31 +0100)]
netfilter: nftables: report EOPNOTSUPP on unsupported flowtable flags

Error was not set accordingly.

Fixes: 8bb69f3b2918 ("netfilter: nf_tables: add flowtable offload control plane")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
3 years agonetfilter: conntrack: Fix gre tunneling over ipv6
Ludovic Senecaux [Thu, 4 Mar 2021 09:10:50 +0000 (04:10 -0500)]
netfilter: conntrack: Fix gre tunneling over ipv6

This fix permits gre connections to be tracked within ip6tables rules

Signed-off-by: Ludovic Senecaux <linuxludo@free.fr>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
3 years agobpf: Fix fexit trampoline.
Alexei Starovoitov [Tue, 16 Mar 2021 21:00:07 +0000 (14:00 -0700)]
bpf: Fix fexit trampoline.

The fexit/fmod_ret programs can be attached to kernel functions that can sleep.
The synchronize_rcu_tasks() will not wait for such tasks to complete.
In such case the trampoline image will be freed and when the task
wakes up the return IP will point to freed memory causing the crash.
Solve this by adding percpu_ref_get/put for the duration of trampoline
and separate trampoline vs its image life times.
The "half page" optimization has to be removed, since
first_half->second_half->first_half transition cannot be guaranteed to
complete in deterministic time. Every trampoline update becomes a new image.
The image with fmod_ret or fexit progs will be freed via percpu_ref_kill and
call_rcu_tasks. Together they will wait for the original function and
trampoline asm to complete. The trampoline is patched from nop to jmp to skip
fexit progs. They are freed independently from the trampoline. The image with
fentry progs only will be freed via call_rcu_tasks_trace+call_rcu_tasks which
will wait for both sleepable and non-sleepable progs to complete.

Fixes: fec56f5890d9 ("bpf: Introduce BPF trampoline")
Reported-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Paul E. McKenney <paulmck@kernel.org> # for RCU
Link: https://lore.kernel.org/bpf/20210316210007.38949-1-alexei.starovoitov@gmail.com
3 years agonet: fix race between napi kthread mode and busy poll
Wei Wang [Tue, 16 Mar 2021 22:36:47 +0000 (15:36 -0700)]
net: fix race between napi kthread mode and busy poll

Currently, napi_thread_wait() checks for NAPI_STATE_SCHED bit to
determine if the kthread owns this napi and could call napi->poll() on
it. However, if socket busy poll is enabled, it is possible that the
busy poll thread grabs this SCHED bit (after the previous napi->poll()
invokes napi_complete_done() and clears SCHED bit) and tries to poll
on the same napi. napi_disable() could grab the SCHED bit as well.
This patch tries to fix this race by adding a new bit
NAPI_STATE_SCHED_THREADED in napi->state. This bit gets set in
____napi_schedule() if the threaded mode is enabled, and gets cleared
in napi_complete_done(), and we only poll the napi in kthread if this
bit is set. This helps distinguish the ownership of the napi between
kthread and other scenarios and fixes the race issue.

Fixes: 29863d41bb6e ("net: implement threaded-able napi poll loop support")
Reported-by: Martin Zaharinov <micron10@gmail.com>
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Wei Wang <weiwan@google.com>
Cc: Alexander Duyck <alexanderduyck@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agobpf, selftests: Fix up some test_verifier cases for unprivileged
Piotr Krysiuk [Tue, 16 Mar 2021 10:44:42 +0000 (11:44 +0100)]
bpf, selftests: Fix up some test_verifier cases for unprivileged

Fix up test_verifier error messages for the case where the original error
message changed, or for the case where pointer alu errors differ between
privileged and unprivileged tests. Also, add alternative tests for keeping
coverage of the original verifier rejection error message (fp alu), and
newly reject map_ptr += rX where rX == 0 given we now forbid alu on these
types for unprivileged. All test_verifier cases pass after the change. The
test case fixups were kept separate to ease backporting of core changes.

Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
3 years agobpf: Add sanity check for upper ptr_limit
Piotr Krysiuk [Tue, 16 Mar 2021 08:47:02 +0000 (09:47 +0100)]
bpf: Add sanity check for upper ptr_limit

Given we know the max possible value of ptr_limit at the time of retrieving
the latter, add basic assertions, so that the verifier can bail out if
anything looks odd and reject the program. Nothing triggered this so far,
but it also does not hurt to have these.

Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
3 years agoMerge tag 'mac80211-for-net-2021-03-17' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Wed, 17 Mar 2021 19:21:15 +0000 (12:21 -0700)]
Merge tag 'mac80211-for-net-2021-03-17' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211

Johannes Berg says:

====================
First round of fixes for 5.12-rc:
 * HE (802.11ax) elements can be extended, handle that
 * fix locking in network namespace changes that was
   broken due to the RTNL-redux work
 * various other small fixes
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/sched: cls_flower: fix only mask bit check in the validate_ct_state
wenxu [Wed, 17 Mar 2021 04:02:43 +0000 (12:02 +0800)]
net/sched: cls_flower: fix only mask bit check in the validate_ct_state

The ct_state validate should not only check the mask bit and also
check mask_bit & key_bit..
For the +new+est case example, The 'new' and 'est' bits should be
set in both state_mask and state flags. Or the -new-est case also
will be reject by kernel.
When Openvswitch with two flows
ct_state=+trk+new,action=commit,forward
ct_state=+trk+est,action=forward

A packet go through the kernel  and the contrack state is invalid,
The ct_state will be +trk-inv. Upcall to the ovs-vswitchd, the
finally dp action will be drop with -new-est+trk.

Fixes: 1bcc51ac0731 ("net/sched: cls_flower: Reject invalid ct_state flags rules")
Fixes: 3aed8b63336c ("net/sched: cls_flower: validate ct_state for invalid and reply flags")
Signed-off-by: wenxu <wenxu@ucloud.cn>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoionic: linearize tso skb with too many frags
Shannon Nelson [Wed, 17 Mar 2021 00:07:47 +0000 (17:07 -0700)]
ionic: linearize tso skb with too many frags

We were linearizing non-TSO skbs that had too many frags, but
we weren't checking number of frags on TSO skbs.  This could
lead to a bad page reference when we received a TSO skb with
more frags than the Tx descriptor could support.

v2: use gso_segs rather than yet another division
    don't rework the check on the nr_frags

Fixes: 0f3154e6bcb3 ("ionic: Add Tx and Rx handling")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agobpf: Simplify alu_limit masking for pointer arithmetic
Piotr Krysiuk [Tue, 16 Mar 2021 07:26:25 +0000 (08:26 +0100)]
bpf: Simplify alu_limit masking for pointer arithmetic

Instead of having the mov32 with aux->alu_limit - 1 immediate, move this
operation to retrieve_ptr_limit() instead to simplify the logic and to
allow for subsequent sanity boundary checks inside retrieve_ptr_limit().
This avoids in future that at the time of the verifier masking rewrite
we'd run into an underflow which would not sign extend due to the nature
of mov32 instruction.

Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
3 years agobpf: Fix off-by-one for area size in creating mask to left
Piotr Krysiuk [Tue, 16 Mar 2021 07:20:16 +0000 (08:20 +0100)]
bpf: Fix off-by-one for area size in creating mask to left

retrieve_ptr_limit() computes the ptr_limit for registers with stack and
map_value type. ptr_limit is the size of the memory area that is still
valid / in-bounds from the point of the current position and direction
of the operation (add / sub). This size will later be used for masking
the operation such that attempting out-of-bounds access in the speculative
domain is redirected to remain within the bounds of the current map value.

When masking to the right the size is correct, however, when masking to
the left, the size is off-by-one which would lead to an incorrect mask
and thus incorrect arithmetic operation in the non-speculative domain.
Piotr found that if the resulting alu_limit value is zero, then the
BPF_MOV32_IMM() from the fixup_bpf_calls() rewrite will end up loading
0xffffffff into AX instead of sign-extending to the full 64 bit range,
and as a result, this allows abuse for executing speculatively out-of-
bounds loads against 4GB window of address space and thus extracting the
contents of kernel memory via side-channel.

Fixes: 979d63d50c0c ("bpf: prevent out of bounds speculation on pointer arithmetic")
Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
3 years agobpf: Prohibit alu ops for pointer types not defining ptr_limit
Piotr Krysiuk [Tue, 16 Mar 2021 08:47:02 +0000 (09:47 +0100)]
bpf: Prohibit alu ops for pointer types not defining ptr_limit

The purpose of this patch is to streamline error propagation and in particular
to propagate retrieve_ptr_limit() errors for pointer types that are not defining
a ptr_limit such that register-based alu ops against these types can be rejected.

The main rationale is that a gap has been identified by Piotr in the existing
protection against speculatively out-of-bounds loads, for example, in case of
ctx pointers, unprivileged programs can still perform pointer arithmetic. This
can be abused to execute speculatively out-of-bounds loads without restrictions
and thus extract contents of kernel memory.

Fix this by rejecting unprivileged programs that attempt any pointer arithmetic
on unprotected pointer types. The two affected ones are pointer to ctx as well
as pointer to map. Field access to a modified ctx' pointer is rejected at a
later point in time in the verifier, and 7c6967326267 ("bpf: Permit map_ptr
arithmetic with opcode add and offset 0") only relevant for root-only use cases.
Risk of unprivileged program breakage is considered very low.

Fixes: 7c6967326267 ("bpf: Permit map_ptr arithmetic with opcode add and offset 0")
Fixes: b2157399cc98 ("bpf: prevent out-of-bounds speculation")
Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
3 years agoftrace: Fix modify_ftrace_direct.
Alexei Starovoitov [Tue, 16 Mar 2021 19:58:15 +0000 (12:58 -0700)]
ftrace: Fix modify_ftrace_direct.

The following sequence of commands:
  register_ftrace_direct(ip, addr1);
  modify_ftrace_direct(ip, addr1, addr2);
  unregister_ftrace_direct(ip, addr2);
will cause the kernel to warn:
[   30.179191] WARNING: CPU: 2 PID: 1961 at kernel/trace/ftrace.c:5223 unregister_ftrace_direct+0x130/0x150
[   30.180556] CPU: 2 PID: 1961 Comm: test_progs    W  O      5.12.0-rc2-00378-g86bc10a0a711-dirty #3246
[   30.182453] RIP: 0010:unregister_ftrace_direct+0x130/0x150

When modify_ftrace_direct() changes the addr from old to new it should update
the addr stored in ftrace_direct_funcs. Otherwise the final
unregister_ftrace_direct() won't find the address and will cause the splat.

Fixes: 0567d6809182 ("ftrace: Add modify_ftrace_direct()")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lore.kernel.org/bpf/20210316195815.34714-1-alexei.starovoitov@gmail.com
3 years agoMAINTAINERS: Update Spidernet network driver
Geoff Levand [Tue, 16 Mar 2021 17:13:52 +0000 (10:13 -0700)]
MAINTAINERS: Update Spidernet network driver

Change the Spidernet network driver from supported to
maintained, add the linuxppc-dev ML, and add myself as
a 'maintainer'.

Cc: Ishizaki Kou <kou.ishizaki@toshiba.co.jp>
Signed-off-by: Geoff Levand <geoff@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'nfp-fixes'
David S. Miller [Tue, 16 Mar 2021 22:36:44 +0000 (15:36 -0700)]
Merge branch 'nfp-fixes'

Simon Horman says:

====================
Fixes for nfp pre_tunnel code

Louis Peens says:

The following set of patches fixes up a few bugs in the pre_tun
decap code paths which has been hiding for a while.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonfp: flower: fix pre_tun mask id allocation
Louis Peens [Tue, 16 Mar 2021 18:13:10 +0000 (19:13 +0100)]
nfp: flower: fix pre_tun mask id allocation

pre_tun_rule flows does not follow the usual add-flow path, instead
they are used to update the pre_tun table on the firmware. This means
that if the mask-id gets allocated here the firmware will never see the
"NFP_FL_META_FLAG_MANAGE_MASK" flag for the specific mask id, which
triggers the allocation on the firmware side. This leads to the firmware
mask being corrupted and causing all sorts of strange behaviour.

Fixes: f12725d98cbe ("nfp: flower: offload pre-tunnel rules")
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonfp: flower: add ipv6 bit to pre_tunnel control message
Louis Peens [Tue, 16 Mar 2021 18:13:09 +0000 (19:13 +0100)]
nfp: flower: add ipv6 bit to pre_tunnel control message

Differentiate between ipv4 and ipv6 flows when configuring the pre_tunnel
table to prevent them trampling each other in the table.

Fixes: 783461604f7e ("nfp: flower: update flow merge code to support IPv6 tunnels")
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonfp: flower: fix unsupported pre_tunnel flows
Louis Peens [Tue, 16 Mar 2021 18:13:08 +0000 (19:13 +0100)]
nfp: flower: fix unsupported pre_tunnel flows

There are some pre_tunnel flows combinations which are incorrectly being
offloaded without proper support, fix these.

- Matching on MPLS is not supported for pre_tun.
- Match on IPv4/IPv6 layer must be present.
- Destination MAC address must match pre_tun.dev MAC

Fixes: 120ffd84a9ec ("nfp: flower: verify pre-tunnel rules")
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: broadcom: BCM4908_ENET should not default to y, unconditionally
Geert Uytterhoeven [Tue, 16 Mar 2021 14:03:41 +0000 (15:03 +0100)]
net: broadcom: BCM4908_ENET should not default to y, unconditionally

Merely enabling compile-testing should not enable additional code.
To fix this, restrict the automatic enabling of BCM4908_ENET to
ARCH_BCM4908.

Fixes: 4feffeadbcb2e5b1 ("net: broadcom: bcm4908enet: add BCM4908 controller driver")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Rafał Miłecki <rafal@milecki.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/sched: act_api: fix miss set post_ct for ovs after do conntrack in act_ct
wenxu [Tue, 16 Mar 2021 08:33:54 +0000 (16:33 +0800)]
net/sched: act_api: fix miss set post_ct for ovs after do conntrack in act_ct

When openvswitch conntrack offload with act_ct action. The first rule
do conntrack in the act_ct in tc subsystem. And miss the next rule in
the tc and fallback to the ovs datapath but miss set post_ct flag
which will lead the ct_state_key with -trk flag.

Fixes: 7baf2429a1a9 ("net/sched: cls_flower add CT_FLAGS_INVALID flag support")
Signed-off-by: wenxu <wenxu@ucloud.cn>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge tag 'linux-can-fixes-for-5.12-20210316' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Tue, 16 Mar 2021 22:10:43 +0000 (15:10 -0700)]
Merge tag 'linux-can-fixes-for-5.12-20210316' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can

Marc Kleine-Budde says:

====================
pull-request: can 2021-03-16

this is a pull request of 11 patches for net/master.

The first patch is by Martin Willi and fixes the deletion of network
name spaces with physical CAN interfaces in them.

The next two patches are by me an fix the ISOTP protocol, to ensure
that unused flags in classical CAN frames are properly initialized to
zero.

Stephane Grosjean contributes a patch for the pcan_usb_fd driver,
which add MODULE_SUPPORTED_DEVICE lines for two supported devices.

Angelo Dureghello's patch for the flexcan driver fixes a potential div
by zero, if the bitrate is not set during driver probe.

Jimmy Assarsson's patch for the kvaser_pciefd disables bus load
reporting in the device, if it was previously enabled by the vendor's
out of tree drier. A patch for the kvaser_usb adds support for a new
device, by adding the appropriate USB product ID.

Tong Zhang contributes two patches for the c_can driver. First a
use-after-free in the c_can_pci driver is fixed, in the second patch
the runtime PM for the c_can_pci is fixed by moving the runtime PM
enable/disable from the core driver to the platform driver.

The last two patches are by Torin Cooper-Bennun for the m_can driver.
First a extraneous msg loss warning is removed then he fixes the
RX-path, which might be blocked by errors.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodocs: net: ena: Fix ena_start_xmit() function name typo
Zenghui Yu [Tue, 16 Mar 2021 03:27:37 +0000 (11:27 +0800)]
docs: net: ena: Fix ena_start_xmit() function name typo

The ena.rst documentation referred to end_start_xmit() when it should refer
to ena_start_xmit(). Fix the typo.

Signed-off-by: Zenghui Yu <yuzenghui@huawei.com>
Acked-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoselftests/net: fix warnings on reuseaddr_ports_exhausted
Carlos Llamas [Tue, 16 Mar 2021 01:04:29 +0000 (01:04 +0000)]
selftests/net: fix warnings on reuseaddr_ports_exhausted

Fix multiple warnings seen with gcc 10.2.1:
reuseaddr_ports_exhausted.c:32:41: warning: missing braces around initializer [-Wmissing-braces]
   32 | struct reuse_opts unreusable_opts[12] = {
      |                                         ^
   33 |  {0, 0, 0, 0},
      |   {   } {   }

Fixes: 7f204a7de8b0 ("selftests: net: Add SO_REUSEADDR test to check if 4-tuples are fully utilized.")
Signed-off-by: Carlos Llamas <cmllamas@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agowireless/nl80211: fix wdev_id may be used uninitialized
Jarod Wilson [Fri, 12 Mar 2021 16:36:51 +0000 (11:36 -0500)]
wireless/nl80211: fix wdev_id may be used uninitialized

Build currently fails with -Werror=maybe-uninitialized set:

net/wireless/nl80211.c: In function '__cfg80211_wdev_from_attrs':
net/wireless/nl80211.c:124:44: error: 'wdev_id' may be used
uninitialized in this function [-Werror=maybe-uninitialized]

Easy fix is to just initialize wdev_id to 0, since it's value doesn't
otherwise matter unless have_wdev_id is true.

Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
CC: Johannes Berg <johannes@sipsolutions.net>
CC: "David S. Miller" <davem@davemloft.net>
CC: Jakub Kicinski <kuba@kernel.org>
CC: linux-wireless@vger.kernel.org
CC: netdev@vger.kernel.org
Signed-off-by: Jarod Wilson <jarod@redhat.com>
Link: https://lore.kernel.org/r/20210312163651.1398207-1-jarod@redhat.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: choose first enabled channel for monitor
Karthikeyan Kathirvel [Thu, 11 Mar 2021 05:29:07 +0000 (10:59 +0530)]
mac80211: choose first enabled channel for monitor

Even if the first channel from sband channel list is invalid
or disabled mac80211 ends up choosing it as the default channel
for monitor interfaces, making them not usable.

Fix this by assigning the first available valid or enabled
channel instead.

Signed-off-by: Karthikeyan Kathirvel <kathirve@codeaurora.org>
Link: https://lore.kernel.org/r/1615440547-7661-1-git-send-email-kathirve@codeaurora.org
[reword commit message, comment, code cleanups]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agonl80211: fix locking for wireless device netns change
Johannes Berg [Wed, 10 Mar 2021 20:58:40 +0000 (21:58 +0100)]
nl80211: fix locking for wireless device netns change

We have all the network interfaces marked as netns-local
since the only reasonable thing to do right now is to set
a whole device, including all netdevs, into a different
network namespace. For this reason, we also have our own
way of changing the network namespace.

Unfortunately, the RTNL locking changes broke this, and
it now results in many RTNL assertions. The trivial fix
for those (just hold RTNL for the changes) however leads
to deadlocks in the cfg80211 netdev notifier.

Since we only need the wiphy, and that's still protected
by the RTNL, add a new NL80211_FLAG_NO_WIPHY_MTX flag to
the nl80211 ops and use it to _not_ take the wiphy mutex
but only the RTNL. This way, the notifier does all the
work necessary during unregistration/registration of the
netdevs from the old and in the new namespace.

Reported-by: Sid Hayn <sidhayn@gmail.com>
Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Link: https://lore.kernel.org/r/20210310215839.eadf7c43781b.I5fc6cf6676f800ab8008e03bbea9c3349b02d804@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: Check crypto_aead_encrypt for errors
Daniel Phan [Tue, 9 Mar 2021 20:41:36 +0000 (12:41 -0800)]
mac80211: Check crypto_aead_encrypt for errors

crypto_aead_encrypt returns <0 on error, so if these calls are not checked,
execution may continue with failed encrypts.  It also seems that these two
crypto_aead_encrypt calls are the only instances in the codebase that are
not checked for errors.

Signed-off-by: Daniel Phan <daniel.phan36@gmail.com>
Link: https://lore.kernel.org/r/20210309204137.823268-1-daniel.phan36@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: Allow HE operation to be longer than expected.
Brian Norris [Tue, 23 Feb 2021 05:19:26 +0000 (13:19 +0800)]
mac80211: Allow HE operation to be longer than expected.

We observed some Cisco APs sending the following HE Operation IE in
associate response:

  ff 0a 24 f4 3f 00 01 fc ff 00 00 00

Its HE operation parameter is 0x003ff4, so the expected total length is
7 which does not match the actual length = 10. This causes association
failing with "HE AP is missing HE Capability/operation."

According to P802.11ax_D4 Table9-94, HE operation is extensible, and
according to 802.11-2016 10.27.8, STA should discard the part beyond
the maximum length and parse the truncated element.

Allow HE operation element to be longer than expected to handle this
case and future extensions.

Fixes: e4d005b80dee ("mac80211: refactor extended element parsing")
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Yen-lin Lai <yenlinlai@chromium.org>
Link: https://lore.kernel.org/r/20210223051926.2653301-1-yenlinlai@chromium.org
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: minstrel_ht: remove unused variable 'mg'
Johannes Berg [Fri, 19 Feb 2021 09:57:45 +0000 (10:57 +0100)]
mac80211: minstrel_ht: remove unused variable 'mg'

This probably came in through some refactoring and what is
now a call to minstrel_ht_group_min_rate_offset(), remove
the unused variable.

Reported-by: kernel test robot <lkp@intel.com>
Acked-by: Felix Fietkau <nbd@nbd.name>
Link: https://lore.kernel.org/r/20210219105744.f2538a80f6cf.I3d53554c158d5b896ac07ea546bceac67372ec28@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: fix double free in ibss_leave
Markus Theil [Sat, 13 Feb 2021 13:36:53 +0000 (14:36 +0100)]
mac80211: fix double free in ibss_leave

Clear beacon ie pointer and ie length after free
in order to prevent double free.

==================================================================
BUG: KASAN: double-free or invalid-free \
in ieee80211_ibss_leave+0x83/0xe0 net/mac80211/ibss.c:1876

CPU: 0 PID: 8472 Comm: syz-executor100 Not tainted 5.11.0-rc6-syzkaller #0
Call Trace:
 __dump_stack lib/dump_stack.c:79 [inline]
 dump_stack+0x107/0x163 lib/dump_stack.c:120
 print_address_description.constprop.0.cold+0x5b/0x2c6 mm/kasan/report.c:230
 kasan_report_invalid_free+0x51/0x80 mm/kasan/report.c:355
 ____kasan_slab_free+0xcc/0xe0 mm/kasan/common.c:341
 kasan_slab_free include/linux/kasan.h:192 [inline]
 __cache_free mm/slab.c:3424 [inline]
 kfree+0xed/0x270 mm/slab.c:3760
 ieee80211_ibss_leave+0x83/0xe0 net/mac80211/ibss.c:1876
 rdev_leave_ibss net/wireless/rdev-ops.h:545 [inline]
 __cfg80211_leave_ibss+0x19a/0x4c0 net/wireless/ibss.c:212
 __cfg80211_leave+0x327/0x430 net/wireless/core.c:1172
 cfg80211_leave net/wireless/core.c:1221 [inline]
 cfg80211_netdev_notifier_call+0x9e8/0x12c0 net/wireless/core.c:1335
 notifier_call_chain+0xb5/0x200 kernel/notifier.c:83
 call_netdevice_notifiers_info+0xb5/0x130 net/core/dev.c:2040
 call_netdevice_notifiers_extack net/core/dev.c:2052 [inline]
 call_netdevice_notifiers net/core/dev.c:2066 [inline]
 __dev_close_many+0xee/0x2e0 net/core/dev.c:1586
 __dev_close net/core/dev.c:1624 [inline]
 __dev_change_flags+0x2cb/0x730 net/core/dev.c:8476
 dev_change_flags+0x8a/0x160 net/core/dev.c:8549
 dev_ifsioc+0x210/0xa70 net/core/dev_ioctl.c:265
 dev_ioctl+0x1b1/0xc40 net/core/dev_ioctl.c:511
 sock_do_ioctl+0x148/0x2d0 net/socket.c:1060
 sock_ioctl+0x477/0x6a0 net/socket.c:1177
 vfs_ioctl fs/ioctl.c:48 [inline]
 __do_sys_ioctl fs/ioctl.c:753 [inline]
 __se_sys_ioctl fs/ioctl.c:739 [inline]
 __x64_sys_ioctl+0x193/0x200 fs/ioctl.c:739
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Reported-by: syzbot+93976391bf299d425f44@syzkaller.appspotmail.com
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
Link: https://lore.kernel.org/r/20210213133653.367130-1-markus.theil@tu-ilmenau.de
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agomac80211: fix rate mask reset
Johannes Berg [Fri, 12 Feb 2021 10:22:14 +0000 (11:22 +0100)]
mac80211: fix rate mask reset

Coverity reported the strange "if (~...)" condition that's
always true. It suggested that ! was intended instead of ~,
but upon further analysis I'm convinced that what really was
intended was a comparison to 0xff/0xffff (in HT/VHT cases
respectively), since this indicates that all of the rates
are enabled.

Change the comparison accordingly.

I'm guessing this never really mattered because a reset to
not having a rate mask is basically equivalent to having a
mask that enables all rates.

Reported-by: Colin Ian King <colin.king@canonical.com>
Fixes: 2ffbe6d33366 ("mac80211: fix and optimize MCS mask handling")
Fixes: b119ad6e726c ("mac80211: add rate mask logic for vht rates")
Reviewed-by: Colin Ian King <colin.king@canonical.com>
Link: https://lore.kernel.org/r/20210212112213.36b38078f569.I8546a20c80bc1669058eb453e213630b846e107b@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
3 years agocan: m_can: m_can_rx_peripheral(): fix RX being blocked by errors
Torin Cooper-Bennun [Wed, 3 Mar 2021 14:43:51 +0000 (14:43 +0000)]
can: m_can: m_can_rx_peripheral(): fix RX being blocked by errors

For M_CAN peripherals, m_can_rx_handler() was called with quota = 1,
which caused any error handling to block RX from taking place until
the next time the IRQ handler is called. This had been observed to
cause RX to be blocked indefinitely in some cases.

This is fixed by calling m_can_rx_handler with a sensibly high quota.

Fixes: f524f829b75a ("can: m_can: Create a m_can platform framework")
Link: https://lore.kernel.org/r/20210303144350.4093750-1-torin@maxiluxsystems.com
Suggested-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Torin Cooper-Bennun <torin@maxiluxsystems.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: m_can: m_can_do_rx_poll(): fix extraneous msg loss warning
Torin Cooper-Bennun [Wed, 3 Mar 2021 10:31:52 +0000 (10:31 +0000)]
can: m_can: m_can_do_rx_poll(): fix extraneous msg loss warning

Message loss from RX FIFO 0 is already handled in
m_can_handle_lost_msg(), with netdev output included.

Removing this warning also improves driver performance under heavy
load, where m_can_do_rx_poll() may be called many times before this
interrupt is cleared, causing this message to be output many
times (thanks Mariusz Madej for this report).

Fixes: e0d1f4816f2a ("can: m_can: add Bosch M_CAN controller support")
Link: https://lore.kernel.org/r/20210303103151.3760532-1-torin@maxiluxsystems.com
Reported-by: Mariusz Madej <mariusz.madej@xtrack.com>
Signed-off-by: Torin Cooper-Bennun <torin@maxiluxsystems.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: c_can: move runtime PM enable/disable to c_can_platform
Tong Zhang [Tue, 2 Mar 2021 02:55:40 +0000 (21:55 -0500)]
can: c_can: move runtime PM enable/disable to c_can_platform

Currently doing modprobe c_can_pci will make the kernel complain:

    Unbalanced pm_runtime_enable!

this is caused by pm_runtime_enable() called before pm is initialized.

This fix is similar to 227619c3ff7c, move those pm_enable/disable code
to c_can_platform.

Fixes: 4cdd34b26826 ("can: c_can: Add runtime PM support to Bosch C_CAN/D_CAN controller")
Link: http://lore.kernel.org/r/20210302025542.987600-1-ztong0001@gmail.com
Signed-off-by: Tong Zhang <ztong0001@gmail.com>
Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: c_can_pci: c_can_pci_remove(): fix use-after-free
Tong Zhang [Mon, 1 Mar 2021 02:45:11 +0000 (21:45 -0500)]
can: c_can_pci: c_can_pci_remove(): fix use-after-free

There is a UAF in c_can_pci_remove(). dev is released by
free_c_can_dev() and is used by pci_iounmap(pdev, priv->base) later.
To fix this issue, save the mmio address before releasing dev.

Fixes: 5b92da0443c2 ("c_can_pci: generic module for C_CAN/D_CAN on PCI")
Link: https://lore.kernel.org/r/20210301024512.539039-1-ztong0001@gmail.com
Signed-off-by: Tong Zhang <ztong0001@gmail.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: kvaser_usb: Add support for USBcan Pro 4xHS
Jimmy Assarsson [Tue, 9 Mar 2021 09:17:24 +0000 (10:17 +0100)]
can: kvaser_usb: Add support for USBcan Pro 4xHS

Add support for Kvaser USBcan Pro 4xHS.

Link: https://lore.kernel.org/r/20210309091724.31262-2-jimmyassarsson@gmail.com
Signed-off-by: Jimmy Assarsson <extja@kvaser.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: kvaser_pciefd: Always disable bus load reporting
Jimmy Assarsson [Tue, 9 Mar 2021 09:17:23 +0000 (10:17 +0100)]
can: kvaser_pciefd: Always disable bus load reporting

Under certain circumstances, when switching from Kvaser's linuxcan driver
(kvpciefd) to the SocketCAN driver (kvaser_pciefd), the bus load reporting
is not disabled.
This is flooding the kernel log with prints like:
[3485.574677] kvaser_pciefd 0000:02:00.0: Received unexpected packet type 0x00000009

Always put the controller in the expected state, instead of assuming that
bus load reporting is inactive.

Note: If bus load reporting is enabled when the driver is loaded, you will
      still get a number of bus load packages (and printouts), before it is
      disabled.

Fixes: 26ad340e582d ("can: kvaser_pciefd: Add driver for Kvaser PCIEcan devices")
Link: https://lore.kernel.org/r/20210309091724.31262-1-jimmyassarsson@gmail.com
Signed-off-by: Jimmy Assarsson <extja@kvaser.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>