linux-2.6-block.git
12 months agoselftests: mptcp: join: support local endpoint being tracked or not
Matthieu Baerts [Sat, 10 Jun 2023 16:11:41 +0000 (18:11 +0200)]
selftests: mptcp: join: support local endpoint being tracked or not

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

At some points, a new feature caused internal behaviour changes we are
verifying in the selftests, see the Fixes tag below. It was not a uAPI
change but because in these selftests, we check some internal
behaviours, it is normal we have to adapt them from time to time after
having added some features.

It is possible to look for "mptcp_pm_subflow_check_next" in kallsyms
because it was needed to introduce the mentioned feature. So we can know
in advance what the behaviour we are expecting here instead of
supporting the two behaviours.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: join: skip test if iptables/tc cmds fail
Matthieu Baerts [Sat, 10 Jun 2023 16:11:40 +0000 (18:11 +0200)]
selftests: mptcp: join: skip test if iptables/tc cmds fail

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

Some tests are using IPTables and/or TC commands to force some
behaviours. If one of these commands fails -- likely because some
features are not available due to missing kernel config -- we should
intercept the error and skip the tests requiring these features.

Note that if we expect to have these features available and if
SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var is set to 1, the tests
will be marked as failed instead of skipped.

This patch also replaces the 'exit 1' by 'return 1' not to stop the
selftest in the middle without the conclusion if there is an issue with
NF or TC.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 8d014eaa9254 ("selftests: mptcp: add ADD_ADDR timeout test case")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: join: skip check if MIB counter not supported
Matthieu Baerts [Sat, 10 Jun 2023 16:11:39 +0000 (18:11 +0200)]
selftests: mptcp: join: skip check if MIB counter not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the MPTCP MIB counters introduced in commit fc518953bc9c
("mptcp: add and use MIB counter infrastructure") and more later. The
MPTCP Join selftest heavily relies on these counters.

If a counter is not supported by the kernel, it is not displayed when
using 'nstat -z'. We can then detect that and skip the verification. A
new helper (get_counter()) has been added to do the required checks and
return an error if the counter is not available.

Note that if we expect to have these features available and if
SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var is set to 1, the tests
will be marked as failed instead of skipped.

This new helper also makes sure we get the exact counter we want to
avoid issues we had in the past, e.g. with MPTcpExtRmAddr and
MPTcpExtRmAddrDrop sharing the same prefix. While at it, we uniform the
way we fetch a MIB counter.

Note for the backports: we rarely change these modified blocks so if
there is are conflicts, it is very likely because a counter is not used
in the older kernels and we don't need that chunk.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: b08fbf241064 ("selftests: add test-cases for MPTCP MP_JOIN")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: join: helpers to skip tests
Matthieu Baerts [Sat, 10 Jun 2023 16:11:38 +0000 (18:11 +0200)]
selftests: mptcp: join: helpers to skip tests

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

Here are some helpers that will be used to mark subtests as skipped if a
feature is not supported. Marking as a fix for the commit introducing
this selftest to help with the backports.

While at it, also check if kallsyms feature is available as it will also
be used in the following commits to check if MPTCP features are
available before starting a test.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: b08fbf241064 ("selftests: add test-cases for MPTCP MP_JOIN")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: join: use 'iptables-legacy' if available
Matthieu Baerts [Sat, 10 Jun 2023 16:11:37 +0000 (18:11 +0200)]
selftests: mptcp: join: use 'iptables-legacy' if available

IPTables commands using 'iptables-nft' fail on old kernels, at least
5.15 because it doesn't see the default IPTables chains:

  $ iptables -L
  iptables/1.8.2 Failed to initialize nft: Protocol not supported

As a first step before switching to NFTables, we can use iptables-legacy
if available.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 8d014eaa9254 ("selftests: mptcp: add ADD_ADDR timeout test case")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: lib: skip if not below kernel version
Matthieu Baerts [Sat, 10 Jun 2023 16:11:36 +0000 (18:11 +0200)]
selftests: mptcp: lib: skip if not below kernel version

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

A new function is now available to easily detect if a feature is
missing by looking at the kernel version. That's clearly not ideal and
this kind of check should be avoided as soon as possible. But sometimes,
there are no external sign that a "feature" is available or not:
internal behaviours can change without modifying the uAPI and these
selftests are verifying the internal behaviours. Sometimes, the only
(easy) way to verify if the feature is present is to run the test but
then the validation cannot determine if there is a failure with the
feature or if the feature is missing. Then it looks better to check the
kernel version instead of having tests that can never fail. In any case,
we need a solution not to have a whole selftest being marked as failed
just because one sub-test has failed.

Note that this env var car be set to 1 not to do such check and run the
linked sub-test: SELFTESTS_MPTCP_LIB_NO_KVERSION_CHECK.

This new helper is going to be used in the following commits. In order
to ease the backport of such future patches, it would be good if this
patch is backported up to the introduction of MPTCP selftests, hence the
Fixes tag below: this type of check was supposed to be done from the
beginning.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 048d19d444be ("mptcp: add basic kselftest for mptcp")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoMerge branch 'fixes-for-q-usgmii-speeds-and-autoneg'
Jakub Kicinski [Mon, 12 Jun 2023 23:51:22 +0000 (16:51 -0700)]
Merge branch 'fixes-for-q-usgmii-speeds-and-autoneg'

Maxime Chevallier says:

====================
fixes for Q-USGMII speeds and autoneg

This is the second version of a small changeset for QUSGMII support,
fixing inconsistencies in reported max speed and control word parsing.

As reported here [1], there are some inconsistencies for the Q-USGMII
mode speeds and configuration. The first patch in this fixup series
makes so that we correctly report the max speed of 1Gbps for this mode.

The second patch uses a dedicated helper to decode the control word.
This is necessary as although USGMII control words are close to USXGMII,
they don't support the same speeds.

[1] : https://lore.kernel.org/netdev/ZHnd+6FUO77XFJvQ@shell.armlinux.org.uk/
====================

Link: https://lore.kernel.org/r/20230609080305.546028-1-maxime.chevallier@bootlin.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonet: phylink: use a dedicated helper to parse usgmii control word
Maxime Chevallier [Fri, 9 Jun 2023 08:03:05 +0000 (10:03 +0200)]
net: phylink: use a dedicated helper to parse usgmii control word

Q-USGMII is a derivative of USGMII, that uses a specific formatting for
the control word. The layout is close to the USXGMII control word, but
doesn't support speeds over 1Gbps. Use a dedicated decoding logic for
the USGMII control word, re-using USXGMII definitions but only considering
10/100/1000Mbps speeds

Fixes: 5e61fe157a27 ("net: phy: Introduce QUSGMII PHY mode")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonet: phylink: report correct max speed for QUSGMII
Maxime Chevallier [Fri, 9 Jun 2023 08:03:04 +0000 (10:03 +0200)]
net: phylink: report correct max speed for QUSGMII

Q-USGMII is the quad port version of USGMII, and supports a max speed of
1Gbps on each line. Make so that phylink_interface_max_speed() reports
this information correctly.

Fixes: ae0e4bb2a0e0 ("net: phylink: Adjust link settings based on rate matching")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agosctp: fix an error code in sctp_sf_eat_auth()
Dan Carpenter [Fri, 9 Jun 2023 11:05:19 +0000 (14:05 +0300)]
sctp: fix an error code in sctp_sf_eat_auth()

The sctp_sf_eat_auth() function is supposed to enum sctp_disposition
values and returning a kernel error code will cause issues in the
caller.  Change -ENOMEM to SCTP_DISPOSITION_NOMEM.

Fixes: 65b07e5d0d09 ("[SCTP]: API updates to suport SCTP-AUTH extensions.")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agosctp: handle invalid error codes without calling BUG()
Dan Carpenter [Fri, 9 Jun 2023 11:04:43 +0000 (14:04 +0300)]
sctp: handle invalid error codes without calling BUG()

The sctp_sf_eat_auth() function is supposed to return enum sctp_disposition
values but if the call to sctp_ulpevent_make_authkey() fails, it returns
-ENOMEM.

This results in calling BUG() inside the sctp_side_effects() function.
Calling BUG() is an over reaction and not helpful.  Call WARN_ON_ONCE()
instead.

This code predates git.

Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoipvlan: fix bound dev checking for IPv6 l3s mode
Hangbin Liu [Fri, 9 Jun 2023 09:15:02 +0000 (17:15 +0800)]
ipvlan: fix bound dev checking for IPv6 l3s mode

The commit 59a0b022aa24 ("ipvlan: Make skb->skb_iif track skb->dev for l3s
mode") fixed ipvlan bonded dev checking by updating skb skb_iif. This fix
works for IPv4, as in raw_v4_input() the dif is from inet_iif(skb), which
is skb->skb_iif when there is no route.

But for IPv6, the fix is not enough, because in ipv6_raw_deliver() ->
raw_v6_match(), the dif is inet6_iif(skb), which is returns IP6CB(skb)->iif
instead of skb->skb_iif if it's not a l3_slave. To fix the IPv6 part
issue. Let's set IP6CB(skb)->iif to correct ifindex.

BTW, ipvlan handles NS/NA specifically. Since it works fine, I will not
reset IP6CB(skb)->iif when addr->atype is IPVL_ICMPV6.

Fixes: c675e06a98a4 ("ipvlan: decouple l3s mode dependencies from other modes")
Link: https://bugzilla.redhat.com/show_bug.cgi?id=2196710
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: ethtool: correct MAX attribute value for stats
Jakub Kicinski [Thu, 8 Jun 2023 16:23:44 +0000 (09:23 -0700)]
net: ethtool: correct MAX attribute value for stats

When compiling YNL generated code compiler complains about
array-initializer-out-of-bounds. Turns out the MAX value
for STATS_GRP uses the value for STATS.

This may lead to random corruptions in user space (kernel
itself doesn't use this value as it never parses stats).

Fixes: f09ea6fb1272 ("ethtool: add a new command for reading standard stats")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agobnx2x: fix page fault following EEH recovery
David Christensen [Thu, 8 Jun 2023 20:01:43 +0000 (16:01 -0400)]
bnx2x: fix page fault following EEH recovery

In the last step of the EEH recovery process, the EEH driver calls into
bnx2x_io_resume() to re-initialize the NIC hardware via the function
bnx2x_nic_load().  If an error occurs during bnx2x_nic_load(), OS and
hardware resources are released and an error code is returned to the
caller.  When called from bnx2x_io_resume(), the return code is ignored
and the network interface is brought up unconditionally.  Later attempts
to send a packet via this interface result in a page fault due to a null
pointer reference.

This patch checks the return code of bnx2x_nic_load(), prints an error
message if necessary, and does not enable the interface.

Signed-off-by: David Christensen <drc@linux.vnet.ibm.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoMerge tag 'nf-23-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
David S. Miller [Sat, 10 Jun 2023 18:57:03 +0000 (19:57 +0100)]
Merge tag 'nf-23-06-08' of git://git./linux/kernel/git/netfilter/nf

netfilter pull request 23-06-08

Pablo Neira Ayuso says:

====================
The following patchset contains Netfilter fixes for net:

1) Add commit and abort set operation to pipapo set abort path.

2) Bail out immediately in case of ENOMEM in nfnetlink batch.

3) Incorrect error path handling when creating a new rule leads to
   dangling pointer in set transaction list.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonetlabel: fix shift wrapping bug in netlbl_catmap_setlong()
Dmitry Mastykin [Thu, 8 Jun 2023 13:57:54 +0000 (16:57 +0300)]
netlabel: fix shift wrapping bug in netlbl_catmap_setlong()

There is a shift wrapping bug in this code on 32-bit architectures.
NETLBL_CATMAP_MAPTYPE is u64, bitmap is unsigned long.
Every second 32-bit word of catmap becomes corrupted.

Signed-off-by: Dmitry Mastykin <dmastykin@astralinux.ru>
Acked-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoMerge branch 'octeontx2-af-fixes'
David S. Miller [Sat, 10 Jun 2023 18:53:02 +0000 (19:53 +0100)]
Merge branch 'octeontx2-af-fixes'

Naveen Mamindlapalli says:

====================
RVU NIX AF driver fixes

This patch series contains few fixes to the RVU NIX AF driver.

The first patch modifies the driver check to validate whether the req count
for contiguous and non-contiguous arrays is less than the maximum limit.

The second patch fixes HW lbk interface link credit programming.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoocteontx2-af: fix lbk link credits on cn10k
Nithin Dabilpuram [Thu, 8 Jun 2023 11:42:01 +0000 (17:12 +0530)]
octeontx2-af: fix lbk link credits on cn10k

Fix LBK link credits on CN10K to be same as CN9K i.e
16 * MAX_LBK_DATA_RATE instead of current scheme of
calculation based on LBK buf length / FIFO size.

Fixes: 6e54e1c5399a ("octeontx2-af: cn10K: Add MTU configuration")
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoocteontx2-af: fixed resource availability check
Satha Rao [Thu, 8 Jun 2023 11:42:00 +0000 (17:12 +0530)]
octeontx2-af: fixed resource availability check

txschq_alloc response have two different arrays to store continuous
and non-continuous schedulers of each level. Requested count should
be checked for each array separately.

Fixes: 5d9b976d4480 ("octeontx2-af: Support fixed transmit scheduler topology")
Signed-off-by: Satha Rao <skoteshwar@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoMerge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net...
Jakub Kicinski [Sat, 10 Jun 2023 07:11:18 +0000 (00:11 -0700)]
Merge branch '100GbE' of git://git./linux/kernel/git/tnguy/net-queue

Tony Nguyen says:

====================
Intel Wired LAN Driver Updates 2023-06-08 (ice)

This series contains updates to ice driver only.

Simon Horman stops null pointer dereference for GNSS error path.

Kamil fixes memory leak when downing interface when XDP is enabled.

* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
  ice: Fix XDP memory leak when NIC is brought up and down
  ice: Don't dereference NULL in ice_gnss_read error path
====================

Link: https://lore.kernel.org/r/20230608200051.451752-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoiavf: remove mask from iavf_irq_enable_queues()
Ahmed Zaki [Thu, 8 Jun 2023 20:02:26 +0000 (13:02 -0700)]
iavf: remove mask from iavf_irq_enable_queues()

Enable more than 32 IRQs by removing the u32 bit mask in
iavf_irq_enable_queues(). There is no need for the mask as there are no
callers that select individual IRQs through the bitmask. Also, if the PF
allocates more than 32 IRQs, this mask will prevent us from using all of
them.

Modify the comment in iavf_register.h to show that the maximum number
allowed for the IRQ index is 63 as per the iAVF standard 1.0 [1].

link: [1] https://www.intel.com/content/dam/www/public/us/en/documents/product-specifications/ethernet-adaptive-virtual-function-hardware-spec.pdf
Fixes: 5eae00c57f5e ("i40evf: main driver core")
Signed-off-by: Ahmed Zaki <ahmed.zaki@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://lore.kernel.org/r/20230608200226.451861-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoMerge branch 'selftests-mptcp-skip-tests-not-supported-by-old-kernels-part-2'
Jakub Kicinski [Sat, 10 Jun 2023 07:02:17 +0000 (00:02 -0700)]
Merge branch 'selftests-mptcp-skip-tests-not-supported-by-old-kernels-part-2'

Matthieu Baerts says:

====================
selftests: mptcp: skip tests not supported by old kernels (part 2)

After a few years of increasing test coverage in the MPTCP selftests, we
realised [1] the last version of the selftests is supposed to run on old
kernels without issues.

Supporting older versions is not that easy for this MPTCP case: these
selftests are often validating the internals by checking packets that
are exchanged, when some MIB counters are incremented after some
actions, how connections are getting opened and closed in some cases,
etc. In other words, it is not limited to the socket interface between
the userspace and the kernelspace.

In addition to that, the current MPTCP selftests run a lot of different
sub-tests but the TAP13 protocol used in the selftests don't support
sub-tests: one failure in sub-tests implies that the whole selftest is
seen as failed at the end because sub-tests are not tracked. It is then
important to skip sub-tests not supported by old kernels.

To minimise the modifications and reduce the complexity to support old
versions, the idea is to look at external signs and skip the whole
selftests or just some sub-tests before starting them. This cannot be
applied in all cases.

This second part focuses on marking different sub-tests as skipped if
some MPTCP features are not supported. A few techniques are used here:

- Before starting some tests:

  - Check if a file (sysctl knob) is present: that's what patch 13/14 is
    doing for the userspace PM feature.

  - Check if a symbol is present in /proc/kallsyms: patch 1/14 adds some
    helpers in mptcp_lib.sh to ease its use. Then these helpers are used
    in patches 2, 3, 4, 10, 11 and 14/14.

  - Set a flag and get the status to check if a feature is supported:
    patch 8/14 is doing that with the 'fullmesh' flag.

- After having launched the tests:

  - Retrieve the counters after a test and check if they are different
    than 0. Similar to the check with the flag, that's not ideal but in
    this case, the counters were already present before the introduction
    of MPTCP but they have been supported by MPTCP sockets only later.
    Patches 5 and 6/14 are using this technique.

Before skipping tests, SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var
value is checked: if it is set to 1, the test is marked as "failed"
instead of "skipped". MPTCP public CI expects to have all features
supported and it sets this env var to 1 to catch regressions in these
new checks.

Patches 7/14 and 9/14 are a bit different because they don't skip tests:

- Patch 7/14 retrieves the default values instead of using hardcoded
  ones because these default values have been modified at some points.
  Then the comparisons are done with the default values.

- patch 9/14 relaxes the expected returned size from MPTCP's getsockopt
  because the different structures gathering various info can get new
  fields and get bigger over time. We cannot expect that the userspace
  is using the same structure as the kernel.

Patch 12/14 marks the test as "skipped" instead of "failed" if the "ip"
tool is not available.

In this second part, the "mptcp_join" selftest is not modified yet. This
will come soon after in the third part with quite a few patches.

Link: https://lore.kernel.org/stable/CA+G9fYtDGpgT4dckXD-y-N92nqUxuvue_7AtDdBcHrbOMsDZLg@mail.gmail.com/
Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
====================

Link: https://lore.kernel.org/r/20230608-upstream-net-20230608-mptcp-selftests-support-old-kernels-part-2-v1-0-20997a6fd841@tessares.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: userspace pm: skip PM listener events tests if unavailable
Matthieu Baerts [Thu, 8 Jun 2023 16:38:56 +0000 (18:38 +0200)]
selftests: mptcp: userspace pm: skip PM listener events tests if unavailable

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the new listener events linked to the path-manager
introduced by commit f8c9dfbd875b ("mptcp: add pm listener events").

It is possible to look for "mptcp_event_pm_listener" in kallsyms to know
in advance if the kernel supports this feature and skip these sub-tests
if the feature is not supported.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 6c73008aa301 ("selftests: mptcp: listener test for userspace PM")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: userspace pm: skip if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:55 +0000 (18:38 +0200)]
selftests: mptcp: userspace pm: skip if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the MPTCP Userspace PM introduced by commit 4638de5aefe5
("mptcp: handle local addrs announced by userspace PMs").

We can skip all these tests if the feature is not supported simply by
looking for the MPTCP pm_type's sysctl knob.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 259a834fadda ("selftests: mptcp: functional tests for the userspace PM type")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: userspace pm: skip if 'ip' tool is unavailable
Matthieu Baerts [Thu, 8 Jun 2023 16:38:54 +0000 (18:38 +0200)]
selftests: mptcp: userspace pm: skip if 'ip' tool is unavailable

When a required tool is missing, the return code 4 (SKIP) should be
returned instead of 1 (FAIL).

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 259a834fadda ("selftests: mptcp: functional tests for the userspace PM type")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: sockopt: skip TCP_INQ checks if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:53 +0000 (18:38 +0200)]
selftests: mptcp: sockopt: skip TCP_INQ checks if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is TCP_INQ cmsg support introduced in commit 2c9e77659a0c
("mptcp: add TCP_INQ cmsg support").

It is possible to look for "mptcp_ioctl" in kallsyms because it was
needed to introduce the mentioned feature. We can skip these tests and
not set TCPINQ option if the feature is not supported.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 5cbd886ce2a9 ("selftests: mptcp: add TCP_INQ support")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: sockopt: skip getsockopt checks if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:52 +0000 (18:38 +0200)]
selftests: mptcp: sockopt: skip getsockopt checks if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the getsockopt(SOL_MPTCP) to get info about the MPTCP
connections introduced by commit 55c42fa7fa33 ("mptcp: add MPTCP_INFO
getsockopt") and the following ones.

It is possible to look for "mptcp_diag_fill_info" in kallsyms because
it is introduced by the mentioned feature. So we can know in advance if
the feature is supported and skip the sub-test if not.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: ce9979129a0b ("selftests: mptcp: add mptcp getsockopt test cases")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: sockopt: relax expected returned size
Matthieu Baerts [Thu, 8 Jun 2023 16:38:51 +0000 (18:38 +0200)]
selftests: mptcp: sockopt: relax expected returned size

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the getsockopt(SOL_MPTCP) to get info about the MPTCP
connections introduced by commit 55c42fa7fa33 ("mptcp: add MPTCP_INFO
getsockopt") and the following ones.

We cannot guess in advance which sizes the kernel will returned: older
kernel can returned smaller sizes, e.g. recently the tcp_info structure
has been modified in commit 71fc704768f6 ("tcp: add rcv_wnd and
plb_rehash to TCP_INFO") where a new field has been added.

The userspace can also expect a smaller size if it is compiled with old
uAPI kernel headers.

So for these sizes, we can only check if they are above a certain
threshold, 0 for the moment. We can also only compared sizes with the
ones set by the kernel.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: ce9979129a0b ("selftests: mptcp: add mptcp getsockopt test cases")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: pm nl: skip fullmesh flag checks if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:50 +0000 (18:38 +0200)]
selftests: mptcp: pm nl: skip fullmesh flag checks if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the fullmesh flag that can be given to the MPTCP
in-kernel path-manager and introduced in commit 2843ff6f36db ("mptcp:
remote addresses fullmesh").

If the flag is not visible in the dump after having set it, we don't
check the content. Note that if we expect to have this feature and
SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var is set to 1, we always
check the content to avoid regressions.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 6da1dfdd037e ("selftests: mptcp: add set_flags tests in pm_netlink.sh")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: pm nl: remove hardcoded default limits
Matthieu Baerts [Thu, 8 Jun 2023 16:38:49 +0000 (18:38 +0200)]
selftests: mptcp: pm nl: remove hardcoded default limits

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the checks of the default limits returned by the MPTCP
in-kernel path-manager. The default values have been modified by commit
72bcbc46a5c3 ("mptcp: increase default max additional subflows to 2").
Instead of comparing with hardcoded values, we can get the default one
and compare with them.

Note that if we expect to have the latest version, we continue to check
the hardcoded values to avoid unexpected behaviour changes.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: eedbc685321b ("selftests: add PM netlink functional tests")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: diag: skip inuse tests if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:48 +0000 (18:38 +0200)]
selftests: mptcp: diag: skip inuse tests if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the reporting of the MPTCP sockets being used, introduced
by commit c558246ee73e ("mptcp: add statistics for mptcp socket in use").

Similar to the parent commit, it looks like there is no good pre-check
to do here, i.e. dedicated function available in kallsyms. Instead, we
try to get info and if nothing is returned, the test is marked as
skipped.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: e04a30f78809 ("selftest: mptcp: add test for mptcp socket in use")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: diag: skip listen tests if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:47 +0000 (18:38 +0200)]
selftests: mptcp: diag: skip listen tests if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the listen diag dump support introduced by
commit 4fa39b701ce9 ("mptcp: listen diag dump support").

It looks like there is no good pre-check to do here, i.e. dedicated
function available in kallsyms. Instead, we try to get info if nothing
is returned, the test is marked as skipped.

That's not ideal because something could be wrong with the feature and
instead of reporting an error, the test could be marked as skipped. If
we know in advanced that the feature is supposed to be supported, the
tester can set SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var to 1: in
this case the test will report an error instead of marking the test as
skipped if nothing is returned.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: f2ae0fa68e28 ("selftests/mptcp: add diag listen tests")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: connect: skip TFO tests if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:46 +0000 (18:38 +0200)]
selftests: mptcp: connect: skip TFO tests if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the support of TCP_FASTOPEN socket option with MPTCP
connections introduced by commit 4ffb0a02346c ("mptcp: add TCP_FASTOPEN
sock option").

It is possible to look for "mptcp_fastopen_" in kallsyms to know if the
feature is supported or not.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: ca7ae8916043 ("selftests: mptcp: mptfo Initiator/Listener")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: connect: skip disconnect tests if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:45 +0000 (18:38 +0200)]
selftests: mptcp: connect: skip disconnect tests if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the full support of disconnections from the userspace
introduced by commit b29fcfb54cd7 ("mptcp: full disconnect
implementation").

It is possible to look for "mptcp_pm_data_reset" in kallsyms because a
preparation patch added it to ease the introduction of the mentioned
feature.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 05be5e273c84 ("selftests: mptcp: add disconnect tests")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: connect: skip transp tests if not supported
Matthieu Baerts [Thu, 8 Jun 2023 16:38:44 +0000 (18:38 +0200)]
selftests: mptcp: connect: skip transp tests if not supported

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

One of them is the support of IP(V6)_TRANSPARENT socket option with
MPTCP connections introduced by commit c9406a23c116 ("mptcp: sockopt:
add SOL_IP freebind & transparent options").

It is possible to look for "__ip_sock_set_tos" in kallsyms because
IP(V6)_TRANSPARENT socket option support has been added after TOS
support which came with the required infrastructure in MPTCP sockopt
code. To support TOS, the following function has been exported (T). Not
great but better than checking for a specific kernel version.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 5fb62e9cd3ad ("selftests: mptcp: add tproxy test case")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoselftests: mptcp: lib: skip if missing symbol
Matthieu Baerts [Thu, 8 Jun 2023 16:38:43 +0000 (18:38 +0200)]
selftests: mptcp: lib: skip if missing symbol

Selftests are supposed to run on any kernels, including the old ones not
supporting all MPTCP features.

New functions are now available to easily detect if a certain feature is
missing by looking at kallsyms.

These new helpers are going to be used in the following commits. In
order to ease the backport of such future patches, it would be good if
this patch is backported up to the introduction of MPTCP selftests,
hence the Fixes tag below: this type of check was supposed to be done
from the beginning.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 048d19d444be ("mptcp: add basic kselftest for mptcp")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonet/sched: cls_u32: Fix reference counter leak leading to overflow
Lee Jones [Thu, 8 Jun 2023 07:29:03 +0000 (08:29 +0100)]
net/sched: cls_u32: Fix reference counter leak leading to overflow

In the event of a failure in tcf_change_indev(), u32_set_parms() will
immediately return without decrementing the recently incremented
reference counter.  If this happens enough times, the counter will
rollover and the reference freed, leading to a double free which can be
used to do 'bad things'.

In order to prevent this, move the point of possible failure above the
point where the reference counter is incremented.  Also save any
meaningful return values to be applied to the return data at the
appropriate point in time.

This issue was caught with KASAN.

Fixes: 705c7091262d ("net: sched: cls_u32: no need to call tcf_exts_change for newly allocated struct")
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Lee Jones <lee@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet/sched: taprio: fix slab-out-of-bounds Read in taprio_dequeue_from_txq
Zhengchao Shao [Thu, 8 Jun 2023 06:27:56 +0000 (14:27 +0800)]
net/sched: taprio: fix slab-out-of-bounds Read in taprio_dequeue_from_txq

As shown in [1], out-of-bounds access occurs in two cases:
1)when the qdisc of the taprio type is used to replace the previously
configured taprio, count and offset in tc_to_txq can be set to 0. In this
case, the value of *txq in taprio_next_tc_txq() will increases
continuously. When the number of accessed queues exceeds the number of
queues on the device, out-of-bounds access occurs.
2)When packets are dequeued, taprio can be deleted. In this case, the tc
rule of dev is cleared. The count and offset values are also set to 0. In
this case, out-of-bounds access is also caused.

Now the restriction on the queue number is added.

[1] https://groups.google.com/g/syzkaller-bugs/c/_lYOKgkBVMg
Fixes: 2f530df76c8c ("net/sched: taprio: give higher priority to higher TCs in software dequeue mode")
Reported-by: syzbot+04afcb3d2c840447559a@syzkaller.appspotmail.com
Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
Tested-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoocteontx2-af: Fix promiscuous mode
Ratheesh Kannoth [Thu, 8 Jun 2023 05:16:25 +0000 (10:46 +0530)]
octeontx2-af: Fix promiscuous mode

CN10KB silicon introduced a new exact match feature,
which is used for DMAC filtering. The state of installed
DMAC filters in this exact match table is getting corrupted
when promiscuous mode is toggled. Fix this by not touching
Exact match related config when promiscuous mode is toggled.

Fixes: 2dba9459d2c9 ("octeontx2-af: Wrapper functions for MAC addr add/del/update/reset")
Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: renesas: rswitch: Fix timestamp feature after all descriptors are used
Yoshihiro Shimoda [Thu, 8 Jun 2023 01:57:27 +0000 (10:57 +0900)]
net: renesas: rswitch: Fix timestamp feature after all descriptors are used

The timestamp descriptors were intended to act cyclically. Descriptors
from index 0 through gq->ring_size - 1 contain actual information, and
the last index (gq->ring_size) should have LINKFIX to indicate
the first index 0 descriptor. However, the LINKFIX value is missing,
causing the timestamp feature to stop after all descriptors are used.
To resolve this issue, set the LINKFIX to the timestamp descritors.

Reported-by: Phong Hoang <phong.hoang.wz@renesas.com>
Fixes: 33f5d733b589 ("net: renesas: rswitch: Improve TX timestamp accuracy")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet/sched: act_pedit: Parse L3 Header for L4 offset
Max Tottenham [Wed, 7 Jun 2023 16:23:54 +0000 (12:23 -0400)]
net/sched: act_pedit: Parse L3 Header for L4 offset

Instead of relying on skb->transport_header being set correctly, opt
instead to parse the L3 header length out of the L3 headers for both
IPv4/IPv6 when the Extended Layer Op for tcp/udp is used. This fixes a
bug if GRO is disabled, when GRO is disabled skb->transport_header is
set by __netif_receive_skb_core() to point to the L3 header, it's later
fixed by the upper protocol layers, but act_pedit will receive the SKB
before the fixups are completed. The existing behavior causes the
following to edit the L3 header if GRO is disabled instead of the UDP
header:

    tc filter add dev eth0 ingress protocol ip flower ip_proto udp \
 dst_ip 192.168.1.3 action pedit ex munge udp set dport 18053

Also re-introduce a rate-limited warning if we were unable to extract
the header offset when using the 'ex' interface.

Fixes: 71d0ed7079df ("net/act_pedit: Support using offset relative to
the conventional network headers")
Signed-off-by: Max Tottenham <mtottenh@akamai.com>
Reviewed-by: Josh Hunt <johunt@akamai.com>
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202305261541.N165u9TZ-lkp@intel.com/
Reviewed-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: usb: qmi_wwan: add support for Compal RXM-G1
Wes Huang [Thu, 8 Jun 2023 03:01:42 +0000 (11:01 +0800)]
net: usb: qmi_wwan: add support for Compal RXM-G1

Add support for Compal RXM-G1 which is based on Qualcomm SDX55 chip.
This patch adds support for two compositions:

0x9091: DIAG + MODEM + QMI_RMNET + ADB
0x90db: DIAG + DUN + RMNET + DPL + QDSS(Trace) + ADB

T:  Bus=03 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#=  2 Spd=5000 MxCh= 0
D:  Ver= 3.20 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs=  1
P:  Vendor=05c6 ProdID=9091 Rev= 4.14
S:  Manufacturer=QCOM
S:  Product=SDXPRAIRIE-MTP _SN:719AB680
S:  SerialNumber=719ab680
C:* #Ifs= 4 Cfg#= 1 Atr=80 MxPwr=896mA
I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=(none)
E:  Ad=81(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=01(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=(none)
E:  Ad=83(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
E:  Ad=82(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan
E:  Ad=84(I) Atr=03(Int.) MxPS=   8 Ivl=32ms
E:  Ad=8e(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=0f(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=42 Prot=01 Driver=(none)
E:  Ad=03(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=85(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms

T:  Bus=03 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#=  2 Spd=5000 MxCh= 0
D:  Ver= 3.20 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs=  1
P:  Vendor=05c6 ProdID=90db Rev= 4.14
S:  Manufacturer=QCOM
S:  Product=SDXPRAIRIE-MTP _SN:719AB680
S:  SerialNumber=719ab680
C:* #Ifs= 6 Cfg#= 1 Atr=80 MxPwr=896mA
I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=(none)
E:  Ad=81(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=01(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=(none)
E:  Ad=83(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
E:  Ad=82(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan
E:  Ad=84(I) Atr=03(Int.) MxPS=   8 Ivl=32ms
E:  Ad=8e(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=0f(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
E:  Ad=8f(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 4 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
E:  Ad=85(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
I:* If#= 5 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=42 Prot=01 Driver=(none)
E:  Ad=03(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E:  Ad=86(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms

Cc: stable@vger.kernel.org
Signed-off-by: Wes Huang <wes.huang@moxa.com>
Acked-by: Bjørn Mork <bjorn@mork.no>
Link: https://lore.kernel.org/r/20230608030141.3546-1-wes.huang@moxa.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoigb: Fix extts capture value format for 82580/i354/i350
Yuezhen Luan [Wed, 7 Jun 2023 16:41:16 +0000 (09:41 -0700)]
igb: Fix extts capture value format for 82580/i354/i350

82580/i354/i350 features circle-counter-like timestamp registers
that are different with newer i210. The EXTTS capture value in
AUXTSMPx should be converted from raw circle counter value to
timestamp value in resolution of 1 nanosec by the driver.

This issue can be reproduced on i350 nics, connecting an 1PPS
signal to a SDP pin, and run 'ts2phc' command to read external
1PPS timestamp value. On i210 this works fine, but on i350 the
extts is not correctly converted.

The i350/i354/82580's SYSTIM and other timestamp registers are
40bit counters, presenting time range of 2^40 ns, that means these
registers overflows every about 1099s. This causes all these regs
can't be used directly in contrast to the newer i210/i211s.

The igb driver needs to convert these raw register values to
valid time stamp format by using kernel timecounter apis for i350s
families. Here the igb_extts() just forgot to do the convert.

Fixes: 38970eac41db ("igb: support EXTTS on 82580/i354/i350")
Signed-off-by: Yuezhen Luan <eggcar.luan@gmail.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230607164116.3768175-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoping6: Fix send to link-local addresses with VRF.
Guillaume Nault [Wed, 7 Jun 2023 16:05:02 +0000 (18:05 +0200)]
ping6: Fix send to link-local addresses with VRF.

Ping sockets can't send packets when they're bound to a VRF master
device and the output interface is set to a slave device.

For example, when net.ipv4.ping_group_range is properly set, so that
ping6 can use ping sockets, the following kind of commands fails:
  $ ip vrf exec red ping6 fe80::854:e7ff:fe88:4bf1%eth1

What happens is that sk->sk_bound_dev_if is set to the VRF master
device, but 'oif' is set to the real output device. Since both are set
but different, ping_v6_sendmsg() sees their value as inconsistent and
fails.

Fix this by allowing 'oif' to be a slave device of ->sk_bound_dev_if.

This fixes the following kselftest failure:
  $ ./fcnal-test.sh -t ipv6_ping
  [...]
  TEST: ping out, vrf device+address bind - ns-B IPv6 LLA        [FAIL]

Reported-by: Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr>
Closes: https://lore.kernel.org/netdev/b6191f90-ffca-dbca-7d06-88a9788def9c@alu.unizg.hr/
Tested-by: Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr>
Fixes: 5e457896986e ("net: ipv6: Fix ping to link-local addresses.")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/6c8b53108816a8d0d5705ae37bdc5a8322b5e3d9.1686153846.git.gnault@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonet: enetc: correct the indexes of highest and 2nd highest TCs
Wei Fang [Wed, 7 Jun 2023 09:10:48 +0000 (17:10 +0800)]
net: enetc: correct the indexes of highest and 2nd highest TCs

For ENETC hardware, the TCs are numbered from 0 to N-1, where N
is the number of TCs. Numerically higher TC has higher priority.
It's obvious that the highest priority TC index should be N-1 and
the 2nd highest priority TC index should be N-2.

However, the previous logic uses netdev_get_prio_tc_map() to get
the indexes of highest priority and 2nd highest priority TCs, it
does not make sense and is incorrect to give a "tc" argument to
netdev_get_prio_tc_map(). So the driver may get the wrong indexes
of the two highest priotiry TCs which would lead to failed to set
the CBS for the two highest priotiry TCs.

e.g.
$ tc qdisc add dev eno0 parent root handle 100: mqprio num_tc 6 \
map 0 0 1 1 2 3 4 5 queues 1@0 1@1 1@2 1@3 2@4 2@6 hw 1
$ tc qdisc replace dev eno0 parent 100:6 cbs idleslope 100000 \
sendslope -900000 hicredit 12 locredit -113 offload 1
$ Error: Specified device failed to setup cbs hardware offload.
  ^^^^^

In this example, the previous logic deems the indexes of the two
highest priotiry TCs should be 3 and 2. Actually, the indexes are
5 and 4, because the number of TCs is 6. So it would be failed to
configure the CBS for the two highest priority TCs.

Fixes: c431047c4efe ("enetc: add support Credit Based Shaper(CBS) for hardware offload")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonetfilter: nf_tables: incorrect error path handling with NFT_MSG_NEWRULE
Pablo Neira Ayuso [Thu, 8 Jun 2023 00:32:02 +0000 (02:32 +0200)]
netfilter: nf_tables: incorrect error path handling with NFT_MSG_NEWRULE

In case of error when adding a new rule that refers to an anonymous set,
deactivate expressions via NFT_TRANS_PREPARE state, not NFT_TRANS_RELEASE.
Thus, the lookup expression marks anonymous sets as inactive in the next
generation to ensure it is not reachable in this transaction anymore and
decrement the set refcount as introduced by c1592a89942e ("netfilter:
nf_tables: deactivate anonymous set from preparation phase"). The abort
step takes care of undoing the anonymous set.

This is also consistent with rule deletion, where NFT_TRANS_PREPARE is
used. Note that this error path is exercised in the preparation step of
the commit protocol. This patch replaces nf_tables_rule_release() by the
deactivate and destroy calls, this time with NFT_TRANS_PREPARE.

Due to this incorrect error handling, it is possible to access a
dangling pointer to the anonymous set that remains in the transaction
list.

[1009.379054] BUG: KASAN: use-after-free in nft_set_lookup_global+0x147/0x1a0 [nf_tables]
[1009.379106] Read of size 8 at addr ffff88816c4c8020 by task nft-rule-add/137110
[1009.379116] CPU: 7 PID: 137110 Comm: nft-rule-add Not tainted 6.4.0-rc4+ #256
[1009.379128] Call Trace:
[1009.379132]  <TASK>
[1009.379135]  dump_stack_lvl+0x33/0x50
[1009.379146]  ? nft_set_lookup_global+0x147/0x1a0 [nf_tables]
[1009.379191]  print_address_description.constprop.0+0x27/0x300
[1009.379201]  kasan_report+0x107/0x120
[1009.379210]  ? nft_set_lookup_global+0x147/0x1a0 [nf_tables]
[1009.379255]  nft_set_lookup_global+0x147/0x1a0 [nf_tables]
[1009.379302]  nft_lookup_init+0xa5/0x270 [nf_tables]
[1009.379350]  nf_tables_newrule+0x698/0xe50 [nf_tables]
[1009.379397]  ? nf_tables_rule_release+0xe0/0xe0 [nf_tables]
[1009.379441]  ? kasan_unpoison+0x23/0x50
[1009.379450]  nfnetlink_rcv_batch+0x97c/0xd90 [nfnetlink]
[1009.379470]  ? nfnetlink_rcv_msg+0x480/0x480 [nfnetlink]
[1009.379485]  ? __alloc_skb+0xb8/0x1e0
[1009.379493]  ? __alloc_skb+0xb8/0x1e0
[1009.379502]  ? entry_SYSCALL_64_after_hwframe+0x46/0xb0
[1009.379509]  ? unwind_get_return_address+0x2a/0x40
[1009.379517]  ? write_profile+0xc0/0xc0
[1009.379524]  ? avc_lookup+0x8f/0xc0
[1009.379532]  ? __rcu_read_unlock+0x43/0x60

Fixes: 958bee14d071 ("netfilter: nf_tables: use new transaction infrastructure to handle sets")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agoice: Fix XDP memory leak when NIC is brought up and down
Kamil Maziarz [Tue, 6 Jun 2023 10:33:58 +0000 (12:33 +0200)]
ice: Fix XDP memory leak when NIC is brought up and down

Fix the buffer leak that occurs while switching
the port up and down with traffic and XDP by
checking for an active XDP program and freeing all empty TX buffers.

Fixes: efc2214b6047 ("ice: Add support for XDP")
Signed-off-by: Kamil Maziarz <kamil.maziarz@intel.com>
Tested-by: Chandan Kumar Rout <chandanx.rout@intel.com> (A Contingent Worker at Intel)
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
12 months agoMerge tag 'net-6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 8 Jun 2023 16:27:19 +0000 (09:27 -0700)]
Merge tag 'net-6.4-rc6' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Paolo Abeni:
 "Including fixes from can, wifi, netfilter, bluetooth and ebpf.

  Current release - regressions:

   - bpf: sockmap: avoid potential NULL dereference in
     sk_psock_verdict_data_ready()

   - wifi: iwlwifi: fix -Warray-bounds bug in iwl_mvm_wait_d3_notif()

   - phylink: actually fix ksettings_set() ethtool call

   - eth: dwmac-qcom-ethqos: fix a regression on EMAC < 3

  Current release - new code bugs:

   - wifi: mt76: fix possible NULL pointer dereference in
     mt7996_mac_write_txwi()

  Previous releases - regressions:

   - netfilter: fix NULL pointer dereference in nf_confirm_cthelper

   - wifi: rtw88/rtw89: correct PS calculation for SUPPORTS_DYNAMIC_PS

   - openvswitch: fix upcall counter access before allocation

   - bluetooth:
      - fix use-after-free in hci_remove_ltk/hci_remove_irk
      - fix l2cap_disconnect_req deadlock

   - nic: bnxt_en: prevent kernel panic when receiving unexpected
     PHC_UPDATE event

  Previous releases - always broken:

   - core: annotate rfs lockless accesses

   - sched: fq_pie: ensure reasonable TCA_FQ_PIE_QUANTUM values

   - netfilter: add null check for nla_nest_start_noflag() in
     nft_dump_basechain_hook()

   - bpf: fix UAF in task local storage

   - ipv4: ping_group_range: allow GID from 2147483648 to 4294967294

   - ipv6: rpl: fix route of death.

   - tcp: gso: really support BIG TCP

   - mptcp: fixes for user-space PM address advertisement

   - smc: avoid to access invalid RMBs' MRs in SMCRv1 ADD LINK CONT

   - can: avoid possible use-after-free when j1939_can_rx_register fails

   - batman-adv: fix UaF while rescheduling delayed work

   - eth: qede: fix scheduling while atomic

   - eth: ice: make writes to /dev/gnssX synchronous"

* tag 'net-6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (83 commits)
  bnxt_en: Implement .set_port / .unset_port UDP tunnel callbacks
  bnxt_en: Prevent kernel panic when receiving unexpected PHC_UPDATE event
  bnxt_en: Skip firmware fatal error recovery if chip is not accessible
  bnxt_en: Query default VLAN before VNIC setup on a VF
  bnxt_en: Don't issue AP reset during ethtool's reset operation
  bnxt_en: Fix bnxt_hwrm_update_rss_hash_cfg()
  net: bcmgenet: Fix EEE implementation
  eth: ixgbe: fix the wake condition
  eth: bnxt: fix the wake condition
  lib: cpu_rmap: Fix potential use-after-free in irq_cpu_rmap_release()
  bpf: Add extra path pointer check to d_path helper
  net: sched: fix possible refcount leak in tc_chain_tmplt_add()
  net: sched: act_police: fix sparse errors in tcf_police_dump()
  net: openvswitch: fix upcall counter access before allocation
  net: sched: move rtm_tca_policy declaration to include file
  ice: make writes to /dev/gnssX synchronous
  net: sched: add rcu annotations around qdisc->qdisc_sleeping
  rfs: annotate lockless accesses to RFS sock flow table
  rfs: annotate lockless accesses to sk->sk_rxhash
  virtio_net: use control_buf for coalesce params
  ...

12 months agoMerge tag 'xfs-6.4-rc5-fixes' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Linus Torvalds [Thu, 8 Jun 2023 15:46:58 +0000 (08:46 -0700)]
Merge tag 'xfs-6.4-rc5-fixes' of git://git./fs/xfs/xfs-linux

Pull xfs fixes from Dave Chinner:
 "These are a set of regression fixes discovered on recent kernels. I
  was hoping to send this to you a week and half ago, but events out of
  my control delayed finalising the changes until early this week.

  Whilst the diffstat looks large for this stage of the merge window, a
  large chunk of it comes from moving the guts of one function from one
  file to another i.e. it's the same code, it is just run in a different
  context where it is safe to hold a specific lock. Otherwise the
  individual changes are relatively small and straigtht forward.

  Summary:

   - Propagate unlinked inode list corruption back up to log recovery
     (regression fix)

   - improve corruption detection for AGFL entries, AGFL indexes and
     XEFI extents (syzkaller fuzzer oops report)

   - Avoid double perag reference release (regression fix)

   - Improve extent merging detection in scrub (regression fix)

   - Fix a new undefined high bit shift (regression fix)

   - Fix for AGF vs inode cluster buffer deadlock (regression fix)"

* tag 'xfs-6.4-rc5-fixes' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  xfs: collect errors from inodegc for unlinked inode recovery
  xfs: validate block number being freed before adding to xefi
  xfs: validity check agbnos on the AGFL
  xfs: fix agf/agfl verification on v4 filesystems
  xfs: fix double xfs_perag_rele() in xfs_filestream_pick_ag()
  xfs: fix broken logic when detecting mergeable bmap records
  xfs: Fix undefined behavior of shift into sign bit
  xfs: fix AGF vs inode cluster buffer deadlock
  xfs: defered work could create precommits
  xfs: restore allocation trylock iteration
  xfs: buffer pins need to hold a buffer reference

12 months agoice: Don't dereference NULL in ice_gnss_read error path
Simon Horman [Thu, 25 May 2023 10:52:58 +0000 (12:52 +0200)]
ice: Don't dereference NULL in ice_gnss_read error path

If pf is NULL in ice_gnss_read() then it will be dereferenced
in the error path by a call to dev_dbg(ice_pf_to_dev(pf), ...).

Avoid this by simply returning in this case.
If logging is desired an alternate approach might be to
use pr_err() before returning.

Flagged by Smatch as:

  .../ice_gnss.c:196 ice_gnss_read() error: we previously assumed 'pf' could be null (see line 131)

Fixes: 43113ff73453 ("ice: add TTY for GNSS module for E810T device")
Signed-off-by: Simon Horman <horms@kernel.org>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Tested-by: Sunitha Mekala <sunithax.d.mekala@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
12 months agoMerge branch 'bnxt_en-bug-fixes'
Paolo Abeni [Thu, 8 Jun 2023 08:52:48 +0000 (10:52 +0200)]
Merge branch 'bnxt_en-bug-fixes'

Michael Chan says:

====================
bnxt_en: Bug fixes

This patchset has the following fixes for bnxt_en:

1. Add missing VNIC ID parameter in the FW message when getting an
updated RSS configuration from the FW.

2. Fix a warning when doing ethtool reset on newer chips.

3. Fix VLAN issue on a VF when a default VLAN is assigned.

4. Fix a problem during DPC (Downstream Port containment) scenario.

5. Fix a NULL pointer dereference when receiving a PTP event from FW.

6. Fix VXLAN/Geneve UDP port delete/add with newer FW.
====================

Link: https://lore.kernel.org/r/20230607075409.228450-1-michael.chan@broadcom.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Implement .set_port / .unset_port UDP tunnel callbacks
Somnath Kotur [Wed, 7 Jun 2023 07:54:09 +0000 (00:54 -0700)]
bnxt_en: Implement .set_port / .unset_port UDP tunnel callbacks

As per the new udp tunnel framework, drivers which need to know the
details of a port entry (i.e. port type) when it gets deleted should
use the .set_port / .unset_port callbacks.

Implementing the current .udp_tunnel_sync callback would mean that the
deleted tunnel port entry would be all zeros.  This used to work on
older firmware because it would not check the input when deleting a
tunnel port.  With newer firmware, the delete will now fail and
subsequent tunnel port allocation will fail as a result.

Fixes: 442a35a5a7aa ("bnxt: convert to new udp_tunnel_nic infra")
Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Prevent kernel panic when receiving unexpected PHC_UPDATE event
Pavan Chebbi [Wed, 7 Jun 2023 07:54:08 +0000 (00:54 -0700)]
bnxt_en: Prevent kernel panic when receiving unexpected PHC_UPDATE event

The firmware can send PHC_RTC_UPDATE async event on a PF that may not
have PTP registered. In such a case, there will be a null pointer
deference for bp->ptp_cfg when we try to handle the event.

Fix it by not registering for this event with the firmware if !bp->ptp_cfg.
Also, check that bp->ptp_cfg is valid before proceeding when we receive
the event.

Fixes: 8bcf6f04d4a5 ("bnxt_en: Handle async event when the PHC is updated in RTC mode")
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Skip firmware fatal error recovery if chip is not accessible
Vikas Gupta [Wed, 7 Jun 2023 07:54:07 +0000 (00:54 -0700)]
bnxt_en: Skip firmware fatal error recovery if chip is not accessible

Driver starts firmware fatal error recovery by detecting
heartbeat failure or fw reset count register changing.  But
these checks are not reliable if the device is not accessible.
This can happen while DPC (Downstream Port containment) is in
progress.  Skip firmware fatal recovery if pci_device_is_present()
returns false.

Fixes: acfb50e4e773 ("bnxt_en: Add FW fatal devlink_health_reporter.")
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Vikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Query default VLAN before VNIC setup on a VF
Somnath Kotur [Wed, 7 Jun 2023 07:54:06 +0000 (00:54 -0700)]
bnxt_en: Query default VLAN before VNIC setup on a VF

We need to call bnxt_hwrm_func_qcfg() on a VF to query the default
VLAN that may be setup by the PF.  If a default VLAN is enabled,
the VF cannot support VLAN acceleration on the receive side and
the VNIC must be setup to strip out the default VLAN tag.  If a
default VLAN is not enabled, the VF can support VLAN acceleration
on the receive side.  The VNIC should be set up to strip or not
strip the VLAN based on the RX VLAN acceleration setting.

Without this call to determine the default VLAN before calling
bnxt_setup_vnic(), the VNIC may not be set up correctly.  For
example, bnxt_setup_vnic() may set up to strip the VLAN tag based
on stale default VLAN information.  If RX VLAN acceleration is
not enabled, the VLAN tag will be incorrectly stripped and the
RX data path will not work correctly.

Fixes: cf6645f8ebc6 ("bnxt_en: Add function for VF driver to query default VLAN.")
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Don't issue AP reset during ethtool's reset operation
Sreekanth Reddy [Wed, 7 Jun 2023 07:54:05 +0000 (00:54 -0700)]
bnxt_en: Don't issue AP reset during ethtool's reset operation

Only older NIC controller's firmware uses the PROC AP reset type.
Firmware on 5731X/5741X and newer chips does not support this reset
type.  When bnxt_reset() issues a series of resets, this PROC AP
reset may actually fail on these newer chips because the firmware
is not ready to accept this unsupported command yet.  Avoid this
unnecessary error by skipping this reset type on chips that don't
support it.

Fixes: 7a13240e3718 ("bnxt_en: fix ethtool_reset_flags ABI violations")
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Sreekanth Reddy <sreekanth.reddy@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agobnxt_en: Fix bnxt_hwrm_update_rss_hash_cfg()
Pavan Chebbi [Wed, 7 Jun 2023 07:54:04 +0000 (00:54 -0700)]
bnxt_en: Fix bnxt_hwrm_update_rss_hash_cfg()

We must specify the vnic id of the vnic in the input structure of this
firmware message.  Otherwise we will get an error from the firmware.

Fixes: 98a4322b70e8 ("bnxt_en: update RSS config using difference algorithm")
Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agoMerge tag 'batadv-net-pullrequest-20230607' of git://git.open-mesh.org/linux-merge
Jakub Kicinski [Thu, 8 Jun 2023 04:56:01 +0000 (21:56 -0700)]
Merge tag 'batadv-net-pullrequest-20230607' of git://git.open-mesh.org/linux-merge

Simon Wunderlich says:

====================
Here is a batman-adv bugfix:

 - fix a broken sync while rescheduling delayed work,
   by Vladislav Efanov

* tag 'batadv-net-pullrequest-20230607' of git://git.open-mesh.org/linux-merge:
  batman-adv: Broken sync while rescheduling delayed work
====================

Link: https://lore.kernel.org/r/20230607155515.548120-1-sw@simonwunderlich.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonet: bcmgenet: Fix EEE implementation
Florian Fainelli [Tue, 6 Jun 2023 21:43:47 +0000 (14:43 -0700)]
net: bcmgenet: Fix EEE implementation

We had a number of short comings:

- EEE must be re-evaluated whenever the state machine detects a link
  change as wight be switching from a link partner with EEE
  enabled/disabled

- tx_lpi_enabled controls whether EEE should be enabled/disabled for the
  transmit path, which applies to the TBUF block

- We do not need to forcibly enable EEE upon system resume, as the PHY
  state machine will trigger a link event that will do that, too

Fixes: 6ef398ea60d9 ("net: bcmgenet: add EEE support")
Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Link: https://lore.kernel.org/r/20230606214348.2408018-1-florian.fainelli@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoeth: ixgbe: fix the wake condition
Jakub Kicinski [Wed, 7 Jun 2023 01:08:26 +0000 (18:08 -0700)]
eth: ixgbe: fix the wake condition

Flip the netif_carrier_ok() condition in queue wake logic.
When I moved it to inside __netif_txq_completed_wake()
I missed negating it.

This made the condition ineffective and could probably
lead to crashes.

Fixes: 301f227fc860 ("net: piggy back on the memory barrier in bql when waking queues")
Reviewed-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://lore.kernel.org/r/20230607010826.960226-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoeth: bnxt: fix the wake condition
Jakub Kicinski [Wed, 7 Jun 2023 01:08:25 +0000 (18:08 -0700)]
eth: bnxt: fix the wake condition

The down condition should be the negation of the wake condition,
IOW when I moved it from:

 if (cond && wake())
to
 if (__netif_txq_completed_wake(cond))

Cond should have been negated. Flip it now.

This bug leads to occasional crashes with netconsole.
It may also lead to queue never waking up in case BQL is not enabled.

Reported-by: David Wei <davidhwei@meta.com>
Fixes: 08a096780d92 ("bnxt: use new queue try_stop/try_wake macros")
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/20230607010826.960226-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoMerge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Jakub Kicinski [Thu, 8 Jun 2023 04:47:11 +0000 (21:47 -0700)]
Merge tag 'for-netdev' of https://git./linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2023-06-07

We've added 7 non-merge commits during the last 7 day(s) which contain
a total of 12 files changed, 112 insertions(+), 7 deletions(-).

The main changes are:

1) Fix a use-after-free in BPF's task local storage, from KP Singh.

2) Make struct path handling more robust in bpf_d_path, from Jiri Olsa.

3) Fix a syzbot NULL-pointer dereference in sockmap, from Eric Dumazet.

4) UAPI fix for BPF_NETFILTER before final kernel ships,
   from Florian Westphal.

5) Fix map-in-map array_map_gen_lookup code generation where elem_size was
   not being set for inner maps, from Rhys Rustad-Elliott.

6) Fix sockopt_sk selftest's NETLINK_LIST_MEMBERSHIPS assertion,
   from Yonghong Song.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
  bpf: Add extra path pointer check to d_path helper
  selftests/bpf: Fix sockopt_sk selftest
  bpf: netfilter: Add BPF_NETFILTER bpf_attach_type
  selftests/bpf: Add access_inner_map selftest
  bpf: Fix elem_size not being set for inner maps
  bpf: Fix UAF in task local storage
  bpf, sockmap: Avoid potential NULL dereference in sk_psock_verdict_data_ready()
====================

Link: https://lore.kernel.org/r/20230607220514.29698-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agolib: cpu_rmap: Fix potential use-after-free in irq_cpu_rmap_release()
Ben Hutchings [Fri, 2 Jun 2023 18:28:15 +0000 (20:28 +0200)]
lib: cpu_rmap: Fix potential use-after-free in irq_cpu_rmap_release()

irq_cpu_rmap_release() calls cpu_rmap_put(), which may free the rmap.
So we need to clear the pointer to our glue structure in rmap before
doing that, not after.

Fixes: 4e0473f1060a ("lib: cpu_rmap: Avoid use after free on rmap->obj array entries")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/ZHo0vwquhOy3FaXc@decadent.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonetfilter: nfnetlink: skip error delivery on batch in case of ENOMEM
Pablo Neira Ayuso [Wed, 7 Jun 2023 22:19:12 +0000 (00:19 +0200)]
netfilter: nfnetlink: skip error delivery on batch in case of ENOMEM

If caller reports ENOMEM, then stop iterating over the batch and send a
single netlink message to userspace to report OOM.

Fixes: cbb8125eb40b ("netfilter: nfnetlink: deliver netlink errors on batch completion")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agonetfilter: nf_tables: integrate pipapo into commit protocol
Pablo Neira Ayuso [Wed, 7 Jun 2023 09:55:42 +0000 (11:55 +0200)]
netfilter: nf_tables: integrate pipapo into commit protocol

The pipapo set backend follows copy-on-update approach, maintaining one
clone of the existing datastructure that is being updated. The clone
and current datastructures are swapped via rcu from the commit step.

The existing integration with the commit protocol is flawed because
there is no operation to clean up the clone if the transaction is
aborted. Moreover, the datastructure swap happens on set element
activation.

This patch adds two new operations for sets: commit and abort, these new
operations are invoked from the commit and abort steps, after the
transactions have been digested, and it updates the pipapo set backend
to use it.

This patch adds a new ->pending_update field to sets to maintain a list
of sets that require this new commit and abort operations.

Fixes: 3c4287f62044 ("nf_tables: Add set type for arbitrary concatenation of ranges")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agoMerge tag 'input-for-v6.4-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor...
Linus Torvalds [Wed, 7 Jun 2023 20:49:42 +0000 (13:49 -0700)]
Merge tag 'input-for-v6.4-rc5' of git://git./linux/kernel/git/dtor/input

Pull input fixes from Dmitry Torokhov:

 - a fix for unbalanced open count for inhibited input devices

 - fixups in Elantech PS/2 and Cyppress TTSP v5 drivers

 - a quirk to soc_button_array driver to make it work with Lenovo
   Yoga Book X90F / X90L

 - a removal of erroneous entry from xpad driver

* tag 'input-for-v6.4-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: xpad - delete a Razer DeathAdder mouse VID/PID entry
  Input: psmouse - fix OOB access in Elantech protocol
  Input: soc_button_array - add invalid acpi_index DMI quirk handling
  Input: fix open count when closing inhibited device
  Input: cyttsp5 - fix array length

12 months agoMAINTAINERS: Add entry for debug objects
Thomas Gleixner [Tue, 6 Jun 2023 23:14:05 +0000 (01:14 +0200)]
MAINTAINERS: Add entry for debug objects

This is overdue and an oversight.

Add myself to this file deespite the fact that I'm trying to reduce the
number of entries in this file which have my name attached, but in the
hope that patches wont get picked up elsewhere completely unreviewed and
unnoticed.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 months agoafs: Fix setting of mtime when creating a file/dir/symlink
David Howells [Wed, 7 Jun 2023 08:47:13 +0000 (09:47 +0100)]
afs: Fix setting of mtime when creating a file/dir/symlink

kafs incorrectly passes a zero mtime (ie. 1st Jan 1970) to the server when
creating a file, dir or symlink because the mtime recorded in the
afs_operation struct gets passed to the server by the marshalling routines,
but the afs_mkdir(), afs_create() and afs_symlink() functions don't set it.

This gets masked if a file or directory is subsequently modified.

Fix this by filling in op->mtime before calling the create op.

Fixes: e49c7b2f6de7 ("afs: Build an abstraction around an "operation" concept")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
Reviewed-by: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 months agobpf: Add extra path pointer check to d_path helper
Jiri Olsa [Tue, 6 Jun 2023 18:17:14 +0000 (11:17 -0700)]
bpf: Add extra path pointer check to d_path helper

Anastasios reported crash on stable 5.15 kernel with following
BPF attached to lsm hook:

  SEC("lsm.s/bprm_creds_for_exec")
  int BPF_PROG(bprm_creds_for_exec, struct linux_binprm *bprm)
  {
          struct path *path = &bprm->executable->f_path;
          char p[128] = { 0 };

          bpf_d_path(path, p, 128);
          return 0;
  }

But bprm->executable can be NULL, so bpf_d_path call will crash:

  BUG: kernel NULL pointer dereference, address: 0000000000000018
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  PGD 0 P4D 0
  Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC NOPTI
  ...
  RIP: 0010:d_path+0x22/0x280
  ...
  Call Trace:
   <TASK>
   bpf_d_path+0x21/0x60
   bpf_prog_db9cf176e84498d9_bprm_creds_for_exec+0x94/0x99
   bpf_trampoline_6442506293_0+0x55/0x1000
   bpf_lsm_bprm_creds_for_exec+0x5/0x10
   security_bprm_creds_for_exec+0x29/0x40
   bprm_execve+0x1c1/0x900
   do_execveat_common.isra.0+0x1af/0x260
   __x64_sys_execve+0x32/0x40

It's problem for all stable trees with bpf_d_path helper, which was
added in 5.9.

This issue is fixed in current bpf code, where we identify and mark
trusted pointers, so the above code would fail even to load.

For the sake of the stable trees and to workaround potentially broken
verifier in the future, adding the code that reads the path object from
the passed pointer and verifies it's valid in kernel space.

Fixes: 6e22ab9da793 ("bpf: Add d_path helper")
Reported-by: Anastasios Papagiannis <tasos.papagiannnis@gmail.com>
Suggested-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Stanislav Fomichev <sdf@google.com>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20230606181714.532998-1-jolsa@kernel.org
12 months agonet: sched: fix possible refcount leak in tc_chain_tmplt_add()
Hangyu Hua [Wed, 7 Jun 2023 02:23:01 +0000 (10:23 +0800)]
net: sched: fix possible refcount leak in tc_chain_tmplt_add()

try_module_get will be called in tcf_proto_lookup_ops. So module_put needs
to be called to drop the refcount if ops don't implement the required
function.

Fixes: 9f407f1768d3 ("net: sched: introduce chain templates")
Signed-off-by: Hangyu Hua <hbh25y@gmail.com>
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: sched: act_police: fix sparse errors in tcf_police_dump()
Eric Dumazet [Tue, 6 Jun 2023 13:13:04 +0000 (13:13 +0000)]
net: sched: act_police: fix sparse errors in tcf_police_dump()

Fixes following sparse errors:

net/sched/act_police.c:360:28: warning: dereference of noderef expression
net/sched/act_police.c:362:45: warning: dereference of noderef expression
net/sched/act_police.c:362:45: warning: dereference of noderef expression
net/sched/act_police.c:368:28: warning: dereference of noderef expression
net/sched/act_police.c:370:45: warning: dereference of noderef expression
net/sched/act_police.c:370:45: warning: dereference of noderef expression
net/sched/act_police.c:376:45: warning: dereference of noderef expression
net/sched/act_police.c:376:45: warning: dereference of noderef expression

Fixes: d1967e495a8d ("net_sched: act_police: add 2 new attributes to support police 64bit rate and peakrate")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: openvswitch: fix upcall counter access before allocation
Eelco Chaudron [Tue, 6 Jun 2023 11:56:35 +0000 (13:56 +0200)]
net: openvswitch: fix upcall counter access before allocation

Currently, the per cpu upcall counters are allocated after the vport is
created and inserted into the system. This could lead to the datapath
accessing the counters before they are allocated resulting in a kernel
Oops.

Here is an example:

  PID: 59693    TASK: ffff0005f4f51500  CPU: 0    COMMAND: "ovs-vswitchd"
   #0 [ffff80000a39b5b0] __switch_to at ffffb70f0629f2f4
   #1 [ffff80000a39b5d0] __schedule at ffffb70f0629f5cc
   #2 [ffff80000a39b650] preempt_schedule_common at ffffb70f0629fa60
   #3 [ffff80000a39b670] dynamic_might_resched at ffffb70f0629fb58
   #4 [ffff80000a39b680] mutex_lock_killable at ffffb70f062a1388
   #5 [ffff80000a39b6a0] pcpu_alloc at ffffb70f0594460c
   #6 [ffff80000a39b750] __alloc_percpu_gfp at ffffb70f05944e68
   #7 [ffff80000a39b760] ovs_vport_cmd_new at ffffb70ee6961b90 [openvswitch]
   ...

  PID: 58682    TASK: ffff0005b2f0bf00  CPU: 0    COMMAND: "kworker/0:3"
   #0 [ffff80000a5d2f40] machine_kexec at ffffb70f056a0758
   #1 [ffff80000a5d2f70] __crash_kexec at ffffb70f057e2994
   #2 [ffff80000a5d3100] crash_kexec at ffffb70f057e2ad8
   #3 [ffff80000a5d3120] die at ffffb70f0628234c
   #4 [ffff80000a5d31e0] die_kernel_fault at ffffb70f062828a8
   #5 [ffff80000a5d3210] __do_kernel_fault at ffffb70f056a31f4
   #6 [ffff80000a5d3240] do_bad_area at ffffb70f056a32a4
   #7 [ffff80000a5d3260] do_translation_fault at ffffb70f062a9710
   #8 [ffff80000a5d3270] do_mem_abort at ffffb70f056a2f74
   #9 [ffff80000a5d32a0] el1_abort at ffffb70f06297dac
  #10 [ffff80000a5d32d0] el1h_64_sync_handler at ffffb70f06299b24
  #11 [ffff80000a5d3410] el1h_64_sync at ffffb70f056812dc
  #12 [ffff80000a5d3430] ovs_dp_upcall at ffffb70ee6963c84 [openvswitch]
  #13 [ffff80000a5d3470] ovs_dp_process_packet at ffffb70ee6963fdc [openvswitch]
  #14 [ffff80000a5d34f0] ovs_vport_receive at ffffb70ee6972c78 [openvswitch]
  #15 [ffff80000a5d36f0] netdev_port_receive at ffffb70ee6973948 [openvswitch]
  #16 [ffff80000a5d3720] netdev_frame_hook at ffffb70ee6973a28 [openvswitch]
  #17 [ffff80000a5d3730] __netif_receive_skb_core.constprop.0 at ffffb70f06079f90

We moved the per cpu upcall counter allocation to the existing vport
alloc and free functions to solve this.

Fixes: 95637d91fefd ("net: openvswitch: release vport resources on failure")
Fixes: 1933ea365aa7 ("net: openvswitch: Add support to count upcall packets")
Signed-off-by: Eelco Chaudron <echaudro@redhat.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: sched: move rtm_tca_policy declaration to include file
Eric Dumazet [Tue, 6 Jun 2023 11:42:33 +0000 (11:42 +0000)]
net: sched: move rtm_tca_policy declaration to include file

rtm_tca_policy is used from net/sched/sch_api.c and net/sched/cls_api.c,
thus should be declared in an include file.

This fixes the following sparse warning:
net/sched/sch_api.c:1434:25: warning: symbol 'rtm_tca_policy' was not declared. Should it be static?

Fixes: e331473fee3d ("net/sched: cls_api: add missing validation of netlink attributes")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoice: make writes to /dev/gnssX synchronous
Michal Schmidt [Tue, 6 Jun 2023 17:12:53 +0000 (10:12 -0700)]
ice: make writes to /dev/gnssX synchronous

The current ice driver's GNSS write implementation buffers writes and
works through them asynchronously in a kthread. That's bad because:
 - The GNSS write_raw operation is supposed to be synchronous[1][2].
 - There is no upper bound on the number of pending writes.
   Userspace can submit writes much faster than the driver can process,
   consuming unlimited amounts of kernel memory.

A patch that's currently on review[3] ("[v3,net] ice: Write all GNSS
buffers instead of first one") would add one more problem:
 - The possibility of waiting for a very long time to flush the write
   work when doing rmmod, softlockups.

To fix these issues, simplify the implementation: Drop the buffering,
the write_work, and make the writes synchronous.

I tested this with gpsd and ubxtool.

[1] https://events19.linuxfoundation.org/wp-content/uploads/2017/12/The-GNSS-Subsystem-Johan-Hovold-Hovold-Consulting-AB.pdf
    "User interface" slide.
[2] A comment in drivers/gnss/core.c:gnss_write():
        /* Ignoring O_NONBLOCK, write_raw() is synchronous. */
[3] https://patchwork.ozlabs.org/project/intel-wired-lan/patch/20230217120541.16745-1-karol.kolacinski@intel.com/

Fixes: d6b98c8d242a ("ice: add write functionality for GNSS TTY")
Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Tested-by: Sunitha Mekala <sunithax.d.mekala@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agonet: sched: add rcu annotations around qdisc->qdisc_sleeping
Eric Dumazet [Tue, 6 Jun 2023 11:19:29 +0000 (11:19 +0000)]
net: sched: add rcu annotations around qdisc->qdisc_sleeping

syzbot reported a race around qdisc->qdisc_sleeping [1]

It is time we add proper annotations to reads and writes to/from
qdisc->qdisc_sleeping.

[1]
BUG: KCSAN: data-race in dev_graft_qdisc / qdisc_lookup_rcu

read to 0xffff8881286fc618 of 8 bytes by task 6928 on cpu 1:
qdisc_lookup_rcu+0x192/0x2c0 net/sched/sch_api.c:331
__tcf_qdisc_find+0x74/0x3c0 net/sched/cls_api.c:1174
tc_get_tfilter+0x18f/0x990 net/sched/cls_api.c:2547
rtnetlink_rcv_msg+0x7af/0x8c0 net/core/rtnetlink.c:6386
netlink_rcv_skb+0x126/0x220 net/netlink/af_netlink.c:2546
rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:6413
netlink_unicast_kernel net/netlink/af_netlink.c:1339 [inline]
netlink_unicast+0x56f/0x640 net/netlink/af_netlink.c:1365
netlink_sendmsg+0x665/0x770 net/netlink/af_netlink.c:1913
sock_sendmsg_nosec net/socket.c:724 [inline]
sock_sendmsg net/socket.c:747 [inline]
____sys_sendmsg+0x375/0x4c0 net/socket.c:2503
___sys_sendmsg net/socket.c:2557 [inline]
__sys_sendmsg+0x1e3/0x270 net/socket.c:2586
__do_sys_sendmsg net/socket.c:2595 [inline]
__se_sys_sendmsg net/socket.c:2593 [inline]
__x64_sys_sendmsg+0x46/0x50 net/socket.c:2593
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

write to 0xffff8881286fc618 of 8 bytes by task 6912 on cpu 0:
dev_graft_qdisc+0x4f/0x80 net/sched/sch_generic.c:1115
qdisc_graft+0x7d0/0xb60 net/sched/sch_api.c:1103
tc_modify_qdisc+0x712/0xf10 net/sched/sch_api.c:1693
rtnetlink_rcv_msg+0x807/0x8c0 net/core/rtnetlink.c:6395
netlink_rcv_skb+0x126/0x220 net/netlink/af_netlink.c:2546
rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:6413
netlink_unicast_kernel net/netlink/af_netlink.c:1339 [inline]
netlink_unicast+0x56f/0x640 net/netlink/af_netlink.c:1365
netlink_sendmsg+0x665/0x770 net/netlink/af_netlink.c:1913
sock_sendmsg_nosec net/socket.c:724 [inline]
sock_sendmsg net/socket.c:747 [inline]
____sys_sendmsg+0x375/0x4c0 net/socket.c:2503
___sys_sendmsg net/socket.c:2557 [inline]
__sys_sendmsg+0x1e3/0x270 net/socket.c:2586
__do_sys_sendmsg net/socket.c:2595 [inline]
__se_sys_sendmsg net/socket.c:2593 [inline]
__x64_sys_sendmsg+0x46/0x50 net/socket.c:2593
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

Reported by Kernel Concurrency Sanitizer on:
CPU: 0 PID: 6912 Comm: syz-executor.5 Not tainted 6.4.0-rc3-syzkaller-00190-g0d85b27b0cc6 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/16/2023

Fixes: 3a7d0d07a386 ("net: sched: extend Qdisc with rcu")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Vlad Buslov <vladbu@nvidia.com>
Acked-by: Jamal Hadi Salim<jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoMerge branch 'rfs-lockless-annotate'
David S. Miller [Wed, 7 Jun 2023 09:09:05 +0000 (10:09 +0100)]
Merge branch 'rfs-lockless-annotate'

Eric Dumazet says:

====================
rfs: annotate lockless accesses

rfs runs without locks held, so we should annotate
read and writes to shared variables.

It should prevent compilers forcing writes
in the following situation:

  if (var != val)
     var = val;

A compiler could indeed simply avoid the conditional:

    var = val;

This matters if var is shared between many cpus.

v2: aligns one closing bracket (Simon)
    adds Fixes: tags (Jakub)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agorfs: annotate lockless accesses to RFS sock flow table
Eric Dumazet [Tue, 6 Jun 2023 07:41:15 +0000 (07:41 +0000)]
rfs: annotate lockless accesses to RFS sock flow table

Add READ_ONCE()/WRITE_ONCE() on accesses to the sock flow table.

This also prevents a (smart ?) compiler to remove the condition in:

if (table->ents[index] != newval)
        table->ents[index] = newval;

We need the condition to avoid dirtying a shared cache line.

Fixes: fec5e652e58f ("rfs: Receive Flow Steering")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agorfs: annotate lockless accesses to sk->sk_rxhash
Eric Dumazet [Tue, 6 Jun 2023 07:41:14 +0000 (07:41 +0000)]
rfs: annotate lockless accesses to sk->sk_rxhash

Add READ_ONCE()/WRITE_ONCE() on accesses to sk->sk_rxhash.

This also prevents a (smart ?) compiler to remove the condition in:

if (sk->sk_rxhash != newval)
sk->sk_rxhash = newval;

We need the condition to avoid dirtying a shared cache line.

Fixes: fec5e652e58f ("rfs: Receive Flow Steering")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 months agoMerge tag 'for-net-2023-06-05' of git://git.kernel.org/pub/scm/linux/kernel/git/bluet...
Jakub Kicinski [Wed, 7 Jun 2023 04:36:57 +0000 (21:36 -0700)]
Merge tag 'for-net-2023-06-05' of git://git./linux/kernel/git/bluetooth/bluetooth

Luiz Augusto von Dentz says:

====================
bluetooth pull request for net:

 - Fixes to debugfs registration
 - Fix use-after-free in hci_remove_ltk/hci_remove_irk
 - Fixes to ISO channel support
 - Fix missing checks for invalid L2CAP DCID
 - Fix l2cap_disconnect_req deadlock
 - Add lock to protect HCI_UNREGISTER

* tag 'for-net-2023-06-05' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
  Bluetooth: L2CAP: Add missing checks for invalid DCID
  Bluetooth: ISO: use correct CIS order in Set CIG Parameters event
  Bluetooth: ISO: don't try to remove CIG if there are bound CIS left
  Bluetooth: Fix l2cap_disconnect_req deadlock
  Bluetooth: hci_qca: fix debugfs registration
  Bluetooth: fix debugfs registration
  Bluetooth: hci_sync: add lock to protect HCI_UNREGISTER
  Bluetooth: Fix use-after-free in hci_remove_ltk/hci_remove_irk
  Bluetooth: ISO: Fix CIG auto-allocation to select configurable CIG
  Bluetooth: ISO: consider right CIS when removing CIG at cleanup
====================

Link: https://lore.kernel.org/r/20230606003454.2392552-1-luiz.dentz@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoMerge tag 'nf-23-06-07' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Jakub Kicinski [Wed, 7 Jun 2023 04:23:49 +0000 (21:23 -0700)]
Merge tag 'nf-23-06-07' of git://git./linux/kernel/git/netfilter/nf

Pablo Neira Ayuso says:

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

The following patchset contains Netfilter fixes for net:

1) Missing nul-check in basechain hook netlink dump path, from Gavrilov Ilia.

2) Fix bitwise register tracking, from Jeremy Sowden.

3) Null pointer dereference when accessing conntrack helper,
   from Tijs Van Buggenhout.

4) Add schedule point to ipset's call_ad, from Kuniyuki Iwashima.

5) Incorrect boundary check when building chain blob.

* tag 'nf-23-06-07' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
  netfilter: nf_tables: out-of-bound check in chain blob
  netfilter: ipset: Add schedule point in call_ad().
  netfilter: conntrack: fix NULL pointer dereference in nf_confirm_cthelper
  netfilter: nft_bitwise: fix register tracking
  netfilter: nf_tables: Add null check for nla_nest_start_noflag() in nft_dump_basechain_hook()
====================

Link: https://lore.kernel.org/r/20230606225851.67394-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoMerge tag 'wireless-2023-06-06' of git://git.kernel.org/pub/scm/linux/kernel/git...
Jakub Kicinski [Wed, 7 Jun 2023 04:16:52 +0000 (21:16 -0700)]
Merge tag 'wireless-2023-06-06' of git://git./linux/kernel/git/wireless/wireless

Kalle Valo says:

====================
wireless fixes for v6.4

Both rtw88 and rtw89 have a 802.11 powersave fix for a regression
introduced in v6.0. mt76 fixes a race and a null pointer dereference.
iwlwifi fixes an issue where not enough memory was allocated for a
firmware event. And finally the stack has several smaller fixes all
over.

* tag 'wireless-2023-06-06' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
  wifi: cfg80211: fix locking in regulatory disconnect
  wifi: cfg80211: fix locking in sched scan stop work
  wifi: iwlwifi: mvm: Fix -Warray-bounds bug in iwl_mvm_wait_d3_notif()
  wifi: mac80211: fix switch count in EMA beacons
  wifi: mac80211: don't translate beacon/presp addrs
  wifi: mac80211: mlme: fix non-inheritence element
  wifi: cfg80211: reject bad AP MLD address
  wifi: mac80211: use correct iftype HE cap
  wifi: mt76: mt7996: fix possible NULL pointer dereference in mt7996_mac_write_txwi()
  wifi: rtw89: remove redundant check of entering LPS
  wifi: rtw89: correct PS calculation for SUPPORTS_DYNAMIC_PS
  wifi: rtw88: correct PS calculation for SUPPORTS_DYNAMIC_PS
  wifi: mt76: mt7615: fix possible race in mt7615_mac_sta_poll
====================

Link: https://lore.kernel.org/r/20230606150817.EC133C433D2@smtp.kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agovirtio_net: use control_buf for coalesce params
Brett Creeley [Mon, 5 Jun 2023 19:59:25 +0000 (12:59 -0700)]
virtio_net: use control_buf for coalesce params

Commit 699b045a8e43 ("net: virtio_net: notifications coalescing
support") added coalescing command support for virtio_net. However,
the coalesce commands are using buffers on the stack, which is causing
the device to see DMA errors. There should also be a complaint from
check_for_stack() in debug_dma_map_xyz(). Fix this by adding and using
coalesce params from the control_buf struct, which aligns with other
commands.

Cc: stable@vger.kernel.org
Fixes: 699b045a8e43 ("net: virtio_net: notifications coalescing support")
Reviewed-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/r/20230605195925.51625-1-brett.creeley@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agopds_core: Fix FW recovery detection
Brett Creeley [Mon, 5 Jun 2023 19:51:16 +0000 (12:51 -0700)]
pds_core: Fix FW recovery detection

Commit 523847df1b37 ("pds_core: add devcmd device interfaces") included
initial support for FW recovery detection. Unfortunately, the ordering
in pdsc_is_fw_good() was incorrect, which was causing FW recovery to be
undetected by the driver. Fix this by making sure to update the cached
fw_status by calling pdsc_is_fw_running() before setting the local FW
gen.

Fixes: 523847df1b37 ("pds_core: add devcmd device interfaces")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230605195116.49653-1-brett.creeley@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agotcp: gso: really support BIG TCP
Eric Dumazet [Mon, 5 Jun 2023 16:16:47 +0000 (16:16 +0000)]
tcp: gso: really support BIG TCP

We missed that tcp_gso_segment() was assuming skb->len was smaller than 65535 :

oldlen = (u16)~skb->len;

This part came with commit 0718bcc09b35 ("[NET]: Fix CHECKSUM_HW GSO problems.")

This leads to wrong TCP checksum.

Adapt the code to accept arbitrary packet length.

v2:
  - use two csum_add() instead of csum_fold() (Alexander Duyck)
  - Change delta type to __wsum to reduce casts (Alexander Duyck)

Fixes: 09f3d1a3a52c ("ipv6/gso: remove temporary HBH/jumbo header")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230605161647.3624428-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agoipv6: rpl: Fix Route of Death.
Kuniyuki Iwashima [Mon, 5 Jun 2023 18:06:17 +0000 (11:06 -0700)]
ipv6: rpl: Fix Route of Death.

A remote DoS vulnerability of RPL Source Routing is assigned CVE-2023-2156.

The Source Routing Header (SRH) has the following format:

  0                   1                   2                   3
  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |  Next Header  |  Hdr Ext Len  | Routing Type  | Segments Left |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  | CmprI | CmprE |  Pad  |               Reserved                |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |                                                               |
  .                                                               .
  .                        Addresses[1..n]                        .
  .                                                               .
  |                                                               |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

The originator of an SRH places the first hop's IPv6 address in the IPv6
header's IPv6 Destination Address and the second hop's IPv6 address as
the first address in Addresses[1..n].

The CmprI and CmprE fields indicate the number of prefix octets that are
shared with the IPv6 Destination Address.  When CmprI or CmprE is not 0,
Addresses[1..n] are compressed as follows:

  1..n-1 : (16 - CmprI) bytes
       n : (16 - CmprE) bytes

Segments Left indicates the number of route segments remaining.  When the
value is not zero, the SRH is forwarded to the next hop.  Its address
is extracted from Addresses[n - Segment Left + 1] and swapped with IPv6
Destination Address.

When Segment Left is greater than or equal to 2, the size of SRH is not
changed because Addresses[1..n-1] are decompressed and recompressed with
CmprI.

OTOH, when Segment Left changes from 1 to 0, the new SRH could have a
different size because Addresses[1..n-1] are decompressed with CmprI and
recompressed with CmprE.

Let's say CmprI is 15 and CmprE is 0.  When we receive SRH with Segment
Left >= 2, Addresses[1..n-1] have 1 byte for each, and Addresses[n] has
16 bytes.  When Segment Left is 1, Addresses[1..n-1] is decompressed to
16 bytes and not recompressed.  Finally, the new SRH will need more room
in the header, and the size is (16 - 1) * (n - 1) bytes.

Here the max value of n is 255 as Segment Left is u8, so in the worst case,
we have to allocate 3825 bytes in the skb headroom.  However, now we only
allocate a small fixed buffer that is IPV6_RPL_SRH_WORST_SWAP_SIZE (16 + 7
bytes).  If the decompressed size overflows the room, skb_push() hits BUG()
below [0].

Instead of allocating the fixed buffer for every packet, let's allocate
enough headroom only when we receive SRH with Segment Left 1.

[0]:
skbuff: skb_under_panic: text:ffffffff81c9f6e2 len:576 put:576 head:ffff8880070b5180 data:ffff8880070b4fb0 tail:0x70 end:0x140 dev:lo
kernel BUG at net/core/skbuff.c:200!
invalid opcode: 0000 [#1] PREEMPT SMP PTI
CPU: 0 PID: 154 Comm: python3 Not tainted 6.4.0-rc4-00190-gc308e9ec0047 #7
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
RIP: 0010:skb_panic (net/core/skbuff.c:200)
Code: 4f 70 50 8b 87 bc 00 00 00 50 8b 87 b8 00 00 00 50 ff b7 c8 00 00 00 4c 8b 8f c0 00 00 00 48 c7 c7 80 6e 77 82 e8 ad 8b 60 ff <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 90 90 90 90 90 90 90 90 90
RSP: 0018:ffffc90000003da0 EFLAGS: 00000246
RAX: 0000000000000085 RBX: ffff8880058a6600 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff88807dc1c540 RDI: ffff88807dc1c540
RBP: ffffc90000003e48 R08: ffffffff82b392c8 R09: 00000000ffffdfff
R10: ffffffff82a592e0 R11: ffffffff82b092e0 R12: ffff888005b1c800
R13: ffff8880070b51b8 R14: ffff888005b1ca18 R15: ffff8880070b5190
FS:  00007f4539f0b740(0000) GS:ffff88807dc00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000055670baf3000 CR3: 0000000005b0e000 CR4: 00000000007506f0
PKRU: 55555554
Call Trace:
 <IRQ>
 skb_push (net/core/skbuff.c:210)
 ipv6_rthdr_rcv (./include/linux/skbuff.h:2880 net/ipv6/exthdrs.c:634 net/ipv6/exthdrs.c:718)
 ip6_protocol_deliver_rcu (net/ipv6/ip6_input.c:437 (discriminator 5))
 ip6_input_finish (./include/linux/rcupdate.h:805 net/ipv6/ip6_input.c:483)
 __netif_receive_skb_one_core (net/core/dev.c:5494)
 process_backlog (./include/linux/rcupdate.h:805 net/core/dev.c:5934)
 __napi_poll (net/core/dev.c:6496)
 net_rx_action (net/core/dev.c:6565 net/core/dev.c:6696)
 __do_softirq (./arch/x86/include/asm/jump_label.h:27 ./include/linux/jump_label.h:207 ./include/trace/events/irq.h:142 kernel/softirq.c:572)
 do_softirq (kernel/softirq.c:472 kernel/softirq.c:459)
 </IRQ>
 <TASK>
 __local_bh_enable_ip (kernel/softirq.c:396)
 __dev_queue_xmit (net/core/dev.c:4272)
 ip6_finish_output2 (./include/net/neighbour.h:544 net/ipv6/ip6_output.c:134)
 rawv6_sendmsg (./include/net/dst.h:458 ./include/linux/netfilter.h:303 net/ipv6/raw.c:656 net/ipv6/raw.c:914)
 sock_sendmsg (net/socket.c:724 net/socket.c:747)
 __sys_sendto (net/socket.c:2144)
 __x64_sys_sendto (net/socket.c:2156 net/socket.c:2152 net/socket.c:2152)
 do_syscall_64 (arch/x86/entry/common.c:50 arch/x86/entry/common.c:80)
 entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:120)
RIP: 0033:0x7f453a138aea
Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb b8 0f 1f 00 f3 0f 1e fa 41 89 ca 64 8b 04 25 18 00 00 00 85 c0 75 15 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 7e c3 0f 1f 44 00 00 41 54 48 83 ec 30 44 89
RSP: 002b:00007ffcc212a1c8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 00007ffcc212a288 RCX: 00007f453a138aea
RDX: 0000000000000060 RSI: 00007f4539084c20 RDI: 0000000000000003
RBP: 00007f4538308e80 R08: 00007ffcc212a300 R09: 000000000000001c
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: ffffffffc4653600 R14: 0000000000000001 R15: 00007f4539712d1b
 </TASK>
Modules linked in:

Fixes: 8610c7c6e3bd ("net: ipv6: add support for rpl sr exthdr")
Reported-by: Max VA
Closes: https://www.interruptlabs.co.uk/articles/linux-ipv6-route-of-death
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230605180617.67284-1-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonetlink: specs: ethtool: fix random typos
Jakub Kicinski [Mon, 5 Jun 2023 23:32:57 +0000 (16:32 -0700)]
netlink: specs: ethtool: fix random typos

Working on the code gen for C reveals typos in the ethtool spec
as the compiler tries to find the names in the existing uAPI
header. Fix the mistakes.

Fixes: a353318ebf24 ("tools: ynl: populate most of the ethtool spec")
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20230605233257.843977-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
12 months agonetfilter: nf_tables: out-of-bound check in chain blob
Pablo Neira Ayuso [Tue, 6 Jun 2023 14:32:44 +0000 (16:32 +0200)]
netfilter: nf_tables: out-of-bound check in chain blob

Add current size of rule expressions to the boundary check.

Fixes: 2c865a8a28a1 ("netfilter: nf_tables: add rule blob layout")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agonetfilter: ipset: Add schedule point in call_ad().
Kuniyuki Iwashima [Thu, 18 May 2023 17:33:00 +0000 (10:33 -0700)]
netfilter: ipset: Add schedule point in call_ad().

syzkaller found a repro that causes Hung Task [0] with ipset.  The repro
first creates an ipset and then tries to delete a large number of IPs
from the ipset concurrently:

  IPSET_ATTR_IPADDR_IPV4 : 172.20.20.187
  IPSET_ATTR_CIDR        : 2

The first deleting thread hogs a CPU with nfnl_lock(NFNL_SUBSYS_IPSET)
held, and other threads wait for it to be released.

Previously, the same issue existed in set->variant->uadt() that could run
so long under ip_set_lock(set).  Commit 5e29dc36bd5e ("netfilter: ipset:
Rework long task execution when adding/deleting entries") tried to fix it,
but the issue still exists in the caller with another mutex.

While adding/deleting many IPs, we should release the CPU periodically to
prevent someone from abusing ipset to hang the system.

Note we need to increment the ipset's refcnt to prevent the ipset from
being destroyed while rescheduling.

[0]:
INFO: task syz-executor174:268 blocked for more than 143 seconds.
      Not tainted 6.4.0-rc1-00145-gba79e9a73284 #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:syz-executor174 state:D stack:0     pid:268   ppid:260    flags:0x0000000d
Call trace:
 __switch_to+0x308/0x714 arch/arm64/kernel/process.c:556
 context_switch kernel/sched/core.c:5343 [inline]
 __schedule+0xd84/0x1648 kernel/sched/core.c:6669
 schedule+0xf0/0x214 kernel/sched/core.c:6745
 schedule_preempt_disabled+0x58/0xf0 kernel/sched/core.c:6804
 __mutex_lock_common kernel/locking/mutex.c:679 [inline]
 __mutex_lock+0x6fc/0xdb0 kernel/locking/mutex.c:747
 __mutex_lock_slowpath+0x14/0x20 kernel/locking/mutex.c:1035
 mutex_lock+0x98/0xf0 kernel/locking/mutex.c:286
 nfnl_lock net/netfilter/nfnetlink.c:98 [inline]
 nfnetlink_rcv_msg+0x480/0x70c net/netfilter/nfnetlink.c:295
 netlink_rcv_skb+0x1c0/0x350 net/netlink/af_netlink.c:2546
 nfnetlink_rcv+0x18c/0x199c net/netfilter/nfnetlink.c:658
 netlink_unicast_kernel net/netlink/af_netlink.c:1339 [inline]
 netlink_unicast+0x664/0x8cc net/netlink/af_netlink.c:1365
 netlink_sendmsg+0x6d0/0xa4c net/netlink/af_netlink.c:1913
 sock_sendmsg_nosec net/socket.c:724 [inline]
 sock_sendmsg net/socket.c:747 [inline]
 ____sys_sendmsg+0x4b8/0x810 net/socket.c:2503
 ___sys_sendmsg net/socket.c:2557 [inline]
 __sys_sendmsg+0x1f8/0x2a4 net/socket.c:2586
 __do_sys_sendmsg net/socket.c:2595 [inline]
 __se_sys_sendmsg net/socket.c:2593 [inline]
 __arm64_sys_sendmsg+0x80/0x94 net/socket.c:2593
 __invoke_syscall arch/arm64/kernel/syscall.c:38 [inline]
 invoke_syscall+0x84/0x270 arch/arm64/kernel/syscall.c:52
 el0_svc_common+0x134/0x24c arch/arm64/kernel/syscall.c:142
 do_el0_svc+0x64/0x198 arch/arm64/kernel/syscall.c:193
 el0_svc+0x2c/0x7c arch/arm64/kernel/entry-common.c:637
 el0t_64_sync_handler+0x84/0xf0 arch/arm64/kernel/entry-common.c:655
 el0t_64_sync+0x190/0x194 arch/arm64/kernel/entry.S:591

Reported-by: syzkaller <syzkaller@googlegroups.com>
Fixes: a7b4f989a629 ("netfilter: ipset: IP set core support")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Acked-by: Jozsef Kadlecsik <kadlec@netfilter.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agonetfilter: conntrack: fix NULL pointer dereference in nf_confirm_cthelper
Tijs Van Buggenhout [Thu, 25 May 2023 10:25:26 +0000 (12:25 +0200)]
netfilter: conntrack: fix NULL pointer dereference in nf_confirm_cthelper

An nf_conntrack_helper from nf_conn_help may become NULL after DNAT.

Observed when TCP port 1720 (Q931_PORT), associated with h323 conntrack
helper, is DNAT'ed to another destination port (e.g. 1730), while
nfqueue is being used for final acceptance (e.g. snort).

This happenned after transition from kernel 4.14 to 5.10.161.

Workarounds:
 * keep the same port (1720) in DNAT
 * disable nfqueue
 * disable/unload h323 NAT helper

$ linux-5.10/scripts/decode_stacktrace.sh vmlinux < /tmp/kernel.log
BUG: kernel NULL pointer dereference, address: 0000000000000084
[..]
RIP: 0010:nf_conntrack_update (net/netfilter/nf_conntrack_core.c:2080 net/netfilter/nf_conntrack_core.c:2134) nf_conntrack
[..]
nfqnl_reinject (net/netfilter/nfnetlink_queue.c:237) nfnetlink_queue
nfqnl_recv_verdict (net/netfilter/nfnetlink_queue.c:1230) nfnetlink_queue
nfnetlink_rcv_msg (net/netfilter/nfnetlink.c:241) nfnetlink
[..]

Fixes: ee04805ff54a ("netfilter: conntrack: make conntrack userspace helpers work again")
Signed-off-by: Tijs Van Buggenhout <tijs.van.buggenhout@axsguard.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agonetfilter: nft_bitwise: fix register tracking
Jeremy Sowden [Thu, 25 May 2023 14:07:24 +0000 (15:07 +0100)]
netfilter: nft_bitwise: fix register tracking

At the end of `nft_bitwise_reduce`, there is a loop which is intended to
update the bitwise expression associated with each tracked destination
register.  However, currently, it just updates the first register
repeatedly.  Fix it.

Fixes: 34cc9e52884a ("netfilter: nf_tables: cancel tracking for clobbered destination registers")
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agonetfilter: nf_tables: Add null check for nla_nest_start_noflag() in nft_dump_basechai...
Gavrilov Ilia [Wed, 24 May 2023 12:25:27 +0000 (12:25 +0000)]
netfilter: nf_tables: Add null check for nla_nest_start_noflag() in nft_dump_basechain_hook()

The nla_nest_start_noflag() function may fail and return NULL;
the return value needs to be checked.

Found by InfoTeCS on behalf of Linux Verification Center
(linuxtesting.org) with SVACE.

Fixes: d54725cd11a5 ("netfilter: nf_tables: support for multiple devices per netdev hook")
Signed-off-by: Gavrilov Ilia <Ilia.Gavrilov@infotecs.ru>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
12 months agoselftests/bpf: Fix sockopt_sk selftest
Yonghong Song [Tue, 6 Jun 2023 17:22:02 +0000 (10:22 -0700)]
selftests/bpf: Fix sockopt_sk selftest

Commit f4e4534850a9 ("net/netlink: fix NETLINK_LIST_MEMBERSHIPS length report")
fixed NETLINK_LIST_MEMBERSHIPS length report which caused
selftest sockopt_sk failure. The failure log looks like

  test_sockopt_sk:PASS:join_cgroup /sockopt_sk 0 nsec
  run_test:PASS:skel_load 0 nsec
  run_test:PASS:setsockopt_link 0 nsec
  run_test:PASS:getsockopt_link 0 nsec
  getsetsockopt:FAIL:Unexpected NETLINK_LIST_MEMBERSHIPS value unexpected Unexpected NETLINK_LIST_MEMBERSHIPS value: actual 8 != expected 4
  run_test:PASS:getsetsockopt 0 nsec
  #201     sockopt_sk:FAIL

In net/netlink/af_netlink.c, function netlink_getsockopt(), for NETLINK_LIST_MEMBERSHIPS,
nlk->ngroups equals to 36. Before Commit f4e4534850a9, the optlen is calculated as
  ALIGN(nlk->ngroups / 8, sizeof(u32)) = 4
After that commit, the optlen is
  ALIGN(BITS_TO_BYTES(nlk->ngroups), sizeof(u32)) = 8

Fix the test by setting the expected optlen to be 8.

Fixes: f4e4534850a9 ("net/netlink: fix NETLINK_LIST_MEMBERSHIPS length report")
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230606172202.1606249-1-yhs@fb.com
12 months agoMerge tag 'spi-fix-v6.4-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi
Linus Torvalds [Tue, 6 Jun 2023 13:18:28 +0000 (06:18 -0700)]
Merge tag 'spi-fix-v6.4-rc5' of git://git./linux/kernel/git/broonie/spi

Pull spi fixes from Mark Brown:
 "A small collection of driver specific fixes, none of them particularly
  remarkable or severe"

* tag 'spi-fix-v6.4-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: qup: Request DMA before enabling clocks
  spi: mt65xx: make sure operations completed before unloading
  spi: lpspi: disable lpspi module irq in DMA mode

12 months agowifi: cfg80211: fix locking in regulatory disconnect
Johannes Berg [Tue, 6 Jun 2023 12:34:48 +0000 (14:34 +0200)]
wifi: cfg80211: fix locking in regulatory disconnect

This should use wiphy_lock() now instead of requiring the
RTNL, since __cfg80211_leave() via cfg80211_leave() is now
requiring that lock to be held.

Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
12 months agowifi: cfg80211: fix locking in sched scan stop work
Johannes Berg [Tue, 6 Jun 2023 12:34:47 +0000 (14:34 +0200)]
wifi: cfg80211: fix locking in sched scan stop work

This should use wiphy_lock() now instead of acquiring the
RTNL, since cfg80211_stop_sched_scan_req() now needs that.

Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
12 months agoMerge tag 'gfs2-v6.4-rc4-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2...
Linus Torvalds [Tue, 6 Jun 2023 12:49:06 +0000 (05:49 -0700)]
Merge tag 'gfs2-v6.4-rc4-fix' of git://git./linux/kernel/git/gfs2/linux-gfs2

Pull gfs2 fix from Andreas Gruenbacher:

 - Don't get stuck writing page onto itself under direct I/O

* tag 'gfs2-v6.4-rc4-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
  gfs2: Don't get stuck writing page onto itself under direct I/O

12 months agoMerge tag 'platform-drivers-x86-v6.4-4' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 6 Jun 2023 12:42:21 +0000 (05:42 -0700)]
Merge tag 'platform-drivers-x86-v6.4-4' of git://git./linux/kernel/git/pdx86/platform-drivers-x86

Pull x86 platform driver fixes from Hans de Goede:

 - various Microsoft Surface support fixes

 - one fix for the INT3472 driver

* tag 'platform-drivers-x86-v6.4-4' of git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86:
  platform/x86: int3472: Avoid crash in unregistering regulator gpio
  platform/surface: aggregator_tabletsw: Add support for book mode in POS subsystem
  platform/surface: aggregator_tabletsw: Add support for book mode in KIP subsystem
  platform/surface: aggregator: Allow completion work-items to be executed in parallel
  platform/surface: aggregator: Make to_ssam_device_driver() respect constness

12 months agoqed/qede: Fix scheduling while atomic
Manish Chopra [Mon, 5 Jun 2023 11:26:00 +0000 (16:56 +0530)]
qed/qede: Fix scheduling while atomic

Statistics read through bond interface via sysfs causes
below bug and traces as it triggers the bonding module to
collect the slave device statistics while holding the spinlock,
beneath that qede->qed driver statistics flow gets scheduled out
due to usleep_range() used in PTT acquire logic

[ 3673.988874] Hardware name: HPE ProLiant DL365 Gen10 Plus/ProLiant DL365 Gen10 Plus, BIOS A42 10/29/2021
[ 3673.988878] Call Trace:
[ 3673.988891]  dump_stack_lvl+0x34/0x44
[ 3673.988908]  __schedule_bug.cold+0x47/0x53
[ 3673.988918]  __schedule+0x3fb/0x560
[ 3673.988929]  schedule+0x43/0xb0
[ 3673.988932]  schedule_hrtimeout_range_clock+0xbf/0x1b0
[ 3673.988937]  ? __hrtimer_init+0xc0/0xc0
[ 3673.988950]  usleep_range+0x5e/0x80
[ 3673.988955]  qed_ptt_acquire+0x2b/0xd0 [qed]
[ 3673.988981]  _qed_get_vport_stats+0x141/0x240 [qed]
[ 3673.989001]  qed_get_vport_stats+0x18/0x80 [qed]
[ 3673.989016]  qede_fill_by_demand_stats+0x37/0x400 [qede]
[ 3673.989028]  qede_get_stats64+0x19/0xe0 [qede]
[ 3673.989034]  dev_get_stats+0x5c/0xc0
[ 3673.989045]  netstat_show.constprop.0+0x52/0xb0
[ 3673.989055]  dev_attr_show+0x19/0x40
[ 3673.989065]  sysfs_kf_seq_show+0x9b/0xf0
[ 3673.989076]  seq_read_iter+0x120/0x4b0
[ 3673.989087]  new_sync_read+0x118/0x1a0
[ 3673.989095]  vfs_read+0xf3/0x180
[ 3673.989099]  ksys_read+0x5f/0xe0
[ 3673.989102]  do_syscall_64+0x3b/0x90
[ 3673.989109]  entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 3673.989115] RIP: 0033:0x7f8467d0b082
[ 3673.989119] Code: c0 e9 b2 fe ff ff 50 48 8d 3d ca 05 08 00 e8 35 e7 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24
[ 3673.989121] RSP: 002b:00007ffffb21fd08 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
[ 3673.989127] RAX: ffffffffffffffda RBX: 000000000100eca0 RCX: 00007f8467d0b082
[ 3673.989128] RDX: 00000000000003ff RSI: 00007ffffb21fdc0 RDI: 0000000000000003
[ 3673.989130] RBP: 00007f8467b96028 R08: 0000000000000010 R09: 00007ffffb21ec00
[ 3673.989132] R10: 00007ffffb27b170 R11: 0000000000000246 R12: 00000000000000f0
[ 3673.989134] R13: 0000000000000003 R14: 00007f8467b92000 R15: 0000000000045a05
[ 3673.989139] CPU: 30 PID: 285188 Comm: read_all Kdump: loaded Tainted: G        W  OE

Fix this by collecting the statistics asynchronously from a periodic
delayed work scheduled at default stats coalescing interval and return
the recent copy of statisitcs from .ndo_get_stats64(), also add ability
to configure/retrieve stats coalescing interval using below commands -

ethtool -C ethx stats-block-usecs <val>
ethtool -c ethx

Fixes: 133fac0eedc3 ("qede: Add basic ethtool support")
Cc: Sudarsana Kalluru <skalluru@marvell.com>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Manish Chopra <manishc@marvell.com>
Link: https://lore.kernel.org/r/20230605112600.48238-1-manishc@marvell.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
12 months agoMerge tag 'for-linus-2023060501' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 6 Jun 2023 11:36:02 +0000 (04:36 -0700)]
Merge tag 'for-linus-2023060501' of git://git./linux/kernel/git/hid/hid

Pull HID fix from Jiri Kosina:

 - Final, confirmed fix for regression causing some devices connected
   via Logitech HID++ Unifying receiver take too long to initialize
   (Benjamin Tissoires)

* tag 'for-linus-2023060501' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:
  HID: hidpp: terminate retry loop on success

12 months agowifi: iwlwifi: mvm: Fix -Warray-bounds bug in iwl_mvm_wait_d3_notif()
Gustavo A. R. Silva [Fri, 2 Jun 2023 19:42:47 +0000 (13:42 -0600)]
wifi: iwlwifi: mvm: Fix -Warray-bounds bug in iwl_mvm_wait_d3_notif()

kmemdup() at line 2735 is not duplicating enough memory for
notif->tid_tear_down and notif->station_id. As it only duplicates
612 bytes: up to offsetofend(struct iwl_wowlan_info_notif,
received_beacons), this is the range of [0, 612) bytes.

2735 notif = kmemdup(notif_v1,
2736 offsetofend(struct iwl_wowlan_info_notif,
2737     received_beacons),
2738 GFP_ATOMIC);

which evidently does not cover bytes 612 and 613 for members
tid_tear_down and station_id in struct iwl_wowlan_info_notif.
See below:

$ pahole -C iwl_wowlan_info_notif drivers/net/wireless/intel/iwlwifi/mvm/d3.o
struct iwl_wowlan_info_notif {
struct iwl_wowlan_gtk_status_v3 gtk[2];          /*     0   488 */
/* --- cacheline 7 boundary (448 bytes) was 40 bytes ago --- */
struct iwl_wowlan_igtk_status igtk[2];           /*   488    80 */
/* --- cacheline 8 boundary (512 bytes) was 56 bytes ago --- */
__le64                     replay_ctr;           /*   568     8 */
/* --- cacheline 9 boundary (576 bytes) --- */
__le16                     pattern_number;       /*   576     2 */
__le16                     reserved1;            /*   578     2 */
__le16                     qos_seq_ctr[8];       /*   580    16 */
__le32                     wakeup_reasons;       /*   596     4 */
__le32                     num_of_gtk_rekeys;    /*   600     4 */
__le32                     transmitted_ndps;     /*   604     4 */
__le32                     received_beacons;     /*   608     4 */
u8                         tid_tear_down;        /*   612     1 */
u8                         station_id;           /*   613     1 */
u8                         reserved2[2];         /*   614     2 */

/* size: 616, cachelines: 10, members: 13 */
/* last cacheline: 40 bytes */
};

Therefore, when the following assignments take place, actually no memory
has been allocated for those objects:

2743 notif->tid_tear_down = notif_v1->tid_tear_down;
2744 notif->station_id = notif_v1->station_id;

Fix this by allocating space for the whole notif object and zero out the
remaining space in memory after member station_id.

This also fixes the following -Warray-bounds issues:
 CC      drivers/net/wireless/intel/iwlwifi/mvm/d3.o
drivers/net/wireless/intel/iwlwifi/mvm/d3.c: In function ‘iwl_mvm_wait_d3_notif’:
drivers/net/wireless/intel/iwlwifi/mvm/d3.c:2743:30: warning: array subscript ‘struct iwl_wowlan_info_notif[0]’ is partly outside array bounds of ‘unsigned char[612]’ [-Warray-bounds=]
 2743 |                         notif->tid_tear_down = notif_v1->tid_tear_down;
      |
                 from drivers/net/wireless/intel/iwlwifi/mvm/d3.c:7:
In function ‘kmemdup’,
    inlined from ‘iwl_mvm_wait_d3_notif’ at drivers/net/wireless/intel/iwlwifi/mvm/d3.c:2735:12:
include/linux/fortify-string.h:765:16: note: object of size 612 allocated by ‘__real_kmemdup’
  765 |         return __real_kmemdup(p, size, gfp);
      |                ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/wireless/intel/iwlwifi/mvm/d3.c: In function ‘iwl_mvm_wait_d3_notif’:
drivers/net/wireless/intel/iwlwifi/mvm/d3.c:2744:30: warning: array subscript ‘struct iwl_wowlan_info_notif[0]’ is partly outside array bounds of ‘unsigned char[612]’ [-Warray-bounds=]
 2744 |                         notif->station_id = notif_v1->station_id;
      |                              ^~
In function ‘kmemdup’,
    inlined from ‘iwl_mvm_wait_d3_notif’ at drivers/net/wireless/intel/iwlwifi/mvm/d3.c:2735:12:
include/linux/fortify-string.h:765:16: note: object of size 612 allocated by ‘__real_kmemdup’
  765 |         return __real_kmemdup(p, size, gfp);
      |                ^~~~~~~~~~~~~~~~~~~~~~~~~~~~

Link: https://github.com/KSPP/linux/issues/306
Fixes: 905d50ddbc83 ("wifi: iwlwifi: mvm: support wowlan info notification version 2")
Cc: stable@vger.kernel.org
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Acked-by: Gregory Greenman <gregory.greenman@intel.com>
Link: https://lore.kernel.org/r/ZHpGN555FwAKGduH@work
Signed-off-by: Johannes Berg <johannes.berg@intel.com>