| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
There are sleep in atomic bug that could cause kernel panic during
firmware download process. The root cause is that nlmsg_new with
GFP_KERNEL parameter is called in fw_dnld_timeout which is a timer
handler. The call trace is shown below:
BUG: sleeping function called from invalid context at include/linux/sched/mm.h:265
Call Trace:
kmem_cache_alloc_node
__alloc_skb
nfc_genl_fw_download_done
call_timer_fn
__run_timers.part.0
run_timer_softirq
__do_softirq
...
The nlmsg_new with GFP_KERNEL parameter may sleep during memory
allocation process, and the timer handler is run as the result of
a "software interrupt" that should not call any other function
that could sleep.
This patch changes allocation mode of netlink message from GFP_KERNEL
to GFP_ATOMIC in order to prevent sleep in atomic bug. The GFP_ATOMIC
flag makes memory allocation operation could be used in atomic context.
Fixes: 9674da8759df ("NFC: Add firmware upload netlink command")
Fixes: 9ea7187c53f6 ("NFC: netlink: Rename CMD_FW_UPLOAD to CMD_FW_DOWNLOAD")
Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20220504055847.38026-1-duoming@zju.edu.cn
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
OCELOT_POLICER_DISCARD helps "kill dropped packets dead" since a
PERMIT/DENY mask mode with a port mask of 0 isn't enough to stop the CPU
port from receiving packets removed from the forwarding path.
The hardcoded initialization done for it in ocelot_vcap_init() is
confusing. All we need from it is to have a rate and a burst size of 0.
Reuse qos_policer_conf_set() for that purpose.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The "port" argument is used for nothing else except printing on the
error path. Print errors on behalf of the policer index, which is less
confusing anyway.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Unify the code paths for adding to an empty list and to a list with
elements by keeping a "pos" list_head element that indicates where to
insert. Initialize "pos" with the list head itself in case
list_for_each_entry() doesn't iterate over any element.
Note that list_for_each_safe() isn't needed because no element is
removed from the list while iterating.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This makes no functional difference but helps in minimizing the delta
for a future change.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
list_add(..., pos->prev) and list_add_tail(..., pos) are equivalent, use
the later form to unify with the case where the list is empty later.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
In commit 4fdabd509df3 ("dt-bindings: net: lan966x: remove PHY reset")
the PHY reset was removed, but I failed to remove it from the example.
Fix it.
Fixes: 4fdabd509df3 ("dt-bindings: net: lan966x: remove PHY reset")
Reported-by: Rob Herring <robh@kernel.org>
Signed-off-by: Michael Walle <michael@walle.cc>
Acked-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20220503132038.2714128-1-michael@walle.cc
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
As discussed here with Ido Schimmel:
https://patchwork.kernel.org/project/netdevbpf/patch/20220224102908.5255-2-jianbol@nvidia.com/
the default conform-exceed action is "reclassify", for a reason we don't
really understand.
The point is that hardware can't offload that police action, so not
specifying "conform-exceed" was always wrong, even though the command
used to work in hardware (but not in software) until the kernel started
adding validation for it.
Fix the command used by the selftest by making the policer drop on
exceed, and pass the packet to the next action (goto) on conform.
Fixes: 8cd6b020b644 ("selftests: ocelot: add some example VCAP IS1, IS2 and ES0 tc offloads")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://lore.kernel.org/r/20220503121428.842906-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
In commit 190cc82489f4 ("tcp: change source port randomizarion at
connect() time"), the table_perturb[] array was introduced and an
index was taken from the port_offset via hash_32(). But it turns
out that hash_32() performs a multiplication while the input here
comes from the output of SipHash in secure_seq, that is well
distributed enough to avoid the need for yet another hash.
Suggested-by: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Moshe Kol, Amit Klein, and Yossi Gilad reported being able to accurately
identify a client by forcing it to emit only 40 times more connections
than there are entries in the table_perturb[] table. The previous two
improvements consisting in resalting the secret every 10s and adding
randomness to each port selection only slightly improved the situation,
and the current value of 2^8 was too small as it's not very difficult
to make a client emit 10k connections in less than 10 seconds.
Thus we're increasing the perturb table from 2^8 to 2^16 so that the
same precision now requires 2.6M connections, which is more difficult in
this time frame and harder to hide as a background activity. The impact
is that the table now uses 256 kB instead of 1 kB, which could mostly
affect devices making frequent outgoing connections. However such
components usually target a small set of destinations (load balancers,
database clients, perf assessment tools), and in practice only a few
entries will be visited, like before.
A live test at 1 million connections per second showed no performance
difference from the previous value.
Reported-by: Moshe Kol <moshe.kol@mail.huji.ac.il>
Reported-by: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Reported-by: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
We'll need to further increase the size of this table and it's likely
that at some point its size will not be suitable anymore for a static
table. Let's allocate it on boot from inet_hashinfo2_init(), which is
called from tcp_init().
Cc: Moshe Kol <moshe.kol@mail.huji.ac.il>
Cc: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Cc: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Here we're randomly adding between 0 and 7 random increments to the
selected source port in order to add some noise in the source port
selection that will make the next port less predictable.
With the default port range of 32768-60999 this means a worst case
reuse scenario of 14116/8=1764 connections between two consecutive
uses of the same port, with an average of 14116/4.5=3137. This code
was stressed at more than 800000 connections per second to a fixed
target with all connections closed by the client using RSTs (worst
condition) and only 2 connections failed among 13 billion, despite
the hash being reseeded every 10 seconds, indicating a perfectly
safe situation.
Cc: Moshe Kol <moshe.kol@mail.huji.ac.il>
Cc: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Cc: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
In order to limit the ability for an observer to recognize the source
ports sequence used to contact a set of destinations, we should
periodically shuffle the secret. 10 seconds looks effective enough
without causing particular issues.
Cc: Moshe Kol <moshe.kol@mail.huji.ac.il>
Cc: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Cc: Amit Klein <aksecurity@gmail.com>
Cc: Jason A. Donenfeld <Jason@zx2c4.com>
Tested-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Amit Klein suggests that we use different parts of port_offset for the
table's index and the port offset so that there is no direct relation
between them.
Cc: Jason A. Donenfeld <Jason@zx2c4.com>
Cc: Moshe Kol <moshe.kol@mail.huji.ac.il>
Cc: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Cc: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
SipHash replaced MD5 in secure_ipv{4,6}_port_ephemeral() via commit
7cd23e5300c1 ("secure_seq: use SipHash in place of MD5"), but the output
remained truncated to 32-bit only. In order to exploit more bits from the
hash, let's make the functions return the full 64-bit of siphash_3u32().
We also make sure the port offset calculation in __inet_hash_connect()
remains done on 32-bit to avoid the need for div_u64_rem() and an extra
cost on 32-bit systems.
Cc: Jason A. Donenfeld <Jason@zx2c4.com>
Cc: Moshe Kol <moshe.kol@mail.huji.ac.il>
Cc: Yossi Gilad <yossi.gilad@mail.huji.ac.il>
Cc: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Creating a new netdevice allocates at least ~50Kb of memory for various
kernel objects, but only ~5Kb of them are accounted to memcg. As a result,
creating an unlimited number of netdevice inside a memcg-limited container
does not fall within memcg restrictions, consumes a significant part
of the host's memory, can cause global OOM and lead to random kills of
host processes.
The main consumers of non-accounted memory are:
~10Kb 80+ kernfs nodes
~6Kb ipv6_add_dev() allocations
6Kb __register_sysctl_table() allocations
4Kb neigh_sysctl_register() allocations
4Kb __devinet_sysctl_register() allocations
4Kb __addrconf_sysctl_register() allocations
Accounting of these objects allows to increase the share of memcg-related
memory up to 60-70% (~38Kb accounted vs ~54Kb total for dummy netdevice
on typical VM with default Fedora 35 kernel) and this should be enough
to somehow protect the host from misuse inside container.
Other related objects are quite small and may not be taken into account
to minimize the expected performance degradation.
It should be separately mentonied ~300 bytes of percpu allocation
of struct ipstats_mib in snmp6_alloc_dev(), on huge multi-cpu nodes
it can become the main consumer of memory.
This patch does not enables kernfs accounting as it affects
other parts of the kernel and should be discussed separately.
However, even without kernfs, this patch significantly improves the
current situation and allows to take into account more than half
of all netdevice allocations.
Signed-off-by: Vasily Averin <vvs@openvz.org>
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Link: https://lore.kernel.org/r/354a0a5f-9ec3-a25c-3215-304eab2157bc@openvz.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Rather than setting this once init is running, set panic_on_warn from
the kernel command line, so that it catches splats from WireGuard
initialization code and the various crypto selftests.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Use newer, more reliable package dependencies. These should hopefully
reduce flakes. However, we keep the old iputils package, as it
accumulated bugs after resulting in flakes on slow machines.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
When moving to non-system toolchains, we inadvertantly killed the
ability to use ccache. So instead, build ccache support into the test
harness directly.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Rather than relying on the system to have cross toolchains available,
simply download musl.cc's ones and use that libc.so, and then we use it
to fill in a few missing platforms, such as riscv64, riscv64, powerpc64,
and s390x.
Since riscv doesn't have a second serial port in its device description,
we have to use virtio's vport. This is actually the same situation on
ARM, but we were previously hacking QEMU up to work around this, which
required a custom QEMU. Instead just do the vport trick on ARM too.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The parallel tests were added to catch queueing issues from multiple
cores. But what happens in reality when testing tons of processes is
that these separate threads wind up fighting with the scheduler, and we
wind up with contention in places we don't care about that decrease the
chances of hitting a bug. So just do a test with the number of CPU
cores, rather than trying to scale up arbitrarily.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
I hate to do this, but I still do not have a good solution to actually
fix this bug across architectures. So just disable it for now, so that
the CI can still deliver actionable results. This commit adds a large
red warning, so that at least the failure isn't lost forever, and
hopefully this can be revisited down the line.
Link: https://lore.kernel.org/netdev/CAHmME9pv1x6C4TNdL6648HydD8r+txpV4hTUXOBVkrapBXH4QQ@mail.gmail.com/
Link: https://lore.kernel.org/netdev/YmszSXueTxYOC41G@zx2c4.com/
Link: https://lore.kernel.org/wireguard/CAHmME9rNnBiNvBstb7MPwK-7AmAN0sOfnhdR=eeLrowWcKxaaQ@mail.gmail.com/
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Groups created by VFIO backends outside the core IOMMU API should never
be passed directly into the API itself, however they still expose their
standard sysfs attributes, so we can still stumble across them that way.
Take care to consider those cases before jumping into our normal
assumptions of a fully-initialised core API group.
Fixes: 3f6634d997db ("iommu: Use right way to retrieve iommu_ops")
Reported-by: Jan Stancek <jstancek@redhat.com>
Tested-by: Jan Stancek <jstancek@redhat.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/86ada41986988511a8424e84746dfe9ba7f87573.1651667683.git.robin.murphy@arm.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
|
|
The driver periodically queries the device for activity of neighbour
entries in order to report it to the kernel's neighbour table.
Avoid unnecessary activity query when no neighbours are installed. Use
an atomic variable to track the number of neighbours, as it is read
without any locks.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver periodically queries the device for FDB notifications (e.g.,
learned, aged-out) in order to update the bridge driver. These
notifications can only be generated when bridges are offloaded to the
device.
Avoid unnecessary queries by starting to query upon installation of the
first bridge and stop querying upon removal of the last bridge.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver periodically queries the device for activity of ACL rules in
order to report it to tc upon 'FLOW_CLS_STATS'.
In Spectrum-2 and later ASICs, multicast routes are programmed as ACL
rules, but unlike rules installed by tc, their activity is of no
interest.
Avoid unnecessary activity query for such rules by always reporting them
as inactive.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When trapping packets for on-CPU processing, Spectrum machines
differentiate between control and non-control traps. Traffic trapped
through non-control traps is treated as data and kept in shared buffer in
pools 0-4. Traffic trapped through control traps is kept in the dedicated
control buffer 9. The advantage of marking traps as control is that
pressure in the data plane does not prevent the control traffic to be
processed.
When the LLDP trap was introduced, it was marked as a control trap. But
then in commit aed4b5721143 ("mlxsw: spectrum: PTP: Hook into packet
receive path"), PTP traps were introduced. Because Ethernet-encapsulated
PTP packets look to the Spectrum-1 ASIC as LLDP traffic and are trapped
under the LLDP trap, this trap was reconfigured as non-control, in sync
with the PTP traps.
There is however no requirement that PTP traffic be handled as data.
Besides, the usual encapsulation for PTP traffic is UDP, not bare Ethernet,
and that is in deployments that even need PTP, which is far less common
than LLDP. This is reflected by the default policer, which was not bumped
up to the 19Kpps / 24Kpps that is the expected load of a PTP-enabled
Spectrum-1 switch.
Marking of LLDP trap as non-control was therefore probably misguided. In
this patch, change it back to control.
Reported-by: Maksym Yaremchuk <maksymy@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The idea behind the warnings is that the user would get warned in case when
more than one priority is configured for a given DSCP value on a netdevice.
The warning is currently wrong, because dcb_ieee_getapp_mask() returns
the first matching entry, not all of them, and the warning will then claim
that some priority is "current", when in fact it is not.
But more importantly, the warning is misleading in general. Consider the
following commands:
# dcb app flush dev swp19 dscp-prio
# dcb app add dev swp19 dscp-prio 24:3
# dcb app replace dev swp19 dscp-prio 24:2
The last command will issue the following warning:
mlxsw_spectrum3 0000:07:00.0 swp19: Ignoring new priority 2 for DSCP 24 in favor of current value of 3
The reason is that the "replace" command works by first adding the new
value, and then removing all old values. This is the only way to make the
replacement without causing the traffic to be prioritized to whatever the
chip defaults to. The warning is issued in response to adding the new
priority, and then no warning is shown when the old priority is removed.
The upshot is that the canonical way to change traffic prioritization
always produces a warning about ignoring the new priority, but what gets
configured is in fact what the user intended.
An option to just emit warning every time that the prioritization changes
just to make it clear that it happened is obviously unsatisfactory.
Therefore, in this patch, remove the warnings.
Reported-by: Maksym Yaremchuk <maksymy@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
It is customary for selftests to have a comment with a topology diagram,
which serves to illustrate the situation in which the test is done. This
selftest lacks it. Add it.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
It is customary for selftests to have a comment with a topology diagram,
which serves to illustrate the situation in which the test is done. This
selftest lacks it. Add it.
While at it, fix the list of tests so that the test names are enumerated
one at a line.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
A number of mlxsw-specific QoS tests use manual QoS DCB management. As
such, they need to make sure lldpad is not running, because it would
override the configuration the test has applied using other tools. To that
end, these selftests invoke the bail_on_lldpad() helper, which terminates
the selftest if th lldpad is running.
Some of these tests however first install the bash exit trap, which invokes
a cleanup() at the test exit. If bail_on_lldpad() has terminated the script
even before the setup part was run, the cleanup part will be very confused.
Therefore make sure bail_on_lldpad() is invoked before the cleanup is
registered.
While there are still edge cases where the user terminates the script
before the setup was fully done, this takes care of a common situation
where the cleanup would be invoked in an inconsistent state.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
Duoming Zhou
Vladimir Oltean
Michael Walle
Willy Tarreau
Eric Dumazet
Vasily Averin
Jason A. Donenfeld
Robin Murphy
Ido Schimmel
Petr Machata