In their .attach callback, mq[prio] only add the qdiscs of the currently
active TX queues to the device's qdisc hash list.
If a user later increases the number of active TX queues, their qdiscs
are not visible via eg. 'tc qdisc show'.
Add a hook to netif_set_real_num_tx_queues() that walks all active
TX queues and adds those which are missing to the hash list.
CC: Eric Dumazet <edumazet@google.com>
CC: Jamal Hadi Salim <jhs@mojatatu.com>
CC: Cong Wang <xiyou.wangcong@gmail.com>
CC: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 105e808c1d ("pie: remove pie_vars->accu_prob_overflows")
changes the scale of probability values in PIE from (2^64 - 1) to
(2^56 - 1). This affects the precision of tc_pie_xstats->prob in
user space.
This patch ensures user space is unaffected.
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, user who is adding an action expects HW to report stats,
however it does not have exact expectations about the stats types.
That is aligned with TCA_ACT_HW_STATS_TYPE_ANY.
Allow user to specify the type of HW stats for an action and require it.
Pass the information down to flow_offload layer.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Invoke ndo_setup_tc() as appropriate to signal init / replacement,
destroying and dumping of pFIFO / bFIFO Qdisc.
A lot of the FIFO logic is used for pFIFO_head_drop as well, but that's a
semantically very different Qdisc that isn't really in the same boat as
pFIFO / bFIFO. Split some of the functions to keep the Qdisc intact.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The variable pie_vars->accu_prob is used as an accumulator for
probability values. Since probabilty values are scaled using the
MAX_PROB macro denoting (2^64 - 1), pie_vars->accu_prob is
likely to overflow as it is of type u64.
The variable pie_vars->accu_prob_overflows counts the number of
times the variable pie_vars->accu_prob overflows.
The MAX_PROB macro needs to be equal to at least (2^39 - 1) in
order to do precise calculations without any underflow. Thus
MAX_PROB can be reduced to (2^56 - 1) without affecting the
precision in calculations drastically. Doing so will eliminate
the need for the variable pie_vars->accu_prob_overflows as the
variable pie_vars->accu_prob will never overflow.
Removing the variable pie_vars->accu_prob_overflows also reduces
the size of the structure pie_vars to exactly 64 bytes.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In function pie_calculate_probability(), the variables alpha and
beta are of type u64. The variables qdelay, qdelay_old and
params->target are of type psched_time_t (which is also u64).
The explicit type casting done when calculating the value for
the variable delta is redundant and not required.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove ambiguity by using the term backlog instead of qlen when
representing the queue length in bytes.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To make the filler functions more generic, use network
relative skb pulling.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When checking the protocol number tcf_ct_flow_table_lookup() handles
the flow as if it's always ipv4, while it can be ipv6.
Instead, refactor the code to fetch the tcp header, if available,
in the relevant family (ipv4/ipv6) filler function, and do the
check on the returned tcp header.
Fixes: 46475bb20f ("net/sched: act_ct: Software offload of established flows")
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Offload nf conntrack processing by looking up the 5-tuple in the
zone's flow table.
The nf conntrack module will process the packets until a connection is
in established state. Once in established state, the ct state pointer
(nf_conn) will be restored on the skb from a successful ft lookup.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a ft entry when connections enter an established state and delete
the connections when they leave the established state.
The flow table assumes ownership of the connection. In the following
patch act_ct will lookup the ct state from the FT. In future patches,
drivers will register for callbacks for ft add/del events and will be
able to use the information to offload the connections.
Note that connection aging is managed by the FT.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use the NF flow tables infrastructure for CT offload.
Create a nf flow table per zone.
Next patches will add FT entries to this table, and do
the software offload.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add missing attribute validation for TCA_TAPRIO_ATTR_TXTIME_DELAY
to the netlink policy.
Fixes: 4cfd5779bd ("taprio: Add support for txtime-assist mode")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add missing attribute validation for TCA_FQ_ORPHAN_MASK
to the netlink policy.
Fixes: 06eb395fa9 ("pkt_sched: fq: better control of DDOS traffic")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The mptcp conflict was overlapping additions.
The SMC conflict was an additional and removal happening at the same
time.
Signed-off-by: David S. Miller <davem@davemloft.net>
The put of the flags was added by the commit referenced in fixes tag,
however the size of the message was not extended accordingly.
Fix this by adding size of the flags bitfield to the message size.
Fixes: e382267860 ("net: sched: update action implementations to support flags")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend struct flow_action_entry in order to hold TC action cookie
specified by user inserting the action.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set the starting chain from the tc skb ext chain value. Once we read
the tc skb ext, delete it, so cloned/redirect packets won't inherit it.
In order to lookup a chain by the chain index on the ingress block
at ingress classification, provide a lookup function.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
On ingress and cls_act qdiscs init, save the block on ingress
mini_Qdisc and and pass it on to ingress classification, so it
can be used for the looking up a specified chain index.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
TC multi chain configuration can cause offloaded tc chains to miss in
hardware after jumping to some chain. In such cases the software should
continue from the chain that missed in hardware, as the hardware may
have manipulated the packet and updated some counters.
Currently a single tcf classification function serves both ingress and
egress. However, multi chain miss processing (get tc skb extension on
hw miss, set tc skb extension on tc miss) should happen only on
ingress.
Refactor the code to use ingress classification function, and move setting
the tc skb extension from general classification to it, as a prestep
for supporting the hw miss scenario.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
tc flower rules that are based on src or dst port blocking are sometimes
ineffective due to uninitialized stack data. __skb_flow_dissect() extracts
ports from the skb for tc flower to match against. However, the port
dissection is not done when when the FLOW_DIS_IS_FRAGMENT bit is set in
key_control->flags. All callers of __skb_flow_dissect(), zero-out the
key_control field except for fl_classify() as used by the flower
classifier. Thus, the FLOW_DIS_IS_FRAGMENT may be set on entry to
__skb_flow_dissect(), since key_control is allocated on the stack
and may not be initialized.
Since key_basic and key_control are present for all flow keys, let's
make sure they are initialized.
Fixes: 62230715fd ("flow_dissector: do not dissect l4 ports for fragments")
Co-developed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor tc_setup_flow_action() function not to use rtnl lock and remove
'rtnl_held' argument that is no longer needed.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to remove dependency on rtnl lock, take action's tcfa_lock when
constructing its representation as flow_action_entry structure.
Refactor tcf_sample_get_group() to assume that caller holds tcf_lock and
don't take it manually. This callback is only called from flow_action infra
representation translator which now calls it with tcf_lock held, so this
refactoring is necessary to prevent deadlock.
Allocate memory with GFP_ATOMIC flag for ip_tunnel_info copy because
tcf_tunnel_info_copy() is only called from flow_action representation infra
code with tcf_lock spinlock taken.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
unlike other classifiers that can be offloaded (i.e. users can set flags
like 'skip_hw' and 'skip_sw'), 'cls_flower' doesn't validate the size of
netlink attribute 'TCA_FLOWER_FLAGS' provided by user: add a proper entry
to fl_policy.
Fixes: 5b33f48842 ("net/flower: Introduce hardware offload support")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
unlike other classifiers that can be offloaded (i.e. users can set flags
like 'skip_hw' and 'skip_sw'), 'cls_matchall' doesn't validate the size
of netlink attribute 'TCA_MATCHALL_FLAGS' provided by user: add a proper
entry to mall_policy.
Fixes: b87f7936a9 ("net/sched: Add match-all classifier hw offloading.")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When using taprio offloading together with ETF offloading, configured
like this, for example:
$ tc qdisc replace dev $IFACE parent root handle 100 taprio \
num_tc 4 \
map 2 2 1 0 3 2 2 2 2 2 2 2 2 2 2 2 \
queues 1@0 1@1 1@2 1@3 \
base-time $BASE_TIME \
sched-entry S 01 1000000 \
sched-entry S 0e 1000000 \
flags 0x2
$ tc qdisc replace dev $IFACE parent 100:1 etf \
offload delta 300000 clockid CLOCK_TAI
During enqueue, it works out that the verification added for the
"txtime" assisted mode is run when using taprio + ETF offloading, the
only thing missing is initializing the 'next_txtime' of all the cycle
entries. (if we don't set 'next_txtime' all packets from SO_TXTIME
sockets are dropped)
Fixes: 4cfd5779bd ("taprio: Add support for txtime-assist mode")
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When destroying the current taprio instance, which can happen when the
creation of one fails, we should reset the traffic class configuration
back to the default state.
netdev_reset_tc() is a better way because in addition to setting the
number of traffic classes to zero, it also resets the priority to
traffic classes mapping to the default value.
Fixes: 5a781ccbd1 ("tc: Add support for configuring the taprio scheduler")
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netlink policy validation for the 'flags' argument was missing.
Fixes: 4cfd5779bd ("taprio: Add support for txtime-assist mode")
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the driver implementing taprio offloading depends on the value of
the network device number of traffic classes (dev->num_tc) for
whatever reason, it was going to receive the value zero. The value was
only set after the offloading function is called.
So, moving setting the number of traffic classes to before the
offloading function is called fixes this issue. This is safe because
this only happens when taprio is instantiated (we don't allow this
configuration to be changed without first removing taprio).
Fixes: 9c66d15646 ("taprio: Add support for hardware offloading")
Reported-by: Po Liu <po.liu@nxp.com>
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Acked-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The bug is that we call kfree_skb(skb) and then pass "skb" to
qdisc_pkt_len(skb) on the next line, which is a use after free.
Also Cong Wang points out that it's better to delay the actual
frees until we drop the rtnl lock so we should use rtnl_kfree_skbs()
instead of kfree_skb().
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Fixes: ec97ecf1eb ("net: sched: add Flow Queue PIE packet scheduler")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub noticed there is a potential resource leak in
tcindex_set_parms(): when tcindex_filter_result_init() fails
and it jumps to 'errout1' which doesn't release the memory
and resources allocated by tcindex_alloc_perfect_hash().
We should just jump to 'errout_alloc' which calls
tcindex_free_perfect_hash().
Fixes: b9a24bb76b ("net_sched: properly handle failure case of tcf_exts_init()")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As Eric noticed, tcindex_alloc_perfect_hash() uses cp->hash
to compute the size of memory allocation, but cp->hash is
set again after the allocation, this caused an out-of-bound
access.
So we have to move all cp->hash initialization and computation
before the memory allocation. Move cp->mask and cp->shift together
as cp->hash may need them for computation too.
Reported-and-tested-by: syzbot+35d4dea36c387813ed31@syzkaller.appspotmail.com
Fixes: 331b72922c ("net: sched: RCU cls_tcindex")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert the use of kvmalloc_array with __GFP_ZERO to
the equivalent kvcalloc.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In a complex TC class hierarchy like this:
tc qdisc add dev eth0 root handle 1:0 cbq bandwidth 100Mbit \
avpkt 1000 cell 8
tc class add dev eth0 parent 1:0 classid 1:1 cbq bandwidth 100Mbit \
rate 6Mbit weight 0.6Mbit prio 8 allot 1514 cell 8 maxburst 20 \
avpkt 1000 bounded
tc filter add dev eth0 parent 1:0 protocol ip prio 1 u32 match ip \
sport 80 0xffff flowid 1:3
tc filter add dev eth0 parent 1:0 protocol ip prio 1 u32 match ip \
sport 25 0xffff flowid 1:4
tc class add dev eth0 parent 1:1 classid 1:3 cbq bandwidth 100Mbit \
rate 5Mbit weight 0.5Mbit prio 5 allot 1514 cell 8 maxburst 20 \
avpkt 1000
tc class add dev eth0 parent 1:1 classid 1:4 cbq bandwidth 100Mbit \
rate 3Mbit weight 0.3Mbit prio 5 allot 1514 cell 8 maxburst 20 \
avpkt 1000
where filters are installed on qdisc 1:0, so we can't merely
search from class 1:1 when creating class 1:3 and class 1:4. We have
to walk through all the child classes of the direct parent qdisc.
Otherwise we would miss filters those need reverse binding.
Fixes: 07d79fc7d9 ("net_sched: add reverse binding for tc class")
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current implementations of ops->bind_class() are merely
searching for classid and updating class in the struct tcf_result,
without invoking either of cl_ops->bind_tcf() or
cl_ops->unbind_tcf(). This breaks the design of them as qdisc's
like cbq use them to count filters too. This is why syzbot triggered
the warning in cbq_destroy_class().
In order to fix this, we have to call cl_ops->bind_tcf() and
cl_ops->unbind_tcf() like the filter binding path. This patch does
so by refactoring out two helper functions __tcf_bind_filter()
and __tcf_unbind_filter(), which are lockless and accept a Qdisc
pointer, then teaching each implementation to call them correctly.
Note, we merely pass the Qdisc pointer as an opaque pointer to
each filter, they only need to pass it down to the helper
functions without understanding it at all.
Fixes: 07d79fc7d9 ("net_sched: add reverse binding for tc class")
Reported-and-tested-by: syzbot+0a0596220218fcb603a8@syzkaller.appspotmail.com
Reported-and-tested-by: syzbot+63bdb6006961d8c917c6@syzkaller.appspotmail.com
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Invoke ndo_setup_tc as appropriate to signal init / replacement, destroying
and dumping of TBF Qdisc.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In 2011, in commit b0460e4484 ("sch_tbf: report backlog information"),
TBF started copying backlog depth from the child Qdisc before dumping, with
the motivation that the backlog was otherwise not visible in "tc -s qdisc
show".
Later, in 2016, in commit 8d5958f424 ("sch_tbf: update backlog as well"),
TBF got a full-blown backlog tracking. However it kept copying the child's
backlog over before dumping.
That line is now unnecessary, so remove it.
As shown in the following example, backlog is still reported correctly:
# tc -s qdisc show dev veth0 invisible
qdisc tbf 1: root refcnt 2 rate 1Mbit burst 128Kb lat 82.8s
Sent 505475370 bytes 406985 pkt (dropped 0, overlimits 812544 requeues 0)
backlog 81972b 66p requeues 0
qdisc bfifo 0: parent 1:1 limit 10Mb
Sent 505475370 bytes 406985 pkt (dropped 0, overlimits 0 requeues 0)
backlog 81972b 66p requeues 0
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
syzbot reported an out-of-bound access in em_nbyte. As initially
analyzed by Eric, this is because em_nbyte sets its own em->datalen
in em_nbyte_change() other than the one specified by user, but this
value gets overwritten later by its caller tcf_em_validate().
We should leave em->datalen untouched to respect their choices.
I audit all the in-tree ematch users, all of those implement
->change() set em->datalen, so we can just avoid setting it twice
in this case.
Reported-and-tested-by: syzbot+5af9a90dad568aa9f611@syzkaller.appspotmail.com
Reported-by: syzbot+2f07903a5b05e7f36410@syzkaller.appspotmail.com
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Principles:
- Packets are classified on flows.
- This is a Stochastic model (as we use a hash, several flows might
be hashed to the same slot)
- Each flow has a PIE managed queue.
- Flows are linked onto two (Round Robin) lists,
so that new flows have priority on old ones.
- For a given flow, packets are not reordered.
- Drops during enqueue only.
- ECN capability is off by default.
- ECN threshold (if ECN is enabled) is at 10% by default.
- Uses timestamps to calculate queue delay by default.
Usage:
tc qdisc ... fq_pie [ limit PACKETS ] [ flows NUMBER ]
[ target TIME ] [ tupdate TIME ]
[ alpha NUMBER ] [ beta NUMBER ]
[ quantum BYTES ] [ memory_limit BYTES ]
[ ecnprob PERCENTAGE ] [ [no]ecn ]
[ [no]bytemode ] [ [no_]dq_rate_estimator ]
defaults:
limit: 10240 packets, flows: 1024
target: 15 ms, tupdate: 15 ms (in jiffies)
alpha: 1/8, beta : 5/4
quantum: device MTU, memory_limit: 32 Mb
ecnprob: 10%, ecn: off
bytemode: off, dq_rate_estimator: off
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Sachin D. Patil <sdp.sachin@gmail.com>
Signed-off-by: V. Saicharan <vsaicharan1998@gmail.com>
Signed-off-by: Mohit Bhasi <mohitbhasi1998@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch makes the drop_early(), calculate_probability() and
pie_process_dequeue() functions generic enough to be used by
both PIE and FQ-PIE (to be added in a future commit). The major
change here is in the way the functions take in arguments. This
patch exports these functions and makes FQ-PIE dependent on
sch_pie.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make the alignment in the initialization of the struct instances
consistent in the file.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix punctuation and logical mistakes in the comments. The
logical mistake was that "dequeue_rate" is no longer the default
way to calculate queuing delay and is not needed. The default
way to calculate queue delay was changed in commit cec2975f2b
("net: sched: pie: enable timestamp based delay calculation").
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rearrange the members of the structure such that closely
referenced members appear together and/or fit in the same
cacheline. Also, change the order of their initializations to
match the order in which they appear in the structure.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch moves macros, structures and small functions common
to PIE and FQ-PIE (to be added in a future commit) from the file
net/sched/sch_pie.c to the header file include/net/pie.h.
All the moved functions are made inline.
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is a straight-forward conversion case for the new function, keeping
the flow of the existing code as intact as possible.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ungrafting from PRIO bug fixes in net, when merged into net-next,
merge cleanly but create a build failure. The resolution used here is
from Petr Machata.
Signed-off-by: David S. Miller <davem@davemloft.net>
When a child Qdisc is removed from one of the PRIO Qdisc's bands, it is
replaced unconditionally by a NOOP qdisc. As a result, any traffic hitting
that band gets dropped. That is incorrect--no Qdisc was explicitly added
when PRIO was created, and after removal, none should have to be added
either.
Fix PRIO by first attempting to create a default Qdisc and only falling
back to noop when that fails. This pattern of attempting to create an
invisible FIFO, using NOOP only as a fallback, is also seen in other
Qdiscs.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As diagnosed by Florian :
If TCA_FQ_QUANTUM is set to 0x80000000, fq_deueue()
can loop forever in :
if (f->credit <= 0) {
f->credit += q->quantum;
goto begin;
}
... because f->credit is either 0 or -2147483648.
Let's limit TCA_FQ_QUANTUM to no more than 1 << 20 :
This max value should limit risks of breaking user setups
while fixing this bug.
Fixes: afe4fd0624 ("pkt_sched: fq: Fair Queue packet scheduler")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Diagnosed-by: Florian Westphal <fw@strlen.de>
Reported-by: syzbot+dc9071cc5a85950bdfce@syzkaller.appspotmail.com
Signed-off-by: David S. Miller <davem@davemloft.net>
The variables 'window_interval' is u64 and do_div()
truncates it to 32 bits, which means it can test
non-zero and be truncated to zero for division.
The unit of window_interval is nanoseconds,
so its lower 32-bit is relatively easy to exceed.
Fix this issue by using div64_u64() instead.
Fixes: 7298de9cd7 ("sch_cake: Add ingress mode")
Signed-off-by: Wen Yang <wenyang@linux.alibaba.com>
Cc: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Cc: Toke Høiland-Jørgensen <toke@redhat.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: cake@lists.bufferbloat.net
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Revert "net/sched: cls_u32: fix refcount leak in the error path of
u32_change()", and fix the u32 refcount leak in a more generic way that
preserves the semantic of rule dumping.
On tc filters that don't support lockless insertion/removal, there is no
need to guard against concurrent insertion when a removal is in progress.
Therefore, for most of them we can avoid a full walk() when deleting, and
just decrease the refcount, like it was done on older Linux kernels.
This fixes situations where walk() was wrongly detecting a non-empty
filter, like it happened with cls_u32 in the error path of change(), thus
leading to failures in the following tdc selftests:
6aa7: (filter, u32) Add/Replace u32 with source match and invalid indev
6658: (filter, u32) Add/Replace u32 with custom hash table and invalid handle
74c2: (filter, u32) Add/Replace u32 filter with invalid hash table id
On cls_flower, and on (future) lockless filters, this check is necessary:
move all the check_empty() logic in a callback so that each filter
can have its own implementation. For cls_flower, it's sufficient to check
if no IDRs have been allocated.
This reverts commit 275c44aa19.
Changes since v1:
- document the need for delete_empty() when TCF_PROTO_OPS_DOIT_UNLOCKED
is used, thanks to Vlad Buslov
- implement delete_empty() without doing fl_walk(), thanks to Vlad Buslov
- squash revert and new fix in a single patch, to be nice with bisect
tests that run tdc on u32 filter, thanks to Dave Miller
Fixes: 275c44aa19 ("net/sched: cls_u32: fix refcount leak in the error path of u32_change()")
Fixes: 6676d5e416 ("net: sched: set dedicated tcf_walker flag when tp is empty")
Suggested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Suggested-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Tested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's no skb_pull performed when a mirred action is set at egress of a
mac device, with a target device/action that expects skb->data to point
at the network header.
As a result, either the target device is errornously given an skb with
data pointing to the mac (egress case), or the net stack receives the
skb with data pointing to the mac (ingress case).
E.g:
# tc qdisc add dev eth9 root handle 1: prio
# tc filter add dev eth9 parent 1: prio 9 protocol ip handle 9 basic \
action mirred egress redirect dev tun0
(tun0 is a tun device. result: tun0 errornously gets the eth header
instead of the iph)
Revise the push/pull logic of tcf_mirred_act() to not rely on the
skb_at_tc_ingress() vs tcf_mirred_act_wants_ingress() comparison, as it
does not cover all "pull" cases.
Instead, calculate whether the required action on the target device
requires the data to point at the network header, and compare this to
whether skb->data points to network header - and make the push/pull
adjustments as necessary.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Shmulik Ladkani <sladkani@proofpoint.com>
Tested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If fq_classify() recycles a struct fq_flow because
a socket structure has been reallocated, we do not
set sk->sk_pacing_status immediately, but later if the
flow becomes detached.
This means that any flow requiring pacing (BBR, or SO_MAX_PACING_RATE)
might fallback to TCP internal pacing, which requires a per-socket
high resolution timer, and therefore more cpu cycles.
Fixes: 218af599fa ("tcp: internal implementation for pacing")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking fixes from David Miller:
1) Several nf_flow_table_offload fixes from Pablo Neira Ayuso,
including adding a missing ipv6 match description.
2) Several heap overflow fixes in mwifiex from qize wang and Ganapathi
Bhat.
3) Fix uninit value in bond_neigh_init(), from Eric Dumazet.
4) Fix non-ACPI probing of nxp-nci, from Stephan Gerhold.
5) Fix use after free in tipc_disc_rcv(), from Tuong Lien.
6) Enforce limit of 33 tail calls in mips and riscv JIT, from Paul
Chaignon.
7) Multicast MAC limit test is off by one in qede, from Manish Chopra.
8) Fix established socket lookup race when socket goes from
TCP_ESTABLISHED to TCP_LISTEN, because there lacks an intervening
RCU grace period. From Eric Dumazet.
9) Don't send empty SKBs from tcp_write_xmit(), also from Eric Dumazet.
10) Fix active backup transition after link failure in bonding, from
Mahesh Bandewar.
11) Avoid zero sized hash table in gtp driver, from Taehee Yoo.
12) Fix wrong interface passed to ->mac_link_up(), from Russell King.
13) Fix DSA egress flooding settings in b53, from Florian Fainelli.
14) Memory leak in gmac_setup_txqs(), from Navid Emamdoost.
15) Fix double free in dpaa2-ptp code, from Ioana Ciornei.
16) Reject invalid MTU values in stmmac, from Jose Abreu.
17) Fix refcount leak in error path of u32 classifier, from Davide
Caratti.
18) Fix regression causing iwlwifi firmware crashes on boot, from Anders
Kaseorg.
19) Fix inverted return value logic in llc2 code, from Chan Shu Tak.
20) Disable hardware GRO when XDP is attached to qede, frm Manish
Chopra.
21) Since we encode state in the low pointer bits, dst metrics must be
at least 4 byte aligned, which is not necessarily true on m68k. Add
annotations to fix this, from Geert Uytterhoeven.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (160 commits)
sfc: Include XDP packet headroom in buffer step size.
sfc: fix channel allocation with brute force
net: dst: Force 4-byte alignment of dst_metrics
selftests: pmtu: fix init mtu value in description
hv_netvsc: Fix unwanted rx_table reset
net: phy: ensure that phy IDs are correctly typed
mod_devicetable: fix PHY module format
qede: Disable hardware gro when xdp prog is installed
net: ena: fix issues in setting interrupt moderation params in ethtool
net: ena: fix default tx interrupt moderation interval
net/smc: unregister ib devices in reboot_event
net: stmmac: platform: Fix MDIO init for platforms without PHY
llc2: Fix return statement of llc_stat_ev_rx_null_dsap_xid_c (and _test_c)
net: hisilicon: Fix a BUG trigered by wrong bytes_compl
net: dsa: ksz: use common define for tag len
s390/qeth: don't return -ENOTSUPP to userspace
s390/qeth: fix promiscuous mode after reset
s390/qeth: handle error due to unsupported transport mode
cxgb4: fix refcount init for TC-MQPRIO offload
tc-testing: initial tdc selftests for cls_u32
...
when users replace cls_u32 filters with new ones having wrong parameters,
so that u32_change() fails to validate them, the kernel doesn't roll-back
correctly, and leaves semi-configured rules.
Fix this in u32_walk(), avoiding a call to the walker function on filters
that don't have a match rule connected. The side effect is, these "empty"
filters are not even dumped when present; but that shouldn't be a problem
as long as we are restoring the original behaviour, where semi-configured
filters were not even added in the error path of u32_change().
Fixes: 6676d5e416 ("net: sched: set dedicated tcf_walker flag when tp is empty")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add hooks at appropriate points to make it possible to offload the ETS
Qdisc.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduces a new Qdisc, which is based on 802.1Q-2014 wording. It is
PRIO-like in how it is configured, meaning one needs to specify how many
bands there are, how many are strict and how many are dwrr, quanta for the
latter, and priomap.
The new Qdisc operates like the PRIO / DRR combo would when configured as
per the standard. The strict classes, if any, are tried for traffic first.
When there's no traffic in any of the strict queues, the ETS ones (if any)
are treated in the same way as in DRR.
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Turns out tin_quantum_prio isn't used anymore and is a leftover from a
previous implementation of diffserv tins. Since the variable isn't used
in any calculations it can be eliminated.
Drop variable and places where it was set. Rename remaining variable
and consolidate naming of intermediate variables that set it.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace all the occurrences of FIELD_SIZEOF() with sizeof_field() except
at places where these are defined. Later patches will remove the unused
definition of FIELD_SIZEOF().
This patch is generated using following script:
EXCLUDE_FILES="include/linux/stddef.h|include/linux/kernel.h"
git grep -l -e "\bFIELD_SIZEOF\b" | while read file;
do
if [[ "$file" =~ $EXCLUDE_FILES ]]; then
continue
fi
sed -i -e 's/\bFIELD_SIZEOF\b/sizeof_field/g' $file;
done
Signed-off-by: Pankaj Bharadiya <pankaj.laxminarayan.bharadiya@intel.com>
Link: https://lore.kernel.org/r/20190924105839.110713-3-pankaj.laxminarayan.bharadiya@intel.com
Co-developed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: David Miller <davem@davemloft.net> # for net
When a device is bound to a clsact qdisc, bind events are triggered to
registered drivers for both ingress and egress. However, if a driver
registers to such a device using the indirect block routines then it is
assumed that it is only interested in ingress offload and so only replays
ingress bind/unbind messages.
The NFP driver supports the offload of some egress filters when
registering to a block with qdisc of type clsact. However, on unregister,
if the block is still active, it will not receive an unbind egress
notification which can prevent proper cleanup of other registered
callbacks.
Modify the indirect block callback command in TC to send messages of
ingress and/or egress bind depending on the qdisc in use. NFP currently
supports egress offload for TC flower offload so the changes are only
added to TC.
Fixes: 4d12ba4278 ("nfp: flower: allow offloading of matches on 'internal' ports")
Signed-off-by: John Hurley <john.hurley@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With indirect blocks, a driver can register for callbacks from a device
that is does not 'own', for example, a tunnel device. When registering to
or unregistering from a new device, a callback is triggered to generate
a bind/unbind event. This, in turn, allows the driver to receive any
existing rules or to properly clean up installed rules.
When first added, it was assumed that all indirect block registrations
would be for ingress offloads. However, the NFP driver can, in some
instances, support clsact qdisc binds for egress offload.
Change the name of the indirect block callback command in flow_offload to
remove the 'ingress' identifier from it. While this does not change
functionality, a follow up patch will implement a more more generic
callback than just those currently just supporting ingress offload.
Fixes: 4d12ba4278 ("nfp: flower: allow offloading of matches on 'internal' ports")
Signed-off-by: John Hurley <john.hurley@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When user runs a command like
tc qdisc add dev eth1 root mqprio
KASAN stack-out-of-bounds warning is emitted.
Currently, NLA_ALIGN macro used in mqprio_dump provides too large
buffer size as argument for nla_put and memcpy down the call stack.
The flow looks like this:
1. nla_put expects exact object size as an argument;
2. Later it provides this size to memcpy;
3. To calculate correct padding for SKB, nla_put applies NLA_ALIGN
macro itself.
Therefore, NLA_ALIGN should not be applied to the nla_put parameter.
Otherwise it will lead to out-of-bounds memory access in memcpy.
Fixes: 4e8b86c062 ("mqprio: Introduce new hardware offload mode and shaper in mqprio")
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The skb_mpls_push was not updating ethertype of an ethernet packet if
the packet was originally received from a non ARPHRD_ETHER device.
In the below OVS data path flow, since the device corresponding to
port 7 is an l3 device (ARPHRD_NONE) the skb_mpls_push function does
not update the ethertype of the packet even though the previous
push_eth action had added an ethernet header to the packet.
recirc_id(0),in_port(7),eth_type(0x0800),ipv4(tos=0/0xfc,ttl=64,frag=no),
actions:push_eth(src=00:00:00:00:00:00,dst=00:00:00:00:00:00),
push_mpls(label=13,tc=0,ttl=64,bos=1,eth_type=0x8847),4
Fixes: 8822e270d6 ("net: core: move push MPLS functionality from OvS to core helper")
Signed-off-by: Martin Varghese <martin.varghese@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The act_ct TC module shares a common conntrack and NAT infrastructure
exposed via netfilter. It's possible that a packet needs both SNAT and
DNAT manipulation, due to e.g. tuple collision. Netfilter can support
this because it runs through the NAT table twice - once on ingress and
again after egress. The act_ct action doesn't have such capability.
Like netfilter hook infrastructure, we should run through NAT twice to
keep the symmetry.
Fixes: b57dc7c13e ("net/sched: Introduce action ct")
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The recent commit 5c72299fba ("net: sched: cls_flower: Classify
packets using port ranges") had added filtering based on port ranges
to tc flower. However the commit missed necessary changes in hw-offload
code, so the feature gave rise to generating incorrect offloaded flow
keys in NIC.
One more detailed example is below:
$ tc qdisc add dev eth0 ingress
$ tc filter add dev eth0 ingress protocol ip flower ip_proto tcp \
dst_port 100-200 action drop
With the setup above, an exact match filter with dst_port == 0 will be
installed in NIC by hw-offload. IOW, the NIC will have a rule which is
equivalent to the following one.
$ tc qdisc add dev eth0 ingress
$ tc filter add dev eth0 ingress protocol ip flower ip_proto tcp \
dst_port 0 action drop
The behavior was caused by the flow dissector which extracts packet
data into the flow key in the tc flower. More specifically, regardless
of exact match or specified port ranges, fl_init_dissector() set the
FLOW_DISSECTOR_KEY_PORTS flag in struct flow_dissector to extract port
numbers from skb in skb_flow_dissect() called by fl_classify(). Note
that device drivers received the same struct flow_dissector object as
used in skb_flow_dissect(). Thus, offloaded drivers could not identify
which of these is used because the FLOW_DISSECTOR_KEY_PORTS flag was
set to struct flow_dissector in either case.
This patch adds the new FLOW_DISSECTOR_KEY_PORTS_RANGE flag and the new
tp_range field in struct fl_flow_key to recognize which filters are applied
to offloaded drivers. At this point, when filters based on port ranges
passed to drivers, drivers return the EOPNOTSUPP error because they do
not support the feature (the newly created FLOW_DISSECTOR_KEY_PORTS_RANGE
flag).
Fixes: 5c72299fba ("net: sched: cls_flower: Classify packets using port ranges")
Signed-off-by: Yoshiki Komachi <komachi.yoshiki@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sch->q.len hasn't been set if the subqueue is a NOLOCK qdisc
in mq_dump() and mqprio_dump().
Fixes: ce679e8df7 ("net: sched: add support for TCQ_F_NOLOCK subqueues to sch_mqprio")
Signed-off-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: Tony Lu <tonylu@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The skb_mpls_pop was not updating ethertype of an ethernet packet if the
packet was originally received from a non ARPHRD_ETHER device.
In the below OVS data path flow, since the device corresponding to port 7
is an l3 device (ARPHRD_NONE) the skb_mpls_pop function does not update
the ethertype of the packet even though the previous push_eth action had
added an ethernet header to the packet.
recirc_id(0),in_port(7),eth_type(0x8847),
mpls(label=12/0xfffff,tc=0/0,ttl=0/0x0,bos=1/1),
actions:push_eth(src=00:00:00:00:00:00,dst=00:00:00:00:00:00),
pop_mpls(eth_type=0x800),4
Fixes: ed246cee09 ("net: core: move pop MPLS functionality from OvS to core helper")
Signed-off-by: Martin Varghese <martin.varghese@nokia.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This field has never been checked since introduction in mainline kernel
Signed-off-by: Victorien Molle <victorien.molle@wifirst.fr>
Signed-off-by: Florent Fourcot <florent.fourcot@wifirst.fr>
Fixes: 2db6dc2662 "sch_cake: Make gso-splitting configurable from userspace"
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a classful qdisc's child qdisc has set the flag
TCQ_F_CPUSTATS (pfifo_fast for example), the child qdisc's
cpu_bstats should be passed to gnet_stats_copy_basic(),
but many classful qdisc didn't do that. As a result,
`tc -s class show dev DEV` always return 0 for bytes and
packets in this case.
Pass the child qdisc's cpu_bstats to gnet_stats_copy_basic()
to fix this issue.
The qstats also has this problem, but it has been fixed
in 5dd431b6b9 ("net: sched: introduce and use qstats read...")
and bstats still remains buggy.
Fixes: 22e0f8b932 ("net: sched: make bstats per cpu and estimator RCU safe")
Signed-off-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: Tony Lu <tonylu@linux.alibaba.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull RCU updates from Ingo Molnar:
"The main changes in this cycle were:
- Dynamic tick (nohz) updates, perhaps most notably changes to force
the tick on when needed due to lengthy in-kernel execution on CPUs
on which RCU is waiting.
- Linux-kernel memory consistency model updates.
- Replace rcu_swap_protected() with rcu_prepace_pointer().
- Torture-test updates.
- Documentation updates.
- Miscellaneous fixes"
* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (51 commits)
security/safesetid: Replace rcu_swap_protected() with rcu_replace_pointer()
net/sched: Replace rcu_swap_protected() with rcu_replace_pointer()
net/netfilter: Replace rcu_swap_protected() with rcu_replace_pointer()
net/core: Replace rcu_swap_protected() with rcu_replace_pointer()
bpf/cgroup: Replace rcu_swap_protected() with rcu_replace_pointer()
fs/afs: Replace rcu_swap_protected() with rcu_replace_pointer()
drivers/scsi: Replace rcu_swap_protected() with rcu_replace_pointer()
drm/i915: Replace rcu_swap_protected() with rcu_replace_pointer()
x86/kvm/pmu: Replace rcu_swap_protected() with rcu_replace_pointer()
rcu: Upgrade rcu_swap_protected() to rcu_replace_pointer()
rcu: Suppress levelspread uninitialized messages
rcu: Fix uninitialized variable in nocb_gp_wait()
rcu: Update descriptions for rcu_future_grace_period tracepoint
rcu: Update descriptions for rcu_nocb_wake tracepoint
rcu: Remove obsolete descriptions for rcu_barrier tracepoint
rcu: Ensure that ->rcu_urgent_qs is set before resched IPI
workqueue: Convert for_each_wq to use built-in list check
rcu: Several rcu_segcblist functions can be static
rcu: Remove unused function hlist_bl_del_init_rcu()
Documentation: Rename rcu_node_context_switch() to rcu_note_context_switch()
...
Minor conflict in drivers/s390/net/qeth_l2_main.c, kept the lock
from commit c8183f5489 ("s390/qeth: fix potential deadlock on
workqueue flush"), removed the code which was removed by commit
9897d583b0 ("s390/qeth: consolidate some duplicated HW cmd code").
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
ct_policy and mpls_policy are parsed with nla_parse_nested(), which
does NL_VALIDATE_STRICT validation, strict_start_type is not needed
to set as it is actually trying to make some attributes parsed with
NL_VALIDATE_STRICT.
This patch is to remove it, and do the same on rtm_nh_policy which
is parsed by nlmsg_parse().
Suggested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is to allow matching options in erspan.
The options can be described in the form:
VER:INDEX:DIR:HWID/VER:INDEX_MASK:DIR_MASK:HWID_MASK.
When ver is set to 1, index will be applied while dir
and hwid will be ignored, and when ver is set to 2,
dir and hwid will be used while index will be ignored.
Different from geneve, only one option can be set. And
also, geneve options, vxlan options or erspan options
can't be set at the same time.
# ip link add name erspan1 type erspan external
# tc qdisc add dev erspan1 ingress
# tc filter add dev erspan1 protocol ip parent ffff: \
flower \
enc_src_ip 10.0.99.192 \
enc_dst_ip 10.0.99.193 \
enc_key_id 11 \
erspan_opts 1:12:0:0/1:ffff:0:0 \
ip_proto udp \
action mirred egress redirect dev eth0
v1->v2:
- improve some err msgs of extack.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is to allow matching gbp option in vxlan.
The options can be described in the form GBP/GBP_MASK,
where GBP is represented as a 32bit hexadecimal value.
Different from geneve, only one option can be set. And
also, geneve options and vxlan options can't be set at
the same time.
# ip link add name vxlan0 type vxlan dstport 0 external
# tc qdisc add dev vxlan0 ingress
# tc filter add dev vxlan0 protocol ip parent ffff: \
flower \
enc_src_ip 10.0.99.192 \
enc_dst_ip 10.0.99.193 \
enc_key_id 11 \
vxlan_opts 01020304/ffffffff \
ip_proto udp \
action mirred egress redirect dev eth0
v1->v2:
- add .strict_start_type for enc_opts_policy as Jakub noticed.
- use Duplicate instead of Wrong in err msg for extack as Jakub
suggested.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is to allow setting erspan options using the
act_tunnel_key action. Different from geneve options,
only one option can be set. And also, geneve options,
vxlan options or erspan options can't be set at the
same time.
Options are expressed as ver:index:dir:hwid, when ver
is set to 1, index will be applied while dir and hwid
will be ignored, and when ver is set to 2, dir and
hwid will be used while index will be ignored.
# ip link add name erspan1 type erspan external
# tc qdisc add dev eth0 ingress
# tc filter add dev eth0 protocol ip parent ffff: \
flower indev eth0 \
ip_proto udp \
action tunnel_key \
set src_ip 10.0.99.192 \
dst_ip 10.0.99.193 \
dst_port 6081 \
id 11 \
erspan_opts 1:2:0:0 \
action mirred egress redirect dev erspan1
v1->v2:
- do the validation when dst is not yet allocated as Jakub suggested.
- use Duplicate instead of Wrong in err msg for extack.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is to allow setting vxlan options using the
act_tunnel_key action. Different from geneve options,
only one option can be set. And also, geneve options
and vxlan options can't be set at the same time.
gbp is the only param for vxlan options:
# ip link add name vxlan0 type vxlan dstport 0 external
# tc qdisc add dev eth0 ingress
# tc filter add dev eth0 protocol ip parent ffff: \
flower indev eth0 \
ip_proto udp \
action tunnel_key \
set src_ip 10.0.99.192 \
dst_ip 10.0.99.193 \
dst_port 6081 \
id 11 \
vxlan_opts 01020304 \
action mirred egress redirect dev vxlan0
v1->v2:
- add .strict_start_type for enc_opts_policy as Jakub noticed.
- use Duplicate instead of Wrong in err msg for extack as Jakub
suggested.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RFC 8033 suggests an alternative approach to calculate the queue
delay in PIE by using a timestamp on every enqueued packet. This
patch adds an implementation of that approach and sets it as the
default method to calculate queue delay. The previous method (based
on Little's law) to calculate queue delay is set as optional.
Signed-off-by: Gautam Ramakrishnan <gautamramk@gmail.com>
Signed-off-by: Leslie Monis <lesliemonis@gmail.com>
Signed-off-by: Mohit P. Tahiliani <tahiliani@nitk.edu.in>
Acked-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The taprio qdisc allows to set mqprio setting but only once. In case
if mqprio settings are provided next time the error is returned as
it's not allowed to change traffic class mapping in-flignt and that
is normal. But if configuration is absolutely the same - no need to
return error. It allows to provide same command couple times,
changing only base time for instance, or changing only scheds maps,
but leaving mqprio setting w/o modification. It more corresponds the
message: "Changing the traffic mapping of a running schedule is not
supported", so reject mqprio if it's really changed.
Also corrected TC_BITMASK + 1 for consistency, as proposed.
Fixes: a3d43c0d56 ("taprio: Add support adding an admin schedule")
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Tested-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
info->options_len is 'u8' type, and when opts_len with a value >
IP_TUNNEL_OPTS_MAX, 'info->options_len = opts_len' will cast int
to u8 and set a wrong value to info->options_len.
Kernel crashed in my test when doing:
# opts="0102:80:00800022"
# for i in {1..99}; do opts="$opts,0102:80:00800022"; done
# ip link add name geneve0 type geneve dstport 0 external
# tc qdisc add dev eth0 ingress
# tc filter add dev eth0 protocol ip parent ffff: \
flower indev eth0 ip_proto udp action tunnel_key \
set src_ip 10.0.99.192 dst_ip 10.0.99.193 \
dst_port 6081 id 11 geneve_opts $opts \
action mirred egress redirect dev geneve0
So we should do the similar check as cls_flower does, return error
when opts_len > IP_TUNNEL_OPTS_MAX in tunnel_key_copy_opts().
Fixes: 0ed5269f9e ("net/sched: add tunnel option support to act_tunnel_key")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
after commit 4097e9d250 ("net: sched: don't use tc_action->order during
action dump"), 'act->order' is initialized but then it's no more read, so
we can just remove this member of struct tc_action.
CC: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
One conflict in the BPF samples Makefile, some fixes in 'net' whilst
we were converting over to Makefile.target rules in 'net-next'.
Signed-off-by: David S. Miller <davem@davemloft.net>
sk->sk_max_ack_backlog can be read without any lock being held
at least in TCP/DCCP cases.
We need to use READ_ONCE()/WRITE_ONCE() to avoid load/store tearing
and/or potential KCSAN warnings.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sk->sk_ack_backlog can be read without any lock being held.
We need to use READ_ONCE()/WRITE_ONCE() to avoid load/store tearing
and/or potential KCSAN warnings.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now the kernel uses 64bit packet counters in scheduler layer,
we want to export these counters to user space.
Instead risking breaking user space by adding fields
to struct gnet_stats_basic, add a new TCA_STATS_PKT64.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After this change, qdisc packet counter is no longer
a 32bit quantity. We still export 32bit values to user.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a new filter is added to cls_api, the function
tcf_chain_tp_insert_unique() looks up the protocol/priority/chain to
determine if the tcf_proto is duplicated in the chain's hashtable. It then
creates a new entry or continues with an existing one. In cls_flower, this
allows the function fl_ht_insert_unque to determine if a filter is a
duplicate and reject appropriately, meaning that the duplicate will not be
passed to drivers via the offload hooks. However, when a tcf_proto is
destroyed it is removed from its chain before a hardware remove hook is
hit. This can lead to a race whereby the driver has not received the
remove message but duplicate flows can be accepted. This, in turn, can
lead to the offload driver receiving incorrect duplicate flows and out of
order add/delete messages.
Prevent duplicates by utilising an approach suggested by Vlad Buslov. A
hash table per block stores each unique chain/protocol/prio being
destroyed. This entry is only removed when the full destroy (and hardware
offload) has completed. If a new flow is being added with the same
identiers as a tc_proto being detroyed, then the add request is replayed
until the destroy is complete.
Fixes: 8b64678e0a ("net: sched: refactor tp insert/delete for concurrent execution")
Signed-off-by: John Hurley <john.hurley@netronome.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reported-by: Louis Peens <louis.peens@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The same code which recognizes ICMP error packets is duplicated several
times. Use the icmp_is_err() and icmpv6_is_err() helpers instead, which
do the same thing.
ip_multipath_l3_keys() and tcf_nat_act() didn't check for all the error types,
assume that they should instead.
Signed-off-by: Matteo Croce <mcroce@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Don't swap oper and admin schedules too early, it's not correct and
causes crash.
Steps to reproduce:
1)
tc qdisc replace dev eth0 parent root handle 100 taprio \
num_tc 3 \
map 2 2 1 0 2 2 2 2 2 2 2 2 2 2 2 2 \
queues 1@0 1@1 1@2 \
base-time $SOME_BASE_TIME \
sched-entry S 01 80000 \
sched-entry S 02 15000 \
sched-entry S 04 40000 \
flags 2
2)
tc qdisc replace dev eth0 parent root handle 100 taprio \
base-time $SOME_BASE_TIME \
sched-entry S 01 90000 \
sched-entry S 02 20000 \
sched-entry S 04 40000 \
flags 2
3)
tc qdisc replace dev eth0 parent root handle 100 taprio \
base-time $SOME_BASE_TIME \
sched-entry S 01 150000 \
sched-entry S 02 200000 \
sched-entry S 04 40000 \
flags 2
Do 2 3 2 .. steps more times if not happens and observe:
[ 305.832319] Unable to handle kernel write to read-only memory at
virtual address ffff0000087ce7f0
[ 305.910887] CPU: 0 PID: 0 Comm: swapper/0 Not tainted
[ 305.919306] Hardware name: Texas Instruments AM654 Base Board (DT)
[...]
[ 306.017119] x1 : ffff800848031d88 x0 : ffff800848031d80
[ 306.022422] Call trace:
[ 306.024866] taprio_free_sched_cb+0x4c/0x98
[ 306.029040] rcu_process_callbacks+0x25c/0x410
[ 306.033476] __do_softirq+0x10c/0x208
[ 306.037132] irq_exit+0xb8/0xc8
[ 306.040267] __handle_domain_irq+0x64/0xb8
[ 306.044352] gic_handle_irq+0x7c/0x178
[ 306.048092] el1_irq+0xb0/0x128
[ 306.051227] arch_cpu_idle+0x10/0x18
[ 306.054795] do_idle+0x120/0x138
[ 306.058015] cpu_startup_entry+0x20/0x28
[ 306.061931] rest_init+0xcc/0xd8
[ 306.065154] start_kernel+0x3bc/0x3e4
[ 306.068810] Code: f2fbd5b7 f2fbd5b6 d503201f f9400422 (f9000662)
[ 306.074900] ---[ end trace 96c8e2284a9d9d6e ]---
[ 306.079507] Kernel panic - not syncing: Fatal exception in interrupt
[ 306.085847] SMP: stopping secondary CPUs
[ 306.089765] Kernel Offset: disabled
Try to explain one of the possible crash cases:
The "real" admin list is assigned when admin_sched is set to
new_admin, it happens after "swap", that assigns to oper_sched NULL.
Thus if call qdisc show it can crash.
Farther, next second time, when sched list is updated, the admin_sched
is not NULL and becomes the oper_sched, previous oper_sched was NULL so
just skipped. But then admin_sched is assigned new_admin, but schedules
to free previous assigned admin_sched (that already became oper_sched).
Farther, next third time, when sched list is updated,
while one more swap, oper_sched is not null, but it was happy to be
freed already (while prev. admin update), so while try to free
oper_sched the kernel panic happens at taprio_free_sched_cb().
So, move the "swap emulation" where it should be according to function
comment from code.
Fixes: 9c66d15646 ("taprio: Add support for hardware offloading")
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Tested-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The only slightly tricky merge conflict was the netdevsim because the
mutex locking fix overlapped a lot of driver reload reorganization.
The rest were (relatively) trivial in nature.
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 4011921137 ("net: sched: refactor block offloads counter
usage") missed the fact that either new prog or old prog may be
NULL.
Fixes: 4011921137 ("net: sched: refactor block offloads counter usage")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull RCU and LKMM changes from Paul E. McKenney:
- Documentation updates.
- Miscellaneous fixes.
- Dynamic tick (nohz) updates, perhaps most notably changes to
force the tick on when needed due to lengthy in-kernel execution
on CPUs on which RCU is waiting.
- Replace rcu_swap_protected() with rcu_prepace_pointer().
- Torture-test updates.
- Linux-kernel memory consistency model updates.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Extend struct tc_action with new "tcfa_flags" field. Set the field in
tcf_idr_create() function and provide new helper
tcf_idr_create_from_flags() that derives 'cpustats' boolean from flags
value. Update individual hardware-offloaded actions init() to pass their
"flags" argument to new helper in order to skip percpu stats allocation
when user requested it through flags.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend TCA_ACT space with nla_bitfield32 flags. Add
TCA_ACT_FLAGS_NO_PERCPU_STATS as the only allowed flag. Parse the flags in
tcf_action_init_1() and pass resulting value as additional argument to
a_o->init().
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Modify stats update helper functions introduced in previous patches in this
series to fallback to regular tc_action->tcfa_{b|q}stats if cpu stats are
not allocated for the action argument. If regular non-percpu allocated
counters are in use, then obtain action tcfa_lock while modifying them.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Previous commit introduced helper function for updating qstats and
refactored set of actions to use the helpers, instead of modifying qstats
directly. However, one of the affected action exposes its qstats to
skb_tc_reinsert(), which then modifies it.
Refactor skb_tc_reinsert() to return integer error code and don't increment
overlimit qstats in case of error, and use the returned error code in
tcf_mirred_act() to manually increment the overlimit counter with new
helper function.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extract common code that increments cpu_qstats counters into standalone act
API functions. Change hardware offloaded actions that use percpu counter
allocation to use the new functions instead of accessing cpu_qstats
directly.
This commit doesn't change functionality.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extract common code that increments cpu_bstats counter into standalone act
API function. Change hardware offloaded actions that use percpu counter
allocation to use the new function instead of incrementing cpu_bstats
directly.
This commit doesn't change functionality.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, all implementations of tc_action_ops->stats_update() callback
have almost exactly the same implementation of counters update
code (besides gact which also updates drop counter). In order to simplify
support for using both percpu-allocated and regular action counters
depending on run-time flag in following patches, extract action counters
update code into standalone function in act API.
This commit doesn't change functionality.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit replaces the use of rcu_swap_protected() with the more
intuitively appealing rcu_replace_pointer() as a step towards removing
rcu_swap_protected().
Link: https://lore.kernel.org/lkml/CAHk-=wiAsJLw1egFEE=Z7-GGtM6wcvtyytXZA1+BHqta4gg6Hw@mail.gmail.com/
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
[ paulmck: From rcu_replace() to rcu_replace_pointer() per Ingo Molnar. ]
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: <netdev@vger.kernel.org>
Cc: <linux-kernel@vger.kernel.org>
There is networking hardware that isn't based on Ethernet for layers 1 and 2.
For example CAN.
CAN is a multi-master serial bus standard for connecting Electronic Control
Units [ECUs] also known as nodes. A frame on the CAN bus carries up to 8 bytes
of payload. Frame corruption is detected by a CRC. However frame loss due to
corruption is possible, but a quite unusual phenomenon.
While fq_codel works great for TCP/IP, it doesn't for CAN. There are a lot of
legacy protocols on top of CAN, which are not build with flow control or high
CAN frame drop rates in mind.
When using fq_codel, as soon as the queue reaches a certain delay based length,
skbs from the head of the queue are silently dropped. Silently meaning that the
user space using a send() or similar syscall doesn't get an error. However
TCP's flow control algorithm will detect dropped packages and adjust the
bandwidth accordingly.
When using fq_codel and sending raw frames over CAN, which is the common use
case, the user space thinks the package has been sent without problems, because
send() returned without an error. pfifo_fast will drop skbs, if the queue
length exceeds the maximum. But with this scheduler the skbs at the tail are
dropped, an error (-ENOBUFS) is propagated to user space. So that the user
space can slow down the package generation.
On distributions, where fq_codel is made default via CONFIG_DEFAULT_NET_SCH
during compile time, or set default during runtime with sysctl
net.core.default_qdisc (see [1]), we get a bad user experience. In my test case
with pfifo_fast, I can transfer thousands of million CAN frames without a frame
drop. On the other hand with fq_codel there is more then one lost CAN frame per
thousand frames.
As pointed out fq_codel is not suited for CAN hardware, so this patch changes
attach_one_default_qdisc() to use pfifo_fast for "ARPHRD_CAN" network devices.
During transition of a netdev from down to up state the default queuing
discipline is attached by attach_default_qdiscs() with the help of
attach_one_default_qdisc(). This patch modifies attach_one_default_qdisc() to
attach the pfifo_fast (pfifo_fast_ops) if the network device type is
"ARPHRD_CAN".
[1] https://github.com/systemd/systemd/issues/9194
Suggested-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Vincent Prince <vincent.prince.fr@gmail.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is networking hardware that isn't based on Ethernet for layers 1 and 2.
For example CAN.
CAN is a multi-master serial bus standard for connecting Electronic Control
Units [ECUs] also known as nodes. A frame on the CAN bus carries up to 8 bytes
of payload. Frame corruption is detected by a CRC. However frame loss due to
corruption is possible, but a quite unusual phenomenon.
While fq_codel works great for TCP/IP, it doesn't for CAN. There are a lot of
legacy protocols on top of CAN, which are not build with flow control or high
CAN frame drop rates in mind.
When using fq_codel, as soon as the queue reaches a certain delay based length,
skbs from the head of the queue are silently dropped. Silently meaning that the
user space using a send() or similar syscall doesn't get an error. However
TCP's flow control algorithm will detect dropped packages and adjust the
bandwidth accordingly.
When using fq_codel and sending raw frames over CAN, which is the common use
case, the user space thinks the package has been sent without problems, because
send() returned without an error. pfifo_fast will drop skbs, if the queue
length exceeds the maximum. But with this scheduler the skbs at the tail are
dropped, an error (-ENOBUFS) is propagated to user space. So that the user
space can slow down the package generation.
On distributions, where fq_codel is made default via CONFIG_DEFAULT_NET_SCH
during compile time, or set default during runtime with sysctl
net.core.default_qdisc (see [1]), we get a bad user experience. In my test case
with pfifo_fast, I can transfer thousands of million CAN frames without a frame
drop. On the other hand with fq_codel there is more then one lost CAN frame per
thousand frames.
As pointed out fq_codel is not suited for CAN hardware, so this patch changes
attach_one_default_qdisc() to use pfifo_fast for "ARPHRD_CAN" network devices.
During transition of a netdev from down to up state the default queuing
discipline is attached by attach_default_qdiscs() with the help of
attach_one_default_qdisc(). This patch modifies attach_one_default_qdisc() to
attach the pfifo_fast (pfifo_fast_ops) if the network device type is
"ARPHRD_CAN".
[1] https://github.com/systemd/systemd/issues/9194
Signed-off-by: Vincent Prince <vincent.prince.fr@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some interface types could be nested.
(VLAN, BONDING, TEAM, MACSEC, MACVLAN, IPVLAN, VIRT_WIFI, VXLAN, etc..)
These interface types should set lockdep class because, without lockdep
class key, lockdep always warn about unexisting circular locking.
In the current code, these interfaces have their own lockdep class keys and
these manage itself. So that there are so many duplicate code around the
/driver/net and /net/.
This patch adds new generic lockdep keys and some helper functions for it.
This patch does below changes.
a) Add lockdep class keys in struct net_device
- qdisc_running, xmit, addr_list, qdisc_busylock
- these keys are used as dynamic lockdep key.
b) When net_device is being allocated, lockdep keys are registered.
- alloc_netdev_mqs()
c) When net_device is being free'd llockdep keys are unregistered.
- free_netdev()
d) Add generic lockdep key helper function
- netdev_register_lockdep_key()
- netdev_unregister_lockdep_key()
- netdev_update_lockdep_key()
e) Remove unnecessary generic lockdep macro and functions
f) Remove unnecessary lockdep code of each interfaces.
After this patch, each interface modules don't need to maintain
their lockdep keys.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
UDP IPv6 packets auto flowlabels are using a 32bit secret
(static u32 hashrnd in net/core/flow_dissector.c) and
apply jhash() over fields known by the receivers.
Attackers can easily infer the 32bit secret and use this information
to identify a device and/or user, since this 32bit secret is only
set at boot time.
Really, using jhash() to generate cookies sent on the wire
is a serious security concern.
Trying to change the rol32(hash, 16) in ip6_make_flowlabel() would be
a dead end. Trying to periodically change the secret (like in sch_sfq.c)
could change paths taken in the network for long lived flows.
Let's switch to siphash, as we did in commit df453700e8
("inet: switch IP ID generator to siphash")
Using a cryptographically strong pseudo random function will solve this
privacy issue and more generally remove other weak points in the stack.
Packet schedulers using skb_get_hash_perturb() benefit from this change.
Fixes: b56774163f ("ipv6: Enable auto flow labels by default")
Fixes: 42240901f7 ("ipv6: Implement different admin modes for automatic flow labels")
Fixes: 67800f9b1f ("ipv6: Call skb_get_hash_flowi6 to get skb->hash in ip6_make_flowlabel")
Fixes: cb1ce2ef38 ("ipv6: Implement automatic flow label generation on transmit")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Jonathan Berger <jonathann1@walla.com>
Reported-by: Amit Klein <aksecurity@gmail.com>
Reported-by: Benny Pinkas <benny@pinkas.net>
Cc: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit 342db22182 ("sched: Call skb_get_hash_perturb
in sch_fq_codel") we no longer need anything from this file.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
We get one warnings when build kernel W=1:
net/sched/sch_taprio.c:1155:6: warning: no previous prototype for ‘taprio_offload_config_changed’ [-Wmissing-prototypes]
Make the function static to fix this.
Fixes: 9c66d15646 ("taprio: Add support for hardware offloading")
Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Use tcf_tm_dump(), instead of an open coded variant (no functional change
in this patch).
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If packet corruption failed we jump to finish_segs and return
NET_XMIT_SUCCESS. Seeing success will make the parent qdisc
increment its backlog, that's incorrect - we need to return
NET_XMIT_DROP.
Fixes: 6071bd1aa1 ("netem: Segment GSO packets on enqueue")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To corrupt a GSO frame we first perform segmentation. We then
proceed using the first segment instead of the full GSO skb and
requeue the rest of the segments as separate packets.
If there are any issues with processing the first segment we
still want to process the rest, therefore we jump to the
finish_segs label.
Commit 177b800746 ("net: netem: fix backlog accounting for
corrupted GSO frames") started using the pointer to the first
segment in the "rest of segments processing", but as mentioned
above the first segment may had already been freed at this point.
Backlog corrections for parent qdiscs have to be adjusted.
Fixes: 177b800746 ("net: netem: fix backlog accounting for corrupted GSO frames")
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reported-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With threaded interrupts enabled, the interrupt thread runs as SCHED_RR
with priority 50. If a user application with a higher priority preempts
the interrupt thread and tries to shutdown the network interface then it
will loop forever. The kernel will spin in the loop waiting for the
device to become idle and the scheduler will never consider the
interrupt thread because its priority is lower.
Avoid the problem by sleeping for a jiffy giving other tasks,
including the interrupt thread, a chance to run and make progress.
In the original thread it has been suggested to use wait_event() and
properly waiting for the state to occur. DaveM explained that this would
require to add expensive checks in the fast paths of packet processing.
Link: https://lkml.kernel.org/r/1393976987-23555-1-git-send-email-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
[bigeasy: Rewrite commit message, add comment, use
schedule_timeout_uninterruptible()]
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a application sends many packets with the same txtime, they may
be transmitted out of order (different from the order in which they
were enqueued).
This happens because when inserting elements into the tree, when the
txtime of two packets are the same, the new packet is inserted at the
left side of the tree, causing the reordering. The only effect of this
change should be that packets with the same txtime will be transmitted
in the order they are enqueued.
The application in question (the AVTP GStreamer plugin, still in
development) is sending video traffic, in which each video frame have
a single presentation time, the problem is that when packetizing,
multiple packets end up with the same txtime.
The receiving side was rejecting packets because they were being
received out of order.
Fixes: 25db26a913 ("net/sched: Introduce the ETF Qdisc")
Reported-by: Ederson de Souza <ederson.desouza@intel.com>
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After commit eeb84aa0d0 ("net_sched: sch_fq: do not assume EDT
packets are ordered"), all skbs get a non zero time_to_send
in flow_queue_add()
This means @time_next_packet variable in fq_dequeue()
can no longer be zero.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
the following script:
# tc qdisc add dev eth0 clsact
# tc filter add dev eth0 egress protocol ip matchall \
> action mpls push protocol mpls_uc label 0x355aa bos 1
causes corruption of all IP packets transmitted by eth0. On TC egress, we
can't rely on the value of skb->mac_len, because it's 0 and a MPLS 'push'
operation will result in an overwrite of the first 4 octets in the packet
L2 header (e.g. the Destination Address if eth0 is an Ethernet); the same
error pattern is present also in the MPLS 'pop' operation. Fix this error
in act_mpls data plane, computing 'mac_len' as the difference between the
network header and the mac header (when not at TC ingress), and use it in
MPLS 'push'/'pop' core functions.
v2: unbreak 'make htmldocs' because of missing documentation of 'mac_len'
in skb_mpls_pop(), reported by kbuild test robot
CC: Lorenzo Bianconi <lorenzo@kernel.org>
Fixes: 2a2ea50870 ("net: sched: add mpls manipulation actions to TC")
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Acked-by: John Hurley <john.hurley@netronome.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the sake of tcp_poll(), there are few places where we fetch
sk->sk_wmem_queued while this field can change from IRQ or other cpu.
We need to add READ_ONCE() annotations, and also make sure write
sides use corresponding WRITE_ONCE() to avoid store-tearing.
sk_wmem_queued_add() helper is added so that we can in
the future convert to ADD_ONCE() or equivalent if/when
available.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sock_rcvlowat() or int_sk_rcvlowat() might be called without the socket
lock for example from tcp_poll().
Use READ_ONCE() to document the fact that other cpus might change
sk->sk_rcvlowat under us and avoid KCSAN splats.
Use WRITE_ONCE() on write sides too.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
If tcf_register_action failed, mirred_device_notifier
should be unregistered.
Fixes: 3b87956ea6 ("net sched: fix race in mirred device removal")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
When configuring a taprio instance if "flags" is not specified (or
it's zero), taprio currently replies with an "Invalid argument" error.
So, set the return value to zero after we are done with all the
checks.
Fixes: 9c66d15646 ("taprio: Add support for hardware offloading")
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Acked-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
For TCA_ACT_KIND, we have to keep the backward compatibility too,
and rely on nla_strlcpy() to check and terminate the string with
a NUL.
Note for TC actions, nla_strcmp() is already used to compare kind
strings, so we don't need to fix other places.
Fixes: 199ce850ce ("net_sched: add policy validation for action attributes")
Reported-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Marcelo noticed a backward compatibility issue of TCA_KIND
after we move from NLA_STRING to NLA_NUL_STRING, so it is probably
too late to change it.
Instead, to make everyone happy, we can just insert a NUL to
terminate the string with nla_strlcpy() like we do for TC actions.
Fixes: 62794fc4fb ("net_sched: add max len check for TCA_KIND")
Reported-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
The introduction of this schedule point was done in commit
2ba2506ca7 ("[NET]: Add preemption point in qdisc_run")
at a time the loop was not bounded.
Then later in commit d5b8aa1d24 ("net_sched: fix dequeuer fairness")
we added a limit on the number of packets.
Now is the time to remove the schedule point, since the default
limit of 64 packets matches the number of packets a typical NAPI
poll can process in a row.
This solves a latency problem for most TCP receivers under moderate load :
1) host receives a packet.
NET_RX_SOFTIRQ is raised by NIC hard IRQ handler
2) __do_softirq() does its first loop, handling NET_RX_SOFTIRQ
and calling the driver napi->loop() function
3) TCP stores the skb in socket receive queue:
4) TCP calls sk->sk_data_ready() and wakeups a user thread
waiting for EPOLLIN (as a result, need_resched() might now be true)
5) TCP cooks an ACK and sends it.
6) qdisc_run() processes one packet from qdisc, and sees need_resched(),
this raises NET_TX_SOFTIRQ (even if there are no more packets in
the qdisc)
Then we go back to the __do_softirq() in 2), and we see that new
softirqs were raised. Since need_resched() is true, we end up waking
ksoftirqd in this path :
if (pending) {
if (time_before(jiffies, end) && !need_resched() &&
--max_restart)
goto restart;
wakeup_softirqd();
}
So we have many wakeups of ksoftirqd kernel threads,
and more calls to qdisc_run() with associated lock overhead.
Note that another way to solve the issue would be to change TCP
to first send the ACK packet, then signal the EPOLLIN,
but this changes P99 latencies, as sending the ACK packet
can add a long delay.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As explained in the "net: sched: taprio: Avoid division by zero on
invalid link speed" commit, it is legal for the ethtool API to return
zero as a link speed. So guard against it to ensure we don't perform a
division by zero in kernel.
Fixes: e0a7683d30 ("net/sched: cbs: fix port_rate miscalculation")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The check in taprio_set_picos_per_byte is currently not robust enough
and will trigger this division by zero, due to e.g. PHYLINK not setting
kset->base.speed when there is no PHY connected:
[ 27.109992] Division by zero in kernel.
[ 27.113842] CPU: 1 PID: 198 Comm: tc Not tainted 5.3.0-rc5-01246-gc4006b8c2637-dirty #212
[ 27.121974] Hardware name: Freescale LS1021A
[ 27.126234] [<c03132e0>] (unwind_backtrace) from [<c030d8b8>] (show_stack+0x10/0x14)
[ 27.133938] [<c030d8b8>] (show_stack) from [<c10b21b0>] (dump_stack+0xb0/0xc4)
[ 27.141124] [<c10b21b0>] (dump_stack) from [<c10af97c>] (Ldiv0_64+0x8/0x18)
[ 27.148052] [<c10af97c>] (Ldiv0_64) from [<c0700260>] (div64_u64+0xcc/0xf0)
[ 27.154978] [<c0700260>] (div64_u64) from [<c07002d0>] (div64_s64+0x4c/0x68)
[ 27.161993] [<c07002d0>] (div64_s64) from [<c0f3d890>] (taprio_set_picos_per_byte+0xe8/0xf4)
[ 27.170388] [<c0f3d890>] (taprio_set_picos_per_byte) from [<c0f3f614>] (taprio_change+0x668/0xcec)
[ 27.179302] [<c0f3f614>] (taprio_change) from [<c0f2bc24>] (qdisc_create+0x1fc/0x4f4)
[ 27.187091] [<c0f2bc24>] (qdisc_create) from [<c0f2c0c8>] (tc_modify_qdisc+0x1ac/0x6f8)
[ 27.195055] [<c0f2c0c8>] (tc_modify_qdisc) from [<c0ee9604>] (rtnetlink_rcv_msg+0x268/0x2dc)
[ 27.203449] [<c0ee9604>] (rtnetlink_rcv_msg) from [<c0f4fef0>] (netlink_rcv_skb+0xe0/0x114)
[ 27.211756] [<c0f4fef0>] (netlink_rcv_skb) from [<c0f4f6cc>] (netlink_unicast+0x1b4/0x22c)
[ 27.219977] [<c0f4f6cc>] (netlink_unicast) from [<c0f4fa84>] (netlink_sendmsg+0x284/0x340)
[ 27.228198] [<c0f4fa84>] (netlink_sendmsg) from [<c0eae5fc>] (sock_sendmsg+0x14/0x24)
[ 27.235988] [<c0eae5fc>] (sock_sendmsg) from [<c0eaedf8>] (___sys_sendmsg+0x214/0x228)
[ 27.243863] [<c0eaedf8>] (___sys_sendmsg) from [<c0eb015c>] (__sys_sendmsg+0x50/0x8c)
[ 27.251652] [<c0eb015c>] (__sys_sendmsg) from [<c0301000>] (ret_fast_syscall+0x0/0x54)
[ 27.259524] Exception stack(0xe8045fa8 to 0xe8045ff0)
[ 27.264546] 5fa0: b6f608c8 000000f8 00000003 bed7e2f0 00000000 00000000
[ 27.272681] 5fc0: b6f608c8 000000f8 004ce54c 00000128 5d3ce8c7 00000000 00000026 00505c9c
[ 27.280812] 5fe0: 00000070 bed7e298 004ddd64 b6dd1e64
Russell King points out that the ethtool API says zero is a valid return
value of __ethtool_get_link_ksettings:
* If it is enabled then they are read-only; if the link
* is up they represent the negotiated link mode; if the link is down,
* the speed is 0, %SPEED_UNKNOWN or the highest enabled speed and
* @duplex is %DUPLEX_UNKNOWN or the best enabled duplex mode.
So, it seems that taprio is not following the API... I'd suggest either
fixing taprio, or getting agreement to change the ethtool API.
The chosen path was to fix taprio.
Fixes: 7b9eba7ba0 ("net/sched: taprio: fix picos_per_byte miscalculation")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The speed divisor is used in a context expecting an s64, but it is
evaluated using 32-bit arithmetic.
To avoid that happening, instead of multiplying by 1,000,000 in the
first place, simplify the fraction and do a standard 32 bit division
instead.
Fixes: f04b514c0c ("taprio: Set default link speed to 10 Mbps in taprio_set_picos_per_byte")
Reported-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This a new feature, it is preferred that it defaults to N.
We will probe the feature support from userspace before actually using it.
Fixes: 95a7233c45 ('net: openvswitch: Set OvS recirc_id from tc chain index')
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Recent changes that removed rtnl dependency from rules update path of tc
also made tcf_block_put() function sleeping. This function is called from
ops->destroy() of several Qdisc implementations, which in turn is called by
qdisc_put(). Some Qdiscs call qdisc_put() while holding sch tree spinlock,
which results sleeping-while-atomic BUG.
Steps to reproduce for multiq:
tc qdisc add dev ens1f0 root handle 1: multiq
tc qdisc add dev ens1f0 parent 1:10 handle 50: sfq perturb 10
ethtool -L ens1f0 combined 2
tc qdisc change dev ens1f0 root handle 1: multiq
Resulting dmesg:
[ 5539.419344] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:909
[ 5539.420945] in_atomic(): 1, irqs_disabled(): 0, pid: 27658, name: tc
[ 5539.422435] INFO: lockdep is turned off.
[ 5539.423904] CPU: 21 PID: 27658 Comm: tc Tainted: G W 5.3.0-rc8+ #721
[ 5539.425400] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017
[ 5539.426911] Call Trace:
[ 5539.428380] dump_stack+0x85/0xc0
[ 5539.429823] ___might_sleep.cold+0xac/0xbc
[ 5539.431262] __mutex_lock+0x5b/0x960
[ 5539.432682] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0
[ 5539.434103] ? __nla_validate_parse+0x51/0x840
[ 5539.435493] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0
[ 5539.436903] tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0
[ 5539.438327] tcf_block_put_ext.part.0+0x21/0x50
[ 5539.439752] tcf_block_put+0x50/0x70
[ 5539.441165] sfq_destroy+0x15/0x50 [sch_sfq]
[ 5539.442570] qdisc_destroy+0x5f/0x160
[ 5539.444000] multiq_tune+0x14a/0x420 [sch_multiq]
[ 5539.445421] tc_modify_qdisc+0x324/0x840
[ 5539.446841] rtnetlink_rcv_msg+0x170/0x4b0
[ 5539.448269] ? netlink_deliver_tap+0x95/0x400
[ 5539.449691] ? rtnl_dellink+0x2d0/0x2d0
[ 5539.451116] netlink_rcv_skb+0x49/0x110
[ 5539.452522] netlink_unicast+0x171/0x200
[ 5539.453914] netlink_sendmsg+0x224/0x3f0
[ 5539.455304] sock_sendmsg+0x5e/0x60
[ 5539.456686] ___sys_sendmsg+0x2ae/0x330
[ 5539.458071] ? ___sys_recvmsg+0x159/0x1f0
[ 5539.459461] ? do_wp_page+0x9c/0x790
[ 5539.460846] ? __handle_mm_fault+0xcd3/0x19e0
[ 5539.462263] __sys_sendmsg+0x59/0xa0
[ 5539.463661] do_syscall_64+0x5c/0xb0
[ 5539.465044] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 5539.466454] RIP: 0033:0x7f1fe08177b8
[ 5539.467863] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83 ec 28 89 5
4
[ 5539.470906] RSP: 002b:00007ffe812de5d8 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[ 5539.472483] RAX: ffffffffffffffda RBX: 000000005d8135e3 RCX: 00007f1fe08177b8
[ 5539.474069] RDX: 0000000000000000 RSI: 00007ffe812de640 RDI: 0000000000000003
[ 5539.475655] RBP: 0000000000000000 R08: 0000000000000001 R09: 000000000182e9b0
[ 5539.477203] R10: 0000000000404eda R11: 0000000000000246 R12: 0000000000000001
[ 5539.478699] R13: 000000000047f640 R14: 0000000000000000 R15: 0000000000000000
Rearrange locking in multiq_tune() in following ways:
- In loop that removes Qdiscs from disabled queues, call
qdisc_purge_queue() instead of qdisc_tree_flush_backlog() on Qdisc that
is being destroyed. Save the Qdisc in temporary allocated array and call
qdisc_put() on each element of the array after sch tree lock is released.
This is safe to do because Qdiscs have already been reset by
qdisc_purge_queue() inside sch tree lock critical section.
- Do the same change for second loop that initializes Qdiscs for newly
enabled queues in multiq_tune() function. Since sch tree lock is obtained
and released on each iteration of this loop, just call qdisc_put()
directly outside of critical section. Don't verify that old Qdisc is not
noop_qdisc before releasing reference to it because such check is already
performed by qdisc_put*() functions.
Fixes: c266f64dbf ("net: sched: protect block state with mutex")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adjust indentation from spaces to tab (+optional two spaces) as in
coding style with command like:
$ sed -e 's/^ /\t/' -i */Kconfig
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Similar to commit 8b4c3cdd9d
("net: sched: Add policy validation for tc attributes"), we need
to add proper policy validation for TC action attributes too.
Cc: David Ahern <dsahern@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
The TCA_KIND attribute is of NLA_STRING which does not check
the NUL char. KMSAN reported an uninit-value of TCA_KIND which
is likely caused by the lack of NUL.
Change it to NLA_NUL_STRING and add a max len too.
Fixes: 8b4c3cdd9d ("net: sched: Add policy validation for tc attributes")
Reported-and-tested-by: syzbot+618aacd49e8c8b8486bd@syzkaller.appspotmail.com
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
syzbot managed to crash the kernel in tabledist() loading
an empty distribution table.
t = dist->table[rnd % dist->size];
Simply return an error when such load is attempted.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
current 'sample' action doesn't push the mac header of ingress packets if
they are received by a layer 3 tunnel (like gre or sit); but it forgot to
check for gre over ipv6, so the following script:
# tc q a dev $d clsact
# tc f a dev $d ingress protocol ip flower ip_proto icmp action sample \
> group 100 rate 1
# psample -v -g 100
dumps everything, including outer header and mac, when $d is a gre tunnel
over ipv6. Fix this adding a missing label for ARPHRD_IP6GRE devices.
Fixes: 5c5670fae4 ("net/sched: Introduce sample tc action")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Yotam Gigi <yotam.gi@gmail.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
This allows taprio to offload the schedule enforcement to capable
network cards, resulting in more precise windows and less CPU usage.
The gate mask acts on traffic classes (groups of queues of same
priority), as specified in IEEE 802.1Q-2018, and following the existing
taprio and mqprio semantics.
It is up to the driver to perform conversion between tc and individual
netdev queues if for some reason it needs to make that distinction.
Full offload is requested from the network interface by specifying
"flags 2" in the tc qdisc creation command, which in turn corresponds to
the TCA_TAPRIO_ATTR_FLAG_FULL_OFFLOAD bit.
The important detail here is the clockid which is implicitly /dev/ptpN
for full offload, and hence not configurable.
A reference counting API is added to support the use case where Ethernet
drivers need to keep the taprio offload structure locally (i.e. they are
a multi-port switch driver, and configuring a port depends on the
settings of other ports as well). The refcount_t variable is kept in a
private structure (__tc_taprio_qopt_offload) and not exposed to drivers.
In the future, the private structure might also be expanded with a
backpointer to taprio_sched *q, to implement the notification system
described in the patch (of when admin became oper, or an error occurred,
etc, so the offload can be monitored with 'tc qdisc show').
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Voon Weifeng <weifeng.voon@intel.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When filling in hardware intermediate representation tc_setup_flow_action()
directly obtains, checks and takes reference to dev used by mirred action,
instead of using act->ops->get_dev() API created specifically for this
purpose. In order to remove code duplication, refactor flow_action infra to
use action API when obtaining mirred action target dev. Extend get_dev()
with additional argument that is used to provide dev destructor to the
user.
Fixes: 5a6ff4b13d ("net: sched: take reference to action dev before calling offloads")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With recent patch set that removed rtnl lock dependency from cls hardware
offload API rtnl lock is only taken when reading action data and can be
released after action-specific data is parsed into intermediate
representation. However, sample action psample group is passed by pointer
without obtaining reference to it first, which makes it possible to
concurrently overwrite the action and deallocate object pointed by
psample_group pointer after rtnl lock is released but before driver
finished using the pointer.
To prevent such race condition, obtain reference to psample group while it
is used by flow_action infra. Extend psample API with function
psample_group_take() that increments psample group reference counter.
Extend struct tc_action_ops with new get_psample_group() API. Implement the
API for action sample using psample_group_take() and already existing
psample_group_put() as a destructor. Use it in tc_setup_flow_action() to
take reference to psample group pointed to by entry->sample.psample_group
and release it in tc_cleanup_flow_action().
Disable bh when taking psample_groups_lock. The lock is now taken while
holding action tcf_lock that is used by data path and requires bh to be
disabled, so doing the same for psample_groups_lock is necessary to
preserve SOFTIRQ-irq-safety.
Fixes: 918190f50e ("net: sched: flower: don't take rtnl lock for cls hw offloads API")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Generalize flow_action_entry cleanup by extending the structure with
pointer to destructor function. Set the destructor in
tc_setup_flow_action(). Refactor tc_cleanup_flow_action() to call
entry->destructor() instead of using switch that dispatches by entry->id
and manually executes cleanup.
This refactoring is necessary for following patches in this series that
require destructor to use tc_action->ops callbacks that can't be easily
obtained in tc_cleanup_flow_action().
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When tcf_block_get() fails in sfb_init(), q->qdisc is still a NULL
pointer which leads to a crash in sfb_destroy(). Similar for
sch_dsmark.
Instead of fixing each separately, Linus suggested to just accept
NULL pointer in qdisc_put(), which would make callers easier.
(For sch_dsmark, the bug probably exists long before commit
6529eaba33f0.)
Fixes: 6529eaba33 ("net: sched: introduce tcf block infractructure")
Reported-by: syzbot+d5870a903591faaca4ae@syzkaller.appspotmail.com
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Include some headers in files which require them, and remove others
which are not required.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In case of TCA_HHF_NON_HH_WEIGHT or TCA_HHF_QUANTUM is zero,
it would make no progress inside the loop in hhf_dequeue() thus
kernel would get stuck.
Fix this by checking this corner case in hhf_change().
Fixes: 10239edf86 ("net-qdisc-hhf: Heavy-Hitter Filter (HHF) qdisc")
Reported-by: syzbot+bc6297c11f19ee807dc2@syzkaller.appspotmail.com
Reported-by: syzbot+041483004a7f45f1f20a@syzkaller.appspotmail.com
Reported-by: syzbot+55be5f513bed37fc4367@syzkaller.appspotmail.com
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: Terry Lam <vtlam@google.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
At least sch_red and sch_tbf don't implement ->tcf_block()
while still have a non-zero tc "class".
Instead of adding nop implementations to each of such qdisc's,
we can just relax the check of cops->tcf_block() in
tc_bind_tclass(). They don't support TC filter anyway.
Reported-by: syzbot+21b29db13c065852f64b@syzkaller.appspotmail.com
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Whenever MQ is not used on a multiqueue device, we experience
serious reordering problems. Bisection found the cited
commit.
The issue can be described this way :
- A single qdisc hierarchy is shared by all transmit queues.
(eg : tc qdisc replace dev eth0 root fq_codel)
- When/if try_bulk_dequeue_skb_slow() dequeues a packet targetting
a different transmit queue than the one used to build a packet train,
we stop building the current list and save the 'bad' skb (P1) in a
special queue. (bad_txq)
- When dequeue_skb() calls qdisc_dequeue_skb_bad_txq() and finds this
skb (P1), it checks if the associated transmit queues is still in frozen
state. If the queue is still blocked (by BQL or NIC tx ring full),
we leave the skb in bad_txq and return NULL.
- dequeue_skb() calls q->dequeue() to get another packet (P2)
The other packet can target the problematic queue (that we found
in frozen state for the bad_txq packet), but another cpu just ran
TX completion and made room in the txq that is now ready to accept
new packets.
- Packet P2 is sent while P1 is still held in bad_txq, P1 might be sent
at next round. In practice P2 is the lead of a big packet train
(P2,P3,P4 ...) filling the BQL budget and delaying P1 by many packets :/
To solve this problem, we have to block the dequeue process as long
as the first packet in bad_txq can not be sent. Reordering issues
disappear and no side effects have been seen.
Fixes: a53851e2c3 ("net: sched: explicit locking in gso_cpu fallback")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For high speed adapter like Mellanox CX-5 card, it can reach upto
100 Gbits per second bandwidth. Currently htb already supports 64bit rate
in tc utility. However police action rate and peakrate are still limited
to 32bit value (upto 32 Gbits per second). Add 2 new attributes
TCA_POLICE_RATE64 and TCA_POLICE_RATE64 in kernel for 64bit support
so that tc utility can use them for 64bit rate and peakrate value to
break the 32bit limit, and still keep the backward binary compatibility.
Tested-by: David Dai <zdai@linux.vnet.ibm.com>
Signed-off-by: David Dai <zdai@linux.vnet.ibm.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Offloaded OvS datapath rules are translated one to one to tc rules,
for example the following simplified OvS rule:
recirc_id(0),in_port(dev1),eth_type(0x0800),ct_state(-trk) actions:ct(),recirc(2)
Will be translated to the following tc rule:
$ tc filter add dev dev1 ingress \
prio 1 chain 0 proto ip \
flower tcp ct_state -trk \
action ct pipe \
action goto chain 2
Received packets will first travel though tc, and if they aren't stolen
by it, like in the above rule, they will continue to OvS datapath.
Since we already did some actions (action ct in this case) which might
modify the packets, and updated action stats, we would like to continue
the proccessing with the correct recirc_id in OvS (here recirc_id(2))
where we left off.
To support this, introduce a new skb extension for tc, which
will be used for translating tc chain to ovs recirc_id to
handle these miss cases. Last tc chain index will be set
by tc goto chain action and read by OvS datapath.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Variable port_rate is being initialized with a value that is never read
and is being re-assigned a little later on. The assignment is redundant
and hence can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The discussion to be made is absolutely the same as in the case of
previous patch ("taprio: Set default link speed to 10 Mbps in
taprio_set_picos_per_byte"). Nothing is lost when setting a default.
Cc: Leandro Dorileo <leandro.maciel.dorileo@intel.com>
Fixes: e0a7683d30 ("net/sched: cbs: fix port_rate miscalculation")
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The taprio budget needs to be adapted at runtime according to interface
link speed. But that handling is problematic.
For one thing, installing a qdisc on an interface that doesn't have
carrier is not illegal. But taprio prints the following stack trace:
[ 31.851373] ------------[ cut here ]------------
[ 31.856024] WARNING: CPU: 1 PID: 207 at net/sched/sch_taprio.c:481 taprio_dequeue+0x1a8/0x2d4
[ 31.864566] taprio: dequeue() called with unknown picos per byte.
[ 31.864570] Modules linked in:
[ 31.873701] CPU: 1 PID: 207 Comm: tc Not tainted 5.3.0-rc5-01199-g8838fe023cd6 #1689
[ 31.881398] Hardware name: Freescale LS1021A
[ 31.885661] [<c03133a4>] (unwind_backtrace) from [<c030d8cc>] (show_stack+0x10/0x14)
[ 31.893368] [<c030d8cc>] (show_stack) from [<c10ac958>] (dump_stack+0xb4/0xc8)
[ 31.900555] [<c10ac958>] (dump_stack) from [<c0349d04>] (__warn+0xe0/0xf8)
[ 31.907395] [<c0349d04>] (__warn) from [<c0349d64>] (warn_slowpath_fmt+0x48/0x6c)
[ 31.914841] [<c0349d64>] (warn_slowpath_fmt) from [<c0f38db4>] (taprio_dequeue+0x1a8/0x2d4)
[ 31.923150] [<c0f38db4>] (taprio_dequeue) from [<c0f227b0>] (__qdisc_run+0x90/0x61c)
[ 31.930856] [<c0f227b0>] (__qdisc_run) from [<c0ec82ac>] (net_tx_action+0x12c/0x2bc)
[ 31.938560] [<c0ec82ac>] (net_tx_action) from [<c0302298>] (__do_softirq+0x130/0x3c8)
[ 31.946350] [<c0302298>] (__do_softirq) from [<c03502a0>] (irq_exit+0xbc/0xd8)
[ 31.953536] [<c03502a0>] (irq_exit) from [<c03a4808>] (__handle_domain_irq+0x60/0xb4)
[ 31.961328] [<c03a4808>] (__handle_domain_irq) from [<c0754478>] (gic_handle_irq+0x58/0x9c)
[ 31.969638] [<c0754478>] (gic_handle_irq) from [<c0301a8c>] (__irq_svc+0x6c/0x90)
[ 31.977076] Exception stack(0xe8167b20 to 0xe8167b68)
[ 31.982100] 7b20: e9d4bd80 00000cc0 000000cf 00000000 e9d4bd80 c1f38958 00000cc0 c1f38960
[ 31.990234] 7b40: 00000001 000000cf 00000004 e9dc0800 00000000 e8167b70 c0f478ec c0f46d94
[ 31.998363] 7b60: 60070013 ffffffff
[ 32.001833] [<c0301a8c>] (__irq_svc) from [<c0f46d94>] (netlink_trim+0x18/0xd8)
[ 32.009104] [<c0f46d94>] (netlink_trim) from [<c0f478ec>] (netlink_broadcast_filtered+0x34/0x414)
[ 32.017930] [<c0f478ec>] (netlink_broadcast_filtered) from [<c0f47cec>] (netlink_broadcast+0x20/0x28)
[ 32.027102] [<c0f47cec>] (netlink_broadcast) from [<c0eea378>] (rtnetlink_send+0x34/0x88)
[ 32.035238] [<c0eea378>] (rtnetlink_send) from [<c0f25890>] (notify_and_destroy+0x2c/0x44)
[ 32.043461] [<c0f25890>] (notify_and_destroy) from [<c0f25e08>] (qdisc_graft+0x398/0x470)
[ 32.051595] [<c0f25e08>] (qdisc_graft) from [<c0f27a00>] (tc_modify_qdisc+0x3a4/0x724)
[ 32.059470] [<c0f27a00>] (tc_modify_qdisc) from [<c0ee4c84>] (rtnetlink_rcv_msg+0x260/0x2ec)
[ 32.067864] [<c0ee4c84>] (rtnetlink_rcv_msg) from [<c0f4a988>] (netlink_rcv_skb+0xb8/0x110)
[ 32.076172] [<c0f4a988>] (netlink_rcv_skb) from [<c0f4a170>] (netlink_unicast+0x1b4/0x22c)
[ 32.084392] [<c0f4a170>] (netlink_unicast) from [<c0f4a5e4>] (netlink_sendmsg+0x33c/0x380)
[ 32.092614] [<c0f4a5e4>] (netlink_sendmsg) from [<c0ea9f40>] (sock_sendmsg+0x14/0x24)
[ 32.100403] [<c0ea9f40>] (sock_sendmsg) from [<c0eaa780>] (___sys_sendmsg+0x214/0x228)
[ 32.108279] [<c0eaa780>] (___sys_sendmsg) from [<c0eabad0>] (__sys_sendmsg+0x50/0x8c)
[ 32.116068] [<c0eabad0>] (__sys_sendmsg) from [<c0301000>] (ret_fast_syscall+0x0/0x54)
[ 32.123938] Exception stack(0xe8167fa8 to 0xe8167ff0)
[ 32.128960] 7fa0: b6fa68c8 000000f8 00000003 bea142d0 00000000 00000000
[ 32.137093] 7fc0: b6fa68c8 000000f8 0052154c 00000128 5d6468a2 00000000 00000028 00558c9c
[ 32.145224] 7fe0: 00000070 bea14278 00530d64 b6e17e64
[ 32.150659] ---[ end trace 2139c9827c3e5177 ]---
This happens because the qdisc ->dequeue callback gets called. Which
again is not illegal, the qdisc will dequeue even when the interface is
up but doesn't have carrier (and hence SPEED_UNKNOWN), and the frames
will be dropped further down the stack in dev_direct_xmit().
And, at the end of the day, for what? For calculating the initial budget
of an interface which is non-operational at the moment and where frames
will get dropped anyway.
So if we can't figure out the link speed, default to SPEED_10 and move
along. We can also remove the runtime check now.
Cc: Leandro Dorileo <leandro.maciel.dorileo@intel.com>
Fixes: 7b9eba7ba0 ("net/sched: taprio: fix picos_per_byte miscalculation")
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
taprio_init may fail earlier than this line:
list_add(&q->taprio_list, &taprio_list);
i.e. due to the net device not being multi queue.
Attempting to remove q from the global taprio_list when it is not part
of it will result in a kernel panic.
Fix it by matching list_add and list_del better to one another in the
order of operations. This way we can keep the deletion unconditional
and with lower complexity - O(1).
Cc: Leandro Dorileo <leandro.maciel.dorileo@intel.com>
Fixes: 7b9eba7ba0 ("net/sched: taprio: fix picos_per_byte miscalculation")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Recent rtnl lock removal patch changed flow_action infra to require proper
cleanup besides simple memory deallocation. However, matchall classifier
was not updated to call tc_cleanup_flow_action(). Add proper cleanup to
mall_replace_hw_filter() and mall_reoffload().
Fixes: 5a6ff4b13d ("net: sched: take reference to action dev before calling offloads")
Reported-by: Ido Schimmel <idosch@mellanox.com>
Tested-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Implement this callback in order to get the offloaded stats added to the
kernel stats.
Reported-by: Pengfei Liu <pengfeil@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Action sample doesn't properly handle psample_group pointer in overwrite
case. Following issues need to be fixed:
- In tcf_sample_init() function RCU_INIT_POINTER() is used to set
s->psample_group, even though we neither setting the pointer to NULL, nor
preventing concurrent readers from accessing the pointer in some way.
Use rcu_swap_protected() instead to safely reset the pointer.
- Old value of s->psample_group is not released or deallocated in any way,
which results resource leak. Use psample_group_put() on non-NULL value
obtained with rcu_swap_protected().
- The function psample_group_put() that released reference to struct
psample_group pointed by rcu-pointer s->psample_group doesn't respect rcu
grace period when deallocating it. Extend struct psample_group with rcu
head and use kfree_rcu when freeing it.
Fixes: 5c5670fae4 ("net/sched: Introduce sample tc action")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that 'TCQ_F_CPUSTATS' bit can be cleared, depending on the value of
'TCQ_F_NOLOCK' bit in the parent qdisc, we need to be sure that per-cpu
counters are present when 'reset()' is called for pfifo_fast qdiscs.
Otherwise, the following script:
# tc q a dev lo handle 1: root htb default 100
# tc c a dev lo parent 1: classid 1:100 htb \
> rate 95Mbit ceil 100Mbit burst 64k
[...]
# tc f a dev lo parent 1: protocol arp basic classid 1:100
[...]
# tc q a dev lo parent 1:100 handle 100: pfifo_fast
[...]
# tc q d dev lo root
can generate the following splat:
Unable to handle kernel paging request at virtual address dfff2c01bd148000
Mem abort info:
ESR = 0x96000004
Exception class = DABT (current EL), IL = 32 bits
SET = 0, FnV = 0
EA = 0, S1PTW = 0
Data abort info:
ISV = 0, ISS = 0x00000004
CM = 0, WnR = 0
[dfff2c01bd148000] address between user and kernel address ranges
Internal error: Oops: 96000004 [#1] SMP
[...]
pstate: 80000005 (Nzcv daif -PAN -UAO)
pc : pfifo_fast_reset+0x280/0x4d8
lr : pfifo_fast_reset+0x21c/0x4d8
sp : ffff800d09676fa0
x29: ffff800d09676fa0 x28: ffff200012ee22e4
x27: dfff200000000000 x26: 0000000000000000
x25: ffff800ca0799958 x24: ffff1001940f332b
x23: 0000000000000007 x22: ffff200012ee1ab8
x21: 0000600de8a40000 x20: 0000000000000000
x19: ffff800ca0799900 x18: 0000000000000000
x17: 0000000000000002 x16: 0000000000000000
x15: 0000000000000000 x14: 0000000000000000
x13: 0000000000000000 x12: ffff1001b922e6e2
x11: 1ffff001b922e6e1 x10: 0000000000000000
x9 : 1ffff001b922e6e1 x8 : dfff200000000000
x7 : 0000000000000000 x6 : 0000000000000000
x5 : 1fffe400025dc45c x4 : 1fffe400025dc357
x3 : 00000c01bd148000 x2 : 0000600de8a40000
x1 : 0000000000000007 x0 : 0000600de8a40004
Call trace:
pfifo_fast_reset+0x280/0x4d8
qdisc_reset+0x6c/0x370
htb_reset+0x150/0x3b8 [sch_htb]
qdisc_reset+0x6c/0x370
dev_deactivate_queue.constprop.5+0xe0/0x1a8
dev_deactivate_many+0xd8/0x908
dev_deactivate+0xe4/0x190
qdisc_graft+0x88c/0xbd0
tc_get_qdisc+0x418/0x8a8
rtnetlink_rcv_msg+0x3a8/0xa78
netlink_rcv_skb+0x18c/0x328
rtnetlink_rcv+0x28/0x38
netlink_unicast+0x3c4/0x538
netlink_sendmsg+0x538/0x9a0
sock_sendmsg+0xac/0xf8
___sys_sendmsg+0x53c/0x658
__sys_sendmsg+0xc8/0x140
__arm64_sys_sendmsg+0x74/0xa8
el0_svc_handler+0x164/0x468
el0_svc+0x10/0x14
Code: 910012a0 92400801 d343fc03 11000c21 (38fb6863)
Fix this by testing the value of 'TCQ_F_CPUSTATS' bit in 'qdisc->flags',
before dereferencing 'qdisc->cpu_qstats'.
Changes since v1:
- coding style improvements, thanks to Stefano Brivio
Fixes: 8a53e616de ("net: sched: when clearing NOLOCK, clear TCQ_F_CPUSTATS, too")
CC: Paolo Abeni <pabeni@redhat.com>
Reported-by: Li Shuang <shuali@redhat.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The net pointer in struct xt_tgdtor_param is not explicitly
initialized therefore is still NULL when dereferencing it.
So we have to find a way to pass the correct net pointer to
ipt_destroy_target().
The best way I find is just saving the net pointer inside the per
netns struct tcf_idrinfo, which could make this patch smaller.
Fixes: 0c66dc1ea3 ("netfilter: conntrack: register hooks in netns when needed by ruleset")
Reported-and-tested-by: itugrok@yahoo.com
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Don't manually take rtnl lock in flower classifier before calling cls
hardware offloads API. Instead, pass rtnl lock status via 'rtnl_held'
parameter.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to remove dependency on rtnl lock, modify tc_setup_flow_action()
to copy tunnel info, instead of just saving pointer to tunnel_key action
tunnel info. This is necessary to prevent concurrent action overwrite from
releasing tunnel info while it is being used by rtnl-unlocked driver.
Implement helper tcf_tunnel_info_copy() that is used to copy tunnel info
with all its options to dynamically allocated memory block. Modify
tc_cleanup_flow_action() to free dynamically allocated tunnel info.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to remove dependency on rtnl lock when calling hardware offload
API, take reference to action mirred dev when initializing flow_action
structure in tc_setup_flow_action(). Implement function
tc_cleanup_flow_action(), use it to release the device after hardware
offload API is done using it.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to allow using new flow_action infrastructure from unlocked
classifiers, modify tc_setup_flow_action() to accept new 'rtnl_held'
argument. Take rtnl lock before accessing tc_action data. This is necessary
to protect from concurrent action replace.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to remove dependency on rtnl lock from offloads code of
classifiers, take rtnl lock conditionally before executing driver
callbacks. Only obtain rtnl lock if block is bound to devices that require
it.
Block bind/unbind code is rtnl-locked and obtains block->cb_lock while
holding rtnl lock. Obtain locks in same order in tc_setup_cb_*() functions
to prevent deadlock.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend struct flow_block_offload with "unlocked_driver_cb" flag to allow
registering and unregistering block hardware offload callbacks that do not
require caller to hold rtnl lock. Extend tcf_block with additional
lockeddevcnt counter that is incremented for each non-unlocked driver
callback attached to device. This counter is necessary to conditionally
obtain rtnl lock before calling hardware callbacks in following patches.
Register mlx5 tc block offload callbacks as "unlocked".
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To remove dependency on rtnl lock, extend classifier ops with new
ops->hw_add() and ops->hw_del() callbacks. Call them from cls API while
holding cb_lock every time filter if successfully added to or deleted from
hardware.
Implement the new API in flower classifier. Use it to manage hw_filters
list under cb_lock protection, instead of relying on rtnl lock to
synchronize with concurrent fl_reoffload() call.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Without rtnl lock protection filters can no longer safely manage block
offloads counter themselves. Refactor cls API to protect block offloadcnt
with tcf_block->cb_lock that is already used to protect driver callback
list and nooffloaddevcnt counter. The counter can be modified by concurrent
tasks by new functions that execute block callbacks (which is safe with
previous patch that changed its type to atomic_t), however, block
bind/unbind code that checks the counter value takes cb_lock in write mode
to exclude any concurrent modifications. This approach prevents race
conditions between bind/unbind and callback execution code but allows for
concurrency for tc rule update path.
Move block offload counter, filter in hardware counter and filter flags
management from classifiers into cls hardware offloads API. Make functions
tcf_block_offload_{inc|dec}() and tc_cls_offload_cnt_update() to be cls API
private. Implement following new cls API to be used instead:
tc_setup_cb_add() - non-destructive filter add. If filter that wasn't
already in hardware is successfully offloaded, increment block offloads
counter, set filter in hardware counter and flag. On failure, previously
offloaded filter is considered to be intact and offloads counter is not
decremented.
tc_setup_cb_replace() - destructive filter replace. Release existing
filter block offload counter and reset its in hardware counter and flag.
Set new filter in hardware counter and flag. On failure, previously
offloaded filter is considered to be destroyed and offload counter is
decremented.
tc_setup_cb_destroy() - filter destroy. Unconditionally decrement block
offloads counter.
tc_setup_cb_reoffload() - reoffload filter to single cb. Execute cb() and
call tc_cls_offload_cnt_update() if cb() didn't return an error.
Refactor all offload-capable classifiers to atomically offload filters to
hardware, change block offload counter, and set filter in hardware counter
and flag by means of the new cls API functions.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As a preparation for running proto ops functions without rtnl lock, change
offload counter type to atomic. This is necessary to allow updating the
counter by multiple concurrent users when offloading filters to hardware
from unlocked classifiers.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to remove dependency on rtnl lock, extend tcf_block with 'cb_lock'
rwsem and use it to protect flow_block->cb_list and related counters from
concurrent modification. The lock is taken in read mode for read-only
traversal of cb_list in tc_setup_cb_call() and write mode in all other
cases. This approach ensures that:
- cb_list is not changed concurrently while filters is being offloaded on
block.
- block->nooffloaddevcnt is checked while holding the lock in read mode,
but is only changed by bind/unbind code when holding the cb_lock in write
mode to prevent concurrent modification.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/sched/sch_taprio.c:680:32: warning:
entry_list_policy defined but not used [-Wunused-const-variable=]
One of the points of commit a3d43c0d56 ("taprio: Add support adding
an admin schedule") is that it removes support (it now returns "not
supported") for schedules using the TCA_TAPRIO_ATTR_SCHED_SINGLE_ENTRY
attribute (which were never used), the parsing of those types of schedules
was the only user of this policy. So removing this policy should be fine.
Reported-by: Hulk Robot <hulkci@huawei.com>
Suggested-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add get_fill_size() routine used to calculate the action size
when building a batch of events.
Fixes: ca9b0e27e ("pkt_action: add new action skbedit")
Signed-off-by: Roman Mashak <mrv@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes gcc '-Wunused-but-set-variable' warning:
net/sched/sch_fq_codel.c: In function fq_codel_dequeue:
net/sched/sch_fq_codel.c:288:23: warning: variable prev_ecn_mark set but not used [-Wunused-but-set-variable]
net/sched/sch_fq_codel.c:288:6: warning: variable prev_drop_count set but not used [-Wunused-but-set-variable]
They are not used since commit 77ddaff218 ("fq_codel: Kill
useless per-flow dropped statistic")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In error case, all entries should be freed from the sched list
before deleting it. For simplicity use rcu way.
Fixes: 5a781ccbd1 ("tc: Add support for configuring the taprio scheduler")
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
It provide a callback list to find the blocks of tc
and nft subsystems
Signed-off-by: wenxu <wenxu@ucloud.cn>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
move tc indirect block to flow_offload and rename
it to flow indirect block.The nf_tables can use the
indr block architecture.
Signed-off-by: wenxu <wenxu@ucloud.cn>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch make indr_block_call don't access struct tc_indr_block_cb
and tc_indr_block_dev directly
Signed-off-by: wenxu <wenxu@ucloud.cn>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove the tcf_block in the tc_indr_block_dev for muti-subsystem
support.
Signed-off-by: wenxu <wenxu@ucloud.cn>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
