Simple cases of overlapping changes in the packet scheduler.
Must easier to resolve this time.
Which probably means that I screwed it up somehow.
Signed-off-by: David S. Miller <davem@davemloft.net>
Hold netns refcnt before call_rcu() and release it after
the tcf_exts_destroy() is done.
Note, on ->destroy() path we have to respect the return value
of tcf_exts_get_net(), on other paths it should always return
true, so we don't need to care.
Cc: Lucas Bates <lucasb@mojatatu.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>
This patch offloads the classid to hardware and uses the classid
reserved in the range :ffe0 - :ffef to identify hardware traffic
classes reported via dev->num_tc.
tcf_result structure contains the class ID of the class to which
the packet belongs and is offloaded to hardware via flower filter.
A new helper function is introduced to represent HW traffic
classes 0 through 15 using the reserved classid values :ffe0 - :ffef.
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Acked-by: Shannon Nelson <shannon.nelson@oracle.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Several conflicts here.
NFP driver bug fix adding nfp_netdev_is_nfp_repr() check to
nfp_fl_output() needed some adjustments because the code block is in
an else block now.
Parallel additions to net/pkt_cls.h and net/sch_generic.h
A bug fix in __tcp_retransmit_skb() conflicted with some of
the rbtree changes in net-next.
The tc action RCU callback fixes in 'net' had some overlap with some
of the recent tcf_block reworking.
Signed-off-by: David S. Miller <davem@davemloft.net>
Defer the tcf_exts_destroy() in RCU callback to
tc filter workqueue and get RTNL lock.
Reported-by: Chris Mi <chrism@mellanox.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
All drivers are converted to use block callbacks for TC_SETUP_CLS*.
So it is now safe to remove the calls to ndo_setup_tc from cls_*
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend the tc_setup_cb_call entrypoint function originally used only for
action egress devices callbacks to call per-block callbacks as well.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 7091d8c '(net/sched: cls_flower: Add offload support using egress
Hardware device') made sure (when fl_hw_replace_filter is called) to put
the egress_dev mark on persisent structure instance. Hence, following calls
into the HW driver for stats and deletion will note it and act accordingly.
With commit de4784ca03 this property is lost and hence when called,
the HW driver failes to operate (stats, delete) on the offloaded flow.
Fix it by setting the egress_dev flag whenever the ingress device is
different from the hw device since this is exactly the condition under
which we're calling into the HW driver through the egress port net-device.
Fixes: de4784ca03 ('net: sched: get rid of struct tc_to_netdev')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roi Dayan <roid@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The only user of cls_flower->egress_dev is mlx5. So do the conversion
there alongside with the code originating the call in cls_flower
function fl_hw_replace_filter to the newly introduced egress device
callback infrastucture.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move dissection of tunnel info from the flower classifier to the flow
dissector where all other dissection occurs. This should not have any
behavioural affect on other users of the flow dissector.
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes: c15ab236d6 ("net/sched: Change cls_flower to use IDR")
Cc: Chris Mi <chrism@mellanox.com>
Cc: Jiri Pirko <jiri@mellanox.com>
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>
TC filters when used as classifiers are bound to TC classes.
However, there is a hidden difference when adding them in different
orders:
1. If we add tc classes before its filters, everything is fine.
Logically, the classes exist before we specify their ID's in
filters, it is easy to bind them together, just as in the current
code base.
2. If we add tc filters before the tc classes they bind, we have to
do dynamic lookup in fast path. What's worse, this happens all
the time not just once, because on fast path tcf_result is passed
on stack, there is no way to propagate back to the one in tc filters.
This hidden difference hurts performance silently if we have many tc
classes in hierarchy.
This patch intends to close this gap by doing the reverse binding when
we create a new class, in this case we can actually search all the
filters in its parent, match and fixup by classid. And because
tcf_result is specific to each type of tc filter, we have to introduce
a new ops for each filter to tell how to bind the class.
Note, we still can NOT totally get rid of those class lookup in
->enqueue() because cgroup and flow filters have no way to determine
the classid at setup time, they still have to go through dynamic lookup.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, all filters with the same priority are linked in a doubly
linked list. Every filter should have a unique handle. To make the
handle unique, we need to iterate the list every time to see if the
handle exists or not when inserting a new filter. It is time-consuming.
For example, it takes about 5m3.169s to insert 64K rules.
This patch changes cls_flower to use IDR. With this patch, it
takes about 0m1.127s to insert 64K rules. The improvement is huge.
But please note that in this testing, all filters share the same action.
If every filter has a unique action, that is another bottleneck.
Follow-up patch in this patchset addresses that.
Signed-off-by: Chris Mi <chrism@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I made a stupid mistake using TC_CLSFLOWER_STATS instead of
TC_SETUP_CLSFLOWER. Funny thing is that both are defined as "2" so it
actually did not cause any harm. Anyway, fixing it now.
Fixes: 2572ac53c4 ("net: sched: make type an argument for ndo_setup_tc")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
cops->tcf_cl_offload is no longer needed, as the drivers check what they
can and cannot offload using the classid identify helpers. So remove this.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now we use 'unsigned long fh' as a pointer in every place,
it is safe to convert it to a void pointer now. This gets
rid of many casts to pointer.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Get rid of struct tc_to_netdev which is now just unnecessary container
and rather pass per-type structures down to drivers directly.
Along with that, consolidate the naming of per-type structure variables
in cls_*.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
prio is not cls_flower specific, but it is meaningful for all
classifiers. Seems that only mlxsw cares about the value. Obviously,
cls offload in other drivers is broken.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As ndo_setup_tc is generic offload op for whole tc subsystem, does not
really make sense to have cls-specific args. So move them under
cls_common structurure which is embedded in all cls structs.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since this is specific to flower now, make it part of the flower offload
struct.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the type is always present, push it to be a separate argument to
ndo_setup_tc. On the way, name the type enum and use it for arg type.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As the f struct was allocated right before fl_set_parms call, no need
to use tcf_exts_change to do atomic change, and we can just fill-up
the unused exts struct directly by tcf_exts_validate.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to push the chain index down to the drivers, so they have the
information to which chain the rule belongs. For now, no driver supports
multichain offload, so only chain 0 is supported. This is needed to
prevent chain squashes during offload for now. Later this will be used
to implement multichain offload.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Benefit from the support of ip header fields dissection and
allow users to set rules matching on ipv4 tos and ttl or
ipv6 traffic-class and hoplimit.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Benefit from the support of tcp flags dissection and allow user to
insert rules matching on tcp flags.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since several of the the netlink attributes used to configure the flower
classifier's MPLS TC, BOS and Label fields have additional bits which are
unused, check those bits to ensure that they are actually 0 as suggested
by Jamal.
Signed-off-by: Benjamin LaHaise <benjamin.lahaise@netronome.com>
Cc: David Miller <davem@davemloft.net>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Simon Horman <simon.horman@netronome.com>
Cc: Jakub Kicinski <kubakici@wp.pl>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support to the tc flower classifier to match based on fields in MPLS
labels (TTL, Bottom of Stack, TC field, Label).
Signed-off-by: Benjamin LaHaise <benjamin.lahaise@netronome.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Simon Horman <simon.horman@netronome.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Hadar Hen Zion <hadarh@mellanox.com>
Cc: Gao Feng <fgao@ikuai8.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We could have a race condition where in ->classify() path we
dereference tp->root and meanwhile a parallel ->destroy() makes it
a NULL. Daniel cured this bug in commit d936377414
("net, sched: respect rcu grace period on cls destruction").
This happens when ->destroy() is called for deleting a filter to
check if we are the last one in tp, this tp is still linked and
visible at that time. The root cause of this problem is the semantic
of ->destroy(), it does two things (for non-force case):
1) check if tp is empty
2) if tp is empty we could really destroy it
and its caller, if cares, needs to check its return value to see if it
is really destroyed. Therefore we can't unlink tp unless we know it is
empty.
As suggested by Daniel, we could actually move the test logic to ->delete()
so that we can safely unlink tp after ->delete() tells us the last one is
just deleted and before ->destroy().
Fixes: 1e052be69d ("net_sched: destroy proto tp when all filters are gone")
Cc: Roi Dayan <roid@mellanox.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pass the new extended ACK reporting struct to all of the generic
netlink parsing functions. For now, pass NULL in almost all callers
(except for some in the core.)
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Flower support for the "in hw" offloading flags.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Amir Vadai <amir@vadai.me>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Dump the classifier flags only if non zero and make sure to check
the return status of the handler that puts them into the netlink msg.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The driver that offloads flower rules needs to know with which priority
user inserted the rules. So add this information into offload struct.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When matching on the ICMPv6 code ICMPV6_CODE rather than
ICMPV4_CODE attributes should be used.
This corrects what appears to be a typo.
Sample usage:
tc qdisc add dev eth0 ingress
tc filter add dev eth0 protocol ipv6 parent ffff: flower \
indev eth0 ip_proto icmpv6 type 128 code 0 action drop
Without this change the code parameter above is effectively ignored.
Fixes: 7b684884fb ("net/sched: cls_flower: Support matching on ICMP type and code")
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The new ARP support has pushed the stack size over the edge on ARM,
as there are two large objects on the stack in this function (mask
and tb) and both have now grown a bit more:
net/sched/cls_flower.c: In function 'fl_change':
net/sched/cls_flower.c:928:1: error: the frame size of 1072 bytes is larger than 1024 bytes [-Werror=frame-larger-than=]
We can solve this by dynamically allocating one or both of them.
I first tried to do it just for the mask, but that only saved
152 bytes on ARM, while this version just does it for the 'tb'
array, bringing the stack size back down to 664 bytes.
Fixes: 99d31326cb ("net/sched: cls_flower: Support matching on ARP")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Flower currently allows having the same filter twice with the same
priority. Actions (and statistics update) will always execute on the
first inserted rule leaving the second rule unused.
This patch disallows that.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Support matching on ARP operation, and hardware and protocol addresses
for Ethernet hardware and IPv4 protocol addresses.
Example usage:
tc qdisc add dev eth0 ingress
tc filter add dev eth0 protocol arp parent ffff: flower indev eth0 \
arp_op request arp_sip 10.0.0.1 action drop
tc filter add dev eth0 protocol rarp parent ffff: flower indev eth0 \
arp_op reply arp_tha 52:54:3f:00:00:00/24 action drop
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since we now use a non zero mask on addr_type, we are matching on its
value (IPV4/IPV6). So before this fix, matching on enc_src_ip/enc_dst_ip
failed in SW/classify path since its value was zero.
This patch sets the proper value of addr_type for encapsulated packets.
Fixes: 970bfcd097 ('net/sched: cls_flower: Use mask for addr_type')
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Hadar Hen Zion <hadarh@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When matching on flags, we should require the user to provide the
mask and avoid using an all-ones mask. Not doing so causes matching
on flags provided w.o mask to hit on the value being unset for all
flags, which may not what the user wanted to happen.
Fixes: faa3ffce78 ('net/sched: cls_flower: Add support for matching on flags')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reported-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Zero bits on the mask signify a "don't care" on the corresponding bits
in key. Some HWs require those bits on the key to be zero. Since these
bits are masked anyway, it's okay to provide the masked key to all
drivers.
Fixes: 5b33f48842 ('net/flower: Introduce hardware offload support')
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When addr_type is set, mask should also be set.
Fixes: 66530bdf85 ('sched,cls_flower: set key address type when present')
Fixes: bc3103f1ed ('net/sched: cls_flower: Classify packet in ip tunnels')
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Support matching on ICMP type and code.
Example usage:
tc qdisc add dev eth0 ingress
tc filter add dev eth0 protocol ip parent ffff: flower \
indev eth0 ip_proto icmp type 8 code 0 action drop
tc filter add dev eth0 protocol ipv6 parent ffff: flower \
indev eth0 ip_proto icmpv6 type 128 code 0 action drop
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add UAPI to provide set of flags for matching, where the flags
provided from user-space are mapped to flow-dissector flags.
The 1st flag allows to match on whether the packet is an
IP fragment and corresponds to the FLOW_DIS_IS_FRAGMENT flag.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Check if the returned device from tcf_exts_get_dev function supports tc
offload and in case the rule can't be offloaded, set the filter hw_dev
parameter to the original device given by the user.
The filter hw_device parameter should always be set by fl_hw_replace_filter
function, since this pointer is used by dump stats and destroy
filter for each flower rule (offloaded or not).
Fixes: 7091d8c705 ('net/sched: cls_flower: Add offload support using egress Hardware device')
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Reported-by: Simon Horman <horms@verge.net.au>
Tested-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Couple conflicts resolved here:
1) In the MACB driver, a bug fix to properly initialize the
RX tail pointer properly overlapped with some changes
to support variable sized rings.
2) In XGBE we had a "CONFIG_PM" --> "CONFIG_PM_SLEEP" fix
overlapping with a reorganization of the driver to support
ACPI, OF, as well as PCI variants of the chip.
3) In 'net' we had several probe error path bug fixes to the
stmmac driver, meanwhile a lot of this code was cleaned up
and reorganized in 'net-next'.
4) The cls_flower classifier obtained a helper function in
'net-next' called __fl_delete() and this overlapped with
Daniel Borkamann's bug fix to use RCU for object destruction
in 'net'. It also overlapped with Jiri's change to guard
the rhashtable_remove_fast() call with a check against
tc_skip_sw().
5) In mlx4, a revert bug fix in 'net' overlapped with some
unrelated changes in 'net-next'.
6) In geneve, a stale header pointer after pskb_expand_head()
bug fix in 'net' overlapped with a large reorganization of
the same code in 'net-next'. Since the 'net-next' code no
longer had the bug in question, there was nothing to do
other than to simply take the 'net-next' hunks.
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to support hardware offloading when the device given by the tc
rule is different from the Hardware underline device, extract the mirred
(egress) device from the tc action when a filter is added, using the new
tc_action_ops, get_dev().
Flower caches the information about the mirred device and use it for
calling ndo_setup_tc in filter change, update stats and delete.
Calling ndo_setup_tc of the mirred (egress) device instead of the
ingress device will allow a resolution between the software ingress
device and the underline hardware device.
The resolution will take place inside the offloading driver using
'egress_device' flag added to tc_to_netdev struct which is provided to
the offloading driver.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of providing many arguments to fl_hw_{replace/destroy}_filter
functions, just provide cls_fl_filter struct that includes all the relevant
args.
This patches doesn't add any new functionality.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Check skip_hw flag isn't set before calling
fl_hw_{replace/destroy}_filter and fl_hw_update_stats functions.
Replace the call to tc_should_offload with tc_can_offload.
tc_can_offload only checks if the device supports offloading, the check for
skip_hw flag is done earlier in the flow.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Be symmetric to hashtable insert and remove filter from hashtable only
in case skip sw flag is not set.
Fixes: e69985c67c ("net/sched: cls_flower: Introduce support in SKIP SW flag")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Amir Vadai <amir@vadai.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Roi reported a crash in flower where tp->root was NULL in ->classify()
callbacks. Reason is that in ->destroy() tp->root is set to NULL via
RCU_INIT_POINTER(). It's problematic for some of the classifiers, because
this doesn't respect RCU grace period for them, and as a result, still
outstanding readers from tc_classify() will try to blindly dereference
a NULL tp->root.
The tp->root object is strictly private to the classifier implementation
and holds internal data the core such as tc_ctl_tfilter() doesn't know
about. Within some classifiers, such as cls_bpf, cls_basic, etc, tp->root
is only checked for NULL in ->get() callback, but nowhere else. This is
misleading and seemed to be copied from old classifier code that was not
cleaned up properly. For example, d3fa76ee6b ("[NET_SCHED]: cls_basic:
fix NULL pointer dereference") moved tp->root initialization into ->init()
routine, where before it was part of ->change(), so ->get() had to deal
with tp->root being NULL back then, so that was indeed a valid case, after
d3fa76ee6b, not really anymore. We used to set tp->root to NULL long
ago in ->destroy(), see 47a1a1d4be ("pkt_sched: remove unnecessary xchg()
in packet classifiers"); but the NULLifying was reintroduced with the
RCUification, but it's not correct for every classifier implementation.
In the cases that are fixed here with one exception of cls_cgroup, tp->root
object is allocated and initialized inside ->init() callback, which is always
performed at a point in time after we allocate a new tp, which means tp and
thus tp->root was not globally visible in the tp chain yet (see tc_ctl_tfilter()).
Also, on destruction tp->root is strictly kfree_rcu()'ed in ->destroy()
handler, same for the tp which is kfree_rcu()'ed right when we return
from ->destroy() in tcf_destroy(). This means, the head object's lifetime
for such classifiers is always tied to the tp lifetime. The RCU callback
invocation for the two kfree_rcu() could be out of order, but that's fine
since both are independent.
Dropping the RCU_INIT_POINTER(tp->root, NULL) for these classifiers here
means that 1) we don't need a useless NULL check in fast-path and, 2) that
outstanding readers of that tp in tc_classify() can still execute under
respect with RCU grace period as it is actually expected.
Things that haven't been touched here: cls_fw and cls_route. They each
handle tp->root being NULL in ->classify() path for historic reasons, so
their ->destroy() implementation can stay as is. If someone actually
cares, they could get cleaned up at some point to avoid the test in fast
path. cls_u32 doesn't set tp->root to NULL. For cls_rsvp, I just added a
!head should anyone actually be using/testing it, so it at least aligns with
cls_fw and cls_route. For cls_flower we additionally need to defer rhashtable
destruction (to a sleepable context) after RCU grace period as concurrent
readers might still access it. (Note that in this case we need to hold module
reference to keep work callback address intact, since we only wait on module
unload for all call_rcu()s to finish.)
This fixes one race to bring RCU grace period guarantees back. Next step
as worked on by Cong however is to fix 1e052be69d ("net_sched: destroy
proto tp when all filters are gone") to get the order of unlinking the tp
in tc_ctl_tfilter() for the RTM_DELTFILTER case right by moving
RCU_INIT_POINTER() before tcf_destroy() and let the notification for
removal be done through the prior ->delete() callback. Both are independant
issues. Once we have that right, we can then clean tp->root up for a number
of classifiers by not making them RCU pointers, which requires a new callback
(->uninit) that is triggered from tp's RCU callback, where we just kfree()
tp->root from there.
Fixes: 1f947bf151 ("net: sched: rcu'ify cls_bpf")
Fixes: 9888faefe1 ("net: sched: cls_basic use RCU")
Fixes: 70da9f0bf9 ("net: sched: cls_flow use RCU")
Fixes: 77b9900ef5 ("tc: introduce Flower classifier")
Fixes: bf3994d2ed ("net/sched: introduce Match-all classifier")
Fixes: 952313bd62 ("net: sched: cls_cgroup use RCU")
Reported-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Roi Dayan <roid@mellanox.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
