This patch makes sure ENOBUFS is always returned from sendmsg if there
is no TX queue configured. This was not the case for zero-copy
mode. With this patch this error reporting is consistent between copy
mode and zero-copy mode.
Fixes: ac98d8aab6 ("xsk: wire upp Tx zero-copy functions")
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
This patch stops returning EAGAIN in TX copy mode when the completion
queue is full as zero-copy does not do this. Instead this situation
can be detected by comparing the head and tail pointers of the
completion queue in both modes. In any case, EAGAIN was not the
correct error code here since no amount of calling sendmsg will solve
the problem. Only consuming one or more messages on the completion
queue will fix this.
With this patch, the error reporting becomes consistent between copy
mode and zero-copy mode.
Fixes: 35fcde7f8d ("xsk: support for Tx")
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
This patch removes the ENXIO return code from TX copy-mode when
someone has forcefully changed the number of queues on the device so
that the queue bound to the socket is no longer available. Just
silently stop sending anything as in zero-copy mode so the error
reporting gets consistent between the two modes.
Fixes: 35fcde7f8d ("xsk: support for Tx")
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Add support for DEVLINK_CMD_REGION_READ_GET used for both reading
and dumping region data. Read allows reading from a region specific
address for given length. Dump allows reading the full region.
If only snapshot ID is provided a snapshot dump will be done.
If snapshot ID, Address and Length are provided a snapshot read
will done.
This is used for both snapshot access and will be used in the same
way to access current data on the region.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for DEVLINK_CMD_REGION_DEL used
for deleting a snapshot from a region. The snapshot ID is required.
Also added notification support for NEW and DEL of snapshots.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend the support for DEVLINK_CMD_REGION_GET command to also
return the IDs of the snapshot currently present on the region.
Each reply will include a nested snapshots attribute that
can contain multiple snapshot attributes each with an ID.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for DEVLINK_CMD_REGION_GET command which is used for
querying for the supported DEV/REGION values of devlink devices.
The support is both for doit and dumpit.
Reply includes:
BUS_NAME, DEVICE_NAME, REGION_NAME, REGION_SIZE
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Each device address region can store multiple snapshots,
each snapshot is identified using a different numerical ID.
This ID is used when deleting a snapshot or showing an address
region specific snapshot. This patch exposes a callback to add
a new snapshot to an address region.
The snapshot will be deleted using the destructor function
when destroying a region or when a snapshot delete command
from devlink user tool.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To restrict the driver with the snapshot ID selection a new callback
is introduced for the driver to get the snapshot ID before creating
a new snapshot. This will also allow giving the same ID for multiple
snapshots taken of different regions on the same time.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This allows a device to register its supported address regions.
Each address region can be accessed directly for example reading
the snapshots taken of this address space.
Drivers are not limited in the name selection for different regions.
An example of a region-name can be: pci cr-space, register-space.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If variable length link layer headers result in a packet shorter
than dev->hard_header_len, reset the network header offset. Else
skb->mac_len may exceed skb->len after skb_mac_reset_len.
packet_sendmsg_spkt already has similar logic.
Fixes: b84bbaf7a6 ("packet: in packet_snd start writing at link layer allocation")
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 5fa12739a5 ("net: ipv4: listify ip_rcv_finish") calling
dst_input(skb) was split-out. The ip_sublist_rcv_finish() just calls
dst_input(skb) in a loop.
The problem is that ip_sublist_rcv_finish() forgot to remove the SKB
from the list before invoking dst_input(). Further more we need to
clear skb->next as other parts of the network stack use another kind
of SKB lists for xmit_more (see dev_hard_start_xmit).
A crash occurs if e.g. dst_input() invoke ip_forward(), which calls
dst_output()/ip_output() that eventually calls __dev_queue_xmit() +
sch_direct_xmit(), and a crash occurs in validate_xmit_skb_list().
This patch only fixes the crash, but there is a huge potential for
a performance boost if we can pass an SKB-list through to ip_forward.
Fixes: 5fa12739a5 ("net: ipv4: listify ip_rcv_finish")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The pfmemalloc flag indicates that the skb was allocated from
the PFMEMALLOC reserves, and the flag is currently copied on skb
copy and clone.
However, an skb copied from an skb flagged with pfmemalloc
wasn't necessarily allocated from PFMEMALLOC reserves, and on
the other hand an skb allocated that way might be copied from an
skb that wasn't.
So we should not copy the flag on skb copy, and rather decide
whether to allow an skb to be associated with sockets unrelated
to page reclaim depending only on how it was allocated.
Move the pfmemalloc flag before headers_start[0] using an
existing 1-bit hole, so that __copy_skb_header() doesn't copy
it.
When cloning, we'll now take care of this flag explicitly,
contravening to the warning comment of __skb_clone().
While at it, restore the newline usage introduced by commit
b193722731 ("net: reorganize sk_buff for faster
__copy_skb_header()") to visually separate bytes used in
bitfields after headers_start[0], that was gone after commit
a9e419dc7b ("netfilter: merge ctinfo into nfct pointer storage
area"), and describe the pfmemalloc flag in the kernel-doc
structure comment.
This doesn't change the size of sk_buff or cacheline boundaries,
but consolidates the 15 bits hole before tc_index into a 2 bytes
hole before csum, that could now be filled more easily.
Reported-by: Patrick Talbert <ptalbert@redhat.com>
Fixes: c93bdd0e03 ("netvm: allow skb allocation to use PFMEMALLOC reserves")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
use RCU instead of spin_{,un}lock_bh, to protect concurrent read/write on
act_skbedit configuration. This reduces the effects of contention in the
data path, in case multiple readers are present.
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
use per-CPU counters, instead of sharing a single set of stats with all
cores: this removes the need of spinlocks when stats are read/updated.
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using get_seconds() for timestamps is deprecated since it can lead
to overflows on 32-bit systems. While the interface generally doesn't
overflow until year 2106, the specific implementation of the TCP PAWS
algorithm breaks in 2038 when the intermediate signed 32-bit timestamps
overflow.
A related problem is that the local timestamps in CLOCK_REALTIME form
lead to unexpected behavior when settimeofday is called to set the system
clock backwards or forwards by more than 24 days.
While the first problem could be solved by using an overflow-safe method
of comparing the timestamps, a nicer solution is to use a monotonic
clocksource with ktime_get_seconds() that simply doesn't overflow (at
least not until 136 years after boot) and that doesn't change during
settimeofday().
To make 32-bit and 64-bit architectures behave the same way here, and
also save a few bytes in the tcp_options_received structure, I'm changing
the type to a 32-bit integer, which is now safe on all architectures.
Finally, the ts_recent_stamp field also (confusingly) gets used to store
a jiffies value in tcp_synq_overflow()/tcp_synq_no_recent_overflow().
This is currently safe, but changing the type to 32-bit requires
some small changes there to keep it working.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of kzalloc/free for aead_request allocation and free, use
functions aead_request_alloc(), aead_request_free(). It ensures that
any sensitive crypto material held in crypto transforms is securely
erased from memory.
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Acked-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Under rare conditions where repair code may be used it is possible that
window probes are either unnecessary or undesired. If the user knows that
window probes are not wanted or needed this change allows them to skip
sending them when a socket comes out of repair.
Signed-off-by: Stefan Baranoff <sbaranoff@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes a bug where the sequence numbers of a socket created using
TCP repair functionality are lower than set after connect is called.
This occurs when the repair socket overlaps with a TIME-WAIT socket and
triggers the re-use code. The amount lower is equal to the number of times
that a particular IP/port set is re-used and then put back into TIME-WAIT.
Re-using the first time the sequence number is 1 lower, closing that socket
and then re-opening (with repair) a new socket with the same addresses/ports
puts the sequence number 2 lower than set via setsockopt. The third time is
3 lower, etc. I have not tested what the limit of this acrewal is, if any.
The fix is, if a socket is in repair mode, to respect the already set
sequence number and timestamp when it would have already re-used the
TIME-WAIT socket.
Signed-off-by: Stefan Baranoff <sbaranoff@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
takes inode, vfsmount, name, O_... flags and file_operations and
either returns a new struct file (in which case inode reference we
held is consumed) or returns ERR_PTR(), in which case no refcounts
are altered.
converted aio_private_file() and sock_alloc_file() to it
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
... so that it could set both ->f_flags and ->f_mode, without callers
having to set ->f_flags manually.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Stefan Schmidt says:
====================
pull-request: ieee802154 for net 2018-07-11
An update from ieee802154 for your *net* tree.
Build system fix for a missing include from Arnd Bergmann.
Setting the IFLA_LINK for the lowpan parent from Lubomir Rintel.
Fixes for some RX corner cases in adf7242 driver by Michael Hennerich.
And some small patches to cleanup our BUG_ON vs WARN_ON usage.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
In some virtual environments we observe a significant higher number of
packet reordering and delays than we have been used to traditionally.
This makes it necessary with stricter checks on incoming link protocol
messages' session number, which until now only has been validated for
RESET messages.
Since the other two message types, ACTIVATE and STATE messages also
carry this number, it is easy to extend the validation check to those
messages.
We also introduce a flag indicating if a link has a valid peer session
number or not. This eliminates the mixing of 32- and 16-bit arithmethics
we are currently using to achieve this.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some switch infrastructures produce huge amounts of packet duplicates.
This becomes a problem if those messages are STATE/NACK protocol
messages, causing unnecessary retransmissions of already accepted
packets.
We now introduce a unique sequence number per STATE protocol message
so that duplicates can be identified and ignored. This will also be
useful when tracing such cases, and to avert replay attacks when TIPC
is encrypted.
For compatibility reasons we have to introduce a new capability flag
TIPC_LINK_PROTO_SEQNO to handle this new feature.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jeff Kirsher says:
====================
L2 Fwd Offload & 10GbE Intel Driver Updates 2018-07-09
This patch series is meant to allow support for the L2 forward offload, aka
MACVLAN offload without the need for using ndo_select_queue.
The existing solution currently requires that we use ndo_select_queue in
the transmit path if we want to associate specific Tx queues with a given
MACVLAN interface. In order to get away from this we need to repurpose the
tc_to_txq array and XPS pointer for the MACVLAN interface and use those as
a means of accessing the queues on the lower device. As a result we cannot
offload a device that is configured as multiqueue, however it doesn't
really make sense to configure a macvlan interfaced as being multiqueue
anyway since it doesn't really have a qdisc of its own in the first place.
The big changes in this set are:
Allow lower device to update tc_to_txq and XPS map of offloaded MACVLAN
Disable XPS for single queue devices
Replace accel_priv with sb_dev in ndo_select_queue
Add sb_dev parameter to fallback function for ndo_select_queue
Consolidated ndo_select_queue functions that appeared to be duplicates
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Congestion control algorithms, which access the rate sample
through the tcp_cong_control function, only have access to the maximum
of the send and receive interval, for cases where the acknowledgment
rate may be inaccurate due to ACK compression or decimation. Algorithms
may want to use send rates and receive rates as separate signals.
Signed-off-by: Deepti Raghavan <deeptir@mit.edu>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix action attribute size calculation function to take rcu read lock and
access act_cookie pointer with rcu dereference.
Fixes: eec94fdb04 ("net: sched: use rcu for action cookie update")
Reported-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This aligns the addr_gen_mode sysctl with the expected behavior of the
"all" variant.
Fixes: d35a00b8e3 ("net/ipv6: allow sysctl to change link-local address generation mode")
Suggested-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
inet6_ifla6_size() is called to check how much space is needed by
inet6_fill_link_af() and inet6_fill_ifinfo(), both of which include
the IFLA_INET6_ADDR_GEN_MODE attribute. Reserve some room for it.
Fixes: bc91b0f07a ("ipv6: addrconf: implement address generation modes")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The value has already been copied from this netns's devconf_dflt, it
shouldn't be reset to the global kernel default.
Fixes: d35a00b8e3 ("net/ipv6: allow sysctl to change link-local address generation mode")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
addrconf_sysctl_addr_gen_mode() has multiple problems. First, it ignores
the errors returned by proc_dointvec().
addrconf_sysctl_addr_gen_mode() calls proc_dointvec() directly, which
writes the value to memory, and then checks if it's valid and may return
EINVAL. If a bad value is given, the value displayed when reading
net.ipv6.conf.foo.addr_gen_mode next time will be invalid. In case the
value provided by the user was valid, addrconf_dev_config() won't be
called since idev->cnf.addr_gen_mode has already been updated.
Fix this in the usual way we deal with values that need to be checked
after the proc_do*() helper has returned: define a local ctl_table and
storage, call proc_dointvec() on that temporary area, then check and
store.
addrconf_sysctl_addr_gen_mode() also writes the new value to the global
ipv6_devconf_dflt, when we're writing to some netns's default, so that
new netns will inherit the value that was set by the change occuring in
any netns. That doesn't make any sense, so let's drop this assignment.
Finally, since addr_gen_mode is a __u32, switch to proc_douintvec().
Fixes: d35a00b8e3 ("net/ipv6: allow sysctl to change link-local address generation mode")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Zahari issued tc vlan command without setting vlan_ethtype, which will
crash kernel. To avoid this, we must check tb[TCA_FLOWER_KEY_VLAN_ETH_TYPE]
is not null before use it.
Also we don't need to dump vlan_ethtype or cvlan_ethtype in this case.
Fixes: d64efd0926 ('net/sched: flower: Add supprt for matching on QinQ vlan headers')
Signed-off-by: Jianbo Liu <jianbol@mellanox.com>
Reported-by: Zahari Doychev <zahari.doychev@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sykzaller triggered several panics similar to the below:
[...]
[ 248.851531] BUG: KASAN: use-after-free in _copy_to_user+0x5c/0x90
[ 248.857656] Read of size 985 at addr ffff8808017ffff2 by task a.out/1425
[...]
[ 248.865902] CPU: 1 PID: 1425 Comm: a.out Not tainted 4.18.0-rc4+ #13
[ 248.865903] Hardware name: Supermicro SYS-5039MS-H12TRF/X11SSE-F, BIOS 2.1a 03/08/2018
[ 248.865905] Call Trace:
[ 248.865910] dump_stack+0xd6/0x185
[ 248.865911] ? show_regs_print_info+0xb/0xb
[ 248.865913] ? printk+0x9c/0xc3
[ 248.865915] ? kmsg_dump_rewind_nolock+0xe4/0xe4
[ 248.865919] print_address_description+0x6f/0x270
[ 248.865920] kasan_report+0x25b/0x380
[ 248.865922] ? _copy_to_user+0x5c/0x90
[ 248.865924] check_memory_region+0x137/0x190
[ 248.865925] kasan_check_read+0x11/0x20
[ 248.865927] _copy_to_user+0x5c/0x90
[ 248.865930] bpf_test_finish.isra.8+0x4f/0xc0
[ 248.865932] bpf_prog_test_run_skb+0x6a0/0xba0
[...]
After scrubbing the BPF prog a bit from the noise, turns out it called
bpf_skb_change_head() for the lwt_xmit prog with headroom of 2. Nothing
wrong in that, however, this was run with repeat >> 0 in bpf_prog_test_run_skb()
and the same skb thus keeps changing until the pskb_expand_head() called
from skb_cow() keeps bailing out in atomic alloc context with -ENOMEM.
So upon return we'll basically have 0 headroom left yet blindly do the
__skb_push() of 14 bytes and keep copying data from there in bpf_test_finish()
out of bounds. Fix to check if we have enough headroom and if pskb_expand_head()
fails, bail out with error.
Another bug independent of this fix (but related in triggering above) is
that BPF_PROG_TEST_RUN should be reworked to reset the skb/xdp buffer to
it's original state from input as otherwise repeating the same test in a
loop won't work for benchmarking when underlying input buffer is getting
changed by the prog each time and reused for the next run leading to
unexpected results.
Fixes: 1cf1cae963 ("bpf: introduce BPF_PROG_TEST_RUN command")
Reported-by: syzbot+709412e651e55ed96498@syzkaller.appspotmail.com
Reported-by: syzbot+54f39d6ab58f39720a55@syzkaller.appspotmail.com
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
bpf_lwt_seg6_* helpers require CONFIG_IPV6_SEG6_BPF, and currently
return -EOPNOTSUPP to indicate unavailability. This patch forces the
BPF verifier to reject programs using these helpers when
!CONFIG_IPV6_SEG6_BPF, allowing users to more easily probe if they are
available or not.
Signed-off-by: Mathieu Xhonneux <m.xhonneux@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
The nes infiniband driver uses current_kernel_time() to get a nanosecond
granunarity timestamp to initialize its tcp sequence counters. This is
one of only a few remaining users of that deprecated function, so we
should try to get rid of it.
Aside from using a deprecated API, there are several problems I see here:
- Using a CLOCK_REALTIME based time source makes it predictable in
case the time base is synchronized.
- Using a coarse timestamp means it only gets updated once per jiffie,
making it even more predictable in order to avoid having to access
the hardware clock source
- The upper 2 bits are always zero because the nanoseconds are at most
999999999.
For the Linux TCP implementation, we use secure_tcp_seq(), which appears
to be appropriate here as well, and solves all the above problems.
i40iw uses a variant of the same code, so I do that same thing there
for ipv4. Unlike nes, i40e also supports ipv6, which needs to call
secure_tcpv6_seq instead.
Acked-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
get_seconds() is deprecated because it can overflow on 32-bit
architectures. For the xfrm_state->lastused member, we treat the data
as a 64-bit number already, so we just need to use the right accessor
that works on both 32-bit and 64-bit machines.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
The lifetime managment uses '__u64' timestamps on the user space
interface, but 'unsigned long' for reading the current time in the kernel
with get_seconds().
While this is probably safe beyond y2038, it will still overflow in 2106,
and the get_seconds() call is deprecated because fo that.
This changes the xfrm time handling to use time64_t consistently, along
with reading the time using the safer ktime_get_real_seconds(). It still
suffers from problems that can happen from a concurrent settimeofday()
call or (to a lesser degree) a leap second update, but since the time
stamps are part of the user API, there is nothing we can do to prevent
that.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
At lower bandwidths, the transmission time of a single GSO segment can add
an unacceptable amount of latency due to HOL blocking. Furthermore, with a
software shaper, any tuning mechanism employed by the kernel to control the
maximum size of GSO segments is thrown off by the artificial limit on
bandwidth. For this reason, we split GSO segments into their individual
packets iff the shaper is active and configured to a bandwidth <= 1 Gbps.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit adds configurable overhead compensation support to the rate
shaper. With this feature, userspace can configure the actual bottleneck
link overhead and encapsulation mode used, which will be used by the shaper
to calculate the precise duration of each packet on the wire.
This feature is needed because CAKE is often deployed one or two hops
upstream of the actual bottleneck (which can be, e.g., inside a DSL or
cable modem). In this case, the link layer characteristics and overhead
reported by the kernel does not match the actual bottleneck. Being able to
set the actual values in use makes it possible to configure the shaper rate
much closer to the actual bottleneck rate (our experience shows it is
possible to get with 0.1% of the actual physical bottleneck rate), thus
keeping latency low without sacrificing bandwidth.
The overhead compensation has three tunables: A fixed per-packet overhead
size (which, if set, will be accounted from the IP packet header), a
minimum packet size (MPU) and a framing mode supporting either ATM or PTM
framing. We include a set of common keywords in TC to help users configure
the right parameters. If no overhead value is set, the value reported by
the kernel is used.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds support for DiffServ-based priority queueing to CAKE. If the
shaper is in use, each priority tier gets its own virtual clock, which
limits that tier's rate to a fraction of the overall shaped rate, to
discourage trying to game the priority mechanism.
CAKE defaults to a simple, three-tier mode that interprets most code points
as "best effort", but places CS1 traffic into a low-priority "bulk" tier
which is assigned 1/16 of the total rate, and a few code points indicating
latency-sensitive or control traffic (specifically TOS4, VA, EF, CS6, CS7)
into a "latency sensitive" high-priority tier, which is assigned 1/4 rate.
The other supported DiffServ modes are a 4-tier mode matching the 802.11e
precedence rules, as well as two 8-tier modes, one of which implements
strict precedence of the eight priority levels.
This commit also adds an optional DiffServ 'wash' mode, which will zero out
the DSCP fields of any packet passing through CAKE. While this can
technically be done with other mechanisms in the kernel, having the feature
available in CAKE significantly decreases configuration complexity; and the
implementation cost is low on top of the other DiffServ-handling code.
Filters and applications can set the skb->priority field to override the
DSCP-based classification into tiers. If TC_H_MAJ(skb->priority) matches
CAKE's qdisc handle, the minor number will be interpreted as a priority
tier if it is less than or equal to the number of configured priority
tiers.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CAKE is deployed on a gateway that also performs NAT (which is a
common deployment mode), the host fairness mechanism cannot distinguish
internal hosts from each other, and so fails to work correctly.
To fix this, we add an optional NAT awareness mode, which will query the
kernel conntrack mechanism to obtain the pre-NAT addresses for each packet
and use that in the flow and host hashing.
When the shaper is enabled and the host is already performing NAT, the cost
of this lookup is negligible. However, in unlimited mode with no NAT being
performed, there is a significant CPU cost at higher bandwidths. For this
reason, the feature is turned off by default.
Cc: netfilter-devel@vger.kernel.org
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds a global netfilter function to extract a conntrack tuple from an
skb. The function uses a new function added to nf_ct_hook, which will try
to get the tuple from skb->_nfct, and do a full lookup if that fails. This
makes it possible to use the lookup function before the skb has passed
through the conntrack init hooks (e.g., in an ingress qdisc). The tuple is
copied to the caller to avoid issues with reference counting.
The function returns false if conntrack is not loaded, allowing it to be
used without incurring a module dependency on conntrack. This is used by
the NAT mode in sch_cake.
Cc: netfilter-devel@vger.kernel.org
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ACK filter is an optional feature of CAKE which is designed to improve
performance on links with very asymmetrical rate limits. On such links
(which are unfortunately quite prevalent, especially for DSL and cable
subscribers), the downstream throughput can be limited by the number of
ACKs capable of being transmitted in the *upstream* direction.
Filtering ACKs can, in general, have adverse effects on TCP performance
because it interferes with ACK clocking (especially in slow start), and it
reduces the flow's resiliency to ACKs being dropped further along the path.
To alleviate these drawbacks, the ACK filter in CAKE tries its best to
always keep enough ACKs queued to ensure forward progress in the TCP flow
being filtered. It does this by only filtering redundant ACKs. In its
default 'conservative' mode, the filter will always keep at least two
redundant ACKs in the queue, while in 'aggressive' mode, it will filter
down to a single ACK.
The ACK filter works by inspecting the per-flow queue on every packet
enqueue. Starting at the head of the queue, the filter looks for another
eligible packet to drop (so the ACK being dropped is always closer to the
head of the queue than the packet being enqueued). An ACK is eligible only
if it ACKs *fewer* bytes than the new packet being enqueued, including any
SACK options. This prevents duplicate ACKs from being filtered, to avoid
interfering with retransmission logic. In addition, we check TCP header
options and only drop those that are known to not interfere with sender
state. In particular, packets with unknown option codes are never dropped.
In aggressive mode, an eligible packet is always dropped, while in
conservative mode, at least two ACKs are kept in the queue. Only pure ACKs
(with no data segments) are considered eligible for dropping, but when an
ACK with data segments is enqueued, this can cause another pure ACK to
become eligible for dropping.
The approach described above ensures that this ACK filter avoids most of
the drawbacks of a naive filtering mechanism that only keeps flow state but
does not inspect the queue. This is the rationale for including the ACK
filter in CAKE itself rather than as separate module (as the TC filter, for
instance).
Our performance evaluation has shown that on a 30/1 Mbps link with a
bidirectional traffic test (RRUL), turning on the ACK filter on the
upstream link improves downstream throughput by ~20% (both modes) and
upstream throughput by ~12% in conservative mode and ~40% in aggressive
mode, at the cost of ~5ms of inter-flow latency due to the increased
congestion.
In *really* pathological cases, the effect can be a lot more; for instance,
the ACK filter increases the achievable downstream throughput on a link
with 100 Kbps in the upstream direction by an order of magnitude (from ~2.5
Mbps to ~25 Mbps).
Finally, even though we consider the ACK filter to be safer than most, we
do not recommend turning it on everywhere: on more symmetrical link
bandwidths the effect is negligible at best.
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ingress mode is meant to be enabled when CAKE runs downlink of the
actual bottleneck (such as on an IFB device). The mode changes the shaper
to also account dropped packets to the shaped rate, as these have already
traversed the bottleneck.
Enabling ingress mode will also tune the AQM to always keep at least two
packets queued *for each flow*. This is done by scaling the minimum queue
occupancy level that will disable the AQM by the number of active bulk
flows. The rationale for this is that retransmits are more expensive in
ingress mode, since dropped packets have to traverse the bottleneck again
when they are retransmitted; thus, being more lenient and keeping a minimum
number of packets queued will improve throughput in cases where the number
of active flows are so large that they saturate the bottleneck even at
their minimum window size.
This commit also adds a separate switch to enable ingress mode rate
autoscaling. If enabled, the autoscaling code will observe the actual
traffic rate and adjust the shaper rate to match it. This can help avoid
latency increases in the case where the actual bottleneck rate decreases
below the shaped rate. The scaling filters out spikes by an EWMA filter.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
sch_cake targets the home router use case and is intended to squeeze the
most bandwidth and latency out of even the slowest ISP links and routers,
while presenting an API simple enough that even an ISP can configure it.
Example of use on a cable ISP uplink:
tc qdisc add dev eth0 cake bandwidth 20Mbit nat docsis ack-filter
To shape a cable download link (ifb and tc-mirred setup elided)
tc qdisc add dev ifb0 cake bandwidth 200mbit nat docsis ingress wash
CAKE is filled with:
* A hybrid Codel/Blue AQM algorithm, "Cobalt", tied to an FQ_Codel
derived Flow Queuing system, which autoconfigures based on the bandwidth.
* A novel "triple-isolate" mode (the default) which balances per-host
and per-flow FQ even through NAT.
* An deficit based shaper, that can also be used in an unlimited mode.
* 8 way set associative hashing to reduce flow collisions to a minimum.
* A reasonable interpretation of various diffserv latency/loss tradeoffs.
* Support for zeroing diffserv markings for entering and exiting traffic.
* Support for interacting well with Docsis 3.0 shaper framing.
* Extensive support for DSL framing types.
* Support for ack filtering.
* Extensive statistics for measuring, loss, ecn markings, latency
variation.
A paper describing the design of CAKE is available at
https://arxiv.org/abs/1804.07617, and will be published at the 2018 IEEE
International Symposium on Local and Metropolitan Area Networks (LANMAN).
This patch adds the base shaper and packet scheduler, while subsequent
commits add the optional (configurable) features. The full userspace API
and most data structures are included in this commit, but options not
understood in the base version will be ignored.
Various versions baking have been available as an out of tree build for
kernel versions going back to 3.10, as the embedded router world has been
running a few years behind mainline Linux. A stable version has been
generally available on lede-17.01 and later.
sch_cake replaces a combination of iptables, tc filter, htb and fq_codel
in the sqm-scripts, with sane defaults and vastly simpler configuration.
CAKE's principal author is Jonathan Morton, with contributions from
Kevin Darbyshire-Bryant, Toke Høiland-Jørgensen, Sebastian Moeller,
Ryan Mounce, Tony Ambardar, Dean Scarff, Nils Andreas Svee, Dave Täht,
and Loganaden Velvindron.
Testing from Pete Heist, Georgios Amanakis, and the many other members of
the cake@lists.bufferbloat.net mailing list.
tc -s qdisc show dev eth2
qdisc cake 8017: root refcnt 2 bandwidth 1Gbit diffserv3 triple-isolate split-gso rtt 100.0ms noatm overhead 38 mpu 84
Sent 51504294511 bytes 37724591 pkt (dropped 6, overlimits 64958695 requeues 12)
backlog 0b 0p requeues 12
memory used: 1053008b of 15140Kb
capacity estimate: 970Mbit
min/max network layer size: 28 / 1500
min/max overhead-adjusted size: 84 / 1538
average network hdr offset: 14
Bulk Best Effort Voice
thresh 62500Kbit 1Gbit 250Mbit
target 5.0ms 5.0ms 5.0ms
interval 100.0ms 100.0ms 100.0ms
pk_delay 5us 5us 6us
av_delay 3us 2us 2us
sp_delay 2us 1us 1us
backlog 0b 0b 0b
pkts 3164050 25030267 9530280
bytes 3227519915 35396974782 12879808898
way_inds 0 8 0
way_miss 21 366 25
way_cols 0 0 0
drops 5 0 1
marks 0 0 0
ack_drop 0 0 0
sp_flows 1 3 0
bk_flows 0 1 1
un_flows 0 0 0
max_len 68130 68130 68130
Tested-by: Pete Heist <peteheist@gmail.com>
Tested-by: Georgios Amanakis <gamanakis@gmail.com>
Signed-off-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
