Up to now several high speed NICs have custom mechanisms of recycling
the allocated memory they use for their payloads.
Our page_pool API already has recycling capabilities that are always
used when we are running in 'XDP mode'. So let's tweak the API and the
kernel network stack slightly and allow the recycling to happen even
during the standard operation.
The API doesn't take into account 'split page' policies used by those
drivers currently, but can be extended once we have users for that.
The idea is to be able to intercept the packet on skb_release_data().
If it's a buffer coming from our page_pool API recycle it back to the
pool for further usage or just release the packet entirely.
To achieve that we introduce a bit in struct sk_buff (pp_recycle:1) and
a field in struct page (page->pp) to store the page_pool pointer.
Storing the information in page->pp allows us to recycle both SKBs and
their fragments.
We could have skipped the skb bit entirely, since identical information
can bederived from struct page. However, in an effort to affect the free path
as less as possible, reading a single bit in the skb which is already
in cache, is better that trying to derive identical information for the
page stored data.
The driver or page_pool has to take care of the sync operations on it's own
during the buffer recycling since the buffer is, after opting-in to the
recycling, never unmapped.
Since the gain on the drivers depends on the architecture, we are not
enabling recycling by default if the page_pool API is used on a driver.
In order to enable recycling the driver must call skb_mark_for_recycle()
to store the information we need for recycling in page->pp and
enabling the recycling bit, or page_pool_store_mem_info() for a fragment.
Co-developed-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Co-developed-by: Matteo Croce <mcroce@microsoft.com>
Signed-off-by: Matteo Croce <mcroce@microsoft.com>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is a prerequisite patch, the next one is enabling recycling of
skbs and fragments. Add an extra argument on __skb_frag_unref() to
handle recycling, and update the current users of the function with that.
Signed-off-by: Matteo Croce <mcroce@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Core:
- bpf:
- allow bpf programs calling kernel functions (initially to
reuse TCP congestion control implementations)
- enable task local storage for tracing programs - remove the
need to store per-task state in hash maps, and allow tracing
programs access to task local storage previously added for
BPF_LSM
- add bpf_for_each_map_elem() helper, allowing programs to
walk all map elements in a more robust and easier to verify
fashion
- sockmap: support UDP and cross-protocol BPF_SK_SKB_VERDICT
redirection
- lpm: add support for batched ops in LPM trie
- add BTF_KIND_FLOAT support - mostly to allow use of BTF
on s390 which has floats in its headers files
- improve BPF syscall documentation and extend the use of kdoc
parsing scripts we already employ for bpf-helpers
- libbpf, bpftool: support static linking of BPF ELF files
- improve support for encapsulation of L2 packets
- xdp: restructure redirect actions to avoid a runtime lookup,
improving performance by 4-8% in microbenchmarks
- xsk: build skb by page (aka generic zerocopy xmit) - improve
performance of software AF_XDP path by 33% for devices
which don't need headers in the linear skb part (e.g. virtio)
- nexthop: resilient next-hop groups - improve path stability
on next-hops group changes (incl. offload for mlxsw)
- ipv6: segment routing: add support for IPv4 decapsulation
- icmp: add support for RFC 8335 extended PROBE messages
- inet: use bigger hash table for IP ID generation
- tcp: deal better with delayed TX completions - make sure we don't
give up on fast TCP retransmissions only because driver is
slow in reporting that it completed transmitting the original
- tcp: reorder tcp_congestion_ops for better cache locality
- mptcp:
- add sockopt support for common TCP options
- add support for common TCP msg flags
- include multiple address ids in RM_ADDR
- add reset option support for resetting one subflow
- udp: GRO L4 improvements - improve 'forward' / 'frag_list'
co-existence with UDP tunnel GRO, allowing the first to take
place correctly even for encapsulated UDP traffic
- micro-optimize dev_gro_receive() and flow dissection, avoid
retpoline overhead on VLAN and TEB GRO
- use less memory for sysctls, add a new sysctl type, to allow using
u8 instead of "int" and "long" and shrink networking sysctls
- veth: allow GRO without XDP - this allows aggregating UDP
packets before handing them off to routing, bridge, OvS, etc.
- allow specifing ifindex when device is moved to another namespace
- netfilter:
- nft_socket: add support for cgroupsv2
- nftables: add catch-all set element - special element used
to define a default action in case normal lookup missed
- use net_generic infra in many modules to avoid allocating
per-ns memory unnecessarily
- xps: improve the xps handling to avoid potential out-of-bound
accesses and use-after-free when XPS change race with other
re-configuration under traffic
- add a config knob to turn off per-cpu netdev refcnt to catch
underflows in testing
Device APIs:
- add WWAN subsystem to organize the WWAN interfaces better and
hopefully start driving towards more unified and vendor-
-independent APIs
- ethtool:
- add interface for reading IEEE MIB stats (incl. mlx5 and
bnxt support)
- allow network drivers to dump arbitrary SFP EEPROM data,
current offset+length API was a poor fit for modern SFP
which define EEPROM in terms of pages (incl. mlx5 support)
- act_police, flow_offload: add support for packet-per-second
policing (incl. offload for nfp)
- psample: add additional metadata attributes like transit delay
for packets sampled from switch HW (and corresponding egress
and policy-based sampling in the mlxsw driver)
- dsa: improve support for sandwiched LAGs with bridge and DSA
- netfilter:
- flowtable: use direct xmit in topologies with IP
forwarding, bridging, vlans etc.
- nftables: counter hardware offload support
- Bluetooth:
- improvements for firmware download w/ Intel devices
- add support for reading AOSP vendor capabilities
- add support for virtio transport driver
- mac80211:
- allow concurrent monitor iface and ethernet rx decap
- set priority and queue mapping for injected frames
- phy: add support for Clause-45 PHY Loopback
- pci/iov: add sysfs MSI-X vector assignment interface
to distribute MSI-X resources to VFs (incl. mlx5 support)
New hardware/drivers:
- dsa: mv88e6xxx: add support for Marvell mv88e6393x -
11-port Ethernet switch with 8x 1-Gigabit Ethernet
and 3x 10-Gigabit interfaces.
- dsa: support for legacy Broadcom tags used on BCM5325, BCM5365
and BCM63xx switches
- Microchip KSZ8863 and KSZ8873; 3x 10/100Mbps Ethernet switches
- ath11k: support for QCN9074 a 802.11ax device
- Bluetooth: Broadcom BCM4330 and BMC4334
- phy: Marvell 88X2222 transceiver support
- mdio: add BCM6368 MDIO mux bus controller
- r8152: support RTL8153 and RTL8156 (USB Ethernet) chips
- mana: driver for Microsoft Azure Network Adapter (MANA)
- Actions Semi Owl Ethernet MAC
- can: driver for ETAS ES58X CAN/USB interfaces
Pure driver changes:
- add XDP support to: enetc, igc, stmmac
- add AF_XDP support to: stmmac
- virtio:
- page_to_skb() use build_skb when there's sufficient tailroom
(21% improvement for 1000B UDP frames)
- support XDP even without dedicated Tx queues - share the Tx
queues with the stack when necessary
- mlx5:
- flow rules: add support for mirroring with conntrack,
matching on ICMP, GTP, flex filters and more
- support packet sampling with flow offloads
- persist uplink representor netdev across eswitch mode
changes
- allow coexistence of CQE compression and HW time-stamping
- add ethtool extended link error state reporting
- ice, iavf: support flow filters, UDP Segmentation Offload
- dpaa2-switch:
- move the driver out of staging
- add spanning tree (STP) support
- add rx copybreak support
- add tc flower hardware offload on ingress traffic
- ionic:
- implement Rx page reuse
- support HW PTP time-stamping
- octeon: support TC hardware offloads - flower matching on ingress
and egress ratelimitting.
- stmmac:
- add RX frame steering based on VLAN priority in tc flower
- support frame preemption (FPE)
- intel: add cross time-stamping freq difference adjustment
- ocelot:
- support forwarding of MRP frames in HW
- support multiple bridges
- support PTP Sync one-step timestamping
- dsa: mv88e6xxx, dpaa2-switch: offload bridge port flags like
learning, flooding etc.
- ipa: add IPA v4.5, v4.9 and v4.11 support (Qualcomm SDX55, SM8350,
SC7280 SoCs)
- mt7601u: enable TDLS support
- mt76:
- add support for 802.3 rx frames (mt7915/mt7615)
- mt7915 flash pre-calibration support
- mt7921/mt7663 runtime power management fixes
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=vcbA
-----END PGP SIGNATURE-----
Merge tag 'net-next-5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking updates from Jakub Kicinski:
"Core:
- bpf:
- allow bpf programs calling kernel functions (initially to
reuse TCP congestion control implementations)
- enable task local storage for tracing programs - remove the
need to store per-task state in hash maps, and allow tracing
programs access to task local storage previously added for
BPF_LSM
- add bpf_for_each_map_elem() helper, allowing programs to walk
all map elements in a more robust and easier to verify fashion
- sockmap: support UDP and cross-protocol BPF_SK_SKB_VERDICT
redirection
- lpm: add support for batched ops in LPM trie
- add BTF_KIND_FLOAT support - mostly to allow use of BTF on
s390 which has floats in its headers files
- improve BPF syscall documentation and extend the use of kdoc
parsing scripts we already employ for bpf-helpers
- libbpf, bpftool: support static linking of BPF ELF files
- improve support for encapsulation of L2 packets
- xdp: restructure redirect actions to avoid a runtime lookup,
improving performance by 4-8% in microbenchmarks
- xsk: build skb by page (aka generic zerocopy xmit) - improve
performance of software AF_XDP path by 33% for devices which don't
need headers in the linear skb part (e.g. virtio)
- nexthop: resilient next-hop groups - improve path stability on
next-hops group changes (incl. offload for mlxsw)
- ipv6: segment routing: add support for IPv4 decapsulation
- icmp: add support for RFC 8335 extended PROBE messages
- inet: use bigger hash table for IP ID generation
- tcp: deal better with delayed TX completions - make sure we don't
give up on fast TCP retransmissions only because driver is slow in
reporting that it completed transmitting the original
- tcp: reorder tcp_congestion_ops for better cache locality
- mptcp:
- add sockopt support for common TCP options
- add support for common TCP msg flags
- include multiple address ids in RM_ADDR
- add reset option support for resetting one subflow
- udp: GRO L4 improvements - improve 'forward' / 'frag_list'
co-existence with UDP tunnel GRO, allowing the first to take place
correctly even for encapsulated UDP traffic
- micro-optimize dev_gro_receive() and flow dissection, avoid
retpoline overhead on VLAN and TEB GRO
- use less memory for sysctls, add a new sysctl type, to allow using
u8 instead of "int" and "long" and shrink networking sysctls
- veth: allow GRO without XDP - this allows aggregating UDP packets
before handing them off to routing, bridge, OvS, etc.
- allow specifing ifindex when device is moved to another namespace
- netfilter:
- nft_socket: add support for cgroupsv2
- nftables: add catch-all set element - special element used to
define a default action in case normal lookup missed
- use net_generic infra in many modules to avoid allocating
per-ns memory unnecessarily
- xps: improve the xps handling to avoid potential out-of-bound
accesses and use-after-free when XPS change race with other
re-configuration under traffic
- add a config knob to turn off per-cpu netdev refcnt to catch
underflows in testing
Device APIs:
- add WWAN subsystem to organize the WWAN interfaces better and
hopefully start driving towards more unified and vendor-
independent APIs
- ethtool:
- add interface for reading IEEE MIB stats (incl. mlx5 and bnxt
support)
- allow network drivers to dump arbitrary SFP EEPROM data,
current offset+length API was a poor fit for modern SFP which
define EEPROM in terms of pages (incl. mlx5 support)
- act_police, flow_offload: add support for packet-per-second
policing (incl. offload for nfp)
- psample: add additional metadata attributes like transit delay for
packets sampled from switch HW (and corresponding egress and
policy-based sampling in the mlxsw driver)
- dsa: improve support for sandwiched LAGs with bridge and DSA
- netfilter:
- flowtable: use direct xmit in topologies with IP forwarding,
bridging, vlans etc.
- nftables: counter hardware offload support
- Bluetooth:
- improvements for firmware download w/ Intel devices
- add support for reading AOSP vendor capabilities
- add support for virtio transport driver
- mac80211:
- allow concurrent monitor iface and ethernet rx decap
- set priority and queue mapping for injected frames
- phy: add support for Clause-45 PHY Loopback
- pci/iov: add sysfs MSI-X vector assignment interface to distribute
MSI-X resources to VFs (incl. mlx5 support)
New hardware/drivers:
- dsa: mv88e6xxx: add support for Marvell mv88e6393x - 11-port
Ethernet switch with 8x 1-Gigabit Ethernet and 3x 10-Gigabit
interfaces.
- dsa: support for legacy Broadcom tags used on BCM5325, BCM5365 and
BCM63xx switches
- Microchip KSZ8863 and KSZ8873; 3x 10/100Mbps Ethernet switches
- ath11k: support for QCN9074 a 802.11ax device
- Bluetooth: Broadcom BCM4330 and BMC4334
- phy: Marvell 88X2222 transceiver support
- mdio: add BCM6368 MDIO mux bus controller
- r8152: support RTL8153 and RTL8156 (USB Ethernet) chips
- mana: driver for Microsoft Azure Network Adapter (MANA)
- Actions Semi Owl Ethernet MAC
- can: driver for ETAS ES58X CAN/USB interfaces
Pure driver changes:
- add XDP support to: enetc, igc, stmmac
- add AF_XDP support to: stmmac
- virtio:
- page_to_skb() use build_skb when there's sufficient tailroom
(21% improvement for 1000B UDP frames)
- support XDP even without dedicated Tx queues - share the Tx
queues with the stack when necessary
- mlx5:
- flow rules: add support for mirroring with conntrack, matching
on ICMP, GTP, flex filters and more
- support packet sampling with flow offloads
- persist uplink representor netdev across eswitch mode changes
- allow coexistence of CQE compression and HW time-stamping
- add ethtool extended link error state reporting
- ice, iavf: support flow filters, UDP Segmentation Offload
- dpaa2-switch:
- move the driver out of staging
- add spanning tree (STP) support
- add rx copybreak support
- add tc flower hardware offload on ingress traffic
- ionic:
- implement Rx page reuse
- support HW PTP time-stamping
- octeon: support TC hardware offloads - flower matching on ingress
and egress ratelimitting.
- stmmac:
- add RX frame steering based on VLAN priority in tc flower
- support frame preemption (FPE)
- intel: add cross time-stamping freq difference adjustment
- ocelot:
- support forwarding of MRP frames in HW
- support multiple bridges
- support PTP Sync one-step timestamping
- dsa: mv88e6xxx, dpaa2-switch: offload bridge port flags like
learning, flooding etc.
- ipa: add IPA v4.5, v4.9 and v4.11 support (Qualcomm SDX55, SM8350,
SC7280 SoCs)
- mt7601u: enable TDLS support
- mt76:
- add support for 802.3 rx frames (mt7915/mt7615)
- mt7915 flash pre-calibration support
- mt7921/mt7663 runtime power management fixes"
* tag 'net-next-5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2451 commits)
net: selftest: fix build issue if INET is disabled
net: netrom: nr_in: Remove redundant assignment to ns
net: tun: Remove redundant assignment to ret
net: phy: marvell: add downshift support for M88E1240
net: dsa: ksz: Make reg_mib_cnt a u8 as it never exceeds 255
net/sched: act_ct: Remove redundant ct get and check
icmp: standardize naming of RFC 8335 PROBE constants
bpf, selftests: Update array map tests for per-cpu batched ops
bpf: Add batched ops support for percpu array
bpf: Implement formatted output helpers with bstr_printf
seq_file: Add a seq_bprintf function
sfc: adjust efx->xdp_tx_queue_count with the real number of initialized queues
net:nfc:digital: Fix a double free in digital_tg_recv_dep_req
net: fix a concurrency bug in l2tp_tunnel_register()
net/smc: Remove redundant assignment to rc
mpls: Remove redundant assignment to err
llc2: Remove redundant assignment to rc
net/tls: Remove redundant initialization of record
rds: Remove redundant assignment to nr_sig
dt-bindings: net: mdio-gpio: add compatible for microchip,mdio-smi0
...
the commit 1ddc3229ad ("skbuff: remove some unnecessary operation
in skb_segment_list()") introduces an issue very similar to the
one already fixed by commit 53475c5dd8 ("net: fix use-after-free when
UDP GRO with shared fraglist").
If the GSO skb goes though skb_clone() and pskb_expand_head() before
entering skb_segment_list(), the latter will unshare the frag_list
skbs and will release the old list. With the reverted commit in place,
when skb_segment_list() completes, skb->next points to the just
released list, and later on the kernel will hit UaF.
Note that since commit e0e3070a9b ("udp: properly complete L4 GRO
over UDP tunnel packet") the critical scenario can be reproduced also
receiving UDP over vxlan traffic with:
NIC (NETIF_F_GRO_FRAGLIST enabled) -> vxlan -> UDP sink
Attaching a packet socket to the NIC will cause skb_clone() and the
tunnel decapsulation will call pskb_expand_head().
Fixes: 1ddc3229ad ("skbuff: remove some unnecessary operation in skb_segment_list()")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We only have skb_send_sock_locked() which requires callers
to use lock_sock(). Introduce a variant skb_send_sock()
which locks on its own, callers do not need to lock it
any more. This will save us from adding a ->sendmsg_locked
for each protocol.
To reuse the code, pass function pointers to __skb_send_sock()
and build skb_send_sock() and skb_send_sock_locked() on top.
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20210331023237.41094-4-xiyou.wangcong@gmail.com
gro list uses skb_shinfo(skb)->frag_list to link two skb together,
and NAPI_GRO_CB(p)->last->next is used when there are more skb,
see skb_gro_receive_list(). gso expects that each segmented skb is
linked together using skb->next, so only the first skb->next need
to set to skb_shinfo(skb)-> frag_list when doing gso list segment.
It is the same reason that nskb->next does not need to be set to
list_skb before goto the error handling, because nskb->next already
pointers to list_skb.
And nskb is also the last skb at the end of loop, so remove tail
variable and use nskb instead.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The recent addition of in_serving_softirq() to kconv.h results in
compile failure on PREEMPT_RT because it requires
task_struct::softirq_disable_cnt. This is not available if kconv.h is
included from sched.h.
It is not needed to include kconv.h from sched.h. All but the net/ user
already include the kconv header file.
Move the include of the kconv.h header from sched.h it its users.
Additionally include sched.h from kconv.h to ensure that everything
task_struct related is available.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Andrey Konovalov <andreyknvl@google.com>
Link: https://lkml.kernel.org/r/20210218173124.iy5iyqv3a4oia4vv@linutronix.de
The referenced commit expands the skb_seq_state used by
skb_find_text with a 4B frag_off field, growing it to 48B.
This exceeds container ts_state->cb, causing a stack corruption:
[ 73.238353] Kernel panic - not syncing: stack-protector: Kernel stack
is corrupted in: skb_find_text+0xc5/0xd0
[ 73.247384] CPU: 1 PID: 376 Comm: nping Not tainted 5.11.0+ #4
[ 73.252613] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS 1.14.0-2 04/01/2014
[ 73.260078] Call Trace:
[ 73.264677] dump_stack+0x57/0x6a
[ 73.267866] panic+0xf6/0x2b7
[ 73.270578] ? skb_find_text+0xc5/0xd0
[ 73.273964] __stack_chk_fail+0x10/0x10
[ 73.277491] skb_find_text+0xc5/0xd0
[ 73.280727] string_mt+0x1f/0x30
[ 73.283639] ipt_do_table+0x214/0x410
The struct is passed between skb_find_text and its callbacks
skb_prepare_seq_read, skb_seq_read and skb_abort_seq read through
the textsearch interface using TS_SKB_CB.
I assumed that this mapped to skb->cb like other .._SKB_CB wrappers.
skb->cb is 48B. But it maps to ts_state->cb, which is only 40B.
skb->cb was increased from 40B to 48B after ts_state was introduced,
in commit 3e3850e989 ("[NETFILTER]: Fix xfrm lookup in
ip_route_me_harder/ip6_route_me_harder").
Increase ts_state.cb[] to 48 to fit the struct.
Also add a BUILD_BUG_ON to avoid a repeat.
The alternative is to directly add a dependency from textsearch onto
linux/skbuff.h, but I think the intent is textsearch to have no such
dependencies on its callers.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=211911
Fixes: 97550f6fa5 ("net: compound page support in skb_seq_read")
Reported-by: Kris Karas <bugs-a17@moonlit-rail.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
napi_frags_finish() and napi_skb_finish() can only be called inside
NAPI Rx context, so we can feed NAPI cache with skbuff_heads that
got NAPI_MERGED_FREE verdict instead of immediate freeing.
Replace __kfree_skb() with __kfree_skb_defer() in napi_skb_finish()
and move napi_skb_free_stolen_head() to skbuff.c, so it can drop skbs
to NAPI cache.
As many drivers call napi_alloc_skb()/napi_get_frags() on their
receive path, this becomes especially useful.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
{,__}napi_alloc_skb() is mostly used either for optional non-linear
receive methods (usually controlled via Ethtool private flags and off
by default) and/or for Rx copybreaks.
Use __napi_build_skb() here for obtaining skbuff_heads from NAPI cache
instead of inplace allocations. This includes both kmalloc and page
frag paths.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reuse the old and forgotten SKB_ALLOC_NAPI to add an option to get
an skbuff_head from the NAPI cache instead of inplace allocation
inside __alloc_skb().
This implies that the function is called from softirq or BH-off
context, not for allocating a clone or from a distant node.
Cc: Alexander Duyck <alexander.duyck@gmail.com> # Simplified flags check
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of just bulk-flushing skbuff_heads queued up through
napi_consume_skb() or __kfree_skb_defer(), try to reuse them
on allocation path.
If the cache is empty on allocation, bulk-allocate the first
16 elements, which is more efficient than per-skb allocation.
If the cache is full on freeing, bulk-wipe the second half of
the cache (32 elements).
This also includes custom KASAN poisoning/unpoisoning to be
double sure there are no use-after-free cases.
To not change current behaviour, introduce a new function,
napi_build_skb(), to optionally use a new approach later
in drivers.
Note on selected bulk size, 16:
- this equals to XDP_BULK_QUEUE_SIZE, DEV_MAP_BULK_SIZE
and especially VETH_XDP_BATCH, which is also used to
bulk-allocate skbuff_heads and was tested on powerful
setups;
- this also showed the best performance in the actual
test series (from the array of {8, 16, 32}).
Suggested-by: Edward Cree <ecree.xilinx@gmail.com> # Divide on two halves
Suggested-by: Eric Dumazet <edumazet@google.com> # KASAN poisoning
Cc: Dmitry Vyukov <dvyukov@google.com> # Help with KASAN
Cc: Paolo Abeni <pabeni@redhat.com> # Reduced batch size
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
NAPI cache structures will be used for allocating skbuff_heads,
so move their declarations a bit upper.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
This function isn't much needed as NAPI skb queue gets bulk-freed
anyway when there's no more room, and even may reduce the efficiency
of bulk operations.
It will be even less needed after reusing skb cache on allocation path,
so remove it and this way lighten network softirqs a bit.
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Just call __build_skb_around() instead of open-coding it.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use unlikely() annotations for skbuff_head and data similarly to the
two other allocation functions and remove totally redundant goto.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
__build_skb_around() can never fail and always returns passed skb.
Make it return void to simplify and optimize the code.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eversince the introduction of __kmalloc_reserve(), "ip" argument
hasn't been used. _RET_IP_ is embedded inside
kmalloc_node_track_caller().
Remove the redundant macro and rename the function after it.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation before reusing several functions in all three skb
allocation variants, move __alloc_skb() next to the
__netdev_alloc_skb() and __napi_alloc_skb().
No functional changes.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the current implementation of {netdev,napi}_alloc_frag(), it doesn't
have any align guarantee for the returned buffer address, But for some
hardwares they do require the DMA buffer to be aligned correctly,
so we would have to use some workarounds like below if the buffers
allocated by the {netdev,napi}_alloc_frag() are used by these hardwares
for DMA.
buf = napi_alloc_frag(really_needed_size + align);
buf = PTR_ALIGN(buf, align);
These codes seems ugly and would waste a lot of memories if the buffers
are used in a network driver for the TX/RX. We have added the align
support for the page_frag functions, so add the corresponding
{netdev,napi}_frag functions.
Signed-off-by: Kevin Hao <haokexin@gmail.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Avoid the assumption that ksize(kmalloc(S)) == ksize(kmalloc(S)): when
cloning an skb, save and restore truesize after pskb_expand_head(). This
can occur if the allocator decides to service an allocation of the same
size differently (e.g. use a different size class, or pass the
allocation on to KFENCE).
Because truesize is used for bookkeeping (such as sk_wmem_queued), a
modified truesize of a cloned skb may result in corrupt bookkeeping and
relevant warnings (such as in sk_stream_kill_queues()).
Link: https://lkml.kernel.org/r/X9JR/J6dMMOy1obu@elver.google.com
Reported-by: syzbot+7b99aafdcc2eedea6178@syzkaller.appspotmail.com
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20210201160420.2826895-1-elver@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch adds TCP_NLA_TTL to SCM_TIMESTAMPING_OPT_STATS that exports
the time-to-live or hop limit of the latest incoming packet with
SCM_TSTAMP_ACK. The value exported may not be from the packet that acks
the sequence when incoming packets are aggregated. Exporting the
time-to-live or hop limit value of incoming packets helps to estimate
the hop count of the path of the flow that may change over time.
Signed-off-by: Yousuk Seung <ysseung@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Link: https://lore.kernel.org/r/20210120204155.552275-1-ysseung@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Conflicts:
drivers/net/can/dev.c
commit 03f16c5075 ("can: dev: can_restart: fix use after free bug")
commit 3e77f70e73 ("can: dev: move driver related infrastructure into separate subdir")
Code move.
drivers/net/dsa/b53/b53_common.c
commit 8e4052c32d ("net: dsa: b53: fix an off by one in checking "vlan->vid"")
commit b7a9e0da2d ("net: switchdev: remove vid_begin -> vid_end range from VLAN objects")
Field rename.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The commit dbd50f238d ("net: move the hsize check to the else
block in skb_segment") introduced a data corruption for devices
supporting scatter-gather.
The problem boils down to signed/unsigned comparison given
unexpected results: if signed 'hsize' is negative, it will be
considered greater than a positive 'len', which is unsigned.
This commit addresses resorting to the old checks order, so that
'hsize' never has a negative value when compared with 'len'.
v1 -> v2:
- reorder hsize checks instead of explicit cast (Alex)
Bisected-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Fixes: dbd50f238d ("net: move the hsize check to the else block in skb_segment")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Link: https://lore.kernel.org/r/861947c2d2d087db82af93c21920ce8147d15490.1611074818.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After commit 89319d3801 ("net: Add frag_list support to skb_segment"),
it goes to process frag_list when !hsize in skb_segment(). However, when
using skb frag_list, sg normally should not be set. In this case, hsize
will be set with len right before !hsize check, then it won't go to
frag_list processing code.
So the right thing to do is move the hsize check to the else block, so
that it won't affect the !hsize check for frag_list processing.
v1->v2:
- change to do "hsize <= 0" check instead of "!hsize", and also move
"hsize < 0" into else block, to save some cycles, as Alex suggested.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 3226b158e6 ("net: avoid 32 x truesize under-estimation for
tiny skbs") ensured that skbs with data size lower than 1025 bytes
will be kmalloc'ed to avoid excessive page cache fragmentation and
memory consumption.
However, the fix adressed only __napi_alloc_skb() (primarily for
virtio_net and napi_get_frags()), but the issue can still be achieved
through __netdev_alloc_skb(), which is still used by several drivers.
Drivers often allocate a tiny skb for headers and place the rest of
the frame to frags (so-called copybreak).
Mirror the condition to __netdev_alloc_skb() to handle this case too.
Since v1 [0]:
- fix "Fixes:" tag;
- refine commit message (mention copybreak usecase).
[0] https://lore.kernel.org/netdev/20210114235423.232737-1-alobakin@pm.me
Fixes: a1c7fff7e1 ("net: netdev_alloc_skb() use build_skb()")
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Link: https://lore.kernel.org/r/20210115150354.85967-1-alobakin@pm.me
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Both virtio net and napi_get_frags() allocate skbs
with a very small skb->head
While using page fragments instead of a kmalloc backed skb->head might give
a small performance improvement in some cases, there is a huge risk of
under estimating memory usage.
For both GOOD_COPY_LEN and GRO_MAX_HEAD, we can fit at least 32 allocations
per page (order-3 page in x86), or even 64 on PowerPC
We have been tracking OOM issues on GKE hosts hitting tcp_mem limits
but consuming far more memory for TCP buffers than instructed in tcp_mem[2]
Even if we force napi_alloc_skb() to only use order-0 pages, the issue
would still be there on arches with PAGE_SIZE >= 32768
This patch makes sure that small skb head are kmalloc backed, so that
other objects in the slab page can be reused instead of being held as long
as skbs are sitting in socket queues.
Note that we might in the future use the sk_buff napi cache,
instead of going through a more expensive __alloc_skb()
Another idea would be to use separate page sizes depending
on the allocated length (to never have more than 4 frags per page)
I would like to thank Greg Thelen for his precious help on this matter,
analysing crash dumps is always a time consuming task.
Fixes: fd11a83dd3 ("net: Pull out core bits of __netdev_alloc_skb and add __napi_alloc_skb")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/r/20210113161819.1155526-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
skb_seq_read iterates over an skb, returning pointer and length of
the next data range with each call.
It relies on kmap_atomic to access highmem pages when needed.
An skb frag may be backed by a compound page, but kmap_atomic maps
only a single page. There are not enough kmap slots to always map all
pages concurrently.
Instead, if kmap_atomic is needed, iterate over each page.
As this increases the number of calls, avoid this unless needed.
The necessary condition is captured in skb_frag_must_loop.
I tried to make the change as obvious as possible. It should be easy
to verify that nothing changes if skb_frag_must_loop returns false.
Tested:
On an x86 platform with
CONFIG_HIGHMEM=y
CONFIG_DEBUG_KMAP_LOCAL_FORCE_MAP=y
CONFIG_NETFILTER_XT_MATCH_STRING=y
Run
ip link set dev lo mtu 1500
iptables -A OUTPUT -m string --string 'badstring' -algo bm -j ACCEPT
dd if=/dev/urandom of=in bs=1M count=20
nc -l -p 8000 > /dev/null &
nc -w 1 -q 0 localhost 8000 < in
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
skbs in fraglist could be shared by a BPF filter loaded at TC. If TC
writes, it will call skb_ensure_writable -> pskb_expand_head to create
a private linear section for the head_skb. And then call
skb_clone_fraglist -> skb_get on each skb in the fraglist.
skb_segment_list overwrites part of the skb linear section of each
fragment itself. Even after skb_clone, the frag_skbs share their
linear section with their clone in PF_PACKET.
Both sk_receive_queue of PF_PACKET and PF_INET (or PF_INET6) can have
a link for the same frag_skbs chain. If a new skb (not frags) is
queued to one of the sk_receive_queue, multiple ptypes can see and
release this. It causes use-after-free.
[ 4443.426215] ------------[ cut here ]------------
[ 4443.426222] refcount_t: underflow; use-after-free.
[ 4443.426291] WARNING: CPU: 7 PID: 28161 at lib/refcount.c:190
refcount_dec_and_test_checked+0xa4/0xc8
[ 4443.426726] pstate: 60400005 (nZCv daif +PAN -UAO)
[ 4443.426732] pc : refcount_dec_and_test_checked+0xa4/0xc8
[ 4443.426737] lr : refcount_dec_and_test_checked+0xa0/0xc8
[ 4443.426808] Call trace:
[ 4443.426813] refcount_dec_and_test_checked+0xa4/0xc8
[ 4443.426823] skb_release_data+0x144/0x264
[ 4443.426828] kfree_skb+0x58/0xc4
[ 4443.426832] skb_queue_purge+0x64/0x9c
[ 4443.426844] packet_set_ring+0x5f0/0x820
[ 4443.426849] packet_setsockopt+0x5a4/0xcd0
[ 4443.426853] __sys_setsockopt+0x188/0x278
[ 4443.426858] __arm64_sys_setsockopt+0x28/0x38
[ 4443.426869] el0_svc_common+0xf0/0x1d0
[ 4443.426873] el0_svc_handler+0x74/0x98
[ 4443.426880] el0_svc+0x8/0xc
Fixes: 3a1296a38d (net: Support GRO/GSO fraglist chaining.)
Signed-off-by: Dongseok Yi <dseok.yi@samsung.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/r/1610072918-174177-1-git-send-email-dseok.yi@samsung.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Unlike the rest of the skb_zcopy_ functions, these routines
operate on a 'struct ubuf', not a skb. Remove the 'skb_'
prefix from the naming to make things clearer.
Suggested-by: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, when an ubuf is attached to a new skb, the shared
flags word is initialized to a fixed value. Instead of doing
this, set the default flags in the ubuf, and have new skbs
inherit from this default.
This is needed when setting up different zerocopy types.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In preparation for expanded zerocopy (TX and RX), move
the zerocopy related bits out of tx_flags into their own
flag word.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
At Willem's suggestion, rename the sock_zerocopy_* functions
so that they match the MSG_ZEROCOPY flag, which makes it clear
they are specific to this zerocopy implementation.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
RX zerocopy fragment pages which are not allocated from the
system page pool require special handling. Give the callback
in skb_zcopy_clear() a chance to process them first.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The sock_zerocopy_put_abort function contains logic which is
specific to the current zerocopy implementation. Add a wrapper
which checks the callback and dispatches apppropriately.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add an optional skb parameter to the zerocopy callback parameter,
which is passed down from skb_zcopy_clear(). This gives access
to the original skb, which is needed for upcoming RX zero-copy
error handling.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Replace sock_zerocopy_put with the generic skb_zcopy_put()
function. Pass 'true' as the success argument, as this
is identical to no change.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Before this change, the caller of sock_zerocopy_callback would
need to save the zerocopy status, decrement and check the refcount,
and then call the callback function - the callback was only invoked
when the refcount reached zero.
Now, the caller just passes the status into the callback function,
which saves the status and handles its own refcounts.
This makes the behavior of the sock_zerocopy_callback identical
to the tpacket and vhost callbacks.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
All 'struct ubuf_info' users should have a callback defined
as of commit 0a4a060bb2 ("sock: fix zerocopy_success regression
with msg_zerocopy").
Remove the dead code path to consume_skb(), which makes
assumptions about how the structure was allocated.
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
syzbot reproduces BUG_ON in skb_checksum_help():
tun creates (bogus) skb with huge partial-checksummed area and
small ip packet inside. Then ip_rcv trims the skb based on size
of internal ip packet, after that csum offset points beyond of
trimmed skb. Then checksum_tg() called via netfilter hook
triggers BUG_ON:
offset = skb_checksum_start_offset(skb);
BUG_ON(offset >= skb_headlen(skb));
To work around the problem this patch forces pskb_trim_rcsum_slow()
to return -EINVAL in described scenario. It allows its callers to
drop such kind of packets.
Link: https://syzkaller.appspot.com/bug?id=b419a5ca95062664fe1a60b764621eb4526e2cd0
Reported-by: syzbot+7010af67ced6105e5ab6@syzkaller.appspotmail.com
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/1b2494af-2c56-8ee2-7bc0-923fcad1cdf8@virtuozzo.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
skb_mpls_dec_ttl() reads the LSE without ensuring that it is contained in
the skb "linear" area. Fix this calling pskb_may_pull() before reading the
current ttl.
Found by code inspection.
Fixes: 2a2ea50870 ("net: sched: add mpls manipulation actions to TC")
Reported-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Link: https://lore.kernel.org/r/53659f28be8bc336c113b5254dc637cc76bbae91.1606987074.git.dcaratti@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
It turns out that usage of skb extensions can cause memory leaks. Ido
Schimmel reported: "[...] there are instances that blindly overwrite
'skb->extensions' by invoking skb_copy_header() after __alloc_skb()."
Therefore, give up on using skb extensions for KCOV handle, and instead
directly store kcov_handle in sk_buff.
Fixes: 6370cc3bbd ("net: add kcov handle to skb extensions")
Fixes: 85ce50d337 ("net: kcov: don't select SKB_EXTENSIONS when there is no NET")
Fixes: 97f53a08cb ("net: linux/skbuff.h: combine SKB_EXTENSIONS + KCOV handling")
Link: https://lore.kernel.org/linux-wireless/20201121160941.GA485907@shredder.lan/
Reported-by: Ido Schimmel <idosch@idosch.org>
Signed-off-by: Marco Elver <elver@google.com>
Link: https://lore.kernel.org/r/20201125224840.2014773-1-elver@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Trivial conflict in CAN, keep the net-next + the byteswap wrapper.
Conflicts:
drivers/net/can/usb/gs_usb.c
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When setting sk_err, set it to ee_errno, not ee_origin.
Commit f5f99309fa ("sock: do not set sk_err in
sock_dequeue_err_skb") disabled updating sk_err on errq dequeue,
which is correct for most error types (origins):
- sk->sk_err = err;
Commit 38b257938a ("sock: reset sk_err when the error queue is
empty") reenabled the behavior for IMCP origins, which do require it:
+ if (icmp_next)
+ sk->sk_err = SKB_EXT_ERR(skb_next)->ee.ee_origin;
But read from ee_errno.
Fixes: 38b257938a ("sock: reset sk_err when the error queue is empty")
Reported-by: Ayush Ranjan <ayushranjan@google.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Link: https://lore.kernel.org/r/20201126151220.2819322-1-willemdebruijn.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use napi_consume_skb() to assert the case when it is not called
in a atomic softirq context.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some identifiers have different names between their prototypes
and the kernel-doc markup.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Reviewed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
