Alexei Starovoitov says:
====================
pull-request: bpf-next 2023-12-18
This PR is larger than usual and contains changes in various parts
of the kernel.
The main changes are:
1) Fix kCFI bugs in BPF, from Peter Zijlstra.
End result: all forms of indirect calls from BPF into kernel
and from kernel into BPF work with CFI enabled. This allows BPF
to work with CONFIG_FINEIBT=y.
2) Introduce BPF token object, from Andrii Nakryiko.
It adds an ability to delegate a subset of BPF features from privileged
daemon (e.g., systemd) through special mount options for userns-bound
BPF FS to a trusted unprivileged application. The design accommodates
suggestions from Christian Brauner and Paul Moore.
Example:
$ sudo mkdir -p /sys/fs/bpf/token
$ sudo mount -t bpf bpffs /sys/fs/bpf/token \
-o delegate_cmds=prog_load:MAP_CREATE \
-o delegate_progs=kprobe \
-o delegate_attachs=xdp
3) Various verifier improvements and fixes, from Andrii Nakryiko, Andrei Matei.
- Complete precision tracking support for register spills
- Fix verification of possibly-zero-sized stack accesses
- Fix access to uninit stack slots
- Track aligned STACK_ZERO cases as imprecise spilled registers.
It improves the verifier "instructions processed" metric from single
digit to 50-60% for some programs.
- Fix verifier retval logic
4) Support for VLAN tag in XDP hints, from Larysa Zaremba.
5) Allocate BPF trampoline via bpf_prog_pack mechanism, from Song Liu.
End result: better memory utilization and lower I$ miss for calls to BPF
via BPF trampoline.
6) Fix race between BPF prog accessing inner map and parallel delete,
from Hou Tao.
7) Add bpf_xdp_get_xfrm_state() kfunc, from Daniel Xu.
It allows BPF interact with IPSEC infra. The intent is to support
software RSS (via XDP) for the upcoming ipsec pcpu work.
Experiments on AWS demonstrate single tunnel pcpu ipsec reaching
line rate on 100G ENA nics.
8) Expand bpf_cgrp_storage to support cgroup1 non-attach, from Yafang Shao.
9) BPF file verification via fsverity, from Song Liu.
It allows BPF progs get fsverity digest.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (164 commits)
bpf: Ensure precise is reset to false in __mark_reg_const_zero()
selftests/bpf: Add more uprobe multi fail tests
bpf: Fail uprobe multi link with negative offset
selftests/bpf: Test the release of map btf
s390/bpf: Fix indirect trampoline generation
selftests/bpf: Temporarily disable dummy_struct_ops test on s390
x86/cfi,bpf: Fix bpf_exception_cb() signature
bpf: Fix dtor CFI
cfi: Add CFI_NOSEAL()
x86/cfi,bpf: Fix bpf_struct_ops CFI
x86/cfi,bpf: Fix bpf_callback_t CFI
x86/cfi,bpf: Fix BPF JIT call
cfi: Flip headers
selftests/bpf: Add test for abnormal cnt during multi-kprobe attachment
selftests/bpf: Don't use libbpf_get_error() in kprobe_multi_test
selftests/bpf: Add test for abnormal cnt during multi-uprobe attachment
bpf: Limit the number of kprobes when attaching program to multiple kprobes
bpf: Limit the number of uprobes when attaching program to multiple uprobes
bpf: xdp: Register generic_kfunc_set with XDP programs
selftests/bpf: utilize string values for delegate_xxx mount options
...
====================
Link: https://lore.kernel.org/r/20231219000520.34178-1-alexei.starovoitov@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Kalle Valo says:
====================
wireless-next patches for v6.8
The second features pull request for v6.8. A bigger one this time with
changes both to stack and drivers. We have a new Wifi band RFI (WBRF)
mitigation feature for which we pulled an immutable branch shared with
other subsystems. And, as always, other new features and bug fixes all
over.
Major changes:
cfg80211/mac80211
* AMD ACPI based Wifi band RFI (WBRF) mitigation feature
* Basic Service Set (BSS) usage reporting
* TID to link mapping support
* mac80211 hardware flag to disallow puncturing
iwlwifi
* new debugfs file fw_dbg_clear
mt76
* NVMEM EEPROM improvements
* mt7996 Extremely High Throughpu (EHT) improvements
* mt7996 Wireless Ethernet Dispatcher (WED) support
* mt7996 36-bit DMA support
ath12k
* support one MSI vector
* WCN7850: support AP mode
* tag 'wireless-next-2023-12-18' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next: (207 commits)
wifi: mt76: mt7996: Use DECLARE_FLEX_ARRAY() and fix -Warray-bounds warnings
wifi: ath11k: workaround too long expansion sparse warnings
Revert "wifi: ath12k: use ATH12K_PCI_IRQ_DP_OFFSET for DP IRQ"
wifi: rt2x00: remove useless code in rt2x00queue_create_tx_descriptor()
wifi: rtw89: only reset BB/RF for existing WiFi 6 chips while starting up
wifi: rtw89: add DBCC H2C to notify firmware the status
wifi: rtw89: mac: add suffix _ax to MAC functions
wifi: rtw89: mac: add flags to check if CMAC and DMAC are enabled
wifi: rtw89: 8922a: add power on/off functions
wifi: rtw89: add XTAL SI for WiFi 7 chips
wifi: rtw89: phy: print out RFK log with formatted string
wifi: rtw89: parse and print out RFK log from C2H events
wifi: rtw89: add C2H event handlers of RFK log and report
wifi: rtw89: load RFK log format string from firmware file
wifi: rtw89: fw: add version field to BB MCU firmware element
wifi: rtw89: fw: load TX power track tables from fw_element
wifi: mwifiex: configure BSSID consistently when starting AP
wifi: mwifiex: add extra delay for firmware ready
wifi: mac80211: sta_info.c: fix sentence grammar
wifi: mac80211: rx.c: fix sentence grammar
...
====================
Link: https://lore.kernel.org/r/20231218163900.C031DC433C9@smtp.kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With the introduction of the rcu_replace_pointer_rtnl helper,
cleanup the rtnl_unregister_* functions to use the helper instead
of open coding it.
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to address the issues encountered with commit 1effe8ca4e
("skbuff: fix coalescing for page_pool fragment recycling"), the
combination of the following condition was excluded from skb coalescing:
from->pp_recycle = 1
from->cloned = 1
to->pp_recycle = 1
However, with page pool environments, the aforementioned combination can
be quite common(ex. NetworkMananger may lead to the additional
packet_type being registered, thus the cloning). In scenarios with a
higher number of small packets, it can significantly affect the success
rate of coalescing. For example, considering packets of 256 bytes size,
our comparison of coalescing success rate is as follows:
Without page pool: 70%
With page pool: 13%
Consequently, this has an impact on performance:
Without page pool: 2.57 Gbits/sec
With page pool: 2.26 Gbits/sec
Therefore, it seems worthwhile to optimize this scenario and enable
coalescing of this particular combination. To achieve this, we need to
ensure the correct increment of the "from" SKB page's page pool
reference count (pp_ref_count).
Following this optimization, the success rate of coalescing measured in
our environment has improved as follows:
With page pool: 60%
This success rate is approaching the rate achieved without using page
pool, and the performance has also been improved:
With page pool: 2.52 Gbits/sec
Below is the performance comparison for small packets before and after
this optimization. We observe no impact to packets larger than 4K.
packet size before after improved
(bytes) (Gbits/sec) (Gbits/sec)
128 1.19 1.27 7.13%
256 2.26 2.52 11.75%
512 4.13 4.81 16.50%
1024 6.17 6.73 9.05%
2048 14.54 15.47 6.45%
4096 25.44 27.87 9.52%
Signed-off-by: Liang Chen <liangchen.linux@gmail.com>
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Suggested-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Mina Almasry <almasrymina@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Up to now, we were only subtracting from the number of used page fragments
to figure out when a page could be freed or recycled. A following patch
introduces support for multiple users referencing the same fragment. So
reduce the initial page fragments value to half to avoid overflowing.
Signed-off-by: Liang Chen <liangchen.linux@gmail.com>
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Reviewed-by: Mina Almasry <almarsymina@google.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 1580ab63fc ("tcp/dccp: better use of ephemeral ports in connect()")
we added an heuristic to select even ports for connect() and odd ports for bind().
This was nice because no applications changes were needed.
But it added more costs when all even ports are in use,
when there are few listeners and many active connections.
Since then, IP_LOCAL_PORT_RANGE has been added to permit an application
to partition ephemeral port range at will.
This patch extends the idea so that if IP_LOCAL_PORT_RANGE is set on
a socket before accept(), port selection no longer favors even ports.
This means that connect() can find a suitable source port faster,
and applications can use a different split between connect() and bind()
users.
This should give more entropy to Toeplitz hash used in RSS: Using even
ports was wasting one bit from the 16bit sport.
A similar change can be done in inet_csk_find_open_port() if needed.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jakub Sitnicki <jakub@cloudflare.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Jason Xing <kerneljasonxing@gmail.com>
Link: https://lore.kernel.org/r/20231214192939.1962891-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ensure the various dtor functions match their prototype and retain
their CFI signatures, since they don't have their address taken, they
are prone to not getting CFI, making them impossible to call
indirectly.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20231215092707.799451071@infradead.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
BPF struct_ops uses __arch_prepare_bpf_trampoline() to write
trampolines for indirect function calls. These tramplines much have
matching CFI.
In order to obtain the correct CFI hash for the various methods, add a
matching structure that contains stub functions, the compiler will
generate correct CFI which we can pilfer for the trampolines.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20231215092707.566977112@infradead.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
We assume in handful of places that the name of the spec is
the same as the name of the family. We could fix that but
it seems like a fair assumption to make. Rename the MPTCP
spec instead.
Reviewed-by: Mat Martineau <martineau@kernel.org>
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
optmem_max being used in tx zerocopy,
we want to be able to control it on a netns basis.
Following patch changes two tests.
Tested:
oqq130:~# cat /proc/sys/net/core/optmem_max
131072
oqq130:~# echo 1000000 >/proc/sys/net/core/optmem_max
oqq130:~# cat /proc/sys/net/core/optmem_max
1000000
oqq130:~# unshare -n
oqq130:~# cat /proc/sys/net/core/optmem_max
131072
oqq130:~# exit
logout
oqq130:~# cat /proc/sys/net/core/optmem_max
1000000
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For many years, /proc/sys/net/core/optmem_max default value
on a 64bit kernel has been 20 KB.
Regular usage of TCP tx zerocopy needs a bit more.
Google has used 128KB as the default value for 7 years without
any problem.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Send credit update message when SO_RCVLOWAT is updated and it is bigger
than number of bytes in rx queue. It is needed, because 'poll()' will
wait until number of bytes in rx queue will be not smaller than
O_RCVLOWAT, so kick sender to send more data. Otherwise mutual hungup
for tx/rx is possible: sender waits for free space and receiver is
waiting data in 'poll()'.
Rename 'set_rcvlowat' callback to 'notify_set_rcvlowat' and set
'sk->sk_rcvlowat' only in one place (i.e. 'vsock_set_rcvlowat'), so the
transport doesn't need to do it.
Fixes: b89d882dc9 ("vsock/virtio: reduce credit update messages")
Signed-off-by: Arseniy Krasnov <avkrasnov@salutedevices.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add one more condition for sending credit update during dequeue from
stream socket: when number of bytes in the rx queue is smaller than
SO_RCVLOWAT value of the socket. This is actual for non-default value
of SO_RCVLOWAT (e.g. not 1) - idea is to "kick" peer to continue data
transmission, because we need at least SO_RCVLOWAT bytes in our rx
queue to wake up user for reading data (in corner case it is also
possible to stuck both tx and rx sides, this is why 'Fixes' is used).
Fixes: b89d882dc9 ("vsock/virtio: reduce credit update messages")
Signed-off-by: Arseniy Krasnov <avkrasnov@salutedevices.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to support IP_PKTINFO on those packets, we need to call
ipv4_pktinfo_prepare.
When sending mrouted/pimd daemons a cache report IGMP msg, it is
unnecessary to set dst on the newly created skb.
It used to be necessary on older versions until
commit d826eb14ec ("ipv4: PKTINFO doesnt need dst reference") which
changed the way IP_PKTINFO struct is been retrieved.
Changes from v1:
1. Undo changes in ipv4_pktinfo_prepare function. use it directly
and copy the control block.
Fixes: d826eb14ec ("ipv4: PKTINFO doesnt need dst reference")
Signed-off-by: Leone Fernando <leone4fernando@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit adds an unstable kfunc helper to access internal xfrm_state
associated with an SA. This is intended to be used for the upcoming
IPsec pcpu work to assign special pcpu SAs to a particular CPU. In other
words: for custom software RSS.
That being said, the function that this kfunc wraps is fairly generic
and used for a lot of xfrm tasks. I'm sure people will find uses
elsewhere over time.
This commit also adds a corresponding bpf_xdp_xfrm_state_release() kfunc
to release the refcnt acquired by bpf_xdp_get_xfrm_state(). The verifier
will require that all acquired xfrm_state's are released.
Co-developed-by: Antony Antony <antony.antony@secunet.com>
Signed-off-by: Antony Antony <antony.antony@secunet.com>
Acked-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
Link: https://lore.kernel.org/r/a29699c42f5fad456b875c98dd11c6afc3ffb707.1702593901.git.dxu@dxuuu.xyz
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
The build can become unreproducible if the list of files
found by $(wildcard ...) differs. Sort the list to avoid
this.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Because atalk_ioctl() accesses sk->sk_receive_queue
without holding a sk->sk_receive_queue.lock, it can
cause a race with atalk_recvmsg().
A use-after-free for skb occurs with the following flow.
```
atalk_ioctl() -> skb_peek()
atalk_recvmsg() -> skb_recv_datagram() -> skb_free_datagram()
```
Add sk->sk_receive_queue.lock to atalk_ioctl() to fix this issue.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Hyunwoo Kim <v4bel@theori.io>
Link: https://lore.kernel.org/r/20231213041056.GA519680@v4bel-B760M-AORUS-ELITE-AX
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Symmetric RSS hash functions are beneficial in applications that monitor
both Tx and Rx packets of the same flow (IDS, software firewalls, ..etc).
Getting all traffic of the same flow on the same RX queue results in
higher CPU cache efficiency.
A NIC that supports "symmetric-xor" can achieve this RSS hash symmetry
by XORing the source and destination fields and pass the values to the
RSS hash algorithm.
The user may request RSS hash symmetry for a specific algorithm, via:
# ethtool -X eth0 hfunc <hash_alg> symmetric-xor
or turn symmetry off (asymmetric) by:
# ethtool -X eth0 hfunc <hash_alg>
The specific fields for each flow type should then be specified as usual
via:
# ethtool -N|-U eth0 rx-flow-hash <flow_type> s|d|f|n
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Signed-off-by: Ahmed Zaki <ahmed.zaki@intel.com>
Link: https://lore.kernel.org/r/20231213003321.605376-4-ahmed.zaki@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The get/set_rxfh ethtool ops currently takes the rxfh (RSS) parameters
as direct function arguments. This will force us to change the API (and
all drivers' functions) every time some new parameters are added.
This is part 1/2 of the fix, as suggested in [1]:
- First simplify the code by always providing a pointer to all params
(indir, key and func); the fact that some of them may be NULL seems
like a weird historic thing or a premature optimization.
It will simplify the drivers if all pointers are always present.
- Then make the functions take a dev pointer, and a pointer to a
single struct wrapping all arguments. The set_* should also take
an extack.
Link: https://lore.kernel.org/netdev/20231121152906.2dd5f487@kernel.org/ [1]
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Suggested-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Ahmed Zaki <ahmed.zaki@intel.com>
Link: https://lore.kernel.org/r/20231213003321.605376-2-ahmed.zaki@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Releasing the DMA mapping will be useful for other types
of pages, so factor it out. Make sure compiler inlines it,
to avoid any regressions.
Signed-off-by: Mina Almasry <almasrymina@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Follow up commit 9690ae6042 ("ethtool: add header/data split
indication") and add the set part of Ethtool's header split, i.e.
ability to enable/disable header split via the Ethtool Netlink
interface. This might be helpful to optimize the setup for particular
workloads, for example, to avoid XDP frags, and so on.
A driver should advertise ``ETHTOOL_RING_USE_TCP_DATA_SPLIT`` in its
ops->supported_ring_params to allow doing that. "Unknown" passed from
the userspace when the header split is supported means the driver is
free to choose the preferred state.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Link: https://lore.kernel.org/r/20231212142752.935000-2-aleksander.lobakin@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tcf_idr_insert_many will replace the allocated -EBUSY pointer in
tcf_idr_check_alloc with the real action pointer, exposing it
to all operations. This operation is only needed when the action pointer
is created (ACT_P_CREATED). For actions which are bound to (returned 0),
the pointer already resides in the idr making such operation a nop.
Even though it's a nop, it's still not a cheap operation as internally
the idr code walks the idr and then does a replace on the appropriate slot.
So if the action was bound, better skip the idr replace entirely.
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Link: https://lore.kernel.org/r/20231211181807.96028-3-pctammela@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Instead of relying only on the idrinfo->lock mutex for
bind/alloc logic, rely on a combination of rcu + mutex + atomics
to better scale the case where multiple rtnl-less filters are
binding to the same action object.
Action binding happens when an action index is specified explicitly and
an action exists which such index exists. Example:
tc actions add action drop index 1
tc filter add ... matchall action drop index 1
tc filter add ... matchall action drop index 1
tc filter add ... matchall action drop index 1
tc filter ls ...
filter protocol all pref 49150 matchall chain 0 filter protocol all pref 49150 matchall chain 0 handle 0x1
not_in_hw
action order 1: gact action drop
random type none pass val 0
index 1 ref 4 bind 3
filter protocol all pref 49151 matchall chain 0 filter protocol all pref 49151 matchall chain 0 handle 0x1
not_in_hw
action order 1: gact action drop
random type none pass val 0
index 1 ref 4 bind 3
filter protocol all pref 49152 matchall chain 0 filter protocol all pref 49152 matchall chain 0 handle 0x1
not_in_hw
action order 1: gact action drop
random type none pass val 0
index 1 ref 4 bind 3
When no index is specified, as before, grab the mutex and allocate
in the idr the next available id. In this version, as opposed to before,
it's simplified to store the -EBUSY pointer instead of the previous
alloc + replace combination.
When an index is specified, rely on rcu to find if there's an object in
such index. If there's none, fallback to the above, serializing on the
mutex and reserving the specified id. If there's one, it can be an -EBUSY
pointer, in which case we just try again until it's an action, or an action.
Given the rcu guarantees, the action found could be dead and therefore
we need to bump the refcount if it's not 0, handling the case it's
in fact 0.
As bind and the action refcount are already atomics, these increments can
happen without the mutex protection while many tcf_idr_check_alloc race
to bind to the same action instance.
In case binding encounters a parallel delete or add, it will return
-EAGAIN in order to try again. Both filter and action apis already
have the retry machinery in-place. In case it's an unlocked filter it
retries under the rtnl lock.
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Link: https://lore.kernel.org/r/20231211181807.96028-2-pctammela@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Based on the tcp man page, if TCP_NODELAY is set, it disables Nagle's algorithm
and packets are sent as soon as possible. However in the `tcp_push` function
where autocorking is evaluated the `nonagle` value set by TCP_NODELAY is not
considered which can trigger unexpected corking of packets and induce delays.
For example, if two packets are generated as part of a server's reply, if the
first one is not transmitted on the wire quickly enough, the second packet can
trigger the autocorking in `tcp_push` and be delayed instead of sent as soon as
possible. It will either wait for additional packets to be coalesced or an ACK
from the client before transmitting the corked packet. This can interact badly
if the receiver has tcp delayed acks enabled, introducing 40ms extra delay in
completion times. It is not always possible to control who has delayed acks
set, but it is possible to adjust when and how autocorking is triggered.
Patch prevents autocorking if the TCP_NODELAY flag is set on the socket.
Patch has been tested using an AWS c7g.2xlarge instance with Ubuntu 22.04 and
Apache Tomcat 9.0.83 running the basic servlet below:
import java.io.IOException;
import java.io.OutputStreamWriter;
import java.io.PrintWriter;
import javax.servlet.ServletException;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
public class HelloWorldServlet extends HttpServlet {
@Override
protected void doGet(HttpServletRequest request, HttpServletResponse response)
throws ServletException, IOException {
response.setContentType("text/html;charset=utf-8");
OutputStreamWriter osw = new OutputStreamWriter(response.getOutputStream(),"UTF-8");
String s = "a".repeat(3096);
osw.write(s,0,s.length());
osw.flush();
}
}
Load was applied using wrk2 (https://github.com/kinvolk/wrk2) from an AWS
c6i.8xlarge instance. With the current auto-corking behavior and TCP_NODELAY
set an additional 40ms latency from P99.99+ values are observed. With the
patch applied we see no occurrences of 40ms latencies. The patch has also been
tested with iperf and uperf benchmarks and no regression was observed.
# No patch with tcp_autocorking=1 and TCP_NODELAY set on all sockets
./wrk -t32 -c128 -d40s --latency -R10000 http://172.31.49.177:8080/hello/hello'
...
50.000% 0.91ms
75.000% 1.12ms
90.000% 1.46ms
99.000% 1.73ms
99.900% 1.96ms
99.990% 43.62ms <<< 40+ ms extra latency
99.999% 48.32ms
100.000% 49.34ms
# With patch
./wrk -t32 -c128 -d40s --latency -R10000 http://172.31.49.177:8080/hello/hello'
...
50.000% 0.89ms
75.000% 1.13ms
90.000% 1.44ms
99.000% 1.67ms
99.900% 1.78ms
99.990% 2.27ms <<< no 40+ ms extra latency
99.999% 3.71ms
100.000% 4.57ms
Fixes: f54b311142 ("tcp: auto corking")
Signed-off-by: Salvatore Dipietro <dipiets@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
syzkaller report:
kernel BUG at net/core/skbuff.c:3452!
invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.7.0-rc4-00009-gbee0e7762ad2-dirty #135
RIP: 0010:skb_copy_and_csum_bits (net/core/skbuff.c:3452)
Call Trace:
icmp_glue_bits (net/ipv4/icmp.c:357)
__ip_append_data.isra.0 (net/ipv4/ip_output.c:1165)
ip_append_data (net/ipv4/ip_output.c:1362 net/ipv4/ip_output.c:1341)
icmp_push_reply (net/ipv4/icmp.c:370)
__icmp_send (./include/net/route.h:252 net/ipv4/icmp.c:772)
ip_fragment.constprop.0 (./include/linux/skbuff.h:1234 net/ipv4/ip_output.c:592 net/ipv4/ip_output.c:577)
__ip_finish_output (net/ipv4/ip_output.c:311 net/ipv4/ip_output.c:295)
ip_output (net/ipv4/ip_output.c:427)
__ip_queue_xmit (net/ipv4/ip_output.c:535)
__tcp_transmit_skb (net/ipv4/tcp_output.c:1462)
__tcp_retransmit_skb (net/ipv4/tcp_output.c:3387)
tcp_retransmit_skb (net/ipv4/tcp_output.c:3404)
tcp_retransmit_timer (net/ipv4/tcp_timer.c:604)
tcp_write_timer (./include/linux/spinlock.h:391 net/ipv4/tcp_timer.c:716)
The panic issue was trigered by tcp simultaneous initiation.
The initiation process is as follows:
TCP A TCP B
1. CLOSED CLOSED
2. SYN-SENT --> <SEQ=100><CTL=SYN> ...
3. SYN-RECEIVED <-- <SEQ=300><CTL=SYN> <-- SYN-SENT
4. ... <SEQ=100><CTL=SYN> --> SYN-RECEIVED
5. SYN-RECEIVED --> <SEQ=100><ACK=301><CTL=SYN,ACK> ...
// TCP B: not send challenge ack for ack limit or packet loss
// TCP A: close
tcp_close
tcp_send_fin
if (!tskb && tcp_under_memory_pressure(sk))
tskb = skb_rb_last(&sk->tcp_rtx_queue); //pick SYN_ACK packet
TCP_SKB_CB(tskb)->tcp_flags |= TCPHDR_FIN; // set FIN flag
6. FIN_WAIT_1 --> <SEQ=100><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ...
// TCP B: send challenge ack to SYN_FIN_ACK
7. ... <SEQ=301><ACK=101><CTL=ACK> <-- SYN-RECEIVED //challenge ack
// TCP A: <SND.UNA=101>
8. FIN_WAIT_1 --> <SEQ=101><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ... // retransmit panic
__tcp_retransmit_skb //skb->len=0
tcp_trim_head
len = tp->snd_una - TCP_SKB_CB(skb)->seq // len=101-100
__pskb_trim_head
skb->data_len -= len // skb->len=-1, wrap around
... ...
ip_fragment
icmp_glue_bits //BUG_ON
If we use tcp_trim_head() to remove acked SYN from packet that contains data
or other flags, skb->len will be incorrectly decremented. We can remove SYN
flag that has been acked from rtx_queue earlier than tcp_trim_head(), which
can fix the problem mentioned above.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Co-developed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com>
Link: https://lore.kernel.org/r/20231210020200.1539875-1-dongchenchen2@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Because rose_ioctl() accesses sk->sk_receive_queue
without holding a sk->sk_receive_queue.lock, it can
cause a race with rose_accept().
A use-after-free for skb occurs with the following flow.
```
rose_ioctl() -> skb_peek()
rose_accept() -> skb_dequeue() -> kfree_skb()
```
Add sk->sk_receive_queue.lock to rose_ioctl() to fix this issue.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Hyunwoo Kim <v4bel@theori.io>
Link: https://lore.kernel.org/r/20231209100538.GA407321@v4bel-B760M-AORUS-ELITE-AX
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Because do_vcc_ioctl() accesses sk->sk_receive_queue
without holding a sk->sk_receive_queue.lock, it can
cause a race with vcc_recvmsg().
A use-after-free for skb occurs with the following flow.
```
do_vcc_ioctl() -> skb_peek()
vcc_recvmsg() -> skb_recv_datagram() -> skb_free_datagram()
```
Add sk->sk_receive_queue.lock to do_vcc_ioctl() to fix this issue.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Hyunwoo Kim <v4bel@theori.io>
Link: https://lore.kernel.org/r/20231209094210.GA403126@v4bel-B760M-AORUS-ELITE-AX
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Sometimes there may be reasons for which a BSS that's
actually found in scan cannot be used to connect to,
for example a nonprimary link of an NSTR mobile AP MLD
cannot be used for normal direct connections to it.
Not indicating these to userspace as we do now of course
avoids being able to connect to them, but it's better if
they're shown to userspace and it can make an appropriate
decision, without e.g. doing an additional ML probe.
Thus add an indication of what a BSS can be used for,
currently "normal" and "MLD link", including a reason
bitmap for it being not usable.
The latter can be extended later for certain BSSes if there
are other reasons they cannot be used.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Reviewed-by: Ilan Peer <ilan.peer@intel.com>
Reviewed-by: Gregory Greenman <gregory.greenman@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://msgid.link/20231211085121.0464f25e0b1d.I9f70ca9f1440565ad9a5207d0f4d00a20cca67e7@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
