linux/include
Eric Dumazet 86c2bc293b tcp: use RCU lookup in __inet_hash_connect()
When __inet_hash_connect() has to try many 4-tuples before
finding an available one, we see a high spinlock cost from
the many spin_lock_bh(&head->lock) performed in its loop.

This patch adds an RCU lookup to avoid the spinlock cost.

check_established() gets a new @rcu_lookup argument.
First reason is to not make any changes while head->lock
is not held.
Second reason is to not make this RCU lookup a second time
after the spinlock has been acquired.

Tested:

Server:

ulimit -n 40000; neper/tcp_crr -T 200 -F 30000 -6 --nolog

Client:

ulimit -n 40000; neper/tcp_crr -T 200 -F 30000 -6 --nolog -c -H server

Before series:

  utime_start=0.288582
  utime_end=1.548707
  stime_start=20.637138
  stime_end=2002.489845
  num_transactions=484453
  latency_min=0.156279245
  latency_max=20.922042756
  latency_mean=1.546521274
  latency_stddev=3.936005194
  num_samples=312537
  throughput=47426.00

perf top on the client:

 49.54%  [kernel]       [k] _raw_spin_lock
 25.87%  [kernel]       [k] _raw_spin_lock_bh
  5.97%  [kernel]       [k] queued_spin_lock_slowpath
  5.67%  [kernel]       [k] __inet_hash_connect
  3.53%  [kernel]       [k] __inet6_check_established
  3.48%  [kernel]       [k] inet6_ehashfn
  0.64%  [kernel]       [k] rcu_all_qs

After this series:

  utime_start=0.271607
  utime_end=3.847111
  stime_start=18.407684
  stime_end=1997.485557
  num_transactions=1350742
  latency_min=0.014131929
  latency_max=17.895073144
  latency_mean=0.505675853  # Nice reduction of latency metrics
  latency_stddev=2.125164772
  num_samples=307884
  throughput=139866.80      # 190 % increase

perf top on client:

 56.86%  [kernel]       [k] __inet6_check_established
 17.96%  [kernel]       [k] __inet_hash_connect
 13.88%  [kernel]       [k] inet6_ehashfn
  2.52%  [kernel]       [k] rcu_all_qs
  2.01%  [kernel]       [k] __cond_resched
  0.41%  [kernel]       [k] _raw_spin_lock

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jason Xing <kerneljasonxing@gmail.com>
Tested-by: Jason Xing <kerneljasonxing@gmail.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250302124237.3913746-5-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-04 17:46:27 -08:00
..
acpi LoongArch changes for v6.14 2025-01-28 08:52:01 -08:00
asm-generic kbuild: keep symbols for symbol_get() even with CONFIG_TRIM_UNUSED_KSYMS 2025-02-06 01:08:58 +09:00
clocksource KVM/arm64 updates for 6.14 2025-01-28 09:01:36 -08:00
crypto
cxl
drm drm: Fix DSC BPP increment decoding 2025-02-13 10:20:30 +02:00
dt-bindings dt-bindings: clock: qcom: Add CAMCC clocks for QCS8300 2025-02-02 20:59:04 -06:00
hyperv
keys
kunit linux_kselftest-kunit-6.14-rc1 2025-01-22 12:32:39 -08:00
kvm
linux ppp: use IFF_NO_QUEUE in virtual interfaces 2025-03-04 17:11:17 -08:00
math-emu
media
memory
misc
net tcp: use RCU lookup in __inet_hash_connect() 2025-03-04 17:46:27 -08:00
pcmcia
ras
rdma
rv rv: Reset per-task monitors also for idle tasks 2025-01-23 12:16:04 -05:00
scsi Driver core and debugfs updates 2025-01-28 12:25:12 -08:00
soc soc: driver updates for 6.14 2025-01-24 14:56:59 -08:00
sound ASoC: cs35l56: Prevent races when soft-resetting using SPI control 2025-02-25 15:06:58 +00:00
target
trace Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2025-02-27 10:20:58 -08:00
uapi First 6.15 material: 2025-03-04 08:50:42 -08:00
ufs scsi: ufs: core: Fix use-after free in init error and remove paths 2025-02-03 17:20:01 -05:00
vdso
video
xen