MCS/CLH algorithm performance collapsed from 1 core to 2 cores under X86, worse than spinlock

[killpigger]

OS: Centos Linux performance2 3.10.0-1062.el7.x86_64 #1 SMP Wed Aug 7 18:08:02 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
CPU: Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz

1. All tests using the same args with: -l 1 -d 2000 -w 1 -c 1 -p 0 except -n, I have test -n with 1,2,4,8,16, the results are as fllows:
2. The results come from binary stress_test_clhs/spinlocks/tickets/ttass/mcss/hclhs
3. My question : Most articles said that the clh/mcs are more scalable than spinlock, but as the results show bellow，their performance are all collapsed, the clh's speed from 58M/s decrease to 5M/s rapidly, but I just increase the test threads from 1 to 2, even worse than spinlock, I want know what happens

--------------------------------------------------------Performance results-------------------------------------------------------------
1 thread:
spinlocks :#acquires : 132742143 ( 66371071 / s)
tickets :#acquires : 110548289 ( 55274144 / s)
ttass :#acquires : 131053147 ( 65526573 / s)
mcss :#acquires : 88319239 ( 44159619 / s)
clhs :#acquires : 117235434 ( 58588422 / s) /* Only 1 core, the mcss and clhs performance seems OK comparing with spinlocks */
hclhs :#acquires : 36983691 ( 18491845 / s)

2 threads:
spinlocks :#acquires : 23118253 ( 11559126 / s)
tickets :#acquires : 17053137 ( 8526568 / s)
ttass :#acquires : 71154038 ( 35577019 / s)
mcss :#acquires : 8048977 ( 4024488 / s)
clhs :#acquires : 10789801 ( 5394900 / s) /* Only 2 cores, the mcss and clhs performance had collapsed comparing with spinlocks */
hclhs :#acquires : 8898379 ( 4449189 / s)

4 threads:
spinlocks :#acquires : 13412203 ( 6706101 / s)
tickets :#acquires : 9639068 ( 4819534 / s)
ttass :#acquires : 35791617 ( 17895808 / s)
mcss :#acquires : 9743361 ( 4871680 / s)
clhs :#acquires : 10513970 ( 5256985 / s) /* Still slower than spinlocks, but the performance is stable */
hclhs :#acquires : 9422532 ( 4711266 / s)

8 threads:
spinlocks :#acquires : 6723061 ( 3361530 / s)
tickets :#acquires : 9840261 ( 4920130 / s)
ttass :#acquires : 12843745 ( 6418663 / s)
mcss :#acquires : 10730546 ( 5365273 / s)
clhs :#acquires : 10306623 ( 5153311 / s)
hclhs :#acquires : 10541255 ( 5270627 / s)

16 threads:
spinlocks :#acquires : 3204828 ( 1602414 / s)
tickets :#acquires : 5435362 ( 2717681 / s)
ttass :#acquires : 3680548 ( 1840274 / s)
mcss :#acquires : 7855023 ( 3927511 / s)
clhs :#acquires : 8642363 ( 4321181 / s)
hclhs :#acquires : 6773022 ( 3386511 / s)