[P2P] EFA support

[YangZhou1997]

Description

Please include a summary of the changes and the related issue.

Fixes # (issue)

Type of Change

• Bug fix
• New feature
• Documentation update

How Has This Been Tested?

Include any tests here.

• Unit tests
• Integration tests
• Manual testing

Checklist

• My code follows the style guidelines, e.g. format.sh.
• I have run build_and_install.sh to verify compilation.
• I have removed redundant variables and comments.
• I have updated the documentation.
• I have added tests.

[xuzhiying9510]

• benchmark_uccl_write.py

[Client]    256 B :   0.00 Gbps |   0.00 GB/s | 0.002518 s
[Client]   1.0 KB :   0.01 Gbps |   0.00 GB/s | 0.001270 s
[Client]   4.0 KB :   0.03 Gbps |   0.00 GB/s | 0.000960 s
[Client]  16.0 KB :   0.14 Gbps |   0.02 GB/s | 0.000964 s
[Client]  64.0 KB :   0.54 Gbps |   0.07 GB/s | 0.000970 s
[Client] 256.0 KB :   5.86 Gbps |   0.73 GB/s | 0.000358 s
[Client]   1.0 MB :  97.12 Gbps |  12.14 GB/s | 0.000086 s
[Client]  10.0 MB : 247.22 Gbps |  30.90 GB/s | 0.000339 s
[Client]  64.0 MB : 327.58 Gbps |  40.95 GB/s | 0.001639 s
[Client] 100.0 MB : 352.25 Gbps |  44.03 GB/s | 0.002381 s

• benchmark_uccl_read.py

[Client]    256 B :   0.00 Gbps |   0.00 GB/s | 0.006268 s
[Client]   1.0 KB :   0.00 Gbps |   0.00 GB/s | 0.006257 s
[Client]   4.0 KB :   0.01 Gbps |   0.00 GB/s | 0.003152 s
[Client]  16.0 KB :   3.23 Gbps |   0.40 GB/s | 0.000041 s
[Client]  64.0 KB :  11.92 Gbps |   1.49 GB/s | 0.000044 s
[Client] 256.0 KB :  42.60 Gbps |   5.33 GB/s | 0.000049 s
[Client]   1.0 MB : 101.01 Gbps |  12.63 GB/s | 0.000083 s
[Client]  10.0 MB : 312.85 Gbps |  39.11 GB/s | 0.000268 s
[Client]  64.0 MB : 366.40 Gbps |  45.80 GB/s | 0.001465 s
[Client] 100.0 MB : 369.11 Gbps |  46.14 GB/s | 0.002273 s

[YangZhou1997]

@xuzhiying9510 will change the implementation to use function pointers, which get set up one time during Endpoint creation time.