go
/
gvisor
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
4,045 Commits 2 Branches 96 Tags 81 MiB
Commit Graph

3 Commits

Author SHA1 Message Date
Bhasker Hariharan 6b43cf791a Replace calculateChecksum w/ the unrolled version. 
Fixes #1656

PiperOrigin-RevId: 291703760
 2020-01-27 05:34:34 -08:00
Bhasker Hariharan 68514d4ba3 Unroll checksum computation loop. 
Checksum computation is one of the most expensive bits of
packet processing. Manual unrolling of the loop provides
significant improvement in checksum speed.

Updates #1656

BenchmarkChecksum/checksum_64-12                49834124                23.6 ns/op
BenchmarkChecksum/checksum_128-12               27111997                44.1 ns/op
BenchmarkChecksum/checksum_256-12               11416683                91.5 ns/op
BenchmarkChecksum/checksum_512-12                6375298               174 ns/op
BenchmarkChecksum/checksum_1024-12               3403852               338 ns/op
BenchmarkChecksum/checksum_1500-12               2343576               493 ns/op
BenchmarkChecksum/checksum_2048-12               1730521               656 ns/op
BenchmarkChecksum/checksum_4096-12                920469              1327 ns/op
BenchmarkChecksum/checksum_8192-12                445885              2637 ns/op
BenchmarkChecksum/checksum_16384-12               226342              5268 ns/op
BenchmarkChecksum/checksum_32767-12               114210             10503 ns/op
BenchmarkChecksum/checksum_32768-12                99138             10610 ns/op
BenchmarkChecksum/checksum_65535-12                53438             21158 ns/op
BenchmarkChecksum/checksum_65536-12                52993             21067 ns/op
BenchmarkUnrolledChecksum/checksum_64-12        61035639                19.1 ns/op
BenchmarkUnrolledChecksum/checksum_128-12               36067015                33.6 ns/op
BenchmarkUnrolledChecksum/checksum_256-12               19731220                60.4 ns/op
BenchmarkUnrolledChecksum/checksum_512-12                9091291               116 ns/op
BenchmarkUnrolledChecksum/checksum_1024-12               4976406               226 ns/op
BenchmarkUnrolledChecksum/checksum_1500-12               3685224               328 ns/op
BenchmarkUnrolledChecksum/checksum_2048-12               2579108               447 ns/op
BenchmarkUnrolledChecksum/checksum_4096-12               1350475               887 ns/op
BenchmarkUnrolledChecksum/checksum_8192-12                658248              1780 ns/op
BenchmarkUnrolledChecksum/checksum_16384-12               335869              3534 ns/op
BenchmarkUnrolledChecksum/checksum_32767-12               168650              7095 ns/op
BenchmarkUnrolledChecksum/checksum_32768-12               168075              7098 ns/op
BenchmarkUnrolledChecksum/checksum_65535-12                75085             14277 ns/op
BenchmarkUnrolledChecksum/checksum_65536-12                75921             14127 ns/op

PiperOrigin-RevId: 291643290
 2020-01-26 18:35:01 -08:00
Andrei Vagin 8720bd643e netstack/tcp: software segmentation offload 
Right now, we send each tcp packet separately, we call one system
call per-packet. This patch allows to generate multiple tcp packets
and send them by sendmmsg.

The arguable part of this CL is a way how to handle multiple headers.
This CL adds the next field to the Prepandable buffer.

Nginx test results:

Server Software:        nginx/1.15.9
Server Hostname:        10.138.0.2
Server Port:            8080

Document Path:          /10m.txt
Document Length:        10485760 bytes

w/o gso:
Concurrency Level:      5
Time taken for tests:   5.491 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      1048600200 bytes
HTML transferred:       1048576000 bytes
Requests per second:    18.21 [#/sec] (mean)
Time per request:       274.525 [ms] (mean)
Time per request:       54.905 [ms] (mean, across all concurrent requests)
Transfer rate:          186508.03 [Kbytes/sec] received

sw-gso:

Concurrency Level:      5
Time taken for tests:   3.852 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      1048600200 bytes
HTML transferred:       1048576000 bytes
Requests per second:    25.96 [#/sec] (mean)
Time per request:       192.576 [ms] (mean)
Time per request:       38.515 [ms] (mean, across all concurrent requests)
Transfer rate:          265874.92 [Kbytes/sec] received

w/o gso:
$ ./tcp_benchmark --client --duration 15  --ideal
[SUM]  0.0-15.1 sec  2.20 GBytes  1.25 Gbits/sec

software gso:
$ tcp_benchmark --client --duration 15  --ideal --gso $((1<<16)) --swgso
[SUM]  0.0-15.1 sec  3.99 GBytes  2.26 Gbits/sec

PiperOrigin-RevId: 276112677
 2019-10-22 11:55:56 -07:00