Adding Benchmark test for Checksums

[joviegas]

Motivation and Context

• Need to benchmark the different checksums

Modifications

• Added new Test cases

Testing

updateEntireByteArrayChecksum (Java8)

@Warmup(iterations = 3, time = 15, timeUnit = TimeUnit.SECONDS)
@Measurement(iterations = 5, time = 10, timeUnit = TimeUnit.SECONDS)
@Fork(2)

Checksum Provider	Size	Score (ns/op)	Error (ns/op)
CRC32	128B	54.359	0.149
CRC32C (Crt)	128B	168.957	12.198
CRC64NVME	128B	238.858	3.05
MD5	128B	656.694	9.134
SHA1	128B	948.87	6.687
SHA256	128B	1407.35	15.987
CRC32	1MB	63937.4	311.145
CRC32C	1MB	142089	5188.44
CRC64NVME	1MB	779647	6751.18
MD5	1MB	3.23534e+06	1510.63
SHA1	1MB	4.81596e+06	6461.77
SHA256	1MB	7.17205e+06	60232.7
CRC32	4KB	327.245	37.799
CRC32C	4KB	553.963	25.966
CRC64NVME	4KB	3191.97	27.827
MD5	4KB	13246.6	10.639
SHA1	4KB	19684.2	26.322
SHA256	4KB	28974.7	124.396

updateIndividualByteChecksumOneByteATime (Java8)

Checksum Provider	Size	Score (ns/op)	Error (ns/op)
CRC32	128B	503.246	0.383
CRC32C (CRT)	128B	19749.9	4416.36
CRC64NVME	128B	17260.7	78.228
MD5	128B	2821.84	97.619
SHA1	128B	3187.29	140.14
SHA256	128B	3642.74	172.055
CRC32	1MB	3.8796e+06	1601.41
CRC32C	1MB	1.35245e+08	695891
CRC64NVME	1MB	1.35547e+08	373876
MD5	1MB	1.80819e+07	16197.2
SHA1	1MB	1.97299e+07	30513.9
SHA256	1MB	2.21067e+07	119922
CRC32	4KB	15508.9	21.904
CRC32C	4KB	539955	7495.23
CRC64NVME	4KB	545065	4185.17
MD5	4KB	83687	190.44
SHA1	4KB	90607.2	463.367
SHA256	4KB	97773.3	9369.09

@Warmup(iterations = 3, time = 3, timeUnit = TimeUnit.SECONDS)
@Measurement(iterations = 5, time = 5, timeUnit = TimeUnit.SECONDS)
@Fork(1)

updateEntireByteArrayChecksum (Java8)

Checksum Provider	Size	Score (ns/op)	Error (ns/op)
CRC32	128B	108.077	0.748
CRC32C (CRT)	128B	161.075	0.321
CRC64NVME	128B	226.943	0.964
MD5	128B	659.158	2.358
SHA1	128B	954.494	18.984
SHA256	128B	1406	8.164
CRC32	1MB	64046.1	122.424
CRC32C	1MB	145312	799.353
CRC64NVME	1MB	776510	3767.16
MD5	1MB	3.234e+06	3911.2
SHA1	1MB	4.80487e+06	21984
SHA256	1MB	7.11034e+06	24152.7
CRC32	4KB	306.098	26.504
CRC32C	4KB	529.352	2.096
CRC64NVME	4KB	3189.51	60.239
MD5	4KB	13177.7	8.595
SHA1	4KB	19576.1	41.6
SHA256	4KB	28992.4	48.096

updateIndividualByteChecksumOneByteATime (Java8)

Checksum Provider	Size	Score (ns/op)	Error (ns/op)
CRC32	128B	503.304	0.966
CRC32C (CRT)	128B	16636.7	168.581
CRC64NVME	128B	16931.1	105.514
MD5	128B	2843.91	51.346
SHA1	128B	3255.72	234.994
SHA256	128B	3422.36	36.194
CRC32	1MB	3.88485e+06	3175.98
CRC32C	1MB	1.35303e+08	1.05541e+06
CRC64NVME	1MB	1.3671e+08	4.81374e+06
MD5	1MB	1.81059e+07	8601.73
SHA1	1MB	1.98254e+07	79692.5
SHA256	1MB	2.20228e+07	57341.4
CRC32	4KB	15571.7	8.456
CRC32C	4KB	521963	3788.89
CRC64NVME	4KB	539288	9694.32
MD5	4KB	84142.6	342.75
SHA1	4KB	83252.4	458.621
SHA256	4KB	93326.5	5752.15

License

• I confirm that this pull request can be released under the Apache 2 license

[zoewangg]

Can we do reset in tearDown methond @TearDown?

[joviegas]

Done

[zoewangg]

Is 3 seconds sufficient for warmup? Is the result consistent after warming up?

[joviegas]

I did test with both 3 and 15 , the relative Score is same .
However changed it to 15 to be consistent with other test

[sonarcloud]

Quality Gate failed

Failed conditions
0.0% Coverage on New Code (required ≥ 80%)

See analysis details on SonarQube Cloud