Use hardware-accelerated Adler32 computation

[bgrainger]

https://github.com/SixLabors/ImageSharp/blob/master/src/ImageSharp/Formats/Png/Zlib/Adler32.cs is an Apache 2.0-licensed file that implements a hardware-accelerated implementation of Adler32 in C#.

[novak-oleksandr]

Not that if I had doubts about benchmark results, but still:

BenchmarkDotNet=v0.12.1, OS=Windows 10.0.19041.804 (2004/?/20H1)
Intel Core i7-8850H CPU 2.60GHz (Coffee Lake), 1 CPU, 12 logical and 6 physical cores
.NET Core SDK=5.0.103
  [Host]     : .NET Core 5.0.3 (CoreCLR 5.0.321.7212, CoreFX 5.0.321.7212), X64 RyuJIT
  Job-QQDQQD : .NET Core 5.0.3 (CoreCLR 5.0.321.7212, CoreFX 5.0.321.7212), X64 RyuJIT

WarmupCount=10  

Method	data	Mean	Error	StdDev
B_AdlerNaive	Syste(...)Byte] [34]	288.89 ns	0.362 ns	0.302 ns
B_AdlerNaive	Syste(...)Byte] [34]	3,717.99 ns	9.278 ns	8.225 ns
B_AdlerNaive	Syste(...)Byte] [34]	3,852,480.23 ns	3,814.165 ns	3,185.000 ns
B_AdlerNaive	Syste(...)Byte] [34]	61,852,477.78 ns	116,195.446 ns	108,689.292 ns
B_AdlerSse	Syste(...)Byte] [34]	13.72 ns	0.052 ns	0.049 ns
B_AdlerSse	Syste(...)Byte] [34]	41.93 ns	0.835 ns	0.781 ns
B_AdlerSse	Syste(...)Byte] [34]	41,533.86 ns	111.976 ns	93.505 ns
B_AdlerSse	Syste(...)Byte] [34]	1,018,107.32 ns	5,609.412 ns	4,379.460 ns
B_AdlerScalar	Syste(...)Byte] [34]	30.99 ns	0.083 ns	0.069 ns
B_AdlerScalar	Syste(...)Byte] [34]	381.93 ns	0.870 ns	0.771 ns
B_AdlerScalar	Syste(...)Byte] [34]	390,690.20 ns	508.281 ns	450.578 ns
B_AdlerScalar	Syste(...)Byte] [34]	6,345,381.67 ns	56,812.787 ns	47,441.250 ns

array sizes were: 80 (min threshold to use compression), 1024, 1,048,576 and 16,777,215 (i.e max protocol packet size)

[bgrainger]

Added in 1.3.0.