Process "multi" strings in compiled regexes 4 or 2 chars at a time when possible

[stephentoub]

When we encounter a series of characters in a regex pattern, we emit a comparison check per character. This updates that codegen to compare 2 or 4 characters at a time, when possible. In general these sequences are not long, but they can easily be a handful of characters, and comparing with ints and longs instead of chars slightly improves both throughput and the size of the IL and JIT'd asm.

@jkotas, any concerns with the change (philosophically due to the Unsafe usage, or otherwise)? Note that I first used StartsWith, but the overhead from that was very pronounced, with the cross-over point not being until around 40 or so characters.

Trivial microbenchmark:

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Diagnosers;
using BenchmarkDotNet.Running;
using System.Text.RegularExpressions;

[MemoryDiagnoser]
public class Program
{
    static void Main(string[] args) => BenchmarkSwitcher.FromAssemblies(new[] { typeof(Program).Assembly }).Run(args);

    private readonly Regex _regex = new Regex(@"Oh what a beautiful morning.  Oh what a beautiful day.  I've got a beautiful feeling everything's going my way.", RegexOptions.Compiled);

    [Benchmark] public bool Match() =>   _regex.IsMatch("Oh what a beautiful morning.  Oh what a beautiful day.  I've got a beautiful feeling everything's going my way.");
}

Method	Toolchain	Mean	Error	StdDev	Ratio
Match	\master\corerun.exe	74.27 ns	0.155 ns	0.130 ns	1.00
Match	\pr\corerun.exe	61.73 ns	0.195 ns	0.182 ns	0.83

Contributes to #1349

[jkotas]

@jkotas, any concerns with the change (philosophically due to the Unsafe usage, or otherwise)?

It sounds reasonable to me. The high-performance parsers always end up using pattern like this. The convenience methods like StartsWith won't cut it.