JIT: Optimize bit-wise and operation and compare with same constant value

[varelen]

This pull request resolves #101000 and is based on the feedback that was given on the original PR #103868. Thanks for the previous work, it was a really great way to get started.

The code emitted for:

void Foo(int x)
{
    if ((x & 0xC000000) == 0xC000000)
        Console.WriteLine();
}

x64 (before):

and     edi, 0xC000000
cmp     edi, 0xC000000
jne     SHORT G_M000_IG04

x64 (after):

not      edi
and      edi, 0xC000000
jne      SHORT G_M000_IG04

ARM64 (before):

and	w0, w0, #0xC000000
mov	w1, #0xC000000
cmp	w0, w1
bne	G_M29168_IG04

ARM64 (after):

mvn	w0, w0
and	w0, w0, #0xC000000
cbnz	w0, G_M29168_IG04

[jakobbotsch]

Thanks for the contribution! I closed/reopened this to kick off a fresh run of the CI.

cc @dotnet/jit-contrib

[EgorBo]

can you leave a comment of what this transformation does?

[varelen]

I added a comment, hope it's okay.

[EgorBo]

/azp run Fuzzlyn

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[JulieLeeMSFT]

@EgorBo, please review this community PR again.

[JulieLeeMSFT]

@BruceForstall, please review this community PR.

[BruceForstall]

/azp run Fuzzlyn, runtime-coreclr outerloop

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[BruceForstall]

/ba-g unrelated failures