[SingleSource/Vectorizer] Add unit tests for FindLastIV pattern.

[Mel-Chen]

This patch adds runtime test case for vectorization of FindLastIV reduction idiom:

int32_t Rdx = -1;
for (int32_t I = 0; I < TC; I++) {
  Rdx = A[I] > B[I] ? I : Rdx;
}
return Rdx;

Improving test coverage for llvm/llvm-project#67812 and llvm/llvm-project#120395.

[alexey-bataev]

Looks good to me, but wait for couple days for others to review

[Mel-Chen]

@fhahn @Meinersbur Do you have any suggestions and comments?

[fhahn]

Same as for, do we need smaller types to trigger epilogue vectorization on more platforms?

IIUC llvm/llvm-project#120395 will need to be fixed first, otherwise the new tests will fail?

[Mel-Chen]

Same as for, do we need smaller types to trigger epilogue vectorization on more platforms?

Sure, could you take a look at AnyOf #195 to see if the approach is appropriate? I will modify this patch in the same way if you feel good.

IIUC llvm/llvm-project#120395 will need to be fixed first, otherwise the new tests will fail?

Yes, the current implementation still produces incorrect output in the following datasets after enabling epilogue vectorization.

    // Check with only Src1[N - 1] > Src2[N - 1].
    for (unsigned I = 0; I != N; ++I)
      Src1[I] = Src2[I] = std::numeric_limits<Ty>::min();
    Src1[N - 1] = std::numeric_limits<Ty>::max();

llvm/llvm-project#120395 can fix this error.

[Mel-Chen]

Add test for FFindLastIV, and s16 type for triggering epilogue vectorization easier.
Please take a look again, thanks.

[fhahn]

LGTM, thanks!

[AaronBallman]

FYI, but this test fails with GCC 15.x because they enabled -D_GLIBCXX_ASSERTIONS=1 libstdc++ by default. Here's a reproducer on godbolt: https://godbolt.org/z/qh4EGo3s7

We ran into this with our downstream at Intel because we also enable _GLIBCXX_ASSERTIONS, but upstream Clang does not currently do that nor does libc++ currently have the same assertions, so that's likely why we've not started hitting this failure in upstream's tree yet.