LGTM.

We could perhaps commit it without enabling it to begin with.

@SjoerdMeijer, I could confirm that D51160 with this scheduler model shows the improvement and it had been successfully applied to our in-house compiler.
Also, there was no issue about performance & regressions with this patch. So, I think that you can commit this with minor fixes @evgeny777 mentioned before.

Hi Javed.

When I tried to test this patch, it showed some improvements on several benchmarks like Spec2000/2006.
But other benchmarks like dhrystone, commercial benchmark's sub workloads show performance degradation.

I will update benchmark results until next week.
benchmark list : dhrystone, spec2000, spec2006, one commercial benchmark.

Addressed SjoerdMeijer' new comments. Thanks.

Thanks Benjamin.
Committed in r298895.

@evandro
This is code review about the case "Some targets using CSEL is more prefer than CSINV, CSINC.".

Change the feature name PreferCSEL to FastCSEL.

Addressing Renato, James's comments.
Thanks.

Addressing Benjamin's comment.
Thanks.

Thanks for your opinion & example Renato.

After internal(Samsung only) discussion, revert patch about Exynos-M3 part.

Thanks for comment Renato.

Addressing Matthias's comment.
Thanks.

Thanks. Hal, Tim.

OK Tim.

Or how about give up inserting prefetch intrinsic when the loop has inline asm?
I don't think users are stupid that they don't know prefetch is necessary or not when they use inline asm.

Thanks for the comment Tim.

ping2.

ping?

Is Diff 64111 too much modification?

Addressed Hal's comments.

Thanks for review Hal.!

LGTM.

Commited in r267502.

Hi Chad.

I think this optimization will affect all ARM Architectures. Is this optimization also good for cortex-a57?

Addressed Gerolf's comments.

sure, sorry I missed that. I looked at this too long, I guess :-). It is principally the same ‘better ILP' story as for integers. The prototypical idea is this: imagine two fmul operands feeding the fadd. When the two fmul can execute in parallel it can be faster to issue fmul, fmul, fadd rather than fmul, fmadd.

Hi Gerolf.

LGTM, too. with a few minor nits.

Hi Gerolf.

Evandro, we still wait your reply over 2weeks.
Could you please answer or review about our questions?

LGTM! too.

I also think this is reasonable patch.

I also think this is reasonable change.

Kindly ping ...

ping?

Rebase patch against latest trunk change & modiy comments.

Hi Evandro.

Thanks for review Chad.!

Hi Sanjay.
I already explained that we got improvement on commercial benchmark by e-mail.

Modify the patch which can consider uArch information.

Thanks Chad.!
Commited in r262580.