This broke glamor with the radeonsi driver, hits unreachable():

Do you have a link the broken builder? I'm happy to take a look at fixing this.

Remove another check for MAP_FAILED.

This change does several things, and I think most of them need further thought. Can we discuss Monday?

Note that if you add a line like:

In D52300#1241754, @kbobyrev wrote:

I addressed the easiest issues. I'll try to implement separate storage structure for Heads and Payloads which would potentially make the implementation cleaner and easier to understand (and also more maintainable since that would be way easier to go for SIMD instructions speedups and other encoding schemes if we do that).

That doesn't sound more maintainable, that sounds like a performance hack that will hurt the layering.
Which is ok :-) but please don't do that until you measure a nontrivial performance improvement from it.

Update error reporting in one path.

Renaming them all will take a bit more work (for everyone) - I'm not sure it's especially urgent. Personally I'm happy as long as they all have underscores, and aren't wildly inconsistent.

@RKSimon should I make any other changes to this?

You can always write an unit-test, too.

Whoops, that acceptance was accidental. Pretending I want changes just to make it clear in Phab. :-)

I really like this idea in general, thank you for proposing this patch!

Thanks for the reviews.
Now take code size into account, and removed outdated comment.

It looks like an explicit FNEG operation isn't going to fly. I propose we drop that pursuit and continue forward with a constrained FNEG intrinsic.

In D52318#1241940, @andreadb wrote:

Unfortunately, I also found this regression:

define <8 x i32> @bar(<8 x i32> %A, <8 x i32> %B, <8 x i32> %Mask) {
  %1 = and <8 x i32> %A, %Mask
  %2 = xor <8 x i32> %1, %Mask
  %3 = add <8 x i32> %2, %B
  ret <8 x i32> %3
}

I think this one is addressed in the VByte patch, so I'm closing this revision in order to prevent conflicts between these two.

Hi Sanjay,

• Be more explicit about the nature of "benchmarks" with memory tracking logic via State::SetLabel(...).
• Force single iteration for "benchmarks" with memory usage tracking
• Add another "benchmark" with Dex memory overhead over plain SymbolSlab

Re-landed as r342739.