Include Harmony Array/TypedArray methods unconditionally

Include Harmony Array/TypedArray methods unconditionally

Conditionally including Array and TypedArray methods seems to cause
a slowdown in V8 context creation, possibly due to the new code added.

BUG=chromium:504629
R=adamk@chromium.org
LOG=Y

Committed: https://crrev.com/a1dcac3df7a23a311dea12b8478e3c7517877496
Cr-Commit-Position: refs/heads/master@{#29430}

[Dan Ehrenberg, 2015-06-29 05:32:58]

Is there a better approach?

Should this patch remove the --harmony-arrays flag or leave it in as a vestige
that doesn't do anything?

[adamk, 2015-06-29 21:30:26]

adamk@chromium.org changed reviewers:
+ rossberg@chromium.org

[adamk, 2015-06-29 21:30:26]

I've confirmed locally that this fixes the performance test. So the only
question is whether it's the right approach.

I think there's not much point in leaving the flag in, given that there are so
few references to it in code.

As for removing the ability to flag flip this off, that's unfortunate, but I
don't think the perf regression is worth the ease-of-disabling. Adding rossberg
for his thoughts on this question.

[rossberg, 2015-06-30 12:11:52]

On 2015/06/29 21:30:26, adamk wrote:
> I've confirmed locally that this fixes the performance test. So the only
> question is whether it's the right approach.
> 
> I think there's not much point in leaving the flag in, given that there are so
> few references to it in code.
> 
> As for removing the ability to flag flip this off, that's unfortunate, but I
> don't think the perf regression is worth the ease-of-disabling. Adding
rossberg
> for his thoughts on this question.

Sounds reasonable to skip the flag toggling in this case. But I'd prefer to
remove the flag entirely then, instead of leaving it in zombie state. (That also
includes moving relevant tests from harmony/ to es6/.)

[Dan Ehrenberg, 2015-07-01 22:06:35]

On 2015/06/30 12:11:52, rossberg wrote:
> On 2015/06/29 21:30:26, adamk wrote:
> > I've confirmed locally that this fixes the performance test. So the only
> > question is whether it's the right approach.
> > 
> > I think there's not much point in leaving the flag in, given that there are
so
> > few references to it in code.
> > 
> > As for removing the ability to flag flip this off, that's unfortunate, but I
> > don't think the perf regression is worth the ease-of-disabling. Adding
> rossberg
> > for his thoughts on this question.
> 
> Sounds reasonable to skip the flag toggling in this case. But I'd prefer to
> remove the flag entirely then, instead of leaving it in zombie state. (That
also
> includes moving relevant tests from harmony/ to es6/.)

Removed the flag and moved the test, PTAL.

[adamk, 2015-07-01 22:10:03]

The CQ bit was checked by adamk@chromium.org

[adamk, 2015-07-01 22:10:04]

lgtm

[commit-bot: I haz the power, 2015-07-01 22:10:13]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1215863003/20001

[commit-bot: I haz the power, 2015-07-01 23:10:20]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-07-01 23:10:30]

Patchset 2 (id:??) landed as
https://crrev.com/a1dcac3df7a23a311dea12b8478e3c7517877496
Cr-Commit-Position: refs/heads/master@{#29430}