Reduce ErrorsFutureStrictReservedWords test to make it faster

Reduce ErrorsFutureStrictReservedWords test to make it faster

Since checking all possible combinations is taking so long this
reduces the test to test the odd cases (let, yield and static) as
well as a single ordinary case.

BUG=v8:3707
LOG=n

Committed: https://chromium.googlesource.com/v8/v8/+/624150a3696da61e1014ffa2e3350436b768199e

[arv (Not doing code reviews), 2014-11-19 18:32:58]

arv@chromium.org changed reviewers:
+ bmeurer@chromium.org, jkummerow@chromium.org

[arv (Not doing code reviews), 2014-11-19 18:32:59]

PTAL

[Jakob Kummerow, 2014-11-19 19:27:06]

LGTM.

I've changed the description to LOG=n. The rule for "y" is: "is this change
interesting to external users/embedders who might read the ChangeLog?", where
making a test faster is a clear counter-example.

The biggest improvement I can think of would come from introducing a
"kAllowAllExperimentalFeatures" parser flag, to be used by tests, that would
imply the flags for individual harmony features that the various parser tests
currently pass in, and thereby get the number of flag combinations from 1 << 8
down to 1 << 3 or so (if kAllowLazy and kAllowNativesSyntax remain separate).

[arv (Not doing code reviews), 2014-11-19 19:38:25]

The CQ bit was checked by arv@chromium.org

[arv (Not doing code reviews), 2014-11-19 19:38:53]

The CQ bit was unchecked by arv@chromium.org

[arv (Not doing code reviews), 2014-11-19 19:39:17]

On 2014/11/19 19:27:06, Jakob wrote:
> LGTM.
> 
> I've changed the description to LOG=n. The rule for "y" is: "is this change
> interesting to external users/embedders who might read the ChangeLog?", where
> making a test faster is a clear counter-example.
> 
> The biggest improvement I can think of would come from introducing a
> "kAllowAllExperimentalFeatures" parser flag, to be used by tests, that would
> imply the flags for individual harmony features that the various parser tests
> currently pass in, and thereby get the number of flag combinations from 1 << 8
> down to 1 << 3 or so (if kAllowLazy and kAllowNativesSyntax remain separate).

In this case, the different combinations of the flags is not important. Let me
split it up.

[arv (Not doing code reviews), 2014-11-19 20:03:06]

On 2014/11/19 19:39:17, arv wrote:
> On 2014/11/19 19:27:06, Jakob wrote:
> > LGTM.
> > 
> > I've changed the description to LOG=n. The rule for "y" is: "is this change
> > interesting to external users/embedders who might read the ChangeLog?",
where
> > making a test faster is a clear counter-example.
> > 
> > The biggest improvement I can think of would come from introducing a
> > "kAllowAllExperimentalFeatures" parser flag, to be used by tests, that would
> > imply the flags for individual harmony features that the various parser
tests
> > currently pass in, and thereby get the number of flag combinations from 1 <<
8
> > down to 1 << 3 or so (if kAllowLazy and kAllowNativesSyntax remain
separate).
> 
> In this case, the different combinations of the flags is not important. Let me
> split it up.

Splitting them up makes the test take longer. Committing as is.

[arv (Not doing code reviews), 2014-11-19 20:03:15]

The CQ bit was checked by arv@chromium.org

[commit-bot: I haz the power, 2014-11-19 20:03:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/724713004/1

[commit-bot: I haz the power, 2014-11-19 20:31:07]

Committed patchset #1 (id:1)