[strong] Check constructor returns

[strong] Check constructor returns

R=dslomov@chromium.org
BUG=v8:3956
LOG=N

Committed: https://crrev.com/74cf8e0122559680fd94dffa9a322edf4b34260d
Cr-Commit-Position: refs/heads/master@{#27320}

[Dmitry Lomov (no reviews), 2015-03-19 13:59:41]

lgtm % comments

https://codereview.chromium.org/1019283002/diff/1/src/messages.js
File src/messages.js (right):

https://codereview.chromium.org/1019283002/diff/1/src/messages.js#newcode176
src/messages.js:176: strong_constructor_return_value: ["Please do not return a
value from a constructor in strong mode"],
Why do we keep pleading with users here?

https://codereview.chromium.org/1019283002/diff/1/test/cctest/test-parsing.cc
File test/cctest/test-parsing.cc (right):

https://codereview.chromium.org/1019283002/diff/1/test/cctest/test-parsing.cc...
test/cctest/test-parsing.cc:5632: RunParserSyncTest(strong_context_data, data,
kError, NULL, 0, always_flags,
Please add positive tests for strong mode:
 'constructor() { super(); if (x) return; ... }' 
etc

[adamk, 2015-03-19 14:04:11]

https://codereview.chromium.org/1019283002/diff/1/src/messages.js
File src/messages.js (right):

https://codereview.chromium.org/1019283002/diff/1/src/messages.js#newcode176
src/messages.js:176: strong_constructor_return_value: ["Please do not return a
value from a constructor in strong mode"],
On 2015/03/19 13:59:41, Dmitry Lomov (chromium) wrote:
> Why do we keep pleading with users here?

Drive-by +1 here. The "please" sounds odd for something that's a hard
requirement of the language. That is, it reads more like a deprecation message
than an error.

[rossberg, 2015-03-19 16:02:57]

https://codereview.chromium.org/1019283002/diff/1/src/messages.js
File src/messages.js (right):

https://codereview.chromium.org/1019283002/diff/1/src/messages.js#newcode176
src/messages.js:176: strong_constructor_return_value: ["Please do not return a
value from a constructor in strong mode"],
On 2015/03/19 14:04:11, adamk wrote:
> On 2015/03/19 13:59:41, Dmitry Lomov (chromium) wrote:
> > Why do we keep pleading with users here?
> 
> Drive-by +1 here. The "please" sounds odd for something that's a hard
> requirement of the language. That is, it reads more like a deprecation message
> than an error.

Strong mode intentionally has the mostly friendly semantics in the world. :)

Alright, these should be reformulated. But rather consistently in a separate CL.

https://codereview.chromium.org/1019283002/diff/1/test/cctest/test-parsing.cc
File test/cctest/test-parsing.cc (right):

https://codereview.chromium.org/1019283002/diff/1/test/cctest/test-parsing.cc...
test/cctest/test-parsing.cc:5632: RunParserSyncTest(strong_context_data, data,
kError, NULL, 0, always_flags,
On 2015/03/19 13:59:41, Dmitry Lomov (chromium) wrote:
> Please add positive tests for strong mode:
>  'constructor() { super(); if (x) return; ... }' 
> etc

We don't have any 'positive' strong mode tests at all, so far. We need to think
about a proper way of ensuring those, but that's also for separate CL.

[rossberg, 2015-03-19 16:03:17]

The CQ bit was checked by rossberg@chromium.org

[commit-bot: I haz the power, 2015-03-19 16:03:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1019283002/1

[commit-bot: I haz the power, 2015-03-19 19:40:00]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-03-19 19:41:03]

Patchset 1 (id:??) landed as
https://crrev.com/74cf8e0122559680fd94dffa9a322edf4b34260d
Cr-Commit-Position: refs/heads/master@{#27320}