Handle ES2015 Function.name in CallSite::GetMethodName

Handle ES2015 Function.name in CallSite::GetMethodName

CallSite depends on using the function name to get ahold of the property
name from which an exception was thrown. This fix properly handles the
ES2015 names for getters and setters. The new tests pass both with
--harmony-function-name off and on.

BUG=v8:3699
LOG=n

Committed: https://crrev.com/045fa997b77f335e251dc8167bb2f409b4abded1
Cr-Commit-Position: refs/heads/master@{#34469}

[adamk, 2016-03-03 01:29:51]

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

[adamk, 2016-03-03 01:29:52]

Here's the patch I mentioned in email. Happy to take a different approach, but
this at least seems to work (I'm not really clear on how robust getMethodName()
needs to be).

[Yang, 2016-03-03 05:40:29]

On 2016/03/03 01:29:52, adamk wrote:
> Here's the patch I mentioned in email. Happy to take a different approach, but
> this at least seems to work (I'm not really clear on how robust
getMethodName()
> needs to be).

LGTM.

I havent followed this ES6 change. Iirc we used to not store the get/set prefix
to the function name, and add these later. What happened to that? Is that only
done when formatting stack traces? If yes, we might want to change that to not
first strip and later readd the prefix.

[adamk, 2016-03-03 19:53:39]

On 2016/03/03 05:40:29, Yang wrote:
> On 2016/03/03 01:29:52, adamk wrote:
> > Here's the patch I mentioned in email. Happy to take a different approach,
but
> > this at least seems to work (I'm not really clear on how robust
> getMethodName()
> > needs to be).
> 
> LGTM.
> 
> I havent followed this ES6 change. Iirc we used to not store the get/set
prefix
> to the function name, and add these later. What happened to that? Is that only
> done when formatting stack traces? If yes, we might want to change that to not
> first strip and later readd the prefix.

I haven't found any code in V8 for adding "get" or "set" prefixes, but the
devtools do do that formatting in a few places; I plan to fix that once this
lands, on the Blink side.

[adamk, 2016-03-03 20:03:26]

The CQ bit was checked by adamk@chromium.org

[commit-bot: I haz the power, 2016-03-03 20:03:34]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1751403004/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1751403004/1

[commit-bot: I haz the power, 2016-03-03 20:18:38]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-03-03 20:19:51]

Description was changed from

==========
Handle ES2015 Function.name in CallSite::GetMethodName

CallSite depends on using the function name to get ahold of the property
name from which an exception was thrown. This fix properly handles the
ES2015 names for getters and setters. The new tests pass both with
--harmony-function-name off and on.

BUG=v8:3699
LOG=n
==========

to

==========
Handle ES2015 Function.name in CallSite::GetMethodName

CallSite depends on using the function name to get ahold of the property
name from which an exception was thrown. This fix properly handles the
ES2015 names for getters and setters. The new tests pass both with
--harmony-function-name off and on.

BUG=v8:3699
LOG=n

Committed: https://crrev.com/045fa997b77f335e251dc8167bb2f409b4abded1
Cr-Commit-Position: refs/heads/master@{#34469}
==========

[commit-bot: I haz the power, 2016-03-03 20:19:52]

Patchset 1 (id:??) landed as
https://crrev.com/045fa997b77f335e251dc8167bb2f409b4abded1
Cr-Commit-Position: refs/heads/master@{#34469}