ES2015 web compat workaround: RegExp.prototype.flags => ""

ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
but omits a UseCounter to make backports easier.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel

Committed: https://crrev.com/b22b2588740b059b26b6687095514c03dea0feaf
Cr-Commit-Position: refs/heads/master@{#34201}

[Dan Ehrenberg, 2016-01-27 20:06:10]

The CQ bit was checked by littledan@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-27 20:06:18]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1640803003/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1640803003/1

[Dan Ehrenberg, 2016-01-27 20:08:56]

This patch includes UseCounters. It might need to be backported to earlier
versions of V8, in which case the UseCounter will probably have to be stripped
out for the backport.

[adamk, 2016-01-27 20:14:29]

Please leave the usecounter for a separate patch; that'd make merging this more
straightforward (it's easier to land two patches on trunk and merge one than
land one patch and partially-merge another).

Were you thinking of landing this now or waiting for more web developer feedback
from M48?

[commit-bot: I haz the power, 2016-01-27 20:21:13]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-27 20:21:14]

Dry run: Try jobs failed on following builders:
  v8_linux64_avx2_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel/builds/...)

[Dan Ehrenberg, 2016-01-27 21:04:07]

On 2016/01/27 at 20:21:14, commit-bot wrote:
> Dry run: Try jobs failed on following builders:
>   v8_linux64_avx2_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel/builds/...)

I was thinking we could land this now and backport to 48; it's always possible
to back it out later, but we don't know how wide an impact this compat issue has
yet. The UseCounters will give us that data. I'll separate them out.

[adamk, 2016-01-27 21:06:07]

Justifying a merge to stable usually requires a significant impact assessment;
it's not clear we have that yet. Interested to see if we hear back from the dev
on that bug.

[Dan Ehrenberg, 2016-01-27 22:19:34]

The CQ bit was checked by littledan@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-27 22:19:41]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1640803003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1640803003/20001

[Dan Ehrenberg, 2016-01-27 22:22:11]

Apparently all websites using es6shim before the version from early November are
subject to this issue. That seems like a fairly big impact. Maybe we should
include this patch now, and try again in a year or so once those installations
have cleared out. The new version removes the UseCounter.

[commit-bot: I haz the power, 2016-01-27 22:30:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-27 22:30:58]

Dry run: Try jobs failed on following builders:
  v8_mac_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel/builds/14654)

[Dan Ehrenberg, 2016-01-28 00:21:43]

Description was changed from

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
and adds UseCounters for how often the feature is invoked.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
==========

to

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
and adds UseCounters for how often the feature is invoked.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
==========

[Dan Ehrenberg, 2016-01-28 00:21:48]

The CQ bit was checked by littledan@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-28 00:21:53]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1640803003/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1640803003/40001

[commit-bot: I haz the power, 2016-01-28 01:45:26]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-01-28 01:45:27]

Dry run: This issue passed the CQ dry run.

[Dan Ehrenberg, 2016-01-28 04:56:49]

Description was changed from

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
and adds UseCounters for how often the feature is invoked.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
==========

to

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
but omits a UseCounter to make backports easier.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
==========

[Dan Ehrenberg, 2016-01-28 04:57:03]

No UseCounters, and fixed tests. PTAL.

[adamk, 2016-01-28 23:47:51]

Code looks fine, but I'd still like to hold off on landing this until we have
more evidence that the specced behavior is causing a problem in the wild.

To me, the downside of landing this is that it's just more work we have to do in
the future, checking UseCounters and removing the workaround if they're low
enough. If it turns out that there are no further reports of breakage in Chrome
48, then we needn't do any more work, either now or in the future, around this
change.

[adamk, 2016-02-22 23:59:12]

Turns out the non-web-compatible behavior, contrary to our previous
understanding, was not shipped to stable in Chrome 48 (it was still behind a
flag).

Given that we're not seeing more evidence of breakage during the M49 release
process (https://bugs.chromium.org/p/chromium/issues/detail?id=588625), this
change lgtm, and we should merge it to the 4.9 branch as well.

[Dan Ehrenberg, 2016-02-23 00:34:47]

The CQ bit was checked by littledan@chromium.org

[commit-bot: I haz the power, 2016-02-23 00:34:57]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1640803003/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1640803003/40001

[commit-bot: I haz the power, 2016-02-23 01:48:38]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-02-23 01:49:12]

Description was changed from

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
but omits a UseCounter to make backports easier.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
==========

to

==========
ES2015 web compat workaround: RegExp.prototype.flags => ""

It turns out that some old polyfill library uses
RegExp.prototype.flags as a way of feature testing. It's not clear
how widespread this is. For now, as a minimal workaround, we can
return undefined from getters like RegExp.prototype.global when
the receiver is RegExp.prototype. This patch implements that strategy
but omits a UseCounter to make backports easier.

R=adamk
CC=yangguo@chromium.org
BUG=chromium:581577
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel

Committed: https://crrev.com/b22b2588740b059b26b6687095514c03dea0feaf
Cr-Commit-Position: refs/heads/master@{#34201}
==========

[commit-bot: I haz the power, 2016-02-23 01:49:13]

Patchset 3 (id:??) landed as
https://crrev.com/b22b2588740b059b26b6687095514c03dea0feaf
Cr-Commit-Position: refs/heads/master@{#34201}