PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler:
  https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694
Review-Url: https://codereview.chromium.org/2640293002
Cr-Commit-Position: refs/heads/master@{#445050}
Committed: https://chromium.googlesource.com/chromium/src/+/ff4edc10916a681cb2f6028c149f9755d8024bf4

[zino, 2017-01-19 14:18:41]

Description was changed from

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's UserInteraction timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

BUG=624694
==========

to

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's UserInteraction timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

BUG=624694
==========

[zino, 2017-01-19 14:18:41]

jinho.bang@samsung.com changed reviewers:
+ haraken@chromium.org, rouslan@chromium.org

[zino, 2017-01-19 14:18:49]

PTAL

[haraken, 2017-01-19 14:27:32]

Yeah, the spec is not clear about the task source. rouslan@?

If we're not sure, we can use the MiscPlatformAPI at the moment.

[please use gerrit instead, 2017-01-19 15:02:05]

What problem does this solve or which advantage does this provide? Here's the
relevant text from the spec, by the way:

"Implementations may choose to implement a timeout for pending updates if
detailsPromise doesn't settle in a reasonable amount of time. If an
implementation chooses to implement a timeout, they must execute the steps
listed below in the "upon rejection" path. Such a timeout is a fatal error for
the payment request"

[haraken, 2017-01-19 15:29:26]

On 2017/01/19 15:02:05, rouslan wrote:
> What problem does this solve or which advantage does this provide? Here's the
> relevant text from the spec, by the way:
> 
> "Implementations may choose to implement a timeout for pending updates if
> detailsPromise doesn't settle in a reasonable amount of time. If an
> implementation chooses to implement a timeout, they must execute the steps
> listed below in the "upon rejection" path. Such a timeout is a fatal error for
> the payment request"

This is a part of the per-frame scheduler work:

https://groups.google.com/a/chromium.org/d/topic/blink-dev/80dsbEEOcWY/discus...

[zino, 2017-01-19 15:32:41]

On 2017/01/19 15:02:05, rouslan wrote:
> What problem does this solve or which advantage does this provide? Here's the
> relevant text from the spec, by the way:

This is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler.
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

It's not for only Payment Request API and it's changing in overall Blink.
According to the link, the per-frame scheduler enables us to flexibly schedule
tasks/timers per frame.
For example, it enables to throttle offscreen iframes, which is a key technology
to save power and improve responsiveness of the main frame.

> 
> "Implementations may choose to implement a timeout for pending updates if
> detailsPromise doesn't settle in a reasonable amount of time. If an
> implementation chooses to implement a timeout, they must execute the steps
> listed below in the "upon rejection" path. Such a timeout is a fatal error for
> the payment request"

I think we can use MiscPlatformAPI. (But I'm not sure if it's proper.)
I found a similar case: https://codereview.chromium.org/2642863002

[zino, 2017-01-19 15:37:06]

Description was changed from

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's UserInteraction timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

BUG=624694
==========

to

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

BUG=624694
==========

[zino, 2017-01-19 15:37:56]

Description was changed from

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

BUG=624694
==========

to

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler.
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694
==========

[zino, 2017-01-19 15:38:10]

Description was changed from

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler.
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694
==========

to

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler:
  https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694
==========

[please use gerrit instead, 2017-01-19 15:38:51]

Let's use MiscPlatformAPI. Thank you for the context.

[zino, 2017-01-19 15:40:47]

On 2017/01/19 15:38:51, rouslan wrote:
> Let's use MiscPlatformAPI. Thank you for the context.

Done.

[zino, 2017-01-19 15:41:00]

On 2017/01/19 15:40:47, zino wrote:
> On 2017/01/19 15:38:51, rouslan wrote:
> > Let's use MiscPlatformAPI. Thank you for the context.
> 
> Done.

Thank you for review.

PTAL

[please use gerrit instead, 2017-01-19 16:33:19]

lgtm

[haraken, 2017-01-20 00:12:33]

LGTM

[zino, 2017-01-20 13:01:48]

The CQ bit was checked by jinho.bang@samsung.com

[commit-bot: I haz the power, 2017-01-20 13:01:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-20 14:30:16]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1484917308268500,
"parent_rev": "a7a3f277b70327d2fd2f383acba1b1c2c78f018c", "commit_rev":
"ff4edc10916a681cb2f6028c149f9755d8024bf4"}

[commit-bot: I haz the power, 2017-01-20 14:30:43]

Description was changed from

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler:
  https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694
==========

to

==========
PaymentRequest: Move m_abortTimer to frame-specific TaskRunnerTimer.

Move m_abortTimer to frame-specific TaskRunnerTimer. This associates it with
the frame's MiscPlatformAPI timer task. This patch also makes the signature of
constructor take ExecutionContext to get the frame object. This CL includes
IDL change but it's not breaking existing web.

The spec link:
  https://w3c.github.io/browser-payment-api/#updatewith-method

This CL is migrating tasks/timers from a per-renderer scheduler to a per-frame
scheduler:
  https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/80dsbEEOcWY

BUG=624694

Review-Url: https://codereview.chromium.org/2640293002
Cr-Commit-Position: refs/heads/master@{#445050}
Committed:
https://chromium.googlesource.com/chromium/src/+/ff4edc10916a681cb2f6028c149f...
==========

[commit-bot: I haz the power, 2017-01-20 14:30:45]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/ff4edc10916a681cb2f6028c149f...