Convert Devtools protocol Input timestamp to monotonic clock

Convert Devtools protocol Input timestamp to monotonic clock.

Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837
Committed: https://crrev.com/9ca5a011a3c5b1ff2856a17f60b30f7d5e977584
Cr-Commit-Position: refs/heads/master@{#380201}

[majidvp, 2016-03-02 19:52:33]

Description was changed from

==========
Fix devtools timestamp coversion

BUG=590837
==========

to

==========
Convert Devtools protocol Input timestamp to monotonic clock.


Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837
==========

[majidvp, 2016-03-02 19:54:20]

majidvp@chromium.org changed reviewers:
+ dgozman@chromium.org

[dgozman, 2016-03-02 20:11:38]

Should we also change InspectorInputAgent::dispatchTouchEvent?

[majidvp, 2016-03-07 13:42:53]

On 2016/03/02 20:11:38, dgozman wrote:
> Should we also change InspectorInputAgent::dispatchTouchEvent?

Yes we do. Added logic to cover that case.

[majidvp, 2016-03-07 13:45:49]

https://codereview.chromium.org/1747383002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/inspector/InspectorInputAgent.cpp (right):

https://codereview.chromium.org/1747383002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/inspector/InspectorInputAgent.cpp:83: return
timestamp - epochToMonotonicTimeDelta;
We don't need this type of conversion anywhere else so I feel it is best to keep
it here and not add it to wtf/CurrentTime.h

[dgozman, 2016-03-07 17:59:37]

dgozman@chromium.org changed reviewers:
+ samuong@chromium.org

[dgozman, 2016-03-07 17:59:38]

@samoung: could you please take a look? Won't chromedriver break because of
this?

https://codereview.chromium.org/1747383002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/inspector/InspectorInputAgent.cpp (right):

https://codereview.chromium.org/1747383002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/inspector/InspectorInputAgent.cpp:83: return
timestamp - epochToMonotonicTimeDelta;
On 2016/03/07 13:45:49, majidvp wrote:
> We don't need this type of conversion anywhere else so I feel it is best to
keep
> it here and not add it to wtf/CurrentTime.h

Agreed.

[samuong, 2016-03-07 19:22:15]

ChromeDriver doesn't set the timestamp when it dispatches key/mouse/touch
events, so I don't think this will break us. Will let you know if this breaks
our tests once we get a build going.

[dgozman, 2016-03-07 19:28:18]

lgtm

[samuong, 2016-03-07 20:06:56]

btw does this need to be rebased to fix the compile errors?

[samuong, 2016-03-07 23:40:42]

On 2016/03/07 20:06:56, samuong wrote:
> btw does this need to be rebased to fix the compile errors?

gmanikpure@ just ran this against our chromedriver tests and didn't see any
failures, so lgtm

[majidvp, 2016-03-09 16:13:26]

On 2016/03/07 23:40:42, samuong wrote:
> On 2016/03/07 20:06:56, samuong wrote:
> > btw does this need to be rebased to fix the compile errors?
> 
> gmanikpure@ just ran this against our chromedriver tests and didn't see any
> failures, so lgtm

Thanks for testing and confirming.

Meanwhile I have also added a layout test that covers the clock conversion
logic. 

dgozman@: can you please take a look again? Only difference is the 
layout test.

[dgozman, 2016-03-09 17:08:22]

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html
(right):

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:3:
window.addEventListener("keydown", logEvent);
style: no indentation in <script> tag

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:9:
function logEvent(event) {
style: { on next line for function body

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:112:
InspectorTest.completeTest();
You have to wait for Runtime.evaluate to finish before completing test (do it in
a callback).

[majidvp, 2016-03-09 20:09:08]

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html
(right):

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:3:
window.addEventListener("keydown", logEvent);
On 2016/03/09 17:08:21, dgozman wrote:
> style: no indentation in <script> tag

Done.

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:9:
function logEvent(event) {
On 2016/03/09 17:08:21, dgozman wrote:
> style: { on next line for function body

Done.

https://codereview.chromium.org/1747383002/diff/80001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/inspector-protocol/input/eventTimestamp.html:112:
InspectorTest.completeTest();
On 2016/03/09 17:08:21, dgozman wrote:
> You have to wait for Runtime.evaluate to finish before completing test (do it
in
> a callback).

Done.

[dgozman, 2016-03-09 20:10:13]

still lgtm

[majidvp, 2016-03-09 20:16:56]

The CQ bit was checked by majidvp@chromium.org

[majidvp, 2016-03-09 20:16:56]

The patchset sent to the CQ was uploaded after l-g-t-m from samuong@chromium.org
Link to the patchset: https://codereview.chromium.org/1747383002/#ps100001
(title: "Address feedback")

[commit-bot: I haz the power, 2016-03-09 20:17:14]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1747383002/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1747383002/100001

[commit-bot: I haz the power, 2016-03-09 20:27:24]

Description was changed from

==========
Convert Devtools protocol Input timestamp to monotonic clock.


Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837
==========

to

==========
Convert Devtools protocol Input timestamp to monotonic clock.


Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837
==========

[commit-bot: I haz the power, 2016-03-09 20:27:26]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2016-03-09 20:29:44]

Description was changed from

==========
Convert Devtools protocol Input timestamp to monotonic clock.


Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837
==========

to

==========
Convert Devtools protocol Input timestamp to monotonic clock.

Devtools protocol specifies input timestamp as a unix epoch time [1]. While
chrome internally expects a monotonic timestamp that is measured since system
start time.

The current patch satisfies both contracts by converting received value from
wall clock (base::Time) to monotonic clock (base::TimeTicks) before using it for
|WebInputInput.timeStampSeconds|.

[1] https://developer.chrome.com/devtools/docs/protocol/1.1/input

BUG=590837

Committed: https://crrev.com/9ca5a011a3c5b1ff2856a17f60b30f7d5e977584
Cr-Commit-Position: refs/heads/master@{#380201}
==========

[commit-bot: I haz the power, 2016-03-09 20:29:46]

Patchset 6 (id:??) landed as
https://crrev.com/9ca5a011a3c5b1ff2856a17f60b30f7d5e977584
Cr-Commit-Position: refs/heads/master@{#380201}