Bug 223844 – Allow DisplayRefreshMonitors to be more long-lived objects

• Component: WebCore Misc.
• Status: RESOLVED FIXED

[Simon Fraser (smfr)]

Allow DisplayRefreshMonitor to be more long-lived objects

[Simon Fraser (smfr)]

Created attachment 424470 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 424471 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 424472 [details]
Patch

[Chris Dumez]

Comment on attachment 424472 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=424472&action=review

R=me assuming bots are happy

> Source/WebCore/platform/graphics/DisplayRefreshMonitor.cpp:106
> +    LockHolder lock(mutex());

I think we usually use locker for such variables, e.g:
Auto locker = holdLock(lock());

> Source/WebCore/platform/graphics/DisplayRefreshMonitor.cpp:111
> +    auto started = startNotificationMechanism();

I don’t think we need this local. Could write:
If(!startNotifocationMechanism())
    return false;

> Source/WebCore/platform/graphics/DisplayRefreshMonitor.cpp:128
> +    ++m_unscheduledFireCount;

Could be merged into the next line.

> Source/WebCore/platform/graphics/DisplayRefreshMonitor.cpp:135
>          LockHolder lock(m_mutex);

Same comment about locker.

> Source/WebCore/platform/graphics/DisplayRefreshMonitor.h:75
> +    Lock& mutex() { return m_mutex; }

Why don’t we call this lock instead of mutex since this is what we call it nowadays?

[Simon Fraser (smfr)]

Created attachment 424473 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 424477 [details]
Patch

[EWS]

Committed r275144: <https://commits.webkit.org/r275144>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 424477 [details].

[Radar WebKit Bug Importer]

<rdar://problem/75926530>