Bug 196475

Summary: [WebCore] Put most of derived classes of ScriptWrappable into IsoHeap
Product: WebKit Reporter: Yusuke Suzuki <ysuzuki>
Component: DOMAssignee: Yusuke Suzuki <ysuzuki>
Status: RESOLVED FIXED    
Severity: Normal CC: ddkilzer, saam, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Patch
none
Patch
none
Patch
none
Patch
none
Patch
none
Patch
saam: review+
Patch none

Description Yusuke Suzuki 2019-04-01 18:39:42 PDT 
...
Comment 1 Yusuke Suzuki 2019-04-01 19:26:44 PDT 
Created attachment 366456 [details]
Patch
Comment 2 Yusuke Suzuki 2019-04-01 19:38:51 PDT 
Created attachment 366458 [details]
Patch
Comment 3 Yusuke Suzuki 2019-04-01 19:39:31 PDT 
I'll put some special MACRO to avoid accidental instantiation of template classes with non-iso-heap thing...
Comment 4 Yusuke Suzuki 2019-04-02 20:16:39 PDT 
Created attachment 366572 [details]
Patch
Comment 5 Yusuke Suzuki 2019-04-02 21:11:14 PDT 
Created attachment 366575 [details]
Patch
Comment 6 Yusuke Suzuki 2019-04-02 22:14:22 PDT 
Because derived classes of Event exist in Internal repository (see WebCore/dom/mac/GestureEvents.cpp), we do not include Event class hierarchy for now.
Comment 7 Yusuke Suzuki 2019-04-02 22:29:24 PDT 
Created attachment 366577 [details]
Patch
Comment 8 Yusuke Suzuki 2019-04-02 22:55:49 PDT 
Created attachment 366580 [details]
Patch
Comment 9 Yusuke Suzuki 2019-04-02 22:56:13 PDT 
The purpose of this patch is putting many classes into IsoHeap and watching the bots.
1. If the bots say it is not memory-footprint sensitive, we continue doing this and expanding IsoHeap converage.
2. If the bots say it is memory-footprint sensitive, investigate the scalability problem of current IsoHeap and fix this.
Currently, this patch does not include Event and derived classes because they are also inherited in Internal repository.
Comment 10 Saam Barati 2019-04-03 11:18:24 PDT 
Comment on attachment 366580 [details]
Patch

r=me
Comment 11 Yusuke Suzuki 2019-04-04 01:20:41 PDT 
Created attachment 366702 [details]
Patch

Patch for landing
Comment 12 Yusuke Suzuki 2019-04-04 12:08:16 PDT 
Committed r243887: <https://trac.webkit.org/changeset/243887>
Comment 13 Radar WebKit Bug Importer 2019-04-04 12:09:25 PDT 
<rdar://problem/49616980>
Comment 14 Yusuke Suzuki 2019-04-04 12:57:34 PDT 
I'm looking into RTCIceTransport linker errors.
Comment 15 Yusuke Suzuki 2019-04-04 13:08:45 PDT 
Committed r243890: <https://trac.webkit.org/changeset/243890>
Comment 16 Michael Catanzaro 2019-04-04 13:13:58 PDT 
Seems this requires a clean build. I've forced a clean build on a few bots. No doubt more will be needed.
Comment 17 Yusuke Suzuki 2019-04-04 13:27:38 PDT 
(In reply to Michael Catanzaro from comment #16)
> Seems this requires a clean build. I've forced a clean build on a few bots.
> No doubt more will be needed.

Yeah, I agree with you. It seems that forwarding headers are not updated, and clean build will solve it in GTK & WPE ports :)
Comment 18 Yusuke Suzuki 2019-04-04 18:45:33 PDT 
Committed r243924: <https://trac.webkit.org/changeset/243924>
Comment 19 David Kilzer (:ddkilzer) 2019-04-05 05:25:54 PDT 
(In reply to Yusuke Suzuki from comment #18)
> Committed r243924: <https://trac.webkit.org/changeset/243924>

Follow-up build fix attempt for Windows (may not have fixed all issues):

Committed r243931: <https://trac.webkit.org/changeset/243931>
Comment 20 Yusuke Suzuki 2019-04-08 15:26:23 PDT 
Committed r244052: <https://trac.webkit.org/changeset/244052>