Bug 156782 – Modern IDB (Workers): Enable INDEXED_DATABASE_IN_WORKERS compile time flag, but disabled in RuntimeEnabledFeatures

• Status: RESOLVED
• Resolution: FIXED
• Component: WebCore Misc.

[Brady Eidson]

Modern IDB (Workers): Enable INDEXED_DATABASE_IN_WORKERS compile time flag, but disabled in RuntimeEnabledFeatures

[Brady Eidson]

Created attachment 276805 [details]
WIP

This WIP compiles and is probably almost done, minus a test.

[WebKit Commit Bot]

Attachment 276805 [details] did not pass style-queue:


ERROR: Source/WebKit/mac/Configurations/FeatureDefines.xcconfig:0:  Any changes made to FeatureDefines should be made to all of them (changed file does not match Tools/TestWebKitAPI/Configurations/FeatureDefines.xcconfig).  [featuredefines/equality] [5]
ERROR: Source/JavaScriptCore/Configurations/FeatureDefines.xcconfig:0:  Any changes made to FeatureDefines should be made to all of them (changed file does not match Tools/TestWebKitAPI/Configurations/FeatureDefines.xcconfig).  [featuredefines/equality] [5]
ERROR: Source/WebCore/Configurations/FeatureDefines.xcconfig:0:  Any changes made to FeatureDefines should be made to all of them (changed file does not match Tools/TestWebKitAPI/Configurations/FeatureDefines.xcconfig).  [featuredefines/equality] [5]
ERROR: Source/WebKit2/Configurations/FeatureDefines.xcconfig:0:  Any changes made to FeatureDefines should be made to all of them (changed file does not match Tools/TestWebKitAPI/Configurations/FeatureDefines.xcconfig).  [featuredefines/equality] [5]
Total errors found: 4 in 13 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Brady Eidson]

Created attachment 276834 [details]
Patch v1

[Brady Eidson]

Created attachment 276835 [details]
Patch v2

[Alex Christensen]

Comment on attachment 276835 [details]
Patch v2

View in context: https://bugs.webkit.org/attachment.cgi?id=276835&action=review

> LayoutTests/storage/indexeddb/modern/workers-enable-expected.txt:9
> +FAIL [Worker] self.indexedDB should be non-null. Was null
> +FAIL [Worker] self.indexedDB instanceof IDBFactory should be true. Was false.

fail?

[Brady Eidson]

(In reply to comment #5)
> Comment on attachment 276835 [details]
> Patch v2
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=276835&action=review
> 
> > LayoutTests/storage/indexeddb/modern/workers-enable-expected.txt:9
> > +FAIL [Worker] self.indexedDB should be non-null. Was null
> > +FAIL [Worker] self.indexedDB instanceof IDBFactory should be true. Was false.
> 
> fail?

From the ChangeLog:

+        Note: One test has some "FAIL" lines in the expected results, which is intentional.
+        As work on this progresses, those FAILs will become PASSes and the expectations will be updated.

[Alex Christensen]

../../Source/WebCore/testing/InternalSettings.cpp:547:46: error: 'class WebCore::RuntimeEnabledFeatures' has no member named 'setIndexedDBWorkersEnabled'
     RuntimeEnabledFeatures::sharedFeatures().setIndexedDBWorkersEnabled(enabled);

I think you need some more compile time flags.

[Brady Eidson]

(In reply to comment #7)
> ../../Source/WebCore/testing/InternalSettings.cpp:547:46: error: 'class
> WebCore::RuntimeEnabledFeatures' has no member named
> 'setIndexedDBWorkersEnabled'
>     
> RuntimeEnabledFeatures::sharedFeatures().setIndexedDBWorkersEnabled(enabled);
> 
> I think you need some more compile time flags.

Yes, I planned to rely on EWS to tell me where, which is why I didn't cq?

[Alex Christensen]

Comment on attachment 276835 [details]
Patch v2

View in context: https://bugs.webkit.org/attachment.cgi?id=276835&action=review

> Source/WebCore/Modules/notifications/WorkerGlobalScopeNotifications.cpp:59
> -        provideTo(context, supplementName(), WTFMove(newSupplement));
> +        provideTo(&scope, supplementName(), WTFMove(newSupplement));

Outside the scope of this patch, but we should make provideTo take a reference.

[Brady Eidson]

(In reply to comment #9)
> Comment on attachment 276835 [details]
> Patch v2
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=276835&action=review
> 
> > Source/WebCore/Modules/notifications/WorkerGlobalScopeNotifications.cpp:59
> > -        provideTo(context, supplementName(), WTFMove(newSupplement));
> > +        provideTo(&scope, supplementName(), WTFMove(newSupplement));
> 
> Outside the scope of this patch, but we should make provideTo take a
> reference.

Considered doing it in this patch before it become obvious just how many sites would have to change.

[Brady Eidson]

http://trac.webkit.org/changeset/199779

[Csaba Osztrogonác]

(In reply to comment #11)
> http://trac.webkit.org/changeset/199779

It broke the non INDEXED_DATABASE_IN_WORKERS build, for example EFL and GTK. And not in Webkit2, but in Webcore.

[Brady Eidson]

(In reply to comment #12)
> (In reply to comment #11)
> > http://trac.webkit.org/changeset/199779
> 
> It broke the non INDEXED_DATABASE_IN_WORKERS build, for example EFL and GTK.
> And not in Webkit2, but in Webcore.

Link to failure? I saw that breakage in EWS and attempted to fix.

[Brady Eidson]

A link to the failure so others don't have to go hunting:
https://build.webkit.org/builders/GTK%20Linux%2064-bit%20Release%20%28Build%29/builds/68616/steps/compile-webkit/logs/stdio

Working on a fix.

[Brady Eidson]

Fix attempt in http://trac.webkit.org/changeset/199782