Bug 209940

Summary: Web Inspector: increase the auto-inspect debugger timeout delay to account for slower networks/devices
Product: WebKit Reporter: Devin Rousso <hi>
Component: Web InspectorAssignee: Devin Rousso <hi>
Status: RESOLVED FIXED    
Severity: Normal CC: bburg, ews-watchlist, hi, inspector-bugzilla-changes, joepeck, keith_miller, mark.lam, msaboff, saam, tzagallo, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: WebKit Local Build   
Hardware: All   
OS: All   
Attachments:
Description Flags
Patch
none
Patch none

Description Devin Rousso 2020-04-02 16:49:11 PDT 
Right now, when auto-inspecting, we only pause for a maximum of 1s which may not be enough time for messages from the frontend to be received, especially if there are a lot of things that need to be done (e.g. loading breakpoints from `indexedDB`).
Comment 1 Devin Rousso 2020-04-02 16:49:23 PDT 
<rdar://problem/61120919>
Comment 2 Devin Rousso 2020-04-02 16:50:34 PDT 
Created attachment 395322 [details]
Patch
Comment 3 Joseph Pecoraro 2020-04-02 19:17:03 PDT 
Comment on attachment 395322 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=395322&action=review

r=me

> Source/JavaScriptCore/inspector/remote/cocoa/RemoteInspectorCocoa.mm:160
> +        dispatch_after(dispatch_time(DISPATCH_TIME_NOW, 10 * NSEC_PER_SEC), dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_DEFAULT, 0), ^{

Would still be good to pull this out as a constant so its easier to read.
Comment 4 Devin Rousso 2020-04-03 11:08:29 PDT 
Created attachment 395397 [details]
Patch
Comment 5 EWS 2020-04-03 11:30:36 PDT 
Committed r259479: <https://trac.webkit.org/changeset/259479>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 395397 [details].
Comment 6 BJ Burg 2020-04-03 11:50:25 PDT 
Comment on attachment 395397 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=395397&action=review

> Source/JavaScriptCore/inspector/remote/cocoa/RemoteInspectorCocoa.mm:-145
> -            LOG_ERROR("Skipping Automatic Inspection Candidate with pageId(%u) because we are already paused waiting for pageId(%u)", targetIdentifier, m_automaticInspectionCandidateTargetIdentifier);

I don't think this change is correct. Why WTFLogAlways? We have RELEASE_LOG_ERROR and RELEASE_LOG_ERROR_IF, which should do the same thing but go through os_log and won't get lost in stderr/stdout.
Comment 7 Devin Rousso 2020-04-03 12:14:01 PDT 
Comment on attachment 395397 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=395397&action=review

>> Source/JavaScriptCore/inspector/remote/cocoa/RemoteInspectorCocoa.mm:-145
>> -            LOG_ERROR("Skipping Automatic Inspection Candidate with pageId(%u) because we are already paused waiting for pageId(%u)", targetIdentifier, m_automaticInspectionCandidateTargetIdentifier);
> 
> I don't think this change is correct. Why WTFLogAlways? We have RELEASE_LOG_ERROR and RELEASE_LOG_ERROR_IF, which should do the same thing but go through os_log and won't get lost in stderr/stdout.

JSC doesn't have any of the necessary things to make `RELEASE_LOG_ERROR` work, namely a `LOG_CHANNEL_PREFIX` and any logging channels.  I suppose I could make one, but that seemed like a lot of overkill for something that should hopefully not happen.
Comment 8 BJ Burg 2020-04-03 14:39:48 PDT 
Comment on attachment 395397 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=395397&action=review

>>> Source/JavaScriptCore/inspector/remote/cocoa/RemoteInspectorCocoa.mm:-145
>>> -            LOG_ERROR("Skipping Automatic Inspection Candidate with pageId(%u) because we are already paused waiting for pageId(%u)", targetIdentifier, m_automaticInspectionCandidateTargetIdentifier);
>> 
>> I don't think this change is correct. Why WTFLogAlways? We have RELEASE_LOG_ERROR and RELEASE_LOG_ERROR_IF, which should do the same thing but go through os_log and won't get lost in stderr/stdout.
> 
> JSC doesn't have any of the necessary things to make `RELEASE_LOG_ERROR` work, namely a `LOG_CHANNEL_PREFIX` and any logging channels.  I suppose I could make one, but that seemed like a lot of overkill for something that should hopefully not happen.

Oh, forgot about that. Carry on!