WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
181942
SW: Make sure ServiceWorker loading and requests are correctly cleared by ITP
https://bugs.webkit.org/show_bug.cgi?id=181942
Summary
SW: Make sure ServiceWorker loading and requests are correctly cleared by ITP
youenn fablet
Reported
2018-01-22 13:10:26 PST
SW: Make sure ServiceWorker loading and requests are correctly cleared by ITP
Attachments
Patch
(1.95 KB, patch)
2018-01-22 13:15 PST
,
youenn fablet
no flags
Details
Formatted Diff
Diff
View All
Add attachment
proposed patch, testcase, etc.
youenn fablet
Comment 1
2018-01-22 13:10:52 PST
<
rdar://problem/35132091
>
youenn fablet
Comment 2
2018-01-22 13:15:26 PST
Created
attachment 331955
[details]
Patch
John Wilander
Comment 3
2018-01-22 13:17:24 PST
Comment on
attachment 331955
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=331955&action=review
Looks good to me. Not a reviewer though.
> Source/WebKit/UIProcess/WebResourceLoadStatisticsStore.cpp:62 > + WebsiteDataType::DOMCache,
This will make ITP start clearing DOM cache on non-SW too. I assume that's what you want.
youenn fablet
Comment 4
2018-01-22 13:23:10 PST
(In reply to John Wilander from
comment #3
)
> Comment on
attachment 331955
[details]
> Patch > > View in context: >
https://bugs.webkit.org/attachment.cgi?id=331955&action=review
> > Looks good to me. Not a reviewer though. > > > Source/WebKit/UIProcess/WebResourceLoadStatisticsStore.cpp:62 > > + WebsiteDataType::DOMCache, > > This will make ITP start clearing DOM cache on non-SW too. I assume that's > what you want.
Yes, DOMCache is not very different from IDB except that it is request/response specific.
Chris Dumez
Comment 5
2018-01-22 13:25:45 PST
Comment on
attachment 331955
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=331955&action=review
> Source/WebKit/UIProcess/WebResourceLoadStatisticsStore.cpp:73 > + WebsiteDataType::ServiceWorkerRegistrations,
Why do we want to clear SW registrations? To maintain any state, the SW would have to use an API such as DOMCache or IDB no?
youenn fablet
Comment 6
2018-01-22 13:27:38 PST
(In reply to Chris Dumez from
comment #5
)
> Comment on
attachment 331955
[details]
> Patch > > View in context: >
https://bugs.webkit.org/attachment.cgi?id=331955&action=review
> > > Source/WebKit/UIProcess/WebResourceLoadStatisticsStore.cpp:73 > > + WebsiteDataType::ServiceWorkerRegistrations, > > Why do we want to clear SW registrations? To maintain any state, the SW > would have to use an API such as DOMCache or IDB no?
John might know more but I would think that the JavaScript might contain a token that is specific for every fetch of the JavaScript that is not conditional.
John Wilander
Comment 7
2018-01-22 13:28:08 PST
(In reply to Chris Dumez from
comment #5
)
> Comment on
attachment 331955
[details]
> Patch > > View in context: >
https://bugs.webkit.org/attachment.cgi?id=331955&action=review
> > > Source/WebKit/UIProcess/WebResourceLoadStatisticsStore.cpp:73 > > + WebsiteDataType::ServiceWorkerRegistrations, > > Why do we want to clear SW registrations? To maintain any state, the SW > would have to use an API such as DOMCache or IDB no?
Can the registration itself can be used as state? Can a third party make registrations look different in any way across sites?
youenn fablet
Comment 8
2018-01-22 13:28:20 PST
> John might know more but I would think that the JavaScript might contain a > token that is specific for every fetch of the JavaScript that is not > conditional.
When JS fetch is conditional, it would obviously return a 304.
WebKit Commit Bot
Comment 9
2018-01-22 14:15:27 PST
Comment on
attachment 331955
[details]
Patch Clearing flags on attachment: 331955 Committed
r227358
: <
https://trac.webkit.org/changeset/227358
>
WebKit Commit Bot
Comment 10
2018-01-22 14:15:28 PST
All reviewed patches have been landed. Closing bug.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug