Closed
Bug 1383876
Opened 7 years ago
Closed 6 years ago
Remove remaining link "prerender" code
Categories
(Core :: DOM: Core & HTML, enhancement, P3)
Core
DOM: Core & HTML
Tracking
()
RESOLVED
FIXED
mozilla59
Tracking | Status | |
---|---|---|
firefox59 | --- | fixed |
People
(Reporter: nika, Assigned: nika)
References
Details
(Keywords: dev-doc-complete)
Attachments
(5 files, 4 obsolete files)
210.87 KB,
patch
|
smaug
:
review+
|
Details | Diff | Splinter Review |
94.32 KB,
patch
|
Details | Diff | Splinter Review | |
8.94 KB,
patch
|
Details | Diff | Splinter Review | |
11.92 KB,
patch
|
Details | Diff | Splinter Review | |
975 bytes,
patch
|
freesamael
:
review+
|
Details | Diff | Splinter Review |
We don't intend to finish the implementation which was added in bug 1315105, so we should remove the code which was added in that bug and related bugs. However, we do want to keep the GroupedSHistory work which :freesamael worked on, as we may want to extend it in the future to improve cross-process loads in e10s-multi. This bug is to track removing the rest of that code.
Updated•7 years ago
|
Priority: -- → P3
Updated•7 years ago
|
Keywords: dev-doc-needed,
site-compat
Updated•7 years ago
|
Keywords: site-compat
Assignee | ||
Comment 1•7 years ago
|
||
This was never shipped or documented anywhere as far as I know - I doubt we'll need any dev-doc work.
Comment 2•7 years ago
|
||
(In reply to Michael Layzell [:mystor] from comment #1) > This was never shipped or documented anywhere as far as I know - I doubt > we'll need any dev-doc work. See https://bugzilla.mozilla.org/show_bug.cgi?id=1315105#c63
Keywords: dev-doc-needed → dev-doc-complete
Assignee | ||
Comment 3•6 years ago
|
||
I'm taking this to remove this unused code to start working towards bug 1419857.
Assignee: nobody → nika
Assignee | ||
Comment 4•6 years ago
|
||
This is the first part of removing as much of the prerendering logic as I could find from the code. It would be good to go over this & make sure that we actually wanted to remove everything which I remove here. The second patch also removes a bunch of tests which we also need to make sure we actually want to remove. If you notice something I totaly missed let me know. I'm doing a try push with this right now. MozReview-Commit-ID: 2aHA6NcQPGk
Attachment #8936944 -
Flags: review?(sawang)
Assignee | ||
Comment 5•6 years ago
|
||
MozReview-Commit-ID: KwvOcgQKheK
Attachment #8936945 -
Flags: review?(sawang)
Assignee | ||
Comment 6•6 years ago
|
||
I found more ^.^ MozReview-Commit-ID: IuJHv3n0Uob
Attachment #8936949 -
Flags: review?(sawang)
Assignee | ||
Comment 7•6 years ago
|
||
MozReview-Commit-ID: 5j8fNeKuHAM
Attachment #8936950 -
Flags: review?(sawang)
Comment 8•6 years ago
|
||
Comment on attachment 8936944 [details] [diff] [review] Part 1: Remove GroupedSHistory and Prerendering logic from C++ code Review of attachment 8936944 [details] [diff] [review]: ----------------------------------------------------------------- I tried my best... ::: docshell/shistory/nsSHistory.cpp @@ +1651,1 @@ > } else { Remove the `else` since it early returns above. ::: dom/base/nsFrameLoader.cpp @@ +364,2 @@ > bool > nsFrameLoader::SwapBrowsersAndNotify(nsFrameLoader* aOther) Do you plan to keep `BrowserWillChangeProcess` & `BrowserChangedProcess` stuff? Otherwise this can be removed I think. @@ +401,2 @@ > already_AddRefed<Promise> > nsFrameLoader::FireWillChangeProcessEvent() Same here. @@ +437,2 @@ > void > nsFrameLoader::AddProcessChangeBlockingPromise(Promise& aPromise, ErrorResult& aRv) And here. And the whole `GroupedHistoryEvent` class + webidl. ::: dom/base/nsFrameLoader.h @@ +468,3 @@ > // A stack-maintained reference to an array of promises which are blocking > // grouped history navigation > nsTArray<RefPtr<mozilla::dom::Promise>>* mBrowserChangingProcessBlockers; Same here. ::: dom/ipc/TabChild.cpp @@ +3510,2 @@ > already_AddRefed<nsISHistory> > TabChild::GetRelatedSHistory() Looks like we removed all its callers. ::: dom/webidl/FrameLoader.webidl @@ -12,2 @@ > interface nsIMessageSender; > -interface nsIPartialSHistory; We can remove these 2 bindings: https://searchfox.org/mozilla-central/rev/110706c3c09d457dc70293b213d7bccb4f6f5643/dom/bindings/Bindings.conf#1742-1743,1750-1751 ::: dom/xul/nsXULElement.cpp @@ -1491,5 @@ > (new AsyncEventDispatcher(this, > NS_LITERAL_STRING("XULFrameLoaderCreated"), > /* aBubbles */ true))->RunDOMEventWhenSafe(); > - > - if (AttrValueIs(kNameSpaceID_None, nsGkAtoms::prerendered, And the GkAtoms. Do you happen to know how the rust atoms mapping are generated? We may want to also remove this one: https://searchfox.org/mozilla-central/rev/110706c3c09d457dc70293b213d7bccb4f6f5643/servo/components/style/gecko/generated/atom_macro.rs#2220
Attachment #8936944 -
Flags: review?(sawang) → review+
Comment 9•6 years ago
|
||
Comment on attachment 8936949 [details] [diff] [review] Part 3: Remove prerendering methods from nsIWebBrowserChrome3.idl Review of attachment 8936949 [details] [diff] [review]: ----------------------------------------------------------------- One missing implementation: https://searchfox.org/mozilla-central/rev/110706c3c09d457dc70293b213d7bccb4f6f5643/toolkit/components/extensions/ext-browser-content.js#315-320
Attachment #8936949 -
Flags: review?(sawang) → review+
Assignee | ||
Comment 10•6 years ago
|
||
FYI: I want to keep around the GroupedHistoryEvent, BrowserChangedProcess, BrowserWillChangeProcess and the process blockers, because I'm going to want to use them in the future. I will probably rename GroupedHistoryEvent.
Updated•6 years ago
|
Attachment #8936945 -
Flags: review?(sawang) → review+
Updated•6 years ago
|
Attachment #8936950 -
Flags: review?(sawang) → review+
Assignee | ||
Comment 11•6 years ago
|
||
MozReview-Commit-ID: 2aHA6NcQPGk
Assignee | ||
Updated•6 years ago
|
Attachment #8936944 -
Attachment is obsolete: true
Assignee | ||
Updated•6 years ago
|
Attachment #8936945 -
Attachment is obsolete: true
Attachment #8936949 -
Attachment is obsolete: true
Attachment #8936950 -
Attachment is obsolete: true
Assignee | ||
Comment 12•6 years ago
|
||
MozReview-Commit-ID: KwvOcgQKheK
Assignee | ||
Comment 13•6 years ago
|
||
MozReview-Commit-ID: IuJHv3n0Uob
Assignee | ||
Comment 14•6 years ago
|
||
MozReview-Commit-ID: 5j8fNeKuHAM
Assignee | ||
Comment 15•6 years ago
|
||
MozReview-Commit-ID: 1HKNAeI7hSV
Assignee | ||
Updated•6 years ago
|
Attachment #8937131 -
Flags: review?(sawang)
Updated•6 years ago
|
Attachment #8937131 -
Flags: review?(sawang) → review+
Assignee | ||
Comment 16•6 years ago
|
||
Comment on attachment 8937127 [details] [diff] [review] Part 1: Remove GroupedSHistory and Prerendering logic from C++ code I forgot that Samael isn't a DOM peer (and neither am I), so I'm going to need you to sign off on my webidl changes. Nothing web visible has been changed.
Attachment #8937127 -
Flags: review?(bugs)
Comment 17•6 years ago
|
||
So we don't want to use this stuff to handle process switches properly? Or do we not want to switch process ever for a tab?
Comment 18•6 years ago
|
||
Comment on attachment 8937127 [details] [diff] [review] Part 1: Remove GroupedSHistory and Prerendering logic from C++ code r+ for the .webidl.
Attachment #8937127 -
Flags: review?(bugs) → review+
Assignee | ||
Comment 19•6 years ago
|
||
I went through this code again, and I'm pretty sure we don't want any of this stuff even if we want to have bfcache in the first version of improved process switching. I've updated the patch to apply on the new tree & I'm going to land it.
Comment 20•6 years ago
|
||
Pushed by nika@thelayzells.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/156c6bee1752 Part 1: Remove GroupedSHistory and Prerendering logic from C++ code, r=freesamael, r=smaug https://hg.mozilla.org/integration/mozilla-inbound/rev/156088fcf75b Part 2: Remove GroupedSHistory and Prerendering logic from JS code, r=freesamael https://hg.mozilla.org/integration/mozilla-inbound/rev/e89b001fd9f5 Part 3: Remove prerendering methods from nsIWebBrowserChrome3.idl, r=freesamael https://hg.mozilla.org/integration/mozilla-inbound/rev/c61976e2d3ac Part 4: Remove even more tests which no longer work, r=freesamael, r=smaug https://hg.mozilla.org/integration/mozilla-inbound/rev/6d9855f7bfdb Part 5: Clean up in moz.build files, r=freesamael
Comment 21•6 years ago
|
||
Pushed by nika@thelayzells.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/e0683e2979c0 Part 6: Fix eslint failures, a=bustage
Comment 22•6 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/156c6bee1752 https://hg.mozilla.org/mozilla-central/rev/156088fcf75b https://hg.mozilla.org/mozilla-central/rev/e89b001fd9f5 https://hg.mozilla.org/mozilla-central/rev/c61976e2d3ac https://hg.mozilla.org/mozilla-central/rev/6d9855f7bfdb https://hg.mozilla.org/mozilla-central/rev/e0683e2979c0
Status: NEW → RESOLVED
Closed: 6 years ago
status-firefox59:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla59
Updated•5 years ago
|
Component: DOM → DOM: Core & HTML
You need to log in
before you can comment on or make changes to this bug.
Description
•