1589821 - Many errors of the form - WARNING: '!inner', file .../dom/ipc/JSWindowActorService.cpp, line 172

Status: NEW

(Thunderbird :: Testing Infrastructure, defect)

Description

[ISHIKAWA, Chiaki]

During |make mozmill| test suite run of C-C TB (Full Debug version) locally, I noticed MANY
errors of the form
WARNING: '!inner', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/dom/ipc/JSWindowActorService.cpp, line 172

(I also noticed that it happens in |mach xpcshell-test| but much fewer, only a few?)

This error was not seen on Oct 11th locally, so very new.

I looked at the source code and it did not seem to have changed lately. But something did elsewhere.

158 /**
159 * This listener only listens for events for the child side of the protocol.
160 * This will work in both content and parent processes.
161 /
162 NS_IMETHODIMP JSWindowActorProtocol::HandleEvent(Event aEvent) {
163 // Determine which inner window we're associated with, and get its
164 // WindowGlobalChild actor.
165 EventTarget* target = aEvent->GetOriginalTarget();
166 if (NS_WARN_IF(!target)) {
167 return NS_ERROR_FAILURE;
168 }
169
170 nsCOMPtr<nsPIDOMWindowInner> inner =
171 do_QueryInterface(target->GetOwnerGlobal());
172 if (NS_WARN_IF(!inner)) { <==== THIS
173 return NS_ERROR_FAILURE;
174 }
175
176 RefPtr<WindowGlobalChild> wgc = inner->GetWindowGlobalChild();
177 if (NS_WARN_IF(!wgc)) {
178 return NS_ERROR_FAILURE;
179 }
180

In my last count of the error in my LOCAL log, it appears 671 times and just too many to ignore as a newcomer.
The error can be observed in tryserver job as well.
For example,
The submission: https://treeherder.mozilla.org/#/jobs?repo=try-comm-central&revision=66ef3c2b0a7daf6d91aba1af924ca769f8cd282e

linux x64 debug mozmill log (z1): live.log
https://taskcluster-artifacts.net/S0wCWlHyRV2RuoC3B3I-nA/0/public/logs/live_backing.log
In this particular log, it may look to be related to editing task.
But if you look at z2, z3, and z4 log, I am not so sure.

In any case, since the error just has begun to pop up, I think it is worthwhile to record this.

I have no idea which component against which I should file this bugzilla.
I also suspect that this happens on other architecture, etc.

Comment 1

[ISHIKAWA, Chiaki]

I also suspect that this happens on other architecture, etc.

It does. For example, you can check Windows 10 x64 debug version live.log
https://taskcluster-artifacts.net/FWJBw21aSWa7VaQiNEcW8Q/0/public/logs/live_backing.log
There, it does not seem to be related to editing task so strongly, for example.

Comment 2

[Jorg K (CEST = GMT+2)]

Nika, this warning
https://searchfox.org/mozilla-central/rev/696cf3b52a9aa04aa5013008a8b448904a298356/dom/ipc/JSWindowActorService.cpp#172
which was added in Feb. 2019 here
https://hg.mozilla.org/mozilla-central/rev/3575c9a372e7#l3.182
has become quite noisy. Is that expected? Looks like it happens in Thunderbirds compose window.

Comment 3

[Nika Layzell [:nika] (ni? for response)]

That warning isn't really an error, and should probably not emit log-spew. The NS_WARN_IF should probably be removed.

It's likely caused by some event being fired directly on the InProcessBrowserChildMessageManager object, rather than being fired on the window.

Comment 4

[ISHIKAWA, Chiaki]

(In reply to :Nika Layzell (ni? for response) from comment #3)

That warning isn't really an error, and should probably not emit log-spew. The NS_WARN_IF should probably be removed.

It's likely caused by some event being fired directly on the InProcessBrowserChildMessageManager object, rather than being fired on the window.

So that means something is incorrectly firing events to unexpecting object?

Comment 5

[Chris Peterson [:cpeterson]]

Tentatively moving all bugs whose summaries mention "Fission" (or other Fission-related keywords) but are not assigned to a Fission Milestone to the "?" triage milestone.

This will generate a lot of bugmail, so you can filter your bugmail for the following UUID and delete them en masse:

0ee3c76a-bc79-4eb2-8d12-05dc0b68e732

Comment 6

[ISHIKAWA, Chiaki]

As of today, the number warnings recorded from |make mozmill| local run of C-C TB (full debug version) is
No.4 in the most frequent warning messages:

========================================
Warning list

992: Main Thread] WARNING: Failed to retarget HTML data delivery to the parser thread.: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/parser/html/nsHtml5StreamParser.cpp, line 1142

990: Main Thread] WARNING: '!ClientIsValidPrincipalInfo(mClientInfo.PrincipalInfo())', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/dom/clients/manager/ClientSource.cpp, line 176

812: Main Thread] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/xpcom/base/nsTraceRefcnt.cpp, line 198

X 616: Main Thread] WARNING: '!inner', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/dom/ipc/JSWindowActorService.cpp, line 181

272: Main Thread] WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80550006: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/comm/mailnews/local/src/nsMailboxService.cpp, line 616

272: Main Thread] WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80550006: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/comm/mailnews/base/util/nsMsgDBFolder.cpp, line 4794

---

Comment 7

[ISHIKAWA, Chiaki]

|make mozmill| is gone from TB test suite. mochitest is here to stay.

In terms of frequency of occurrences, this warning is still no.4 in terms of frequency of appearances in mochitest log.

In the following, I tried to separate the warnings observed with /libeditor/ files and other files.

========================================
Warning list
except for "Failed to open" which is checked elsewhere
========================================


   ========================================
   Firstly, warnings under /libeditor/ 
   ========================================

2883: Main Thread] WARNING: HTMLEditor::GetAbsolutelyPositionedSelectionContainer() reached <html> element: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/editor/libeditor/HTMLAbsPositionEditor.cpp, line 91
314: Main Thread] WARNING: mozInlineSpellChecker::Cleanup() failed, but ignored: 'NS_SUCCEEDED(rvIgnored)', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/editor/libeditor/EditorBase.cpp, line 548
139: Main Thread] WARNING: '!aNode', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/editor/libeditor/EditorBase.cpp, line 3049

... omission ...

    ========================================
    Other than /libeditor/
    ========================================

1009: Main Thread] WARNING: '!ClientIsValidPrincipalInfo(mClientInfo.PrincipalInfo())', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/dom/clients/manager/ClientSource.cpp, line 175
1004: Main Thread] WARNING: Failed to retarget HTML data delivery to the parser thread.: file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/parser/html/nsHtml5StreamParser.cpp, line 1131
900: Main Thread] WARNING: '!inner', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/dom/ipc/JSWindowActorProtocol.cpp, line 172
314: Main Thread] WARNING: '!mTextEditor', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/extensions/spellcheck/src/mozInlineSpellChecker.cpp, line 660
314: Main Thread] WARNING: '!mTextEditor', file /NEW-SSD/NREF-COMM-CENTRAL/mozilla/extensions/spellcheck/src/mozInlineSpellChecker.cpp, line 1644
 ... omission.

Just FYI as of May 13.