Crash in mozilla::dom::ServiceWorkerRegistrar::GetShutdownPhase[STALLED]
Categories
(Core :: DOM: Service Workers, defect, P2)
Tracking
()
Tracking | Status | |
---|---|---|
firefox-esr52 | --- | unaffected |
firefox-esr68 | --- | wontfix |
firefox55 | --- | unaffected |
firefox56 | --- | unaffected |
firefox57 | --- | wontfix |
firefox58 | --- | wontfix |
firefox59 | --- | wontfix |
firefox60 | --- | wontfix |
firefox61 | --- | wontfix |
firefox62 | --- | wontfix |
firefox63 | --- | wontfix |
firefox64 | --- | wontfix |
firefox65 | --- | wontfix |
firefox66 | --- | wontfix |
firefox67 | --- | wontfix |
firefox68 | --- | wontfix |
firefox69 | --- | wontfix |
firefox70 | --- | wontfix |
firefox71 | --- | wontfix |
firefox72 | --- | wontfix |
People
(Reporter: philipp, Assigned: kmag)
References
(Blocks 1 open bug)
Details
(4 keywords)
Crash Data
Attachments
(4 files, 3 obsolete files)
Assignee | ||
Comment 1•7 years ago
|
||
Updated•7 years ago
|
Comment 2•7 years ago
|
||
Assignee | ||
Comment 3•7 years ago
|
||
Comment hidden (mozreview-request) |
Comment 5•7 years ago
|
||
mozreview-review |
Comment 6•7 years ago
|
||
mozreview-review |
Assignee | ||
Comment 7•7 years ago
|
||
Assignee | ||
Comment 8•7 years ago
|
||
mozreview-review-reply |
Updated•7 years ago
|
Comment 9•7 years ago
|
||
mozreview-review |
Assignee | ||
Comment 10•7 years ago
|
||
Assignee | ||
Comment 11•7 years ago
|
||
Assignee | ||
Updated•7 years ago
|
Comment 12•7 years ago
|
||
bugherder |
Comment 13•7 years ago
|
||
Assignee | ||
Comment 14•7 years ago
|
||
Assignee | ||
Comment 15•7 years ago
|
||
Comment 17•7 years ago
|
||
uplift |
Updated•7 years ago
|
Comment 18•7 years ago
|
||
Assignee | ||
Comment 19•7 years ago
|
||
Assignee | ||
Comment 20•7 years ago
|
||
Assignee | ||
Comment 21•7 years ago
|
||
Comment 22•7 years ago
|
||
Assignee | ||
Comment 23•7 years ago
|
||
Comment hidden (mozreview-request) |
Assignee | ||
Updated•7 years ago
|
Comment hidden (mozreview-request) |
Assignee | ||
Updated•7 years ago
|
Assignee | ||
Updated•7 years ago
|
Comment hidden (mozreview-request) |
Comment 27•7 years ago
|
||
mozreview-review |
Assignee | ||
Comment 28•7 years ago
|
||
Assignee | ||
Comment 29•7 years ago
|
||
Comment 30•7 years ago
|
||
Assignee | ||
Comment 31•7 years ago
|
||
Comment 32•7 years ago
|
||
bugherder |
Comment 33•7 years ago
|
||
mozreview-review |
Comment 35•7 years ago
|
||
uplift |
Assignee | ||
Updated•7 years ago
|
Assignee | ||
Comment 36•7 years ago
|
||
Comment hidden (mozreview-request) |
Comment hidden (mozreview-request) |
Comment hidden (mozreview-request) |
Comment 40•7 years ago
|
||
mozreview-review |
Comment 41•7 years ago
|
||
mozreview-review |
Updated•7 years ago
|
Assignee | ||
Comment 42•7 years ago
|
||
Assignee | ||
Comment 43•7 years ago
|
||
mozreview-review-reply |
Assignee | ||
Comment 44•7 years ago
|
||
Comment 45•7 years ago
|
||
bugherder |
Comment 46•7 years ago
|
||
mozreview-review |
Comment 47•7 years ago
|
||
Comment 48•7 years ago
|
||
Assignee | ||
Comment 49•7 years ago
|
||
Assignee | ||
Comment 50•7 years ago
|
||
Comment 51•7 years ago
|
||
Assignee | ||
Updated•7 years ago
|
Assignee | ||
Comment 52•7 years ago
|
||
Comment 53•7 years ago
|
||
bugherder |
Updated•7 years ago
|
Assignee | ||
Comment 54•7 years ago
|
||
Comment 55•7 years ago
|
||
Assignee | ||
Comment 56•7 years ago
|
||
Comment 57•6 years ago
|
||
Assignee | ||
Comment 58•6 years ago
|
||
Comment 59•6 years ago
|
||
Comment 60•6 years ago
|
||
Assignee | ||
Comment 61•6 years ago
|
||
Assignee | ||
Comment 62•6 years ago
|
||
Updated•6 years ago
|
Comment 64•6 years ago
|
||
Comment 65•6 years ago
|
||
Updated•6 years ago
|
Assignee | ||
Updated•6 years ago
|
Updated•6 years ago
|
Updated•6 years ago
|
Updated•6 years ago
|
Comment 68•5 years ago
|
||
Hey :kmag, in bug 1553401 when looking at the raw crash JSON for https://crash-stats.mozilla.org/report/index/aee3cc64-bb79-4e9a-a03f-1e09f0190522 I noticed that we still seem to be trying to annotate the contents of "nsAsyncShutdown.js"[1] but you renamed that[2] in bug 1524688. The crash report in question seems to think it has the contents of nsAsyncShutdown.js.
I guess my Q is:
- Is there anything informative in that bug's crash/friend crashes.
- Should we be updating the annotation code?
1: https://searchfox.org/mozilla-central/rev/7556a400affa9eb99e522d2d17c40689fa23a729/js/xpconnect/loader/mozJSComponentLoader.cpp#371
2: https://hg.mozilla.org/mozilla-central/rev/fe9b6695212ec3960ab9e730ccc446a8c3da9375#l6.1
Updated•5 years ago
|
Assignee | ||
Comment 69•5 years ago
|
||
(In reply to Andrew Sutherland [:asuth] (he/him) from comment #68)
Hey :kmag, in bug 1553401 when looking at the raw crash JSON for https://crash-stats.mozilla.org/report/index/aee3cc64-bb79-4e9a-a03f-1e09f0190522 I noticed that we still seem to be trying to annotate the contents of "nsAsyncShutdown.js"[1] but you renamed that[2] in bug 1524688. The crash report in question seems to think it has the contents of nsAsyncShutdown.js.
I guess my Q is:
- Is there anything informative in that bug's crash/friend crashes.
- Should we be updating the annotation code?
I don't know, at this point. I'd still like to see the contents of those files, but I asked Andrew McCreight to get the annotations for me and he said they weren't there. If there actually are crash reports with that annotation data now, then I'll update the annotation code and maybe we can actually get some useful information from them.
If not, I guess the annotation code should just be removed.
Assignee | ||
Comment 70•5 years ago
|
||
(In reply to Andrew Sutherland [:asuth] (he/him) from comment #68)
Hey :kmag, in bug 1553401 when looking at the raw crash JSON for https://crash-stats.mozilla.org/report/index/aee3cc64-bb79-4e9a-a03f-1e09f0190522 I noticed that we still seem to be trying to annotate the contents of "nsAsyncShutdown.js"[1] but you renamed that[2] in bug 1524688. The crash report in question seems to think it has the contents of nsAsyncShutdown.js.
So, interestingly, based on this crash, it looks like the user is running a newer Firefox binary with an omni.ja from an older version. It has the actual contents of the old version of nsAsyncShutdown.js
. Which would explain why loading the component would fail, since the JSM wouldn't actually exist at the new location we try to load it from.
I suppose that may be the result of an incomplete update attempt or a botched system restore of some sort.
But it's also different from the other errors I've seen, where a lot of the time we're failing with an undefined symbol error for a symbol which is clearly a corrupted version of the symbol the script is supposed to be referencing.
I suppose maybe both types of errors could possibly be explained by a partial MAR update gone wrong...
Comment 71•5 years ago
|
||
This is the #5 overall crash so far in 68.0b13.
Assignee | ||
Comment 72•5 years ago
|
||
Just to be clear, at this point we have pretty high confidence that this crash is caused by some sort of corruption that just happens to show up here. It doesn't actually have anything to do with the ServiceWorker code. There's omni.ja (or possibly in some cases memory) corruption that's going to make the browser crash or at least be unusably broken.
It's still not clear whether there's more than one source of corruption, though.
Comment 73•5 years ago
|
||
Good day. I was having this problem with starting my Firefox 32 bit (I think)
https://crash-stats.mozilla.org/report/index/17dfacfa-391a-4883-b5b8-58ef80190803
As a test, I installed the 64 bit and it works.
Comment 74•5 years ago
|
||
Adding updated affected branches, based on the fact this is seen in the top 25 crashes in 70.0b5.
Updated•5 years ago
|
Comment 75•5 years ago
|
||
Stalled and unassigned, marking as fix-optional for 71 as we are in beta now.
Updated•5 years ago
|
Comment 77•5 years ago
|
||
(In reply to Kris Maglione [:kmag] from comment #72)
Just to be clear, at this point we have pretty high confidence that this crash is caused by some sort of corruption that just happens to show up here. It doesn't actually have anything to do with the ServiceWorker code. There's omni.ja (or possibly in some cases memory) corruption that's going to make the browser crash or at least be unusably broken.
Does this mean, we can/should change the components assignment accordingly?
Comment 78•5 years ago
|
||
:rstrong, would it make more sense for this bug to be moved to some installer component where it can hopefully make more progress given that the corruption presumably involves app update?
The general situation here is that, at startup, the ServiceWorkerRegistrar tries to get the the nsIAsyncShutdownService ("@mozilla.org/async-shutdown-service;1") so it can register itself as a blocker. The service is implemented in JS and fails to load and by all indications the load is corrupt. This manifests itself in ServiceWorkers in crash-stats, but the underlying problem is systemic
Comment 79•5 years ago
|
||
I just checked crash-stats and this bug also affects FennecAndroid which doesn't use Application Update so there must be something else causing this other than Application Update... perhaps a cache hasn't been invalidated.
Updated•5 years ago
|
Updated•5 years ago
|
Comment 80•4 years ago
|
||
Duping to bug 1471720 where there is a plan at https://bugzilla.mozilla.org/show_bug.cgi?id=1471720#c18 to help move the attribution out of ServiceWorkers, but which won't magically fix omni.ja corruption.
Updated•4 years ago
|
Comment 81•4 years ago
|
||
Since the bug is closed, the stalled keyword is now meaningless.
For more information, please visit auto_nag documentation.
Updated•4 years ago
|
Assignee | ||
Comment 82•4 years ago
|
||
They've done their job. It's clear at this point that these failures are
happening because of omni.ja corruption.
Updated•4 years ago
|
Comment 83•4 years ago
|
||
Comment 84•4 years ago
|
||
Backed out for causing xpc failures in AsyncShutdown
Backout link: https://hg.mozilla.org/integration/autoland/rev/f830bcabd026282e14d48aabe0fedbe210509c82
Failure log: https://treeherder.mozilla.org/logviewer?job_id=330821338&repo=autoland&lineNumber=3792
Comment 85•4 years ago
|
||
Comment 86•4 years ago
|
||
bugherder |
Assignee | ||
Updated•4 years ago
|
Updated•2 years ago
|
Description
•