1113957 - ServiceWorker unregistration should also use the FIFO job queue

Status: RESOLVED FIXED

(Core :: DOM: Workers, defect)

Description

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

Otherwise

register('/blah', ...).then(function(swr) { swr.unregister(); }) has no-op behaviour due to the SW still being in the installing stage.

Comment 1

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

Attachment: ServiceWorker unregistration uses job queue

Comment 2

[Andrea Marchesini [:baku]]

Comment on attachment 8553348 [details] [diff] [review]
ServiceWorker unregistration uses job queue

Review of attachment 8553348 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/workers/ServiceWorkerManager.cpp
@@ +1446,5 @@
> +  {
> +    AssertIsOnMainThread();
> +    nsCOMPtr<nsIRunnable> r =
> +      NS_NewRunnableMethod(this, &ServiceWorkerUnregisterJob::UnregisterAndDone);
> +    DebugOnly<nsresult> ok = NS_DispatchToMainThread(r);

MOZ_ALWAYS_TRUE(NS_SUCCEEDED(NS_DispatchToMainThread(r)));

It's exactly like yours. Up to you. Lately I prefer this way to write this check.

::: dom/workers/test/serviceworkers/test_unregister.html
@@ -86,5 @@
>  
>    function runTest() {
>      simpleRegister()
> -      .then(function(v) {
> -        info("simpleRegister() promise resolved");

Why have you removed all this useful info messages?

Comment 3

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

(In reply to Andrea Marchesini (:baku) from comment #2)
> Comment on attachment 8553348 [details] [diff] [review]
> ServiceWorker unregistration uses job queue
> 
> Review of attachment 8553348 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> ::: dom/workers/ServiceWorkerManager.cpp
> @@ +1446,5 @@
> > +  {
> > +    AssertIsOnMainThread();
> > +    nsCOMPtr<nsIRunnable> r =
> > +      NS_NewRunnableMethod(this, &ServiceWorkerUnregisterJob::UnregisterAndDone);
> > +    DebugOnly<nsresult> ok = NS_DispatchToMainThread(r);
> 
> MOZ_ALWAYS_TRUE(NS_SUCCEEDED(NS_DispatchToMainThread(r)));
> 
> It's exactly like yours. Up to you. Lately I prefer this way to write this
> check.
> 
> ::: dom/workers/test/serviceworkers/test_unregister.html
> @@ -86,5 @@
> >  
> >    function runTest() {
> >      simpleRegister()
> > -      .then(function(v) {
> > -        info("simpleRegister() promise resolved");
> 
> Why have you removed all this useful info messages?

They were debugging messages to ensure the promises were getting resolved. I never intended to land them in the first place. They are just noise unless a test is consistently failing/timing out.

Comment 4

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

Attachment: MozReview Request: bz://1113957/nsm

/r/2921 - Bug 1113957 - ServiceWorker unregistration uses job queue. r=baku
/r/2923 - Bug 1081293 - Silently ignore lack of ServiceWorkerManager when shutting down ServiceWorkerContainer.

Pull down these commits:

hg pull review -r 1c5b76b58327ba8c6c536d18af43f9788ce477ee

Comment 6

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

Attachment: Do not assert job queue empty at shutdown

Comment 7

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/67d15be29a07

Comment 8

[Phil Ringnalda (:philor)]

https://hg.mozilla.org/mozilla-central/rev/67d15be29a07

Comment 9

[Nikhil Marathe [:nsm] (No longer reading bugmail, please needinfo?)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/237edb7e4ae5

Comment 10

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/237edb7e4ae5