1170493 - b2g process does not boot anymore when rebooting the device if the user installs a service worker enabled app

Status: RESOLVED DUPLICATE of bug 1171486

(Core :: DOM: Service Workers, defect)

Description

[José Antonio Olivera Ortega [:jaoo]]

STR required b2g:

1. In the browser app, opens appA.
2. appA registers a SW.
3. Install AppA.
4. Open appA directly, appA registers SW.
5. Reboot the device

Expected

The device boots correctly

Actual result

The device doesn't boot correctly. I see:

F/libc    ( 6320): Fatal signal 11 (SIGSEGV) at 0x00000028 (code=1), thread 6320 (b2g)

This issue needs some debugging with gdb.

Note: patches from bug 1162088 need to be applied.

Comment 1

[(no longer active)]

I found a startup crash that I filed as bug 1171486.  Could this be a dupe?

Comment 2

[José Antonio Olivera Ortega [:jaoo]]

(In reply to :Ehsan Akhgari (not reading bugmail, needinfo? me!) from comment #1)
> I found a startup crash that I filed as bug 1171486.  Could this be a dupe?

Likely. See what I see with a B2G debug build.

I/Gecko   ( 1194): [Parent 1194] ###!!! ASSERTION: Recursive GetService!: 'Error', file ../../../xpcom/components/nsComponentManager.cpp, line 1546
F/MOZ_Assert( 1194): Assertion failure: swm, at /Volumes/firefoxos/dev/mozilla-central/dom/workers/ServiceWorkerManager.cpp:2614
F/libc    ( 1194): Fatal signal 11 (SIGSEGV) at 0x00000000 (code=1), thread 1194 (b2g)

Let me check the patch and I'll let you know.

Comment 3

[(no longer active)]

(In reply to José Antonio Olivera Ortega [:jaoo] from comment #2)
> (In reply to :Ehsan Akhgari (not reading bugmail, needinfo? me!) from
> comment #1)
> > I found a startup crash that I filed as bug 1171486.  Could this be a dupe?
> 
> Likely. See what I see with a B2G debug build.
> 
> I/Gecko   ( 1194): [Parent 1194] ###!!! ASSERTION: Recursive GetService!:
> 'Error', file ../../../xpcom/components/nsComponentManager.cpp, line 1546
> F/MOZ_Assert( 1194): Assertion failure: swm, at
> /Volumes/firefoxos/dev/mozilla-central/dom/workers/ServiceWorkerManager.cpp:
> 2614
> F/libc    ( 1194): Fatal signal 11 (SIGSEGV) at 0x00000000 (code=1), thread
> 1194 (b2g)

Yep, that is exactly the same bug!

Comment 4

[José Antonio Olivera Ortega [:jaoo]]

(In reply to José Antonio Olivera Ortega [:jaoo] from comment #2)
> (In reply to :Ehsan Akhgari (not reading bugmail, needinfo? me!) from
> comment #1)
> > I found a startup crash that I filed as bug 1171486.  Could this be a dupe?
> 
> Likely. See what I see with a B2G debug build.

> Let me check the patch and I'll let you know.

Yes, patches from bug 1171486 fix the issue. Thanks!

Comment 5

[Noemí Freire (:noemi)]

Just adding that this issues is not reproducible anymore in m-c once Bug 1171486 has been landed. Tested with today's (6/8) master build and the device successfully boots (step 5 in the description). Environmental variables:
Flame device
Build Id: 20150608073530
Gecko: b83b4f3
Gaia: 4d1f2a4
Platform version: 41.0a1

Thanks!