crash in js::ion::Cannon (from ROTOGAME app)

RESOLVED INCOMPLETE

Status

()

--
critical
RESOLVED INCOMPLETE
6 years ago
3 years ago

People

(Reporter: eviljeff, Unassigned)

Tracking

({crash, regression})

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: [native-crash], crash signature)

(Reporter)

Description

6 years ago
This bug was filed from the Socorro interface and is 
report bp-1bb6ea40-fc2b-4968-82e6-31cde2121116 .
============================================================= 


the app content can be accessed directly at http://sq.rotogame.com

Updated

6 years ago
Assignee: nobody → general
Component: Web Apps → JavaScript Engine
Product: Firefox for Android → Core
QA Contact: aaron.train

Updated

6 years ago
Hardware: All → ARM
Whiteboard: [native-crash]
Version: unspecified → 18 Branch
Andrew - Can you consistently reproduce this crash?
Hardware: ARM → All
Whiteboard: [native-crash]
Version: 18 Branch → unspecified
(Reporter)

Comment 2

6 years ago
(In reply to Jason Smith [:jsmith] from comment #1)
> Andrew - Can you consistently reproduce this crash?

Yeah.  Just playing the game for a short while on Android* produces it.  Doesn't appear to happen on Desktop or FxOS.

*(on Transformer TF101 Tablet and a Galaxy S2)
Probably going to need a regression window.
Keywords: regressionwindow-wanted

Updated

6 years ago
Keywords: regression
Hardware: All → ARM
Whiteboard: [native-crash]

Comment 4

6 years ago
As the author of the app (ROTOGAMEsq), if I can help with anything, then let me know. I only have an Android 2.3.5 device at hand, though: Here the app runs fine, except that *sometimes* I am sent back to the home screen right during (or maybe directly after) the splash screen (icon + rotating animation). I am using Firefox Aurora for ARMv6, which is experimental:

https://wiki.mozilla.org/Mobile/Platforms/Android#ARMv6_.28experimental.29
> I am using Firefox Aurora for ARMv6, which is experimental:
> 
> https://wiki.mozilla.org/Mobile/Platforms/Android#ARMv6_.28experimental.29

Does the game crash for you? IonMonkey should be disabled on ARMv6 builds, so I would not expect you to ever see this particular crash.

Comment 6

6 years ago
> Does the game crash for you?

I just checked and right during or at the end of the splash screen I got
a message: "Aurora has crashed. [...]" (agreed to send data to Mozilla).
Also sometimes, as said above, I get sent back to the home screen from
the splash screen. Once the game is up, however, I have never seen a
crash.

I don't know if this is specific to my app, or if there is a general
issue with what happens while the splash screen is showing.

Comment 7

6 years ago
(In reply to Felix E. Klee from comment #6)
> > Does the game crash for you?
> I just checked and right during or at the end of the splash screen I got
> a message: "Aurora has crashed. [...]" (agreed to send data to Mozilla).
What is the crash ID (bp...) in the about:crashes page?

Comment 9

6 years ago
(In reply to Felix E. Klee from comment #8)
> It should be: bp-d4fd4fd1-cb9e-4d4f-ae6c-9b5e12121120

That's a different signature, which is tracked in bug 786815 - but thanks for noting this!
See also bp-91393daa-a2f7-457d-b372-6f39f2121214 and bug 821625, for the same crash (I believe) with a different game.

Comment 11

6 years ago
It think it's fixed by bug 805421.
Does it happen in Nightly?
Flags: needinfo?(awilliamson)
(Reporter)

Comment 12

6 years ago
(In reply to Scoobidiver from comment #11)
> It think it's fixed by bug 805421.
> Does it happen in Nightly?

The game doesn't crash any more on Aurora for me (has the patch been applied to that branch also?).  Its unresponsive but this is possibly/probably a different bug.
Flags: needinfo?(awilliamson)

Comment 13

6 years ago
(In reply to Andrew Williamson [:eviljeff] from comment #12)
> has the patch been applied to that branch also?
No, only in Nightly.
Can you check in Nightly (http://nightly.mozilla.org/)?

Comment 14

6 years ago
As the author of ROTOGAMEsq, I cannot reproduce the unresponsiveness on experimental Firefox Aurora for ARMv6 (Huawei Sonic). What do you mean by unresponsive? Does the game load slowly, or are animations slow, etc.?

Comment 15

6 years ago
I just installed Nightly for ARMv6. Right after installation, when loading ROTOGAMEsq for the first time, it was indeed unresponsive. That *may* have have been caused by background processes running on that particular device. Because:

After restarting the device and clearing the cache in Nightly, I cannot reproduce the problem anymore: ROTOGAMEsq runs smoothly in Firefox Nightly on a Huawei Sonic (ARMv6).
(Reporter)

Comment 16

6 years ago
(In reply to Felix E. Klee from comment #14)
> As the author of ROTOGAMEsq, I cannot reproduce the unresponsiveness on
> experimental Firefox Aurora for ARMv6 (Huawei Sonic). What do you mean by
> unresponsive? Does the game load slowly, or are animations slow, etc.?

The app didn't respond to any touches or drags.  Occasionally I'd glance back at the screen and something had registered (e.g. a rotation of some of the blocks, or a puzzle change) but it was so far after a series of random gestures I wasn't sure which worked.  When I tested it worked well on the (lower spec'd) FirefoxOS device, or I might have assumed it was broken entirely.

@jsmith - can you replicate?

Comment 17

6 years ago
Weird. The app is specifically designed to run on low speced devices and to be energy efficient. For animations, `requestAnimationFrame` is used, and redraws should only happen where necessary.

Updated

6 years ago
Depends on: 805421
(Assignee)

Updated

4 years ago
Assignee: general → nobody
Firefox for Android no longer supports armv6. Game appears to work correctly without crashing on my HTC M8. Closing as incomplete.
Status: NEW → RESOLVED
Last Resolved: 3 years ago
Keywords: regressionwindow-wanted
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.