On Sep 11, 11:14 am, fadden wrote:
> On Sep 10, 6:20 pm, DaveG wrote:
>
> > With forcecopy enabled, I now get the crash in a different location:
>
> Interesting. Could be that whatever was trashing the virtual heap is
> now trashing the native heap. The guard areas
On Sep 11, 11:14 am, fadden wrote:
> On Sep 10, 6:20 pm, DaveG wrote:
>
> > With forcecopy enabled, I now get the crash in a different location:
>
> Interesting. Could be that whatever was trashing the virtual heap is
> now trashing the native heap. The guard areas
On Sep 10, 11:30 am, fadden wrote:
> An invalid DeleteLocalRef should cause the CheckJNI code to scream.
> And probably not cause a failure due to the way local refs are
> currently implemented.
Ok; no wonder I couldn't find any DeleteLocalRef problems.
> In my experience, crashes during GC are
On Sep 8, 1:25 pm, fadden wrote:
> On Sep 8, 12:26 pm, DaveG wrote:
>
> > My app is running into the following crash and I'm having a hard time
> > diagnosing it. We have multiple background threads (including a couple
> > pthreads started via JNI that are
My app is running into the following crash and I'm having a hard time
diagnosing it. We have multiple background threads (including a couple
pthreads started via JNI that are making calls back up to Java
occasionally), but none of them are reporting errors . Any suggestions
on how to go about isol
On Aug 24, 12:34 pm, Hong wrote:
> If you do: adb uninstall your.package.name.mainactivity,
> do you ever get any crash/error etc?
>
> If not, the upgrade/installation process will uninstall your app cleanly.
We do get crashes (that we're working on), but they don't matter
since our data files a
On Aug 21, 5:09 pm, Mark Murphy wrote:
> David Golombek wrote:
> > We could kill our app proactively, as soon
> > as we launch the ACTION_VIEW activity, but then we're left in a bad
> > state if the user aborts.
>
> Why? In other words, what difference is there between:
>
> -- the user exits your
I'm working on the upgrade process for an Android app and am having a
hard time ensuring a graceful shutdown of the old version. Can someone
explain how the old version of the app is killed and if there is
anything we can do to hook into the process?
In more detail:
We notice that a new version
8 matches
Mail list logo
