Request for review (S): 7022037: Pause when exiting if a debugger is attached on windows
coleen.phillimore at oracle.com
Mon Feb 28 10:28:24 PST 2011
Never mind. I'm not quite caught up with reading email.
On 2/28/2011 1:19 PM, Coleen Phillimore wrote:
> In os/posix/vm/os_posix.cpp the function is named
> is_debugger_present() rather than is_debugger_attached(). You've
> probably already found this out.
> Why can't you just enable this by default if someone has
> ShowMessageBoxOnError rather than adding a new option? The two seem
> compatible to me.
> On 2/24/2011 9:27 AM, Staffan Larsen wrote:
>>> -----Original Message-----
>>> From: Dmitry Samersoff
>>> Sent: den 24 februari 2011 3:16
>>> To: Staffan Larsen
>>> Cc: hotspot-runtime-dev at openjdk.java.net
>>> Subject: Re: Request for review (S): 7022037: Pause when exiting if a
>>> debugger is attached on windows
>>> 1. Is it possible to put this functionality under command line flag?
>>> We have PauseAtStartup and I think it's better to have similar
>>> PauseAtShutdown than autodetect debugger.
>> I would prefer not to since the beauty is to have the auto-detection
>> do "the right thing". This is functionality you want to use every
>> time you run from Visual Studio. Requiring a separate command line
>> flag kind of defeats the purpose. We could add a flag to turn it off,
>>> 2. It's better to create separate function for it.
>> Good point. Should I have a separate function in
>> src/share/vm/runtime/java.cpp, or should I have separate functions in
>> the os classes (to remove the current #ifdef).
More information about the hotspot-runtime-dev