<AWT Dev> <AWT dev>[9] Review request for JDK-8160453: [macosx] scary native stack trace in closed/java/awt/FileDialog/FileDialogBufferOverflowTest/FileNameTest.java

[Avik Niyogi]

Hi Manajit,

A few suggestions/queries: 
1) Please check the same situation where a FileDialog_SAVE is used instead. Does the fix impact this case? 
If there is no impact in that case (which is the most likely situation), can the else block be removed as no other case should be handled then because as far as I understand, the FileDialog_SAVE is already handled?
2) Where are fDirectory and fFile being set? Can a change be done to this instead within safeSaveOrLoad method?
3) What I gather from the actual situation is that safeSaveOrLoad is called multiple times even before the defaults are retrieved and this change will run Modally only when the details are available for display.
Is it possible to trace what is calling it before the object is created? Maybe there is a bigger problem altogether and even though this fixes it neatly, a clean solution might possibly exist. This is just a recommendation though.

With Regards,
Avik Niyogi

> On 10-Nov-2016, at 11:00 pm, Manajit Halder <manajit.halder at oracle.com> wrote:
> 
> Hi All,
> 
> Kindly review the fix for JDK9.
> 
> Bug: 
> https://bugs.openjdk.java.net/browse/JDK-8160453 <https://bugs.openjdk.java.net/browse/JDK-8160453>
> 
> Webrev: 
> http://cr.openjdk.java.net/~mhalder/8160453/webrev.00/ <http://cr.openjdk.java.net/~mhalder/8160453/webrev.00/>
> 
> Issue: 
> Test passed but stack trace was generated because of trying to pass nil directory as parameter in method runModalForDirectory.
> 
> Fix: 
> Added nil check for directory and file before calling the method runModalForDirectory.
> 
> Regards,
> Manajit

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/awt-dev/attachments/20161111/800a2d86/attachment.html>