JDK-5077866 : [Cinnabar] REG: JTable crashes when focus moved out from a cell
  • Type: Bug
  • Component: client-libs
  • Sub-Component: java.awt:i18n
  • Affected Version: 5.0
  • Priority: P2
  • Status: Resolved
  • Resolution: Fixed
  • OS: linux
  • CPU: generic
  • Submitted: 2004-07-23
  • Updated: 2004-12-13
  • Resolved: 2004-12-13
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
Other JDK 6
5.0u4Fixed 6 betaFixed
Related Reports
Relates :  
JDK-6203815 - Checkbox in JTable makes application crash if launch from a XDMCP session
Description
This is a regression in Tiger. Worked fine with 1.4.2 
The bug is reproduceable only on Cinnbar B13. Works fine with windows, sparc and Redhat AS 3.0. Not specific to any L&F.  Haven't tried with any older builds on Cinnabar.

STEPS TO REPRODUCE:

-  Launch SwingSet2 demo
-  Click on JTable tab
-  Double click on any cell in JTable
-  Click on any other cell
-  If you find a crash, the bug is reproduced.

The log file is attached.

I am able to reproduce the bug from Tiger B18. Haven't tried the builds before b18 yet.

But there's an interesting thing noted here:

After I run SwingSet2 on 1.4.2 where it works fine, I'm later not able to reproduce the bug on any of the Tiger builds.
But when I reboot the machine again and run on any of the Tiger builds, the crash is reproduced. This happens consistently.

Also talked with Shannon about the bug. 

----------

Cinnabar B13 is being tested here.
Output of cat /etc/SuSE-release
VERSION = 2.0
RELEASE = 1.0

Hardware info:

Number of cpus: 1
CPU model name: 6  , vendor: AuthenticAMD
Operating at  : 598.853  Mhz
Swap info     : Total: 1052216 Used: 45348 Free: 1006868
Memory size   : 126204 Kb
Shared Memory : SHMSEG=4096 SHMMNI= SHMMAX=32768 SHMMAX=32768 SHMTOT=8388608
--------- Video ---------
BOARD 0 = *(1:5:0) ATI Technologies Inc 3D Rage Pro AGP 1X/2X rev 92
BOARD_NAME=(--) PCI:*(1:5:0) ATI Technologies Inc 3D Rage Pro AGP 1X/2X rev 92,
Mem @ 0xfd000000/24, 0xfeaff000/12, I/O @ 0xe800/8, BIOS @ 0xfeac0000/17


###@###.### 2004-07-29
Comments
EVALUATION Name: yn153226 Date: 07/26/2004 Cannot reproduce with SuSE 9.0, either. My guess is, there's an issue with java installations in Cinnabar. Perhaps java 1.5 cannot load a shared library while 1.4.2 can; once libabry is loaded, there's no crash anymore. Anyway, it seems to me rather Cinnabar issue. Did you try another _Cinnabar_ build:-)? ###@###.### ====================================================================== Name: yn153226 Date: 07/26/2004 an attached stack trace shows that the crash occurs on disposing of an input method context, so it is most probably IM issue. ###@###.### ====================================================================== No crash using Cinnabar B14 and Tiger B59. Tests with SwingSet2.jar and Font2DTest.jar do not reproduce bug report. ###@###.### 2004-07-27 ====================================================================== I have tested Tiger b58 on Cinnabar b14 (Linux, Solaris10/Sparc and Solaris10/Intel version) with C and de_DE locale. SwingSet2.jar and Font2DTest.jar worked fine everywhere, no crash. Which Cinnabar version do you use and which HW? ###@###.### 2004-07-29 ====================================================================== Dear the submitter, Could you please check if this bug is still reproducible on the latest Cinnabar environment and update this bug report with the status? Thanks in advance. ###@###.### 2004-09-08 Got information. See Comments. ###@###.### 2004-09-13 The root cause of this bug is that the function createStatusWindow calls XCreateFontSet (in j2se/src/solaris/native/sun/awt/awt_InputMethod.c), and that function call may return NULL due to no requetsed font found or some other reasons (no enough memory etc). As a result, when we try to call XFreeFontSet to free the fontset object returned by XCreateFontSet during resource release, on some linux platform, it cause crash. The crash is actually by random. The fix is to call XCreateFontSet with the right font name. Also calls XFreeStringList when XCreateFontSet fails to find the font name specified. And do a NULL check before calling XFreeFontSet just for safty purpose. ###@###.### 2004-11-17 02:21:41 GMT ###@###.### 2004-11-17 17:42:00 GMT ###@###.### 2004-11-20 02:48:21 GMT ###@###.### 2004-11-23 01:19:16 GMT
17-11-2004