Bug ID: JDK-6558151 10 tests pass jdk6.0u01b06 but fail on jdk6.0u2b01 and jdk6.0u2b02

Type: Bug
Component: client-libs
Sub-Component: 2d
Affected Version: 6u2

Priority: P2
Status: Closed
Resolution: Duplicate
OS: windows_2003
CPU: x86

Submitted: 2007-05-16
Updated: 2011-01-19
Resolved: 2007-05-17

JDK              : jvm64 - JDK6.0_u2b01, JDK6.0_u2b02
Platforms[s]     : Windows 2003 x64
testbase         : /net/vice.russia/export/home0/dtf/dtf_ws/suites/6.0
Falling tests    : 10 testcases:  
Tested on        : preved.russia.sun.com

Log for JDK6.0_u2b01,JDK6.0_u2b02:

SUITE RUN PREPARATION 2D_PrintingMerlin: READY
TEST Testcase CancelablePrintJobTest 1/4 1/10 FAIL
TEST Testcase CollatePageRangesPrinterNameTest 2/4 2/10 FAIL
TEST Testcase Type1F3FontTest 3/4 3/10 FAIL
TEST Testcase PrinterNameTest 4/4 4/10 FAIL
SUITE RUN PREPARATION 2D_Font: READY
TEST Testcase AntialiasEnabledTest 1/6 5/10 FAIL
TEST Testcase AntialiasGASPTest 2/6 6/10 FAIL
TEST Testcase AntialiasLCDTextTest 3/6 7/10 FAIL
TEST Testcase AntialiasLCDContrastTest 4/6 8/10 FAIL
TEST Testcase SwingAntialiasStartupTest 5/6 9/10 FAIL
TEST Testcase SwingAntialiasLCDTextTest 6/6 10/10 FAIL

Log for jdk6.0u01b06:

SUITE RUN PREPARATION 2D_PrintingMerlin: READY
TEST Testcase CancelablePrintJobTest 1/4 1/10 PASS
TEST Testcase CollatePageRangesPrinterNameTest 2/4 2/10 PASS
TEST Testcase Type1F3FontTest 3/4 3/10 PASS
TEST Testcase PrinterNameTest 4/4 4/10 PASS
SUITE RUN PREPARATION 2D_Font: READY
TEST Testcase AntialiasEnabledTest 1/6 5/10 PASS
TEST Testcase AntialiasGASPTest 2/6 6/10 PASS
TEST Testcase AntialiasLCDTextTest 3/6 7/10 PASS
TEST Testcase AntialiasLCDContrastTest 4/6 8/10 PASS
TEST Testcase SwingAntialiasStartupTest 5/6 9/10 PASS
TEST Testcase SwingAntialiasLCDTextTest 6/6 10/10 PASS


Result:
===============
JDK6.0_u2b01,
/net/vice.russia/export/home0/dtf/dtf_ws/dtf3-exec/6.0-reurn_2d-win2003x64-jvm64-2007-05-16-16-41-50-0790

JDK6.0_u2b02:
/net/vice.russia/export/home0/dtf/dtf_ws/dtf3-exec/6.0-reurn_2d-win2003x64-jvm64-2007-05-16-16-00-46-0937

jdk6.0u01b06:
/net/vice.russia/export/home0/dtf/dtf_ws/dtf3-exec/6.0-reurn_2d-win2003x64-jvm64-2007-05-16-15-57-11-0692

Steps to reproduce:
===============
1 Login to preved.russia.sun.com: http://preved.russia.sun.com:5800/
2 cd k:/dtf/dtf_ws/dtf3-exec/6.0-reurn_2d-win2003x64-jvm64-2007-05-16-15-57-11-0692/results/6.0-reurn_2d-win2003x64-jvm64-2007-05-16-15-57-11-0692
(/net/vice.russia/export/home0 = k:)
3 run tonga.sh

EVALUATION The submitter confirms that the fix for 6548757 resolves all of these. Closing as a duplicate.

17-05-2007

EVALUATION Its very probable these are all related to 6548757: Crash in fontmanager.dll on windows-amd64 All the printing failures look as if they would execute the code path that has the bug and the rest of the tests are font related so could also trigger it. However a couple of other tests that should trigger the bug weren't reported to fail. Its possible these other tests passed because no important memory location was corrupted in those runs. But another condition to trigger this bug is that a font name must be >= 32 chars Looking in the registry I don't see that on that system. But the system appears to have its language for non-unicode programs set to Russian, and sure enough the names that enumerate seem to be longer localised names. Also I was able to prove that this system is vulnerable to 6548757 with a small test program, and that a build containing the fix for that bug helps that test program complete properly. The only other thing I see is that all the JDK's fonts that are part of a JDK1.5.0_11 install are in the windows registry. I can't think why that would cause a crash, but the reason they are there is due to an amd64 installer bug : 6266292 java/awt/FontClass/CreateFont/NoRegister.java fails which is fixed in JDK 6 but not 1.5.X I need the submitter to re-run these tests on that system using the fixed build. Marking incomplete pending this additional info.

16-05-2007