We've ran into an issue with the EditJTI class in JCK when running
this with our VM as the harness VM. The behaviour of it differs
between 32-bit Windows as well as on 32/64-bit Linux from 64-bit
Windows. The reason for this seems to be that (quoting from our
internal bug report):
"The difference between HotSpot and our VM on Win64 is that
for System.getProperty("os.name") we return the correct
'Windows Server 2003' and HotSpot returns 'Windows XP'. The
EditJTI program does not support this and gets confused, it
thinks that the OS is an unrecognized one."
This can cause problems to the programs determining the version of
the OS from Java. Question is: will there be something done about
this?
this with our VM as the harness VM. The behaviour of it differs
between 32-bit Windows as well as on 32/64-bit Linux from 64-bit
Windows. The reason for this seems to be that (quoting from our
internal bug report):
"The difference between HotSpot and our VM on Win64 is that
for System.getProperty("os.name") we return the correct
'Windows Server 2003' and HotSpot returns 'Windows XP'. The
EditJTI program does not support this and gets confused, it
thinks that the OS is an unrecognized one."
This can cause problems to the programs determining the version of
the OS from Java. Question is: will there be something done about
this?
- duplicates
-
JDK-4850771 os.name is wrong in win2003 64bits in mantis and tiger
-
- Resolved
-