Hi Volker,

Yes, it looks fine. I did'nt know if it was a reasonable request across os versions.

Thanks, Roger


On 7/27/2015 1:54 PM, Volker Simonis wrote:
Hi Roger,

I've added the requested test. See:

http://cr.openjdk.java.net/~simonis/webrevs/2015/8132374.v1/

I've tested it on Solaris 10/11, Linux SLES 12/Ubuntu 12.04, MaxOS X
10.9 and 10.10, Windows 7 and AIX 7.1
I think it should work on all supported platforms.

OK now?

Regards,
Volker

On Mon, Jul 27, 2015 at 3:31 PM, Roger Riggs <roger.ri...@oracle.com> wrote:
Hi Volker,

I doubt there is a test for any of the OS's but is it possible to write a
test that would
compare that string to the version string produced by one of the os provided
tools.
For example uname?

Thanks, Roger


On 7/27/2015 9:29 AM, Volker Simonis wrote:
Thanks, Alan.

We've done this correctly in our SAP JVM as well as IBM in their J9
but for some reason we've missed the fixed in the initial OpenJDK
port.

Regards,
Volker


On Mon, Jul 27, 2015 at 3:02 PM, Alan Bateman <alan.bate...@oracle.com>
wrote:

On 27/07/2015 13:54, Volker Simonis wrote:
Hi,

can somebody please review this tiny fix to get the value of the
'os.version' property correct on AIX:

http://cr.openjdk.java.net/~simonis/webrevs/2015/8132374/
https://bugs.openjdk.java.net/browse/JDK-8132374


This looks okay to me, assuming applications on AIX looking at os.version
won't be surprised.

-Alan


Reply via email to