On 27/10/2012 01:03, David Holmes wrote:
On 26/10/2012 10:23 PM, Staffan Larsen wrote:
Some awk scripts in the tests check for
-- process information unavailable
as valid output. However, there are other outputs that are also 'ok'.
For example:
-- main class information unavailable
webrev: http://cr.openjdk.java.net/~sla/8001621/webrev.01/
The wildcards seems a little too open to me - is there a risk of
matching things they should not?
I'm a little puzzled by the problem - is this something new, something
related to a specific platform ?
David
I don't think this is a new problem, the jps tests have been known to
fail when run after other tests that have left VMs running. I think
Staffan's changes are okay, an alternative would have been to able
handle strings such as " -- main class information unavailable" but I
think that just complicates things, and there may be others too.
-Alan
