Still looking for an official Review of this change. Thanks, /Staffan
On 11 Nov 2013, at 21:25, Staffan Larsen <staffan.lar...@oracle.com> wrote: > Ah, good catch. > > Here is an updated review: http://cr.openjdk.java.net/~sla/6606002/webrev.01/ > > Thanks, > /Staffan > > On 11 Nov 2013, at 21:06, serguei.spit...@oracle.com wrote: > >> It looks good. >> >> The only one nit is about the line: >> 160 System.out.println("Command line: "+Arguments.getJVMArgs() + >> Arguments.getJVMFlags()); >> >> If one or both of Arguments.getJVMArgs() and Arguments.getJVMFlags() return >> null >> (not sure if it is the case) then the output will have "null" like this: >> "Command line: nullnull" >> >> Would it make sense to keep the original checks for non-null strings? >> >> >> Thanks, >> Serguei >> >> >> On 11/11/13 11:28 AM, Staffan Larsen wrote: >>> The problem here is that ‘jinfo -flags’ only looks at the command line, but >>> if a user has changed a flag after the VM started (for example by using >>> ‘jinfo -flag’) that new value does not show up. >>> >>> I am changing the output so that ‘jinfo -flags’ now prints something like >>> this: >>> >>> Non-default VM flags: -XX:InitialHeapSize=268435456 >>> -XX:MaxHeapSize=4294967296 -XX:+PrintGCDetails >>> -XX:+UseCompressedClassPointers -XX:+UseCompressedOops -XX:+UseParallelGC >>> Command line: -XX:+PrintGCDetails >>> >>> >>> webrev: http://cr.openjdk.java.net/~sla/6606002/webrev.00/ >>> bug: https://bugs.openjdk.java.net/browse/JDK-6606002 >>> >>> Thanks, >>> /Staffan >> >
