From: Steve B. <Ste...@an...> - 2004-12-28 13:46:24
|
Steven Augart wrote: > Oh, I had only meant for the -X:gc:initialHeap and -X:gc:boundedHeap > arguments to be synonyms for -Xms and -Xmx. There's no question in my > mind that it would be a Bad Thing to remove support for -Xms and -Xmx. I don't think this is a good idea. As Dave said, -Xms & -Xmx are something we just have to live with because they are standard JVM options. Having other options that alias them is confusing. Furthermore there are currently technical reasons why we can't really do that. The -Xms & -Xmx options are parsed very early in the bootstrap (in C code), before the allocators are up and running (there is a chicken and egg reason for this). On the other hand, the -X:gc: options (and most other Jikes RVM options) are parsed using heavyweight parsing mechanisms that depend on the allocators having been booted. >> I agree that it makes sense for -X:gc:printOptions to display the >> initial and max heap sizes, but it might be that the cleanest way to >> do this is to modify it's printing code to print those values (which >> it must know anyways) instead of adding duplicate options. > We I think the more appropriate thing to do would be to get one of the gc verbosity levels to print it out. --Steve |