JBoss has no involvement with these settings. They are directives to the JVM on the setting of the JVM heap size limits. Most, if not all JVMs, do not automatically block-off this memory from the OS as this selfish strategy may be detrimental to operation, particularly in a shared environment. However, it does signal to the OS that it may need that much contiguous memory as a minimum. This reduces the churn, the need to incrementally add blocks as memory consumption grows and so on.
However, all this is for use by the OS and the JVM. I'm sure that someone more qualified can give a better explanation of OS memory management.
I understand, basically my question raised up, because I have the webMethods integration server running on another server and when I add those parameters for the jvm, it does allocate the memory as soon as the process starts.
But based on those settings for the JVM, if my JBoss app server needs to use more memory, it will be allocated until I have defined the maximum?
If JBoss is in need of memory, like any Java app, it will allocated memory up to the defined maximum, yes. This is completely a JVM and not a JBoss issue, check your favorite JVM's documentation for more information