Martin Buchholz martinrb at google.com
Tue Apr 2 22:29:55 UTC 2013

On Tue, Apr 2, 2013 at 3:20 PM, Mark Sheppard <mark.sheppard at oracle.com>wrote:

>  Hi Martin,
>   are there any concerns that the string literal "." might be on the stack
> and not in data segment or heap?


> Also, is it assured that the returned parentPathv will never be
> de-allocated? I don't see parentPathv exposed beyond **
> that of  UNIXProcess_md.c, at least not directly.

It is not visible anywhere else.  It's "permanently" allocated.  Those of
you thinking about shutting down and restarting a JVM within the same
process will cringe - this is one of many "little leaks" in such an
unsupported scenario.

More information about the core-libs-dev mailing list