synchronization in gcj

Brian Miller bnmille@uswest.com
Fri Jan 15 08:48:00 GMT 1999


Jerry Kramskoy wrote:
>> ...freeing up all monitors acquired during method execution when an 
> exception goes off.

Perhaps the bytecode verifier should insist that all monitor releases
be done at the beginning of a finally block. Wouldn't enforcing that
rule be enough to guarantee robustness?


More information about the Java mailing list

AltStyle によって変換されたページ (->オリジナル) /