RFR: 8195744: Avoid calling ClassLoader.checkPackageAccess if security manager is not installed [v4]
Coleen Phillimore
coleenp at openjdk.java.net
Fri Feb 5 16:28:02 UTC 2021
On Fri, 5 Feb 2021 03:14:52 GMT, David Holmes <dholmes at openjdk.org> wrote:
>> Yes, why not? It's very useful in logging.
>
> There's no context for this final bit out of output. It could appear many lines after the earlier logging, and unless there is some locking around all this, could be interleaved with different PD checks. It seems to me that all the logging could occur after the Java call and include the success/failure. It could also include whether the check trivially succeeded due to there being no SM.
Ok, I see what you're suggesting. Yes, I'll move all the logging after the check. I don't see any reason to have it if we're not going to make the upcall. Also fixed the test.
-------------
PR: https://git.openjdk.java.net/jdk/pull/2410
More information about the hotspot-runtime-dev
mailing list