RFR: 8259242: vmTestbase/vm/mlvm/mixed/stress/java/findDeadlock/TestDescription.java timed out
Coleen Phillimore
coleenp at openjdk.java.net
Wed Apr 7 12:30:52 UTC 2021
On Wed, 7 Apr 2021 11:55:56 GMT, Coleen Phillimore <coleenp at openjdk.org> wrote:
>> src/hotspot/share/classfile/protectionDomainCache.hpp line 115:
>>
>>> 113:
>>> 114: ProtectionDomainEntry* next() { return Atomic::load(&_next); }
>>> 115: void set_next(ProtectionDomainEntry* entry) { Atomic::store(&_next, entry); }
>>
>> Again these may need to be load_acquire/release_store for correct lock-free traversal. With suitable name changes for the methods.
>
> Thanks, I'll add acquire/release and change the names to match the implementation details.
Did we collectively agree on the naming convention that set_next has to be release_set_next because the implementation uses release_store, or is this just your preference? I find this noisy in the context where these functions are being called.
-------------
PR: https://git.openjdk.java.net/jdk/pull/3362
More information about the hotspot-dev
mailing list