On Tue, 24 Jan 2023 00:16:10 GMT, Alex Menkov <amen...@openjdk.org> wrote:
> classFileParser drops stack map frames for JDK classes (when verification is > not required). > As a result JvmtiClassFileReconstituter cannot restore the attribute for > class redefinition. > Note that if the class is in CDS archive, the frames are restored from CDS, > so this issue affects only JDK classes which are not in CDS. > This code is old (from "initial load") and I don't understand the reason it > was implemented this way. > > Testing: tier1-tier6 Now it becomes clearer why it was implemented this way. I suppose there is a reason why CDS does not elide the attributes when verification is not required. Looking at the code history I see that in former times the method allocated memory for the returned attribute (now it returns pointer in the source stream), so this code reduced the footprint. I agree that this inconsistency is not a bug itself - RetransformClasses spec says "Some attributes may not be present" (in the class bytes passed to CFLH), and JDK-8228604 may be closed as "not an issue". But I still think it make sense to simplify the logic and just keep stackmap table for all classes is class bytes contain them. @dholmes-ora what do you think? ------------- PR: https://git.openjdk.org/jdk/pull/12155
