> On Mar 20, 2017, at 2:58 AM, Peter Levart <peter.lev...@gmail.com> wrote: > > > > On 03/19/2017 11:42 PM, Peter Levart wrote: >> No problem. Here it is (against tip of jake/jdk): >> >> http://cr.openjdk.java.net/~plevart/jdk9-jake/AccessibleObject.canAccess_caching/webrev.02/ >> >> >> This patch changes a little when the stacktrace requested by >> sun.reflect.debugModuleAccessChecks system property is printed. In original >> code it is printed when the access (invocation, get/set) itself fails as >> well as when AccessibleObject.canAccess() returns false. Is that what was >> intended? Patched code only prints when the actual access attempt fails and >> not when canAccess() returns false. If you want canAccess() to also print >> stacktrace, I can add it to canAccess() method. > > ...I was wrong about that claim. Original code does exactly what the patch > does - it only prints stack trace when checkAccess is called and never when > canAccess. As slowCheckMemberAccess (which printed the stacktrace) was > renamed to slowVerifyAccess (which is used from canAccess too), the > printing/throwing had to be moved to the checkAccess method itself to > preserve the behavior.
Looks like webrev.02 was updated to include what you described above. Anyway, the patch looks okay to me. Mandy