The fix looks good to me.
In a case the bug is reproduced again we will have some details (module
name and ID) to understand what is wrong.
On 10/17/16 06:08, Alexander Kulyakhtin wrote:
Could you, please, review these test-only changes
CR: https://bugs.openjdk.java.net/browse/JDK-8166289 "RuntimeException: canRead()
reports false for reading from the same module: expected true, was false"
The fix restores the changes for
https://bugs.openjdk.java.net/browse/JDK-8165017, which were backed out with
https://bugs.openjdk.java.net/browse/JDK-8166312, due to the test failure.
The test failure have never been reproduced after extensive testing. This fix
is trying to restore the changes adding a more correct verification of the
CANREAD command, as well as some additional logging in case the test fails