RFR(M): 8199852: Print more information about class loaders in LinkageErrors.

[David Holmes]

On 21/03/2018 11:34 PM, Lindenmaier, Goetz wrote:
> Hi David,
> 
> I put that description into the bug:
> https://bugs.openjdk.java.net/browse/JDK-8199852
> I thought it should be better archived than in the mail thread.
> 
> The change you mention below is intentional. There are two
> different loaders involved, and so far only one of them was printed.
> I tried to put "different type" info more detailed wording, too (different
> class with the same name).

I need to look into this in more detail tomorrow (sorry nearly midnight 
here). I thought that message was for a case involving only one loader. 
(I added these detailed messages way back when.)

David

> Unfortunately I could not come up with a test for the messages
> in klassVtable.cpp, but there I just removed the "instance of",
> that appears in the printed message if the classloader has a name.
> 
> Any advice when there should be "" in the message? I normalized
> it so that only the user-definable names are quoted. I removed
> the "" from around classnames, methods etc.
> 
> Best regards,
>    Goetz.
> 
> 
>> -----Original Message-----
>> From: David Holmes [mailto:david.holmes at oracle.com]
>> Sent: Mittwoch, 21. März 2018 14:12
>> To: Lindenmaier, Goetz <goetz.lindenmaier at sap.com>; hotspot-runtime-
>> dev at openjdk.java.net
>> Subject: Re: RFR(M): 8199852: Print more information about class loaders in
>> LinkageErrors.
>>
>> Hi Goetz,
>>
>> Can you provide a detailed comparison of all the before and after
>> messages please. It seems to me you are rewriting some of the messages
>> in a way that changes the meaning.
>>
>> src/hotspot/share/classfile/systemDictionary.cpp
>>
>> Before:
>>
>> !         linkage_error1 = "loader constraint violation: loader
>> (instance of ";
>> !         linkage_error2 = ") previously initiated loading for a
>> different type with name \"";
>>
>> after:
>>
>> !         linkage_error1 = "loader constraint violation: loader ";
>> !         linkage_error2 = " wants to load class ";
>> !         linkage_error3 = ". A different class with the same name was
>> previously loaded by ";
>>
>> this does not describe the same situation!
>>
>> Thanks,
>> David
>>
>> On 21/03/2018 9:55 PM, Lindenmaier, Goetz wrote:
>>> Hi,
>>>
>>> ClassLoaders have a field containing a name since Java 9.
>>>
>>> This change improves some of the messages of LinkageErrors to
>>> print the name of the loader involved, as well as the parent of the loader.
>>> This simplifies tracking down the component causing an error in large
>>> Java systems.
>>>
>>> Please review:
>>> http://cr.openjdk.java.net/~goetz/wr18/8199852-
>> exMsg_Linkage/01/index.html
>>>
>>> I ran this through all our tests (jck, jtreg, spec etc).  I will run it through
>> submit-hs
>>> (or jdk/submit) once it's reviewed.
>>>
>>> Best regards,
>>>     Goetz.
>>>
>>>
>>>
>>>