PING: Re: RFR: 8219904: ClassCastException when calling FlightRecorderMXBean#getRecordings()

Erik Gahlin erik.gahlin at oracle.com
Wed Dec 11 20:05:13 UTC 2019


Chihiro, you must finalize the CSR. It is still in the draft state.

I also think you could remove the changes to the copyright in the 
specification.

Thanks
Erik

On 2019-11-29 17:08, Erik Gahlin wrote:
> Hi Chihiro,
>
> I have marked the CSR as reviewed.
>
> Erik
>
> On 2019-11-29 13:34, Chihiro Ito wrote:
>> Hi Erik,
>>
>> Thank you for reviewing and advice.
>> I had referred to other CSR,  I appended the source code and the 
>> webrev link to the specification in the CSR.
>>
>> Could you review the CSR, please?
>>
>> CSR : https://bugs.openjdk.java.net/browse/JDK-8234305
>>
>> Regards,
>> Chihiro
>>
>> 2019年11月29日(金) 0:46 Erik Gahlin <erik.gahlin at oracle.com 
>> <mailto:erik.gahlin at oracle.com>>:
>>
>>     Hi Chihiro,
>>
>>     Looks good.
>>
>>     Please add spec. change to the CSR and I will review it. You can
>>     then finalize the CSR and then it can take up to a week for a reply.
>>
>>     If it is accepted, I can then sponsor your fix.
>>
>>     Thanks
>>     Erik
>>
>>     On 2019-11-28 16:36, Chihiro Ito wrote:
>>>     Hi Erik,
>>>
>>>     Thank you for your review.
>>>
>>>     I fixed the connection method to use attach and checked the
>>>     property "toDisk" is existing.
>>>     Could you please review this again?
>>>
>>>     Webrev : http://cr.openjdk.java.net/~cito/JDK-8219904/webrev.03/
>>>
>>>     Regards,
>>>     Chihiro
>>>
>>>
>>>     2019年11月27日(水) 0:07 Erik Gahlin <erik.gahlin at oracle.com
>>>     <mailto:erik.gahlin at oracle.com>>:
>>>
>>>         Hi,
>>>
>>>         If you can test this without connecting over the network it
>>>         would be good.
>>>
>>>         Not sure if it  helps, but it is possible to attach to the
>>>         same process if you set the system property at startup,
>>>         -Djdk.attach.allowAttachSelf=true
>>>
>>>         Could you also check if the property "toDisk" is missing, and
>>>         if so check if "disk" is set, and if so use it. This is in
>>>         the case somebody called the RecordingInfo:: from with "disk"
>>>         (as the javadoc stated previously). Not sure why anyone would
>>>         do so, but that way we will prevent code from breaking.
>>>
>>>         Thanks
>>>         Erik
>>>
>>>         On 2019-11-26 15:45, Chihiro Ito wrote:
>>>>         Hi,
>>>>
>>>>         Could you please review this?
>>>>
>>>>         Regards,
>>>>         Chihiro
>>>>
>>>>         2019年11月19日(火) 0:54 Chihiro Ito <chiroito107 at gmail.com
>>>>         <mailto:chiroito107 at gmail.com>>:
>>>>
>>>>             Hi Erik,
>>>>
>>>>             I fixed the implementation. Could you review this, please?
>>>>
>>>>             After your review, may I copy the diff of RecordingInfo
>>>>             class into the specification section?
>>>>             This is my first time to do CSR, so I'm sorry if I made
>>>>             a mistake in the procedure.
>>>>
>>>>             Webrev :
>>>> http://cr.openjdk.java.net/~cito/JDK-8219904/webrev.02/
>>>>             JBS : https://bugs.openjdk.java.net/browse/JDK-8219904
>>>>             CSR : https://bugs.openjdk.java.net/browse/JDK-8234305
>>>>
>>>>             Regards,
>>>>             Chihiro
>>>>
>>>>
>>>>             2019年11月18日(月) 9:24 Erik Gahlin <erik.gahlin at oracle.com 
>>>>
>>>>             <mailto:erik.gahlin at oracle.com>>:
>>>>
>>>>                 Hi Chihiro,
>>>>
>>>>                 I created a CSR draft:
>>>>
>>>> https://bugs.openjdk.java.net/browse/JDK-8234305
>>>>
>>>>                 If you fix the implementation, we can then copy the
>>>>                 changes to the RecordingInfo class (in a diff
>>>>                 format) in the specification section.
>>>>
>>>>                 Thanks
>>>>                 Erik
>>>>
>>>>
>>>>>                 On 17 Nov 2019, at 16:14, Chihiro Ito
>>>>>                 <chiroito107 at gmail.com
>>>>>                 <mailto:chiroito107 at gmail.com>> wrote:
>>>>>
>>>>>                 Hi Erik,
>>>>>
>>>>>                 I also think It is better.
>>>>>                 I would like you to help me submit CSR.
>>>>>
>>>>>                 Regards,
>>>>>                 Chihiro
>>>>>
>>>>>                 2019年11月14日(木) 2:46 Erik Gahlin
>>>>>                 <erik.gahlin at oracle.com
>>>>>                 <mailto:erik.gahlin at oracle.com>>:
>>>>>
>>>>>                     Hi Chihiro,
>>>>>
>>>>>                     I think it would be better to change the
>>>>>                     javadoc from "disk" to "toDisk" and not add a
>>>>>                     new method the RecordingInfo class as this is
>>>>>                     less of an intrusive change, to JMC and others.
>>>>>
>>>>>                     Still, a CSR needs to be filed, as this impacts
>>>>>                     compatibility. Do you want help with this?
>>>>>
>>>>>                     The constructor could support both disk and
>>>>>                     toDisk, similar to what you have today.
>>>>>
>>>>>                     Thanks
>>>>>                     Erik
>>>>>
>>>>>                     On 2019-11-13 14:12, Chihiro Ito wrote:
>>>>>>                     Hi Erik,
>>>>>>
>>>>>>                     Could you please review this?
>>>>>>
>>>>>>                     Regards,
>>>>>>                     Chihiro
>>>>>>
>>>>>>                     2019年10月25日(金) 0:43 Chihiro Ito
>>>>>>                     <chiroito107 at gmail.com
>>>>>> <mailto:chiroito107 at gmail.com>>:
>>>>>>
>>>>>>                         Hi Erik,
>>>>>>
>>>>>>                         I leave the isToDisk method to maintain
>>>>>>                         compatibility so that I can connect to
>>>>>>                         Java processes that this issue does not
>>>>>>                         resolve.
>>>>>>                         The isToDisk method was deprecated.
>>>>>>
>>>>>>                         Could you review this, please?
>>>>>>
>>>>>>                         JBS
>>>>>> https://bugs.openjdk.java.net/browse/JDK-8219904
>>>>>>                         webrev
>>>>>> http://cr.openjdk.java.net/~cito/JDK-8219904/webrev.01/
>>>>>>
>>>>>>                         Regards,
>>>>>>                         Chihiro
>>>>>>
>>>>>>                         2019年10月24日(木) 1:12 Chihiro Ito
>>>>>>                         <chiroito107 at gmail.com
>>>>>> <mailto:chiroito107 at gmail.com>>:
>>>>>>
>>>>>>                             Hi Erik,
>>>>>>
>>>>>>                             Thank you for prompt reply.
>>>>>>
>>>>>>                             I also checked the Javadoc. The key
>>>>>>                             should be disk.
>>>>>>
>>>>>>                             > Is "toDisk" synthesized from the
>>>>>>                             method name "isToDisk" in
>>>>>>                             > RecordingInfo, or is it set
>>>>>>                             somewhere else?
>>>>>>
>>>>>>                             Yes, it appears to be composited from
>>>>>>                             the isToDisk method.
>>>>>>                             For compatibility, I think you should
>>>>>>                             leave the isToDisk method and add
>>>>>>                             @Deprecate, what do you think?
>>>>>>
>>>>>>                             This problem occurs when attaching
>>>>>>                             over a network or to a process. The
>>>>>>                             process is connected over the network
>>>>>>                             because it cannot be attached to its
>>>>>>                             own process.
>>>>>>
>>>>>>                             Regards,
>>>>>>                             Chihiro
>>>>>>
>>>>>>                             2019年10月21日(月) 21:13 Erik Gahlin
>>>>>>                             <erik.gahlin at oracle.com
>>>>>> <mailto:erik.gahlin at oracle.com>>:
>>>>>>
>>>>>>                                 Hi,
>>>>>>
>>>>>>                                 The javadoc states that the key
>>>>>>                                 should be "disk" (which is the
>>>>>>                                 same used
>>>>>>                                 by jcmd and 
>>>>>> -XX:StartFlightRecording).
>>>>>>
>>>>>>                                 Is "toDisk" synthesized from the
>>>>>>                                 method name "isToDisk" in
>>>>>>                                 RecordingInfo, or is it set
>>>>>>                                 somewhere else?
>>>>>>
>>>>>>                                 Do the problem only occur if you
>>>>>>                                 access RecordingInfo over network?
>>>>>>                                 I am
>>>>>>                                 asking since you added a port to
>>>>>>                                 the test.
>>>>>>
>>>>>>                                 Erik
>>>>>>                                 > Hi
>>>>>>                                 >
>>>>>>                                 > I fixed a problem with
>>>>>>                                 ClassCastException. Also, I fixed
>>>>>>                                 the CompositeData
>>>>>>                                 > field name because it was 
>>>>>> incorrect.
>>>>>>                                 >
>>>>>>                                 > Could you please review it?
>>>>>>                                 >
>>>>>>                                 > JBS
>>>>>> https://bugs.openjdk.java.net/browse/JDK-8219904
>>>>>>                                 > Webrev
>>>>>> http://cr.openjdk.java.net/~cito/JDK-8219904/webrev.00/
>>>>>>                                 >
>>>>>>                                 > The comment on the JBS mentions
>>>>>> UneclaredThrowableException, which
>>>>>>                                 is a
>>>>>>                                 > different issue.
>>>>>>                                 >
>>>>>>                                 > Regards,
>>>>>>                                 > Chihiro
>>>>>>
>>>>


More information about the hotspot-jfr-dev mailing list