RRF: JEP-271: Unified GC Logging

[Bengt Rutisson]

Hi Kirk,

On 2015-11-23 14:26, kirk.pepperdine at gmail.com wrote:
> Hi Bengt,
>
> I’ve not seen anything egregious as of yet though I’m no where near done.

Thanks for looking at this change!

> One comment for ostream.cpp.
>
> -// log_name comes from -XX:LogFile=log_name, -Xloggc:log_name or
> +// log_name comes from -XX:LogFile=log_name or
>   // -XX:DumpLoadedClassList=<file_name>
>
>
> IME, -Xloggc:log_name is the most common form used. Not stuck to it, 
> just saying…

Are you saying that it is scary to remove -Xloggc? Yes, I would agree. 
We've discussed it a bit here. It would be nice to get rid of it since 
it comes with a bunch of other issues such as log rotation etc. One way 
to handle the transition would be to map -Xloggc=filename to 
-Xlog:gc*:filename. It wouldn't be 100% correct and it wouldn't handle 
the log rotations flags. But maybe it would be a smoother transition for 
some users.

Regards,
Bengt

>
> Regards,
> Kirk
>
>> On Nov 23, 2015, at 9:43 AM, Bengt Rutisson 
>> <bengt.rutisson at oracle.com <mailto:bengt.rutisson at oracle.com>> wrote:
>>
>>
>> Hi Per,
>>
>> Thanks for looking at this!
>>
>> On 2015-11-23 09:51, Per Liden wrote:
>>> Hi Bengt,
>>>
>>> On 2015-11-19 16:29, Bengt Rutisson wrote:
>>>>
>>>> Hi everyone,
>>>>
>>>> After three pre-reviews it is time for the fist official review request
>>>> for JEP-271 Unified GC Logging.
>>>>
>>>> http://openjdk.java.net/jeps/271
>>>>
>>>> Most code changes are in the hotspot code:
>>>> http://cr.openjdk.java.net/~brutisso/JEP-271/review.00/webrev.00/
>>>
>>> This is not a complete review, just a few things I noticed while 
>>> browsing the webrev.
>>>
>>> ----------------------------------------------
>>> /src/share/vm/gc/g1/g1StringDedupQueue.cpp
>>>
>>> 157   log_trace(gc, stringdedup)(
>>> 158     "   [Queue]\n"
>>> 159     "      [Dropped: " UINTX_FORMAT "]", _queue->_dropped);
>>>
>>> Multi-line loging.
>>>
>>> ----------------------------------------------
>>> src/share/vm/gc/g1/g1StringDedupTable.cpp
>>>
>>> 573   log_trace(gc, stringdedup)(
>>> 574     "   [Table]\n"
>>> 575     "      [Memory Usage: " G1_STRDEDUP_BYTES_FORMAT_NS "]\n"
>>> 576     "      [Size: " SIZE_FORMAT ", Min: " SIZE_FORMAT ", Max: " 
>>> SIZE_FORMAT "]\n"
>>> 577     "      [Entries: " UINTX_FORMAT ", Load: " 
>>> G1_STRDEDUP_PERCENT_FORMAT_NS ", Cached: " UINTX_FORMAT ", Added: " 
>>> UINTX_FORMAT ", Removed: " UINTX_FORMAT "]\n"
>>> 578     "      [Resize Count: " UINTX_FORMAT ", Shrink Threshold: " 
>>> UINTX_FORMAT "(" G1_STRDEDUP_PERCENT_FORMAT_NS "), Grow Threshold: " 
>>> UINTX_FORMAT "(" G1_STRDEDUP_PERCENT_FORMAT_NS ")]\n"
>>> 579     "      [Rehash Count: " UINTX_FORMAT ", Rehash Threshold: " 
>>> UINTX_FORMAT ", Hash Seed: 0x%x]\n"
>>> 580     "      [Age Threshold: " UINTX_FORMAT "]",
>>>
>>> Multi-line logging.
>>>
>>> ----------------------------------------------
>>
>> Good catch. Will fix.
>>
>>> src/share/vm/gc/g1/g1StringDedup*.cpp
>>>
>>> In general, I think the string dedup logging currently done with 
>>> log_trace() should be log_debug() instead.
>>
>> Yes, that makes sense. Will change that.
>>
>> Bengt
>>
>>
>>>
>>> cheers,
>>> /Per
>>>
>>>>
>>>> Some tests in the JDK repo have been updated:
>>>> http://cr.openjdk.java.net/~brutisso/JEP-271/review.00/jdk-webrev.00/ 
>>>> <http://cr.openjdk.java.net/%7Ebrutisso/JEP-271/review.00/jdk-webrev.00/>
>>>>
>>>> As with the pre-reviews I have put togther some examples of what 
>>>> the new
>>>> logging looks like:
>>>> http://cr.openjdk.java.net/~brutisso/JEP-271/review.00/compare.html
>>>>
>>>> The intent is that this should cover the bulk of the logging changes.
>>>> There will most definitely be some follow up changes where we fix
>>>> details in the log messages etc.
>>>>
>>>> Among many other old logging flags this changeset removes the two flags
>>>> PringGC and PrintGCDetails. These two will be added back with a follow
>>>> up changeset, but when they are added back they will be marked as
>>>> deprecated.
>>>>
>>>> The reason for first removing them and then adding them back is to get
>>>> testing without these flags. That way we will know that we clean 
>>>> out all
>>>> usages of these flags from our testing.
>>>>
>>>> Thanks,
>>>> Bengt
>>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/hotspot-gc-dev/attachments/20151124/effc7122/attachment.htm>