RFR: 8247536: Support for pre-generated java.lang.invoke classes in CDS static archive

sundararajan.athijegannathan at oracle.com sundararajan.athijegannathan at oracle.com
Tue Aug 18 02:25:21 UTC 2020


Not a full review of fresh changes. But couple of comments:

* src/hotspot/share/memory/lambdaFormInvokers.cpp and 
src/hotspot/share/memory/lambdaFormInvokers.hpp miss "Classpath 
exception" clause in the copyright header

* I had suggested a change GenerateJLIClassesPlugin.java in last round 
of review. That's still applicable.

https://mail.openjdk.java.net/pipermail/core-libs-dev/2020-August/068160.html

-Sundar

On 18/08/20 1:07 am, Yumin Qi wrote:
> Hi, Ioi
>
>   Thanks for review/suggestion. I have updated the webrev at the 
> following link:
>
> http://cr.openjdk.java.net/~minqi/2020/8247536/webrev-02/
>
>   Following changes done in this patch:
>
>   1) move regenerating holder classes into new added file: 
> lambdaFormInvokers.[ch]pp
>
>   "@lambda-form-invoker" tag only observed at jvm side.
>
>   2) remove InvokerBytecodeGeneratorHelper.java, move related 
> functions to existing class, GenerateJLIClassesHelper.java
>
>   3) add tracing for SPECIES_RESOLVE, so archive more classes.
>
>   4) remove InvokerGenerateBytesException.java, using RuntimeException 
> instead.
>
>   5) Changed make file to exclude the new added lamdaFormInvokers.cpp 
> from the non-cds building list.
>
>   6) There is a bug in previous patch, jobject (typeArrayOop) should 
> not be used directly for loading class since GC will move the java 
> object. Fixed by copying the bytes from handle to C heap.
>
>   7) Tested with tier1,2
>
>
> Thanks
>
> Yumin
>
>
>
> On 8/15/20 6:27 PM, Ioi Lam wrote:
>> On 8/15/20 6:19 PM, Ioi Lam wrote:
>>> To better capture what we're trying to do in this RFE, I've changed 
>>> the RFE title (and the subject of this email thread) to
>>>
>>> https://bugs.openjdk.java.net/browse/JDK-8247536
>>> Support for pre-generated java.lang.invoke classes in CDS static 
>>> archive
>>>
>>> I also update the RFE Description to give an overview of the problem 
>>> and the solution.
>>>
>>> The design document is also updated to reflect Yumin's latest 
>>> implementation
>>>
>>
>> Oops, sent too quickly. The design doc's title is also updated:
>>
>> https://wiki.openjdk.java.net/display/HotSpot/Support+for+pre-generated+java.lang.invoke+classes+in+CDS+archive 
>>
>>
>>> Thanks
>>> - Ioi
>>>
>>> On 8/12/20 11:54 AM, calvin.cheung at oracle.com wrote:
>>>> Hi Yumin,
>>>>
>>>> I reviewed mostly the native code. Below are my comments:
>>>>
>>>> 1) classListParser.hpp
>>>>
>>>> 71   bool                _lambda_format;
>>>>
>>>> The above name is too generic. How about _lambda_form or 
>>>> _is_lambda_form?
>>>> If you rename the above, please also rename the function which 
>>>> returns the above bool.
>>>>
>>>> 2) jvm.cpp
>>>>
>>>> 3850 JVM_ENTRY(void, JVM_CDSTraceResolve(JNIEnv* env, jclass 
>>>> ignore, jstring line))
>>>>
>>>> ignore -> ignored
>>>>
>>>> 3) jvm.hpp
>>>>
>>>> 210 JVM_CDSTraceResolve(JNIEnv* env, jclass ignore, jstring line);
>>>>
>>>> Same comment as for jvm.cpp
>>>>
>>>> 4) metaspaceShared.cpp
>>>>
>>>> 2017   size_t i = 0;
>>>> 2018   while (i < size) {
>>>> 2019     full_name[i++] = *start++;
>>>> 2020   }
>>>>
>>>> Could the above be simplified to the following?
>>>>
>>>>     strncpy(full_name, start, size - 1);
>>>>
>>>> 2029   char* class_name = get_full_class_name(path_name);
>>>>
>>>> Should os::free(class_name) be called before the function returns?
>>>>
>>>> 1870 static GrowableArray<char *>* lambda_list = NULL;
>>>>
>>>> The name lambda_list is a bit generic, how about lambda_form_list?
>>>>
>>>> 2112       lambda_list->append(parser.current_line());
>>>>
>>>> Since parser.current_line() does a strdup, do those buffer need to 
>>>> be freed after its use? (i.e. after the call to 
>>>> regenerate_holder_classes()?)
>>>>
>>>> In MetaspaceShared::regenerate_holder_classes, before calling up to 
>>>> java, I think it's better to strip the prefix 
>>>> "@lambda-form-invoker" from the strings. So that the java 
>>>> InvokerBytecodeGeneratorHelper.readTraceConfig method doesn't need 
>>>> to handle it:
>>>>
>>>>  143                     case "[LF_RESOLVE]":
>>>>  144                     case InvokerBytecodeGenerator.CDS_LOG_PREFIX:
>>>>
>>>> 5) DumpSymbolAndStringTable.java
>>>>
>>>> 37     private static final String my_string = 
>>>> "DumpSymbolAndStringTable";
>>>>
>>>> Unused variable?
>>>>
>>>> thanks,
>>>>
>>>> Calvin
>>>>
>>>> On 8/11/20 10:36 AM, Yumin Qi wrote:
>>>>> Forget to send to @core-lib-dev, the patch changed jdk code.
>>>>>
>>>>>
>>>>> Thanks
>>>>>
>>>>> Yumin
>>>>>
>>>>>
>>>>>
>>>>> -------- Forwarded Message --------
>>>>> Subject:     RFR: 8247536: Support pre-generated MethodHandle 
>>>>> lambda forms in CDS
>>>>> Date:     Tue, 11 Aug 2020 07:44:34 -0700
>>>>> From:     Yumin Qi <yumin.qi at oracle.com>
>>>>> To:     hotspot-runtime-dev at openjdk.java.net
>>>>>
>>>>>
>>>>>
>>>>> Hi, Please reivew
>>>>>
>>>>>   bug: https://bugs.openjdk.java.net/browse/JDK-8247536
>>>>>   Webrev: http://cr.openjdk.java.net/~minqi/2020/8247536/webrev-01/
>>>>>
>>>>>   Summary: CDS does not archive pre-generated lambda form classes 
>>>>> for method handles:
>>>>>
>>>>> [0.142s][info][class,load] 
>>>>> java.lang.invoke.LambdaForm$MH/0x0000000800066c40 source: 
>>>>> __JVM_LookupDefineClass__
>>>>>
>>>>> The method handle resolution if not found in the holder class, a 
>>>>> class with the method will be generated and loaded as vm internal 
>>>>> created class and not archived like above. Those class names are 
>>>>> mangled as combination of the class name and the class instance 
>>>>> address.
>>>>>
>>>>> In this patch, collect those method holder class names and their 
>>>>> associated methods' signatures when user specify 
>>>>> DumpLoadedClassList, and record them in the log file in a format 
>>>>> similar to the output format from traced by 
>>>>> -Djava.lang.invoke.MethodHandle.TRACE_RESOLVE=true, but here use 
>>>>> another name for CDS: 
>>>>> -Djava.lang.invoke.MethodHandle.CDS_TRACE_RESOLVE=true instead. 
>>>>> The line prefix also changed from "[LF_RESOLVE]" to 
>>>>> "@lambda-invoke-handle".  At dump time, regenerate the holder 
>>>>> class with those methods and replace the existing holder class and 
>>>>> archived it. At runtime, the resolution of the holder class which 
>>>>> contains those methods are loaded from CDS archive so avoid 
>>>>> regenerating them again. The patch reorganized the code for Jlink 
>>>>> plugin, so the new added InvokerBytecodeGeneratorHelper can be 
>>>>> shared both by JLink plugin and CDS code. Also change made to the 
>>>>> mangled generated class name at static dump, the class name 
>>>>> combining the class name and a sequentially ordered number instead 
>>>>> of the class instance address, which looks like a random.
>>>>>
>>>>>
>>>>> To use this feature, one can do the dump/run just like done for a 
>>>>> static dump/run, so no extra steps required.
>>>>>
>>>>>   1) java -XX:DumpLoadedClassList=myclasslist JavaApp
>>>>>
>>>>>       DumpLoadedClassList will turn on 
>>>>> -Djava.lang.invoke.MethodHandle.CDS_TRACE_RESOLVE=true
>>>>>
>>>>>   2) java -Xshare:dump -XX:SharedClassListFile=myclasslist 
>>>>> -XX:SharedArchiveFile=my.jsa JavaApp
>>>>>
>>>>>   3) java -Xshare:on -XX:SharedArchiveFile=my.jsa JavaApp
>>>>>
>>>>>   The performance on javac HelloWorld.java (javac-benchmark), no 
>>>>> patch vs patch:
>>>>>
>>>>>    1:   2689285002  2641821474 (-47463528)      ---- 391.720 
>>>>> 382.990 ( -8.730)      ----
>>>>>    2:   2687495085  2632969688 (-54525397)      ---- 391.030 
>>>>> 381.480 ( -9.550)      -----
>>>>>    3:   2694664066  2636523114 (-58140952)      ----- 390.610 
>>>>> 382.550 ( -8.060)      ----
>>>>>    4:   2686554164  2639355233 (-47198931)      ---- 390.700 
>>>>> 383.390 ( -7.310)      ---
>>>>>    5:   2691072338  2633016687 (-58055651)      ----- 388.990 
>>>>> 382.360 ( -6.630)      ---
>>>>>    6:   2684448174  2644191854 (-40256320)      --- 389.450 
>>>>> 382.990 ( -6.460)      ---
>>>>>    7:   2694921227  2630505090 (-64416137)      ----- 389.300 
>>>>> 383.160 ( -6.140)      ---
>>>>>    8:   2685209712  2639334320 (-45875392)      ---- 388.370 
>>>>> 381.060 ( -7.310)      ---
>>>>>    9:   2695885942  2640618655 (-55267287)      ---- 389.560 
>>>>> 381.100 ( -8.460)      ----
>>>>>   10:   2689162942  2635658943 (-53503999)      ---- 389.690 
>>>>> 379.110 (-10.580)      -----
>>>>> ============================================================
>>>>>         2689866989  2637396210 (-52470778)      ---- 389.941 
>>>>> 382.017 ( -7.924)      ----
>>>>>
>>>>> instr delta =    -52470778    -1.9507%
>>>>>
>>>>> time  delta =       -7.924 ms -2.0321%
>>>>>
>>>>>   It will show much more improvement if compare with the default 
>>>>> archive:
>>>>>
>>>>>                  no patch                        patch
>>>>>
>>>>> instr         3,996,847,605        3,320,928,995
>>>>>
>>>>> time(s).     0.686835162          0.474933546
>>>>>
>>>>>
>>>>>   Tests:
>>>>>
>>>>> 1)jtreg on jdk/tools and hotspot/runtime local.
>>>>>
>>>>> 2) mach5 ter1,2
>>>>>
>>>>>
>>>>> Thanks
>>>>>
>>>>> Yumin
>>>>>
>>>>>   `
>>>
>>


More information about the core-libs-dev mailing list