RFR(M) JDK-8031819: Remove legacy jdk checks and code
David Holmes
david.holmes at oracle.com
Fri Jun 6 05:29:59 UTC 2014
Typo:
> You must a chunk of code regarding uncaught exception handling:
You _missed_ a chunk of code regarding uncaught exception handling:
David
-----
On 6/06/2014 3:26 PM, David Holmes wrote:
> Hi Harold,
>
> Love to see this cleanup! :) But everyone needs to be very aware that
> once this goes in you can forget about placing a JDK 9 VM anywhere but a
> JDK 9 JDK. (Or did we already hit that?)
>
> A few comments and further cleanups:
>
> In classLoader.* is it worth renaming/absorbing the is_rt_jar13 and
> related "13" items?
>
> ---
>
> javaClasses.cpp:
>
> This comment:
>
> 1732 // For Java 7+ we support the Throwable immutability protocol
> defined for Java 7. This support
> 1733 // was missing in 7u0 so in 7u0 there is a workaround in the
> Throwable class. That workaround
> 1734 // can be removed in a JDK using this JVM version
>
> can be reduced to simply:
>
> 1732 // We support the Throwable immutability protocol defined since
> Java 7.
>
> Or even deleted completely.
>
> ---
>
> systemDictionary.hpp:
>
> The check_klass_Opt_Only_JDK* functions seem to be unused and only
> generated assertion failures when they were used.
>
> ---
>
> linkResolver.cpp
>
> The comment above the change seems irrelevant now the change is made.
>
> ---
>
> reflection.cpp
>
> Comment:
>
> 423 // New (1.4) reflection implementation. Allow all accesses from
> 424 // sun/reflect/MagicAccessorImpl subclasses to succeed trivially.
>
> can drop the first sentence. Ditto #526
>
> ---
>
> reflectionUtils.cpp
>
>
> This comment makes no sense without a specific version reference and can
> be deleted:
>
> 79 // The following class fields do not exist in previous version of
> jdk
>
> ---
>
> thread.cpp:
>
> You must a chunk of code regarding uncaught exception handling:
>
> 1744 // JSR-166: change call from from ThreadGroup.uncaughtException to
> 1745 // java.lang.Thread.dispatchUncaughtException
> 1746 if (uncaught_exception.not_null()) {
> 1747 Handle group(this, java_lang_Thread::threadGroup(threadObj()));
> 1748 {
> 1749 EXCEPTION_MARK;
> 1750 // Check if the method Thread.dispatchUncaughtException()
> exists. If so
> 1751 // call it. Otherwise we have an older library without the
> JSR-166 changes,
> 1752 // so call ThreadGroup.uncaughtException()
>
> We only need the code that calls dispatchUncaughtException.
>
> ---
>
> threadService.cpp
>
> You don't need the block that used to delineate the "if" statement. ie
> lines:
>
> 668 {
> 678 }
>
> can be deleted and the indentation fixed up.
>
> ---
>
> src/share/vm/prims/unsafe.cpp:
>
> There is unused code in here too: Unsafe_SetObject140,
> Unsafe_GetObject140
>
> Cheers,
> David
>
> On 5/06/2014 1:34 AM, harold seigel wrote:
>> Hi,
>>
>> Please review this JDK 9 fix for bug JDK-8031819. The fix removes
>> legacy code for old JDK versions. Although the webrev contains lots of
>> files, the changes are not complicated.
>>
>> Open webrev: http://cr.openjdk.java.net/~hseigel/bug_8031819/
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8031819
>>
>> The fix was tested with the Hotspot JTREG tests, JCK Lang, VM, and
>> API/java_lang tests, nsk Quick tests, and JPRT.
>>
>> Thanks! Harold
More information about the hotspot-runtime-dev
mailing list