[8u-communication] JDK 8u20 released today!

[David Holmes]

On 25/08/2014 7:45 PM, Seán Coffey wrote:
> Please ensure the necessary fixes are now in the JDK 8u codeline. If
> not, please ensure that bug records exist to fix such issues in JDK 8u.

I have marked 8048169 as affecting 8u40 and added a comment that it 
needs backporting.

I have created 8056011 for the second zero-only issue that Andrew 
provides a patch for below, and linked it to 8048169.

David

> regards,
> Sean.
>
> On 20/08/14 16:27, Andrew Hughes wrote:
>> ----- Original Message -----
>>> On 20/08/2014 6:25 AM, Andrew Hughes wrote:
>>>> ----- Original Message -----
>>>>> I'd like to announce that JDK 8u20 has become available for download
>>>>> today [0]. Thanks to all those who have contributed towards it.
>>>>>
>>>>> OpenJDK 8u20 source code is available via the 8u20 stabilization
>>>>> forest
>>>>> [1].  I plan to update the OpenJDK 8u project page with latest
>>>>> status and
>>>>> to
>>>>> also ask the OpenJDK ops team to mark the 8u20 forests as read-only.
>>>>>
>>>>> If you're packaging this release, it could be useful to let subscribed
>>>>> members know about it via communication on this mailing list. Please
>>>>> continue to contribute fixes back to the jdk8u-dev forest [2] which is
>>>>> already gathering changes for the next non-CPU JDK8u release.
>>>>>
>>>>> Regards,
>>>>> Sean.
>>>>>
>>>>> [0] http://www.oracle.com/technetwork/java/javase/downloads/index.html
>>>>> [1] http://hg.openjdk.java.net/jdk8u/jdk8u20/
>>>>> [2] http://hg.openjdk.java.net/jdk8u/jdk8u-dev
>>>>>
>>>> Zero appears to be broken:
>>>>
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2852:114:
>>>> error:
>>>> form\
>>>> at '%lx' expects argument of type 'long unsigned int', but argument
>>>> 4 has
>>>> type 'void*' [-Werror=format=]
>>>>            tty->print_cr("Exception <%s> (" INTPTR_FORMAT ")",
>>>>            except_oop->print_value_string(), (void*)except_oop());
>>>>
>>>> ^
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2856:47:
>>>> error:
>>>> format '%d' expects argument of type 'int', but argument 3 has type
>>>> 'address {aka unsigned char*}' [-Werror=format=]
>>>>                          continuation_bci, THREAD);
>>>>                                                  ^
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2856:47:
>>>> error:
>>>> format '%d' expects argument of type 'int', but argument 4 has type
>>>> 'intptr_t {aka long int}' [-Werror=format=]
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2856:47:
>>>> error:
>>>> format '%lx' expects argument of type 'long unsigned int', but
>>>> argument 5
>>>> has type 'JavaThread*' [-Werror=format=]
>>>> /hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2868:112:
>>>> error:
>>>> format '%lx' expects argument of type 'long unsigned int', but
>>>> argument 4
>>>> has type 'void*' [-Werror=format=]
>>>>          tty->print_cr("Exception <%s> (" INTPTR_FORMAT ")",
>>>>          except_oop->print_value_string(), (void*)except_oop());
>>>>
>>>> ^
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2872:27:
>>>> error:
>>>> format '%d' expects argument of type 'int', but argument 3 has type
>>>> 'address {aka unsigned char*}' [-Werror=format=]
>>>>                        THREAD);
>>>>                              ^
>>>> hotspot/src/share/vm/interpreter/bytecodeInterpreter.cpp:2872:27:
>>>> error:
>>>> format '%lx' expects argument of type 'long unsigned int', but
>>>> argument 4
>>>> has type 'JavaThread*' [-Werror=format=]
>>>>
>>>> 8048169: Change 8037816 breaks HS build on PPC64 and CPP-Interpreter
>>>> platforms
>>>> fixes that but doesn't seem to be in u20.
>>> Unfortunately the need for a 8u backport was not mentioned in the review
>>> thread:
>>>
>>> http://mail.openjdk.java.net/pipermail/hotspot-dev/2014-June/014453.html
>>>
>>> Did anyone flag this since 8048169 was pushed? I can't seem to find any
>>> email thread on it.
>>>
>>> I think we have a hole in the process when the "bug tail" for a
>>> backported fix is not also identified as needing an urgent backport.
>> I couldn't see anything in the thread either. I literally only
>> stumbled across
>> this all yesterday as I found the issue, and u20 was released right as
>> I was doing
>> it :(
>>
>>>> With that applied, it fails further
>>>> on in frame_zero.cpp. I don't see a fix for that in 9 either.
>>> See review thread:
>>>
>>> http://mail.openjdk.java.net/pipermail/hotspot-dev/2014-August/014825.html
>>>
>>>
>>> Not sure if it has been been pushed anywhere yet.
>> It's a different issue, again with format warnings. I have a fix for
>> 8u, but not
>> sure how applicable it is for 9. I haven't had a chance to look at
>> that yet.
>>
>> Actually, it's just this:
>>
>> diff -r 6533ac845e18 -r c75d6ce2e98b src/cpu/zero/vm/frame_zero.cpp
>> --- a/src/cpu/zero/vm/frame_zero.cpp    Mon Jun 30 14:58:52 2014 -0400
>> +++ b/src/cpu/zero/vm/frame_zero.cpp    Tue Aug 19 22:31:00 2014 +0100
>> @@ -216,7 +216,7 @@
>>       valuebuf[buflen - 1] = '\0';
>>       // Print the result
>> -    st->print_cr(" " PTR_FORMAT ": %-21s = %s", addr, fieldbuf,
>> valuebuf);
>> +    st->print_cr(" " PTR_FORMAT ": %-21s = %s", p2i(addr), fieldbuf,
>> valuebuf);
>>     }
>>   }
>>
>> I'm guessing 8048169 covered the PPC port use of the interpreter, but
>> this
>> file is only built as part of Zero.
>>
>>> David
>>> -----
>>>
>