On 12/02/2019 8:28 am, Franco Gastón Pellegrini wrote:
My mistake to not fully describe it but I tried jdk 11 and 12 using the disable warning as error on both, getting similar errors. (Both 32 bits)

The "operator delete" issue has already been encountered:

https://bugs.openjdk.java.net/browse/JDK-8196880

But I can't see how that relates to share/code/codeBlob.cpp and only for 32-bit. That would be a question for hotspot-...@openjdk.java.net

David

El lun., 11 de feb. de 2019 19:26, David Holmes <david.hol...@oracle.com <mailto:david.hol...@oracle.com>> escribió:

    On 12/02/2019 12:03 am, Alexey Ivanov wrote:
     > Hi David,
     >
     > On 11/02/2019 11:33, David Holmes wrote:
     >> Hi Alexey,
     >>
     >> On 11/02/2019 7:12 pm, Alexey Ivanov wrote:
     >>> Hi Franco,
     >>>
     >>> On 11/02/2019 02:03, David Holmes wrote:
     >>>> On 8/02/2019 5:07 am, Franco Gastón Pellegrini wrote:
     >>>>> I tried compiling JDK 12 for 32 bits, and I get similar errors:
     >>>>>
     >>>>> === Output from failing command(s) repeated here ===
     >>>>> * For target
    hotspot_variant-client_libjvm_objs_classFileParser.obj:
     >>>>> classFileParser.cpp
     >>>>>
    
c:/cygwin64/home/franc/java/jdk12/src/hotspot/share/classfile/classFileParser.cpp(312):

     >>>>> error C2220: warning treated as error - no 'object' file
    generated
     >>>>>
    
c:/cygwin64/home/franc/java/jdk12/src/hotspot/share/classfile/classFileParser.cpp(312):

     >>>>> warning C4267: '=': conversion from 'size_t' to 'u2',
    possible loss
     >>>>> of data
     >>>>>     ... (rest of output omitted)
     >>>
     >>> I used --disable-warnings-as-errors option to configure when I
    built
     >>> JDK for 32 bit Windows last time.
     >>
     >> Note there are actual errors in Franco's log not just the
    warnings I
     >> commented on. I think we've already encountered the operate delete
     >> issue so it may be fixed in 13, but not 12. Need to track it down.
     >
     > I must have missed that point.
     > The message above says “error C2220: warning treated as error”, so I
     > decided to let Franco know --disable-warnings-as-errors helped me.

    Yeah sorry - getting confused by the different things Franco tried. He
    tried a JDK11 build with --disable-warnings-as-errors but it still
    failed because of real errors:

    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(229):

    error C2956: sized deallocation function 'operator delete(void*,
    size_t)' would be chosen as placement deallocation function.
    predefined C++ types (compiler internal)(44): note: see declaration of
    'operator delete'

    Then he posted that he'd tried jdk12 as well with "similar errors" but
    only showed the

    error C2220: warning treated as error - no 'object' file generated
    warning C4267: '=': conversion from 'size_t' to 'u2', possible loss
    of data

    which means he was not running with --disable-warnings-as-errors

    David
    -----


     >
     > It's been a long while since I built jdk-dev on 32 bit Windows last
     > time; many things could have changed there.
     >
     > Regards,
     > Alexey
     >
     >>
     >> Cheers,
     >> David
     >>
     >>> David pointed to JDK-8205677 which should be forwardported to
    13 to
     >>> fix the problem.
     >>>
     >>> Regards,
     >>> Alexey
     >>>
     >>>>
     >>>> This is an issue we ran into compiling 8u with VS2017:
     >>>>
     >>>> https://bugs.openjdk.java.net/browse/JDK-8205677
     >>>>
     >>>> We probably never noticed in JDK 12 because we don't do 32-bits
     >>>> builds with VS2017.
     >>>>
     >>>> David
     >>>> -----
     >>>>
     >>>>> * All command lines available in
     >>>>>
    
/cygdrive/c/cygwin64/home/franc/java/jdk12/build/windows-x86-client-fastdebug/make-support/failure-logs.

     >>>>>
     >>>>> === End of repeated output ===
     >>>>>
     >>>>> No indication of failed target found.
     >>>>> Hint: Try searching the build log for '] Error'.
     >>>>> Hint: See doc/building.html#troubleshooting for assistance.
     >>>>>
     >>>>> make[1]: *** [/home/franc/java/jdk12/make/Init.gmk:310: main]
    Error 2
     >>>>> make: *** [/home/franc/java/jdk12/make/Init.gmk:186: default]
    Error 2
     >>>>>
     >>>>>
     >>>>> El mié., 6 de feb. de 2019 a la(s) 19:23, Franco Gastón
    Pellegrini
     >>>>> (francogpellegr...@gmail.com
    <mailto:francogpellegr...@gmail.com>
    <mailto:francogpellegr...@gmail.com
    <mailto:francogpellegr...@gmail.com>>)
     >>>>> escribió:
     >>>>>
     >>>>>     I just tried --disable-warnings-as-errors, and JDK 11
    64bits as a
     >>>>>     bootjdk, but I get a lots of errors, and it refuse to build,
     >>>>> like this:
     >>>>>
     >>>>>
    
c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/classfile/classFileParser.cpp(310):

     >>>>>
     >>>>>     warning C4267: '=': conversion from 'size_t' to 'u2',
    possible
     >>>>> loss
     >>>>>     of data
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(229):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(250):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(289):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(312):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(333):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(372):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(437):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
    c:/cygwin64/home/franc/java/jdk11/src/hotspot/share/code/codeBlob.cpp(541):

     >>>>>
     >>>>>     error C2956: sized deallocation function 'operator
    delete(void*,
     >>>>>     size_t)' would be chosen as placement deallocation function.
     >>>>>     predefined C++ types (compiler internal)(44): note: see
     >>>>> declaration
     >>>>>     of 'operator delete'
     >>>>>
     >>>>> <SNIP>
     >>>
     >

Reply via email to