[9] RFR: JI-9039953 zero variant build is broken.

[Severin Gehwolf]

Hi Coleen,

On Mon, 2014-08-04 at 13:48 -0400, Coleen Phillimore wrote:
> Hi,
> 
> Did you try to build fastdebug?  I found 2 more compilation errors when 
> I did that.
> 
> http://cr.openjdk.java.net/~coleenp/8003426/src/cpu/zero/vm/cppInterpreter_zero.cpp.udiff.html
> http://cr.openjdk.java.net/~coleenp/8003426/src/share/vm/interpreter/bytecodeInterpreter.cpp.udiff.html
> 
> Although the last one could be due to the version of gcc I was using 
> (not sure).

I did not try fastdebug, since both debug builds - fastdebug and
slowdebug - are broken for my platform currently (even for
non-zero-variant builds). See:
http://mail.openjdk.java.net/pipermail/hotspot-dev/2014-August/014820.html
http://mail.openjdk.java.net/pipermail/build-dev/2014-July/012900.html

As for a release type zero variant build I only needed the change in the
webrev below. I'll have a look and see if I can work around the generic
debug build problem and will report back as soon as I know more.

Thanks,
Severin

> Coleen
> 
> On 8/4/14, 11:50 AM, Severin Gehwolf wrote:
> > Hi,
> >
> > Building a zero variant of hotspot fails for the latest hs-comp tree
> > with ("release"-type build):
> > src/cpu/zero/vm/frame_zero.inline.hpp:59:11: error: ‘CodeCache’ has not
> > been declared
> >
> > It appears to have been caused by a change of header files (perhaps
> > [1]?). A file which used to be included implicitly does not seem to be
> > included any longer. An explicit include fixes the problem.
> >
> > [2] shows the build failure.
> >
> > I would need a proper JDK bug and a sponsor for this change. An initial
> > bug has been submitted as JI-9039953 via the web interface.
> >
> > Webrev is here:
> > https://jerboaa.fedorapeople.org/bugs/openjdk/JI-9039953/webrev.01/
> >
> > Thanks,
> > Severin
> >
> > [1] http://hg.openjdk.java.net/jdk9/hs-comp/hotspot/rev/22b98ab2a69f
> > [2]
> > http://builder.classpath.org/jenkins/job/OpenJDK9_hscomp_Zero/47/console
> >
>