Re: [SCM] master: Add a test testMemoryBufferCapacity() to TestByteBuffer.java

[Zhao Shujing <pearly dot zhao at oracle dot com>]

It's been updated and the test is passing now.
On Tue, 2007-11-27 at 09:45 -0500, Andrew Cagney wrote:
> Zhao Shujing wrote:
> > Sorry, this test is added for bug 1538. I think .capacity() can use for
> > test if highWater mark is the top location when map memory to
> > ByteBuffer.
> >   
> True; however did you see:
> 
> > +               = new ExecOffspring(new ExecCommand (new String[] {
> > +                                                "/bin/rm",
> > +                                                tmpFile.toString()
> >   
> >   
> > I'm not sure that ExecOffspring is what you need here as it doesn't put 
> > the task into an attached/stopped state; I suspect:
> >    Task = new DaemonBlockedAtEntry("funit-slave").getMainTask().getISA();
> > be simpler?
> and:
> 
> > +       Task task = child.findTaskUsingRefresh(true);
> > +       if (task.getISA() == ISA.IA32)
> > +               assertEquals("Memory Buffer Capacity: ", 0xffffffffL,
> > +                    task.getMemory().capacity());
> > +       if (task.getISA() == ISA.X8664)
> > +               assertEquals("Memory Buffer Capacity: ",
> > 0xffffffffffffffffL,
> > +                     task.getMemory().capacity());
> >      }
> >   
> >   
> > There is also the PPC;  the attribute is dependent on the ISA's word 
> > size, though, is something like:
> >   switch (task.getISA().wordSize())
> >      case 4: assert ... ; break;
> >     case 8: assert ... ; break;
> >      default: fail("unknown word size");
> >   }
> > simplier?
> >
> > is the test passing or still being skipped?
> can we address these?
> 
> Andrew
>