[rfc][icedtea-web] synced headers between PLUGIN_ERROR, PLUGIN_DEBUG and java

[Omair Majid]

* Jiri Vanek <jvanek at redhat.com> [2013-11-01 07:45]:
> On 10/31/2013 07:10 PM, Omair Majid wrote:
> >Hi Jiri,
> >
> >* Jiri Vanek <jvanek at redhat.com> [2013-10-31 13:12]:
> >>diff -r 14a8ee171687 netx/net/sourceforge/jnlp/util/logging/OutputController.java
> >>--- a/netx/net/sourceforge/jnlp/util/logging/OutputController.java	Wed Oct 30 10:36:03 2013 +0100
> >>+++ b/netx/net/sourceforge/jnlp/util/logging/OutputController.java	Thu Oct 31 17:57:00 2013 +0100
> >>@@ -346,6 +346,7 @@
> >>              if (stack != null) {
> >>                  sb.append('[').append(getCallerClass(stack)).append(']');
> >>              }
> >>+            sb.append(" NETX Thread# ").append(Integer.toHexString(((Object)Thread.currentThread()).hashCode())).append(", name ").append(Thread.currentThread().getName());
> >
> >Some nits, here.
> >
> >You probably don't need to cast Thread.currentThread() to Object,
> >hashCode is defined on all types.
> 
> I think I wont...I wont to be  clear that number I'm showing is
> Object's hash -"pointer". Right now Thread do not override it, but
> can...

Thanks to dynamic dispatch, there's no way to do what you are trying to
do. Casting to Object won't make any difference. If Thread ever overrides
hashCode(), it will be still be Thread.hashCode() that's invoked.

> Afaik gthread, nor "pthread" do not have names....

There is a GNU extension to pthread that allows setting the name.
pthread_setname_np() (np == non-portable). This isn't a recommendation
to use that function :)

Patch itself looks okay to me, but someone more familiar with C++ is
welcome to chime in.

Thanks,
Omair