Request for review (M): 6778669: Patch from Red Hat -- fixes compilation errors

[Christian Thalinger]

On Mon, 2009-02-23 at 11:43 -0800, Tom Rodriguez wrote:
> Given the way that NameList is being abused in the code, I think (int) 
> (intptr_t) preserves the intent of storing ints so I think I prefer  
> your patch below.

Here is the second version of the patch:

http://cr.openjdk.java.net/~twisti/6778669/webrev.01/

I added +w to the SunCC compiler flags to find all bugs that GCC could
probably complain about.  I will remove that flag before commit unless I
can fix the remaining warnings like these:

"/export/home/twisti/hotspot-comp/6778669/src/share/vm/adlc/formssel.hpp", line 1017: Warning: MatchRule::append_components hides the function MatchNode::append_components(FormDict&, ComponentList&, bool) const.
"/export/home/twisti/hotspot-comp/6778669/src/share/vm/adlc/formssel.hpp", line 1017: Warning: MatchRule::cisc_spill_match hides the function MatchNode::cisc_spill_match(FormDict&, RegisterForm*, MatchNode*, const char*&, const char*&).
"/export/home/twisti/hotspot-comp/6778669/src/share/vm/adlc/formssel.hpp", line 1017: Warning: MatchRule::equivalent hides the function MatchNode::equivalent(FormDict&, MatchNode*).
"/export/home/twisti/hotspot-comp/6778669/src/share/vm/adlc/formssel.hpp", line 1017: Warning: MatchRule::swap_commutative_op hides the function MatchNode::swap_commutative_op(bool, int).

Unfortunately I don't know how to fix them properly.  But it would be
nice to have the +w flag in there as it found a lot of "bugs", like the
printf ones.

-- Christian