I'm a little confused. What code did you apply javap to? I dont
think I pushed a copy of the newly revised code to any repos so you'd
have to build it yourself.
The new code looks like this:
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
SubjectCRIKey that = (SubjectCRIKey) o;
if (hashcode != that.hashcode) return false;
if (cri != null ? !cri.equals(that.cri) : that.cri !=
null) return false;
if (subject != null ? !subject.equals(that.subject) :
that.subject != null) return false;
return true;
}
I can believe the old code resembled your deconstruction below with
the addition of a "return".
thanks
david jencks
On May 21, 2009, at 12:08 AM, Bert_nor wrote:
I think this is a compiler bug.
javap gives next code:
...........
26: getfield #2; //Field subject:Ljavax/security/auth/Subject;
29: ifnonnull 42
32: aload_2
33: getfield #2; //Field subject:Ljavax/security/auth/Subject;
36: ifnonnull 91
39: goto 87
42: aload_0
43: getfield #2; //Field subject:Ljavax/security/auth/Subject;
46: aload_2
47: getfield #2; //Field subject:Ljavax/security/auth/Subject;
50: invokevirtual #8; //Method
javax/security/auth/Subject.equals:(Ljava/lang/Object;)Z
53: ifeq 73 <---------------------
if(subject.equals(o.subject))
56: aload_0
57: getfield #3; //Field
cri:Ljavax/resource/spi/ConnectionRequestInfo;
60: ifnonnull 73
63: aload_2
64: getfield #3; //Field
cri:Ljavax/resource/spi/ConnectionRequestInfo;
67: ifnonnull 91
70: goto 87
73: aload_0 <----------------------
74: getfield #3; //Field
cri:Ljavax/resource/spi/ConnectionRequestInfo;
77: aload_2
78: getfield #3; //Field
cri:Ljavax/resource/spi/ConnectionRequestInfo;
81: invokevirtual #9; //Method
java/lang/Object.equals:(Ljava/lang/Object;)Z
84: ifeq 91
87: iconst_1
88: goto 92
91: iconst_0
92: ireturn
The result is the following construction
if(subject.equals(o.subject)){
cri.equals(o.cri);
}
djencks wrote:
I opened GERONIMO-4639 and fixed this in trunk and branches 2.1
I couldn't understand the nested iff statements too well so I just
replaced it with what idea generates for equals :-)
This is a component, not part of geronimo itself. We'll need to push
a release of at least the 2.1 branch.
thanks
david jencks
--
View this message in context:
http://www.nabble.com/NullPointerException-MultiPoolConnectionInterceptor%24SubjectCRIKey.equals-tp23634940s134p23648828.html
Sent from the Apache Geronimo - Users mailing list archive at
Nabble.com.
