SEVERE: IO error -- builtin/javasupport (LoadError)


#1

I am seeing the error below when trying to run jruby 1.2.0 and Rails
2.3.2
application on GlassFish. Any idea why I should see this error? It might
be
some classpath issue but if someone could tell what is that JRuby is
looking
for that is not available will be really helpful.

INFO: Deployment of bookstore done is 1,009 ms
Apr 17, 2009 2:23:09 PM
SEVERE: Exception in thread “pool-14-thread-1”
Apr 17, 2009 2:23:09 PM
SEVERE: :1:
Apr 17, 2009 2:23:09 PM
SEVERE: IO error – builtin/javasupport (LoadError)
Apr 17, 2009 2:23:09 PM
SEVERE: …internal jruby stack elided…
Apr 17, 2009 2:23:09 PM
SEVERE: from (unknown).(unknown)(:1)

-vivek.


#2

Pass -d and you should get a better backtrace. That LoadError usually
means something failed to compile.

Vivek P. wrote:

Apr 17, 2009 2:23:09 PM
SEVERE: IO error – builtin/javasupport (LoadError)
Apr 17, 2009 2:23:09 PM
SEVERE: …internal jruby stack elided…
Apr 17, 2009 2:23:09 PM
SEVERE: from (unknown).(unknown)(:1)

-vivek.


To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

#3

Ok, so it is ClassCircularityError emitted while compiling
builtin/javasupport.rb.

Apr 17, 2009 4:02:40 PM
SEVERE: java.lang.ClassCircularityError:
org/jruby/lexer/yacc/RubyYaccLexer$LexState
Apr 17, 2009 4:02:40 PM
SEVERE: at
org.jruby.lexer.yacc.RubyYaccLexer.yylex(RubyYaccLexer.java:917)
Apr 17, 2009 4:02:40 PM
SEVERE: at
org.jruby.lexer.yacc.RubyYaccLexer.advance(RubyYaccLexer.java:295)
Apr 17, 2009 4:02:40 PM
SEVERE: at
org.jruby.parser.DefaultRubyParser.yyparse(DefaultRubyParser.java:1407)

The full stack trace is at:

http://pastie.org/450344

-vivek.

On Fri, Apr 17, 2009 at 3:14 PM, Charles Oliver N. <


#4

Well that’s bizarre…I don’t even know how that would happen. Do a
little more investigation and if you can’t make it go away toss it into
a bug.

Vivek P. wrote:

Apr 17, 2009 4:02:40 PM

Vivek P. wrote:
    Apr 17, 2009 2:23:09 PM SEVERE: :1:
---------------------------------------------------------------------
To unsubscribe from this list, please visit:

  http://xircles.codehaus.org/manage_email

To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

#5

public enum LexState {
EXPR_BEG, EXPR_END, EXPR_ARG, EXPR_CMDARG, EXPR_ENDARG,
EXPR_MID,
EXPR_FNAME, EXPR_DOT, EXPR_CLASS, EXPR_VALUE
}

Pretty weird. A simple enum is causing this? Not even sure how it
could be circular since I do not know how enums compile down via
javac. Clearly something really weird with classloading?

-Tom

On Fri, Apr 17, 2009 at 6:10 PM, Vivek P. removed_email_address@domain.invalid
wrote:

Apr 17, 2009 4:02:40 PM

Apr 17, 2009 2:23:09 PM SEVERE: …internal jruby stack elided…


Blog: http://www.bloglines.com/blog/ThomasEEnebo
Email: removed_email_address@domain.invalid , removed_email_address@domain.invalid


To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

#6

It is coming from:

Ruby.java:

public void loadFile(String scriptName, InputStream in, boolean wrap) {
IRubyObject self = wrap ? TopSelfFactory.createTopSelf(this) :
getTopSelf();
ThreadContext context = getCurrentContext();
String file = context.getFile();

    try {
        secure(4); /* should alter global state */

        context.setFile(scriptName);
        context.preNodeEval(objectClass, self, scriptName);
  •        parseFile(in, scriptName, null).interpret(this, context, 
    

self,
Block.NULL_BLOCK);*
} catch (JumpException.ReturnJump rj) {
return;
} finally {
context.postNodeEval();
context.setFile(file);
}
}

Problem is jruby runtime can’t get created so nothing works on GlassFish
v3
trunk due to this issue!

-vivek.

On Fri, Apr 17, 2009 at 4:14 PM, Charles Oliver N. <


#7

Whew…I was worried. So we’re off the hook then? Let us know if you
need anything else.

Vivek P. wrote:

 > SEVERE:     at
 > The full stack trace is at:
usually
 >>> is looking for that is not available will be really helpful.
 >>>
 >
To unsubscribe from this list, please visit:

   http://xircles.codehaus.org/manage_email

To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

#8

Yes, of course.

thanks,

-vivek.

On Tue, Apr 21, 2009 at 10:27 AM, Charles Oliver N. <


#9

This appears to be a regression[1] in Felix 1.6.0. Basically, it was
caused
due to the fact that during the classload, felix was trying to introspec
the
inner classes to determine the enclosing class.

-vivek.

[1]https://issues.apache.org/jira/browse/FELIX-1045