Rick McGuire wrote:
The other possibility is you've called RexxCreateInterpreter() on a
thread that was already owned by Rexx. In that case, the
RexxThreadContext returned with the instance pushes down the active
one and can create the same symptoms if you return without destroying
the instance. In that situation, you should either create the
instance on a separate thread or just inherit the instance of the
program that called you in the first place.
In the scenario that rexx.exe is used to run the program (which then
fails), no RexxCreateInterpreter() is carried out by BSF4Rexx.
---rony
On Mon, May 25, 2009 at 5:08 AM, Rony G. Flatscher
rony.flatsc...@wu-wien.ac.at wrote:
Rick,
when invoking that program via Java, then it works without a problem!
If invoked via Java then in the native layer an AttachThread() is invoked, a
Routine object is created and invoked with CallRoutine(), followed by a
matching DetachThread().
---rony
Rony G. Flatscher wrote:
Rick McGuire wrote:
I just thought of one more thing that could cause the error you're
seeing. If you do an AttachThread() at some point and neglect to do a
corresponding DetachThread() before returning to your caller, you'll
end up with a corrupted activaation stack that will result in problems
with legacy callback APIs like RexxVariablePool or RexxStemSort.
What;s showing up in this traceback very much appears to match what
might happen in that sort of situation.
In this particular program an external function BSF(), defined as
RexxRoutine1(RexxObjectPtr, BSF, ARGLIST, argArray) {...}
gets invoked to communicate to the Java side. There are no calls from Java
into BSF4Rexx occurring, hence there AttachThread() and DetachThread() are
not invoked in this case (at least not from BSF4Rexx).
Also, BSF4Rexx in this case does not excercise any variable gets or sets. It
just interacts with Java and returns a (string) value back, which gets
stored in the stem.
---rony
P.S.: Just for completeness, here is the Rexx program in question:
/* classic Rexx version, ---rgf, 2003-02, 2003-05-10, 2003-09-04, 2005-12-28
*/
/* this version works on Java 1.1 too (because using wrapEnumeration()) */
/* load the BSF4Rexx functions and start a JVM, if necessary */
if rxFuncQuery(BSF) = 1 then /* BSF() support not loaded yet ? */
do
call rxFuncAdd BsfLoadFuncs, BSF4Rexx, BsfLoadFuncs
call BsfLoadFuncs /* load all BSF*() functions */
call BsfLoadJava /* load Java */
end
/* 'Class.class' is pre-registered for Rexx, cf. docs */
system=bsf('invoke', 'Class.class', 'forName', 'java.lang.System')
properties=bsf('invoke', system, getProperties) /* get the System
properties */
enum=bsf('invoke', properties, 'propertyNames')/* get an enumeration of
the property names */
/* wrap the Enumeration object, so Java 1.1 can handle this too (overcome
Java inner class access restriction) */
/* the following statement is *not* necessary for Java = 1.2 */
enum=bsf('wrapEnumeration', enum)
say copies(=, 70)
stem. = 0
deli = 010203x
do i=1 to 5000 while bsf('invoke', enum, 'hasMoreElements') /* loop over
enumeration */
key=bsf('invoke', enum, 'nextElement')
stem.i= key || deli || bsf('invoke', properties, 'getProperty', key)
stem.0=i /* remember total of elements */
end
say stem.0 elements, now sorting...
call sysStemSort stem.
l=length(stem.0)
do i=1 to stem.0
parse var stem.i key (deli) val
say i~right(l): pp(key)~left(31)=pp(val)
end
exit
pp: procedure
return [ || arg(1) || ]
--
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing,
iPhoneDevCamp asthey present alongside digital heavyweights like Barbarian
Group, R/GA, Big Spaceship. http://www.creativitycat.com
___
Oorexx-devel mailing list
Oorexx-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/oorexx-devel
