Just filed a bug (with the code attached) at <https://sourceforge.net/p/oorexx/bugs/1789/>.
---rony On 18.11.2021 13:38, Rony G. Flatscher wrote: > For completeness: now had the possibility to create and test three versions > of ooRexx r12308: > "release", "releaseWithDebugInfo" and "Debug": "release" and > "releaseWithDebugInfo" crash always at > the same Rexx instance creation (# 168 on 32-bit Windows). The "debug" > version runs without a crash. > > ---rony > > P.S.: The Tomcat 9 Java web server resides on the same Windows 32-bit machine > using the installed > 32-bit ooRexx to serve RexxServerPages/JSP. The Java exception handler kicks > in and creates > hs_error*.log files, unfortunately without the rexx.dll symbols which should > be available with the > "releaseWithDebugInfo". > > On 17.11.2021 19:22, Rony G. Flatscher wrote: >> Just another observation: tested the Java web server with the same Rexx JSP >> with the 32-bit DEBUG >> version of ooRexx from trunk and it does not crash! >> >> ---rony >> >> >> On 19.10.2021 14:15, Rony G. Flatscher wrote: >>> Tested against a 64-bit ooRexx (r12288) on Linux: the crashes occur there >>> too, however at around >>> 400 RexxCreateInterpreter() invocations. >>> >>> ---rony >>> >>> On 19.10.2021 13:23, Rony G. Flatscher wrote: >>>> While testing a simple Rexx script in a JSP there are crashes in >>>> RexxCreateInterpreter(), always >>>> at the same invocation number (at # 168 plus a primodal interpreter >>>> instance). >>>> >>>> Each of the 168 JSP-requests will cause a RexxInterpreter instance to be >>>> created to run the Rexx >>>> scripts of a particular JSP invocation. >>>> >>>> Here a snippet of the Java (32-bit, Java 8) hs_error logfile after >>>> creating a debug version of >>>> ooRexx (32-bit, r12297): >>>> _______________________________________________ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
