#1244 Program exception under Windows XP SP3

None
open
nobody
None
none
1
2014-02-19
2014-02-12
No

ooRexx 4.2.0 release candidate 2
Windows XP SP3

The following code generates program exception 0xc0000005 at offset 00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3 with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not tested with 4.2.0):

q = .outtrap~exec("assoc")

::class outtrap public subclass rexxqueue
::method exec class
use arg command=""
queue=self~new(self~create)
if command<>"" then command '| rxqueue "'queue~get'"'
return queue
::method uninit
self~delete

Related

Bugs: #1244

Discussion

  • Mark Miesfeld

    Mark Miesfeld - 2014-02-13

    This in essence is a duplicate of [bugs:#1214], I think. Or at least the crash is.

    Rick - The crash seems easy enough to reproduce on Windows 7, by making rxapi unstartable:

    C:\Rexx\ooRexx>sc stop rxapi

    SERVICE_NAME: rxapi
    TYPE : 10 WIN32_OWN_PROCESS
    STATE : 3 STOP_PENDING
    (STOPPABLE, NOT_PAUSABLE, IGNORES_SHUTDOWN)
    WIN32_EXIT_CODE : 0 (0x0)
    SERVICE_EXIT_CODE : 0 (0x0)
    CHECKPOINT : 0x0
    WAIT_HINT : 0x0

    C:\Rexx\ooRexx>ren rxapi.exe txapi.exe

    C:\Rexx\ooRexx>cd ..

    Then run either Staffan's test program, of the simpler test program from bug 1214:

    q = rxqueue('create')

    Crashes at:

    inline SessionID getSession() { return session; }
    

    The this pointer is NULL.

    • this 0x0000000000000000 {restartRequired=??? connectionEstablished=??? session=??? ...} LocalAPIManager * const

    Rick, on my system running the test program brings up the crash dialog with the option to debug the program, but selecting it does not bring up the debugger. Running the test program directly under the debugger works fine:

    devenv /debugexe rexx.exe test.rex

    I think the pid file problem on unix-like systems needs to be addressed at some point, but not now.

    But raising an exception if rxapi can't be started rather than crashing probably should be addressed before we release 4.2.0. I'm not sure exactly where raising the exceptions should be done though.

     

    Related

    Bugs: #1214


    Last edit: Mark Miesfeld 2014-02-13
    • Rick McGuire

      Rick McGuire - 2014-02-13

      Mark,

      I've managed to reproduce this, but I seriously disagree that this is a
      4.2.0 must fix. This problem has existed since 4.0, and fixing it does not
      appear to be a simple task. I'm quite opposed to trying to cram in a fix
      that might prove very disruptive at this point in the release cycle.

      Rick

      On Thu, Feb 13, 2014 at 11:08 AM, Mark Miesfeld miesfeld@users.sf.netwrote:

      This in essence is a duplicate of [bugs:1214], I think. Or at least the
      crash is.

      Rick - The crash seems easy enough to reproduce on Windows 7, by making
      rxapi unstartable:

      C:\Rexx\ooRexx>sc stop rxapi

      SERVICE_NAME: rxapi
      TYPE : 10 WIN32_OWN_PROCESS
      STATE : 3 STOP_PENDING
      (STOPPABLE, NOT_PAUSABLE, IGNORES_SHUTDOWN)
      WIN32_EXIT_CODE : 0 (0x0)
      SERVICE_EXIT_CODE : 0 (0x0)
      CHECKPOINT : 0x0
      WAIT_HINT : 0x0

      C:\Rexx\ooRexx>ren rxapi.exe txapi.exe

      C:\Rexx\ooRexx>cd ..

      Then run either Staffan's test program, of the simpler test program from
      bug 1214:

      q = rxqueue('create')

      Crashes at:

      inline SessionID getSession() { return session; }
      

      The this pointer is NULL.

      • this 0x0000000000000000 {restartRequired=???
        connectionEstablished=??? session=??? ...} LocalAPIManager * const

      Rick, on my system running the test program brings up the crash dialog
      with the option to debug the program, but selecting it does not bring up
      the debugger. Running the test program directly under the debugger works
      fine:

      devenv /debugexe rexx.exe test.rex

      I think the pid file problem on unix-like systems needs to be addressed at
      some point, but not now.

      But raising an exception if rxapi can't be started rather than crashing
      probably should be addressed before we release 4.2.0. I'm not sure exactly
      where raising the exceptions should be done though.


      [bugs:#1244] Program exception under Windows XP SP3

      Status: open
      Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
      Last Updated: Wed Feb 12, 2014 06:52 PM UTC
      Owner: nobody

      ooRexx 4.2.0 release candidate 2
      Windows XP SP3

      The following code generates program exception 0xc0000005 at offset
      00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
      with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
      tested with 4.2.0):

      q = .outtrap~exec("assoc")

      ::class outtrap public subclass rexxqueue
      ::method exec class
      use arg command=""
      queue=self~new(self~create)
      if command<>"" then command '| rxqueue "'queue~get'"'
      return queue
      ::method uninit
      self~delete


      Sent from sourceforge.net because you indicated interest in <
      https://sourceforge.net/p/oorexx/bugs/1244/>

      To unsubscribe from further messages, please visit <
      https://sourceforge.net/auth/subscriptions/>

       

      Related

      Bugs: #1244

      • Rick McGuire

        Rick McGuire - 2014-02-13

        I think I have a fairly simple fix for this, but am still leery about
        pushing this into 4.2.0. I also need to try testing out a few more things
        first to make sure there are no side effects.

        Rick

        On Thu, Feb 13, 2014 at 11:49 AM, Rick McGuire bigrixx@users.sf.net wrote:

        Mark,

        I've managed to reproduce this, but I seriously disagree that this is a
        4.2.0 must fix. This problem has existed since 4.0, and fixing it does not
        appear to be a simple task. I'm quite opposed to trying to cram in a fix
        that might prove very disruptive at this point in the release cycle.

        Rick

        On Thu, Feb 13, 2014 at 11:08 AM, Mark Miesfeld <miesfeld@users.sf.net

        wrote:

        This in essence is a duplicate of [bugs:1214], I think. Or at least the
        crash is.

        Rick - The crash seems easy enough to reproduce on Windows 7, by making
        rxapi unstartable:

        C:\Rexx\ooRexx>sc stop rxapi

        SERVICE_NAME: rxapi
        TYPE : 10 WIN32_OWN_PROCESS
        STATE : 3 STOP_PENDING
        (STOPPABLE, NOT_PAUSABLE,
        IGNORES_SHUTDOWN)
        WIN32_EXIT_CODE : 0 (0x0)
        SERVICE_EXIT_CODE : 0 (0x0)
        CHECKPOINT : 0x0
        WAIT_HINT : 0x0

        C:\Rexx\ooRexx>ren rxapi.exe txapi.exe

        C:\Rexx\ooRexx>cd ..

        Then run either Staffan's test program, of the simpler test program from
        bug 1214:

        q = rxqueue('create')

        Crashes at:

        inline SessionID getSession() { return session; }
        

        The this pointer is NULL.

        • this 0x0000000000000000 {restartRequired=???
          connectionEstablished=??? session=??? ...} LocalAPIManager * const

        Rick, on my system running the test program brings up the crash dialog
        with the option to debug the program, but selecting it does not bring up
        the debugger. Running the test program directly under the debugger works
        fine:

        devenv /debugexe rexx.exe test.rex

        I think the pid file problem on unix-like systems needs to be addressed
        at
        some point, but not now.

        But raising an exception if rxapi can't be started rather than crashing
        probably should be addressed before we release 4.2.0. I'm not sure
        exactly
        where raising the exceptions should be done though.


        [bugs:#1244] Program exception under Windows XP SP3

        Status: open
        Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
        Last Updated: Wed Feb 12, 2014 06:52 PM UTC
        Owner: nobody

        ooRexx 4.2.0 release candidate 2
        Windows XP SP3

        The following code generates program exception 0xc0000005 at offset
        00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
        with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
        tested with 4.2.0):

        q = .outtrap~exec("assoc")

        ::class outtrap public subclass rexxqueue
        ::method exec class
        use arg command=""
        queue=self~new(self~create)
        if command<>"" then command '| rxqueue "'queue~get'"'
        return queue
        ::method uninit
        self~delete


        Sent from sourceforge.net because you indicated interest in <
        https://sourceforge.net/p/oorexx/bugs/1244/>

        To unsubscribe from further messages, please visit <
        https://sourceforge.net/auth/subscriptions/>


        [bugs:#1244] Program exception under Windows XP SP3

        Status: open
        Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
        Last Updated: Thu Feb 13, 2014 04:33 PM UTC
        Owner: nobody

        ooRexx 4.2.0 release candidate 2
        Windows XP SP3

        The following code generates program exception 0xc0000005 at offset
        00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
        with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
        tested with 4.2.0):

        q = .outtrap~exec("assoc")

        ::class outtrap public subclass rexxqueue
        ::method exec class
        use arg command=""
        queue=self~new(self~create)
        if command<>"" then command '| rxqueue "'queue~get'"'
        return queue
        ::method uninit
        self~delete


        Sent from sourceforge.net because you indicated interest in <
        https://sourceforge.net/p/oorexx/bugs/1244/>

        To unsubscribe from further messages, please visit <
        https://sourceforge.net/auth/subscriptions/>

         

        Related

        Bugs: #1244

      • Mark Miesfeld

        Mark Miesfeld - 2014-02-13

        Rick, that's fine, we don't need to fix this for 4.2.0. I had thought it would be a simple fix for you. But, if it isn't let's just let it go to after 4.2.0.

        If you give me a few hints, I'll work on it post 4.2.0.

         
        • Rick McGuire

          Rick McGuire - 2014-02-13

          I'll just commit what I have to the trunk version.

          Rick

          On Thu, Feb 13, 2014 at 12:09 PM, Mark Miesfeld miesfeld@users.sf.netwrote:

          Rick, that's fine, we don't need to fix this for 4.2.0. I had thought it
          would be a simple fix for you. But, if it isn't let's just let it go to
          after 4.2.0.

          If you give me a few hints, I'll work on it post 4.2.0.


          [bugs:#1244] Program exception under Windows XP SP3

          Status: open
          Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
          Last Updated: Thu Feb 13, 2014 04:33 PM UTC
          Owner: nobody

          ooRexx 4.2.0 release candidate 2
          Windows XP SP3

          The following code generates program exception 0xc0000005 at offset
          00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
          with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
          tested with 4.2.0):

          q = .outtrap~exec("assoc")

          ::class outtrap public subclass rexxqueue
          ::method exec class
          use arg command=""
          queue=self~new(self~create)
          if command<>"" then command '| rxqueue "'queue~get'"'
          return queue
          ::method uninit
          self~delete


          Sent from sourceforge.net because you indicated interest in <
          https://sourceforge.net/p/oorexx/bugs/1244/>

          To unsubscribe from further messages, please visit <
          https://sourceforge.net/auth/subscriptions/>

           

          Related

          Bugs: #1244

  • Mark Miesfeld

    Mark Miesfeld - 2014-02-13

    Staffan,

    If I'm correct about the crash on your XP system being due to rxapi is not startable, you should be able to work around this by figuring out why rxapi is not starting.

    I haven't had a chance to test on XP. I don't work on XP any more. I'll try to test it on a XP system tonight.

    It could be that there is a second bug here in the XP installation, or it could be that it is caused by something goofy you have done. ;-)

    If you select to install rxapi as a service when you run the Windows installer, then rxapi should always be running when you boot up. If you select not to install rxapi as a service, then why not? It is really the best option on Windows.

    Would you run a few more tests to pin this down and add a little info to the bug report.

    1.) Did you install rxapi as a service during the installation?

    2.) Ensure that rxapi is running on your system and rerun your test program. pslist is free and a nice utility to have:

    C:\Rexx>pslist rxapi

    pslist v1.28 - Sysinternals PsList
    Copyright ⌐ 2000-2004 Mark Russinovich
    Sysinternals

    Process information for KITE:

    process rxapi was not found on KITE

    C:\Rexx>rxapi

    C:\Rexx>pslist rxapi

    pslist v1.28 - Sysinternals PsList
    Copyright ⌐ 2000-2004 Mark Russinovich
    Sysinternals

    Process information for KITE:

    Name Pid Pri Thd Hnd Priv CPU Time Elapsed Time
    rxapi 2488 8 5 52 1104 0:00:00.015 0:00:13.297

    C:\Rexx>

    You can get pslist if you want from:

    http://technet.microsoft.com/en-us/sysinternals/bb896649.aspx?ppud=4

    Thanks.

     
  • Mark Miesfeld

    Mark Miesfeld - 2014-02-15

    Staffan, could you please follow up on this.

    I finally had a chance to test this on a XP system. There is no problem at all using the test you supplied. I ran it a dozen times. I downloaded the 4.2.0 release candidate 2 directly from SourceForge for the test.

    Could you try your test again. Make sure rxapi is running and see if it still crashes. If it doesn't, then the question is why wasn't rxapi starting for you.

    Thanks.

     
    • Staffan Tylen

      Staffan Tylen - 2014-02-16

      Mark,

      I uninstalled the existing ooRexx (4.2.0 RC2) on my XP system, downloaded
      4.2.0 RC2 again, installed it, ran the test program, and it still crashes
      on statement

      queue = self~new(self~create)

      I thought that Rick had located the error in ooRexx but that you both
      decided not to implement a solution at this stage for 4.2.0, or did I get
      that wrong? Why this works for you I cannot explain, are you on XP SP3?

      Staffan

      On Sat, Feb 15, 2014 at 6:54 PM, Mark Miesfeld miesfeld@users.sf.netwrote:

      Staffan, could you please follow up on this.

      I finally had a chance to test this on a XP system. There is no problem at
      all using the test you supplied. I ran it a dozen times. I downloaded the
      4.2.0 release candidate 2 directly from SourceForge for the test.

      Could you try your test again. Make sure rxapi is running and see if it
      still crashes. If it doesn't, then the question is why wasn't rxapi
      starting for you.

      Thanks.

      Status: open
      Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
      Last Updated: Thu Feb 13, 2014 04:33 PM UTC
      Owner: nobody

      ooRexx 4.2.0 release candidate 2
      Windows XP SP3

      The following code generates program exception 0xc0000005 at offset
      00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
      with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
      tested with 4.2.0):

      q = .outtrap~exec("assoc")

      ::class outtrap public subclass rexxqueue
      ::method exec class
      use arg command=""
      queue=self~new(self~create)
      if command<>"" then command '| rxqueue "'queue~get'"'
      return queue
      ::method uninit
      self~delete


      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/oorexx/bugs/1244/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       

      Related

      Bugs: #1244

  • Mark Miesfeld

    Mark Miesfeld - 2014-02-16

    Hi Staffan,

    I'm sorry if I'm not clear here, maybe I need to take writing lessons. You seem to be missing the point of the exercise. You're right, there is nothing changed that will fix the problem for you. It doesn't crash for me, or Rick, or thousands of other people, on XP with SP3 because we are not running the test on your system.

    1.) The information I'm trying to get from you is, does it crash if you are sure rxapi is running? To do that test you need to determine if rxapi is running at the instant you run the test. You need to use a command prompt utility that will show if rxapi is running or not, it would also be nice if you changed your create queue line to this to your test:

    queue=self~new(self~create); say 'queue:' queue

    and then run this test and copy and paste the information in this bug:

    C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

    pslist v1.28 - Sysinternals PsList
    Copyright ⌐ 2000-2004 Mark Russinovich
    Sysinternals

    Process information for KITE:

    Name Pid Pri Thd Hnd Priv CPU Time Elapsed Time
    rxapi 1620 8 4 74 1328 0:00:00.171 36:49:26.736

    C:\work.ooRexx\bug.work\staffan\queue.crash>staffanTest.rex
    queue: S0000000000000ED0Q00000000000A82D0

    C:\work.ooRexx\bug.work\staffan\queue.crash>

    If #1 shows that rxapi is not running then I have the next question:

    2.) If rxapi is not running, what happens if you manually start rxapi before running the test?

    C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

    pslist v1.28 - Sysinternals PsList
    Copyright ⌐ 2000-2004 Mark Russinovich
    Sysinternals

    Process information for KITE:

    process rxapi was not found on KITE

    C:\work.ooRexx\bug.work\staffan\queue.crash>rxapi

    C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

    pslist v1.28 - Sysinternals PsList
    Copyright ⌐ 2000-2004 Mark Russinovich
    Sysinternals

    Process information for KITE:

    Name Pid Pri Thd Hnd Priv CPU Time Elapsed Time
    rxapi 2456 8 5 50 1224 0:00:00.015 0:00:09.551

    C:\work.ooRexx\bug.work\staffan\queue.crash>staffanTest.rex
    queue: S0000000000000F40Q0000000000AA7D30

    C:\work.ooRexx\bug.work\staffan\queue.crash>

    Note that in this test it is important that after starting rxapi, you check again to be sure that it truly is running.

    Once we have the information from the above tests, then we have what we need to go further.

    • If rxapi is not running and you get the crash, and it doesn't crash when rxapi is running, then we have the explanation for the crash, and the question becomes: Why doesn't rxapi start for you?

    • If you get the crash when rxapi is running for sure, then the crash you see is not what we think it is.

    Until we get this basic information, everything is just pure speculation.

    Oh, to try and prevent so much back and forth can you also please answer these questions.

    3.) Did you install rxapi as a Windows service when you installed?

    4.) If you did install rxapi as a Windows service, does the service show correctly in Administrator Tools -> services ?

    5.) If you didn't install rxapi as a Windows service, what happens if you reinstall ooRexx and select to install rxapi as a Windows service? Does your test still crash?

    Thanks

     
    • Staffan Tylen

      Staffan Tylen - 2014-02-18

      Mark,

      Back from a short break I started up my XP system again and suddenly the
      test program works! I had tried it I don't know how many times before
      without success, the only explanation I can come up with is that after the
      fresh install of ooRexx 4.2.0 RC2 (with rxapi as a windows service) I
      didn't reboot, maybe that was the whole issue. Anyway I began testing my
      code on XP and it worked for a while, then something went wrong (it
      involved compiled code so I didn't get a hint of what went wrong) and after
      that the app no longer works, even after reinstall, reboot, you name it.
      While it was working I found that the GUI was not really what I was hoping
      for so I've now taken the decision to ignore XP completely. Its days are
      counted anyway.

      My apologies for taking up everyone's time with this.

      Staffan

      On Sun, Feb 16, 2014 at 6:59 PM, Mark Miesfeld miesfeld@users.sf.netwrote:

      Hi Staffan,

      I'm sorry if I'm not clear here, maybe I need to take writing lessons. You
      seem to be missing the point of the exercise. You're right, there is
      nothing changed that will fix the problem for you. It doesn't crash for me,
      or Rick, or thousands of other people, on XP with SP3 because we are not
      running the test on your system.

      1.) The information I'm trying to get from you is, does it crash if you
      are sure rxapi is running? To do that test you need to determine if rxapi
      is running at the instant you run the test. You need to use a command
      prompt utility that will show if rxapi is running or not, it would also be
      nice if you changed your create queue line to this to your test:

      queue=self~new(self~create); say 'queue:' queue

      and then run this test and copy and paste the information in this bug:

      C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

      pslist v1.28 - Sysinternals PsList
      Copyright ⌐ 2000-2004 Mark Russinovich
      Sysinternals

      Process information for KITE:

      Name Pid Pri Thd Hnd Priv CPU Time Elapsed Time
      rxapi 1620 8 4 74 1328 0:00:00.171 36:49:26.736

      C:\work.ooRexx\bug.work\staffan\queue.crash>staffanTest.rex
      queue: S0000000000000ED0Q00000000000A82D0

      C:\work.ooRexx\bug.work\staffan\queue.crash>

      If #1 shows that rxapi is not running then I have the next question:

      2.) If rxapi is not running, what happens if you manually start rxapi
      before running the test?

      C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

      pslist v1.28 - Sysinternals PsList
      Copyright ⌐ 2000-2004 Mark Russinovich
      Sysinternals

      Process information for KITE:

      process rxapi was not found on KITE

      C:\work.ooRexx\bug.work\staffan\queue.crash>rxapi

      C:\work.ooRexx\bug.work\staffan\queue.crash>pslist rxapi

      pslist v1.28 - Sysinternals PsList
      Copyright ⌐ 2000-2004 Mark Russinovich
      Sysinternals

      Process information for KITE:

      Name Pid Pri Thd Hnd Priv CPU Time Elapsed Time
      rxapi 2456 8 5 50 1224 0:00:00.015 0:00:09.551

      C:\work.ooRexx\bug.work\staffan\queue.crash>staffanTest.rex
      queue: S0000000000000F40Q0000000000AA7D30

      C:\work.ooRexx\bug.work\staffan\queue.crash>

      Note that in this test it is important that after starting rxapi, you
      check again to be sure that it truly is running.

      Once we have the information from the above tests, then we have what we
      need to go further.

      -

      If rxapi is not running and you get the crash, and it doesn't crash
      when rxapi is running, then we have the explanation for the crash, and the
      question becomes: Why doesn't rxapi start for you?
      -

      If you get the crash when rxapi is running for sure, then the crash
      you see is not what we think it is.

      Until we get this basic information, everything is just pure speculation.

      Oh, to try and prevent so much back and forth can you also please answer
      these questions.

      3.) Did you install rxapi as a Windows service when you installed?

      4.) If you did install rxapi as a Windows service, does the service show
      correctly in Administrator Tools -> services ?

      5.) If you didn't install rxapi as a Windows service, what happens if you
      reinstall ooRexx and select to install rxapi as a Windows service? Does
      your test still crash?

      Thanks

      Status: open
      Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
      Last Updated: Sat Feb 15, 2014 05:54 PM UTC
      Owner: nobody

      ooRexx 4.2.0 release candidate 2
      Windows XP SP3

      The following code generates program exception 0xc0000005 at offset
      00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
      with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
      tested with 4.2.0):

      q = .outtrap~exec("assoc")

      ::class outtrap public subclass rexxqueue
      ::method exec class
      use arg command=""
      queue=self~new(self~create)
      if command<>"" then command '| rxqueue "'queue~get'"'
      return queue
      ::method uninit
      self~delete


      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/oorexx/bugs/1244/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       
  • Mark Miesfeld

    Mark Miesfeld - 2014-02-19

    Hi Staffan.

    It's fine that you decided to ignore XP. But, you've left us with an open bug with no resolution.

    There is a simple explanation for why it worked and why it didn't work. When it worked rxapi was running and when it didn't work rxapi wasn't running and couldn't be started. Unfortunately, you still haven't tried to confirm that the simple explanation is correct or not.

    I'm not sure I understand why you can't check if rxapi is actually running when your program fails. And why you can't test if the program works if you manually start rxapi.

     
    • Staffan Tylen

      Staffan Tylen - 2014-02-19

      Hi Mark,

      Yes, you are right, I need to provide further info on this, my bad.

      I've now tried to recreate the problem and the good or bad thing is that I
      can't! I'm starting to believe that the old XP system that I've been
      testing this on might have had some irregularities that disappeared when I
      uninstalled the existing ooRexx (including the directory tree) and
      reinstalled 4.2.0 RC2. I have no other explanation. As nobody else has had
      any similar problems I guess you need to put this down as user error. The
      windows log still contains records of the crashes if somebody is
      interested. I hope this can be closed now, thanks for your noending support.

      Staffan

      On Wed, Feb 19, 2014 at 4:42 PM, Mark Miesfeld miesfeld@users.sf.netwrote:

      Hi Staffan.

      It's fine that you decided to ignore XP. But, you've left us with an open
      bug with no resolution.

      There is a simple explanation for why it worked and why it didn't work.
      When it worked rxapi was running and when it didn't work rxapi wasn't
      running and couldn't be started. Unfortunately, you still haven't tried to
      confirm that the simple explanation is correct or not.

      I'm not sure I understand why you can't check if rxapi is actually running
      when your program fails. And why you can't test if the program works if you
      manually start rxapi.


      Status: open
      Created: Wed Feb 12, 2014 06:52 PM UTC by Staffan Tylen
      Last Updated: Sun Feb 16, 2014 05:59 PM UTC
      Owner: nobody

      ooRexx 4.2.0 release candidate 2
      Windows XP SP3

      The following code generates program exception 0xc0000005 at offset
      00001e5a in application rexx.exe module rexxapi.dll under Windows XP SP3
      with ooRexx 4.2.0 release candidate 2 but works under Windows 7 (but not
      tested with 4.2.0):

      q = .outtrap~exec("assoc")

      ::class outtrap public subclass rexxqueue
      ::method exec class
      use arg command=""
      queue=self~new(self~create)
      if command<>"" then command '| rxqueue "'queue~get'"'
      return queue
      ::method uninit
      self~delete


      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/oorexx/bugs/1244/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       

      Related

      Bugs: #1244


Anonymous

Cancel  Add attachments





Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks