[LINUX5250] Re: 5250

[Scott Klement <klemscot@xxxxxxxxxxxx>]

I took a look at your tracefile, and I'm wondering if this might be an MDT
issue?   I know that there was some talk about something being wrong with
MDT handling with the AS/36, and maybe this is related.

Here's where the MDT problems were last discussed last time:
    http://archive.midrange.com/linux5250/200107/msg00011.html
and here Jay Felice posted a patch:
    http://archive.midrange.com/linux5250/200107/msg00041.html

Here is the same patch, except this one is updated to use the current
version of tn5250 from the 0.17.x branch:

Index: session.c
===================================================================
RCS file: /cvsroot/tn5250/tn5250/src/session.c,v
retrieving revision 1.56
diff -u -r1.56 session.c
--- session.c   20 Mar 2002 08:27:17 -0000      1.56
+++ session.c   22 Mar 2002 22:21:10 -0000
@@ -354,8 +354,7 @@
    switch (This->read_opcode) {
    case CMD_READ_INPUT_FIELDS:
       TN5250_ASSERT(aidcode != 0);
-      if (tn5250_dbuffer_mdt(dbuffer)
-           && tn5250_dbuffer_send_data_for_aid_key(dbuffer,aidcode)) {
+      if (tn5250_dbuffer_send_data_for_aid_key(dbuffer,aidcode)) {
         field = dbuffer->field_list;
         if (field != NULL) {
            do {
@@ -367,14 +366,12 @@
       break;

    case CMD_READ_IMMEDIATE:
-      if (tn5250_dbuffer_mdt(dbuffer)) {
-        field = dbuffer->field_list;
-        if (field != NULL) {
-           do {
-              tn5250_session_send_field (This, &field_buf, field);
-              field = field->next;
-           } while (field != dbuffer->field_list);
-        }
+      field = dbuffer->field_list;
+      if (field != NULL) {
+            do {
+                tn5250_session_send_field (This, &field_buf, field);
+                field = field->next;
+            } while (field != dbuffer->field_list);
       }
       break;


Heather, if you could try this patch and see if the solves the problem,
that would be great.  Otherwise, I can't really test things myself, since
I'm not sure how to reproduce the problem on my system. (But, if you could
tell me how to reproduce them, that would be even better still!)


On Fri, 22 Mar 2002, Heather Guse wrote:

> I am working remotely on this issue.  What I have done is installed the
> newes tar you had given me on two machines.  The machine I ultimately
> need to fix this on is remote and accessible only by dial-up.  I am
> trying to find the answer before I drive out there, and that way I can
> bring a CD with needed files/configs.  Last night I installed the
> tarballs on two systems that were having the same problems.  One is a RH
> 7.2 (just like the one I am trying to fix) and one is a RH 7.1.  One of
> the things that is not clear to me is what else I need to setup after
> the installation of tn5250.  Such as keyboard mappings and term types.
> I have been using tn5250 3 ways: At the console, with Xwindows, and over
> ssh.  They all have the same result, which you will see when you run the
> debug.  FYI, I can't seem to get fn keys working properly, so I have
> been using the ESC-n combo.
>
> The tracefile:
> You will see in the trace file that just before I sign
> off, I have put in the location code...I then receive a
> description...and if I press enter I should get back...more information
> --I do not.  This is the best example I can give you.
>
> Thank you for looking into this,
> Heather
>