Re: Deep bug

[Scott Klement <klemscot@xxxxxxxxxxxx>]

I tried this... I took a file that has more than 3 screens of fields,
and followed the steps you gave below.  (Yes, the steps make sense)

I was NOT able to crash the emulator, though.  I tried both version
0.17.1 and 0.16.1, neither crashed. (but... see my previous message
about this -- I have a feeling we're not working from the same code)

This sounds suspciously like a problem that I fixed back in June of
last year...   Tell me, does it still do this if you hit enter slowly?

Also... does it happen for new/blank records?  (Even if you have to 
add one to test with?)


On Tue, 17 Apr 2001, Jason Bechtel wrote:

> Howdy,
> 
> Replying to my own initial message...
> 
> My client has now seen what he believes to be the same bug when using
> the common AS/400 "dfu" utility.  He was examining a file and moving
> through screens when the emulator died a similar death.  Here is what he
> said he did:
> 
> strdfu<Enter>
> 5<Enter>
> {choose the file, library(?), and member}
> PgDn  (shows fields for first record)
> Enter (shows more fields)
> Enter (shows more fields)
> Enter (shows more fields)
> * crashes here *
> Enter (shows more fields and loops back to the beginning)
> F3 (End)
> 
> I'm not an AS/400 guy, so I have no idea if this even makes sense.  He
> says you need a table that has many fields in order to have to press
> three <Enter>'s.  He thinks it might be layering the "windows" and that
> the third one is like the third window in their application.  This would
> substantiate his intuition that it is related to multiple layers of
> "windows".
> 
> At least this one is easier to reproduce...  :)
> 
> Jason
> 
> 
> Jason Bechtel wrote:
> > 
> > Hi all,
> > 
> > A client of mine using the xt5250 and tn5250 applications compiled from
> > source (version 0.15.6) has encountered a reproduceable crash with core
> > dump.  From their informal observation, it appears that the problem
> > occurs when they are three "screens" deep (you know, those little AS400
> > ASCII "windows").  Here is the message that appears when tn5250 is run
> > out of a terminal:
> > 
> >         Assertion tn5250_attribute (attr) failed at wtd.c, line 329.  A
> >         borted (core dumped)
> > 
> > Sure enough, that part of the code looks to be an assertion:
> > 
> >     298 /****i* lib5250/tn5250_wtd_context_write_field
> >     299  * NAME
> >     300  *    tn5250_wtd_context_write_field
> >     301  * SYNOPSIS
> >     302  *    tn5250_wtd_context_write_field (This, field, attr);
> >     303  * INPUTS
> >     304  *    Tn5250WTDContext *   This       -
> >     305  *    Tn5250Field *        field      -
> >     306  *    unsigned char        attr       -
> >     307  * DESCRIPTION
> >     308  *    Write the SF order and the field's information.
> >     309  *****/
> >     310 static void tn5250_wtd_context_write_field (Tn5250WTDContext
> > *This, Tn5250Field *field, unsigned char attr)
> >     311 {
> >     312    TN5250_LOG (("Writing SF order in stream data.\n"));
> >     313
> >     314    tn5250_wtd_context_putc (This, SF);
> >     315
> >     316    /* Put the field format word. */
> >     317    if (field->FFW != 0) {
> >     318       tn5250_wtd_context_putc (This, (unsigned char)(field->FFW
> > >> 8));
> >     319       tn5250_wtd_context_putc (This, (unsigned char)(field->FFW
> > & 0x00ff));
> >     320
> >     321       /* Put the field control word(s). */
> >     322       if (field->FCW != 0) {
> >     323          tn5250_wtd_context_putc (This, (unsigned
> > char)(field->FCW >> 8));
> >     324          tn5250_wtd_context_putc (This, (unsigned
> > char)(field->FCW & 0x00ff));
> >     325       }
> >     326    }
> >     327
> >     328    /* Put the screen attribute. */
> > *-> 329    TN5250_ASSERT (tn5250_attribute (attr));
> >     330    tn5250_wtd_context_putc (This, attr);
> >     331
> >     332    /* Put the field length. */
> >     333    tn5250_wtd_context_putc (This, (unsigned char)(field->length
> > >> 8));
> >     334    tn5250_wtd_context_putc (This, (unsigned char)(field->length
> > & 0x00ff));
> >     335 }
> > 
> > Anyone have any ideas what might be causing this error?  Has this
> > already been found and solved (new to the list)?  How difficult would
> > this be to debug?
> 

+---
| This is the LINUX5250 Mailing List!
| To submit a new message, send your mail to LINUX5250@midrange.com.
| To subscribe to this list send email to LINUX5250-SUB@midrange.com.
| To unsubscribe from this list send email to LINUX5250-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---