insight/123: Stepping into a void function and then Finishing it causes Insight to malfunction

[Douglas dot Swanson at compaq dot com]

>Number:         123
>Category:       insight
>Synopsis:       Stepping into a void function and then Finishing it causes Insight to malfunction
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Jan 28 09:33:00 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Douglas.Swanson@compaq.com
>Release:        unknown-1.0
>Organization:
>Environment:
Win2K / MinGW 2.95.3 / Gdb 5 / Insight 
>Description:
I attach a small program that illustrates the failure.
In a Dos window I compile this program with MinGW (2.95.3).
I invoke Insight and step into (s) update().  After a 
few iterations of the nested "for" loops I finish (f)
update().  Insight correctly highlights the invocation
in main(), but now if I try to continue (c) nothing
happens.

I have a larger program from which this small example
was derived.  In that program after Insight returns to
the invocation of update() subsequent next (n) commands
provoke the following dialog box.

Warning
Program received signal SIGSEGV, Segmentation fault.

Note:
 1) Running the same executable under native gdb does
not provoke this problem.
 2) If I do not step into (s) the void function and
finish (f) it, the problem does not occur.
>How-To-Repeat:
Please see Description
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: