[Swig-devel] [ swig-Bugs-1684887 ] Arguments being scrambled (regression 1.3.27 -> 1.3.28)

[SourceForge.net <no...@so...>]

Bugs item #1684887, was opened at 2007-03-20 21:15
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101645&aid=1684887&group_id=1645

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: perl
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Rich Wales (richwales)
Assigned to: Nobody/Anonymous (nobody)
Summary: Arguments being scrambled (regression 1.3.27 -> 1.3.28)

Initial Comment:
I am maintaining a large application in Perl and C++.  It works fine using SWIG 1.3.27, but it broke when I tried rebuilding it using 1.3.28, or 1.3.31.  (I first noticed the problem with 1.3.31, and then I went back to the archives, found the older versions, and tried them one at a time till I determined that the breakage occurred between 1.3.27 and 1.3.28.)

Briefly, the problem I'm seeing involves a C++ routine with a long argument list, which is wrapped by SWIG so it can be called from the Perl part of the program.  The first two arguments are both integers.  For some reason, the second argument is being changed when I use 1.3.28; the value seen by the C++ routine is not the same as the value I passed in the Perl code (and this corruption of the function argument values causes the program to fail).

I tried adding an fprintf(stderr,...) to the generated wrapper.cc file, in the middle of the wrapper function's argument-grabbing sequences, and I confirmed that the second argument being picked up by the wrapper function was indeed NOT the same as what I had passed in from Perl.

Again, this program worked perfectly fine with SWIG 1.3.27 -- but my 100% identical source code dies horribly if I build it with any later SWIG -- so it would seem that something broke with the changes to make 1.3.28.

Does this sound familiar to anyone?  What sorts of additional detail, and/or testing, would be required from me in order to narrow this down sufficiently?


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101645&aid=1684887&group_id=1645