[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Octave-bug-tracker] [bug #51844] [control package] Plotting system resp
From: |
Dmitri A. Sergatskov |
Subject: |
[Octave-bug-tracker] [bug #51844] [control package] Plotting system responses doesn't work anymore |
Date: |
Sat, 26 Aug 2017 15:55:00 -0400 (EDT) |
User-agent: |
Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:55.0) Gecko/20100101 Firefox/55.0 |
Follow-up Comment #1, bug #51844 (project octave):
I think the problem is that "octave_idx_type" has changed.
In sl_td04ad.cc ( around line 154) one can see:
// error indicator
octave_idx_type info;
// SLICOT routine TD04AD
F77_XFCN (td04ad, TD04AD,
(rowcol,
m, p,
index,
dcoeff.fortran_vec (), lddcoe,
ucoeff.fortran_vec (), lduco1, lduco2,
nr,
a.fortran_vec (), lda,
b.fortran_vec (), ldb,
c.fortran_vec (), ldc,
d.fortran_vec (), ldd,
tol,
iwork,
dwork, ldwork,
info));
if (f77_exception_encountered)
error ("tf2ss: __sl_td04ad__: exception in SLICOT subroutine
TD04AD");
if (info != 0)
error ("tf2ss: __sl_td04ad__: TD04AD returned info = %d", info);
"info" in TD04AD is integer, and "info" in sl_td04ad.cc
is octave_idx_type is long int (?), so it contains some trailing garbage. The
quick (and, most likely, improper) fix is to
initialize info to 0L in sl_td04ad.cc.
Dmitri.
--
_______________________________________________________
Reply to this item at:
<http://savannah.gnu.org/bugs/?51844>
_______________________________________________
Message sent via/by Savannah
http://savannah.gnu.org/