help-octave
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

compiling with gcc-2.5.7


From: Mark Odegard
Subject: compiling with gcc-2.5.7
Date: Thu, 6 Jan 94 13:50:54 CST

After fixing the SCHUR class two argument problem I was able to compile
octave with gcc-2.5.7 and libg++2.5.3.  There are no longer the problems
with the "control reaches end of non-void function" warnings, so
Complex.h doesn't have to be edited to remove "~Complex".  However:

There are now lots of warnings like:
DiagMatrix.cc: In method `class RowVector DiagMatrix::row(char *) const':
DiagMatrix.cc:356: warning: this function may return with or without a value
        and:
Matrix-ext.cc: In method `int SCHUR::init(const class Matrix &, const char *)':
Matrix-ext.cc:473: warning: ANSI C++ prohibits conversion from `(double *, 
double *)' to `(...)'

These seem to have no effect.  The process does fail, however, in the
final load phase, with:
/files1/home/itlameo/octave/octave-0.79/src/g-builtins.cc:1639: builtin_givens
(t ree_constant const *, int, int): undefined reference to
`givens(tree_constant const *, int, int)'
collect2: ld returned 1 exit status.

I was able to complete the load by editing the Makefile in the /src
directory by adding "f-givens.o" to the "OBJECTS" variable, as in:
OBJECTS = arith-ops.o builtins.o error.o file-io.o g-builtins.o \
        ETC......
        dynamic-ld.o f-givens.o

The resultant "octave" runs but I haven't tried the "givens" function.
I looked at the "givens" source code but couldn't find anything which
was obviously wrong.

Anyway, good luck to all.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Mark Odegard                             | address@hidden
Unocal Corporation, Frontier Exploration | Telno: (713)287-7666
14141 Southwest Freeway                  | Fax:   (713)287-5403
Sugar Land, TX 77478, USA                | address@hidden
>>>>>>>>>>>>> "No man is truly free who has to work for a living" <<<<<<<<<<


reply via email to

[Prev in Thread] Current Thread [Next in Thread]