[Top][All Lists]

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

Re: libtool support for Intel's icc

From: Roberto Bagnara
Subject: Re: libtool support for Intel's icc
Date: Sat, 23 Nov 2002 00:35:58 +0100
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1) Gecko/20020827

Boehne, Robert wrote:
We have a patch, but no copyright assignment as of yet.
I have been assured that the paperwork has been sent to the FSF, so
it should just be a matter of time.

Thanks Robert.  Still on the subject of supporting as many compilers
as possible: why libtool treats compilers it does not know as if they
were gcc?  Let me explain: I configured a directory with CXX=como
(this is Comeau C/C++ compiler and --disable-shared.
So I don't want shared library in this build and I am not using GCC.
Nonetheless, libtool tries to link with commands like

como -tlocal -I/usr/local/include -o intersection1 intersection1.o \
  ehandlers.o print.o  ../src/.libs/libppl.a -L/opt/como/local/lib/ \
  /usr/local/lib/ -L/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.2 \
  -L/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.2/../../.. \
-L/usr/local/distrib/objdir/i686-pc-linux-gnu/libstdc++-v3/src \
  -L/usr/local/distrib/objdir/gcc -lm -lc -lgcc_s /usr/local/lib/
  --rpath /usr/local/lib --rpath /usr/local/lib

which refer to GCC specific options and libraries that, of course, cannot
work.  Notice that the "right" command to link would simply be

como -o intersection1 intersection1.o ehandlers.o print.o -L../src/.libs
  -lppl -L/opt/como/local/lib/ -lgmpxx -lgmp

Would it be sensible/possible to modify libtool so that, if the compiler
is not known and static libraries have to be built, a "standard" link
command is issued?  By "standard link command" I mean one that is
composed of the compiler name, one -o option, a number of .o and .a
files and the usual -Ldir and -llib options.  I guess this behavior
is likely to work with several compilers (it would work today with
Comeau and Intel compilers).


Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy

reply via email to

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