On 02/ 2/11 04:50 PM, Volker Braun wrote:
On Wednesday, February 2, 2011 3:55:27 PM UTC, Dr. David Kirkby wrote:
There is never any point in setting SAGE64=no. [...] ONLY if it is set
to "yes" will anything different happen. That is usually adding the
compiler flag "-m64", though in some cases it's a bit more complex.
Dave, can you elaborate on what SAGE64 is good for besides adding -m64?
Adding the compiler flag could easily be done in the gcc wrapper:
http://trac.sagemath.org/sage_trac/ticket/10572
Sorry if this is slightly OT.
In many cases the compiler flag -m64 gets added, though more recently the
contents of the variable $CFLAG64 get added, where $CFLAG64 defaults to -m64.
That will allow one to use a compiler where the option for 64 bit builds is not
-m64. (E.g. HP's and IBM's compilers.)
But as John has posted, there are cases where SAGE64 does other things too. For
some packages, one needs to pass the flag to a configure script, so
./configure CFLAGS="-m64" --prefix=$SAGE_LOCAL
or something like that is needed - in fact, I'm told by the autoconf developers
that is the preferred way.
ATLAS also has its own system to indicate 64-bit builds.
One package (I forget which) has a huge number of "targets", so one does
make foobar
The appropiate target would be set for a 64-bit build. It might be something
like "sun-64-bit-gcc"
So there's no way this can be replaced by a compiler flag.
I happen to think #10572 is a particularly bad idea, as I have noted on the
ticket.
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
Dave
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
