Hi,
On Feb 11, 2005, at 1:23, Jeff Whitaker wrote:
Perhaps I'll make a gcc4 package a splitoff of gfortran - what do you
think?
GCC4 has several new features, including auto vectorization and tree-ssa
optimization. It sounds very attractive that fink has gcc4 package.
I have used some
Jeff Whitaker wrote:
I've created a gcc4 package the builds everything
(c,c++,f95,objc,java,ada). I could just add Provides and Conflicts
gfortran and be done with it? Is there any reason to split them up?
As long as you added a prefix or suffix to the gcc and g++ executables so
that they
Peter O'Gorman wrote:
Jeff Whitaker wrote:
I've created a gcc4 package the builds everything
(c,c++,f95,objc,java,ada). I could just add Provides and Conflicts
gfortran and be done with it? Is there any reason to split them up?
As long as you added a prefix or suffix to the gcc and g++
Jeff Whitaker wrote:
I think that there are other options hidden around somewhere, like
enabling the gnu objc runtime as well as the apple objc runtime,
should this be looked into?
I see
--enable-objc-gcenable the use of Boehm's garbage collector with
Peter O'Gorman wrote:
Looks like --enable-threads=posix builds a libobjc-gnu.dylib ... gee,
that's obvious, I spotted that right away (NOT!).
Peter
That's the default I guess, since I've already got that lib. I went
ahead and checked in the gcc4 package if you want to give it a try (best
to
On Feb 11, 2005, at 10:58 AM, Jeff Whitaker wrote:
I don't really know, I'm hoping others will hop in on this thread. I
know Alexander Strange has been building gccs from the apple ppc
branch, maybe a fink gcc4 package should come from there?
Hmm - I'd like to hear about that. I doubt that
On Fri, Feb 11, 2005 at 08:58:45AM -0700, Jeff Whitaker wrote:
Peter O'Gorman wrote:
Jeff Whitaker wrote:
I've created a gcc4 package the builds everything
(c,c++,f95,objc,java,ada). I could just add Provides and Conflicts
gfortran and be done with it? Is there any reason to split them
Daniel Macks wrote:
On Fri, Feb 11, 2005 at 08:58:45AM -0700, Jeff Whitaker wrote:
Can we get rid of the kernel's version-number in its installation
directory? If gcc builds OS-version-specific stuff, we have a package
portability problem...
dan
Dan: I haven't been able to figure out how to
Well, it took a little longer than I'd hoped, but I just sent this message
to gcc-patches: http://gcc.gnu.org/ml/gcc-patches/2005-02/msg00439.html
which solves the issue for me. The bug has been reopened and is at
http://gcc.gnu.org/PR18810.
Jeff, could you please apply the patch in that mail
Peter O'Gorman wrote:
Well, it took a little longer than I'd hoped, but I just sent this
message to gcc-patches:
http://gcc.gnu.org/ml/gcc-patches/2005-02/msg00439.html which solves
the issue for me. The bug has been reopened and is at
http://gcc.gnu.org/PR18810.
Jeff, could you please apply
Jeff Whitaker wrote:
Thanks Peter! That's quite a sleuthing job ..
If I were smarter I'd have found it sooner :)
I've updated g95 to use this patch, and am working on a new gfortran
which uses this and your two-level namespace patch. Perhaps I'll make a
gcc4 package a splitoff of gfortran -
11 matches
Mail list logo
