This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: ___divti3 and ___umodti3 missing on Darwin
- From: howarth at bromo dot msbb dot uc dot edu (Jack Howarth)
- To: geoffk at apple dot com, howarth at bromo dot msbb dot uc dot edu
- Cc: dje at watson dot ibm dot com, echristo at apple dot com, gcc at gcc dot gnu dot org, mrs at apple dot com
- Date: Mon, 7 Aug 2006 10:50:28 -0400 (EDT)
- Subject: Re: ___divti3 and ___umodti3 missing on Darwin
Geoff,
Looking at this from Xcode 2.3, which doesn't have a dylib for ppc64,
"nm libgcc.a | grep ti" doesn't show anything whereas on x86_64 FC5 Linux
this shows all the TImode symbols. This result isn't surprising since
the gcc 4.0.1 in Xcode 2.3 can't resolve the symbols as well when the
static libgcc is used. I guess this implies that 64-bit code generation
hasn't really been tested that hard on Darwin.
Jack
ps As I mentioned before, it would behove Apple to change the regress
tester to use something like...
make -k check RUNTESTFLAGS='--target_board "unix{,-m64}"'
...so that both 32 and 64 bit code generation is tested. After all
you are already doing multilib builds so why not?