This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [bfin, patch] Adding bfin*-rtems*
On Tue, 2007-01-09 at 23:08 +0100, Bernd Schmidt wrote:
> Ian Lance Taylor wrote:
> > -N is a linker option. If you use -N on the gcc command line, gcc
> > will pass it to the linker.
> >
> > (I don't know why bfin uses it.)
>
> Historical baggage, most likely.
OK, so would you be opposed to removing it? We already are using the
patch below for bfin-rtems* for quite a while, but I am sure if we are
correct in doing so (We are using binutils-2.17)
Do you have any idea in which version of binutils the bug the comments
in t-bfin* refer to has been addressed?
> Parts of the port are quite old and
> copied from other ports.
I suspected as much but wasn't sure ;)
Ralf
Index: config/bfin/t-bfin
===================================================================
--- config/bfin/t-bfin (revision 120631)
+++ config/bfin/t-bfin (working copy)
@@ -15,10 +15,6 @@
echo '#define FLOAT' > fp-bit.c
cat $(srcdir)/config/fp-bit.c >> fp-bit.c
-# This shouldn't be needed here. I added it to the specs file for now, until
-# it is fixed in binutils (if it is necessary).
-GCC_CFLAGS += -N
-
# Assemble startup files.
$(T)crti.o: $(srcdir)/config/bfin/crti.s $(GCC_PASSES)
$(GCC_FOR_TARGET) $(GCC_CFLAGS) $(MULTILIB_CFLAGS) $(INCLUDES) \
Index: config/bfin/t-bfin-elf
===================================================================
--- config/bfin/t-bfin-elf (revision 120631)
+++ config/bfin/t-bfin-elf (working copy)
@@ -15,10 +15,6 @@
echo '#define FLOAT' > fp-bit.c
cat $(srcdir)/config/fp-bit.c >> fp-bit.c
-# This shouldn't be needed here. I added it to the specs file for now, until
-# it is fixed in binutils (if it is necessary).
-GCC_CFLAGS += -N
-
CRTSTUFF_T_CFLAGS = -fpic
TARGET_LIBGCC2_CFLAGS = -fpic