[PATCH] Build initfini.s with -fno-unit-at-a-time for ppc/ppc64.
Dwayne Grant McConnell
dgm69@us.ibm.com
Thu Aug 4 21:28:00 GMT 2005
On Wed, 3 Aug 2005, Roland McGrath wrote:
> Does it work if you change those settings to use += instead? That way you
> don't have to copy the setting currently $(fno-unit-at-a-time) and keep
> track of it should we change it in future.
Yes. Here is a modified patch that uses += and removes duplicate flags and
changes the comments to be consistent.
--
Dwayne Grant McConnell <dgm69@us.ibm.com>
Lotus Notes: Dwayne McConnell/Austin/IBM@IBMUS
2005-08-03 Dwayne Grant McConnell <dgm69@us.ibm.com>
* sysdeps/powerpc/powerpc32/Makefile (CFLAGS-initfini.s): Use +=,
not =, to append. Remove duplicate flags.
* sysdeps/powerpc/powerpc64/Makefile (CFLAGS-initfini.s):
Likewise. Fix comments.
diff -urN libc.orig/sysdeps/powerpc/powerpc32/Makefile libc/sysdeps/powerpc/powerpc32/Makefile
--- libc.orig/sysdeps/powerpc/powerpc32/Makefile 2003-10-22 16:30:02.000000000 -0500
+++ libc/sysdeps/powerpc/powerpc32/Makefile 2005-08-04 15:21:11.553672369 -0500
@@ -21,7 +21,7 @@
ifneq ($(elf),no)
# The initfini generation code doesn't work in the presence of -fPIC, so
# we use -fpic instead which is much better.
-CFLAGS-initfini.s = -g0 -fpic -O1
+CFLAGS-initfini.s += -fpic -O1
# There is no benefit to using sdata for these objects, and the user
# of the library should be able to control what goes into sdata.
diff -urN libc.orig/sysdeps/powerpc/powerpc64/Makefile libc/sysdeps/powerpc/powerpc64/Makefile
--- libc.orig/sysdeps/powerpc/powerpc64/Makefile 2004-10-06 17:05:14.000000000 -0500
+++ libc/sysdeps/powerpc/powerpc64/Makefile 2005-08-04 15:19:54.560208219 -0500
@@ -12,9 +12,9 @@
ifeq ($(subdir),csu)
ifneq ($(elf),no)
-# The initfini generation code doesn't work in the presence of -g1 or
-# higher, so we use -g0.
-CFLAGS-initfini.s = -g0 -fpic -O1
+# The initfini generation code doesn't work in the presence of -fPIC, so
+# we use -fpic instead which is much better.
+CFLAGS-initfini.s += -fpic -O1
endif
endif
More information about the Libc-alpha
mailing list