patches/glibc/ports-2.10.1/220-section-comments.patch
author "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Sun Sep 12 23:38:18 2010 +0200 (2010-09-12)
changeset 2120 48de021b3a9e
permissions -rw-r--r--
arch/mips: add experimental mips64 samples

Both toolchains were tested to successfully:
- _build_ busybox, giving respectively an n32 and an n64 binary
although it has not been run-tested.
- build linux-2.6.36-rc3 for Loongson, giving an n64 vmlinux
(which is expected, due to the way the kernel is built),
and the kernel properly boots!

Credits are due to Julien MOUTHINO (julm on freenode#uClibc) for
testing that the kernel boots on his hardware! Cheers! ;-)

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
     1 http://sources.redhat.com/ml/binutils/2004-04/msg00665.html
     2 
     3 fixes building on some architectures (like m68k/arm/cris/etc...) because
     4 it does the right thing
     5 
     6 diff -durN glibc-2.10.1.orig/include/libc-symbols.h glibc-2.10.1/include/libc-symbols.h
     7 --- glibc-2.10.1.orig/include/libc-symbols.h	2009-03-14 00:51:46.000000000 +0100
     8 +++ glibc-2.10.1/include/libc-symbols.h	2009-11-13 00:50:07.000000000 +0100
     9 @@ -239,12 +239,12 @@
    10  #  define __make_section_unallocated(section_string)
    11  # endif
    12  
    13 -/* Tacking on "\n\t#" to the section name makes gcc put it's bogus
    14 +/* Tacking on "\n#APP\n\t#" to the section name makes gcc put it's bogus
    15     section attributes on what looks like a comment to the assembler.  */
    16  # ifdef HAVE_SECTION_QUOTES
    17 -#  define __sec_comment "\"\n\t#\""
    18 +#  define __sec_comment "\"\n#APP\n\t#\""
    19  # else
    20 -#  define __sec_comment "\n\t#"
    21 +#  define __sec_comment "\n#APP\n\t#"
    22  # endif
    23  # define link_warning(symbol, msg) \
    24    __make_section_unallocated (".gnu.warning." #symbol) \