Mercurialcrosstool-ng / file revision
summary | shortlog | changelog | graph | tags | branches | files | changeset | file | latest | revisions | annotate | diff | raw | help
patches/glibc/ports-2.12.1/610-fpu-cw-mips.patch
author Esben Haabendal <esben.haabendal@prevas.dk>
Wed Sep 26 09:09:23 2012 +0200 (2012-09-26)
changeset 3065 603ccf80c48b
permissions -rw-r--r--
kernel/linux: change m68k/no-mmu linux arch tuple to use -uclinux

GCC requires m68k arch tuples to be *-*-uclinux-* to support Linux on
no-mmu m68k (ColdFire) cpus.

Blackfin arch tuple must be *-*-linux-uclibc for FD_PIC_ELF toolchains,
so we cannot just switch to uclinux for no-mmu Linux toolchains.

Signed-off-by: "Esben Haabendal" <esben@haabendal.dk>
Message-Id: <876271s1ee.fsf@arh128.prevas.dk>
PatchWork-Id: 186976 
     1 http://sourceware.org/ml/libc-alpha/2002-10/msg00392.html

     2 

     3 diff -durN glibc-2.12.1.orig/glibc-ports-2.12.1/sysdeps/mips/fpu_control.h glibc-2.12.1/glibc-ports-2.12.1/sysdeps/mips/fpu_control.h

     4 --- glibc-2.12.1.orig/glibc-ports-2.12.1/sysdeps/mips/fpu_control.h	2009-05-16 10:36:20.000000000 +0200

     5 +++ glibc-2.12.1/glibc-ports-2.12.1/sysdeps/mips/fpu_control.h	2009-11-13 00:51:31.000000000 +0100

     6 @@ -86,7 +86,7 @@

     7  #define _FPU_RC_UP      0x2

     8  #define _FPU_RC_DOWN    0x3

     9  

    10 -#define _FPU_RESERVED 0xfe3c0000  /* Reserved bits in cw */

    11 +#define _FPU_RESERVED 0xfebc0000  /* Reserved bits in cw */

    12  

    13  

    14  /* The fdlibm code requires strict IEEE double precision arithmetic,

    15 diff -durN glibc-2.12.1.orig/ports/sysdeps/mips/fpu_control.h glibc-2.12.1/ports/sysdeps/mips/fpu_control.h
crosstool-ng