x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

11 files changed, 1415 insertions, 0 deletions

diff --git a/arch/m68k/lib/Makefile b/arch/m68k/lib/Makefile
new file mode 100644
index 000000000..15278a952
--- /dev/null
+++ b/arch/m68k/lib/Makefile
@@ -0,0 +1,16 @@
+# SPDX-License-Identifier: GPL-2.0
+
+#
+# Makefile for m68k-specific library files..
+#
+
+lib-y	:= memcpy.o memset.o memmove.o
+
+lib-$(CONFIG_MMU) += uaccess.o
+lib-$(CONFIG_CPU_HAS_NO_MULDIV64) += mulsi3.o divsi3.o udivsi3.o
+lib-$(CONFIG_CPU_HAS_NO_MULDIV64) += modsi3.o umodsi3.o
+
+ifndef CONFIG_GENERIC_CSUM
+lib-y	+= checksum.o
+endif
+
diff --git a/arch/m68k/lib/checksum.c b/arch/m68k/lib/checksum.c
new file mode 100644
index 000000000..5acb82184
--- /dev/null
+++ b/arch/m68k/lib/checksum.c
@@ -0,0 +1,369 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * INET		An implementation of the TCP/IP protocol suite for the LINUX
+ *		operating system.  INET is implemented using the  BSD Socket
+ *		interface as the means of communication with the user level.
+ *
+ *		IP/TCP/UDP checksumming routines
+ *
+ * Authors:	Jorge Cwik, <jorge@laser.satlink.net>
+ *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
+ *		Tom May, <ftom@netcom.com>
+ *		Andreas Schwab, <schwab@issan.informatik.uni-dortmund.de>
+ *		Lots of code moved from tcp.c and ip.c; see those files
+ *		for more names.
+ *
+ * 03/02/96	Jes Sorensen, Andreas Schwab, Roman Hodek:
+ *		Fixed some nasty bugs, causing some horrible crashes.
+ *		A: At some points, the sum (%0) was used as
+ *		length-counter instead of the length counter
+ *		(%1). Thanks to Roman Hodek for pointing this out.
+ *		B: GCC seems to mess up if one uses too many
+ *		data-registers to hold input values and one tries to
+ *		specify d0 and d1 as scratch registers. Letting gcc
+ *		choose these registers itself solves the problem.
+ *
+ * 1998/8/31	Andreas Schwab:
+ *		Zero out rest of buffer on exception in
+ *		csum_partial_copy_from_user.
+ */
+
+#include <linux/module.h>
+#include <net/checksum.h>
+
+/*
+ * computes a partial checksum, e.g. for TCP/UDP fragments
+ */
+
+__wsum csum_partial(const void *buff, int len, __wsum sum)
+{
+	unsigned long tmp1, tmp2;
+	  /*
+	   * Experiments with ethernet and slip connections show that buff
+	   * is aligned on either a 2-byte or 4-byte boundary.
+	   */
+	__asm__("movel %2,%3\n\t"
+		"btst #1,%3\n\t"	/* Check alignment */
+		"jeq 2f\n\t"
+		"subql #2,%1\n\t"	/* buff%4==2: treat first word */
+		"jgt 1f\n\t"
+		"addql #2,%1\n\t"	/* len was == 2, treat only rest */
+		"jra 4f\n"
+	     "1:\t"
+		"addw %2@+,%0\n\t"	/* add first word to sum */
+		"clrl %3\n\t"
+		"addxl %3,%0\n"		/* add X bit */
+	     "2:\t"
+		/* unrolled loop for the main part: do 8 longs at once */
+		"movel %1,%3\n\t"	/* save len in tmp1 */
+		"lsrl #5,%1\n\t"	/* len/32 */
+		"jeq 2f\n\t"		/* not enough... */
+		"subql #1,%1\n"
+	     "1:\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"dbra %1,1b\n\t"
+		"clrl %4\n\t"
+		"addxl %4,%0\n\t"	/* add X bit */
+		"clrw %1\n\t"
+		"subql #1,%1\n\t"
+		"jcc 1b\n"
+	     "2:\t"
+		"movel %3,%1\n\t"	/* restore len from tmp1 */
+		"andw #0x1c,%3\n\t"	/* number of rest longs */
+		"jeq 4f\n\t"
+		"lsrw #2,%3\n\t"
+		"subqw #1,%3\n"
+	     "3:\t"
+		/* loop for rest longs */
+		"movel %2@+,%4\n\t"
+		"addxl %4,%0\n\t"
+		"dbra %3,3b\n\t"
+		"clrl %4\n\t"
+		"addxl %4,%0\n"		/* add X bit */
+	     "4:\t"
+		/* now check for rest bytes that do not fit into longs */
+		"andw #3,%1\n\t"
+		"jeq 7f\n\t"
+		"clrl %4\n\t"		/* clear tmp2 for rest bytes */
+		"subqw #2,%1\n\t"
+		"jlt 5f\n\t"
+		"movew %2@+,%4\n\t"	/* have rest >= 2: get word */
+		"swap %4\n\t"		/* into bits 16..31 */
+		"tstw %1\n\t"		/* another byte? */
+		"jeq 6f\n"
+	     "5:\t"
+		"moveb %2@,%4\n\t"	/* have odd rest: get byte */
+		"lslw #8,%4\n\t"	/* into bits 8..15; 16..31 untouched */
+	     "6:\t"
+		"addl %4,%0\n\t"	/* now add rest long to sum */
+		"clrl %4\n\t"
+		"addxl %4,%0\n"		/* add X bit */
+	     "7:\t"
+		: "=d" (sum), "=d" (len), "=a" (buff),
+		  "=&d" (tmp1), "=&d" (tmp2)
+		: "0" (sum), "1" (len), "2" (buff)
+	    );
+	return(sum);
+}
+
+EXPORT_SYMBOL(csum_partial);
+
+
+/*
+ * copy from user space while checksumming, with exception handling.
+ */
+
+__wsum
+csum_and_copy_from_user(const void __user *src, void *dst, int len)
+{
+	/*
+	 * GCC doesn't like more than 10 operands for the asm
+	 * statements so we have to use tmp2 for the error
+	 * code.
+	 */
+	unsigned long tmp1, tmp2;
+	__wsum sum = ~0U;
+
+	__asm__("movel %2,%4\n\t"
+		"btst #1,%4\n\t"	/* Check alignment */
+		"jeq 2f\n\t"
+		"subql #2,%1\n\t"	/* buff%4==2: treat first word */
+		"jgt 1f\n\t"
+		"addql #2,%1\n\t"	/* len was == 2, treat only rest */
+		"jra 4f\n"
+	     "1:\n"
+	     "10:\t"
+		"movesw %2@+,%4\n\t"	/* add first word to sum */
+		"addw %4,%0\n\t"
+		"movew %4,%3@+\n\t"
+		"clrl %4\n\t"
+		"addxl %4,%0\n"		/* add X bit */
+	     "2:\t"
+		/* unrolled loop for the main part: do 8 longs at once */
+		"movel %1,%4\n\t"	/* save len in tmp1 */
+		"lsrl #5,%1\n\t"	/* len/32 */
+		"jeq 2f\n\t"		/* not enough... */
+		"subql #1,%1\n"
+	     "1:\n"
+	     "11:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "12:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "13:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "14:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "15:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "16:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "17:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+	     "18:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"dbra %1,1b\n\t"
+		"clrl %5\n\t"
+		"addxl %5,%0\n\t"	/* add X bit */
+		"clrw %1\n\t"
+		"subql #1,%1\n\t"
+		"jcc 1b\n"
+	     "2:\t"
+		"movel %4,%1\n\t"	/* restore len from tmp1 */
+		"andw #0x1c,%4\n\t"	/* number of rest longs */
+		"jeq 4f\n\t"
+		"lsrw #2,%4\n\t"
+		"subqw #1,%4\n"
+	     "3:\n"
+		/* loop for rest longs */
+	     "19:\t"
+		"movesl %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"dbra %4,3b\n\t"
+		"clrl %5\n\t"
+		"addxl %5,%0\n"		/* add X bit */
+	     "4:\t"
+		/* now check for rest bytes that do not fit into longs */
+		"andw #3,%1\n\t"
+		"jeq 7f\n\t"
+		"clrl %5\n\t"		/* clear tmp2 for rest bytes */
+		"subqw #2,%1\n\t"
+		"jlt 5f\n\t"
+	     "20:\t"
+		"movesw %2@+,%5\n\t"	/* have rest >= 2: get word */
+		"movew %5,%3@+\n\t"
+		"swap %5\n\t"		/* into bits 16..31 */
+		"tstw %1\n\t"		/* another byte? */
+		"jeq 6f\n"
+	     "5:\n"
+	     "21:\t"
+		"movesb %2@,%5\n\t"	/* have odd rest: get byte */
+		"moveb %5,%3@+\n\t"
+		"lslw #8,%5\n\t"	/* into bits 8..15; 16..31 untouched */
+	     "6:\t"
+		"addl %5,%0\n\t"	/* now add rest long to sum */
+		"clrl %5\n\t"
+		"addxl %5,%0\n\t"	/* add X bit */
+	     "7:\t"
+		".section .fixup,\"ax\"\n"
+		".even\n"
+		/* If any exception occurs, return 0 */
+	     "90:\t"
+		"clrl %0\n"
+		"jra 7b\n"
+		".previous\n"
+		".section __ex_table,\"a\"\n"
+		".long 10b,90b\n"
+		".long 11b,90b\n"
+		".long 12b,90b\n"
+		".long 13b,90b\n"
+		".long 14b,90b\n"
+		".long 15b,90b\n"
+		".long 16b,90b\n"
+		".long 17b,90b\n"
+		".long 18b,90b\n"
+		".long 19b,90b\n"
+		".long 20b,90b\n"
+		".long 21b,90b\n"
+		".previous"
+		: "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
+		  "=&d" (tmp1), "=d" (tmp2)
+		: "0" (sum), "1" (len), "2" (src), "3" (dst)
+	    );
+
+	return sum;
+}
+
+
+/*
+ * copy from kernel space while checksumming, otherwise like csum_partial
+ */
+
+__wsum
+csum_partial_copy_nocheck(const void *src, void *dst, int len)
+{
+	unsigned long tmp1, tmp2;
+	__wsum sum = 0;
+	__asm__("movel %2,%4\n\t"
+		"btst #1,%4\n\t"	/* Check alignment */
+		"jeq 2f\n\t"
+		"subql #2,%1\n\t"	/* buff%4==2: treat first word */
+		"jgt 1f\n\t"
+		"addql #2,%1\n\t"	/* len was == 2, treat only rest */
+		"jra 4f\n"
+	     "1:\t"
+		"movew %2@+,%4\n\t"	/* add first word to sum */
+		"addw %4,%0\n\t"
+		"movew %4,%3@+\n\t"
+		"clrl %4\n\t"
+		"addxl %4,%0\n"		/* add X bit */
+	     "2:\t"
+		/* unrolled loop for the main part: do 8 longs at once */
+		"movel %1,%4\n\t"	/* save len in tmp1 */
+		"lsrl #5,%1\n\t"	/* len/32 */
+		"jeq 2f\n\t"		/* not enough... */
+		"subql #1,%1\n"
+	     "1:\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"dbra %1,1b\n\t"
+		"clrl %5\n\t"
+		"addxl %5,%0\n\t"	/* add X bit */
+		"clrw %1\n\t"
+		"subql #1,%1\n\t"
+		"jcc 1b\n"
+	     "2:\t"
+		"movel %4,%1\n\t"	/* restore len from tmp1 */
+		"andw #0x1c,%4\n\t"	/* number of rest longs */
+		"jeq 4f\n\t"
+		"lsrw #2,%4\n\t"
+		"subqw #1,%4\n"
+	     "3:\t"
+		/* loop for rest longs */
+		"movel %2@+,%5\n\t"
+		"addxl %5,%0\n\t"
+		"movel %5,%3@+\n\t"
+		"dbra %4,3b\n\t"
+		"clrl %5\n\t"
+		"addxl %5,%0\n"		/* add X bit */
+	     "4:\t"
+		/* now check for rest bytes that do not fit into longs */
+		"andw #3,%1\n\t"
+		"jeq 7f\n\t"
+		"clrl %5\n\t"		/* clear tmp2 for rest bytes */
+		"subqw #2,%1\n\t"
+		"jlt 5f\n\t"
+		"movew %2@+,%5\n\t"	/* have rest >= 2: get word */
+		"movew %5,%3@+\n\t"
+		"swap %5\n\t"		/* into bits 16..31 */
+		"tstw %1\n\t"		/* another byte? */
+		"jeq 6f\n"
+	     "5:\t"
+		"moveb %2@,%5\n\t"	/* have odd rest: get byte */
+		"moveb %5,%3@+\n\t"
+		"lslw #8,%5\n"		/* into bits 8..15; 16..31 untouched */
+	     "6:\t"
+		"addl %5,%0\n\t"	/* now add rest long to sum */
+		"clrl %5\n\t"
+		"addxl %5,%0\n"		/* add X bit */
+	     "7:\t"
+		: "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
+		  "=&d" (tmp1), "=&d" (tmp2)
+		: "0" (sum), "1" (len), "2" (src), "3" (dst)
+	    );
+    return(sum);
+}
+EXPORT_SYMBOL(csum_partial_copy_nocheck);
diff --git a/arch/m68k/lib/divsi3.S b/arch/m68k/lib/divsi3.S
new file mode 100644
index 000000000..62787b433
--- /dev/null
+++ b/arch/m68k/lib/divsi3.S
@@ -0,0 +1,123 @@
+/* libgcc1 routines for 68000 w/o floating-point hardware.
+   Copyright (C) 1994, 1996, 1997, 1998 Free Software Foundation, Inc.
+
+This file is part of GNU CC.
+
+GNU CC is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation; either version 2, or (at your option) any
+later version.
+
+In addition to the permissions in the GNU General Public License, the
+Free Software Foundation gives you unlimited permission to link the
+compiled version of this file with other programs, and to distribute
+those programs without any restriction coming from the use of this
+file.  (The General Public License restrictions do apply in other
+respects; for example, they cover modification of the file, and
+distribution when not linked into another program.)
+
+This file is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details. */
+
+/* As a special exception, if you link this library with files
+   compiled with GCC to produce an executable, this does not cause
+   the resulting executable to be covered by the GNU General Public License.
+   This exception does not however invalidate any other reasons why
+   the executable file might be covered by the GNU General Public License.  */
+
+/* Use this one for any 680x0; assumes no floating point hardware.
+   The trailing " '" appearing on some lines is for ANSI preprocessors.  Yuk.
+   Some of this code comes from MINIX, via the folks at ericsson.
+   D. V. Henkel-Wallace (gumby@cygnus.com) Fete Bastille, 1992
+*/
+
+#include <linux/export.h>
+
+/* These are predefined by new versions of GNU cpp.  */
+
+#ifndef __USER_LABEL_PREFIX__
+#define __USER_LABEL_PREFIX__ _
+#endif
+
+#ifndef __REGISTER_PREFIX__
+#define __REGISTER_PREFIX__
+#endif
+
+#ifndef __IMMEDIATE_PREFIX__
+#define __IMMEDIATE_PREFIX__ #
+#endif
+
+/* ANSI concatenation macros.  */
+
+#define CONCAT1(a, b) CONCAT2(a, b)
+#define CONCAT2(a, b) a ## b
+
+/* Use the right prefix for global labels.  */
+
+#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)
+
+/* Use the right prefix for registers.  */
+
+#define REG(x) CONCAT1 (__REGISTER_PREFIX__, x)
+
+/* Use the right prefix for immediate values.  */
+
+#define IMM(x) CONCAT1 (__IMMEDIATE_PREFIX__, x)
+
+#define d0 REG (d0)
+#define d1 REG (d1)
+#define d2 REG (d2)
+#define d3 REG (d3)
+#define d4 REG (d4)
+#define d5 REG (d5)
+#define d6 REG (d6)
+#define d7 REG (d7)
+#define a0 REG (a0)
+#define a1 REG (a1)
+#define a2 REG (a2)
+#define a3 REG (a3)
+#define a4 REG (a4)
+#define a5 REG (a5)
+#define a6 REG (a6)
+#define fp REG (fp)
+#define sp REG (sp)
+
+	.text
+	.proc
+	.globl	SYM (__divsi3)
+SYM (__divsi3):
+	movel	d2, sp@-
+
+	moveq	IMM (1), d2	/* sign of result stored in d2 (=1 or =-1) */
+	movel	sp@(12), d1	/* d1 = divisor */
+	jpl	L1
+	negl	d1
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	negb	d2		/* change sign because divisor <0  */
+#else
+	negl	d2		/* change sign because divisor <0  */
+#endif
+L1:	movel	sp@(8), d0	/* d0 = dividend */
+	jpl	L2
+	negl	d0
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	negb	d2
+#else
+	negl	d2
+#endif
+
+L2:	movel	d1, sp@-
+	movel	d0, sp@-
+	jbsr	SYM (__udivsi3)	/* divide abs(dividend) by abs(divisor) */
+	addql	IMM (8), sp
+
+	tstb	d2
+	jpl	L3
+	negl	d0
+
+L3:	movel	sp@+, d2
+	rts
+
+	EXPORT_SYMBOL(__divsi3)
diff --git a/arch/m68k/lib/memcpy.c b/arch/m68k/lib/memcpy.c
new file mode 100644
index 000000000..c1e2dfb20
--- /dev/null
+++ b/arch/m68k/lib/memcpy.c
@@ -0,0 +1,89 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/string.h>
+
+void *memcpy(void *to, const void *from, size_t n)
+{
+	void *xto = to;
+	size_t temp;
+
+	if (!n)
+		return xto;
+	if ((long)to & 1) {
+		char *cto = to;
+		const char *cfrom = from;
+		*cto++ = *cfrom++;
+		to = cto;
+		from = cfrom;
+		n--;
+	}
+#if defined(CONFIG_M68000)
+	if ((long)from & 1) {
+		char *cto = to;
+		const char *cfrom = from;
+		for (; n; n--)
+			*cto++ = *cfrom++;
+		return xto;
+	}
+#endif
+	if (n > 2 && (long)to & 2) {
+		short *sto = to;
+		const short *sfrom = from;
+		*sto++ = *sfrom++;
+		to = sto;
+		from = sfrom;
+		n -= 2;
+	}
+	temp = n >> 2;
+	if (temp) {
+		long *lto = to;
+		const long *lfrom = from;
+#if defined(CONFIG_M68000) || defined(CONFIG_COLDFIRE)
+		for (; temp; temp--)
+			*lto++ = *lfrom++;
+#else
+		size_t temp1;
+		asm volatile (
+			"	movel %2,%3\n"
+			"	andw  #7,%3\n"
+			"	lsrl  #3,%2\n"
+			"	negw  %3\n"
+			"	jmp   %%pc@(1f,%3:w:2)\n"
+			"4:	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"	movel %0@+,%1@+\n"
+			"1:	dbra  %2,4b\n"
+			"	clrw  %2\n"
+			"	subql #1,%2\n"
+			"	jpl   4b"
+			: "=a" (lfrom), "=a" (lto), "=d" (temp), "=&d" (temp1)
+			: "0" (lfrom), "1" (lto), "2" (temp));
+#endif
+		to = lto;
+		from = lfrom;
+	}
+	if (n & 2) {
+		short *sto = to;
+		const short *sfrom = from;
+		*sto++ = *sfrom++;
+		to = sto;
+		from = sfrom;
+	}
+	if (n & 1) {
+		char *cto = to;
+		const char *cfrom = from;
+		*cto = *cfrom;
+	}
+	return xto;
+}
+EXPORT_SYMBOL(memcpy);
diff --git a/arch/m68k/lib/memmove.c b/arch/m68k/lib/memmove.c
new file mode 100644
index 000000000..e33f00b02
--- /dev/null
+++ b/arch/m68k/lib/memmove.c
@@ -0,0 +1,121 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/string.h>
+
+void *memmove(void *dest, const void *src, size_t n)
+{
+	void *xdest = dest;
+	size_t temp;
+
+	if (!n)
+		return xdest;
+
+	if (dest < src) {
+		if ((long)dest & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			*cdest++ = *csrc++;
+			dest = cdest;
+			src = csrc;
+			n--;
+		}
+#if defined(CONFIG_M68000)
+		if ((long)src & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			for (; n; n--)
+				*cdest++ = *csrc++;
+			return xdest;
+		}
+#endif
+		if (n > 2 && (long)dest & 2) {
+			short *sdest = dest;
+			const short *ssrc = src;
+			*sdest++ = *ssrc++;
+			dest = sdest;
+			src = ssrc;
+			n -= 2;
+		}
+		temp = n >> 2;
+		if (temp) {
+			long *ldest = dest;
+			const long *lsrc = src;
+			temp--;
+			do
+				*ldest++ = *lsrc++;
+			while (temp--);
+			dest = ldest;
+			src = lsrc;
+		}
+		if (n & 2) {
+			short *sdest = dest;
+			const short *ssrc = src;
+			*sdest++ = *ssrc++;
+			dest = sdest;
+			src = ssrc;
+		}
+		if (n & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			*cdest = *csrc;
+		}
+	} else {
+		dest = (char *)dest + n;
+		src = (const char *)src + n;
+		if ((long)dest & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			*--cdest = *--csrc;
+			dest = cdest;
+			src = csrc;
+			n--;
+		}
+#if defined(CONFIG_M68000)
+		if ((long)src & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			for (; n; n--)
+				*--cdest = *--csrc;
+			return xdest;
+		}
+#endif
+		if (n > 2 && (long)dest & 2) {
+			short *sdest = dest;
+			const short *ssrc = src;
+			*--sdest = *--ssrc;
+			dest = sdest;
+			src = ssrc;
+			n -= 2;
+		}
+		temp = n >> 2;
+		if (temp) {
+			long *ldest = dest;
+			const long *lsrc = src;
+			temp--;
+			do
+				*--ldest = *--lsrc;
+			while (temp--);
+			dest = ldest;
+			src = lsrc;
+		}
+		if (n & 2) {
+			short *sdest = dest;
+			const short *ssrc = src;
+			*--sdest = *--ssrc;
+			dest = sdest;
+			src = ssrc;
+		}
+		if (n & 1) {
+			char *cdest = dest;
+			const char *csrc = src;
+			*--cdest = *--csrc;
+		}
+	}
+	return xdest;
+}
+EXPORT_SYMBOL(memmove);
diff --git a/arch/m68k/lib/memset.c b/arch/m68k/lib/memset.c
new file mode 100644
index 000000000..8a7639f0a
--- /dev/null
+++ b/arch/m68k/lib/memset.c
@@ -0,0 +1,74 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/string.h>
+
+void *memset(void *s, int c, size_t count)
+{
+	void *xs = s;
+	size_t temp;
+
+	if (!count)
+		return xs;
+	c &= 0xff;
+	c |= c << 8;
+	c |= c << 16;
+	if ((long)s & 1) {
+		char *cs = s;
+		*cs++ = c;
+		s = cs;
+		count--;
+	}
+	if (count > 2 && (long)s & 2) {
+		short *ss = s;
+		*ss++ = c;
+		s = ss;
+		count -= 2;
+	}
+	temp = count >> 2;
+	if (temp) {
+		long *ls = s;
+#if defined(CONFIG_M68000) || defined(CONFIG_COLDFIRE)
+		for (; temp; temp--)
+			*ls++ = c;
+#else
+		size_t temp1;
+		asm volatile (
+			"	movel %1,%2\n"
+			"	andw  #7,%2\n"
+			"	lsrl  #3,%1\n"
+			"	negw  %2\n"
+			"	jmp   %%pc@(2f,%2:w:2)\n"
+			"1:	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"	movel %3,%0@+\n"
+			"2:	dbra  %1,1b\n"
+			"	clrw  %1\n"
+			"	subql #1,%1\n"
+			"	jpl   1b"
+			: "=a" (ls), "=d" (temp), "=&d" (temp1)
+			: "d" (c), "0" (ls), "1" (temp));
+#endif
+		s = ls;
+	}
+	if (count & 2) {
+		short *ss = s;
+		*ss++ = c;
+		s = ss;
+	}
+	if (count & 1) {
+		char *cs = s;
+		*cs = c;
+	}
+	return xs;
+}
+EXPORT_SYMBOL(memset);
diff --git a/arch/m68k/lib/modsi3.S b/arch/m68k/lib/modsi3.S
new file mode 100644
index 000000000..1bcb742d0
--- /dev/null
+++ b/arch/m68k/lib/modsi3.S
@@ -0,0 +1,111 @@
+/* libgcc1 routines for 68000 w/o floating-point hardware.
+   Copyright (C) 1994, 1996, 1997, 1998 Free Software Foundation, Inc.
+
+This file is part of GNU CC.
+
+GNU CC is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation; either version 2, or (at your option) any
+later version.
+
+In addition to the permissions in the GNU General Public License, the
+Free Software Foundation gives you unlimited permission to link the
+compiled version of this file with other programs, and to distribute
+those programs without any restriction coming from the use of this
+file.  (The General Public License restrictions do apply in other
+respects; for example, they cover modification of the file, and
+distribution when not linked into another program.)
+
+This file is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details. */
+
+/* As a special exception, if you link this library with files
+   compiled with GCC to produce an executable, this does not cause
+   the resulting executable to be covered by the GNU General Public License.
+   This exception does not however invalidate any other reasons why
+   the executable file might be covered by the GNU General Public License.  */
+
+/* Use this one for any 680x0; assumes no floating point hardware.
+   The trailing " '" appearing on some lines is for ANSI preprocessors.  Yuk.
+   Some of this code comes from MINIX, via the folks at ericsson.
+   D. V. Henkel-Wallace (gumby@cygnus.com) Fete Bastille, 1992
+*/
+
+#include <linux/export.h>
+
+/* These are predefined by new versions of GNU cpp.  */
+
+#ifndef __USER_LABEL_PREFIX__
+#define __USER_LABEL_PREFIX__ _
+#endif
+
+#ifndef __REGISTER_PREFIX__
+#define __REGISTER_PREFIX__
+#endif
+
+#ifndef __IMMEDIATE_PREFIX__
+#define __IMMEDIATE_PREFIX__ #
+#endif
+
+/* ANSI concatenation macros.  */
+
+#define CONCAT1(a, b) CONCAT2(a, b)
+#define CONCAT2(a, b) a ## b
+
+/* Use the right prefix for global labels.  */
+
+#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)
+
+/* Use the right prefix for registers.  */
+
+#define REG(x) CONCAT1 (__REGISTER_PREFIX__, x)
+
+/* Use the right prefix for immediate values.  */
+
+#define IMM(x) CONCAT1 (__IMMEDIATE_PREFIX__, x)
+
+#define d0 REG (d0)
+#define d1 REG (d1)
+#define d2 REG (d2)
+#define d3 REG (d3)
+#define d4 REG (d4)
+#define d5 REG (d5)
+#define d6 REG (d6)
+#define d7 REG (d7)
+#define a0 REG (a0)
+#define a1 REG (a1)
+#define a2 REG (a2)
+#define a3 REG (a3)
+#define a4 REG (a4)
+#define a5 REG (a5)
+#define a6 REG (a6)
+#define fp REG (fp)
+#define sp REG (sp)
+
+	.text
+	.proc
+	.globl	SYM (__modsi3)
+SYM (__modsi3):
+	movel	sp@(8), d1	/* d1 = divisor */
+	movel	sp@(4), d0	/* d0 = dividend */
+	movel	d1, sp@-
+	movel	d0, sp@-
+	jbsr	SYM (__divsi3)
+	addql	IMM (8), sp
+	movel	sp@(8), d1	/* d1 = divisor */
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	movel	d1, sp@-
+	movel	d0, sp@-
+	jbsr	SYM (__mulsi3)	/* d0 = (a/b)*b */
+	addql	IMM (8), sp
+#else
+	mulsl	d1,d0
+#endif
+	movel	sp@(4), d1	/* d1 = dividend */
+	subl	d0, d1		/* d1 = a - (a/b)*b */
+	movel	d1, d0
+	rts
+
+	EXPORT_SYMBOL(__modsi3)
diff --git a/arch/m68k/lib/mulsi3.S b/arch/m68k/lib/mulsi3.S
new file mode 100644
index 000000000..c28532482
--- /dev/null
+++ b/arch/m68k/lib/mulsi3.S
@@ -0,0 +1,105 @@
+/* libgcc1 routines for 68000 w/o floating-point hardware.
+   Copyright (C) 1994, 1996, 1997, 1998 Free Software Foundation, Inc.
+
+This file is part of GNU CC.
+
+GNU CC is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation; either version 2, or (at your option) any
+later version.
+
+In addition to the permissions in the GNU General Public License, the
+Free Software Foundation gives you unlimited permission to link the
+compiled version of this file with other programs, and to distribute
+those programs without any restriction coming from the use of this
+file.  (The General Public License restrictions do apply in other
+respects; for example, they cover modification of the file, and
+distribution when not linked into another program.)
+
+This file is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details. */
+
+/* As a special exception, if you link this library with files
+   compiled with GCC to produce an executable, this does not cause
+   the resulting executable to be covered by the GNU General Public License.
+   This exception does not however invalidate any other reasons why
+   the executable file might be covered by the GNU General Public License.  */
+
+/* Use this one for any 680x0; assumes no floating point hardware.
+   The trailing " '" appearing on some lines is for ANSI preprocessors.  Yuk.
+   Some of this code comes from MINIX, via the folks at ericsson.
+   D. V. Henkel-Wallace (gumby@cygnus.com) Fete Bastille, 1992
+*/
+#include <linux/export.h>
+/* These are predefined by new versions of GNU cpp.  */
+
+#ifndef __USER_LABEL_PREFIX__
+#define __USER_LABEL_PREFIX__ _
+#endif
+
+#ifndef __REGISTER_PREFIX__
+#define __REGISTER_PREFIX__
+#endif
+
+#ifndef __IMMEDIATE_PREFIX__
+#define __IMMEDIATE_PREFIX__ #
+#endif
+
+/* ANSI concatenation macros.  */
+
+#define CONCAT1(a, b) CONCAT2(a, b)
+#define CONCAT2(a, b) a ## b
+
+/* Use the right prefix for global labels.  */
+
+#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)
+
+/* Use the right prefix for registers.  */
+
+#define REG(x) CONCAT1 (__REGISTER_PREFIX__, x)
+
+/* Use the right prefix for immediate values.  */
+
+#define IMM(x) CONCAT1 (__IMMEDIATE_PREFIX__, x)
+
+#define d0 REG (d0)
+#define d1 REG (d1)
+#define d2 REG (d2)
+#define d3 REG (d3)
+#define d4 REG (d4)
+#define d5 REG (d5)
+#define d6 REG (d6)
+#define d7 REG (d7)
+#define a0 REG (a0)
+#define a1 REG (a1)
+#define a2 REG (a2)
+#define a3 REG (a3)
+#define a4 REG (a4)
+#define a5 REG (a5)
+#define a6 REG (a6)
+#define fp REG (fp)
+#define sp REG (sp)
+
+	.text
+	.proc
+	.globl	SYM (__mulsi3)
+SYM (__mulsi3):
+	movew	sp@(4), d0	/* x0 -> d0 */
+	muluw	sp@(10), d0	/* x0*y1 */
+	movew	sp@(6), d1	/* x1 -> d1 */
+	muluw	sp@(8), d1	/* x1*y0 */
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	addw	d1, d0
+#else
+	addl	d1, d0
+#endif
+	swap	d0
+	clrw	d0
+	movew	sp@(6), d1	/* x1 -> d1 */
+	muluw	sp@(10), d1	/* x1*y1 */
+	addl	d1, d0
+
+	rts
+	EXPORT_SYMBOL(__mulsi3)
diff --git a/arch/m68k/lib/uaccess.c b/arch/m68k/lib/uaccess.c
new file mode 100644
index 000000000..7646e461a
--- /dev/null
+++ b/arch/m68k/lib/uaccess.c
@@ -0,0 +1,142 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/uaccess.h>
+
+unsigned long __generic_copy_from_user(void *to, const void __user *from,
+				       unsigned long n)
+{
+	unsigned long tmp, res;
+
+	asm volatile ("\n"
+		"	tst.l	%0\n"
+		"	jeq	2f\n"
+		"1:	"MOVES".l	(%1)+,%3\n"
+		"	move.l	%3,(%2)+\n"
+		"	subq.l	#1,%0\n"
+		"	jne	1b\n"
+		"2:	btst	#1,%5\n"
+		"	jeq	4f\n"
+		"3:	"MOVES".w	(%1)+,%3\n"
+		"	move.w	%3,(%2)+\n"
+		"4:	btst	#0,%5\n"
+		"	jeq	6f\n"
+		"5:	"MOVES".b	(%1)+,%3\n"
+		"	move.b  %3,(%2)+\n"
+		"6:\n"
+		"	.section .fixup,\"ax\"\n"
+		"	.even\n"
+		"10:	lsl.l	#2,%0\n"
+		"	btst	#1,%5\n"
+		"	jeq	8f\n"
+		"30:	addq.l	#2,%0\n"
+		"8:	btst	#0,%5\n"
+		"	jeq	6b\n"
+		"50:	addq.l	#1,%0\n"
+		"	jra	6b\n"
+		"	.previous\n"
+		"\n"
+		"	.section __ex_table,\"a\"\n"
+		"	.align	4\n"
+		"	.long	1b,10b\n"
+		"	.long	3b,30b\n"
+		"	.long	5b,50b\n"
+		"	.previous"
+		: "=d" (res), "+a" (from), "+a" (to), "=&d" (tmp)
+		: "0" (n / 4), "d" (n & 3));
+
+	return res;
+}
+EXPORT_SYMBOL(__generic_copy_from_user);
+
+unsigned long __generic_copy_to_user(void __user *to, const void *from,
+				     unsigned long n)
+{
+	unsigned long tmp, res;
+
+	asm volatile ("\n"
+		"	tst.l	%0\n"
+		"	jeq	4f\n"
+		"1:	move.l	(%1)+,%3\n"
+		"2:	"MOVES".l	%3,(%2)+\n"
+		"3:	subq.l	#1,%0\n"
+		"	jne	1b\n"
+		"4:	btst	#1,%5\n"
+		"	jeq	6f\n"
+		"	move.w	(%1)+,%3\n"
+		"5:	"MOVES".w	%3,(%2)+\n"
+		"6:	btst	#0,%5\n"
+		"	jeq	8f\n"
+		"	move.b	(%1)+,%3\n"
+		"7:	"MOVES".b  %3,(%2)+\n"
+		"8:\n"
+		"	.section .fixup,\"ax\"\n"
+		"	.even\n"
+		"20:	lsl.l	#2,%0\n"
+		"50:	add.l	%5,%0\n"
+		"	jra	8b\n"
+		"	.previous\n"
+		"\n"
+		"	.section __ex_table,\"a\"\n"
+		"	.align	4\n"
+		"	.long	2b,20b\n"
+		"	.long	3b,20b\n"
+		"	.long	5b,50b\n"
+		"	.long	6b,50b\n"
+		"	.long	7b,50b\n"
+		"	.long	8b,50b\n"
+		"	.previous"
+		: "=d" (res), "+a" (from), "+a" (to), "=&d" (tmp)
+		: "0" (n / 4), "d" (n & 3));
+
+	return res;
+}
+EXPORT_SYMBOL(__generic_copy_to_user);
+
+/*
+ * Zero Userspace
+ */
+
+unsigned long __clear_user(void __user *to, unsigned long n)
+{
+	unsigned long res;
+
+	asm volatile ("\n"
+		"	tst.l	%0\n"
+		"	jeq	3f\n"
+		"1:	"MOVES".l	%2,(%1)+\n"
+		"2:	subq.l	#1,%0\n"
+		"	jne	1b\n"
+		"3:	btst	#1,%4\n"
+		"	jeq	5f\n"
+		"4:	"MOVES".w	%2,(%1)+\n"
+		"5:	btst	#0,%4\n"
+		"	jeq	7f\n"
+		"6:	"MOVES".b	%2,(%1)\n"
+		"7:\n"
+		"	.section .fixup,\"ax\"\n"
+		"	.even\n"
+		"10:	lsl.l	#2,%0\n"
+		"40:	add.l	%4,%0\n"
+		"	jra	7b\n"
+		"	.previous\n"
+		"\n"
+		"	.section __ex_table,\"a\"\n"
+		"	.align	4\n"
+		"	.long	1b,10b\n"
+		"	.long	2b,10b\n"
+		"	.long	4b,40b\n"
+		"	.long	5b,40b\n"
+		"	.long	6b,40b\n"
+		"	.long	7b,40b\n"
+		"	.previous"
+		: "=d" (res), "+a" (to)
+		: "d" (0), "0" (n / 4), "d" (n & 3));
+
+    return res;
+}
+EXPORT_SYMBOL(__clear_user);
diff --git a/arch/m68k/lib/udivsi3.S b/arch/m68k/lib/udivsi3.S
new file mode 100644
index 000000000..39ad70596
--- /dev/null
+++ b/arch/m68k/lib/udivsi3.S
@@ -0,0 +1,157 @@
+/* libgcc1 routines for 68000 w/o floating-point hardware.
+   Copyright (C) 1994, 1996, 1997, 1998 Free Software Foundation, Inc.
+
+This file is part of GNU CC.
+
+GNU CC is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation; either version 2, or (at your option) any
+later version.
+
+In addition to the permissions in the GNU General Public License, the
+Free Software Foundation gives you unlimited permission to link the
+compiled version of this file with other programs, and to distribute
+those programs without any restriction coming from the use of this
+file.  (The General Public License restrictions do apply in other
+respects; for example, they cover modification of the file, and
+distribution when not linked into another program.)
+
+This file is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details. */
+
+/* As a special exception, if you link this library with files
+   compiled with GCC to produce an executable, this does not cause
+   the resulting executable to be covered by the GNU General Public License.
+   This exception does not however invalidate any other reasons why
+   the executable file might be covered by the GNU General Public License.  */
+
+/* Use this one for any 680x0; assumes no floating point hardware.
+   The trailing " '" appearing on some lines is for ANSI preprocessors.  Yuk.
+   Some of this code comes from MINIX, via the folks at ericsson.
+   D. V. Henkel-Wallace (gumby@cygnus.com) Fete Bastille, 1992
+*/
+#include <linux/export.h>
+/* These are predefined by new versions of GNU cpp.  */
+
+#ifndef __USER_LABEL_PREFIX__
+#define __USER_LABEL_PREFIX__ _
+#endif
+
+#ifndef __REGISTER_PREFIX__
+#define __REGISTER_PREFIX__
+#endif
+
+#ifndef __IMMEDIATE_PREFIX__
+#define __IMMEDIATE_PREFIX__ #
+#endif
+
+/* ANSI concatenation macros.  */
+
+#define CONCAT1(a, b) CONCAT2(a, b)
+#define CONCAT2(a, b) a ## b
+
+/* Use the right prefix for global labels.  */
+
+#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)
+
+/* Use the right prefix for registers.  */
+
+#define REG(x) CONCAT1 (__REGISTER_PREFIX__, x)
+
+/* Use the right prefix for immediate values.  */
+
+#define IMM(x) CONCAT1 (__IMMEDIATE_PREFIX__, x)
+
+#define d0 REG (d0)
+#define d1 REG (d1)
+#define d2 REG (d2)
+#define d3 REG (d3)
+#define d4 REG (d4)
+#define d5 REG (d5)
+#define d6 REG (d6)
+#define d7 REG (d7)
+#define a0 REG (a0)
+#define a1 REG (a1)
+#define a2 REG (a2)
+#define a3 REG (a3)
+#define a4 REG (a4)
+#define a5 REG (a5)
+#define a6 REG (a6)
+#define fp REG (fp)
+#define sp REG (sp)
+
+	.text
+	.proc
+	.globl	SYM (__udivsi3)
+SYM (__udivsi3):
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	movel	d2, sp@-
+	movel	sp@(12), d1	/* d1 = divisor */
+	movel	sp@(8), d0	/* d0 = dividend */
+
+	cmpl	IMM (0x10000), d1 /* divisor >= 2 ^ 16 ?   */
+	jcc	L3		/* then try next algorithm */
+	movel	d0, d2
+	clrw	d2
+	swap	d2
+	divu	d1, d2          /* high quotient in lower word */
+	movew	d2, d0		/* save high quotient */
+	swap	d0
+	movew	sp@(10), d2	/* get low dividend + high rest */
+	divu	d1, d2		/* low quotient */
+	movew	d2, d0
+	jra	L6
+
+L3:	movel	d1, d2		/* use d2 as divisor backup */
+L4:	lsrl	IMM (1), d1	/* shift divisor */
+	lsrl	IMM (1), d0	/* shift dividend */
+	cmpl	IMM (0x10000), d1 /* still divisor >= 2 ^ 16 ?  */
+	jcc	L4
+	divu	d1, d0		/* now we have 16 bit divisor */
+	andl	IMM (0xffff), d0 /* mask out divisor, ignore remainder */
+
+/* Multiply the 16 bit tentative quotient with the 32 bit divisor.  Because of
+   the operand ranges, this might give a 33 bit product.  If this product is
+   greater than the dividend, the tentative quotient was too large. */
+	movel	d2, d1
+	mulu	d0, d1		/* low part, 32 bits */
+	swap	d2
+	mulu	d0, d2		/* high part, at most 17 bits */
+	swap	d2		/* align high part with low part */
+	tstw	d2		/* high part 17 bits? */
+	jne	L5		/* if 17 bits, quotient was too large */
+	addl	d2, d1		/* add parts */
+	jcs	L5		/* if sum is 33 bits, quotient was too large */
+	cmpl	sp@(8), d1	/* compare the sum with the dividend */
+	jls	L6		/* if sum > dividend, quotient was too large */
+L5:	subql	IMM (1), d0	/* adjust quotient */
+
+L6:	movel	sp@+, d2
+	rts
+
+#else /* __mcf5200__ || __mcoldfire__ */
+
+/* Coldfire implementation of non-restoring division algorithm from
+   Hennessy & Patterson, Appendix A. */
+	link	a6,IMM (-12)
+	moveml	d2-d4,sp@
+	movel	a6@(8),d0
+	movel	a6@(12),d1
+	clrl	d2		| clear p
+	moveq	IMM (31),d4
+L1:	addl	d0,d0		| shift reg pair (p,a) one bit left
+	addxl	d2,d2
+	movl	d2,d3		| subtract b from p, store in tmp.
+	subl	d1,d3
+	jcs	L2		| if no carry,
+	bset	IMM (0),d0	| set the low order bit of a to 1,
+	movl	d3,d2		| and store tmp in p.
+L2:	subql	IMM (1),d4
+	jcc	L1
+	moveml	sp@,d2-d4	| restore data registers
+	unlk	a6		| and return
+	rts
+#endif /* __mcf5200__ || __mcoldfire__ */
+	EXPORT_SYMBOL(__udivsi3)
diff --git a/arch/m68k/lib/umodsi3.S b/arch/m68k/lib/umodsi3.S
new file mode 100644
index 000000000..6640eaa9e
--- /dev/null
+++ b/arch/m68k/lib/umodsi3.S
@@ -0,0 +1,108 @@
+/* libgcc1 routines for 68000 w/o floating-point hardware.
+   Copyright (C) 1994, 1996, 1997, 1998 Free Software Foundation, Inc.
+
+This file is part of GNU CC.
+
+GNU CC is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation; either version 2, or (at your option) any
+later version.
+
+In addition to the permissions in the GNU General Public License, the
+Free Software Foundation gives you unlimited permission to link the
+compiled version of this file with other programs, and to distribute
+those programs without any restriction coming from the use of this
+file.  (The General Public License restrictions do apply in other
+respects; for example, they cover modification of the file, and
+distribution when not linked into another program.)
+
+This file is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details. */
+
+/* As a special exception, if you link this library with files
+   compiled with GCC to produce an executable, this does not cause
+   the resulting executable to be covered by the GNU General Public License.
+   This exception does not however invalidate any other reasons why
+   the executable file might be covered by the GNU General Public License.  */
+
+/* Use this one for any 680x0; assumes no floating point hardware.
+   The trailing " '" appearing on some lines is for ANSI preprocessors.  Yuk.
+   Some of this code comes from MINIX, via the folks at ericsson.
+   D. V. Henkel-Wallace (gumby@cygnus.com) Fete Bastille, 1992
+*/
+#include <linux/export.h>
+/* These are predefined by new versions of GNU cpp.  */
+
+#ifndef __USER_LABEL_PREFIX__
+#define __USER_LABEL_PREFIX__ _
+#endif
+
+#ifndef __REGISTER_PREFIX__
+#define __REGISTER_PREFIX__
+#endif
+
+#ifndef __IMMEDIATE_PREFIX__
+#define __IMMEDIATE_PREFIX__ #
+#endif
+
+/* ANSI concatenation macros.  */
+
+#define CONCAT1(a, b) CONCAT2(a, b)
+#define CONCAT2(a, b) a ## b
+
+/* Use the right prefix for global labels.  */
+
+#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)
+
+/* Use the right prefix for registers.  */
+
+#define REG(x) CONCAT1 (__REGISTER_PREFIX__, x)
+
+/* Use the right prefix for immediate values.  */
+
+#define IMM(x) CONCAT1 (__IMMEDIATE_PREFIX__, x)
+
+#define d0 REG (d0)
+#define d1 REG (d1)
+#define d2 REG (d2)
+#define d3 REG (d3)
+#define d4 REG (d4)
+#define d5 REG (d5)
+#define d6 REG (d6)
+#define d7 REG (d7)
+#define a0 REG (a0)
+#define a1 REG (a1)
+#define a2 REG (a2)
+#define a3 REG (a3)
+#define a4 REG (a4)
+#define a5 REG (a5)
+#define a6 REG (a6)
+#define fp REG (fp)
+#define sp REG (sp)
+
+	.text
+	.proc
+	.globl	SYM (__umodsi3)
+SYM (__umodsi3):
+	movel	sp@(8), d1	/* d1 = divisor */
+	movel	sp@(4), d0	/* d0 = dividend */
+	movel	d1, sp@-
+	movel	d0, sp@-
+	jbsr	SYM (__udivsi3)
+	addql	IMM (8), sp
+	movel	sp@(8), d1	/* d1 = divisor */
+#if !(defined(__mcf5200__) || defined(__mcoldfire__))
+	movel	d1, sp@-
+	movel	d0, sp@-
+	jbsr	SYM (__mulsi3)	/* d0 = (a/b)*b */
+	addql	IMM (8), sp
+#else
+	mulsl	d1,d0
+#endif
+	movel	sp@(4), d1	/* d1 = dividend */
+	subl	d0, d1		/* d1 = a - (a/b)*b */
+	movel	d1, d0
+	rts
+	EXPORT_SYMBOL(__umodsi3)