sh: math-emu support

This implements initial math-emu support, aimed primarily at SH-3.

Signed-off-by: Takashi YOSHII <takasi-y@ops.dti.ne.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>

3 files changed, 697 insertions, 0 deletions

diff --git a/arch/sh/math-emu/Makefile b/arch/sh/math-emu/Makefile
new file mode 100644
index 0000000..638b342c
--- /dev/null
+++ b/arch/sh/math-emu/Makefile
@@ -0,0 +1 @@
+obj-y	:= math.o
diff --git a/arch/sh/math-emu/math.c b/arch/sh/math-emu/math.c
new file mode 100644
index 0000000..26b6046
--- /dev/null
+++ b/arch/sh/math-emu/math.c
@@ -0,0 +1,624 @@
+/*
+ * arch/sh/math-emu/math.c
+ *
+ * Copyright (C) 2006 Takashi YOSHII <takasi-y@ops.dti.ne.jp>
+ *
+ * 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/config.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/signal.h>
+
+#include <asm/system.h>
+#include <asm/uaccess.h>
+#include <asm/processor.h>
+#include <asm/io.h>
+
+#include "sfp-util.h"
+#include <math-emu/soft-fp.h>
+#include <math-emu/single.h>
+#include <math-emu/double.h>
+
+#define	FPUL		(fregs->fpul)
+#define FPSCR		(fregs->fpscr)
+#define FPSCR_RM	(FPSCR&3)
+#define FPSCR_DN	((FPSCR>>18)&1)
+#define FPSCR_PR	((FPSCR>>19)&1)
+#define FPSCR_SZ	((FPSCR>>20)&1)
+#define FPSCR_FR	((FPSCR>>21)&1)
+#define FPSCR_MASK	0x003fffffUL
+
+#define BANK(n)	(n^(FPSCR_FR?16:0))
+#define FR	((unsigned long*)(fregs->fp_regs))
+#define FR0	(FR[BANK(0)])
+#define FRn	(FR[BANK(n)])
+#define FRm	(FR[BANK(m)])
+#define DR	((unsigned long long*)(fregs->fp_regs))
+#define DRn	(DR[BANK(n)/2])
+#define DRm	(DR[BANK(m)/2])
+
+#define XREG(n)	(n^16)
+#define XFn	(FR[BANK(XREG(n))])
+#define XFm	(FR[BANK(XREG(m))])
+#define XDn	(DR[BANK(XREG(n))/2])
+#define XDm	(DR[BANK(XREG(m))/2])
+
+#define R0	(regs->regs[0])
+#define Rn	(regs->regs[n])
+#define Rm	(regs->regs[m])
+
+#define WRITE(d,a)	({if(put_user(d, (typeof (d)*)a)) return -EFAULT;})
+#define READ(d,a)	({if(get_user(d, (typeof (d)*)a)) return -EFAULT;})
+
+#define PACK_S(r,f)	FP_PACK_SP(&r,f)
+#define UNPACK_S(f,r)	FP_UNPACK_SP(f,&r)
+#define PACK_D(r,f) \
+	{u32 t[2]; FP_PACK_DP(t,f); ((u32*)&r)[0]=t[1]; ((u32*)&r)[1]=t[0];}
+#define UNPACK_D(f,r) \
+	{u32 t[2]; t[0]=((u32*)&r)[1]; t[1]=((u32*)&r)[0]; FP_UNPACK_DP(f,t);}
+
+// 2 args instructions.
+#define BOTH_PRmn(op,x) \
+	FP_DECL_EX; if(FPSCR_PR) op(D,x,DRm,DRn); else op(S,x,FRm,FRn);
+
+#define CMP_X(SZ,R,M,N) do{ \
+	FP_DECL_##SZ(Fm); FP_DECL_##SZ(Fn); \
+	UNPACK_##SZ(Fm, M); UNPACK_##SZ(Fn, N); \
+	FP_CMP_##SZ(R, Fn, Fm, 2); }while(0)
+#define EQ_X(SZ,R,M,N) do{ \
+	FP_DECL_##SZ(Fm); FP_DECL_##SZ(Fn); \
+	UNPACK_##SZ(Fm, M); UNPACK_##SZ(Fn, N); \
+	FP_CMP_EQ_##SZ(R, Fn, Fm); }while(0)
+#define CMP(OP) ({ int r; BOTH_PRmn(OP##_X,r); r; })
+
+static int
+fcmp_gt(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	if (CMP(CMP) > 0)
+		regs->sr |= 1;
+	else
+		regs->sr &= ~1;
+
+	return 0;
+}
+
+static int
+fcmp_eq(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	if (CMP(CMP /*EQ*/) == 0)
+		regs->sr |= 1;
+	else
+		regs->sr &= ~1;
+	return 0;
+}
+
+#define ARITH_X(SZ,OP,M,N) do{ \
+	FP_DECL_##SZ(Fm); FP_DECL_##SZ(Fn); FP_DECL_##SZ(Fr); \
+	UNPACK_##SZ(Fm, M); UNPACK_##SZ(Fn, N); \
+	FP_##OP##_##SZ(Fr, Fn, Fm); \
+	PACK_##SZ(N, Fr); }while(0)
+
+static int
+fadd(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	BOTH_PRmn(ARITH_X, ADD);
+	return 0;
+}
+
+static int
+fsub(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	BOTH_PRmn(ARITH_X, SUB);
+	return 0;
+}
+
+static int
+fmul(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	BOTH_PRmn(ARITH_X, MUL);
+	return 0;
+}
+
+static int
+fdiv(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	BOTH_PRmn(ARITH_X, DIV);
+	return 0;
+}
+
+static int
+fmac(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	FP_DECL_EX;
+	FP_DECL_S(Fr);
+	FP_DECL_S(Ft);
+	FP_DECL_S(F0);
+	FP_DECL_S(Fm);
+	FP_DECL_S(Fn);
+	UNPACK_S(F0, FR0);
+	UNPACK_S(Fm, FRm);
+	UNPACK_S(Fn, FRn);
+	FP_MUL_S(Ft, Fm, F0);
+	FP_ADD_S(Fr, Fn, Ft);
+	PACK_S(FRn, Fr);
+	return 0;
+}
+
+// to process fmov's extention (odd n for DR access XD).
+#define FMOV_EXT(x) if(x&1) x+=16-1
+
+static int
+fmov_idx_reg(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(n);
+		READ(FRn, Rm + R0 + 4);
+		n++;
+		READ(FRn, Rm + R0);
+	} else {
+		READ(FRn, Rm + R0);
+	}
+
+	return 0;
+}
+
+static int
+fmov_mem_reg(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(n);
+		READ(FRn, Rm + 4);
+		n++;
+		READ(FRn, Rm);
+	} else {
+		READ(FRn, Rm);
+	}
+
+	return 0;
+}
+
+static int
+fmov_inc_reg(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(n);
+		READ(FRn, Rm + 4);
+		n++;
+		READ(FRn, Rm);
+		Rm += 8;
+	} else {
+		READ(FRn, Rm);
+		Rm += 4;
+	}
+
+	return 0;
+}
+
+static int
+fmov_reg_idx(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(m);
+		WRITE(FRm, Rn + R0 + 4);
+		m++;
+		WRITE(FRm, Rn + R0);
+	} else {
+		WRITE(FRm, Rn + R0);
+	}
+
+	return 0;
+}
+
+static int
+fmov_reg_mem(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(m);
+		WRITE(FRm, Rn + 4);
+		m++;
+		WRITE(FRm, Rn);
+	} else {
+		WRITE(FRm, Rn);
+	}
+
+	return 0;
+}
+
+static int
+fmov_reg_dec(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(m);
+		Rn -= 8;
+		WRITE(FRm, Rn + 4);
+		m++;
+		WRITE(FRm, Rn);
+	} else {
+		Rn -= 4;
+		WRITE(FRm, Rn);
+	}
+
+	return 0;
+}
+
+static int
+fmov_reg_reg(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m,
+	     int n)
+{
+	if (FPSCR_SZ) {
+		FMOV_EXT(m);
+		FMOV_EXT(n);
+		DRn = DRm;
+	} else {
+		FRn = FRm;
+	}
+
+	return 0;
+}
+
+static int
+fnop_mn(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int m, int n)
+{
+	return -EINVAL;
+}
+
+// 1 arg instructions.
+#define NOTYETn(i) static int i(struct sh_fpu_soft_struct *fregs, int n) \
+	{ printk( #i " not yet done.\n"); return 0; }
+
+NOTYETn(ftrv)
+NOTYETn(fsqrt)
+NOTYETn(fipr)
+NOTYETn(fsca)
+NOTYETn(fsrra)
+
+#define EMU_FLOAT_X(SZ,N) do { \
+	FP_DECL_##SZ(Fn); \
+	FP_FROM_INT_##SZ(Fn, FPUL, 32, int); \
+	PACK_##SZ(N, Fn); }while(0)
+static int ffloat(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FP_DECL_EX;
+
+	if (FPSCR_PR)
+		EMU_FLOAT_X(D, DRn);
+	else
+		EMU_FLOAT_X(S, FRn);
+
+	return 0;
+}
+
+#define EMU_FTRC_X(SZ,N) do { \
+	FP_DECL_##SZ(Fn); \
+	UNPACK_##SZ(Fn, N); \
+	FP_TO_INT_##SZ(FPUL, Fn, 32, 1); }while(0)
+static int ftrc(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FP_DECL_EX;
+
+	if (FPSCR_PR)
+		EMU_FTRC_X(D, DRn);
+	else
+		EMU_FTRC_X(S, FRn);
+
+	return 0;
+}
+
+static int fcnvsd(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FP_DECL_EX;
+	FP_DECL_S(Fn);
+	FP_DECL_D(Fr);
+	UNPACK_S(Fn, FPUL);
+	FP_CONV(D, S, 2, 1, Fr, Fn);
+	PACK_D(DRn, Fr);
+	return 0;
+}
+
+static int fcnvds(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FP_DECL_EX;
+	FP_DECL_D(Fn);
+	FP_DECL_S(Fr);
+	UNPACK_D(Fn, DRn);
+	FP_CONV(S, D, 1, 2, Fr, Fn);
+	PACK_S(FPUL, Fr);
+	return 0;
+}
+
+static int fxchg(struct sh_fpu_soft_struct *fregs, int flag)
+{
+	FPSCR ^= flag;
+	return 0;
+}
+
+static int fsts(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FRn = FPUL;
+	return 0;
+}
+
+static int flds(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FPUL = FRn;
+	return 0;
+}
+
+static int fneg(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FRn ^= (1 << (_FP_W_TYPE_SIZE - 1));
+	return 0;
+}
+
+static int fabs(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FRn &= ~(1 << (_FP_W_TYPE_SIZE - 1));
+	return 0;
+}
+
+static int fld0(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FRn = 0;
+	return 0;
+}
+
+static int fld1(struct sh_fpu_soft_struct *fregs, int n)
+{
+	FRn = (_FP_EXPBIAS_S << (_FP_FRACBITS_S - 1));
+	return 0;
+}
+
+static int fnop_n(struct sh_fpu_soft_struct *fregs, int n)
+{
+	return -EINVAL;
+}
+
+/// Instruction decoders.
+
+static int id_fxfd(struct sh_fpu_soft_struct *, int);
+static int id_fnxd(struct sh_fpu_soft_struct *, struct pt_regs *, int, int);
+
+static int (*fnxd[])(struct sh_fpu_soft_struct *, int) = {
+	fsts, flds, ffloat, ftrc, fneg, fabs, fsqrt, fsrra,
+	fld0, fld1, fcnvsd, fcnvds, fnop_n, fnop_n, fipr, id_fxfd
+};
+
+static int (*fnmx[])(struct sh_fpu_soft_struct *, struct pt_regs *, int, int) = {
+	fadd, fsub, fmul, fdiv, fcmp_eq, fcmp_gt, fmov_idx_reg, fmov_reg_idx,
+	fmov_mem_reg, fmov_inc_reg, fmov_reg_mem, fmov_reg_dec,
+	fmov_reg_reg, id_fnxd, fmac, fnop_mn};
+
+static int id_fxfd(struct sh_fpu_soft_struct *fregs, int x)
+{
+	const int flag[] = { FPSCR_SZ, FPSCR_PR, FPSCR_FR, 0 };
+	switch (x & 3) {
+	case 3:
+		fxchg(fregs, flag[x >> 2]);
+		break;
+	case 1:
+		ftrv(fregs, x - 1);
+		break;
+	default:
+		fsca(fregs, x);
+	}
+	return 0;
+}
+
+static int
+id_fnxd(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, int x, int n)
+{
+	return (fnxd[x])(fregs, n);
+}
+
+static int
+id_fnmx(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, u16 code)
+{
+	int n = (code >> 8) & 0xf, m = (code >> 4) & 0xf, x = code & 0xf;
+	return (fnmx[x])(fregs, regs, m, n);
+}
+
+static int
+id_sys(struct sh_fpu_soft_struct *fregs, struct pt_regs *regs, u16 code)
+{
+	int n = ((code >> 8) & 0xf);
+	unsigned long *reg = (code & 0x0010) ? &FPUL : &FPSCR;
+
+	switch (code & 0xf0ff) {
+	case 0x005a:
+	case 0x006a:
+		Rn = *reg;
+		break;
+	case 0x405a:
+	case 0x406a:
+		*reg = Rn;
+		break;
+	case 0x4052:
+	case 0x4062:
+		Rn -= 4;
+		WRITE(*reg, Rn);
+		break;
+	case 0x4056:
+	case 0x4066:
+		READ(*reg, Rn);
+		Rn += 4;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int fpu_emulate(u16 code, struct sh_fpu_soft_struct *fregs, struct pt_regs *regs)
+{
+	if ((code & 0xf000) == 0xf000)
+		return id_fnmx(fregs, regs, code);
+	else
+		return id_sys(fregs, regs, code);
+}
+
+/**
+ *	denormal_to_double - Given denormalized float number,
+ *	                     store double float
+ *
+ *	@fpu: Pointer to sh_fpu_hard structure
+ *	@n: Index to FP register
+ */
+static void denormal_to_double(struct sh_fpu_hard_struct *fpu, int n)
+{
+	unsigned long du, dl;
+	unsigned long x = fpu->fpul;
+	int exp = 1023 - 126;
+
+	if (x != 0 && (x & 0x7f800000) == 0) {
+		du = (x & 0x80000000);
+		while ((x & 0x00800000) == 0) {
+			x <<= 1;
+			exp--;
+		}
+		x &= 0x007fffff;
+		du |= (exp << 20) | (x >> 3);
+		dl = x << 29;
+
+		fpu->fp_regs[n] = du;
+		fpu->fp_regs[n+1] = dl;
+	}
+}
+
+/**
+ *	ieee_fpe_handler - Handle denormalized number exception
+ *
+ *	@regs: Pointer to register structure
+ *
+ *	Returns 1 when it's handled (should not cause exception).
+ */
+static int ieee_fpe_handler(struct pt_regs *regs)
+{
+	unsigned short insn = *(unsigned short *)regs->pc;
+	unsigned short finsn;
+	unsigned long nextpc;
+	int nib[4] = {
+		(insn >> 12) & 0xf,
+		(insn >> 8) & 0xf,
+		(insn >> 4) & 0xf,
+		insn & 0xf};
+
+	if (nib[0] == 0xb ||
+	    (nib[0] == 0x4 && nib[2] == 0x0 && nib[3] == 0xb)) /* bsr & jsr */
+		regs->pr = regs->pc + 4;
+
+	if (nib[0] == 0xa || nib[0] == 0xb) { /* bra & bsr */
+		nextpc = regs->pc + 4 + ((short) ((insn & 0xfff) << 4) >> 3);
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else if (nib[0] == 0x8 && nib[1] == 0xd) { /* bt/s */
+		if (regs->sr & 1)
+			nextpc = regs->pc + 4 + ((char) (insn & 0xff) << 1);
+		else
+			nextpc = regs->pc + 4;
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else if (nib[0] == 0x8 && nib[1] == 0xf) { /* bf/s */
+		if (regs->sr & 1)
+			nextpc = regs->pc + 4;
+		else
+			nextpc = regs->pc + 4 + ((char) (insn & 0xff) << 1);
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else if (nib[0] == 0x4 && nib[3] == 0xb &&
+		 (nib[2] == 0x0 || nib[2] == 0x2)) { /* jmp & jsr */
+		nextpc = regs->regs[nib[1]];
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else if (nib[0] == 0x0 && nib[3] == 0x3 &&
+		 (nib[2] == 0x0 || nib[2] == 0x2)) { /* braf & bsrf */
+		nextpc = regs->pc + 4 + regs->regs[nib[1]];
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else if (insn == 0x000b) { /* rts */
+		nextpc = regs->pr;
+		finsn = *(unsigned short *) (regs->pc + 2);
+	} else {
+		nextpc = regs->pc + 2;
+		finsn = insn;
+	}
+
+	if ((finsn & 0xf1ff) == 0xf0ad) { /* fcnvsd */
+		struct task_struct *tsk = current;
+
+		if ((tsk->thread.fpu.hard.fpscr & (1 << 17))) {
+			/* FPU error */
+			denormal_to_double (&tsk->thread.fpu.hard,
+					    (finsn >> 8) & 0xf);
+			tsk->thread.fpu.hard.fpscr &=
+				~(FPSCR_CAUSE_MASK | FPSCR_FLAG_MASK);
+			set_tsk_thread_flag(tsk, TIF_USEDFPU);
+		} else {
+			tsk->thread.trap_no = 11;
+			tsk->thread.error_code = 0;
+			force_sig(SIGFPE, tsk);
+		}
+
+		regs->pc = nextpc;
+		return 1;
+	}
+
+	return 0;
+}
+
+asmlinkage void do_fpu_error(unsigned long r4, unsigned long r5,
+			     unsigned long r6, unsigned long r7,
+			     struct pt_regs regs)
+{
+	struct task_struct *tsk = current;
+
+	if (ieee_fpe_handler (&regs))
+		return;
+
+	regs.pc += 2;
+	tsk->thread.trap_no = 11;
+	tsk->thread.error_code = 0;
+	force_sig(SIGFPE, tsk);
+}
+
+/**
+ * fpu_init - Initialize FPU registers
+ * @fpu: Pointer to software emulated FPU registers.
+ */
+static void fpu_init(struct sh_fpu_soft_struct *fpu)
+{
+	int i;
+
+	fpu->fpscr = FPSCR_INIT;
+	fpu->fpul = 0;
+
+	for (i = 0; i < 16; i++) {
+		fpu->fp_regs[i] = 0;
+		fpu->xfp_regs[i]= 0;
+	}
+}
+
+/**
+ * do_fpu_inst - Handle reserved instructions for FPU emulation
+ * @inst: instruction code.
+ * @regs: registers on stack.
+ */
+int do_fpu_inst(unsigned short inst, struct pt_regs *regs)
+{
+	struct task_struct *tsk = current;
+	struct sh_fpu_soft_struct *fpu = &(tsk->thread.fpu.soft);
+
+	if (!test_tsk_thread_flag(tsk, TIF_USEDFPU)) {
+		/* initialize once. */
+		fpu_init(fpu);
+		set_tsk_thread_flag(tsk, TIF_USEDFPU);
+	}
+
+	return fpu_emulate(inst, fpu, regs);
+}
diff --git a/arch/sh/math-emu/sfp-util.h b/arch/sh/math-emu/sfp-util.h
new file mode 100644
index 0000000..8ae1bd3
--- /dev/null
+++ b/arch/sh/math-emu/sfp-util.h
@@ -0,0 +1,72 @@
+/*
+ * These are copied from glibc/stdlib/longlong.h
+ */
+
+#define add_ssaaaa(sh, sl, ah, al, bh, bl) \
+  do {                                                                  \
+    UWtype __x;                                                         \
+    __x = (al) + (bl);                                                  \
+    (sh) = (ah) + (bh) + (__x < (al));                                  \
+    (sl) = __x;                                                         \
+  } while (0)
+
+#define sub_ddmmss(sh, sl, ah, al, bh, bl) \
+  do {                                                                  \
+    UWtype __x;                                                         \
+    __x = (al) - (bl);                                                  \
+    (sh) = (ah) - (bh) - (__x > (al));                                  \
+    (sl) = __x;                                                         \
+  } while (0)
+
+#define umul_ppmm(w1, w0, u, v) \
+  __asm__ ("dmulu.l %2,%3\n\tsts    macl,%1\n\tsts  mach,%0"	\
+	: "=r" ((u32)(w1)), "=r" ((u32)(w0))	\
+	:  "r" ((u32)(u)),   "r" ((u32)(v))	\
+	: "macl", "mach")
+
+#define __ll_B ((UWtype) 1 << (W_TYPE_SIZE / 2))
+#define __ll_lowpart(t) ((UWtype) (t) & (__ll_B - 1))
+#define __ll_highpart(t) ((UWtype) (t) >> (W_TYPE_SIZE / 2))
+
+#define udiv_qrnnd(q, r, n1, n0, d) \
+  do {									\
+    UWtype __d1, __d0, __q1, __q0;					\
+    UWtype __r1, __r0, __m;						\
+    __d1 = __ll_highpart (d);						\
+    __d0 = __ll_lowpart (d);						\
+									\
+    __r1 = (n1) % __d1;							\
+    __q1 = (n1) / __d1;							\
+    __m = (UWtype) __q1 * __d0;						\
+    __r1 = __r1 * __ll_B | __ll_highpart (n0);				\
+    if (__r1 < __m)							\
+      {									\
+	__q1--, __r1 += (d);						\
+	if (__r1 >= (d)) /* i.e. we didn't get carry when adding to __r1 */\
+	  if (__r1 < __m)						\
+	    __q1--, __r1 += (d);					\
+      }									\
+    __r1 -= __m;							\
+									\
+    __r0 = __r1 % __d1;							\
+    __q0 = __r1 / __d1;							\
+    __m = (UWtype) __q0 * __d0;						\
+    __r0 = __r0 * __ll_B | __ll_lowpart (n0);				\
+    if (__r0 < __m)							\
+      {									\
+	__q0--, __r0 += (d);						\
+	if (__r0 >= (d))						\
+	  if (__r0 < __m)						\
+	    __q0--, __r0 += (d);					\
+      }									\
+    __r0 -= __m;							\
+									\
+    (q) = (UWtype) __q1 * __ll_B | __q0;				\
+    (r) = __r0;								\
+  } while (0)
+
+#define abort()	return 0
+
+#define __BYTE_ORDER __LITTLE_ENDIAN
+
+