Remove kcopy(), we don't use it.

2 files changed, 0 insertions, 630 deletions

diff --git a/sys/arm/arm/bcopyinout.S b/sys/arm/arm/bcopyinout.S
index 1fc99f5..fdd57d6 100644
--- a/sys/arm/arm/bcopyinout.S
+++ b/sys/arm/arm/bcopyinout.S
@@ -502,210 +502,8 @@ ENTRY(copyout)
 	RESTORE_REGS
 
 	RET
-
-/*
- * r0 = kernel space source address
- * r1 = kernel space destination address
- * r2 = length
- *
- * Copies bytes from kernel space to kernel space, aborting on page fault
- *
- * Copy of copyout, but without the ldrt/strt instructions.
- */
-
-ENTRY(kcopy)
-	/* Quick exit if length is zero */	
-	teq	r2, #0
-	moveq	r0, #0
-	RETeq
-
-	SAVE_REGS
-#ifdef MULTIPROCESSOR
-	/* XXX Probably not appropriate for non-Hydra SMPs */
-	stmfd	sp!, {r0-r2, r14}
-	bl	_C_LABEL(cpu_number)
-	ldr	r4, .Lcpu_info
-	ldr	r4, [r4, r0, lsl #2]
-	ldr	r4, [r4, #CI_CURPCB]
-	ldmfd	sp!, {r0-r2, r14}
-#else
-	ldr	r4, .Lcurpcb
-	ldr	r4, [r4]
 #endif
 
-	ldr	r5, [r4, #PCB_ONFAULT]
-	adr	r3, .Lcopyfault
-	str	r3, [r4, #PCB_ONFAULT]
-
-	PREFETCH(r0, 0)
-	PREFETCH(r1, 0)
-
-	/*
-	 * If not too many bytes, take the slow path.
-	 */
-	cmp	r2, #0x08
-	blt	.Lkcleanup
-
-	/*
-	 * Align destination to word boundary.
-	 */
-	and	r6, r1, #0x3
-	ldr	pc, [pc, r6, lsl #2]
-	b	.Lkalend
-	.word	.Lkalend
-	.word	.Lkal3
-	.word	.Lkal2
-	.word	.Lkal1
-.Lkal3:	ldrb	r6, [r0], #1
-	sub	r2, r2, #1
-	strb	r6, [r1], #1
-.Lkal2:	ldrb	r7, [r0], #1
-	sub	r2, r2, #1
-	strb	r7, [r1], #1
-.Lkal1:	ldrb	r6, [r0], #1
-	sub	r2, r2, #1
-	strb	r6, [r1], #1
-.Lkalend:
-
-	/*
-	 * If few bytes left, finish slow.
-	 */
-	cmp	r2, #0x08
-	blt	.Lkcleanup
-
-	/*
-	 * If source is not aligned, finish slow.
-	 */
-	ands	r3, r0, #0x03
-	bne	.Lkcleanup
-
-	cmp	r2, #0x60	/* Must be > 0x5f for unrolled cacheline */
-	blt	.Lkcleanup8
-
-	/*
-	 * Align source & destination to cacheline boundary.
-	 */
-	and	r6, r1, #0x1f
-	ldr	pc, [pc, r6]
-	b	.Lkcaligned
-	.word	.Lkcaligned
-	.word	.Lkcal28
-	.word	.Lkcal24
-	.word	.Lkcal20
-	.word	.Lkcal16
-	.word	.Lkcal12
-	.word	.Lkcal8
-	.word	.Lkcal4
-.Lkcal28:ldr	r6, [r0], #4
-	sub	r2, r2, #4
-	str	r6, [r1], #4
-.Lkcal24:ldr	r7, [r0], #4
-	sub	r2, r2, #4
-	str	r7, [r1], #4
-.Lkcal20:ldr	r6, [r0], #4
-	sub	r2, r2, #4
-	str	r6, [r1], #4
-.Lkcal16:ldr	r7, [r0], #4
-	sub	r2, r2, #4
-	str	r7, [r1], #4
-.Lkcal12:ldr	r6, [r0], #4
-	sub	r2, r2, #4
-	str	r6, [r1], #4
-.Lkcal8:ldr	r7, [r0], #4
-	sub	r2, r2, #4
-	str	r7, [r1], #4
-.Lkcal4:ldr	r6, [r0], #4
-	sub	r2, r2, #4
-	str	r6, [r1], #4
-
-	/*
-	 * We start with > 0x40 bytes to copy (>= 0x60 got us into this
-	 * part of the code, and we may have knocked that down by as much
-	 * as 0x1c getting aligned).
-	 *
-	 * This loop basically works out to:
-	 * do {
-	 * 	prefetch-next-cacheline(s)
-	 *	bytes -= 0x20;
-	 *	copy cacheline
-	 * } while (bytes >= 0x40);
-	 * bytes -= 0x20;
-	 * copy cacheline
-	 */
-.Lkcaligned:
-	PREFETCH(r0, 32)
-	PREFETCH(r1, 32)
-
-	sub	r2, r2, #0x20
-
-	/* Copy a cacheline */
-	ldmia	r0!, {r6-r11}
-	stmia	r1!, {r6, r7}
-	ldmia	r0!, {r6, r7}
-	stmia	r1!, {r8-r11}
-	stmia	r1!, {r6, r7}
-
-	cmp	r2, #0x40
-	bge	.Lkcaligned
-
-	sub	r2, r2, #0x20
-
-	/* Copy a cacheline */
-	ldmia	r0!, {r6-r11}
-	stmia	r1!, {r6-r7}
-	ldmia	r0!, {r6-r7}
-	stmia	r1!, {r8-r11}
-	stmia	r1!, {r6-r7}
-
-	cmp	r2, #0x08
-	blt	.Lkprecleanup
-
-.Lkcleanup8:
-	ldmia	r0!, {r8-r9}
-	sub	r2, r2, #8
-	stmia	r1!, {r8-r9}
-	cmp	r2, #8
-	bge	.Lkcleanup8
-
-.Lkprecleanup:
-	/*
-	 * If we're done, bail.
-	 */
-	cmp	r2, #0
-	beq	.Lkout
-
-.Lkcleanup:
-	and	r6, r2, #0x3
-	ldr	pc, [pc, r6, lsl #2]
-	b	.Lkcend
-	.word	.Lkc4
-	.word	.Lkc1
-	.word	.Lkc2
-	.word	.Lkc3
-.Lkc4:	ldrb	r6, [r0], #1
-	sub	r2, r2, #1
-	strb	r6, [r1], #1
-.Lkc3:	ldrb	r7, [r0], #1
-	sub	r2, r2, #1
-	strb	r7, [r1], #1
-.Lkc2:	ldrb	r6, [r0], #1
-	sub	r2, r2, #1
-	strb	r6, [r1], #1
-.Lkc1:	ldrb	r7, [r0], #1
-	subs	r2, r2, #1
-	strb	r7, [r1], #1
-.Lkcend:
-	bne	.Lkcleanup
-
-.Lkout:
-	mov	r0, #0
-
-	str	r5, [r4, #PCB_ONFAULT]
-	RESTORE_REGS
-
-	RET
-#endif	/* !__XSCALE__ */
-
 /*
  * int badaddr_read_1(const uint8_t *src, uint8_t *dest)
  *
diff --git a/sys/arm/arm/bcopyinout_xscale.S b/sys/arm/arm/bcopyinout_xscale.S
index d75a2dd..bb785f8 100644
--- a/sys/arm/arm/bcopyinout_xscale.S
+++ b/sys/arm/arm/bcopyinout_xscale.S
@@ -927,431 +927,3 @@ ENTRY(copyout)
 	ldrb	ip, [r0]
 	strbt	ip, [r1]
 	RET
-
-
-/*
- * r0 = kernel space source address
- * r1 = kernel space destination address
- * r2 = length
- *
- * Copies bytes from kernel space to kernel space, aborting on page fault
- */
-ENTRY(kcopy)
-	cmp	r2, #0x00
-	movle	r0, #0x00
-	movle	pc, lr			/* Bail early if length is <= 0 */
-
-	stmfd	sp!, {r10-r11, lr}
-
-#ifdef MULTIPROCESSOR
-	/* XXX Probably not appropriate for non-Hydra SMPs */
-	stmfd	sp!, {r0-r2}
-	bl	_C_LABEL(cpu_number)
-	ldr	r10, .Lcpu_info
-	ldmfd	sp!, {r0-r2}
-	ldr	r10, [r10, r0, lsl #2]
-	ldr	r10, [r10, #CI_CURPCB]
-#else
-	ldr	r10, .Lcurpcb
-	ldr	r10, [r10]
-#endif
-
-	mov	r3, #0x00
-	adr	ip, .Lkcopy_fault
-	ldr	r11, [r10, #PCB_ONFAULT]
-	str	ip, [r10, #PCB_ONFAULT]
-	bl	.Lkcopy_guts
-	str	r11, [r10, #PCB_ONFAULT]
-	mov	r0, #0x00
-	ldmfd	sp!, {r10-r11, pc}
-
-.Lkcopy_fault:
-	str	r11, [r10, #PCB_ONFAULT]
-	cmp	r3, #0x00
-	ldmgtfd	sp!, {r4-r7}		/* r3 > 0 Restore r4-r7 */
-	ldmltfd	sp!, {r4-r9}		/* r3 < 0 Restore r4-r9 */
-	ldmfd	sp!, {r10-r11, pc}
-
-.Lkcopy_guts:
-	pld	[r0]
-	/* Word-align the destination buffer */
-	ands	ip, r1, #0x03		/* Already word aligned? */
-	beq	.Lkcopy_wordaligned	/* Yup */
-	rsb	ip, ip, #0x04
-	cmp	r2, ip			/* Enough bytes left to align it? */
-	blt	.Lkcopy_bad_endgame2	/* Nope. Just copy bytewise */
-	sub	r2, r2, ip
-	rsbs	ip, ip, #0x03
-	addne	pc, pc, ip, lsl #3
-	ldrb	ip, [r0], #0x01
-	strb	ip, [r1], #0x01
-	ldrb	ip, [r0], #0x01
-	strb	ip, [r1], #0x01
-	ldrb	ip, [r0], #0x01
-	strb	ip, [r1], #0x01
-	cmp	r2, #0x00		/* All done? */
-	RETeq
-
-	/* Destination buffer is now word aligned */
-.Lkcopy_wordaligned:
-	ands	ip, r0, #0x03		/* Is src also word-aligned? */
-	bne	.Lkcopy_bad_align	/* Nope. Things just got bad */
-	cmp	r2, #0x08		/* Less than 8 bytes remaining? */
-	blt	.Lkcopy_w_less_than8
-
-	/* Quad-align the destination buffer */
-	tst	r1, #0x07		/* Already quad aligned? */
-	ldrne	ip, [r0], #0x04
-	stmfd	sp!, {r4-r9}		/* Free up some registers */
-	mov	r3, #-1			/* Signal restore r4-r9 */
-	subne	r2, r2, #0x04
-	strne	ip, [r1], #0x04
-
-	/* Destination buffer quad aligned, source is word aligned */
-	subs	r2, r2, #0x80
-	blt	.Lkcopy_w_lessthan128
-
-	/* Copy 128 bytes at a time */
-.Lkcopy_w_loop128:
-	ldr	r4, [r0], #0x04		/* LD:00-03 */
-	ldr	r5, [r0], #0x04		/* LD:04-07 */
-	pld	[r0, #0x18]		/* Prefetch 0x20 */
-	ldr	r6, [r0], #0x04		/* LD:08-0b */
-	ldr	r7, [r0], #0x04		/* LD:0c-0f */
-	ldr	r8, [r0], #0x04		/* LD:10-13 */
-	ldr	r9, [r0], #0x04		/* LD:14-17 */
-	strd	r4, [r1], #0x08		/* ST:00-07 */
-	ldr	r4, [r0], #0x04		/* LD:18-1b */
-	ldr	r5, [r0], #0x04		/* LD:1c-1f */
-	strd	r6, [r1], #0x08		/* ST:08-0f */
-	ldr	r6, [r0], #0x04		/* LD:20-23 */
-	ldr	r7, [r0], #0x04		/* LD:24-27 */
-	pld	[r0, #0x18]		/* Prefetch 0x40 */
-	strd	r8, [r1], #0x08		/* ST:10-17 */
-	ldr	r8, [r0], #0x04		/* LD:28-2b */
-	ldr	r9, [r0], #0x04		/* LD:2c-2f */
-	strd	r4, [r1], #0x08		/* ST:18-1f */
-	ldr	r4, [r0], #0x04		/* LD:30-33 */
-	ldr	r5, [r0], #0x04		/* LD:34-37 */
-	strd	r6, [r1], #0x08		/* ST:20-27 */
-	ldr	r6, [r0], #0x04		/* LD:38-3b */
-	ldr	r7, [r0], #0x04		/* LD:3c-3f */
-	strd	r8, [r1], #0x08		/* ST:28-2f */
-	ldr	r8, [r0], #0x04		/* LD:40-43 */
-	ldr	r9, [r0], #0x04		/* LD:44-47 */
-	pld	[r0, #0x18]		/* Prefetch 0x60 */
-	strd	r4, [r1], #0x08		/* ST:30-37 */
-	ldr	r4, [r0], #0x04		/* LD:48-4b */
-	ldr	r5, [r0], #0x04		/* LD:4c-4f */
-	strd	r6, [r1], #0x08		/* ST:38-3f */
-	ldr	r6, [r0], #0x04		/* LD:50-53 */
-	ldr	r7, [r0], #0x04		/* LD:54-57 */
-	strd	r8, [r1], #0x08		/* ST:40-47 */
-	ldr	r8, [r0], #0x04		/* LD:58-5b */
-	ldr	r9, [r0], #0x04		/* LD:5c-5f */
-	strd	r4, [r1], #0x08		/* ST:48-4f */
-	ldr	r4, [r0], #0x04		/* LD:60-63 */
-	ldr	r5, [r0], #0x04		/* LD:64-67 */
-	pld	[r0, #0x18]		/* Prefetch 0x80 */
-	strd	r6, [r1], #0x08		/* ST:50-57 */
-	ldr	r6, [r0], #0x04		/* LD:68-6b */
-	ldr	r7, [r0], #0x04		/* LD:6c-6f */
-	strd	r8, [r1], #0x08		/* ST:58-5f */
-	ldr	r8, [r0], #0x04		/* LD:70-73 */
-	ldr	r9, [r0], #0x04		/* LD:74-77 */
-	strd	r4, [r1], #0x08		/* ST:60-67 */
-	ldr	r4, [r0], #0x04		/* LD:78-7b */
-	ldr	r5, [r0], #0x04		/* LD:7c-7f */
-	strd	r6, [r1], #0x08		/* ST:68-6f */
-	strd	r8, [r1], #0x08		/* ST:70-77 */
-	subs	r2, r2, #0x80
-	strd	r4, [r1], #0x08		/* ST:78-7f */
-	bge	.Lkcopy_w_loop128
-
-.Lkcopy_w_lessthan128:
-	adds	r2, r2, #0x80		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
-	RETeq				/* Return now if done */
-	subs	r2, r2, #0x20
-	blt	.Lkcopy_w_lessthan32
-
-	/* Copy 32 bytes at a time */
-.Lkcopy_w_loop32:
-	ldr	r4, [r0], #0x04
-	ldr	r5, [r0], #0x04
-	pld	[r0, #0x18]
-	ldr	r6, [r0], #0x04
-	ldr	r7, [r0], #0x04
-	ldr	r8, [r0], #0x04
-	ldr	r9, [r0], #0x04
-	strd	r4, [r1], #0x08
-	ldr	r4, [r0], #0x04
-	ldr	r5, [r0], #0x04
-	strd	r6, [r1], #0x08
-	strd	r8, [r1], #0x08
-	subs	r2, r2, #0x20
-	strd	r4, [r1], #0x08
-	bge	.Lkcopy_w_loop32
-
-.Lkcopy_w_lessthan32:
-	adds	r2, r2, #0x20		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
-	RETeq				/* Return now if done */
-
-	and	r4, r2, #0x18
-	rsb	r5, r4, #0x18
-	subs	r2, r2, r4
-	add	pc, pc, r5, lsl #1
-	nop
-
-	/* At least 24 bytes remaining */
-	ldr	r4, [r0], #0x04
-	ldr	r5, [r0], #0x04
-	nop
-	strd	r4, [r1], #0x08
-
-	/* At least 16 bytes remaining */
-	ldr	r4, [r0], #0x04
-	ldr	r5, [r0], #0x04
-	nop
-	strd	r4, [r1], #0x08
-
-	/* At least 8 bytes remaining */
-	ldr	r4, [r0], #0x04
-	ldr	r5, [r0], #0x04
-	nop
-	strd	r4, [r1], #0x08
-
-	/* Less than 8 bytes remaining */
-	ldmfd	sp!, {r4-r9}
-	RETeq				/* Return now if done */
-	mov	r3, #0x00
-
-.Lkcopy_w_less_than8:
-	subs	r2, r2, #0x04
-	ldrge	ip, [r0], #0x04
-	strge	ip, [r1], #0x04
-	RETeq				/* Return now if done */
-	addlt	r2, r2, #0x04
-	ldrb	ip, [r0], #0x01
-	cmp	r2, #0x02
-	ldrgeb	r2, [r0], #0x01
-	strb	ip, [r1], #0x01
-	ldrgtb	ip, [r0]
-	strgeb	r2, [r1], #0x01
-	strgtb	ip, [r1]
-	RET
-
-/*
- * At this point, it has not been possible to word align both buffers.
- * The destination buffer (r1) is word aligned, but the source buffer
- * (r0) is not.
- */
-.Lkcopy_bad_align:
-	stmfd	sp!, {r4-r7}
-	mov	r3, #0x01
-	bic	r0, r0, #0x03
-	cmp	ip, #2
-	ldr	ip, [r0], #0x04
-	bgt	.Lkcopy_bad3
-	beq	.Lkcopy_bad2
-	b	.Lkcopy_bad1
-
-.Lkcopy_bad1_loop16:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #8
-#else
-	mov	r4, ip, lsr #8
-#endif
-	ldr	r5, [r0], #0x04
-	pld	[r0, #0x018]
-	ldr	r6, [r0], #0x04
-	ldr	r7, [r0], #0x04
-	ldr	ip, [r0], #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, r5, lsr #24
-	mov	r5, r5, lsl #8
-	orr	r5, r5, r6, lsr #24
-	mov	r6, r6, lsl #8
-	orr	r6, r6, r7, lsr #24
-	mov	r7, r7, lsl #8
-	orr	r7, r7, ip, lsr #24
-#else
-	orr	r4, r4, r5, lsl #24
-	mov	r5, r5, lsr #8
-	orr	r5, r5, r6, lsl #24
-	mov	r6, r6, lsr #8
-	orr	r6, r6, r7, lsl #24
-	mov	r7, r7, lsr #8
-	orr	r7, r7, ip, lsl #24
-#endif
-	str	r4, [r1], #0x04
-	str	r5, [r1], #0x04
-	str	r6, [r1], #0x04
-	str	r7, [r1], #0x04
-.Lkcopy_bad1:
-	subs	r2, r2, #0x10         
-	bge	.Lkcopy_bad1_loop16
-
-	adds	r2, r2, #0x10         
-	ldmeqfd	sp!, {r4-r7}
-	RETeq				/* Return now if done */
-	subs	r2, r2, #0x04
-	sublt	r0, r0, #0x03
-	blt	.Lkcopy_bad_endgame
-
-.Lkcopy_bad1_loop4:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #8
-#else
-	mov	r4, ip, lsr #8
-#endif
-	ldr	ip, [r0], #0x04
-	subs	r2, r2, #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, ip, lsr #24
-#else
-	orr	r4, r4, ip, lsl #24
-#endif
-	str	r4, [r1], #0x04
-	bge	.Lkcopy_bad1_loop4
-	sub	r0, r0, #0x03
-	b	.Lkcopy_bad_endgame
-
-.Lkcopy_bad2_loop16:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #16
-#else
-	mov	r4, ip, lsr #16
-#endif
-	ldr	r5, [r0], #0x04
-	pld	[r0, #0x018]
-	ldr	r6, [r0], #0x04
-	ldr	r7, [r0], #0x04
-	ldr	ip, [r0], #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, r5, lsr #16
-	mov	r5, r5, lsl #16
-	orr	r5, r5, r6, lsr #16
-	mov	r6, r6, lsl #16
-	orr	r6, r6, r7, lsr #16
-	mov	r7, r7, lsl #16
-	orr	r7, r7, ip, lsr #16
-#else
-	orr	r4, r4, r5, lsl #16
-	mov	r5, r5, lsr #16
-	orr	r5, r5, r6, lsl #16
-	mov	r6, r6, lsr #16
-	orr	r6, r6, r7, lsl #16
-	mov	r7, r7, lsr #16
-	orr	r7, r7, ip, lsl #16
-#endif
-	str	r4, [r1], #0x04
-	str	r5, [r1], #0x04
-	str	r6, [r1], #0x04
-	str	r7, [r1], #0x04
-.Lkcopy_bad2:
-	subs	r2, r2, #0x10         
-	bge	.Lkcopy_bad2_loop16
-
-	adds	r2, r2, #0x10         
-	ldmeqfd	sp!, {r4-r7}
-	RETeq				/* Return now if done */
-	subs	r2, r2, #0x04
-	sublt	r0, r0, #0x02
-	blt	.Lkcopy_bad_endgame
-
-.Lkcopy_bad2_loop4:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #16
-#else
-	mov	r4, ip, lsr #16
-#endif
-	ldr	ip, [r0], #0x04
-	subs	r2, r2, #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, ip, lsr #16
-#else
-	orr	r4, r4, ip, lsl #16
-#endif
-	str	r4, [r1], #0x04
-	bge	.Lkcopy_bad2_loop4
-	sub	r0, r0, #0x02
-	b	.Lkcopy_bad_endgame
-
-.Lkcopy_bad3_loop16:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #24
-#else
-	mov	r4, ip, lsr #24
-#endif
-	ldr	r5, [r0], #0x04
-	pld	[r0, #0x018]
-	ldr	r6, [r0], #0x04
-	ldr	r7, [r0], #0x04
-	ldr	ip, [r0], #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, r5, lsr #8
-	mov	r5, r5, lsl #24
-	orr	r5, r5, r6, lsr #8
-	mov	r6, r6, lsl #24
-	orr	r6, r6, r7, lsr #8
-	mov	r7, r7, lsl #24
-	orr	r7, r7, ip, lsr #8
-#else
-	orr	r4, r4, r5, lsl #8
-	mov	r5, r5, lsr #24
-	orr	r5, r5, r6, lsl #8
-	mov	r6, r6, lsr #24
-	orr	r6, r6, r7, lsl #8
-	mov	r7, r7, lsr #24
-	orr	r7, r7, ip, lsl #8
-#endif
-	str	r4, [r1], #0x04
-	str	r5, [r1], #0x04
-	str	r6, [r1], #0x04
-	str	r7, [r1], #0x04
-.Lkcopy_bad3:
-	subs	r2, r2, #0x10         
-	bge	.Lkcopy_bad3_loop16
-
-	adds	r2, r2, #0x10         
-	ldmeqfd	sp!, {r4-r7}
-	RETeq				/* Return now if done */
-	subs	r2, r2, #0x04
-	sublt	r0, r0, #0x01
-	blt	.Lkcopy_bad_endgame
-
-.Lkcopy_bad3_loop4:
-#ifdef __ARMEB__
-	mov	r4, ip, lsl #24
-#else
-	mov	r4, ip, lsr #24
-#endif
-	ldr	ip, [r0], #0x04
-	subs	r2, r2, #0x04
-#ifdef __ARMEB__
-	orr	r4, r4, ip, lsr #8
-#else
-	orr	r4, r4, ip, lsl #8
-#endif
-	str	r4, [r1], #0x04
-	bge	.Lkcopy_bad3_loop4
-	sub	r0, r0, #0x01
-
-.Lkcopy_bad_endgame:
-	ldmfd	sp!, {r4-r7}
-	mov	r3, #0x00
-	adds	r2, r2, #0x04
-	RETeq
-.Lkcopy_bad_endgame2:
-	rsbs	r2, r2, #0x03
-	addne	pc, pc, r2, lsl #3
-	nop
-	ldrb	ip, [r0], #0x01
-	strb	ip, [r1], #0x01
-	ldrb	ip, [r0], #0x01
-	strb	ip, [r1], #0x01
-	ldrb	ip, [r0]
-	strb	ip, [r1]
-	RET