Clean up our ARM assembly:

MFC 275256:

Switch to the ARM unified assembly language as the clang integrated as only
supports it. Binutils supports it when the ".syntax unified" directive
is set.

Sponsored by:   ABT Systems Ltd

MFC 275264:

Update _ENTRY to use _EENTRY to reduce the common code.

MFC 275321:

Remove extra labels, ENTRY_NP already provides them.

Sponsored by:   ABT Systems Ltd

MFC 275322:

Correctly a few incorrect uses of ENTRY/EENTRY and END/EEND

Sponsored by:   ABT Systems Ltd

MFC 275416:

Fix the name of the coprocessor to include the "p" prefix, the clang
integrated assembler expects this.

Sponsored by:   ABT Systems Ltd

MFC 275418:

Switch to unified syntax so these can be built with clang 3.5.

Sponsored by:   ABT Systems Ltd

MFC 275519:

Add missing END macros to some of the xscale functions.

Sponsored by:   ABT Systems Ltd

MFC 275520:

Use the unified syntax in a few more assembly files

Sponsored by:   ABT Systems Ltd

MFC 275521:

Set the alignment to 4-bytes after a string as clang 3.5 can switch to
thumb mode if this is incorrect.

Sponsored by:   ABT Systems Ltd

MFC 275522:

Place the literal pool after a RET otherwise clang 3.5 tries to put it too
far away from a ldr psuedo instruction. With this clang will place the
literal value here where it's close enough to be loaded.

Sponsored by:   ABT Systems Ltd

MFC 275523:

Switch to an armv6k cpu, without this clang 3.5 complains "bx lr" is
unsupported as it needs a newer cpu.

Sponsored by:   ABT Systems Ltd

MFC 275524:

Switch to a .cpu directive. These will work when clang 3.5 is imported
where the .arch directive is a nop.

Sponsored by:   ABT Systems Ltd

13 files changed, 225 insertions, 200 deletions

diff --git a/sys/arm/arm/bcopyinout_xscale.S b/sys/arm/arm/bcopyinout_xscale.S
index 45cfe70..5c1026f 100644
--- a/sys/arm/arm/bcopyinout_xscale.S
+++ b/sys/arm/arm/bcopyinout_xscale.S
@@ -38,6 +38,7 @@
 #include <machine/asm.h>
 __FBSDID("$FreeBSD$");
 
+	.syntax	unified
 	.text
 	.align	0
 
@@ -104,8 +105,8 @@ ENTRY(copyin)
 	ldr	r0, =EFAULT
 	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 */
+	ldmfdgt	sp!, {r4-r7}		/* r3 > 0 Restore r4-r7 */
+	ldmfdlt	sp!, {r4-r9}		/* r3 < 0 Restore r4-r9 */
 	ldmfd	sp!, {r10-r11, pc}
 
 .Lcopyin_guts:
@@ -138,7 +139,7 @@ ENTRY(copyin)
 
 	/* Quad-align the destination buffer */
 	tst	r1, #0x07		/* Already quad aligned? */
-	ldrnet	ip, [r0], #0x04
+	ldrtne	ip, [r0], #0x04
 	strne	ip, [r1], #0x04
 	subne	r2, r2, #0x04
 	stmfd	sp!, {r4-r9}		/* Free up some registers */
@@ -207,7 +208,7 @@ ENTRY(copyin)
 
 .Lcopyin_w_lessthan128:
 	adds	r2, r2, #0x80		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq
 	subs	r2, r2, #0x20
 	blt	.Lcopyin_w_lessthan32
@@ -232,7 +233,7 @@ ENTRY(copyin)
 
 .Lcopyin_w_lessthan32:
 	adds	r2, r2, #0x20		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq				/* Return now if done */
 
 	and	r4, r2, #0x18
@@ -266,17 +267,17 @@ ENTRY(copyin)
 
 .Lcopyin_w_less_than8:
 	subs	r2, r2, #0x04
-	ldrget	ip, [r0], #0x04
+	ldrtge	ip, [r0], #0x04
 	strge	ip, [r1], #0x04
 	RETeq				/* Return now if done */
 	addlt	r2, r2, #0x04
 	ldrbt	ip, [r0], #0x01
 	cmp	r2, #0x02
-	ldrgebt	r2, [r0], #0x01
+	ldrbtge	r2, [r0], #0x01
 	strb	ip, [r1], #0x01
-	ldrgtbt	ip, [r0]
-	strgeb	r2, [r1], #0x01
-	strgtb	ip, [r1]
+	ldrbtgt	ip, [r0]
+	strbge	r2, [r1], #0x01
+	strbgt	ip, [r1]
 	RET
 
 /*
@@ -331,7 +332,7 @@ ENTRY(copyin)
 	bge	.Lcopyin_bad1_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x03
@@ -392,7 +393,7 @@ ENTRY(copyin)
 	bge	.Lcopyin_bad2_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x02
@@ -453,7 +454,7 @@ ENTRY(copyin)
 	bge	.Lcopyin_bad3_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x01
@@ -546,8 +547,8 @@ ENTRY(copyout)
 	ldr	r0, =EFAULT
 	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 */
+	ldmfdgt	sp!, {r4-r7}		/* r3 > 0 Restore r4-r7 */
+	ldmfdlt	sp!, {r4-r9}		/* r3 < 0 Restore r4-r9 */
 	ldmfd	sp!, {r10-r11, pc}
 
 .Lcopyout_guts:
@@ -582,7 +583,7 @@ ENTRY(copyout)
 	tst	r0, #0x07		/* Already quad aligned? */
 	ldrne	ip, [r0], #0x04
 	subne	r2, r2, #0x04
-	strnet	ip, [r1], #0x04
+	strtne	ip, [r1], #0x04
 	
 	stmfd	sp!, {r4-r9}		/* Free up some registers */
 	mov	r3, #-1			/* Signal restore r4-r9 */
@@ -650,7 +651,7 @@ ENTRY(copyout)
 
 .Lcopyout_w_lessthan128:
 	adds	r2, r2, #0x80		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x20
 	blt	.Lcopyout_w_lessthan32
@@ -675,7 +676,7 @@ ENTRY(copyout)
 
 .Lcopyout_w_lessthan32:
 	adds	r2, r2, #0x20		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq				/* Return now if done */
 
 	and	r4, r2, #0x18
@@ -710,16 +711,16 @@ ENTRY(copyout)
 .Lcopyout_w_less_than8:
 	subs	r2, r2, #0x04
 	ldrge	ip, [r0], #0x04
-	strget	ip, [r1], #0x04
+	strtge	ip, [r1], #0x04
 	RETeq				/* Return now if done */
 	addlt	r2, r2, #0x04
 	ldrb	ip, [r0], #0x01
 	cmp	r2, #0x02
-	ldrgeb	r2, [r0], #0x01
+	ldrbge	r2, [r0], #0x01
 	strbt	ip, [r1], #0x01
-	ldrgtb	ip, [r0]
-	strgebt	r2, [r1], #0x01
-	strgtbt	ip, [r1]
+	ldrbgt	ip, [r0]
+	strbtge	r2, [r1], #0x01
+	strbtgt	ip, [r1]
 	RET
 
 /*
@@ -774,7 +775,7 @@ ENTRY(copyout)
 	bge	.Lcopyout_bad1_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x03
@@ -835,7 +836,7 @@ ENTRY(copyout)
 	bge	.Lcopyout_bad2_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x02
@@ -896,7 +897,7 @@ ENTRY(copyout)
 	bge	.Lcopyout_bad3_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq				/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r0, r0, #0x01
diff --git a/sys/arm/arm/blockio.S b/sys/arm/arm/blockio.S
index d121f2c..28f502b 100644
--- a/sys/arm/arm/blockio.S
+++ b/sys/arm/arm/blockio.S
@@ -50,6 +50,8 @@
 
 __FBSDID("$FreeBSD$");
 
+	.syntax	unified
+
 /*
  * Read bytes from an I/O address into a block of memory
  *
@@ -71,10 +73,10 @@ ENTRY(read_multi_1)
 	cmp	r12, #2
 	ldrb	r3, [r0]
 	strb	r3, [r1], #1
-	ldrgeb	r3, [r0]
-	strgeb	r3, [r1], #1
-	ldrgtb	r3, [r0]
-	strgtb	r3, [r1], #1
+	ldrbge	r3, [r0]
+	strbge	r3, [r1], #1
+	ldrbgt	r3, [r0]
+	strbgt	r3, [r1], #1
 	subs	r2, r2, r12
 	blt	.Lrm1_l4
 .Lrm1_main:
@@ -91,15 +93,15 @@ ENTRY(read_multi_1)
 	bge	.Lrm1loop
 .Lrm1_l4:
 	adds	r2, r2, #4			/* r2 = length again */
-	ldmeqdb	fp, {fp, sp, pc}
+	ldmdbeq	fp, {fp, sp, pc}
 	RETeq
 	cmp	r2, #2
 	ldrb	r3, [r0]
 	strb	r3, [r1], #1
-	ldrgeb	r3, [r0]
-	strgeb	r3, [r1], #1
-	ldrgtb	r3, [r0]
-	strgtb	r3, [r1], #1
+	ldrbge	r3, [r0]
+	strbge	r3, [r1], #1
+	ldrbgt	r3, [r0]
+	strbgt	r3, [r1], #1
 	ldmdb	fp, {fp, sp, pc}
 END(read_multi_1)
 
@@ -124,10 +126,10 @@ ENTRY(write_multi_1)
 	cmp	r12, #2
 	ldrb	r3, [r1], #1
 	strb	r3, [r0]
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0]
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0]
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0]
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0]
 	subs	r2, r2, r12
 	blt	.Lwm1_l4
 .Lwm1_main:
@@ -144,14 +146,14 @@ ENTRY(write_multi_1)
 	bge	.Lwm1loop
 .Lwm1_l4:
 	adds	r2, r2, #4			/* r2 = length again */
-	ldmeqdb	fp, {fp, sp, pc}
+	ldmdbeq	fp, {fp, sp, pc}
 	cmp	r2, #2
 	ldrb	r3, [r1], #1
 	strb	r3, [r0]
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0]
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0]
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0]
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0]
 	ldmdb	fp, {fp, sp, pc}
 END(write_multi_1)
 
diff --git a/sys/arm/arm/cpufunc_asm_arm11x6.S b/sys/arm/arm/cpufunc_asm_arm11x6.S
index 63fda9e..4d2697a 100644
--- a/sys/arm/arm/cpufunc_asm_arm11x6.S
+++ b/sys/arm/arm/cpufunc_asm_arm11x6.S
@@ -62,7 +62,7 @@
 #include <machine/asm.h>
 __FBSDID("$FreeBSD$");
 
-	.cpu arm1136js
+	.cpu arm1176jz-s
 
 #if 0
 #define Invalidate_I_cache(Rtmp1, Rtmp2) \
diff --git a/sys/arm/arm/cpufunc_asm_armv5.S b/sys/arm/arm/cpufunc_asm_armv5.S
index 7435b46..582ea50 100644
--- a/sys/arm/arm/cpufunc_asm_armv5.S
+++ b/sys/arm/arm/cpufunc_asm_armv5.S
@@ -194,7 +194,6 @@ ENTRY(armv5_idcache_wbinv_range)
 END(armv5_idcache_wbinv_range)
 
 ENTRY_NP(armv5_idcache_wbinv_all)
-armv5_idcache_wbinv_all:
 .Larmv5_idcache_wbinv_all:
 	/*
 	 * We assume that the code here can never be out of sync with the
diff --git a/sys/arm/arm/cpufunc_asm_xscale_c3.S b/sys/arm/arm/cpufunc_asm_xscale_c3.S
index 0fdb39c..bc4a8d5 100644
--- a/sys/arm/arm/cpufunc_asm_xscale_c3.S
+++ b/sys/arm/arm/cpufunc_asm_xscale_c3.S
@@ -143,7 +143,6 @@ __FBSDID("$FreeBSD$");
 
 
 ENTRY_NP(xscalec3_cache_syncI)
-xscalec3_cache_purgeID:
 EENTRY_NP(xscalec3_cache_purgeID)
 	mcr	p15, 0, r0, c7, c5, 0	/* flush I cache (D cleaned below) */
 EENTRY_NP(xscalec3_cache_cleanID)
diff --git a/sys/arm/arm/fusu.S b/sys/arm/arm/fusu.S
index c70215c..3d2716a 100644
--- a/sys/arm/arm/fusu.S
+++ b/sys/arm/arm/fusu.S
@@ -38,6 +38,8 @@
 #include "assym.s"
 __FBSDID("$FreeBSD$");
 
+	.syntax	unified
+
 #ifdef _ARM_ARCH_6
 #define GET_PCB(tmp) \
 	mrc p15, 0, tmp, c13, c0, 4; \
@@ -83,7 +85,7 @@ EENTRY_NP(casuword32)
 	ldrt	r5, [r0]
 	cmp	r5, r1
 	movne	r0, r5
-	streqt	r2, [r0]
+	strteq	r2, [r0]
 #endif
 	moveq	r0, r1
 2:
@@ -129,7 +131,7 @@ EENTRY_NP(fuword32)
 	str	r1, [r2, #PCB_ONFAULT]
 	mov	r0, r3
 	RET
-END(fuword32)
+EEND(fuword32)
 END(fuword)
 
 /*
@@ -269,7 +271,7 @@ _C_LABEL(fusubailout):
 
 fusupcbfaulttext:
 	.asciz	"Yikes - no valid PCB during fusuxxx() addr=%08x\n"
-	.align	0
+	.align	2
 #endif
 
 /*
@@ -295,7 +297,7 @@ EENTRY_NP(suword32)
 	mov	r0, #0x00000000
 	str	r0, [r2, #PCB_ONFAULT]
 	RET
-END(suword32)
+EEND(suword32)
 END(suword)
 
 /*
diff --git a/sys/arm/arm/in_cksum_arm.S b/sys/arm/arm/in_cksum_arm.S
index 6305caf..5f121d5 100644
--- a/sys/arm/arm/in_cksum_arm.S
+++ b/sys/arm/arm/in_cksum_arm.S
@@ -46,6 +46,7 @@
 #include "assym.s"
 __FBSDID("$FreeBSD$");
 
+	.syntax	unified
 /*
  * int in_cksum(struct mbuf *m, int len)
  *
@@ -128,9 +129,9 @@ ASENTRY_NP(L_cksumdata)
 	blt	.Lcksumdata_endgame
 	cmp	r7, #0x02
 	ldrb	r4, [r0], #0x01		/* Fetch 1st byte */
-	ldrgeb	r5, [r0], #0x01		/* Fetch 2nd byte */
+	ldrbge	r5, [r0], #0x01		/* Fetch 2nd byte */
 	movlt	r5, #0x00
-	ldrgtb	r6, [r0], #0x01		/* Fetch 3rd byte */
+	ldrbgt	r6, [r0], #0x01		/* Fetch 3rd byte */
 	movle	r6, #0x00
 	/* Combine the three bytes depending on endianness and alignment */
 #ifdef __ARMEB__
@@ -196,7 +197,7 @@ ASENTRY_NP(L_cksumdata)
 	adcs	r2, r2, r5
 	adc	r2, r2, #0x00
 	subs	r1, r1, #0x40
-	ldrged	r4, [r0], #0x08
+	ldrdge	r4, [r0], #0x08
 	bge	.Lcksumdata_bigloop
 
 	adds	r2, r2, r6		/* r6/r7 still need summing */
@@ -241,7 +242,7 @@ ASENTRY_NP(L_cksumdata)
 	cmp	r1, #0x20
 
 #ifdef _ARM_ARCH_5E
-	ldrged	r4, [r0], #0x08		/* Avoid stalling pld and result */
+	ldrdge	r4, [r0], #0x08		/* Avoid stalling pld and result */
 	blt	.Lcksumdata_less_than_32
 	pld	[r0, #0x18]
 	ldrd	r6, [r0], #0x08
@@ -319,9 +320,9 @@ ASENTRY_NP(L_cksumdata)
 .Lcksumdata_endgame:
 	ldrb	r3, [r0]		/* Fetch first byte */
 	cmp	r1, #0x02
-	ldrgeb	r4, [r0, #0x01]		/* Fetch 2nd and 3rd as necessary */
+	ldrbge	r4, [r0, #0x01]		/* Fetch 2nd and 3rd as necessary */
 	movlt	r4, #0x00
-	ldrgtb	r5, [r0, #0x02]
+	ldrbgt	r5, [r0, #0x02]
 	movle	r5, #0x00
 	/* Combine the three bytes depending on endianness and alignment */
 	tst	r0, #0x01
diff --git a/sys/arm/arm/locore.S b/sys/arm/arm/locore.S
index ce56efd..2835e17 100644
--- a/sys/arm/arm/locore.S
+++ b/sys/arm/arm/locore.S
@@ -474,8 +474,8 @@ ENTRY_NP(cpu_halt)
 	 * Hurl ourselves into the ROM
 	 */
 	mov	r0, #(CPU_CONTROL_32BP_ENABLE | CPU_CONTROL_32BD_ENABLE)
-	mcr     15, 0, r0, c1, c0, 0
-	mcrne   15, 0, r2, c8, c7, 0 	/* nail I+D TLB on ARMv4 and greater */
+	mcr     p15, 0, r0, c1, c0, 0
+	mcrne   p15, 0, r2, c8, c7, 0 	/* nail I+D TLB on ARMv4 and greater */
 	mov     pc, r4
 
 	/*
diff --git a/sys/arm/arm/support.S b/sys/arm/arm/support.S
index 2a6eec9..b095aed 100644
--- a/sys/arm/arm/support.S
+++ b/sys/arm/arm/support.S
@@ -91,6 +91,8 @@ __FBSDID("$FreeBSD$");
 
 #include "assym.s"
 
+	.syntax	unified
+
 .L_arm_memcpy:
 	.word	_C_LABEL(_arm_memcpy)
 .L_arm_bzero:
@@ -130,7 +132,7 @@ ENTRY(bzero)
 .Lnormal0:
 	mov	r3, #0x00
 	b	do_memset
-EEND(bzero)
+END(bzero)
 /* LINTSTUB: Func: void *memset(void *, int, size_t) */
 ENTRY(memset)
 	and	r3, r1, #0xff		/* We deal with bytes */
@@ -167,39 +169,39 @@ do_memset:
 .Lmemset_loop128:
 	subs	r1, r1, #0x80
 #ifdef _ARM_ARCH_5E
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-#else
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+#else
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
 #endif
 	bgt	.Lmemset_loop128
 	RETeq			/* Zero length so just exit */
@@ -210,15 +212,15 @@ do_memset:
 .Lmemset_loop32:
 	subs	r1, r1, #0x20
 #ifdef _ARM_ARCH_5E
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
 #else
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
 #endif
 	bgt	.Lmemset_loop32
 	RETeq			/* Zero length so just exit */
@@ -227,11 +229,11 @@ do_memset:
 
 	/* Deal with 16 bytes or more */
 #ifdef _ARM_ARCH_5E
-	strged	r2, [ip], #0x08
-	strged	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
+	strdge	r2, [ip], #0x08
 #else
-	stmgeia	ip!, {r2-r3}
-	stmgeia	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
+	stmiage	ip!, {r2-r3}
 #endif
 	RETeq			/* Zero length so just exit */
 
@@ -254,17 +256,17 @@ do_memset:
 #endif
 
 	strb	r3, [ip], #0x01		/* Set 1 byte */
-	strgeb	r3, [ip], #0x01		/* Set another byte */
-	strgtb	r3, [ip]		/* and a third */
+	strbge	r3, [ip], #0x01		/* Set another byte */
+	strbgt	r3, [ip]		/* and a third */
 	RET			/* Exit */
 
 .Lmemset_wordunaligned:
 	rsb	r2, r2, #0x004
 	strb	r3, [ip], #0x01		/* Set 1 byte */
 	cmp	r2, #0x02
-	strgeb	r3, [ip], #0x01		/* Set another byte */
+	strbge	r3, [ip], #0x01		/* Set another byte */
 	sub	r1, r1, r2
-	strgtb	r3, [ip], #0x01		/* and a third */
+	strbgt	r3, [ip], #0x01		/* and a third */
 	cmp	r1, #0x04		/* More than 4 bytes left? */
 	bge	.Lmemset_wordaligned	/* Yup */
 
@@ -273,10 +275,11 @@ do_memset:
 	RETeq			/* Zero length so exit */
 	strb	r3, [ip], #0x01		/* Set 1 byte */
 	cmp	r1, #0x02
-	strgeb	r3, [ip], #0x01		/* Set another byte */
-	strgtb	r3, [ip]		/* and a third */
+	strbge	r3, [ip], #0x01		/* Set another byte */
+	strbgt	r3, [ip]		/* and a third */
 	RET			/* Exit */
-END(memset)
+EEND(memset)
+END(bzero)
 
 ENTRY(bcmp)
 	mov	ip, r0
@@ -286,7 +289,7 @@ ENTRY(bcmp)
 
 	/* Are both addresses aligned the same way? */
 	cmp	r2, #0x00
-	eornes	r3, ip, r1
+	eorsne	r3, ip, r1
 	RETeq			/* len == 0, or same addresses! */
 	tst	r3, #0x03
 	subne	r2, r2, #0x01
@@ -366,23 +369,23 @@ ENTRY(bcmp)
 	ldrb	r0, [ip, #0x00]		/* r0 = b1#0 */
 	ldrb	r2, [r1, #0x01]		/* r2 = b2#1 */
 	subs	r0, r0, r3		/* r0 = b1#0 - b2#0 */
-	ldreqb	r3, [ip, #0x01]		/* r3 = b1#1 */
+	ldrbeq	r3, [ip, #0x01]		/* r3 = b1#1 */
 	RETne			/* Return if mismatch on #0 */
 	subs	r0, r3, r2		/* r0 = b1#1 - b2#1 */
-	ldreqb	r3, [r1, #0x02]		/* r3 = b2#2 */
-	ldreqb	r0, [ip, #0x02]		/* r0 = b1#2 */
+	ldrbeq	r3, [r1, #0x02]		/* r3 = b2#2 */
+	ldrbeq	r0, [ip, #0x02]		/* r0 = b1#2 */
 	RETne			/* Return if mismatch on #1 */
 	ldrb	r2, [r1, #0x03]		/* r2 = b2#3 */
 	subs	r0, r0, r3		/* r0 = b1#2 - b2#2 */
-	ldreqb	r3, [ip, #0x03]		/* r3 = b1#3 */
+	ldrbeq	r3, [ip, #0x03]		/* r3 = b1#3 */
 	RETne			/* Return if mismatch on #2 */
 	subs	r0, r3, r2		/* r0 = b1#3 - b2#3 */
-	ldreqb	r3, [r1, #0x04]		/* r3 = b2#4 */
-	ldreqb	r0, [ip, #0x04]		/* r0 = b1#4 */
+	ldrbeq	r3, [r1, #0x04]		/* r3 = b2#4 */
+	ldrbeq	r0, [ip, #0x04]		/* r0 = b1#4 */
 	RETne			/* Return if mismatch on #3 */
 	ldrb	r2, [r1, #0x05]		/* r2 = b2#5 */
 	subs	r0, r0, r3		/* r0 = b1#4 - b2#4 */
-	ldreqb	r3, [ip, #0x05]		/* r3 = b1#5 */
+	ldrbeq	r3, [ip, #0x05]		/* r3 = b1#5 */
 	RETne			/* Return if mismatch on #4 */
 	sub	r0, r3, r2		/* r0 = b1#5 - b2#5 */
 	RET
@@ -436,8 +439,8 @@ EENTRY(memmove)
 	bge	.Lmemmove_floop32
 
 	cmn	r2, #0x10
-	ldmgeia	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
-	stmgeia	r0!, {r3, r4, r12, lr}
+	ldmiage	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
+	stmiage	r0!, {r3, r4, r12, lr}
 	subge	r2, r2, #0x10
 	ldmia	sp!, {r4}		/* return r4 */
 
@@ -446,9 +449,9 @@ EENTRY(memmove)
 
 	/* blat 12 bytes at a time */
 .Lmemmove_floop12:
-	ldmgeia	r1!, {r3, r12, lr}
-	stmgeia	r0!, {r3, r12, lr}
-	subges	r2, r2, #0x0c
+	ldmiage	r1!, {r3, r12, lr}
+	stmiage	r0!, {r3, r12, lr}
+	subsge	r2, r2, #0x0c
 	bge	.Lmemmove_floop12
 
 .Lmemmove_fl12:
@@ -458,23 +461,23 @@ EENTRY(memmove)
 	subs	r2, r2, #4
 	ldrlt	r3, [r1], #4
 	strlt	r3, [r0], #4
-	ldmgeia	r1!, {r3, r12}
-	stmgeia	r0!, {r3, r12}
+	ldmiage	r1!, {r3, r12}
+	stmiage	r0!, {r3, r12}
 	subge	r2, r2, #4
 
 .Lmemmove_fl4:
 	/* less than 4 bytes to go */
 	adds	r2, r2, #4
-	ldmeqia	sp!, {r0, pc}		/* done */
+	ldmiaeq	sp!, {r0, pc}		/* done */
 
 	/* copy the crud byte at a time */
 	cmp	r2, #2
 	ldrb	r3, [r1], #1
 	strb	r3, [r0], #1
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0], #1
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0], #1
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0], #1
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0], #1
 	ldmia	sp!, {r0, pc}
 
 	/* erg - unaligned destination */
@@ -485,10 +488,10 @@ EENTRY(memmove)
 	/* align destination with byte copies */
 	ldrb	r3, [r1], #1
 	strb	r3, [r0], #1
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0], #1
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0], #1
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0], #1
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0], #1
 	subs	r2, r2, r12
 	blt	.Lmemmove_fl4		/* less the 4 bytes */
 
@@ -703,12 +706,12 @@ EENTRY(memmove)
 
 .Lmemmove_bl32:
 	cmn	r2, #0x10
-	ldmgedb	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
-	stmgedb	r0!, {r3, r4, r12, lr}
+	ldmdbge	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
+	stmdbge	r0!, {r3, r4, r12, lr}
 	subge	r2, r2, #0x10
 	adds	r2, r2, #0x14
-	ldmgedb	r1!, {r3, r12, lr}	/* blat a remaining 12 bytes */
-	stmgedb	r0!, {r3, r12, lr}
+	ldmdbge	r1!, {r3, r12, lr}	/* blat a remaining 12 bytes */
+	stmdbge	r0!, {r3, r12, lr}
 	subge	r2, r2, #0x0c
 	ldmia	sp!, {r4, lr}
 
@@ -718,8 +721,8 @@ EENTRY(memmove)
 	subs	r2, r2, #4
 	ldrlt	r3, [r1, #-4]!
 	strlt	r3, [r0, #-4]!
-	ldmgedb	r1!, {r3, r12}
-	stmgedb	r0!, {r3, r12}
+	ldmdbge	r1!, {r3, r12}
+	stmdbge	r0!, {r3, r12}
 	subge	r2, r2, #4
 
 .Lmemmove_bl4:
@@ -731,10 +734,10 @@ EENTRY(memmove)
 	cmp	r2, #2
 	ldrb	r3, [r1, #-1]!
 	strb	r3, [r0, #-1]!
-	ldrgeb	r3, [r1, #-1]!
-	strgeb	r3, [r0, #-1]!
-	ldrgtb	r3, [r1, #-1]!
-	strgtb	r3, [r0, #-1]!
+	ldrbge	r3, [r1, #-1]!
+	strbge	r3, [r0, #-1]!
+	ldrbgt	r3, [r1, #-1]!
+	strbgt	r3, [r0, #-1]!
 	RET
 
 	/* erg - unaligned destination */
@@ -744,10 +747,10 @@ EENTRY(memmove)
 	/* align destination with byte copies */
 	ldrb	r3, [r1, #-1]!
 	strb	r3, [r0, #-1]!
-	ldrgeb	r3, [r1, #-1]!
-	strgeb	r3, [r0, #-1]!
-	ldrgtb	r3, [r1, #-1]!
-	strgtb	r3, [r0, #-1]!
+	ldrbge	r3, [r1, #-1]!
+	strbge	r3, [r0, #-1]!
+	ldrbgt	r3, [r1, #-1]!
+	strbgt	r3, [r0, #-1]!
 	subs	r2, r2, r12
 	blt	.Lmemmove_bl4		/* less than 4 bytes to go */
 	ands	r12, r1, #3
@@ -994,8 +997,8 @@ ENTRY(memcpy)
 	bge	.Lmemcpy_loop32
 
 	cmn	r2, #0x10
-	ldmgeia	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
-	stmgeia	r0!, {r3, r4, r12, lr}
+	ldmiage	r1!, {r3, r4, r12, lr}	/* blat a remaining 16 bytes */
+	stmiage	r0!, {r3, r4, r12, lr}
 	subge	r2, r2, #0x10
 	ldmia	sp!, {r4}		/* return r4 */
 
@@ -1004,9 +1007,9 @@ ENTRY(memcpy)
 
 	/* blat 12 bytes at a time */
 .Lmemcpy_loop12:
-	ldmgeia	r1!, {r3, r12, lr}
-	stmgeia	r0!, {r3, r12, lr}
-	subges	r2, r2, #0x0c
+	ldmiage	r1!, {r3, r12, lr}
+	stmiage	r0!, {r3, r12, lr}
+	subsge	r2, r2, #0x0c
 	bge	.Lmemcpy_loop12
 
 .Lmemcpy_l12:
@@ -1016,26 +1019,26 @@ ENTRY(memcpy)
 	subs	r2, r2, #4
 	ldrlt	r3, [r1], #4
 	strlt	r3, [r0], #4
-	ldmgeia	r1!, {r3, r12}
-	stmgeia	r0!, {r3, r12}
+	ldmiage	r1!, {r3, r12}
+	stmiage	r0!, {r3, r12}
 	subge	r2, r2, #4
 
 .Lmemcpy_l4:
 	/* less than 4 bytes to go */
 	adds	r2, r2, #4
 #ifdef __APCS_26_
-	ldmeqia sp!, {r0, pc}^		/* done */
+	ldmiaeq sp!, {r0, pc}^		/* done */
 #else
-	ldmeqia	sp!, {r0, pc}		/* done */
+	ldmiaeq	sp!, {r0, pc}		/* done */
 #endif
 	/* copy the crud byte at a time */
 	cmp	r2, #2
 	ldrb	r3, [r1], #1
 	strb	r3, [r0], #1
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0], #1
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0], #1
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0], #1
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0], #1
 	ldmia	sp!, {r0, pc}
 
 	/* erg - unaligned destination */
@@ -1046,10 +1049,10 @@ ENTRY(memcpy)
 	/* align destination with byte copies */
 	ldrb	r3, [r1], #1
 	strb	r3, [r0], #1
-	ldrgeb	r3, [r1], #1
-	strgeb	r3, [r0], #1
-	ldrgtb	r3, [r1], #1
-	strgtb	r3, [r0], #1
+	ldrbge	r3, [r1], #1
+	strbge	r3, [r0], #1
+	ldrbgt	r3, [r1], #1
+	strbgt	r3, [r0], #1
 	subs	r2, r2, r12
 	blt	.Lmemcpy_l4		/* less the 4 bytes */
 
@@ -1212,12 +1215,12 @@ ENTRY(memcpy)
 	ldrb	ip, [r1], #0x01
 	sub	r2, r2, #0x01
 	strb	ip, [r3], #0x01
-	ldrleb	ip, [r1], #0x01
+	ldrble	ip, [r1], #0x01
 	suble	r2, r2, #0x01
-	strleb	ip, [r3], #0x01
-	ldrltb	ip, [r1], #0x01
+	strble	ip, [r3], #0x01
+	ldrblt	ip, [r1], #0x01
 	sublt	r2, r2, #0x01
-	strltb	ip, [r3], #0x01
+	strblt	ip, [r3], #0x01
 
 	/* Destination buffer is now word aligned */
 .Lmemcpy_wordaligned:
@@ -1294,7 +1297,7 @@ ENTRY(memcpy)
 
 .Lmemcpy_w_lessthan128:
 	adds	r2, r2, #0x80		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq			/* Return now if done */
 	subs	r2, r2, #0x20
 	blt	.Lmemcpy_w_lessthan32
@@ -1319,7 +1322,7 @@ ENTRY(memcpy)
 
 .Lmemcpy_w_lessthan32:
 	adds	r2, r2, #0x20		/* Adjust for extra sub */
-	ldmeqfd	sp!, {r4-r9}
+	ldmfdeq	sp!, {r4-r9}
 	RETeq			/* Return now if done */
 
 	and	r4, r2, #0x18
@@ -1355,12 +1358,14 @@ ENTRY(memcpy)
 	addlt	r2, r2, #0x04
 	ldrb	ip, [r1], #0x01
 	cmp	r2, #0x02
-	ldrgeb	r2, [r1], #0x01
+	ldrbge	r2, [r1], #0x01
 	strb	ip, [r3], #0x01
-	ldrgtb	ip, [r1]
-	strgeb	r2, [r3], #0x01
-	strgtb	ip, [r3]
+	ldrbgt	ip, [r1]
+	strbge	r2, [r3], #0x01
+	strbgt	ip, [r3]
 	RET
+/* Place a literal pool here for the above ldr instructions to use */
+.ltorg
 
 
 /*
@@ -1413,7 +1418,7 @@ ENTRY(memcpy)
 	bge	.Lmemcpy_bad1_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq			/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r1, r1, #0x03
@@ -1474,7 +1479,7 @@ ENTRY(memcpy)
 	bge	.Lmemcpy_bad2_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq			/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r1, r1, #0x02
@@ -1535,7 +1540,7 @@ ENTRY(memcpy)
 	bge	.Lmemcpy_bad3_loop16
 
 	adds	r2, r2, #0x10
-	ldmeqfd	sp!, {r4-r7}
+	ldmfdeq	sp!, {r4-r7}
 	RETeq			/* Return now if done */
 	subs	r2, r2, #0x04
 	sublt	r1, r1, #0x01
@@ -1564,11 +1569,11 @@ ENTRY(memcpy)
 	RETeq
 	ldrb	ip, [r1], #0x01
 	cmp	r2, #0x02
-	ldrgeb	r2, [r1], #0x01
+	ldrbge	r2, [r1], #0x01
 	strb	ip, [r3], #0x01
-	ldrgtb	ip, [r1]
-	strgeb	r2, [r3], #0x01
-	strgtb	ip, [r3]
+	ldrbgt	ip, [r1]
+	strbge	r2, [r3], #0x01
+	strbgt	ip, [r3]
 	RET
 
 
@@ -1598,7 +1603,7 @@ ENTRY(memcpy)
 	ldrb	ip, [r1], #0x01
 1:	subs	r2, r2, #0x01
 	strb	ip, [r3], #0x01
-	ldrneb	ip, [r1], #0x01
+	ldrbne	ip, [r1], #0x01
 	bne	1b
 	RET
 
diff --git a/sys/arm/include/asm.h b/sys/arm/include/asm.h
index 231d559..e5f06e7 100644
--- a/sys/arm/include/asm.h
+++ b/sys/arm/include/asm.h
@@ -64,6 +64,18 @@
 #endif
 
 /*
+ * EENTRY()/EEND() mark "extra" entry/exit points from a function.
+ * The unwind info cannot handle the concept of a nested function, or a function
+ * with multiple .fnstart directives, but some of our assembler code is written
+ * with multiple labels to allow entry at several points.  The EENTRY() macro
+ * defines such an extra entry point without a new .fnstart, so that it's
+ * basically just a label that you can jump to.  The EEND() macro does nothing
+ * at all, except document the exit point associated with the same-named entry.
+ */
+#define	_EENTRY(x) 	.globl x; .type x,_ASM_TYPE_FUNCTION; x:
+#define	_EEND(x)	/* nothing */
+
+/*
  * gas/arm uses @ as a single comment character and thus cannot be used here
  * Instead it recognised the # instead of an @ symbols in .type directives
  * We define a couple of macros so that assembly code will not be dependent
@@ -72,22 +84,10 @@
 #define _ASM_TYPE_FUNCTION	#function
 #define _ASM_TYPE_OBJECT	#object
 #define GLOBAL(X) .globl x
-#define _ENTRY(x) \
-	.text; _ALIGN_TEXT; .globl x; .type x,_ASM_TYPE_FUNCTION; x: _FNSTART
+#define	_ENTRY(x) \
+	.text; _ALIGN_TEXT; _EENTRY(x) _FNSTART
 #define	_END(x)	.size x, . - x; _FNEND
 
-/*
- * EENTRY()/EEND() mark "extra" entry/exit points from a function.
- * The unwind info cannot handle the concept of a nested function, or a function
- * with multiple .fnstart directives, but some of our assembler code is written
- * with multiple labels to allow entry at several points.  The EENTRY() macro
- * defines such an extra entry point without a new .fnstart, so that it's
- * basically just a label that you can jump to.  The EEND() macro does nothing
- * at all, except document the exit point associated with the same-named entry.
- */
-#define _EENTRY(x) 	.globl x; .type x,_ASM_TYPE_FUNCTION; x:
-#define _EEND(x)	/* nothing */
-
 #ifdef GPROF
 #  define _PROF_PROLOGUE	\
 	mov ip, lr; bl __mcount
diff --git a/sys/arm/ti/ti_smc.S b/sys/arm/ti/ti_smc.S
index 208e3a0..931ad52 100644
--- a/sys/arm/ti/ti_smc.S
+++ b/sys/arm/ti/ti_smc.S
@@ -26,7 +26,7 @@
 #include <machine/asm.h>
 __FBSDID("$FreeBSD$");
 
-	.arch armv7a
+	.cpu cortex-a8
 	.arch_extension sec
 
 /* Issue a smc #0 call */
diff --git a/sys/arm/xscale/ixp425/ixp425_a4x_io.S b/sys/arm/xscale/ixp425/ixp425_a4x_io.S
index 22ab1b3..bccd19b 100644
--- a/sys/arm/xscale/ixp425/ixp425_a4x_io.S
+++ b/sys/arm/xscale/ixp425/ixp425_a4x_io.S
@@ -56,6 +56,7 @@ ENTRY(a4x_bs_r_1)
 	ldr	r0, [r1, r2, LSL #2]
 	and	r0, r0, #0xff
 	mov	pc, lr
+END(a4x_bs_r_1)
 
 ENTRY(a4x_bs_r_2)
 	ldr	r0, [r1, r2, LSL #2]
@@ -63,10 +64,12 @@ ENTRY(a4x_bs_r_2)
 	orr	r1, r1, r1, lsl #8
 	and	r0, r0, r1
 	mov	pc, lr
+END(a4x_bs_r_2)
 
 ENTRY(a4x_bs_r_4)
 	ldr	r0, [r1, r2, LSL #2]
 	mov	pc, lr
+END(a4x_bs_r_4)
 
 /*
  * Write single
@@ -75,6 +78,7 @@ ENTRY(a4x_bs_w_1)
 	and	r3, r3, #0xff
 	str	r3, [r1, r2, LSL #2]
 	mov	pc, lr
+END(a4x_bs_w_1)
 
 ENTRY(a4x_bs_w_2)
 	mov	r0, #0xff
@@ -82,10 +86,12 @@ ENTRY(a4x_bs_w_2)
 	and	r3, r3, r0
 	str	r3, [r1, r2, LSL #2]
 	mov	pc, lr
+END(a4x_bs_w_2)
 
 ENTRY(a4x_bs_w_4)
 	str	r3, [r1, r2, LSL #2]
 	mov	pc, lr
+END(a4x_bs_w_4)
 
 /*
  * Read multiple
@@ -101,6 +107,7 @@ ENTRY(a4x_bs_rm_1)
 	strb	r3, [r1], #1
 	bne	1b
 	mov	pc, lr
+END(a4x_bs_rm_1)
 
 ENTRY(a4x_bs_rm_2)
 	add	r0, r1, r2, lsl #2
@@ -113,6 +120,7 @@ ENTRY(a4x_bs_rm_2)
 	strh	r3, [r1], #2
 	bne	1b
 	mov	pc, lr
+END(a4x_bs_rm_2)
 
 /*
  * Write multiple
@@ -128,6 +136,7 @@ ENTRY(a4x_bs_wm_1)
 	str	r3, [r0]
 	bne	1b
 	mov	pc, lr
+END(a4x_bs_wm_1)
 
 ENTRY(a4x_bs_wm_2)
 	add	r0, r1, r2, lsl #2
@@ -140,3 +149,4 @@ ENTRY(a4x_bs_wm_2)
 	str	r3, [r0]
 	bne	1b
 	mov	pc, lr
+END(a4x_bs_wm_2)
diff --git a/sys/arm/xscale/ixp425/ixp425_pci_asm.S b/sys/arm/xscale/ixp425/ixp425_pci_asm.S
index f686d65..ce668a3 100644
--- a/sys/arm/xscale/ixp425/ixp425_pci_asm.S
+++ b/sys/arm/xscale/ixp425/ixp425_pci_asm.S
@@ -58,6 +58,7 @@ ENTRY(ixp425_pci_mem_bs_r_1)
 	ldrb	r0, [r1, r2]
 #endif /* __ARMEB__ */
 	mov	pc, lr
+END(ixp425_pci_mem_bs_r_1)
 
 ENTRY(ixp425_pci_mem_bs_r_2)
 #ifdef __ARMEB__
@@ -68,10 +69,12 @@ ENTRY(ixp425_pci_mem_bs_r_2)
 	ldrh	r0, [r1, r2]
 #endif /* __ARMEB__ */
 	mov	pc, lr
+END(ixp425_pci_mem_bs_r_2)
 
 ENTRY(ixp425_pci_mem_bs_r_4)
 	ldr	r0, [r1, r2]
 	mov	pc, lr
+END(ixp425_pci_mem_bs_r_4)
 
 /*
  * write single
@@ -86,6 +89,7 @@ ENTRY(ixp425_pci_mem_bs_w_1)
 	strb	r3, [r1, r2]
 #endif /* __ARMEB__ */
 	mov	pc, lr
+END(ixp425_pci_mem_bs_w_1)
 
 ENTRY(ixp425_pci_mem_bs_w_2)
 #ifdef __ARMEB__
@@ -96,7 +100,9 @@ ENTRY(ixp425_pci_mem_bs_w_2)
 	strh	r3, [r1, r2]
 #endif /* __ARMEB__ */
 	mov	pc, lr
+END(ixp425_pci_mem_bs_w_2)
 
 ENTRY(ixp425_pci_mem_bs_w_4)
 	str	r3, [r1, r2]
 	mov	pc, lr
+END(ixp425_pci_mem_bs_w_4)