Remove sfi_* annotations from ARM assembly files.

This semi-mechanical patch removes all uses and definitions of the
sfi_breg, sfi_pld, and sfi_sp macros from various ARM-specific
assembly files.  These were only used by NaCl.

	* sysdeps/arm/sysdep.h
        (ARM_SFI_MACROS, sfi_breg, sfi_pld, sfi_sp): Delete definitions.

	* sysdeps/arm/__longjmp.S, sysdeps/arm/add_n.S
	* sysdeps/arm/addmul_1.S, sysdeps/arm/arm-mcount.S
	* sysdeps/arm/armv6/rawmemchr.S, sysdeps/arm/armv6/strchr.S
	* sysdeps/arm/armv6/strcpy.S, sysdeps/arm/armv6/strlen.S
	* sysdeps/arm/armv6/strrchr.S, sysdeps/arm/armv6t2/memchr.S
	* sysdeps/arm/armv6t2/strlen.S
	* sysdeps/arm/armv7/multiarch/memcpy_impl.S
	* sysdeps/arm/armv7/strcmp.S, sysdeps/arm/dl-tlsdesc.S
	* sysdeps/arm/memcpy.S, sysdeps/arm/memmove.S
	* sysdeps/arm/memset.S, sysdeps/arm/setjmp.S
	* sysdeps/arm/strlen.S, sysdeps/arm/submul_1.S:
        Remove all uses of sfi_breg, sfi_pld, and sfi_sp.

22 files changed, 460 insertions, 882 deletions

diff --git a/ChangeLog b/ChangeLog
index 704968e4a5..97b80aa23e 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,23 @@
 2017-05-20  Zack Weinberg  <zackw@panix.com>
 
+	* sysdeps/arm/sysdep.h
+	(ARM_SFI_MACROS, sfi_breg, sfi_pld, sfi_sp): Delete definitions.
+
+	* sysdeps/arm/__longjmp.S, sysdeps/arm/add_n.S
+	* sysdeps/arm/addmul_1.S, sysdeps/arm/arm-mcount.S
+	* sysdeps/arm/armv6/rawmemchr.S, sysdeps/arm/armv6/strchr.S
+	* sysdeps/arm/armv6/strcpy.S, sysdeps/arm/armv6/strlen.S
+	* sysdeps/arm/armv6/strrchr.S, sysdeps/arm/armv6t2/memchr.S
+	* sysdeps/arm/armv6t2/strlen.S
+	* sysdeps/arm/armv7/multiarch/memcpy_impl.S
+	* sysdeps/arm/armv7/strcmp.S, sysdeps/arm/dl-tlsdesc.S
+	* sysdeps/arm/memcpy.S, sysdeps/arm/memmove.S
+	* sysdeps/arm/memset.S, sysdeps/arm/setjmp.S
+	* sysdeps/arm/strlen.S, sysdeps/arm/submul_1.S:
+	Remove all uses of sfi_breg, sfi_pld, and sfi_sp.
+
+2017-05-20  Zack Weinberg  <zackw@panix.com>
+
 	Remove the NaCl port.
 	* sysdeps/arm/nacl: Remove directory and contents.
 	* sysdeps/nacl: Remove directory and contents.
diff --git a/sysdeps/arm/__longjmp.S b/sysdeps/arm/__longjmp.S
index 2a567b5bfa..5202c728bc 100644
--- a/sysdeps/arm/__longjmp.S
+++ b/sysdeps/arm/__longjmp.S
@@ -28,8 +28,7 @@ ENTRY (__longjmp)
 	mov	ip, r0
 
 #ifdef CHECK_SP
-	sfi_breg ip, \
-	ldr	r4, [\B]	/* jmpbuf's sp */
+	ldr	r4, [ip]	/* jmpbuf's sp */
 	cfi_undefined (r4)
 #ifdef PTR_DEMANGLE
 	PTR_DEMANGLE (r4, r4, a3, a4)
@@ -38,28 +37,22 @@ ENTRY (__longjmp)
 #endif
 
 #ifdef PTR_DEMANGLE
-	sfi_breg ip, \
-	ldr	a4, [\B], #4
+	ldr	a4, [ip], #4
 	PTR_DEMANGLE (a4, a4, a3, r4)
 	cfi_undefined (r4)
-	sfi_breg ip, \
-	ldr	r4, [\B], #4
+	ldr	r4, [ip], #4
 	PTR_DEMANGLE2 (r4, r4, a3)
 #else
-	sfi_breg ip, \
-	ldr	a4, [\B], #4
-	sfi_breg ip, \
-	ldr	r4, [\B], #4
+	ldr	a4, [ip], #4
+	ldr	r4, [ip], #4
 	cfi_undefined (r4)
 #endif
 	/* longjmp probe expects longjmp first argument (4@r0), second
 	   argument (-4@r1), and target address (4@r4), respectively.  */
 	LIBC_PROBE (longjmp, 3, 4@r0, -4@r1, 4@r4)
-	sfi_sp \
 	mov	sp, a4
 	mov	lr, r4
-	sfi_breg ip, \
-	ldmia	\B!, JMP_BUF_REGLIST
+	ldmia	ip!, JMP_BUF_REGLIST
 	cfi_restore (v1)
 	cfi_restore (v2)
 	cfi_restore (v3)
@@ -97,8 +90,7 @@ ENTRY (__longjmp)
 
 	/* Restore the VFP registers.  */
 	/* Following instruction is vldmia ip!, {d8-d15}.  */
-	sfi_breg r12, \
-	ldc	p11, cr8, [\B], #64
+	ldc	p11, cr8, [r12], #64
 .Lno_vfp:
 
 #ifndef ARM_ASSUME_NO_IWMMXT
@@ -107,18 +99,12 @@ ENTRY (__longjmp)
 
 	/* Restore the call-preserved iWMMXt registers.  */
 	/* Following instructions are wldrd wr10, [ip], #8 (etc.)  */
-	sfi_breg r12, \
-	ldcl	p1, cr10, [\B], #8
-	sfi_breg r12, \
-	ldcl	p1, cr11, [\B], #8
-	sfi_breg r12, \
-	ldcl	p1, cr12, [\B], #8
-	sfi_breg r12, \
-	ldcl	p1, cr13, [\B], #8
-	sfi_breg r12, \
-	ldcl	p1, cr14, [\B], #8
-	sfi_breg r12, \
-	ldcl	p1, cr15, [\B], #8
+	ldcl	p1, cr10, [r12], #8
+	ldcl	p1, cr11, [r12], #8
+	ldcl	p1, cr12, [r12], #8
+	ldcl	p1, cr13, [r12], #8
+	ldcl	p1, cr14, [r12], #8
+	ldcl	p1, cr15, [r12], #8
 .Lno_iwmmxt:
 #endif
 
diff --git a/sysdeps/arm/add_n.S b/sysdeps/arm/add_n.S
index b601357e64..811a769959 100644
--- a/sysdeps/arm/add_n.S
+++ b/sysdeps/arm/add_n.S
@@ -52,40 +52,31 @@ ENTRY (FUNC)
 	add	lr, r1, r3, lsl #2	/* compute end src1 */
 	beq	1f
 
-	sfi_breg r1, \
-	ldr	r4, [\B], #4		/* do one to make count even */
-	sfi_breg r2, \
-	ldr	r5, [\B], #4
+	ldr	r4, [r1], #4		/* do one to make count even */
+	ldr	r5, [r2], #4
 	OPC	r4, r4, r5
 	teq	r1, lr			/* end of count? (preserve carry) */
-	sfi_breg r0, \
-	str	r4, [\B], #4
+	str	r4, [r0], #4
 	beq	9f
 1:
 	tst	r3, #2			/* count & 2 == 2?  */
 	beq	2f
-	sfi_breg r1, \
-	ldm	\B!, { r4, r5 }		/* do two to make count 0 mod 4 */
-	sfi_breg r2, \
-	ldm	\B!, { r6, r7 }
+	ldm	r1!, { r4, r5 }		/* do two to make count 0 mod 4 */
+	ldm	r2!, { r6, r7 }
 	OPC	r4, r4, r6
 	OPC	r5, r5, r7
 	teq	r1, lr			/* end of count? */
-	sfi_breg r0, \
-	stm	\B!, { r4, r5 }
+	stm	r0!, { r4, r5 }
 	beq	9f
 2:
-	sfi_breg r1, \
-	ldm	\B!, { r3, r5, r7, r10 }	/* do four each loop */
-	sfi_breg r2, \
-	ldm	\B!, { r4, r6, r8, ip }
+	ldm	r1!, { r3, r5, r7, r10 }	/* do four each loop */
+	ldm	r2!, { r4, r6, r8, ip }
 	OPC	r3, r3, r4
 	OPC	r5, r5, r6
 	OPC	r7, r7, r8
 	OPC	r10, r10, ip
 	teq	r1, lr
-	sfi_breg r0, \
-	stm	\B!, { r3, r5, r7, r10 }
+	stm	r0!, { r3, r5, r7, r10 }
 	bne	2b
 
 9:
diff --git a/sysdeps/arm/addmul_1.S b/sysdeps/arm/addmul_1.S
index a681c242ef..c4e54f8723 100644
--- a/sysdeps/arm/addmul_1.S
+++ b/sysdeps/arm/addmul_1.S
@@ -37,21 +37,16 @@ ENTRY (__mpn_addmul_1)
 	cfi_rel_offset (r6, 8)
 	cfi_rel_offset (r7, 12)
 
-	sfi_breg r1, \
-	ldr	r6, [\B], #4
-	sfi_breg r0, \
-	ldr	r5, [\B]
+	ldr	r6, [r1], #4
+	ldr	r5, [r0]
 	mov	r4, #0			/* init carry in */
 	b	1f
 0:
-	sfi_breg r1, \
-	ldr	r6, [\B], #4		/* load next ul */
+	ldr	r6, [r1], #4		/* load next ul */
 	adds	r7, r4, r5		/* (out, c) = cl + lpl */
-	sfi_breg r0, \
-	ldr	r5, [\B, #4]		/* load next rl */
+	ldr	r5, [r0, #4]		/* load next rl */
 	adc	r4, ip, #0		/* cl = hpl + c */
-	sfi_breg r0, \
-	str	r7, [\B], #4
+	str	r7, [r0], #4
 1:
 	mov	ip, #0			/* zero-extend rl */
 	umlal	r5, ip, r6, r3		/* (hpl, lpl) = ul * vl + rl */
@@ -59,8 +54,7 @@ ENTRY (__mpn_addmul_1)
 	bne	0b
 
 	adds	r4, r4, r5		/* (out, c) = cl + llpl */
-	sfi_breg r0, \
-	str	r4, [\B]
+	str	r4, [r0]
 	adc	r0, ip, #0		/* return hpl + c */
 
 	pop	{ r4, r5, r6, r7 }
diff --git a/sysdeps/arm/arm-mcount.S b/sysdeps/arm/arm-mcount.S
index b08a8dd0d8..df2601b4a6 100644
--- a/sysdeps/arm/arm-mcount.S
+++ b/sysdeps/arm/arm-mcount.S
@@ -90,8 +90,7 @@ ENTRY(__mcount_arm_compat)
 	cfi_rel_offset (lr, 20)
 	movs r0, fp
 	ittt ne
-	sfi_breg r0, \
-	ldrne r0, [\B, #-4]
+	ldrne r0, [r0, #-4]
 	movsne r1, lr
 	blne __mcount_internal
 # if defined (__ARM_ARCH_4T__) && defined (__THUMB_INTERWORK__)
diff --git a/sysdeps/arm/armv6/rawmemchr.S b/sysdeps/arm/armv6/rawmemchr.S
index b4d4eb2e39..2511f105bd 100644
--- a/sysdeps/arm/armv6/rawmemchr.S
+++ b/sysdeps/arm/armv6/rawmemchr.S
@@ -25,8 +25,7 @@ ENTRY (__rawmemchr)
 	@ r0 = start of string
 	@ r1 = character to match
 	@ returns a pointer to the match, which must be present.
-	sfi_breg r0, \
-	ldrb	r2, [\B]		@ load first byte asap
+	ldrb	r2, [r0]		@ load first byte asap
 
 	@ To cater to long strings, we want to search through a few
 	@ characters until we reach an aligned pointer.  To cater to
@@ -42,8 +41,7 @@ ENTRY (__rawmemchr)
 	bxeq	lr
 
 	@ Loop until we find ...
-1:	sfi_breg r0, \
-	ldrb	r2, [\B, #1]!
+1:	ldrb	r2, [r0, #1]!
 	subs	r3, r3, #1		@ ... the alignment point
 	it	ne
 	cmpne	r2, r1			@ ... or C
@@ -56,16 +54,15 @@ ENTRY (__rawmemchr)
 	add	r0, r0, #1
 
 	@ So now we're aligned.
-	sfi_breg r0, \
-	ldrd	r2, r3, [\B], #8
+	ldrd	r2, r3, [r0], #8
 	orr	r1, r1, r1, lsl #8	@ Replicate C to all bytes
 #ifdef ARCH_HAS_T2
 	movw	ip, #0x0101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 	movt	ip, #0x0101
 #else
 	ldr	ip, =0x01010101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 #endif
 	orr	r1, r1, r1, lsl #16
 
@@ -77,11 +74,10 @@ ENTRY (__rawmemchr)
 	eor	r3, r3, r1
 	uqsub8	r2, ip, r2		@ Find C
 	uqsub8	r3, ip, r3
-	sfi_pld	r0, #128
+	pld	[r0, #128]
 	orrs	r3, r3, r2		@ Test both words for found
 	it	eq
-	sfi_breg r0, \
-	ldrdeq	r2, r3, [\B], #8
+	ldrdeq	r2, r3, [r0], #8
 	beq	2b
 
 	@ Found something.  Disambiguate between first and second words.
diff --git a/sysdeps/arm/armv6/strchr.S b/sysdeps/arm/armv6/strchr.S
index f6fd192144..bfd0a6b237 100644
--- a/sysdeps/arm/armv6/strchr.S
+++ b/sysdeps/arm/armv6/strchr.S
@@ -25,8 +25,7 @@ ENTRY (strchr)
 	@ r0 = start of string
 	@ r1 = character to match
 	@ returns NULL for no match, or a pointer to the match
-	sfi_breg r0, \
-	ldrb	r2, [\B]		@ load the first byte asap
+	ldrb	r2, [r0]		@ load the first byte asap
 	uxtb	r1, r1
 
 	@ To cater to long strings, we want to search through a few
@@ -43,8 +42,7 @@ ENTRY (strchr)
 	beq	99f
 
 	@ Loop until we find ...
-1:	sfi_breg r0, \
-	ldrb	r2, [\B, #1]!
+1:	ldrb	r2, [r0, #1]!
 	subs	r3, r3, #1		@ ... the aligment point
 	it	ne
 	cmpne	r2, r1			@ ... or the character
@@ -67,16 +65,15 @@ ENTRY (strchr)
 	cfi_rel_offset (r6, 8)
 	cfi_rel_offset (r7, 12)
 
-	sfi_breg r0, \
-	ldrd	r2, r3, [\B], #8
+	ldrd	r2, r3, [r0], #8
 	orr	r1, r1, r1, lsl #8	@ Replicate C to all bytes
 #ifdef ARCH_HAS_T2
 	movw	ip, #0x0101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 	movt	ip, #0x0101
 #else
 	ldr	ip, =0x01010101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 #endif
 	orr	r1, r1, r1, lsl #16
 
@@ -90,14 +87,13 @@ ENTRY (strchr)
 	uqsub8	r5, ip, r3
 	eor	r7, r3, r1
 	uqsub8	r6, ip, r6		@ Find C
-	sfi_pld	r0, #128		@ Prefetch 2 lines ahead
+	pld	[r0, #128]		@ Prefetch 2 lines ahead
 	uqsub8	r7, ip, r7
 	orr	r4, r4, r6		@ Combine found for EOS and C
 	orr	r5, r5, r7
 	orrs	r6, r4, r5		@ Combine the two words
 	it	eq
-	sfi_breg r0, \
-	ldrdeq	r2, r3, [\B], #8
+	ldrdeq	r2, r3, [r0], #8
 	beq	2b
 
 	@ Found something.  Disambiguate between first and second words.
diff --git a/sysdeps/arm/armv6/strcpy.S b/sysdeps/arm/armv6/strcpy.S
index 239418c658..1b98dbce30 100644
--- a/sysdeps/arm/armv6/strcpy.S
+++ b/sysdeps/arm/armv6/strcpy.S
@@ -44,8 +44,8 @@ ENTRY (strcpy)
 	@ Signal strcpy with DEST in IP.
 	mov	ip, r0
 0:
-	sfi_pld	r0
-	sfi_pld	r1
+	pld	[r0, #0]
+	pld	[r1, #0]
 
 	@ To cater to long strings, we want 8 byte alignment in the source.
 	@ To cater to small strings, we don't want to start that right away.
@@ -54,11 +54,9 @@ ENTRY (strcpy)
 	rsb	r3, r3, #16
 
 	@ Loop until we find ...
-1:	sfi_breg r1, \
-	ldrb	r2, [\B], #1
+1:	ldrb	r2, [r1], #1
 	subs	r3, r3, #1		@ ... the alignment point
-	sfi_breg r0, \
-	strb	r2, [\B], #1
+	strb	r2, [r0], #1
 	it	ne
 	cmpne	r2, #0			@ ... or EOS
 	bne	1b
@@ -68,10 +66,9 @@ ENTRY (strcpy)
 	beq	.Lreturn
 
 	@ Load the next two words asap
-	sfi_breg r1, \
-	ldrd	r2, r3, [\B], #8
-	sfi_pld	r0, #64
-	sfi_pld	r1, #64
+	ldrd	r2, r3, [r1], #8
+	pld	[r0, #64]
+	pld	[r1, #64]
 
 	@ For longer strings, we actaully need a stack frame.
 	push	{ r4, r5, r6, r7 }
@@ -99,18 +96,15 @@ ENTRY (strcpy)
 	.balign	16
 2:	uqsub8	r4, r7, r2		@ Find EOS
 	uqsub8	r5, r7, r3
-	sfi_pld	r1, #128
+	pld	[r1, #128]
 	cmp	r4, #0			@ EOS in first word?
-	sfi_pld	r0, #128
+	pld	[r0, #128]
 	bne	3f
-	sfi_breg r0, \
-	str	r2, [\B], #4
+	str	r2, [r0], #4
 	cmp	r5, #0			@ EOS in second word?
 	bne	4f
-	sfi_breg r0, \
-	str	r3, [\B], #4
-	sfi_breg r1, \
-	ldrd	r2, r3, [\B], #8
+	str	r3, [r0], #4
+	ldrd	r2, r3, [r1], #8
 	b	2b
 
 3:	sub	r1, r1, #4		@ backup to first word
@@ -120,11 +114,9 @@ ENTRY (strcpy)
 	@ Note that we generally back up and re-read source bytes,
 	@ but we'll not re-write dest bytes.
 .Lbyte_loop:
-	sfi_breg r1, \
-	ldrb	r2, [\B], #1
+	ldrb	r2, [r1], #1
 	cmp	r2, #0
-	sfi_breg r0, \
-	strb	r2, [\B], #1
+	strb	r2, [r0], #1
 	bne	.Lbyte_loop
 
 	pop	{ r4, r5, r6, r7 }
@@ -169,8 +161,7 @@ ENTRY (strcpy)
 	@ Store a few bytes from the first word.
 	@ At the same time we align r0 and shift out bytes from r2.
 .rept	4-\unalign
-	sfi_breg r0, \
-	strb	r2, [\B], #1
+	strb	r2, [r0], #1
 	lsr	r2, r2, #8
 .endr
 #ifdef __ARMEB__
@@ -185,23 +176,20 @@ ENTRY (strcpy)
 	orr	r2, r2, r3, lsh_gt #(\unalign*8)
 	@ Save leftover bytes from the two words
 	lsh_ls	r6, r3, #((4-\unalign)*8)
-	sfi_breg r0, \
-	str	r2, [\B], #4
+	str	r2, [r0], #4
 	@ The "real" start of the unaligned copy loop.
-	sfi_breg r1, \
-	ldrd	r2, r3, [\B], #8	@ Load 8 more bytes
+	ldrd	r2, r3, [r1], #8	@ Load 8 more bytes
 	uqsub8	r4, r7, r2		@ Find EOS
-	sfi_pld	r1, #128
+	pld	[r1, #128]
 	uqsub8	r5, r7, r3
-	sfi_pld	r0, #128
+	pld	[r0, #128]
 	cmp	r4, #0			@ EOS in first word?
 	bne	3f
 	@ Combine the leftover and the first word
 	orr	r6, r6, r2, lsh_gt #(\unalign*8)
 	@ Discard used bytes from the first word.
 	lsh_ls	r2, r2, #((4-\unalign)*8)
-	sfi_breg r0, \
-	str	r6, [\B], #4
+	str	r6, [r0], #4
 	b	1b
 	@ Found EOS in one of the words; adjust backward
 3:	sub	r1, r1, #4
@@ -212,8 +200,7 @@ ENTRY (strcpy)
 	rev	r2, r2
 #endif
 .rept	\unalign
-	sfi_breg r0, \
-	strb	r2, [\B], #1
+	strb	r2, [r0], #1
 	lsr	r2, r2, #8
 .endr
 	b	.Lbyte_loop
diff --git a/sysdeps/arm/armv6/strlen.S b/sysdeps/arm/armv6/strlen.S
index f807f810fb..64a971f2de 100644
--- a/sysdeps/arm/armv6/strlen.S
+++ b/sysdeps/arm/armv6/strlen.S
@@ -23,8 +23,7 @@
 
 ENTRY (strlen)
 	@ r0 = start of string
-	sfi_breg r0, \
-	ldrb	r2, [\B]		@ load the first byte asap
+	ldrb	r2, [r0]		@ load the first byte asap
 
 	@ To cater to long strings, we want to search through a few
 	@ characters until we reach an aligned pointer.  To cater to
@@ -39,8 +38,7 @@ ENTRY (strlen)
 	beq	99f
 
 	@ Loop until we find ...
-1:	sfi_breg r0, \
-	ldrb	r2, [\B, #1]!
+1:	ldrb	r2, [r0, #1]!
 	subs	r3, r3, #1		@ ... the aligment point
 	it	ne
 	cmpne	r2, #0			@ ... or EOS
@@ -52,15 +50,14 @@ ENTRY (strlen)
 	add	r0, r0, #1
 
 	@ So now we're aligned.
-	sfi_breg r0, \
-	ldrd	r2, r3, [\B], #8
+	ldrd	r2, r3, [r0], #8
 #ifdef ARCH_HAS_T2
 	movw	ip, #0x0101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 	movt	ip, #0x0101
 #else
 	ldr	ip, =0x01010101
-	sfi_pld	r0, #64
+	pld	[r0, #64]
 #endif
 
 	@ Loop searching for EOS, 8 bytes at a time.
@@ -70,11 +67,10 @@ ENTRY (strlen)
 	.balign	16
 2:	uqsub8	r2, ip, r2		@ Find EOS
 	uqsub8	r3, ip, r3
-	sfi_pld	r0, #128		@ Prefetch 2 lines ahead
+	pld	[r0, #128]		@ Prefetch 2 lines ahead
 	orrs	r3, r3, r2		@ Combine the two words
 	it	eq
-	sfi_breg r0, \
-	ldrdeq	r2, r3, [\B], #8
+	ldrdeq	r2, r3, [r0], #8
 	beq	2b
 
 	@ Found something.  Disambiguate between first and second words.
diff --git a/sysdeps/arm/armv6/strrchr.S b/sysdeps/arm/armv6/strrchr.S
index d6db032a38..e6eea01816 100644
--- a/sysdeps/arm/armv6/strrchr.S
+++ b/sysdeps/arm/armv6/strrchr.S
@@ -33,8 +33,7 @@ ENTRY (strrchr)
 	@ Loop a few times until we're aligned.
 	tst	r3, #7
 	beq	2f
-1:	sfi_breg r3, \
-	ldrb	r2, [\B], #1
+1:	ldrb	r2, [r3], #1
 	cmp	r2, r1			@ Find the character
 	it	eq
 	subeq	r0, r3, #1
@@ -65,8 +64,7 @@ ENTRY (strrchr)
 	@ Loop searching for EOS and C, 8 bytes at a time.
 	@ Any time we find a match in a word, we copy the address of
 	@ the word to r0, and the found bits to r2.
-3:	sfi_breg r3, \
-	ldrd	r4, r5, [\B], #8
+3:	ldrd	r4, r5, [r3], #8
 	@ Subtracting (unsigned saturating) from 1 means result of 1 for
 	@ any byte that was originally zero and 0 otherwise.  Therefore
 	@ we consider the lsb of each byte the "found" bit.
diff --git a/sysdeps/arm/armv6t2/memchr.S b/sysdeps/arm/armv6t2/memchr.S
index c012d73d38..fb4dc8efa3 100644
--- a/sysdeps/arm/armv6t2/memchr.S
+++ b/sysdeps/arm/armv6t2/memchr.S
@@ -65,8 +65,7 @@ ENTRY(memchr)
 
 	@ Work up to an aligned point
 5:
-	sfi_breg r0, \
-	ldrb	r3, [\B],#1
+	ldrb	r3, [r0],#1
 	subs	r2, r2, #1
 	cmp	r3, r1
 	beq	50f		@ If it matches exit found
@@ -91,8 +90,7 @@ ENTRY(memchr)
 	movs	r3, #0
 
 15:
-	sfi_breg r0, \
-	ldrd 	r4,r5, [\B],#8
+	ldrd 	r4,r5, [r0],#8
 #ifndef NO_THUMB
 	subs	r6, r6, #8
 #endif
@@ -130,8 +128,7 @@ ENTRY(memchr)
 #endif
 
 21:  @ Post aligned section, or just a short call
-	sfi_breg r0, \
-	ldrb	r3,[\B],#1
+	ldrb	r3,[r0],#1
 #ifndef NO_THUMB
 	subs	r2,r2,#1
 	eor	r3,r3,r1	@ r3 = 0 if match - doesn't break flags from sub
diff --git a/sysdeps/arm/armv6t2/strlen.S b/sysdeps/arm/armv6t2/strlen.S
index 4795c54ae5..c72a1e73be 100644
--- a/sysdeps/arm/armv6t2/strlen.S
+++ b/sysdeps/arm/armv6t2/strlen.S
@@ -67,7 +67,7 @@
 	.text
 	.p2align 6
 ENTRY(strlen)
-	sfi_pld	srcin, #0
+	pld	[srcin, #0]
 	strd	r4, r5, [sp, #-8]!
 	cfi_adjust_cfa_offset (8)
 	cfi_rel_offset (r4, 0)
@@ -76,15 +76,14 @@ ENTRY(strlen)
 	bic	src, srcin, #7
 	mvn	const_m1, #0
 	ands	tmp1, srcin, #7		/* (8 - bytes) to alignment.  */
-	sfi_pld	src, #32
+	pld	[src, #32]
 	bne.w	.Lmisaligned8
 	mov	const_0, #0
 	mov	result, #-8
 .Lloop_aligned:
 	/* Bytes 0-7.  */
-	sfi_breg src, \
-	ldrd	data1a, data1b, [\B]
-	sfi_pld	src, #64
+	ldrd	data1a, data1b, [src]
+	pld	[src, #64]
 	add	result, result, #8
 .Lstart_realigned:
 	uadd8	data1a, data1a, const_m1	/* Saturating GE<0:3> set.  */
@@ -94,8 +93,7 @@ ENTRY(strlen)
 	cbnz	data1b, .Lnull_found
 
 	/* Bytes 8-15.  */
-	sfi_breg src, \
-	ldrd	data1a, data1b, [\B, #8]
+	ldrd	data1a, data1b, [src, #8]
 	uadd8	data1a, data1a, const_m1	/* Saturating GE<0:3> set.  */
 	add	result, result, #8
 	sel	data1a, const_0, const_m1	/* Select based on GE<0:3>.  */
@@ -104,8 +102,7 @@ ENTRY(strlen)
 	cbnz	data1b, .Lnull_found
 
 	/* Bytes 16-23.  */
-	sfi_breg src, \
-	ldrd	data1a, data1b, [\B, #16]
+	ldrd	data1a, data1b, [src, #16]
 	uadd8	data1a, data1a, const_m1	/* Saturating GE<0:3> set.  */
 	add	result, result, #8
 	sel	data1a, const_0, const_m1	/* Select based on GE<0:3>.  */
@@ -114,8 +111,7 @@ ENTRY(strlen)
 	cbnz	data1b, .Lnull_found
 
 	/* Bytes 24-31.  */
-	sfi_breg src, \
-	ldrd	data1a, data1b, [\B, #24]
+	ldrd	data1a, data1b, [src, #24]
 	add	src, src, #32
 	uadd8	data1a, data1a, const_m1	/* Saturating GE<0:3> set.  */
 	add	result, result, #8
@@ -143,13 +139,12 @@ ENTRY(strlen)
 
 .Lmisaligned8:
 	cfi_restore_state
-	sfi_breg src, \
-	ldrd	data1a, data1b, [\B]
+	ldrd	data1a, data1b, [src]
 	and	tmp2, tmp1, #3
 	rsb	result, tmp1, #0
 	lsl	tmp2, tmp2, #3			/* Bytes -> bits.  */
 	tst	tmp1, #4
-	sfi_pld	src, #64
+	pld	[src, #64]
 	S2HI	tmp2, const_m1, tmp2
 #ifdef NO_THUMB
 	mvn	tmp1, tmp2
diff --git a/sysdeps/arm/armv7/multiarch/memcpy_impl.S b/sysdeps/arm/armv7/multiarch/memcpy_impl.S
index 5d5a3cefaa..c1b9fb0ab5 100644
--- a/sysdeps/arm/armv7/multiarch/memcpy_impl.S
+++ b/sysdeps/arm/armv7/multiarch/memcpy_impl.S
@@ -226,71 +226,40 @@
 
 #ifdef USE_VFP
 	.macro	cpy_line_vfp vreg, base
-	sfi_breg dst, \
-	vstr	\vreg, [\B, #\base]
-	sfi_breg src, \
-	vldr	\vreg, [\B, #\base]
-	sfi_breg dst, \
-	vstr	d0, [\B, #\base + 8]
-	sfi_breg src, \
-	vldr	d0, [\B, #\base + 8]
-	sfi_breg dst, \
-	vstr	d1, [\B, #\base + 16]
-	sfi_breg src, \
-	vldr	d1, [\B, #\base + 16]
-	sfi_breg dst, \
-	vstr	d2, [\B, #\base + 24]
-	sfi_breg src, \
-	vldr	d2, [\B, #\base + 24]
-	sfi_breg dst, \
-	vstr	\vreg, [\B, #\base + 32]
-	sfi_breg src, \
-	vldr	\vreg, [\B, #\base + prefetch_lines * 64 - 32]
-	sfi_breg dst, \
-	vstr	d0, [\B, #\base + 40]
-	sfi_breg src, \
-	vldr	d0, [\B, #\base + 40]
-	sfi_breg dst, \
-	vstr	d1, [\B, #\base + 48]
-	sfi_breg src, \
-	vldr	d1, [\B, #\base + 48]
-	sfi_breg dst, \
-	vstr	d2, [\B, #\base + 56]
-	sfi_breg src, \
-	vldr	d2, [\B, #\base + 56]
+	vstr	\vreg, [dst, #\base]
+	vldr	\vreg, [src, #\base]
+	vstr	d0, [dst, #\base + 8]
+	vldr	d0, [src, #\base + 8]
+	vstr	d1, [dst, #\base + 16]
+	vldr	d1, [src, #\base + 16]
+	vstr	d2, [dst, #\base + 24]
+	vldr	d2, [src, #\base + 24]
+	vstr	\vreg, [dst, #\base + 32]
+	vldr	\vreg, [src, #\base + prefetch_lines * 64 - 32]
+	vstr	d0, [dst, #\base + 40]
+	vldr	d0, [src, #\base + 40]
+	vstr	d1, [dst, #\base + 48]
+	vldr	d1, [src, #\base + 48]
+	vstr	d2, [dst, #\base + 56]
+	vldr	d2, [src, #\base + 56]
 	.endm
 
 	.macro	cpy_tail_vfp vreg, base
-	sfi_breg dst, \
-	vstr	\vreg, [\B, #\base]
-	sfi_breg src, \
-	vldr	\vreg, [\B, #\base]
-	sfi_breg dst, \
-	vstr	d0, [\B, #\base + 8]
-	sfi_breg src, \
-	vldr	d0, [\B, #\base + 8]
-	sfi_breg dst, \
-	vstr	d1, [\B, #\base + 16]
-	sfi_breg src, \
-	vldr	d1, [\B, #\base + 16]
-	sfi_breg dst, \
-	vstr	d2, [\B, #\base + 24]
-	sfi_breg src, \
-	vldr	d2, [\B, #\base + 24]
-	sfi_breg dst, \
-	vstr	\vreg, [\B, #\base + 32]
-	sfi_breg dst, \
-	vstr	d0, [\B, #\base + 40]
-	sfi_breg src, \
-	vldr	d0, [\B, #\base + 40]
-	sfi_breg dst, \
-	vstr	d1, [\B, #\base + 48]
-	sfi_breg src, \
-	vldr	d1, [\B, #\base + 48]
-	sfi_breg dst, \
-	vstr	d2, [\B, #\base + 56]
-	sfi_breg src, \
-	vldr	d2, [\B, #\base + 56]
+	vstr	\vreg, [dst, #\base]
+	vldr	\vreg, [src, #\base]
+	vstr	d0, [dst, #\base + 8]
+	vldr	d0, [src, #\base + 8]
+	vstr	d1, [dst, #\base + 16]
+	vldr	d1, [src, #\base + 16]
+	vstr	d2, [dst, #\base + 24]
+	vldr	d2, [src, #\base + 24]
+	vstr	\vreg, [dst, #\base + 32]
+	vstr	d0, [dst, #\base + 40]
+	vldr	d0, [src, #\base + 40]
+	vstr	d1, [dst, #\base + 48]
+	vldr	d1, [src, #\base + 48]
+	vstr	d2, [dst, #\base + 56]
+	vldr	d2, [src, #\base + 56]
 	.endm
 #endif
 
@@ -316,26 +285,16 @@ ENTRY(memcpy)
 	vst1.8	{d0}, [\reg]!
 	.endm
 
-	/* These are used by the NaCl sfi_breg macro.  */
-	.macro _sfi_breg_dmask_neon_load_d0 reg
-	_sfi_dmask \reg
-	.endm
-	.macro _sfi_breg_dmask_neon_store_d0 reg
-	_sfi_dmask \reg
-	.endm
-
 	and	tmp1, count, #0x38
 	.macro dispatch_step i
-	sfi_breg src, neon_load_d0 \B
-	sfi_breg dst, neon_store_d0 \B
+	neon_load_d0 src
+	neon_store_d0 dst
 	.endm
 	dispatch_7_dword
 
 	tst	count, #4
-	sfi_breg src, \
-	ldrne	tmp1, [\B], #4
-	sfi_breg dst, \
-	strne	tmp1, [\B], #4
+	ldrne	tmp1, [src], #4
+	strne	tmp1, [dst], #4
 #else
 	/* Copy up to 15 full words of data.  May not be aligned.  */
 	/* Cannot use VFP for unaligned data.  */
@@ -344,23 +303,17 @@ ENTRY(memcpy)
 	add	src, src, tmp1
 	/* Jump directly into the sequence below at the correct offset.  */
 	.macro dispatch_step i
-	sfi_breg src, \
-	ldr	tmp1, [\B, #-(\i * 4)]
-	sfi_breg dst, \
-	str	tmp1, [\B, #-(\i * 4)]
+	ldr	tmp1, [src, #-(\i * 4)]
+	str	tmp1, [dst, #-(\i * 4)]
 	.endm
 	dispatch_15_word
 #endif
 
 	lsls	count, count, #31
-	sfi_breg src, \
-	ldrhcs	tmp1, [\B], #2
-	sfi_breg src, \
-	ldrbne	src, [\B]		/* Src is dead, use as a scratch.  */
-	sfi_breg dst, \
-	strhcs	tmp1, [\B], #2
-	sfi_breg dst, \
-	strbne	src, [\B]
+	ldrhcs	tmp1, [src], #2
+	ldrbne	src, [src]		/* Src is dead, use as a scratch.  */
+	strhcs	tmp1, [dst], #2
+	strbne	src, [dst]
 	bx	lr
 
 .Lcpy_not_short:
@@ -388,19 +341,13 @@ ENTRY(memcpy)
 	beq	1f
 	rsbs	tmp2, tmp2, #0
 	sub	count, count, tmp2, lsr #29
-	sfi_breg src, \
-	ldrmi	tmp1, [\B], #4
-	sfi_breg dst, \
-	strmi	tmp1, [\B], #4
+	ldrmi	tmp1, [src], #4
+	strmi	tmp1, [dst], #4
 	lsls	tmp2, tmp2, #2
-	sfi_breg src, \
-	ldrhcs	tmp1, [\B], #2
-	sfi_breg src, \
-	ldrbne	tmp2, [\B], #1
-	sfi_breg dst, \
-	strhcs	tmp1, [\B], #2
-	sfi_breg dst, \
-	strbne	tmp2, [\B], #1
+	ldrhcs	tmp1, [src], #2
+	ldrbne	tmp2, [src], #1
+	strhcs	tmp1, [dst], #2
+	strbne	tmp2, [dst], #1
 
 1:
 	subs	tmp2, count, #64	/* Use tmp2 for count.  */
@@ -412,40 +359,24 @@ ENTRY(memcpy)
 .Lcpy_body_medium:			/* Count in tmp2.  */
 #ifdef USE_VFP
 1:
-	sfi_breg src, \
-	vldr	d0, [\B, #0]
+	vldr	d0, [src, #0]
 	subs	tmp2, tmp2, #64
-	sfi_breg src, \
-	vldr	d1, [\B, #8]
-	sfi_breg dst, \
-	vstr	d0, [\B, #0]
-	sfi_breg src, \
-	vldr	d0, [\B, #16]
-	sfi_breg dst, \
-	vstr	d1, [\B, #8]
-	sfi_breg src, \
-	vldr	d1, [\B, #24]
-	sfi_breg dst, \
-	vstr	d0, [\B, #16]
-	sfi_breg src, \
-	vldr	d0, [\B, #32]
-	sfi_breg dst, \
-	vstr	d1, [\B, #24]
-	sfi_breg src, \
-	vldr	d1, [\B, #40]
-	sfi_breg dst, \
-	vstr	d0, [\B, #32]
-	sfi_breg src, \
-	vldr	d0, [\B, #48]
-	sfi_breg dst, \
-	vstr	d1, [\B, #40]
-	sfi_breg src, \
-	vldr	d1, [\B, #56]
-	sfi_breg dst, \
-	vstr	d0, [\B, #48]
+	vldr	d1, [src, #8]
+	vstr	d0, [dst, #0]
+	vldr	d0, [src, #16]
+	vstr	d1, [dst, #8]
+	vldr	d1, [src, #24]
+	vstr	d0, [dst, #16]
+	vldr	d0, [src, #32]
+	vstr	d1, [dst, #24]
+	vldr	d1, [src, #40]
+	vstr	d0, [dst, #32]
+	vldr	d0, [src, #48]
+	vstr	d1, [dst, #40]
+	vldr	d1, [src, #56]
+	vstr	d0, [dst, #48]
 	add	src, src, #64
-	sfi_breg dst, \
-	vstr	d1, [\B, #56]
+	vstr	d1, [dst, #56]
 	add	dst, dst, #64
 	bge	1b
 	tst	tmp2, #0x3f
@@ -456,48 +387,30 @@ ENTRY(memcpy)
 	add	dst, dst, tmp1
 	add	src, src, tmp1
 	.macro dispatch_step i
-	sfi_breg src, \
-	vldr	d0, [\B, #-(\i * 8)]
-	sfi_breg dst, \
-	vstr	d0, [\B, #-(\i * 8)]
+	vldr	d0, [src, #-(\i * 8)]
+	vstr	d0, [dst, #-(\i * 8)]
 	.endm
 	dispatch_7_dword
 #else
 	sub	src, src, #8
 	sub	dst, dst, #8
 1:
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #8]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #8]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #16]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #16]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #24]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #24]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #32]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #32]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #40]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #40]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #48]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #48]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #56]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #56]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #64]!
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #64]!
+	ldrd	A_l, A_h, [src, #8]
+	strd	A_l, A_h, [dst, #8]
+	ldrd	A_l, A_h, [src, #16]
+	strd	A_l, A_h, [dst, #16]
+	ldrd	A_l, A_h, [src, #24]
+	strd	A_l, A_h, [dst, #24]
+	ldrd	A_l, A_h, [src, #32]
+	strd	A_l, A_h, [dst, #32]
+	ldrd	A_l, A_h, [src, #40]
+	strd	A_l, A_h, [dst, #40]
+	ldrd	A_l, A_h, [src, #48]
+	strd	A_l, A_h, [dst, #48]
+	ldrd	A_l, A_h, [src, #56]
+	strd	A_l, A_h, [dst, #56]
+	ldrd	A_l, A_h, [src, #64]!
+	strd	A_l, A_h, [dst, #64]!
 	subs	tmp2, tmp2, #64
 	bge	1b
 	tst	tmp2, #0x3f
@@ -524,28 +437,20 @@ ENTRY(memcpy)
 	add	dst, dst, tmp1
 	add	src, src, tmp1
 	.macro dispatch_step i
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #-(\i * 8)]
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #-(\i * 8)]
+	ldrd	A_l, A_h, [src, #-(\i * 8)]
+	strd	A_l, A_h, [dst, #-(\i * 8)]
 	.endm
 	dispatch_7_dword
 #endif
 
 	tst	tmp2, #4
-	sfi_breg src, \
-	ldrne	tmp1, [\B], #4
-	sfi_breg dst, \
-	strne	tmp1, [\B], #4
+	ldrne	tmp1, [src], #4
+	strne	tmp1, [dst], #4
 	lsls	tmp2, tmp2, #31		/* Count (tmp2) now dead. */
-	sfi_breg src, \
-	ldrhcs	tmp1, [\B], #2
-	sfi_breg src, \
-	ldrbne	tmp2, [\B]
-	sfi_breg dst, \
-	strhcs	tmp1, [\B], #2
-	sfi_breg dst, \
-	strbne	tmp2, [\B]
+	ldrhcs	tmp1, [src], #2
+	ldrbne	tmp2, [src]
+	strhcs	tmp1, [dst], #2
+	strbne	tmp2, [dst]
 
 .Ldone:
 	ldr	tmp2, [sp], #FRAME_SIZE
@@ -565,23 +470,15 @@ ENTRY(memcpy)
 	   copy position into a register.  This should act like a PLD
 	   operation but we won't have to repeat the transfer.  */
 
-	sfi_breg src, \
-	vldr	d3, [\B, #0]
-	sfi_breg src, \
-	vldr	d4, [\B, #64]
-	sfi_breg src, \
-	vldr	d5, [\B, #128]
-	sfi_breg src, \
-	vldr	d6, [\B, #192]
-	sfi_breg src, \
-	vldr	d7, [\B, #256]
-
-	sfi_breg src, \
-	vldr	d0, [\B, #8]
-	sfi_breg src, \
-	vldr	d1, [\B, #16]
-	sfi_breg src, \
-	vldr	d2, [\B, #24]
+	vldr	d3, [src, #0]
+	vldr	d4, [src, #64]
+	vldr	d5, [src, #128]
+	vldr	d6, [src, #192]
+	vldr	d7, [src, #256]
+
+	vldr	d0, [src, #8]
+	vldr	d1, [src, #16]
+	vldr	d2, [src, #24]
 	add	src, src, #32
 
 	subs	tmp2, tmp2, #prefetch_lines * 64 * 2
@@ -606,31 +503,19 @@ ENTRY(memcpy)
 	add	src, src, #3 * 64
 	add	dst, dst, #3 * 64
 	cpy_tail_vfp	d6, 0
-	sfi_breg dst, \
-	vstr	d7, [\B, #64]
-	sfi_breg src, \
-	vldr	d7, [\B, #64]
-	sfi_breg dst, \
-	vstr	d0, [\B, #64 + 8]
-	sfi_breg src, \
-	vldr	d0, [\B, #64 + 8]
-	sfi_breg dst, \
-	vstr	d1, [\B, #64 + 16]
-	sfi_breg src, \
-	vldr	d1, [\B, #64 + 16]
-	sfi_breg dst, \
-	vstr	d2, [\B, #64 + 24]
-	sfi_breg src, \
-	vldr	d2, [\B, #64 + 24]
-	sfi_breg dst, \
-	vstr	d7, [\B, #64 + 32]
+	vstr	d7, [dst, #64]
+	vldr	d7, [src, #64]
+	vstr	d0, [dst, #64 + 8]
+	vldr	d0, [src, #64 + 8]
+	vstr	d1, [dst, #64 + 16]
+	vldr	d1, [src, #64 + 16]
+	vstr	d2, [dst, #64 + 24]
+	vldr	d2, [src, #64 + 24]
+	vstr	d7, [dst, #64 + 32]
 	add	src, src, #96
-	sfi_breg dst, \
-	vstr	d0, [\B, #64 + 40]
-	sfi_breg dst, \
-	vstr	d1, [\B, #64 + 48]
-	sfi_breg dst, \
-	vstr	d2, [\B, #64 + 56]
+	vstr	d0, [dst, #64 + 40]
+	vstr	d1, [dst, #64 + 48]
+	vstr	d2, [dst, #64 + 56]
 	add	dst, dst, #128
 	add	tmp2, tmp2, #prefetch_lines * 64
 	b	.Lcpy_body_medium
@@ -641,83 +526,59 @@ ENTRY(memcpy)
 	/* Pre-bias src and dst.  */
 	sub	src, src, #8
 	sub	dst, dst, #8
-	sfi_pld	src, #8
-	sfi_pld	src, #72
+	pld	[src, #8]
+	pld	[src, #72]
 	subs	tmp2, tmp2, #64
-	sfi_pld	src, #136
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #8]
+	pld	[src, #136]
+	ldrd	A_l, A_h, [src, #8]
 	strd	B_l, B_h, [sp, #8]
 	cfi_rel_offset (B_l, 8)
 	cfi_rel_offset (B_h, 12)
-	sfi_breg src, \
-	ldrd	B_l, B_h, [\B, #16]
+	ldrd	B_l, B_h, [src, #16]
 	strd	C_l, C_h, [sp, #16]
 	cfi_rel_offset (C_l, 16)
 	cfi_rel_offset (C_h, 20)
-	sfi_breg src, \
-	ldrd	C_l, C_h, [\B, #24]
+	ldrd	C_l, C_h, [src, #24]
 	strd	D_l, D_h, [sp, #24]
 	cfi_rel_offset (D_l, 24)
 	cfi_rel_offset (D_h, 28)
-	sfi_pld	src, #200
-	sfi_breg src, \
-	ldrd	D_l, D_h, [\B, #32]!
+	pld	[src, #200]
+	ldrd	D_l, D_h, [src, #32]!
 	b	1f
 	.p2align	6
 2:
-	sfi_pld	src, #232
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #40]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #40]
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #48]
-	sfi_breg src, \
-	ldrd	B_l, B_h, [\B, #48]
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #56]
-	sfi_breg src, \
-	ldrd	C_l, C_h, [\B, #56]
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #64]!
-	sfi_breg src, \
-	ldrd	D_l, D_h, [\B, #64]!
+	pld	[src, #232]
+	strd	A_l, A_h, [dst, #40]
+	ldrd	A_l, A_h, [src, #40]
+	strd	B_l, B_h, [dst, #48]
+	ldrd	B_l, B_h, [src, #48]
+	strd	C_l, C_h, [dst, #56]
+	ldrd	C_l, C_h, [src, #56]
+	strd	D_l, D_h, [dst, #64]!
+	ldrd	D_l, D_h, [src, #64]!
 	subs	tmp2, tmp2, #64
 1:
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #8]
-	sfi_breg src, \
-	ldrd	A_l, A_h, [\B, #8]
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #16]
-	sfi_breg src, \
-	ldrd	B_l, B_h, [\B, #16]
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #24]
-	sfi_breg src, \
-	ldrd	C_l, C_h, [\B, #24]
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #32]
-	sfi_breg src, \
-	ldrd	D_l, D_h, [\B, #32]
+	strd	A_l, A_h, [dst, #8]
+	ldrd	A_l, A_h, [src, #8]
+	strd	B_l, B_h, [dst, #16]
+	ldrd	B_l, B_h, [src, #16]
+	strd	C_l, C_h, [dst, #24]
+	ldrd	C_l, C_h, [src, #24]
+	strd	D_l, D_h, [dst, #32]
+	ldrd	D_l, D_h, [src, #32]
 	bcs	2b
 	/* Save the remaining bytes and restore the callee-saved regs.  */
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #40]
+	strd	A_l, A_h, [dst, #40]
 	add	src, src, #40
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #48]
+	strd	B_l, B_h, [dst, #48]
 	ldrd	B_l, B_h, [sp, #8]
 	cfi_restore (B_l)
 	cfi_restore (B_h)
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #56]
+	strd	C_l, C_h, [dst, #56]
 	ldrd	C_l, C_h, [sp, #16]
 	cfi_restore (C_l)
 	cfi_restore (C_h)
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #64]
+	strd	D_l, D_h, [dst, #64]
 	ldrd	D_l, D_h, [sp, #24]
 	cfi_restore (D_l)
 	cfi_restore (D_h)
@@ -734,35 +595,29 @@ ENTRY(memcpy)
 	cfi_remember_state
 
 .Lcpy_notaligned:
-	sfi_pld	src
-	sfi_pld	src, #64
+	pld	[src, #0]
+	pld	[src, #64]
 	/* There's at least 64 bytes to copy, but there is no mutual
 	   alignment.  */
 	/* Bring DST to 64-bit alignment.  */
 	lsls	tmp2, dst, #29
-	sfi_pld	src, #(2 * 64)
+	pld	[src, #(2 * 64)]
 	beq	1f
 	rsbs	tmp2, tmp2, #0
 	sub	count, count, tmp2, lsr #29
-	sfi_breg src, \
-	ldrmi	tmp1, [\B], #4
-	sfi_breg dst, \
-	strmi	tmp1, [\B], #4
+	ldrmi	tmp1, [src], #4
+	strmi	tmp1, [dst], #4
 	lsls	tmp2, tmp2, #2
-	sfi_breg src, \
-	ldrbne	tmp1, [\B], #1
-	sfi_breg src, \
-	ldrhcs	tmp2, [\B], #2
-	sfi_breg dst, \
-	strbne	tmp1, [\B], #1
-	sfi_breg dst, \
-	strhcs	tmp2, [\B], #2
+	ldrbne	tmp1, [src], #1
+	ldrhcs	tmp2, [src], #2
+	strbne	tmp1, [dst], #1
+	strhcs	tmp2, [dst], #2
 1:
-	sfi_pld	src, #(3 * 64)
+	pld	[src, #(3 * 64)]
 	subs	count, count, #64
 	ldrmi	tmp2, [sp], #FRAME_SIZE
 	bmi	.Ltail63unaligned
-	sfi_pld	src, #(4 * 64)
+	pld	[src, #(4 * 64)]
 
 #ifdef USE_NEON
 	/* These need an extra layer of macro just to work around a
@@ -775,132 +630,88 @@ ENTRY(memcpy)
 	vst1.8	{\reglist}, [ALIGN (\basereg, 64)]!
 	.endm
 
-	/* These are used by the NaCl sfi_breg macro.  */
-	.macro _sfi_breg_dmask_neon_load_multi reg
-	_sfi_dmask \reg
-	.endm
-	.macro _sfi_breg_dmask_neon_store_multi reg
-	_sfi_dmask \reg
-	.endm
-
-	sfi_breg src, neon_load_multi d0-d3, \B
-	sfi_breg src, neon_load_multi d4-d7, \B
+	neon_load_multi d0-d3, src
+	neon_load_multi d4-d7, src
 	subs	count, count, #64
 	bmi	2f
 1:
-	sfi_pld	src, #(4 * 64)
-	sfi_breg dst, neon_store_multi d0-d3, \B
-	sfi_breg src, neon_load_multi d0-d3, \B
-	sfi_breg dst, neon_store_multi d4-d7, \B
-	sfi_breg src, neon_load_multi d4-d7, \B
+	pld	[src, #(4 * 64)]
+	neon_store_multi d0-d3, dst
+	neon_load_multi d0-d3, src
+	neon_store_multi d4-d7, dst
+	neon_load_multi d4-d7, src
 	subs	count, count, #64
 	bpl	1b
 2:
-	sfi_breg dst, neon_store_multi d0-d3, \B
-	sfi_breg dst, neon_store_multi d4-d7, \B
+	neon_store_multi d0-d3, dst
+	neon_store_multi d4-d7, dst
 	ands	count, count, #0x3f
 #else
 	/* Use an SMS style loop to maximize the I/O bandwidth.  */
 	sub	src, src, #4
 	sub	dst, dst, #8
 	subs	tmp2, count, #64	/* Use tmp2 for count.  */
-	sfi_breg src, \
-	ldr	A_l, [\B, #4]
-	sfi_breg src, \
-	ldr	A_h, [\B, #8]
+	ldr	A_l, [src, #4]
+	ldr	A_h, [src, #8]
 	strd	B_l, B_h, [sp, #8]
 	cfi_rel_offset (B_l, 8)
 	cfi_rel_offset (B_h, 12)
-	sfi_breg src, \
-	ldr	B_l, [\B, #12]
-	sfi_breg src, \
-	ldr	B_h, [\B, #16]
+	ldr	B_l, [src, #12]
+	ldr	B_h, [src, #16]
 	strd	C_l, C_h, [sp, #16]
 	cfi_rel_offset (C_l, 16)
 	cfi_rel_offset (C_h, 20)
-	sfi_breg src, \
-	ldr	C_l, [\B, #20]
-	sfi_breg src, \
-	ldr	C_h, [\B, #24]
+	ldr	C_l, [src, #20]
+	ldr	C_h, [src, #24]
 	strd	D_l, D_h, [sp, #24]
 	cfi_rel_offset (D_l, 24)
 	cfi_rel_offset (D_h, 28)
-	sfi_breg src, \
-	ldr	D_l, [\B, #28]
-	sfi_breg src, \
-	ldr	D_h, [\B, #32]!
+	ldr	D_l, [src, #28]
+	ldr	D_h, [src, #32]!
 	b	1f
 	.p2align	6
 2:
-	sfi_pld	src, #(5 * 64) - (32 - 4)
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #40]
-	sfi_breg src, \
-	ldr	A_l, [\B, #36]
-	sfi_breg src, \
-	ldr	A_h, [\B, #40]
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #48]
-	sfi_breg src, \
-	ldr	B_l, [\B, #44]
-	sfi_breg src, \
-	ldr	B_h, [\B, #48]
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #56]
-	sfi_breg src, \
-	ldr	C_l, [\B, #52]
-	sfi_breg src, \
-	ldr	C_h, [\B, #56]
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #64]!
-	sfi_breg src, \
-	ldr	D_l, [\B, #60]
-	sfi_breg src, \
-	ldr	D_h, [\B, #64]!
+	pld	[src, #(5 * 64) - (32 - 4)]
+	strd	A_l, A_h, [dst, #40]
+	ldr	A_l, [src, #36]
+	ldr	A_h, [src, #40]
+	strd	B_l, B_h, [dst, #48]
+	ldr	B_l, [src, #44]
+	ldr	B_h, [src, #48]
+	strd	C_l, C_h, [dst, #56]
+	ldr	C_l, [src, #52]
+	ldr	C_h, [src, #56]
+	strd	D_l, D_h, [dst, #64]!
+	ldr	D_l, [src, #60]
+	ldr	D_h, [src, #64]!
 	subs	tmp2, tmp2, #64
 1:
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #8]
-	sfi_breg src, \
-	ldr	A_l, [\B, #4]
-	sfi_breg src, \
-	ldr	A_h, [\B, #8]
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #16]
-	sfi_breg src, \
-	ldr	B_l, [\B, #12]
-	sfi_breg src, \
-	ldr	B_h, [\B, #16]
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #24]
-	sfi_breg src, \
-	ldr	C_l, [\B, #20]
-	sfi_breg src, \
-	ldr	C_h, [\B, #24]
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #32]
-	sfi_breg src, \
-	ldr	D_l, [\B, #28]
-	sfi_breg src, \
-	ldr	D_h, [\B, #32]
+	strd	A_l, A_h, [dst, #8]
+	ldr	A_l, [src, #4]
+	ldr	A_h, [src, #8]
+	strd	B_l, B_h, [dst, #16]
+	ldr	B_l, [src, #12]
+	ldr	B_h, [src, #16]
+	strd	C_l, C_h, [dst, #24]
+	ldr	C_l, [src, #20]
+	ldr	C_h, [src, #24]
+	strd	D_l, D_h, [dst, #32]
+	ldr	D_l, [src, #28]
+	ldr	D_h, [src, #32]
 	bcs	2b
 
 	/* Save the remaining bytes and restore the callee-saved regs.  */
-	sfi_breg dst, \
-	strd	A_l, A_h, [\B, #40]
+	strd	A_l, A_h, [dst, #40]
 	add	src, src, #36
-	sfi_breg dst, \
-	strd	B_l, B_h, [\B, #48]
+	strd	B_l, B_h, [dst, #48]
 	ldrd	B_l, B_h, [sp, #8]
 	cfi_restore (B_l)
 	cfi_restore (B_h)
-	sfi_breg dst, \
-	strd	C_l, C_h, [\B, #56]
+	strd	C_l, C_h, [dst, #56]
 	ldrd	C_l, C_h, [sp, #16]
 	cfi_restore (C_l)
 	cfi_restore (C_h)
-	sfi_breg dst, \
-	strd	D_l, D_h, [\B, #64]
+	strd	D_l, D_h, [dst, #64]
 	ldrd	D_l, D_h, [sp, #24]
 	cfi_restore (D_l)
 	cfi_restore (D_h)
diff --git a/sysdeps/arm/armv7/strcmp.S b/sysdeps/arm/armv7/strcmp.S
index c8fab4ba0f..25d055754e 100644
--- a/sysdeps/arm/armv7/strcmp.S
+++ b/sysdeps/arm/armv7/strcmp.S
@@ -178,10 +178,8 @@
 #endif
 ENTRY (strcmp)
 #if STRCMP_PRECHECK == 1
-	sfi_breg src1, \
-	ldrb	r2, [\B]
-	sfi_breg src2, \
-	ldrb	r3, [\B]
+	ldrb	r2, [src1]
+	ldrb	r3, [src2]
 	cmp	r2, #1
 	it	cs
 	cmpcs	r2, r3
@@ -211,11 +209,9 @@ ENTRY (strcmp)
 	and	tmp2, tmp1, #3
 	bic	src2, src2, #7
 	lsl	tmp2, tmp2, #3	/* Bytes -> bits.  */
-	sfi_breg src1, \
-	ldrd	data1a, data1b, [\B], #16
+	ldrd	data1a, data1b, [src1], #16
 	tst	tmp1, #4
-	sfi_breg src2, \
-	ldrd	data2a, data2b, [\B], #16
+	ldrd	data2a, data2b, [src2], #16
 	prepare_mask tmp1, tmp2
 	apply_mask data1a, tmp1
 	apply_mask data2a, tmp1
@@ -231,10 +227,8 @@ ENTRY (strcmp)
 	.p2align 5,,12  /* Don't start in the tail bytes of a cache line.  */
 	.p2align 2	/* Always word aligned.  */
 .Lloop_aligned8:
-	sfi_breg src1, \
-	ldrd	data1a, data1b, [\B], #16
-	sfi_breg src2, \
-	ldrd	data2a, data2b, [\B], #16
+	ldrd	data1a, data1b, [src1], #16
+	ldrd	data2a, data2b, [src2], #16
 .Lstart_realigned8:
 	uadd8	syndrome_b, data1a, const_m1	/* Only want GE bits,  */
 	eor	syndrome_a, data1a, data2a
@@ -245,10 +239,8 @@ ENTRY (strcmp)
 	sel	syndrome_b, syndrome_b, const_m1
 	cbnz	syndrome_b, .Ldiff_in_b
 
-	sfi_breg src1, \
-	ldrd	data1a, data1b, [\B, #-8]
-	sfi_breg src2, \
-	ldrd	data2a, data2b, [\B, #-8]
+	ldrd	data1a, data1b, [src1, #-8]
+	ldrd	data2a, data2b, [src2, #-8]
 	uadd8	syndrome_b, data1a, const_m1	/* Only want GE bits,  */
 	eor	syndrome_a, data1a, data2a
 	sel	syndrome_a, syndrome_a, const_m1
@@ -279,19 +271,15 @@ ENTRY (strcmp)
 	/* Unrolled by a factor of 2, to reduce the number of post-increment
 	   operations.  */
 .Lloop_aligned4:
-	sfi_breg src1, \
-	ldr	data1, [\B], #8
-	sfi_breg src2, \
-	ldr	data2, [\B], #8
+	ldr	data1, [src1], #8
+	ldr	data2, [src2], #8
 .Lstart_realigned4:
 	uadd8	syndrome, data1, const_m1	/* Only need GE bits.  */
 	eor	syndrome, data1, data2
 	sel	syndrome, syndrome, const_m1
 	cbnz	syndrome, .Laligned4_done
-	sfi_breg src1, \
-	ldr	data1, [\B, #-4]
-	sfi_breg src2, \
-	ldr	data2, [\B, #-4]
+	ldr	data1, [src1, #-4]
+	ldr	data2, [src2, #-4]
 	uadd8	syndrome, data1, const_m1
 	eor	syndrome, data1, data2
 	sel	syndrome, syndrome, const_m1
@@ -307,11 +295,9 @@ ENTRY (strcmp)
 	   masking off the unwanted loaded data to prevent a difference.  */
 	lsl	tmp1, tmp1, #3	/* Bytes -> bits.  */
 	bic	src1, src1, #3
-	sfi_breg src1, \
-	ldr	data1, [\B], #8
+	ldr	data1, [src1], #8
 	bic	src2, src2, #3
-	sfi_breg src2, \
-	ldr	data2, [\B], #8
+	ldr	data2, [src2], #8
 
 	prepare_mask tmp1, tmp1
 	apply_mask data1, tmp1
@@ -324,30 +310,26 @@ ENTRY (strcmp)
 	sub	src2, src2, tmp1
 	bic	src1, src1, #3
 	lsls	tmp1, tmp1, #31
-	sfi_breg src1, \
-	ldr	data1, [\B], #4
+	ldr	data1, [src1], #4
 	beq	.Laligned_m2
 	bcs	.Laligned_m1
 
 #if STRCMP_PRECHECK == 0
-	sfi_breg src2, \
-	ldrb	data2, [\B, #1]
+	ldrb	data2, [src2, #1]
 	uxtb	tmp1, data1, ror #BYTE1_OFFSET
 	subs	tmp1, tmp1, data2
 	bne	.Lmisaligned_exit
 	cbz	data2, .Lmisaligned_exit
 
 .Laligned_m2:
-	sfi_breg src2, \
-	ldrb	data2, [\B, #2]
+	ldrb	data2, [src2, #2]
 	uxtb	tmp1, data1, ror #BYTE2_OFFSET
 	subs	tmp1, tmp1, data2
 	bne	.Lmisaligned_exit
 	cbz	data2, .Lmisaligned_exit
 
 .Laligned_m1:
-	sfi_breg src2, \
-	ldrb	data2, [\B, #3]
+	ldrb	data2, [src2, #3]
 	uxtb	tmp1, data1, ror #BYTE3_OFFSET
 	subs	tmp1, tmp1, data2
 	bne	.Lmisaligned_exit
@@ -356,16 +338,14 @@ ENTRY (strcmp)
 #else  /* STRCMP_PRECHECK */
 	/* If we've done the pre-check, then we don't need to check the
 	   first byte again here.  */
-	sfi_breg src2, \
-	ldrb	data2, [\B, #2]
+	ldrb	data2, [src2, #2]
 	uxtb	tmp1, data1, ror #BYTE2_OFFSET
 	subs	tmp1, tmp1, data2
 	bne	.Lmisaligned_exit
 	cbz	data2, .Lmisaligned_exit
 
 .Laligned_m2:
-	sfi_breg src2, \
-	ldrb	data2, [\B, #3]
+	ldrb	data2, [src2, #3]
 	uxtb	tmp1, data1, ror #BYTE3_OFFSET
 	subs	tmp1, tmp1, data2
 	bne	.Lmisaligned_exit
@@ -391,13 +371,11 @@ ENTRY (strcmp)
 	cfi_restore_state
 	/* src1 is word aligned, but src2 has no common alignment
 	   with it.  */
-	sfi_breg src1, \
-	ldr	data1, [\B], #4
+	ldr	data1, [src1], #4
 	lsls	tmp1, src2, #31		/* C=src2[1], Z=src2[0].  */
 
 	bic	src2, src2, #3
-	sfi_breg src2, \
-	ldr	data2, [\B], #4
+	ldr	data2, [src2], #4
 	bhi	.Loverlap1		/* C=1, Z=0 => src2[1:0] = 0b11.  */
 	bcs	.Loverlap2		/* C=1, Z=1 => src2[1:0] = 0b10.  */
 
@@ -409,13 +387,11 @@ ENTRY (strcmp)
 	sel	syndrome, syndrome, const_m1
 	bne	4f
 	cbnz	syndrome, 5f
-	sfi_breg src2, \
-	ldr	data2, [\B], #4
+	ldr	data2, [src2], #4
 	eor	tmp1, tmp1, data1
 	cmp	tmp1, data2, S2HI #24
 	bne	6f
-	sfi_breg src1, \
-	ldr	data1, [\B], #4
+	ldr	data1, [src1], #4
 	b	.Loverlap3
 4:
 	S2LO	data2, data2, #8
@@ -427,8 +403,7 @@ ENTRY (strcmp)
 
 	/* We can only get here if the MSB of data1 contains 0, so
 	   fast-path the exit.  */
-	sfi_breg src2, \
-	ldrb	result, [\B]
+	ldrb	result, [src2]
 	ldrd	r4, r5, [sp], #16
 	cfi_remember_state
 	cfi_def_cfa_offset (0)
@@ -454,13 +429,11 @@ ENTRY (strcmp)
 	sel	syndrome, syndrome, const_m1
 	bne	4f
 	cbnz	syndrome, 5f
-	sfi_breg src2, \
-	ldr	data2, [\B], #4
+	ldr	data2, [src2], #4
 	eor	tmp1, tmp1, data1
 	cmp	tmp1, data2, S2HI #16
 	bne	6f
-	sfi_breg src1, \
-	ldr	data1, [\B], #4
+	ldr	data1, [src1], #4
 	b	.Loverlap2
 4:
 	S2LO	data2, data2, #16
@@ -469,8 +442,7 @@ ENTRY (strcmp)
 	ands	syndrome, syndrome, const_m1, S2LO #16
 	bne	.Lstrcmp_done_equal
 
-	sfi_breg src2, \
-	ldrh	data2, [\B]
+	ldrh	data2, [src2]
 	S2LO	data1, data1, #16
 #ifdef __ARM_BIG_ENDIAN
 	lsl	data2, data2, #16
@@ -490,13 +462,11 @@ ENTRY (strcmp)
 	sel	syndrome, syndrome, const_m1
 	bne	4f
 	cbnz	syndrome, 5f
-	sfi_breg src2, \
-	ldr	data2, [\B], #4
+	ldr	data2, [src2], #4
 	eor	tmp1, tmp1, data1
 	cmp	tmp1, data2, S2HI #8
 	bne	6f
-	sfi_breg src1, \
-	ldr	data1, [\B], #4
+	ldr	data1, [src1], #4
 	b	.Loverlap1
 4:
 	S2LO	data2, data2, #24
@@ -504,8 +474,7 @@ ENTRY (strcmp)
 5:
 	tst	syndrome, #LSB
 	bne	.Lstrcmp_done_equal
-	sfi_breg src2, \
-	ldr	data2, [\B]
+	ldr	data2, [src2]
 6:
 	S2LO	data1, data1, #8
 	bic	data2, data2, #MSB
diff --git a/sysdeps/arm/dl-tlsdesc.S b/sysdeps/arm/dl-tlsdesc.S
index 5caf2c154d..e7bed02188 100644
--- a/sysdeps/arm/dl-tlsdesc.S
+++ b/sysdeps/arm/dl-tlsdesc.S
@@ -32,8 +32,7 @@
 	eabi_fnstart
 	.align 2
 _dl_tlsdesc_return:
-	sfi_breg r0, \
-	ldr	r0, [\B]
+	ldr	r0, [r0]
 	BX	(lr)
 	eabi_fnend
 	cfi_endproc
@@ -92,30 +91,23 @@ _dl_tlsdesc_dynamic:
 	cfi_rel_offset (r3,4)
 	cfi_rel_offset (r4,8)
 	cfi_rel_offset (lr,12)
-	sfi_breg r0, \
-	ldr	r1, [\B] /* td */
+	ldr	r1, [r0] /* td */
 	GET_TLS (lr)
 	mov	r4, r0 /* r4 = tp */
-	sfi_breg r0, \
-	ldr	r0, [\B]
-	sfi_breg r1, \
-	ldr	r2, [\B, #8] /* gen_count */
-	sfi_breg r0, \
-	ldr	r3, [\B]
+	ldr	r0, [r0]
+	ldr	r2, [r1, #8] /* gen_count */
+	ldr	r3, [r0]
 	cmp	r2, r3
 	bhi	1f
-	sfi_breg r1, \
-	ldr	r3, [\B]
+	ldr	r3, [r1]
 #ifndef ARM_NO_INDEX_REGISTER
 	ldr	r2, [r0, r3, lsl #3]
 #else
 	add	lr, r0, r3, lsl #3
-	sfi_breg lr, \
-	ldr	r2, [\B]
+	ldr	r2, [lr]
 #endif
 	cmn	r2, #1
 	ittt	ne
-	sfi_breg r1, \
 	ldrne	r3, [r1, #4]
 	addne	r3, r2, r3
 	rsbne	r0, r4, r3
@@ -178,8 +170,7 @@ _dl_tlsdesc_lazy_resolver:
 	pop	{r2}
 	cfi_adjust_cfa_offset (-4)
 	cfi_restore (r2)
-	sfi_breg r0, \
-	ldr	r1, [\B, #4]
+	ldr	r1, [r0, #4]
 	BX	(r1)
 	eabi_fnend
 	cfi_endproc
@@ -220,8 +211,7 @@ _dl_tlsdesc_resolve_hold:
 	cfi_restore (r2)
 	cfi_restore (r1)
 	cfi_restore (r0)
-	sfi_breg r0, \
-	ldr     r1, [\B, #4]
+	ldr     r1, [r0, #4]
 	BX      (r1)
 	eabi_fnend
 	cfi_endproc
diff --git a/sysdeps/arm/memcpy.S b/sysdeps/arm/memcpy.S
index bdb830a8fe..62e48c3e4a 100644
--- a/sysdeps/arm/memcpy.S
+++ b/sysdeps/arm/memcpy.S
@@ -70,7 +70,7 @@ ENTRY(memcpy)
 		subs	r2, r2, #4
 		blt	8f
 		ands	ip, r0, #3
-	PLD(	sfi_pld	r1, #0			)
+	PLD(	pld	[r1, #0]		)
 		bne	9f
 		ands	ip, r1, #3
 		bne	10f
@@ -97,19 +97,17 @@ ENTRY(memcpy)
 	CALGN(	bx	r4			)
 #endif
 
-	PLD(	sfi_pld	r1, #0			)
+	PLD(	pld	[r1, #0]		)
 2:	PLD(	subs	r2, r2, #96		)
-	PLD(	sfi_pld	r1, #28			)
+	PLD(	pld	[r1, #28]		)
 	PLD(	blt	4f			)
-	PLD(	sfi_pld	r1, #60			)
-	PLD(	sfi_pld	r1, #92			)
+	PLD(	pld	[r1, #60]		)
+	PLD(	pld	[r1, #92]		)
 
-3:	PLD(	sfi_pld	r1, #124		)
-4:		sfi_breg r1, \
-		ldmia	\B!, {r3, r4, r5, r6, r7, r8, ip, lr}
+3:	PLD(	pld	[r1, #124]		)
+4:		ldmia	r1!, {r3, r4, r5, r6, r7, r8, ip, lr}
 		subs	r2, r2, #32
-		sfi_breg r0, \
-		stmia	\B!, {r3, r4, r5, r6, r7, r8, ip, lr}
+		stmia	r0!, {r3, r4, r5, r6, r7, r8, ip, lr}
 		bge	3b
 	PLD(	cmn	r2, #96			)
 	PLD(	bge	4b			)
@@ -136,26 +134,19 @@ ENTRY(memcpy)
 		.p2align ARM_BX_ALIGN_LOG2
 6:		nop
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r3, [\B], #4
+		ldr	r3, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r4, [\B], #4
+		ldr	r4, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r5, [\B], #4
+		ldr	r5, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r6, [\B], #4
+		ldr	r6, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r7, [\B], #4
+		ldr	r7, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r8, [\B], #4
+		ldr	r8, [r1], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	lr, [\B], #4
+		ldr	lr, [r1], #4
 
 #ifndef ARM_ALWAYS_BX
 		add	pc, pc, ip, lsl #(ARM_BX_ALIGN_LOG2 - 2)
@@ -172,26 +163,19 @@ ENTRY(memcpy)
 		.p2align ARM_BX_ALIGN_LOG2
 66:		nop
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r3, [\B], #4
+		str	r3, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r4, [\B], #4
+		str	r4, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r5, [\B], #4
+		str	r5, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r6, [\B], #4
+		str	r6, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r7, [\B], #4
+		str	r7, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r8, [\B], #4
+		str	r8, [r0], #4
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	lr, [\B], #4
+		str	lr, [r0], #4
 
 #ifdef ARM_ALWAYS_BX
 		pop	{r10}
@@ -209,18 +193,12 @@ ENTRY(memcpy)
 		cfi_restore (r8)
 
 8:		movs	r2, r2, lsl #31
-		sfi_breg r1, \
-		ldrbne	r3, [\B], #1
-		sfi_breg r1, \
-		ldrbcs	r4, [\B], #1
-		sfi_breg r1, \
-		ldrbcs	ip, [\B]
-		sfi_breg r0, \
-		strbne	r3, [\B], #1
-		sfi_breg r0, \
-		strbcs	r4, [\B], #1
-		sfi_breg r0, \
-		strbcs	ip, [\B]
+		ldrbne	r3, [r1], #1
+		ldrbcs	r4, [r1], #1
+		ldrbcs	ip, [r1]
+		strbne	r3, [r0], #1
+		strbcs	r4, [r0], #1
+		strbcs	ip, [r0]
 
 #if ((defined (__ARM_ARCH_4T__) && defined(__THUMB_INTERWORK__)) \
      || defined (ARM_ALWAYS_BX))
@@ -237,27 +215,20 @@ ENTRY(memcpy)
 
 9:		rsb	ip, ip, #4
 		cmp	ip, #2
-		sfi_breg r1, \
-		ldrbgt	r3, [\B], #1
-		sfi_breg r1, \
-		ldrbge	r4, [\B], #1
-		sfi_breg r1, \
-		ldrb	lr, [\B], #1
-		sfi_breg r0, \
-		strbgt	r3, [\B], #1
-		sfi_breg r0, \
-		strbge	r4, [\B], #1
+		ldrbgt	r3, [r1], #1
+		ldrbge	r4, [r1], #1
+		ldrb	lr, [r1], #1
+		strbgt	r3, [r0], #1
+		strbge	r4, [r0], #1
 		subs	r2, r2, ip
-		sfi_breg r0, \
-		strb	lr, [\B], #1
+		strb	lr, [r0], #1
 		blt	8b
 		ands	ip, r1, #3
 		beq	1b
 
 10:		bic	r1, r1, #3
 		cmp	ip, #2
-		sfi_breg r1, \
-		ldr	lr, [\B], #4
+		ldr	lr, [r1], #4
 		beq	17f
 		bgt	18f
 
@@ -281,20 +252,18 @@ ENTRY(memcpy)
 		cfi_rel_offset (r8, 12)
 		cfi_rel_offset (r10, 16)
 
-	PLD(	sfi_pld	r1, #0			)
+	PLD(	pld	[r1, #0]		)
 	PLD(	subs	r2, r2, #96		)
-	PLD(	sfi_pld	r1, #28			)
+	PLD(	pld	[r1, #28]		)
 	PLD(	blt	13f			)
-	PLD(	sfi_pld	r1, #60			)
-	PLD(	sfi_pld	r1, #92			)
+	PLD(	pld	[r1, #60]		)
+	PLD(	pld	[r1, #92]		)
 
-12:	PLD(	sfi_pld	r1, #124		)
-13:		sfi_breg r1, \
-		ldmia	\B!, {r4, r5, r6, r7}
+12:	PLD(	pld	[r1, #124]		)
+13:		ldmia	r1!, {r4, r5, r6, r7}
 		mov	r3, lr, PULL #\pull
 		subs	r2, r2, #32
-		sfi_breg r1, \
-		ldmia	\B!, {r8, r10, ip, lr}
+		ldmia	r1!, {r8, r10, ip, lr}
 		orr	r3, r3, r4, PUSH #\push
 		mov	r4, r4, PULL #\pull
 		orr	r4, r4, r5, PUSH #\push
@@ -310,8 +279,7 @@ ENTRY(memcpy)
 		orr	r10, r10, ip, PUSH #\push
 		mov	ip, ip, PULL #\pull
 		orr	ip, ip, lr, PUSH #\push
-		sfi_breg r0, \
-		stmia	\B!, {r3, r4, r5, r6, r7, r8, r10, ip}
+		stmia	r0!, {r3, r4, r5, r6, r7, r8, r10, ip}
 		bge	12b
 	PLD(	cmn	r2, #96			)
 	PLD(	bge	13b			)
@@ -328,12 +296,10 @@ ENTRY(memcpy)
 		beq	16f
 
 15:		mov	r3, lr, PULL #\pull
-		sfi_breg r1, \
-		ldr	lr, [\B], #4
+		ldr	lr, [r1], #4
 		subs	ip, ip, #4
 		orr	r3, r3, lr, PUSH #\push
-		sfi_breg r0, \
-		str	r3, [\B], #4
+		str	r3, [r0], #4
 		bgt	15b
 	CALGN(	cmp	r2, #0			)
 	CALGN(	bge	11b			)
diff --git a/sysdeps/arm/memmove.S b/sysdeps/arm/memmove.S
index f5247f46d8..b18aa329d3 100644
--- a/sysdeps/arm/memmove.S
+++ b/sysdeps/arm/memmove.S
@@ -87,7 +87,7 @@ ENTRY(memmove)
 		subs	r2, r2, #4
 		blt	8f
 		ands	ip, r0, #3
-	PLD(	sfi_pld	r1, #-4			)
+	PLD(	pld	[r1, #-4]		)
 		bne	9f
 		ands	ip, r1, #3
 		bne	10f
@@ -113,19 +113,17 @@ ENTRY(memmove)
 	CALGN(	bx	r4			)
 #endif
 
-	PLD(	sfi_pld	r1, #-4			)
+	PLD(	pld	[r1, #-4]		)
 2:	PLD(	subs	r2, r2, #96		)
-	PLD(	sfi_pld	r1, #-32		)
+	PLD(	pld	[r1, #-32]		)
 	PLD(	blt	4f			)
-	PLD(	sfi_pld	r1, #-64		)
-	PLD(	sfi_pld	r1, #-96		)
+	PLD(	pld	[r1, #-64]		)
+	PLD(	pld	[r1, #-96]		)
 
-3:	PLD(	sfi_pld	r1, #-128		)
-4:		sfi_breg r1, \
-		ldmdb	\B!, {r3, r4, r5, r6, r7, r8, ip, lr}
+3:	PLD(	pld	[r1, #-128]		)
+4:		ldmdb	r1!, {r3, r4, r5, r6, r7, r8, ip, lr}
 		subs	r2, r2, #32
-		sfi_breg r0, \
-		stmdb	\B!, {r3, r4, r5, r6, r7, r8, ip, lr}
+		stmdb	r0!, {r3, r4, r5, r6, r7, r8, ip, lr}
 		bge	3b
 	PLD(	cmn	r2, #96			)
 	PLD(	bge	4b			)
@@ -152,26 +150,19 @@ ENTRY(memmove)
 		.p2align ARM_BX_ALIGN_LOG2
 6:		nop
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r3, [\B, #-4]!
+		ldr	r3, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r4, [\B, #-4]!
+		ldr	r4, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r5, [\B, #-4]!
+		ldr	r5, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r6, [\B, #-4]!
+		ldr	r6, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r7, [\B, #-4]!
+		ldr	r7, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	r8, [\B, #-4]!
+		ldr	r8, [r1, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r1, \
-		ldr	lr, [\B, #-4]!
+		ldr	lr, [r1, #-4]!
 
 #ifndef ARM_ALWAYS_BX
 		add	pc, pc, ip, lsl #(ARM_BX_ALIGN_LOG2 - 2)
@@ -188,26 +179,19 @@ ENTRY(memmove)
 		.p2align ARM_BX_ALIGN_LOG2
 66:		nop
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r3, [\B, #-4]!
+		str	r3, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r4, [\B, #-4]!
+		str	r4, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r5, [\B, #-4]!
+		str	r5, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r6, [\B, #-4]!
+		str	r6, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r7, [\B, #-4]!
+		str	r7, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	r8, [\B, #-4]!
+		str	r8, [r0, #-4]!
 		.p2align ARM_BX_ALIGN_LOG2
-		sfi_breg r0, \
-		str	lr, [\B, #-4]!
+		str	lr, [r0, #-4]!
 
 #ifdef ARM_ALWAYS_BX
 		pop	{r10}
@@ -225,18 +209,12 @@ ENTRY(memmove)
 		cfi_restore (r8)
 
 8:		movs	r2, r2, lsl #31
-		sfi_breg r1, \
-		ldrbne	r3, [\B, #-1]!
-		sfi_breg r1, \
-		ldrbcs	r4, [\B, #-1]!
-		sfi_breg r1, \
-		ldrbcs	ip, [\B, #-1]
-		sfi_breg r0, \
-		strbne	r3, [\B, #-1]!
-		sfi_breg r0, \
-		strbcs	r4, [\B, #-1]!
-		sfi_breg r0, \
-		strbcs	ip, [\B, #-1]
+		ldrbne	r3, [r1, #-1]!
+		ldrbcs	r4, [r1, #-1]!
+		ldrbcs	ip, [r1, #-1]
+		strbne	r3, [r0, #-1]!
+		strbcs	r4, [r0, #-1]!
+		strbcs	ip, [r0, #-1]
 
 #if ((defined (__ARM_ARCH_4T__) && defined (__THUMB_INTERWORK__)) \
      || defined (ARM_ALWAYS_BX))
@@ -252,27 +230,20 @@ ENTRY(memmove)
 		cfi_restore_state
 
 9:		cmp	ip, #2
-		sfi_breg r1, \
-		ldrbgt	r3, [\B, #-1]!
-		sfi_breg r1, \
-		ldrbge	r4, [\B, #-1]!
-		sfi_breg r1, \
-		ldrb	lr, [\B, #-1]!
-		sfi_breg r0, \
-		strbgt	r3, [\B, #-1]!
-		sfi_breg r0, \
-		strbge	r4, [\B, #-1]!
+		ldrbgt	r3, [r1, #-1]!
+		ldrbge	r4, [r1, #-1]!
+		ldrb	lr, [r1, #-1]!
+		strbgt	r3, [r0, #-1]!
+		strbge	r4, [r0, #-1]!
 		subs	r2, r2, ip
-		sfi_breg r0, \
-		strb	lr, [\B, #-1]!
+		strb	lr, [r0, #-1]!
 		blt	8b
 		ands	ip, r1, #3
 		beq	1b
 
 10:		bic	r1, r1, #3
 		cmp	ip, #2
-		sfi_breg r1, \
-		ldr	r3, [\B, #0]
+		ldr	r3, [r1, #0]
 		beq	17f
 		blt	18f
 
@@ -296,20 +267,18 @@ ENTRY(memmove)
 		cfi_rel_offset (r8, 12)
 		cfi_rel_offset (r10, 16)
 
-	PLD(	sfi_pld	r1, #-4			)
+	PLD(	pld	[r1, #-4]		)
 	PLD(	subs	r2, r2, #96		)
-	PLD(	sfi_pld	r1, #-32		)
+	PLD(	pld	[r1, #-32]		)
 	PLD(	blt	13f			)
-	PLD(	sfi_pld	r1, #-64		)
-	PLD(	sfi_pld	r1, #-96		)
+	PLD(	pld	[r1, #-64]		)
+	PLD(	pld	[r1, #-96]		)
 
-12:	PLD(	sfi_pld	r1, #-128		)
-13:		sfi_breg r1, \
-		ldmdb   \B!, {r7, r8, r10, ip}
+12:	PLD(	pld	[r1, #-128]		)
+13:		ldmdb   r1!, {r7, r8, r10, ip}
 		mov     lr, r3, PUSH #\push
 		subs    r2, r2, #32
-		sfi_breg r1, \
-		ldmdb   \B!, {r3, r4, r5, r6}
+		ldmdb   r1!, {r3, r4, r5, r6}
 		orr     lr, lr, ip, PULL #\pull
 		mov     ip, ip, PUSH #\push
 		orr     ip, ip, r10, PULL #\pull
@@ -325,8 +294,7 @@ ENTRY(memmove)
 		orr     r5, r5, r4, PULL #\pull
 		mov     r4, r4, PUSH #\push
 		orr     r4, r4, r3, PULL #\pull
-		sfi_breg r0, \
-		stmdb   \B!, {r4 - r8, r10, ip, lr}
+		stmdb   r0!, {r4 - r8, r10, ip, lr}
 		bge	12b
 	PLD(	cmn	r2, #96			)
 	PLD(	bge	13b			)
@@ -343,12 +311,10 @@ ENTRY(memmove)
 		beq	16f
 
 15:		mov     lr, r3, PUSH #\push
-		sfi_breg r1, \
-		ldr	r3, [\B, #-4]!
+		ldr	r3, [r1, #-4]!
 		subs	ip, ip, #4
 		orr	lr, lr, r3, PULL #\pull
-		sfi_breg r0, \
-		str	lr, [\B, #-4]!
+		str	lr, [r0, #-4]!
 		bgt	15b
 	CALGN(	cmp	r2, #0			)
 	CALGN(	bge	11b			)
diff --git a/sysdeps/arm/memset.S b/sysdeps/arm/memset.S
index d369c20a80..95946360bf 100644
--- a/sysdeps/arm/memset.S
+++ b/sysdeps/arm/memset.S
@@ -32,8 +32,7 @@ ENTRY(memset)
 
 1:
 	tst	r3, #3		@ aligned yet?
-	sfi_breg r3, \
-	strbne	r1, [\B], #1
+	strbne	r1, [r3], #1
 	subne	r2, r2, #1
 	bne	1b
 
@@ -44,33 +43,25 @@ ENTRY(memset)
 
 1:
 	subs	r2, r2, #8
-	sfi_breg r3, \
-	stmiacs	\B!, {r1, ip}	@ store up to 32 bytes per loop iteration
+	stmiacs	r3!, {r1, ip}	@ store up to 32 bytes per loop iteration
 	subscs	r2, r2, #8
-	sfi_breg r3, \
-	stmiacs	\B!, {r1, ip}
+	stmiacs	r3!, {r1, ip}
 	subscs	r2, r2, #8
-	sfi_breg r3, \
-	stmiacs	\B!, {r1, ip}
+	stmiacs	r3!, {r1, ip}
 	subscs	r2, r2, #8
-	sfi_breg r3, \
-	stmiacs	\B!, {r1, ip}
+	stmiacs	r3!, {r1, ip}
 	bcs	1b
 
 	and	r2, r2, #7
 2:
 	subs	r2, r2, #1	@ store up to 4 bytes per loop iteration
-	sfi_breg r3, \
-	strbcs	r1, [\B], #1
+	strbcs	r1, [r3], #1
 	subscs	r2, r2, #1
-	sfi_breg r3, \
-	strbcs	r1, [\B], #1
+	strbcs	r1, [r3], #1
 	subscs	r2, r2, #1
-	sfi_breg r3, \
-	strbcs	r1, [\B], #1
+	strbcs	r1, [r3], #1
 	subscs	r2, r2, #1
-	sfi_breg r3, \
-	strbcs	r1, [\B], #1
+	strbcs	r1, [r3], #1
 	bcs	2b
 
 	DO_RET(lr)
diff --git a/sysdeps/arm/setjmp.S b/sysdeps/arm/setjmp.S
index c809253d1b..2235890936 100644
--- a/sysdeps/arm/setjmp.S
+++ b/sysdeps/arm/setjmp.S
@@ -36,20 +36,15 @@ ENTRY (__sigsetjmp)
 #ifdef PTR_MANGLE
 	mov	a4, sp
 	PTR_MANGLE2 (a4, a4, a3)
-	sfi_breg ip, \
-	str	a4, [\B], #4
+	str	a4, [ip], #4
 	PTR_MANGLE2 (a4, lr, a3)
-	sfi_breg ip, \
-	str	a4, [\B], #4
+	str	a4, [ip], #4
 #else
-	sfi_breg ip, \
-	str	sp, [\B], #4
-	sfi_breg ip, \
-	str	lr, [\B], #4
+	str	sp, [ip], #4
+	str	lr, [ip], #4
 #endif
 	/* Save registers */
-	sfi_breg ip, \
-	stmia	\B!, JMP_BUF_REGLIST
+	stmia	ip!, JMP_BUF_REGLIST
 
 #if !defined ARM_ASSUME_NO_IWMMXT || defined __SOFTFP__
 # define NEED_HWCAP 1
@@ -80,8 +75,7 @@ ENTRY (__sigsetjmp)
 	   Don't use VFP instructions directly because this code
 	   is used in non-VFP multilibs.  */
 	/* Following instruction is vstmia ip!, {d8-d15}.  */
-	sfi_breg ip, \
-	stc	p11, cr8, [\B], #64
+	stc	p11, cr8, [ip], #64
 .Lno_vfp:
 
 #ifndef ARM_ASSUME_NO_IWMMXT
@@ -90,18 +84,12 @@ ENTRY (__sigsetjmp)
 
 	/* Save the call-preserved iWMMXt registers.  */
 	/* Following instructions are wstrd wr10, [ip], #8 (etc.)  */
-	sfi_breg r12, \
-	stcl	p1, cr10, [\B], #8
-	sfi_breg r12, \
-	stcl	p1, cr11, [\B], #8
-	sfi_breg r12, \
-	stcl	p1, cr12, [\B], #8
-	sfi_breg r12, \
-	stcl	p1, cr13, [\B], #8
-	sfi_breg r12, \
-	stcl	p1, cr14, [\B], #8
-	sfi_breg r12, \
-	stcl	p1, cr15, [\B], #8
+	stcl	p1, cr10, [r12], #8
+	stcl	p1, cr11, [r12], #8
+	stcl	p1, cr12, [r12], #8
+	stcl	p1, cr13, [r12], #8
+	stcl	p1, cr14, [r12], #8
+	stcl	p1, cr15, [r12], #8
 .Lno_iwmmxt:
 #endif
 
diff --git a/sysdeps/arm/strlen.S b/sysdeps/arm/strlen.S
index 009142be70..fb1589bbe6 100644
--- a/sysdeps/arm/strlen.S
+++ b/sysdeps/arm/strlen.S
@@ -30,8 +30,7 @@
 
 ENTRY(strlen)
 	bic     r1, r0, $3              @ addr of word containing first byte
-	sfi_breg r1, \
-	ldr     r2, [\B], $4            @ get the first word
+	ldr     r2, [r1], $4            @ get the first word
 	ands    r3, r0, $3              @ how many bytes are duff?
 	rsb     r0, r3, $0              @ get - that number into counter.
 	beq     Laligned                @ skip into main check routine if no
@@ -55,8 +54,7 @@ Laligned:				@ here, we have a word in r2.  Does it
 	tstne   r2, $0x00ff0000         @
 	tstne   r2, $0xff000000         @
 	addne   r0, r0, $4              @ if not, the string is 4 bytes longer
-	sfi_breg r1, \
-	ldrne   r2, [\B], $4            @ and we continue to the next word
+	ldrne   r2, [r1], $4            @ and we continue to the next word
 	bne     Laligned                @
 Llastword:				@ drop through to here once we find a
 #ifdef __ARMEB__
diff --git a/sysdeps/arm/submul_1.S b/sysdeps/arm/submul_1.S
index 34606dde51..24d39d93b8 100644
--- a/sysdeps/arm/submul_1.S
+++ b/sysdeps/arm/submul_1.S
@@ -37,24 +37,19 @@ ENTRY (__mpn_submul_1)
 	cfi_rel_offset (r6, 8)
 	cfi_rel_offset (r7, 12)
 
-	sfi_breg r1, \
-	ldr	r6, [\B], #4
-	sfi_breg r0, \
-	ldr	r7, [\B]
+	ldr	r6, [r1], #4
+	ldr	r7, [r0]
 	mov	r4, #0			/* init carry in */
 	b	1f
 0:
-	sfi_breg r1, \
-	ldr	r6, [\B], #4		/* load next ul */
+	ldr	r6, [r1], #4		/* load next ul */
 	adds	r5, r5, r4		/* (lpl, c) = lpl + cl */
 	adc	r4, ip, #0		/* cl = hpl + c */
 	subs	r5, r7, r5		/* (lpl, !c) = rl - lpl */
-	sfi_breg r0, \
-	ldr	r7, [\B, #4]		/* load next rl */
+	ldr	r7, [r0, #4]		/* load next rl */
 	it	cc
 	addcc	r4, r4, #1		/* cl += !c */
-	sfi_breg r0, \
-	str	r5, [\B], #4
+	str	r5, [r0], #4
 1:
 	umull	r5, ip, r6, r3		/* (hpl, lpl) = ul * vl */
 	subs	r2, r2, #1
@@ -63,8 +58,7 @@ ENTRY (__mpn_submul_1)
 	adds	r5, r5, r4		/* (lpl, c) = lpl + cl */
 	adc	r4, ip, #0		/* cl = hpl + c */
 	subs	r5, r7, r5		/* (lpl, !c) = rl - lpl */
-	sfi_breg r0, \
-	str	r5, [\B], #4
+	str	r5, [r0], #4
 	it	cc
 	addcc	r4, r4, #1		/* cl += !c */
 	mov	r0, r4			/* return carry */
diff --git a/sysdeps/arm/sysdep.h b/sysdeps/arm/sysdep.h
index 990e636892..6d60c34df1 100644
--- a/sysdeps/arm/sysdep.h
+++ b/sysdeps/arm/sysdep.h
@@ -150,7 +150,6 @@
    second version uses it.  */
 # define LDST_INDEXED_NOINDEX(OP, R, T, X, Y)		\
 	add	T, X, Y;				\
-	sfi_breg T,					\
 	OP	R, [T]
 # define LDST_INDEXED_INDEX(OP, R, X, Y)		\
 	OP	R, [X, Y]
@@ -198,7 +197,7 @@
 #  define LDR_GLOBAL(R, T, SYMBOL, CONSTANT)				\
 	movw	T, #:lower16:SYMBOL;					\
 	movt	T, #:upper16:SYMBOL;					\
-	sfi_breg T, ldr R, [\B, $CONSTANT]
+	ldr	R, [T, $CONSTANT]
 # elif defined (ARCH_HAS_T2) && defined (PIC) && ARM_PCREL_MOVW_OK
 #  define LDR_GLOBAL(R, T, SYMBOL, CONSTANT)				\
 	movw	R, #:lower16:_GLOBAL_OFFSET_TABLE_ - 97f - PC_OFS;	\
@@ -212,7 +211,7 @@
 97:	add	R, R, pc;						\
 98:	LDST_PC_INDEXED (ldr, T, T, T);					\
 	LDST_INDEXED (ldr, R, T, R, T);					\
-	sfi_breg R, ldr	R, [\B, $CONSTANT]
+	ldr	R, [R, $CONSTANT]
 # else
 #  define LDR_GLOBAL(R, T, SYMBOL, CONSTANT)		\
 	ldr	T, 99f;					\
@@ -277,53 +276,6 @@
 	cfi_restore_state
 # endif /* ARCH_HAS_HARD_TP */
 
-# ifndef ARM_SFI_MACROS
-# define ARM_SFI_MACROS 1
-/* This assembly macro is prepended to any load/store instruction,
-   pulling the base register out of the addressing mode syntax and
-   making it the first operand of the macro.  For example:
-	ldr r0, [r1]
-   becomes:
-	sfi_breg r1, ldr r0, [\B]
-   The \B stands in for the base register that is the first operand
-   to the macro, so we can avoid error-prone repetition of the base
-   register in two places on the line.
-
-   This is used for all memory access through a base register other
-   than PC or SP.  It's intended to support SFI schemes such as
-   Native Client, where the OS will enforce that all load/store
-   instructions use a special form.  In any such configuration,
-   another sysdep.h file will have defined ARM_SFI_MACROS and
-   provided its own assembly macros with the same interface.  */
-
-        .macro sfi_breg basereg, insn, operands:vararg
-                .macro _sfi_breg_doit B
-                \insn \operands
-                .endm
-                _sfi_breg_doit \basereg
-                .purgem _sfi_breg_doit
-        .endm
-
-/* This assembly macro replaces the "pld" instruction.
-   The syntax:
-	sfi_pld REGISTER, #OFFSET
-   is exactly equivalent to:
-	sfi_breg REGISTER, pld [\B, #OFFSET]
-   (and ", #OFFSET" is optional).  We have a separate macro
-   only to work around a bug in GAS versions prior to 2.23.2,
-   that misparses the sfi_breg macro expansion in this case.  */
-
-	.macro sfi_pld basereg, offset=#0
-		pld [\basereg, \offset]
-	.endm
-
-/* This macro precedes any instruction that directly changes the SP.
-   It's not needed for push/pop or for any kind of load or store that
-   implicitly changes the SP via the ! syntax.  */
-# define sfi_sp	/* Nothing to do.  */
-
-# endif
-
 /* These are the directives used for EABI unwind info.
    Wrap them in macros so another configuration's sysdep.h
    file can define them away if it doesn't use EABI unwind info.  */