Use p2align instead ALIGN

10 files changed, 301 insertions, 323 deletions

diff --git a/ChangeLog b/ChangeLog
index 297ff42b39..8a39723aca 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,16 @@
+2013-10-08  Ondřej Bílka  <neleai@seznam.cz>
+
+	* sysdeps/x86_64/memset.S (ALIGN): Macro removed.
+	Use .p2align directive instead, throughout.
+	* sysdeps/x86_64/multiarch/memcmp-sse4.S: Likewise.
+	* sysdeps/x86_64/multiarch/memcmp-ssse3.S: Likewise.
+	* sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S: Likewise.
+	* sysdeps/x86_64/multiarch/memcpy-ssse3-back.S: Likewise.
+	* sysdeps/x86_64/multiarch/memcpy-ssse3.S: Likewise.
+	* sysdeps/x86_64/multiarch/strcmp-sse2-unaligned.S: Likewise.
+	* sysdeps/x86_64/strchr.S: Likewise.
+	* sysdeps/x86_64/strrchr.S: Likewise.
+
 2013-10-08  Siddhesh Poyarekar  <siddhesh@redhat.com>
 
 	* sysdeps/ieee754/dbl-64/e_pow.c: Fix code formatting.
diff --git a/sysdeps/x86_64/memset.S b/sysdeps/x86_64/memset.S
index 6c69f4b442..9b1de89d98 100644
--- a/sysdeps/x86_64/memset.S
+++ b/sysdeps/x86_64/memset.S
@@ -19,10 +19,6 @@
 
 #include <sysdep.h>
 
-#ifndef ALIGN
-# define ALIGN(n) .p2align n
-#endif
-
 	.text
 #if !defined NOT_IN_libc
 ENTRY(__bzero)
@@ -71,12 +67,12 @@ L(entry_from_bzero):
 L(return):
 	rep
 	ret
-	ALIGN (4)
+	.p2align 4
 L(between_32_64_bytes):
 	movdqu	%xmm8, 16(%rdi)
 	movdqu	%xmm8, -32(%rdi,%rdx)
 	ret
-	ALIGN (4)
+	.p2align 4
 L(loop_start):
 	leaq	64(%rdi), %rcx
 	movdqu	%xmm8, (%rdi)
@@ -92,7 +88,7 @@ L(loop_start):
 	andq	$-64, %rdx
 	cmpq	%rdx, %rcx
 	je	L(return)
-	ALIGN (4)
+	.p2align 4
 L(loop):
 	movdqa	%xmm8, (%rcx)
 	movdqa	%xmm8, 16(%rcx)
diff --git a/sysdeps/x86_64/multiarch/memcmp-sse4.S b/sysdeps/x86_64/multiarch/memcmp-sse4.S
index 1ed4200f4c..d7b147e5ce 100644
--- a/sysdeps/x86_64/multiarch/memcmp-sse4.S
+++ b/sysdeps/x86_64/multiarch/memcmp-sse4.S
@@ -25,10 +25,6 @@
 #  define MEMCMP	__memcmp_sse4_1
 # endif
 
-# ifndef ALIGN
-#  define ALIGN(n)	.p2align n
-# endif
-
 # define JMPTBL(I, B)	(I - B)
 
 # define BRANCH_TO_JMPTBL_ENTRY(TABLE, INDEX, SCALE)		\
@@ -60,7 +56,7 @@ ENTRY (MEMCMP)
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
 # ifndef USE_AS_WMEMCMP
-	ALIGN (4)
+	.p2align 4
 L(firstbyte):
 	movzbl	(%rdi), %eax
 	movzbl	(%rsi), %ecx
@@ -68,7 +64,7 @@ L(firstbyte):
 	ret
 # endif
 
-	ALIGN (4)
+	.p2align 4
 L(79bytesormore):
 	movdqu	(%rsi), %xmm1
 	movdqu	(%rdi), %xmm2
@@ -316,7 +312,7 @@ L(less32bytesin256):
 	add	%rdx, %rdi
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
-	ALIGN (4)
+	.p2align 4
 L(512bytesormore):
 # ifdef DATA_CACHE_SIZE_HALF
 	mov	$DATA_CACHE_SIZE_HALF, %R8_LP
@@ -329,7 +325,7 @@ L(512bytesormore):
 	cmp	%r8, %rdx
 	ja	L(L2_L3_cache_unaglined)
 	sub	$64, %rdx
-	ALIGN (4)
+	.p2align 4
 L(64bytesormore_loop):
 	movdqu	(%rdi), %xmm2
 	pxor	(%rsi), %xmm2
@@ -361,7 +357,7 @@ L(64bytesormore_loop):
 
 L(L2_L3_cache_unaglined):
 	sub	$64, %rdx
-	ALIGN (4)
+	.p2align 4
 L(L2_L3_unaligned_128bytes_loop):
 	prefetchnta 0x1c0(%rdi)
 	prefetchnta 0x1c0(%rsi)
@@ -396,7 +392,7 @@ L(L2_L3_unaligned_128bytes_loop):
 /*
  * This case is for machines which are sensitive for unaligned instructions.
  */
-	ALIGN (4)
+	.p2align 4
 L(2aligned):
 	cmp	$128, %rdx
 	ja	L(128bytesormorein2aligned)
@@ -444,7 +440,7 @@ L(less32bytesin64in2alinged):
 	add	%rdx, %rdi
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
-	ALIGN (4)
+	.p2align 4
 L(128bytesormorein2aligned):
 	cmp	$512, %rdx
 	ja	L(512bytesormorein2aligned)
@@ -519,7 +515,7 @@ L(less32bytesin128in2aligned):
 	add	%rdx, %rdi
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
-	ALIGN (4)
+	.p2align 4
 L(256bytesormorein2aligned):
 
 	sub	$256, %rdx
@@ -632,7 +628,7 @@ L(less32bytesin256in2alinged):
 	add	%rdx, %rdi
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
-	ALIGN (4)
+	.p2align 4
 L(512bytesormorein2aligned):
 # ifdef DATA_CACHE_SIZE_HALF
 	mov	$DATA_CACHE_SIZE_HALF, %R8_LP
@@ -646,7 +642,7 @@ L(512bytesormorein2aligned):
 	ja	L(L2_L3_cache_aglined)
 
 	sub	$64, %rdx
-	ALIGN (4)
+	.p2align 4
 L(64bytesormore_loopin2aligned):
 	movdqa	(%rdi), %xmm2
 	pxor	(%rsi), %xmm2
@@ -678,7 +674,7 @@ L(64bytesormore_loopin2aligned):
 L(L2_L3_cache_aglined):
 	sub	$64, %rdx
 
-	ALIGN (4)
+	.p2align 4
 L(L2_L3_aligned_128bytes_loop):
 	prefetchnta 0x1c0(%rdi)
 	prefetchnta 0x1c0(%rsi)
@@ -711,7 +707,7 @@ L(L2_L3_aligned_128bytes_loop):
 	BRANCH_TO_JMPTBL_ENTRY(L(table_64bytes), %rdx, 4)
 
 
-	ALIGN (4)
+	.p2align 4
 L(64bytesormore_loop_end):
 	add	$16, %rdi
 	add	$16, %rsi
@@ -806,7 +802,7 @@ L(8bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(12bytes):
 	mov	-12(%rdi), %rax
 	mov	-12(%rsi), %rcx
@@ -827,7 +823,7 @@ L(0bytes):
 
 # ifndef USE_AS_WMEMCMP
 /* unreal case for wmemcmp */
-	ALIGN (4)
+	.p2align 4
 L(65bytes):
 	movdqu	-65(%rdi), %xmm1
 	movdqu	-65(%rsi), %xmm2
@@ -864,7 +860,7 @@ L(9bytes):
 	sub	%edx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(13bytes):
 	mov	-13(%rdi), %rax
 	mov	-13(%rsi), %rcx
@@ -877,7 +873,7 @@ L(13bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(5bytes):
 	mov	-5(%rdi), %eax
 	mov	-5(%rsi), %ecx
@@ -888,7 +884,7 @@ L(5bytes):
 	sub	%edx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(66bytes):
 	movdqu	-66(%rdi), %xmm1
 	movdqu	-66(%rsi), %xmm2
@@ -929,7 +925,7 @@ L(10bytes):
 	sub	%ecx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(14bytes):
 	mov	-14(%rdi), %rax
 	mov	-14(%rsi), %rcx
@@ -942,7 +938,7 @@ L(14bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(6bytes):
 	mov	-6(%rdi), %eax
 	mov	-6(%rsi), %ecx
@@ -958,7 +954,7 @@ L(2bytes):
 	sub	%ecx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(67bytes):
 	movdqu	-67(%rdi), %xmm2
 	movdqu	-67(%rsi), %xmm1
@@ -997,7 +993,7 @@ L(11bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(15bytes):
 	mov	-15(%rdi), %rax
 	mov	-15(%rsi), %rcx
@@ -1010,7 +1006,7 @@ L(15bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(7bytes):
 	mov	-7(%rdi), %eax
 	mov	-7(%rsi), %ecx
@@ -1023,7 +1019,7 @@ L(7bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(3bytes):
 	movzwl	-3(%rdi), %eax
 	movzwl	-3(%rsi), %ecx
@@ -1036,7 +1032,7 @@ L(1bytes):
 	ret
 # endif
 
-	ALIGN (4)
+	.p2align 4
 L(68bytes):
 	movdqu	-68(%rdi), %xmm2
 	movdqu	-68(%rsi), %xmm1
@@ -1079,7 +1075,7 @@ L(20bytes):
 
 # ifndef USE_AS_WMEMCMP
 /* unreal cases for wmemcmp */
-	ALIGN (4)
+	.p2align 4
 L(69bytes):
 	movdqu	-69(%rsi), %xmm1
 	movdqu	-69(%rdi), %xmm2
@@ -1115,7 +1111,7 @@ L(21bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(70bytes):
 	movdqu	-70(%rsi), %xmm1
 	movdqu	-70(%rdi), %xmm2
@@ -1151,7 +1147,7 @@ L(22bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(71bytes):
 	movdqu	-71(%rsi), %xmm1
 	movdqu	-71(%rdi), %xmm2
@@ -1188,7 +1184,7 @@ L(23bytes):
 	ret
 # endif
 
-	ALIGN (4)
+	.p2align 4
 L(72bytes):
 	movdqu	-72(%rsi), %xmm1
 	movdqu	-72(%rdi), %xmm2
@@ -1227,7 +1223,7 @@ L(24bytes):
 
 # ifndef USE_AS_WMEMCMP
 /* unreal cases for wmemcmp */
-	ALIGN (4)
+	.p2align 4
 L(73bytes):
 	movdqu	-73(%rsi), %xmm1
 	movdqu	-73(%rdi), %xmm2
@@ -1265,7 +1261,7 @@ L(25bytes):
 	sub	%ecx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(74bytes):
 	movdqu	-74(%rsi), %xmm1
 	movdqu	-74(%rdi), %xmm2
@@ -1302,7 +1298,7 @@ L(26bytes):
 	movzwl	-2(%rsi), %ecx
 	jmp	L(diffin2bytes)
 
-	ALIGN (4)
+	.p2align 4
 L(75bytes):
 	movdqu	-75(%rsi), %xmm1
 	movdqu	-75(%rdi), %xmm2
@@ -1342,7 +1338,7 @@ L(27bytes):
 	xor	%eax, %eax
 	ret
 # endif
-	ALIGN (4)
+	.p2align 4
 L(76bytes):
 	movdqu	-76(%rsi), %xmm1
 	movdqu	-76(%rdi), %xmm2
@@ -1388,7 +1384,7 @@ L(28bytes):
 
 # ifndef USE_AS_WMEMCMP
 /* unreal cases for wmemcmp */
-	ALIGN (4)
+	.p2align 4
 L(77bytes):
 	movdqu	-77(%rsi), %xmm1
 	movdqu	-77(%rdi), %xmm2
@@ -1430,7 +1426,7 @@ L(29bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(78bytes):
 	movdqu	-78(%rsi), %xmm1
 	movdqu	-78(%rdi), %xmm2
@@ -1470,7 +1466,7 @@ L(30bytes):
 	xor	%eax, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(79bytes):
 	movdqu	-79(%rsi), %xmm1
 	movdqu	-79(%rdi), %xmm2
@@ -1510,7 +1506,7 @@ L(31bytes):
 	xor	%eax, %eax
 	ret
 # endif
-	ALIGN (4)
+	.p2align 4
 L(64bytes):
 	movdqu	-64(%rdi), %xmm2
 	movdqu	-64(%rsi), %xmm1
@@ -1548,7 +1544,7 @@ L(32bytes):
 /*
  * Aligned 8 bytes to avoid 2 branch "taken" in one 16 alinged code block.
  */
-	ALIGN (3)
+	.p2align 3
 L(less16bytes):
 	movsbq	%dl, %rdx
 	mov	(%rsi, %rdx), %rcx
@@ -1585,7 +1581,7 @@ L(diffin2bytes):
 	sub	%ecx, %eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(end):
 	and	$0xff, %eax
 	and	$0xff, %ecx
@@ -1599,7 +1595,7 @@ L(end):
 	neg	%eax
 	ret
 
-	ALIGN (4)
+	.p2align 4
 L(nequal_bigger):
 	ret
 
@@ -1611,7 +1607,7 @@ L(unreal_case):
 END (MEMCMP)
 
 	.section .rodata.sse4.1,"a",@progbits
-	ALIGN (3)
+	.p2align 3
 # ifndef USE_AS_WMEMCMP
 L(table_64bytes):
 	.int	JMPTBL (L(0bytes), L(table_64bytes))
diff --git a/sysdeps/x86_64/multiarch/memcmp-ssse3.S b/sysdeps/x86_64/multiarch/memcmp-ssse3.S
index e319df926e..e04f918dff 100644
--- a/sysdeps/x86_64/multiarch/memcmp-ssse3.S
+++ b/sysdeps/x86_64/multiarch/memcmp-ssse3.S
@@ -25,10 +25,6 @@
 #  define MEMCMP	__memcmp_ssse3
 # endif
 
-# ifndef ALIGN
-#  define ALIGN(n)	.p2align n
-# endif
-
 /* Warning!
 	   wmemcmp has to use SIGNED comparison for elements.
 	   memcmp has to use UNSIGNED comparison for elemnts.
@@ -50,7 +46,7 @@ ENTRY (MEMCMP)
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 /* ECX >= 32.  */
 L(48bytesormore):
 	movdqu	(%rdi), %xmm3
@@ -90,7 +86,7 @@ L(48bytesormore):
 	je	L(shr_6)
 	jmp	L(shr_7)
 
-	ALIGN	(2)
+	.p2align 2
 L(next_unaligned_table):
 	cmp	$8, %edx
 	je	L(shr_8)
@@ -117,7 +113,7 @@ L(next_unaligned_table):
 	jmp	L(shr_12)
 # endif
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_0):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -137,7 +133,7 @@ L(shr_0):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_0_gobble):
 	movdqa	(%rsi), %xmm0
 	xor	%eax, %eax
@@ -180,7 +176,7 @@ L(next):
 
 # ifndef USE_AS_WMEMCMP
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_1):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -207,7 +203,7 @@ L(shr_1):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_1_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -258,7 +254,7 @@ L(shr_1_gobble_next):
 	jmp	L(less48bytes)
 
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_2):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -285,7 +281,7 @@ L(shr_2):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_2_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -335,7 +331,7 @@ L(shr_2_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_3):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -362,7 +358,7 @@ L(shr_3):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_3_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -414,7 +410,7 @@ L(shr_3_gobble_next):
 
 # endif
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_4):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -441,7 +437,7 @@ L(shr_4):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_4_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -493,7 +489,7 @@ L(shr_4_gobble_next):
 
 # ifndef USE_AS_WMEMCMP
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_5):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -520,7 +516,7 @@ L(shr_5):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_5_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -570,7 +566,7 @@ L(shr_5_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_6):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -597,7 +593,7 @@ L(shr_6):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_6_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -647,7 +643,7 @@ L(shr_6_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_7):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -674,7 +670,7 @@ L(shr_7):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_7_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -726,7 +722,7 @@ L(shr_7_gobble_next):
 
 # endif
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_8):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -753,7 +749,7 @@ L(shr_8):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_8_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -805,7 +801,7 @@ L(shr_8_gobble_next):
 
 # ifndef USE_AS_WMEMCMP
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_9):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -832,7 +828,7 @@ L(shr_9):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_9_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -882,7 +878,7 @@ L(shr_9_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_10):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -909,7 +905,7 @@ L(shr_10):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_10_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -959,7 +955,7 @@ L(shr_10_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_11):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -986,7 +982,7 @@ L(shr_11):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_11_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -1038,7 +1034,7 @@ L(shr_11_gobble_next):
 
 # endif
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_12):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -1065,7 +1061,7 @@ L(shr_12):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_12_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -1117,7 +1113,7 @@ L(shr_12_gobble_next):
 
 # ifndef USE_AS_WMEMCMP
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_13):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -1144,7 +1140,7 @@ L(shr_13):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_13_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -1194,7 +1190,7 @@ L(shr_13_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_14):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -1221,7 +1217,7 @@ L(shr_14):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_14_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -1271,7 +1267,7 @@ L(shr_14_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_15):
 	cmp	$80, %rcx
 	lea	-48(%rcx), %rcx
@@ -1298,7 +1294,7 @@ L(shr_15):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 
-	ALIGN	(4)
+	.p2align 4
 L(shr_15_gobble):
 	sub	$32, %rcx
 	movdqa	16(%rsi), %xmm0
@@ -1348,7 +1344,7 @@ L(shr_15_gobble_next):
 	add	%rcx, %rdi
 	jmp	L(less48bytes)
 # endif
-	ALIGN	(4)
+	.p2align 4
 L(exit):
 	pmovmskb %xmm1, %r8d
 	sub	$0xffff, %r8d
@@ -1389,56 +1385,56 @@ L(less16bytes):
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte16):
 	movzbl	-16(%rdi), %eax
 	movzbl	-16(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte17):
 	movzbl	-15(%rdi), %eax
 	movzbl	-15(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte18):
 	movzbl	-14(%rdi), %eax
 	movzbl	-14(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte19):
 	movzbl	-13(%rdi), %eax
 	movzbl	-13(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte20):
 	movzbl	-12(%rdi), %eax
 	movzbl	-12(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte21):
 	movzbl	-11(%rdi), %eax
 	movzbl	-11(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(Byte22):
 	movzbl	-10(%rdi), %eax
 	movzbl	-10(%rsi), %edx
 	sub	%edx, %eax
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(next_24_bytes):
 	lea	8(%rdi), %rdi
 	lea	8(%rsi), %rsi
@@ -1479,14 +1475,14 @@ L(next_24_bytes):
 	jne	L(find_diff)
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(second_double_word):
 	mov	-12(%rdi), %eax
 	cmp	-12(%rsi), %eax
 	jne	L(find_diff)
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(next_two_double_words):
 	and	$15, %dh
 	jz	L(fourth_double_word)
@@ -1495,7 +1491,7 @@ L(next_two_double_words):
 	jne	L(find_diff)
 	ret
 
-	ALIGN	(4)
+	.p2align 4
 L(fourth_double_word):
 	mov	-4(%rdi), %eax
 	cmp	-4(%rsi), %eax
@@ -1503,7 +1499,7 @@ L(fourth_double_word):
 	ret
 # endif
 
-	ALIGN	(4)
+	.p2align 4
 L(less48bytes):
 	cmp	$8, %ecx