x86 asm: Replace .align with .balign

The .align directive used to align storage locations is
ambiguous. On some platforms and assemblers it takes a byte count,
on others the argument is interpreted as a shift value. The current
usage expects the first interpretation.

Replace it with the unambiguous .balign directive which always
expects a byte count, regardless of platform and assembler.

Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Reviewed-by: Tino Reichardt <milky-zfs@mcmilk.de>
Reviewed-by: Richard Yao <richard.yao@alumni.stonybrook.edu>
Signed-off-by: Attila Fülöp <attila@fueloep.org>
Closes #14422

module/icp/asm-x86_64/sha2/sha256_impl.S

@@ -133,7 +133,7 @@ ENTRY_NP(SHA256TransformBlocks)
 	mov	4*7(%rdi),%r11d
 	jmp	.Lloop
 
-.align	16
+.balign	16
 .Lloop:
 	xor	%rdi,%rdi
 	mov	4*0(%rsi),%r12d
@@ -873,7 +873,7 @@ ENTRY_NP(SHA256TransformBlocks)
 
 	add	%r14d,%eax			# h+=Maj(a,b,c)
 	jmp	.Lrounds_16_xx
-.align	16
+.balign	16
 .Lrounds_16_xx:
 	mov	4(%rsp),%r13d
 	mov	56(%rsp),%r12d
@@ -2064,7 +2064,7 @@ ENTRY_NP(SHA256TransformBlocks)
 SET_SIZE(SHA256TransformBlocks)
 
 .section .rodata
-.align	64
+.balign	64
 SET_OBJ(K256)
 K256:
 	.long	0x428a2f98,0x71374491,0xb5c0fbcf,0xe9b5dba5

module/icp/asm-x86_64/sha2/sha512_impl.S

@@ -134,7 +134,7 @@ ENTRY_NP(SHA512TransformBlocks)
 	mov	8*7(%rdi),%r11
 	jmp	.Lloop
 
-.align	16
+.balign	16
 .Lloop:
 	xor	%rdi,%rdi
 	mov	8*0(%rsi),%r12
@@ -874,7 +874,7 @@ ENTRY_NP(SHA512TransformBlocks)
 
 	add	%r14,%rax			# h+=Maj(a,b,c)
 	jmp	.Lrounds_16_xx
-.align	16
+.balign	16
 .Lrounds_16_xx:
 	mov	8(%rsp),%r13
 	mov	112(%rsp),%r12
@@ -2065,7 +2065,7 @@ ENTRY_NP(SHA512TransformBlocks)
 SET_SIZE(SHA512TransformBlocks)
 
 .section .rodata
-.align	64
+.balign	64
 SET_OBJ(K512)
 K512:
 	.quad	0x428a2f98d728ae22,0x7137449123ef65cd