crypto: poly1305 - use unaligned access macros to output digest

Currently the only part of poly1305-generic which is assuming special
alignment is the part where the final digest is written.  Switch this
over to the unaligned access macros so that we'll be able to remove the
cra_alignmask.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto/poly1305_generic.c

@@ -210,7 +210,6 @@ EXPORT_SYMBOL_GPL(crypto_poly1305_update);
 int crypto_poly1305_final(struct shash_desc *desc, u8 *dst)
 {
 	struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
-	__le32 *mac = (__le32 *)dst;
 	u32 h0, h1, h2, h3, h4;
 	u32 g0, g1, g2, g3, g4;
 	u32 mask;
@@ -267,10 +266,10 @@ int crypto_poly1305_final(struct shash_desc *desc, u8 *dst)
 	h3 = (h3 >> 18) | (h4 <<  8);
 
 	/* mac = (h + s) % (2^128) */
-	f = (f >> 32) + h0 + dctx->s[0]; mac[0] = cpu_to_le32(f);
-	f = (f >> 32) + h1 + dctx->s[1]; mac[1] = cpu_to_le32(f);
-	f = (f >> 32) + h2 + dctx->s[2]; mac[2] = cpu_to_le32(f);
-	f = (f >> 32) + h3 + dctx->s[3]; mac[3] = cpu_to_le32(f);
+	f = (f >> 32) + h0 + dctx->s[0]; put_unaligned_le32(f, dst +  0);
+	f = (f >> 32) + h1 + dctx->s[1]; put_unaligned_le32(f, dst +  4);
+	f = (f >> 32) + h2 + dctx->s[2]; put_unaligned_le32(f, dst +  8);
+	f = (f >> 32) + h3 + dctx->s[3]; put_unaligned_le32(f, dst + 12);
 
 	return 0;
 }