crypto: cast6 - use unaligned accessors instead of alignmask

Instead of using an alignmask of 0x3 to ensure 32-bit alignment of the
CAST6 input and output blocks, which propagates to mode drivers, and
results in pointless copying on architectures that don't care about
alignment, use the unaligned accessors, which will do the right thing on
each respective architecture, avoiding the need for double buffering.

Signed-off-by: Ard Biesheuvel <[e-mail]>
Signed-off-by: Herbert Xu <[e-mail]>

crypto/cast6_generic.c

···
  */
 
 
-#include <asm/byteorder.h>
+#include <asm/unaligned.h>
 #include <linux/init.h>
 #include <linux/crypto.h>
 #include <linux/module.h>
···
 void __cast6_encrypt(const void *ctx, u8 *outbuf, const u8 *inbuf)
 {
 	const struct cast6_ctx *c = ctx;
-	const __be32 *src = (const __be32 *)inbuf;
-	__be32 *dst = (__be32 *)outbuf;
 	u32 block[4];
 	const u32 *Km;
 	const u8 *Kr;
 
-	block[0] = be32_to_cpu(src[0]);
-	block[1] = be32_to_cpu(src[1]);
-	block[2] = be32_to_cpu(src[2]);
-	block[3] = be32_to_cpu(src[3]);
+	block[0] = get_unaligned_be32(inbuf);
+	block[1] = get_unaligned_be32(inbuf + 4);
+	block[2] = get_unaligned_be32(inbuf + 8);
+	block[3] = get_unaligned_be32(inbuf + 12);
 
 	Km = c->Km[0]; Kr = c->Kr[0]; Q(block, Kr, Km);
 	Km = c->Km[1]; Kr = c->Kr[1]; Q(block, Kr, Km);
···
 	Km = c->Km[10]; Kr = c->Kr[10]; QBAR(block, Kr, Km);
 	Km = c->Km[11]; Kr = c->Kr[11]; QBAR(block, Kr, Km);
 
-	dst[0] = cpu_to_be32(block[0]);
-	dst[1] = cpu_to_be32(block[1]);
-	dst[2] = cpu_to_be32(block[2]);
-	dst[3] = cpu_to_be32(block[3]);
+	put_unaligned_be32(block[0], outbuf);
+	put_unaligned_be32(block[1], outbuf + 4);
+	put_unaligned_be32(block[2], outbuf + 8);
+	put_unaligned_be32(block[3], outbuf + 12);
 }
 EXPORT_SYMBOL_GPL(__cast6_encrypt);
 
···
 void __cast6_decrypt(const void *ctx, u8 *outbuf, const u8 *inbuf)
 {
 	const struct cast6_ctx *c = ctx;
-	const __be32 *src = (const __be32 *)inbuf;
-	__be32 *dst = (__be32 *)outbuf;
 	u32 block[4];
 	const u32 *Km;
 	const u8 *Kr;
 
-	block[0] = be32_to_cpu(src[0]);
-	block[1] = be32_to_cpu(src[1]);
-	block[2] = be32_to_cpu(src[2]);
-	block[3] = be32_to_cpu(src[3]);
+	block[0] = get_unaligned_be32(inbuf);
+	block[1] = get_unaligned_be32(inbuf + 4);
+	block[2] = get_unaligned_be32(inbuf + 8);
+	block[3] = get_unaligned_be32(inbuf + 12);
 
 	Km = c->Km[11]; Kr = c->Kr[11]; Q(block, Kr, Km);
 	Km = c->Km[10]; Kr = c->Kr[10]; Q(block, Kr, Km);
···
 	Km = c->Km[1]; Kr = c->Kr[1]; QBAR(block, Kr, Km);
 	Km = c->Km[0]; Kr = c->Kr[0]; QBAR(block, Kr, Km);
 
-	dst[0] = cpu_to_be32(block[0]);
-	dst[1] = cpu_to_be32(block[1]);
-	dst[2] = cpu_to_be32(block[2]);
-	dst[3] = cpu_to_be32(block[3]);
+	put_unaligned_be32(block[0], outbuf);
+	put_unaligned_be32(block[1], outbuf + 4);
+	put_unaligned_be32(block[2], outbuf + 8);
+	put_unaligned_be32(block[3], outbuf + 12);
 }
 EXPORT_SYMBOL_GPL(__cast6_decrypt);
 
···
 	.cra_flags = CRYPTO_ALG_TYPE_CIPHER,
 	.cra_blocksize = CAST6_BLOCK_SIZE,
 	.cra_ctxsize = sizeof(struct cast6_ctx),
-	.cra_alignmask = 3,
 	.cra_module = THIS_MODULE,
 	.cra_u = {
 		  .cipher = {